Stainless Steels For Design Engineers

Copyright © 2008 ASM International®. All rights reserved.
Stainless Steels for Design Engineers (#05231G) www.asminternational.org
STAINLESS STEELS
FOR DESIGN ENGINEERS
MICHAEL MCGUIRE
ASM International®
Materials Park, Ohio 44073-0002
www.asminternational.org
Copyright © 2008
by
ASM International®
All rights reserved
No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise, without the written permission of the
copyright owner.
First printing, December 2008
Great care is taken in the compilation and production of this book, but it should be made clear that NO WAR-
RANTIES, EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, WARRANTIES OF MER-
CHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, ARE GIVEN IN CONNECTION WITH
THIS PUBLICATION. Although this information is believed to be accurate by ASM, ASM cannot guarantee
that favorable results will be obtained from the use of this publication alone. This publication is intended for use
by persons having technical skill, at their sole discretion and risk. Since the conditions of product or material use
are outside of ASM’s control, ASM assumes no liability or obligation in connection with any use of this infor-
mation. No claim of any kind, whether as to products or information in this publication, and whether or not
based on negligence, shall be greater in amount than the purchase price of this product or publication in respect
of which damages are claimed. THE REMEDY HEREBY PROVIDED SHALL BE THE EXCLUSIVE AND
SOLE REMEDY OF BUYER, AND IN NO EVENT SHALL EITHER PARTY BE LIABLE FOR SPECIAL,
INDIRECT OR CONSEQUENTIAL DAMAGES WHETHER OR NOT CAUSED BY OR RESULTING
FROM THE NEGLIGENCE OF SUCH PARTY. As with any material, evaluation of the material under end-use
conditions prior to specification is essential. Therefore, specific testing under actual conditions is recommended.
Nothing contained in this book shall be construed as a grant of any right of manufacture, sale, use, or reproduc-
tion, in connection with any method, process, apparatus, product, composition, or system, whether or not cov-
ered by letters patent, copyright, or trademark, and nothing contained in this book shall be construed as a de-
fense against any alleged infringement of letters patent, copyright, or trademark, or as a defense against liability
for such infringement.
Comments, criticisms, and suggestions are invited, and should be forwarded to ASM International.
Prepared under the direction of the ASM International Technical Book Committee (2007–2008), Lichun L.
Chen, Chair.
ASM International staff who worked on this project include Scott Henry, Senior Manager of Product and Service
Development; Steven R. Lampman, Technical Editor; Eileen De Guire, Associate Editor; Ann Britton, Editorial
Assistant; Bonnie Sanders, Manager of Production; Madrid Tramble, Senior Production Coordinator; Diane
Grubbs, Production Coordinator; Patty Conti, Production Coordinator; and Kathryn Muldoon, Production Assistant
Library of Congress Control Number: 2008934669
ISBN-13: 978-0-87170-717-8
ISBN-10: 0-87170-717-9
SAN: 204-7586
ASM International®
Materials Park, OH 44073-0002
Printed in the United States of America
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
METALLURGY
Chapter 1 Metallurgy........................................................................................................1
CORROSION AND OXIDATION

Chapter 2 Corrosion Theory............................................................................................11
Chapter 3 Corrosion Kinetics..........................................................................................19
Chapter 4 Corrosion Types..............................................................................................27
Chapter 5 Oxidation.......................................................................................................57
STAINLESS STEEL ALLOYS

Chapter 6 Austenitic Stainless Steels ..............................................................................69
Chapter 7 Duplex Stainless Steels...................................................................................91
Chapter 8 Ferritic Stainless Steels .................................................................................109
Chapter 9 Martensitic Stainless Steels ..........................................................................123
Chapter 10 Precipitation-Hardening Stainless Steels ......................................................137
PROCESSING
Chapter 11 Casting Alloys...............................................................................................147
Chapter 12 Melting, Casting, and Hot Processing...........................................................155
Chapter 13 Thermal Processing ......................................................................................161

Chapter 14 Forming........................................................................................................173
Chapter 15 Machining ....................................................................................................181
Chapter 16 Surface Finishing ..........................................................................................193
Chapter 17 Welding........................................................................................................201
APPLICATIONS
Chapter 18 Architecture and Construction.....................................................................213
Chapter 19 Automotive and Transportation Applications................................................225
Chapter 20 Commercial and Residential Applications ....................................................233
Chapter 21 Marine Systems Applications........................................................................243
Chapter 22 Petroleum Industry Applications ..................................................................247
Chapter 23 Chemical and Process Industry Applications ................................................257
Chapter 24 Pulp-and-Paper Industry Applications ..........................................................265
APPENDIXES
Appendix 1 Compositions...............................................................................................269
Appendix 2 Physical and Mechanical Properties of Select Alloys....................................279
Appendix 3 Introduction to Thermo-Calc and Instructions for

Accessing Free Demonstration ....................................................................281
Index .................................................................................................................................285
iv
Preface
The rate of growth of stainless steel has outpaced that of other metals and alloys, and by 2010 may
surpass aluminum as the second most widely used metal after carbon steel. The 2007 world produc-
tion of stainless steel was approximately 30,000,000 tons and has nearly doubled in the last ten years.
This growth is occurring at the same time that the production of stainless steel continues to become
more consolidated. One result of this is a more widespread need to understand stainless steel with
fewer resources to provide that information. The concurrent technical evolution in stainless steel and
increasing volatility of raw material prices has made it more important for the engineers and design-
ers who use stainless steel to make sound technical judgments about which stainless steels to use and
how to use them.
This book provides design engineers with an up-to-date source of information at a level useful for
both metallurgists and other engineers and technicians. It seeks to bridge the gap between the inter-
net where much current, but raw information is available and scholarly books and journals that pro-
vide theory that is difficult to put into practice. The content of the book is selected for utility for the
user of stainless steel. The first section gives elementary metallurgy and identification of constituents
of stainless, the effects of alloying elements and a significant section on corrosion. A second section
is oriented toward processes important to users of stainless steel. The third section is about each fam-
ily of stainless alloys and includes the most recent additions that have come to the market. The fourth
section deals in some depth with the major applications for stainless steel. This last part is presented
without the promotional bias which is found in many steel producers’, alloy producers’, and trade as-
sociations’ literature. While a number of steel producers have provided assistance to the author, there
has been no attempt to unfairly bias information in their favor. To the contrary, those producers re-
sponsible for generating factual, useful data for the user community are those who should benefit the
most by books such as this. The author is particularly indebted to Allegheny Ludlum and John
Grubb, and his many colleagues who assisted him, for technical assistance throughout the writing
and to Carnegie Mellon University for their support. The author also wishes to thank Professor Srid-
har Seetharaman at Carnegie Mellon University for his help in writing the corrosion chapter and oth-
ers who helped: Roy Matway of CMU, Vittorio Boneschi of Centro-Inox; Paul Mason of Thermo-
Calc; Bob Drab of Schmolz Bichenbach; Elisabeth Torsner and Chuck Turack Outukumpu, USA;
Scott Balliett of Latrobe Steel; Jim Halliday and Fred Deuschle of Contrarian Metals Resources; Pro-
fessors Tony DeArdo of Pitt and Gerhard Welsch of CWRU; the staffs of Centro-Inox, Euro-Inox,
SSNA, The Nickel Institute; and the editorial staff at ASM International, Scott Henry, Eileen
DeGuire, Charlie Moosbrugger and Steve Lampman. I would also like to thank the many members
of my forum at Eng-tips.com who have contributed much collective knowledge and perspective to
this book.
ASM International is the society for materials
engineers and scientists, a worldwide network
dedicated to advancing industry, technology, and
applications of metals and materials.
ASM International, Materials Park, Ohio, USA

This publication is copyright © ASM International®. All rights reserved.
Publication title Product code
Stainless Steels for Design Engineers #05231G
To order products from ASM International:

Online Visit www.asminternational.org/bookstore
Telephone 1-800-336-5152 (US) or 1-440-338-5151 (Outside US)

Fax 1-440-338-4634
Customer Service, ASM International
Mail
9639 Kinsman Rd, Materials Park, Ohio 44073-0002, USA
Email CustomerService@asminternational.org
American Technical Publishers Ltd.
27-29 Knowl Piece, Wilbury Way, Hitchin Hertfordshire SG4 0SX,
In Europe United Kingdom
Telephone: 01462 437933 (account holders), 01462 431525 (credit card)
www.ameritech.co.uk
Neutrino Inc.
In Japan Takahashi Bldg., 44-3 Fuda 1-chome, Chofu-Shi, Tokyo 182 Japan
Telephone: 81 (0) 424 84 5550
Terms of Use. This publication is being made available in PDF format as a benefit to members and
customers of ASM International. You may download and print a copy of this publication for your
personal use only. Other use and distribution is prohibited without the express written permission of
ASM International.
No warranties, express or implied, including, without limitation, warranties of merchantability or
fitness for a particular purpose, are given in connection with this publication. Although this
information is believed to be accurate by ASM, ASM cannot guarantee that favorable results will be
obtained from the use of this publication alone. This publication is intended for use by persons having
technical skill, at their sole discretion and risk. Since the conditions of product or material use are
outside of ASM's control, ASM assumes no liability or obligation in connection with any use of this
information. As with any material, evaluation of the material under end-use conditions prior to
specification is essential. Therefore, specific testing under actual conditions is recommended.
Nothing contained in this publication shall be construed as a grant of any right of manufacture, sale,
use, or reproduction, in connection with any method, process, apparatus, product, composition, or
system, whether or not covered by letters patent, copyright, or trademark, and nothing contained in this
publication shall be construed as a defense against any alleged infringement of letters patent,
copyright, or trademark, or as a defense against liability for such infringement.



Fax 1-440-338-4634
Mail
www.ameritech.co.uk
Neutrino Inc.
Telephone: 81 (0) 424 84 5550
ASM International.
Stainless Steels for Design Engineers Copyright © 2008 ASM International®
Michael F. McGuire, p 1-10 All rights reserved.
DOI: 10.1361/ssde2008p001 www.asminternational.org
CHAPTER 1
Metallurgy
Summary phases coexisting simultaneously. Stainless

steel is an exceptional alloy system in that it is
COMPARED TO ALLOY STEELS, stainless not a dilute solution. Alloy steels may contain
steels are chemically complex. The large number several percent of alloying elements, such as
of alloying elements makes possible a larger carbon, manganese, nickel, molybdenum,
range of possible phases or basic crystal struc- chromium, and silicon, in addition to the impu-
tures. The large amount of the alloying elements rities sulfur, oxygen, and phosphorus. Alloy
makes the deviation from the behavior of pure steels typically contain very small amounts of
iron greater; consequently, the calculations that titanium, niobium, and aluminum. The total
predict which phases will exist are more difficult. amount* of these alloying elements seldom ex-
The three basic phases of stainless steels are fer- ceeds 5%. The same is true for most aluminum
rite, austenite, and martensite. The wide variety alloys. In contrast, stainless steels contain no
of alloys that exist is based on: less than about 11% chromium alone. Most
stainless alloys have manganese, silicon, car-
• Combinations of these phases
bon, and nickel in thermodynamically mean-
• Altering the composition of these phases
ingful amounts as well as large concentrations
• Adding secondary phases for particular
of nickel and/or molybdenum.
purposes The result of the large number of alloying ele-
Metallurgy, as discussed in this chapter, fo- ments in relatively high concentrations is that
cuses on phases normally encountered in stain- stainless steel can have many stable phases con-
less steels and their characteristics. In subsequent currently. In almost every case, having phases
chapters on types of stainless steel, there are other than the principal one or two phases for
more detailed treatments of the alloys made of which the alloy was designed is undesirable be-
these phases and their properties. cause of the possibility of undesirable variations
in mechanical or corrosion performance. The
producer of stainless steel controls the chemical
composition and thermomechanical processing,
Introduction so that when the processor or end user receives
the product it is usually in the correct condition.
Most widely used alloy systems, such as car-
However, subsequent processing or service con-
bon steels, alloy steels, and aluminum alloys, are
ditions may alter the carefully established phase
relatively dilute solutions of several elements in
structure. Therefore, it is necessary to discuss
the parent matrix. Carbon and alloy steels, with
the phases that can exist in stainless steel and
very few exceptions, are principally of the mag-
the conditions under which they form so that the
netic body-centered cubic (bcc) phase or a
enlightened user will know which phases to
slightly distorted version of it. Aluminum alloys
avoid and how to avoid them.
share the face-centered cubic (fcc) structure of
It is possible to use thermodynamics to calcu-
pure aluminum. A given structure, which can
late which phases may exist at a given tempera-
have a certain range of compositions, is what is
meant by a phase, just as a gas or liquid is a * All compositions are given in weight percent unless
phase. In solid metals, there can be a number of stated otherwise.
2 / Stainless Steels for Design Engineers
ture for a given composition. It is not remotely not a small, pedantic point. Most stainless steels
feasible, however, to give an adequate treatment are used in the metastable condition. For exam-
of the thermodynamics required to do this. The ple, the common alloy 304 (also called 18-8) is
topic alone requires a book. The necessary normally used in the fully austenitic condition.
knowledge has been embedded in proprietary It would “rather” be partly ferritic, but the sub-
computer programs that will be used instead. stitutional diffusion of chromium in austenite
that is required to form a ferrite phase of a sepa-
rate composition is so slow that it cannot occur
Thermodynamics of Stainless Steel in terrestrial time frames. However, if energy is
applied by mechanical shear, the austenite can
Pure metals, from a practical viewpoint, are transform without diffusion to the lower free-
either liquid or solid depending on temperature, energy martensite phase, a quasi-bcc structure
with the possibility of some trivial small gas of lower free energy.
vapor pressure. A law of thermodynamics is that The calculation of which phases exist under
the number of possible condensed (i.e., solid) equilibrium conditions proves to be extraordi-
phases equals the number of elemental con- narily difficult in complicated alloy systems.
stituents plus one. The solid has a crystallo- This is because thermodynamic values can be
graphic structure that may vary with tempera- measured accurately only in the liquid state, so
ture. Many metals have a less-dense bcc the values for the solid state are extrapolations.
structure at high temperature and transform to a Also, the interaction between elements is very
denser fcc structure at lower temperatures. Iron important in nondilute alloys such as stainless
does this. Iron has the curious characteristic of steel. Consequently, most published phase dia-
transforming from fcc back to the low-density grams are experimentally derived. To determine
bcc at still lower temperatures. This is a result of which phases exist at a given composition and
the unpaired 3d orbital electrons (those that give temperature, a sample is made, equilibrated at
rise to ferromagnetism) that are not given up as the appropriate temperature, and quenched to
valence electrons, causing repulsive forces be- room temperature. It is assumed that the charac-
tween atoms and requiring a more widely spaced teristic equilibrium phases have been frozen and
structure. are then identified by various techniques for
All thermodynamic properties are based on structure, composition, and the like. This impor-
interatomic attractions. In metals, the metal tant work is obviously tedious and susceptible
atoms give up valence electrons to the entire to experimental error and applies only to spe-
mass. These electrons are of varying energy cific compositions. Any “what if” extrapolation
states and highly mobile. They are responsible to a different alloy composition carries the risk
for the ability of metals to conduct heat and of error.
electricity well. The attraction, the strength of A practical tool has been developed that per-
the bond, is proportional to the charge differ- mits phase diagrams to be calculated for arbitrary
ence and distance. The attraction determines compositions. These are computer simulated,
such macroscopic properties as melting temper- mathematical models that can perform the com-
ature, density, and elastic modulus. plex thermodynamic calculations. To do this with
In this book, the main concern with thermo- accuracy requires databases of thermodynamic
dynamics is predicting which phases are present values. These values must be derived from com-
both at equilibrium and in the quite frequent puter analysis of experimental phase equilibrium
metastable state. The prediction involves calcu- diagrams. They are expensive to derive and vali-
lating the free energy of the various possible date, and only a few exist. Hence, they are pro-
phases. The phase with the lowest energy is prietary. In Appendix 3, a license to one such
most favored, but others may have free energies program, Thermo-Calc, can be found. The ver-
that permit them to exist. The difference be- sion has a reduced three-element capability but
tween these two is that the equilibrium state, uses the same proprietary thermodynamic data-
that of the lowest free energy, may require base of the full version. The program allows de-
atomic rearrangements to occur for equilibrium termination of which phases can exist for any
compositions to be reached on an atomic scale. composition and temperature. Whether the
If diffusion is too sluggish for these rearrange- phases will form depends also on kinetic factors.
ments to take place, the structure may retain the First, however, it is good to become familiar with
prior metastable structure indefinitely. This is the principal phases found in stainless steel.
Chapter 1: Metallurgy / 3
Phases and oxygen, have no influence on which phase

is favored. Again, it must be emphasized that
Ferrite the influence of an alloying element on structure
has zero bearing on its influence on corrosion
The basis of stainless alloys is, of course, resistance.
iron. Iron, as stated, solidifies as a bcc alloy be- The elements that promote ferrite over austen-
fore transforming to the denser fcc austenite at ite also have the effect, at still lower tempera-
lower temperatures. At still lower temperatures, tures, of promoting intermetallic compounds
it reverts to the bcc structure. It is accurate to generally composed of iron, chromium, and
surmise from this that the free energy of both some of those alloying elements. These are dis-
structures is close. Alloying elements that pro- cussed separately.
mote one structure over the other can therefore Metals are effective solvents in both the liq-
change which one predominates. The element uid and solid states. An important part of steel-
that produces the ability to form the passive film making is refining the molten metal to remove
that makes stainless corrosion resistant, the undesired impurities dissolved in it. The
chromium, has the characteristic of stabilizing normal technique is to add elements that react
the bcc structure. As chromium is added to iron, selectively with the targeted impurities to form
the temperature range over which austenite is an immiscible reactant that can become part of
stable grows smaller until, at about 12% the slag and physically separated from the re-
chromium, ferrite is stable at all temperatures. fined alloy. This is done for the primary impuri-
This is, coincidentally, the approximate level of ties oxygen and sulfur. A third common impu-
chromium needed to keep alloys from rusting rity, phosphorus, is not so easily removed and
under ambient conditions, but this effect is not must be excluded from raw materials to be kept
related to whether the structure is bcc or fcc. under control.
The iron-chromium phase diagram (Fig. 1) In stainless steel, carbon and nitrogen can be
shows the composition and temperature regions detrimental impurities. Both are quite soluble
where ferrite (a), martensite (α' ), austenite (γ), in molten iron-chromium alloys and are fairly
and sigma phase (σ) are stable. soluble in ferrite at high temperatures. This sol-
While chromium is the principal ferrite-pro- ubility decreases exponentially with tempera-
moting alloying element, other elements have ture so that it is essentially zero at room tem-
similar effects, but none produces the quality of perature. These elements have small atomic
stainlessness. Silicon, aluminum, molybdenum, sizes compared to iron and chromium and,
tungsten, niobium, and titanium all favor ferrite. when dissolved, squeeze into interstitial sites
Carbon, nitrogen, manganese, nickel, and cop- within the bcc matrix. Such interstitial solute
per do not and expand the temperature range atoms profoundly distort the structure. They are
over which austenite exists. Elements that are much more soluble in the fcc structure, which,
insoluble in iron at austenite-forming tempera- while denser, has roomier interstitial spaces, so
tures, such as the impurities phosphorus, sulfur, they stabilize that structure. To preserve the fer-
rite structure, carbon and nitrogen must be
eliminated.
There are additional reasons to eliminate car-
bon and nitrogen. During cooling as these ele-
ments become less and less soluble, they must
precipitate. The most thermodynamically favor-
able form in which they can precipitate is as a
compound of chromium, with which they are
very reactive. This occurs at the grain bound-
aries, where nucleation is favored, and depletes
those regions of chromium, rendering them less
corrosion resistant. A second effect is a loss of
toughness due to these precipitates. The diffusion
rates of carbon and nitrogen in ferrite are too
high to prevent this precipitation by quenching.
Fig. 1 The iron chromium phase diagram. Courtesy of Modern refining methods can reduce carbon
Thermo-Calc Software plus nitrogen to under 0.020%, but even this is
too high. So, to avoid the detrimental effects of The bcc structure of ferrite allows more
chromium carbide and nitride formation in fer- rapid diffusion than does the fcc structure of
rite, other benign carbides and nitrides such as austenite. This is true for both the interstitial
those of titanium or niobium are allowed to form diffusion of the elements helium, boron, car-
preferentially. This approach is called stabiliza- bon, nitrogen, and oxygen and the substitu-
tion and is used for most ferritic alloys today. The tional diffusion of all other elements. The rate
older approach, as characterized by alloy 430, is of diffusion of all elements, both interstitial
to permit chromium carbides and nitrides to form and substitutional, in ferrite is about two or
but then to perform a subcritical anneal to reho- three orders of magnitude higher than in
mogenize the chromium and coarsen the pre- austenite. The practical implication of this is
cipitates so that they have only a small negative that precipitation reactions generally cannot be
effect on mechanical properties. suppressed by quenching in ferrite if they in-
Hydrogen and boron are other elements that volve interstitial elements, whereas they can
can be interstitially dissolved in ferrite. Boron is be in austenite. Intermetallic phases can form
normally found at levels of around 5 to 10 ppm. more rapidly in ferrite. This becomes an issue
At higher levels, boron substitutes for carbon in only when total chromium plus molybdenum
carbides. Hydrogen is soluble to several parts exceeds about 20%, above which the sigma
per million by weight. It does not cause hydro- phase appears. This is thus only an issue for
gen embrittlement in annealed ferrite. If the fer- superferritic (high chromium content) alloys or
rite is cold worked, the solubility of hydrogen for the ferrite phase of duplex (ferrite-austen-
increases as the defect structure accommodates ite) alloys.
hydrogen atoms. In this condition, ferrite may The mechanical properties of the ferrite phase
be embrittled by hydrogen, especially if it en- are discussed extensively in Chapter 8, “Ferritic
ters the metal through corrosion processes like Stainless Steels.” Here, it is only necessary to
pitting. This is one explanation of, and the most note that ferrite in stainless steel closely resem-
likely explanation for, stress corrosion cracking. bles low-carbon steel in mechanical behavior. It
While hydrogen is easily removed by argon shares the following characteristics:
oxygen decarburization (AOD), assuming ab- • A toughness transition that occurs around
solutely dry blowing gases and additions are
room temperature
used, it can be picked up during pickling, weld- • Notch sensitivity
ing, or annealing as well as by corrosion. • A yield point phenomenon
All stainless alloys rely on having a uniform • Pronounced crystallographic anisotropy of
level of chromium and the other element,
mechanical properties
molybdenum, which assists in corrosion resist- • High stacking fault energies and low work-
ance, distributed throughout the matrix. If there
hardening rates
are locally low levels of these elements, local-
ized resistance to corrosion is reduced, and lo- These issues are dealt with in the same way as
calized corrosion can occur. This can occur by in carbon steel when these characteristics be-
the precipitation of any phase that is richer in come an issue. The first two are controlled by
chromium or other corrosion-resisting ele- reduction of interstitial levels and refining of
ments. Because chromium is a reactive ele- grain size. The yield point is eliminated by
ment, its success depends to a great degree on slight elongation by temper rolling or elimina-
maintaining the homogeneity required for tion of interstitial carbon and nitrogen, whose
proper corrosion-resistant performance. Incor- interaction with dislocations causes the yield
rect thermal processing is the main way homo- point. The anisotropy is either utilized to advan-
geneity can be lost. Stabilizing makes it much tage by maximizing it, as in the case of deep-
easier to keep chromium from segregating in drawing alloys, or minimized by refining grain
ferritic alloys. size and randomizing grain orientation by spe-
A by-product of stabilization with titanium is cial thermomechanical processing.
that oxygen and sulfur are also eliminated as Ferrite has a greater thermal conductivity
compounds of titanium along with carbon and and lower thermal expansion than austenite. Its
nitrogen. These impurity elements would other- strength decreases with temperature more than
wise also precipitate as compounds containing that of austenite, but the good match in thermal
some chromium, potentially depleting chromium expansion between the ferrite and its oxide
in the vicinity of their precipitation. still makes it an excellent high-temperature
material. Ferrite has very nearly the same cor- basis. Their use is limited by their solubility and
rosion resistance as austenite, but since ferrite their tendency to form precipitating compounds
can hold no nitrogen in solution, it cannot ben- with chromium. Manganese acts largely through
efit from this element. In duplex alloys, the its ability to promote nitrogen solubility. Super-
ferrite is generally the more corrosion resistant austenitic stainless steels, such as S34565, use 4
phase because it is richer in chromium and to 6 % manganese to permit nitrogen levels of
molybdenum. 0.4 to 0.6% to be achieved, resulting in higher
pitting corrosion resistance.
Since all stainless steels contain principally
Austenite
iron and chromium, the addition of a substantial
The second major constituent phase of the amount of austenitizing elements is necessary to
stainless steel alloy system is austenite. Austen- transform the structure to austenite. As a rule of
ite has an fcc atomic structure. The fcc structure thumb, iron alloys require about 17% chromium
is common in many transition metals to the right and 11% nickel (or its equivalents) to remain
of iron in the periodic table. As stated, the fcc austenitic at room temperature. One percent
structure should be considered normal for metals nickel can be replaced by about 2% manganese
well below their melting temperature as it is a as long as nitrogen is present to maintain the
denser structure. The presence of the bcc struc- same phase stability. The omnipresent carbon
ture relates to the unpaired 3d electrons, which and nitrogen have an effect 30 times that of
provide ferromagnetism. Adding elements to nickel, so even in the small amounts in which
iron that causes pairing of the 3d electrons di- they are normally present, they have a signifi-
minishes ferromagnetism and promotes the fcc cant effect. These stabilizing factors are mapped
structure. Nickel and manganese are the most in the Schaeffler diagram of Fig. 2 (Ref 1),
prominent alloying elements that do this, but the whose purpose is to predict the phase makeup
interstitials carbon and nitrogen are the most of weld metal. Since welds solidify relatively
powerful austenite stabilizers on a percentage rapidly, no carbides or intermetallic phases
Fig. 2 Schaeffler-Delong constitution diagram showing phases present in as-solidified stainless steels at room emperature as a func-
tion of composition demonstrating carbon and nitrogen contributions to nickel effects. Adapted from A.L. Schaeffler,
Constitution Diagram for Stainless Steel Weld Metal, Met. Prog., Vol 56, Nov 1949, p 680–688; and W.T. Delong, A Modified Phases
Diagram for Stainless Steel Weld Metals, Met. Prog., Vol 77, Feb 1960, p 98
form, and only ferrite, austenite, and martensite and then frozen in a state of supersaturation in
will be present. Thus, they provide useful infor- the austenite when it forms on cooling. The sul-
mation about the compositional effects on phase fur and oxygen then precipitate during cooling or
development in nonequilibrium situations. The subsequent hot working as isolated inclusions.
nickel equivalent (vertical axis) summarizes The interface between these inclusions and the
how nitrogen, carbon, and other elements com- matrix is the locus of corrosion pit initiation,
bine to create a nickel-like effect. The horizon- quite probably because of chromium depletion
tal axis does the same for chromium and those occurring during and as a result of inclusion
elements that have a similar effect. growth. When an alloy solidifies as austenite,
In most common stainless steels, austenite is sulfur immediately segregates to the grain
normally present in the metastable state, for ex- boundaries because of its low solubility in
ample, the retained austenite in alloy steels. Those austenite, and it forms a low-strength film with a
with carbon above 0.02% would eventually break low melting temperature. This causes poor hot
down into austenite plus carbides, and those with workability and hot cracking of welds.
less than about 30% chromium plus nickel will The diffusion rates in austenite are quite low
form martensite if deformed sufficiently. But in compared to ferrite, so even interstitial elements
the annealed state, the austenite in standard cannot move quickly enough to precipitate below
austenitic stainless steels will remain indefinitely about 400 °C (750 °F). This permits carbon and
as fully austenitic without precipitates unless nitrogen to exist in very high degrees of supersat-
heated above 400 °C (750 °C) for protracted peri- uration if introduced below this temperature, as
ods of time or deformed extensively. is done by various proprietary processes. The
Interstitial elements are much more soluble in low diffusion rates restrict such colossally super-
austenite than in ferrite. Of these, only nitrogen saturated zones to thin surface layers, but they
is considered a beneficial alloying element. It can reach phenomenal hardness of over Rc 70.
both strengthens and improves the pitting corro- The austenite structure does not discourage the
sion resistance of austenite. Carbon has a paral- formation of intermetallic compounds such as
lel effect, but its tendency to form chromium sigma, but it does, fortunately, make their forma-
carbides limits its use and in fact leads to its tion very sluggish, as seen in Fig. 4. The differ-
minimization in most alloys. Before the AOD ence of three orders of magnitude for carbide
was developed and carbon levels in stainless formation reflects the difference between the dif-
steels were higher, austenitic stainless steels fusion of carbon and that of substitutional ele-
were sometimes stabilized by titanium or nio- ments. The formation of sigma in ferrite is about
bium to counter the effects of carbon. Both car- 100 times faster than in austenite. Sigma is al-
bon and nitrogen stabilize the austenite phase, most never seen in commercial 316 alloys.
permitting lower levels of nickel to be used in
austenitic alloys.
Interstitial atoms of carbon and nitrogen dis-
tort the fcc lattice, causing it to expand about
1% linearly per 1 wt% of solute (Fig. 3) (Ref 2).
This produces solid solution hardening of the
austenite. The work hardening of austenite is in-
creased by nitrogen. A third interstitial solute,
hydrogen, produces the same effect but to a
lesser degree. Austenite is not embrittled by hy-
drogen to the extent ferrite or martensite is, but
hydrogen does raise its flow stress and hardness
while lowering its work-hardening rate.
Sulfur and oxygen are considered impurities
because they form inclusions, usually chrome/
manganese silicates and sulfides. If present in
sufficient amounts, sulfur and oxygen precipitate
as primary inclusions before or during solidifica-
tion. In most austenitic stainless alloys, the re-
mainder of these elements are near saturation in
the as-solidified ferrite at very high temperatures Fig. 3 Lattice expansions due to carbon. Source: Ref 2
The mechanical properties of austenite are The physical properties of austenite com-
quite different from those of ferrite. Austenite is pared to ferrite include lower thermal and elec-
characterized by: trical conductivity and greater thermal expan-
sion. It is also, of course, nonmagnetic.
• Low stacking fault energies leading to high
work-hardening rates
• Good toughness even at very low tempera- Martensite
tures Martensite is a phase that forms from the dif-
• Low notch sensitivity fusionless shear of austenite to a distorted cubic
• Lack of a sharp elastic limit or hexagonal structure. This transformation can
• Good high-temperature strength occur spontaneously on cooling or isothermally
• Fairly isotropic mechanical properties with externally applied deformation. It is essen-
tially ferrite that has been formed with a super-
While there is not a great deal of difference saturation of carbon. The resulting structure is
in the yield strengths of austenitic and ferritic very fine and highly faulted, making it quite
alloys of similar alloy levels, austenitic alloys hard. As in carbon steel, the hardness of the
are more ductile, have high work-hardening martensite increases dramatically with intersti-
rates, and therefore have higher tensile tial content because of the huge strain intersti-
strengths. Austenite can be cold worked to ex- tial elements impose on the bcc lattice, distort-
tremely high strengths, around a maximum of ing it into tetragonality.
2000 MPa (290 ksi). Chapter 3, “Austenitic Martensite in stainless steels is restricted to
Stainless Steels,” gives a more thorough and alloy levels at which austenite can form at
quantitative treatment of the mechanical prop- higher temperatures but at which the austenite is
erties of austenite. unstable at ambient temperatures. This gives
In duplex stainless steels, a secondary austen- martensite a fairly narrow composition range.
ite, γ2, can form from ferrite below 650 °C The lowest alloy level is that of the basic 12%
(1200 °F). At this temperature, it has the same chromium steels with 0.1 to 0.2% carbon. The
composition as the ferrite from which it forms most highly alloyed martensites are found in the
and is called type 1. In the 650 to 800 °C (1200 precipitation-hardening grades. Thus, marten-
to 1470 °F) range, a range that can be encoun- sitic stainless steels are inherently limited in
tered in the heat-affected zone (HAZ) at γ/δ corrosion resistance to a level no better than a
boundaries during welding, another type forms. 17 or 18% chromium alloy and often barely
This so-called secondary austenite, γ2, type 2, is qualify as stainless after the chromium tied up
somewhat enriched in nickel over the ferrite as chromium carbide is recognized as not con-
from which it forms but poorer in nitrogen than tributing to the corrosion resistance.
the primary austenite, giving it poorer corrosion The as-formed martensite to the degree it has
resistance. Secondary austenite can also coform significant carbon content is hard and requires
with sigma as γ/δ grain boundaries are depleted tempering to give it adequate toughness. The
of chromium. This secondary austenite is called tempering reaction is the precipitation of car-
type 3 and is also poor in chromium. bon in the form of carbides with the concurrent
loss of internal strain in the martensite lattice.
The complexities of tempering require its dis-
cussion in detail to be found in Chapter 3,
“Martensitic Stainless Steels.” It is worth noting,
however, that all tempering involves carbide for-
mation, thus losing some corrosion-fighting
chromium.
There are two forms of martensite, the ε, ep-
silon, and the α', alpha prime. Epsilon is formed
in steels with low stacking fault energy, which
are primarily the leaner austenitic alloys. Thus, it
forms at cryogenic temperatures or by cold
working. It appears in martensitic alloys of the
precipitation-hardening type. It is nonmagnetic,
Fig. 4 Precipitation kinetics in 316 stainless steel. Source: Ref 3 has a hexagonal close-packed (hcp) structure,
and is very difficult to identify microscopically. chromium or molybdenum, causing localized

The a' martensite is the familiar magnetic vari- lower corrosion resistance. Intermetallic phases
ety known in alloy steels that forms both by form by diffusion of substitutional alloying ele-
quenching and by deformation. ments, which makes their precipitation slower
The mechanical properties of stainless than that of carbides, but they can form in a
martensite are parallel to those of alloy steels. matter of minutes in alloy-rich grades. Defor-
The high quantity of alloying elements in mation, which enhances substitutional diffu-
stainless give an extreme depth of hardening, sion, accelerates their formation. The principal
so there is no concern with ancillary phases intermetallic phases are described next.
such as bainite. The physical properties are Alpha Prime. Not to be confused with
very close to those of ferrite of the same com- martensite, alpha prime is an ordered iron-
position. chromium phase (i.e., iron and chromium atoms
occupy specific, rather than random, sites on
Intermetallic Phases two intersecting superlattices). This structure is
quite brittle. It forms at relatively low tempera-
The number of phases that can coexist in an tures, between 300 and 525 °C (570 and 980
alloy is proportional to the number of alloying °F). Before its true nature was understood, its
elements in the alloy. Table 1 lists data on the presence was known through its causing the
more common precipitates found in stainless phenomenon called 475 embrittlement, origi-
steel. It is not surprising that stainless steel with nally called 885 °F embrittlement. This is some-
iron, chromium, nickel, manganese, silicon, and times confused with temper embrittlement,
often molybdenum, titanium, and niobium which occurs in the same temperature range but
should have numerous ancillary phases. Inter- is caused by phosphide precipitation on prior
metallic phases are normally hard and brittle. austenite grain boundaries of martensite. Alpha
They can render the bulk alloy brittle when they prime precipitation can cause 475 embrittle-
form along grain boundaries. The other concern ment in ferritic or duplex stainless steels and
arising from intermetallic phase formation is the limits their use in this temperature range but not
depletion from the surrounding matrix of at higher temperatures, at which the phase dis-
solves. This phase forms at chromium contents
Table 1 Precipitated phases found in stainless as low as 15%, but fortunately it takes a rela-
steels tively long time to form, on the order of hours,
Precipitate Structure Parameter, A Composition so it will not occur inadvertently during thermal
processing such as welding or annealing.
NbC fcc(a) a = 4.47 NbC
Sigma. Sigma is a brittle tetragonal phase
NbN fcc a = 4.40 NbN richer in chromium and molybdenum than ei-
TiC fcc a = 4.33 TiC ther the ferrite or austenite matrix around it. It
TiN fcc a = 4.24 TiN forms preferentially at ferrite-austenite bound-
Z-phase Tetragonal a = 3.037 c = 7.391 CrNbN aries in the temperature range 600 to 1000 °C
M23C6 fcc a = 10.57–10.68 Cr16Fe5Mo2C (e.g.)
(1110 to 1470 °F) in alloys with more than
about 18% chromium plus molybdenum. Its
M23(C,B)6 fcc a = 10.57–10.68 Cr23(C,B)6
composition is sometimes given as (CrMo)35
M6C Diamond a = 10.62–11.28 (FeCr)21Mo3C; (FeNi)65, but examination of the iron-chromium
cubic Fe3Nb3C; M5SiC
phase diagram shows that it is archetypically an
M2N Hexagonal a = 2.8 c = 4.4 Cr2N equiatomic iron chromium compound. It is
MN Cubic a = 4.13–4.18 CrN strongly promoted by silicon and suppressed by
Gamma fcc a = 3.59 Ni3(Al,Ti) nitrogen. Stabilized alloy grades show more
prime rapid sigma formation than unstabilized alloy
Sigma Tetragonal a = 8.80 c = 4.54 Fe, Ni, Cr, Mo grades (e.g., 347 versus 304). In unstabilized al-
Laves Hexagonal a = 4.73 c = 7.72 Fe2Mo, Fe2Nb
loys the prior precipitation of carbides destabi-
phase lizes austenite, leading to subsequent sigma for-
Chi phase bcc(b) a = 8.807–8.878 Fe36Cr12Mo10
mation. This makes alloys like 310H,
essentially 25Cr-20Ni, especially prone to
G-phase fcc a = 11.2 Ni16Nb6Si7,
Ni16Ti6Si7
sigma formation.
(a) fcc, face-centered cubic. (b) bcc, body-centered cubic.
Sigma forms much more rapidly from ferrite
than from austenite because of the 100-fold
higher diffusion rate of alloy elements in ferrite. The precipitation of the carbide from ferrite
This makes it a much larger issue in superfer- occurs at grain boundaries, is extremely rapid,
ritic and duplex alloys, which have high and cannot be suppressed by quenching. Less
chromium and/or molybdenum levels. Chapter than 20 ppm carbon content is required to pre-
7, “Duplex Stainless Steels,” contains an in- vent its precipitation from ferrite, although up
depth discussion of sigma. to 50 ppm can be effectively kept in solution by
Chi. Chi, χ, is similar to sigma except it con- very vigorous quenching. From austenite, car-
tains more molybdenum and less chromium and bide precipitation occurs below about 900 °C
has a cubic structure. It can coexist with sigma (1650 °F) for carbon levels under 0.10% and at
and forms in the same temperature range. It also 650 °C (1200 °F) for carbon levels below
precipitates at ferrite-austenite boundaries and 0.03%. For practical purposes, precipitation
has the same deleterious effects. ceases below 500 °C (930 °F) due to the slow-
Laves Phase. The laves phase has the struc- ing diffusion of carbon. While carbon is essen-
ture A2B where A is iron or chromium and B is tially insoluble in austenite at room tempera-
molybdenum, niobium, titanium, or silicon. It ture, quenching can easily preserve up to 0.10%
forms at 550 to 650 °C (1020 to 1200 °F) over in supersaturation, as is commonly seen in type
the course of hours. Thus, although its effect 301 stainless.
would be deleterious, it seldom becomes a prac- The carbide precipitation occurs first at grain
tical problem. It is possible for it to form at tem- boundaries. The chromium that combines with
peratures below sigma and above alpha prime, the carbon comes from the matrix in the imme-
but the long times for formation make it rare. diate vicinity and therefore decreases the
chromium content of that region, giving rise to
Carbides, Nitrides, Precipitation Hardening, the phenomenon of sensitization, which comes
from the original phrase “sensitization to inter-
and Inclusions
granular corrosion.” Nickel and molybdenum
Carbon and nitrogen are very important in all decrease the solubility of carbon and thus accel-
steels, but they take on a special significance in erate the precipitation. Nitrogen retards precipi-
stainless steel because chromium, the essential tation. Cold work accelerates precipitation. The
alloying element of stainless steel, reacts more carbide has a hardness of about Rc 72. This
vigorously with carbon and nitrogen than iron makes the phase a useful constituent in wear
does. Except for its role in hardening martensite resistance in martensitic alloys.
and strengthening austenite at high tempera- In higher carbon grades such as the marten-
tures, carbon is almost universally a detrimental sitic stainless alloys, additional, more carbon-
impurity from a corrosion point of view and is rich, carbides may form. These include M7C3
minimized. Its beneficial effect on corrosion re- and M3C. The latter carbide forms during the
sistance when it is in solution is negligible low-temperature tempering of martensite, while
because so little of it can be held in solution. the former precipitates at higher temperatures.
Nitrogen has a lesser tendency to form com- Stabilizing carbides are those that are formed
pounds with chromium, so it is considered a by the intentional addition of elements such as
beneficial alloying element in austenite but not titanium and niobium. These elements form car-
in ferrite, in which it has essentially zero solu- bides of the type MC (metal carbide). The car-
bility. Common carbide and nitride precipitating bon in these compounds may be replaced by ni-
phases are also listed in Table 1. trogen or, in the case of titanium, sulfur. These
Carbides. M23C6 is the main carbide found in carbides form preferentially over chromium car-
stainless steel. Its structure is orthorhombic, and bides and thus prevent sensitization. They pre-
it contains both iron and chromium. It can form cipitate in both the liquid and solid states. In the
at any temperature at which the host austenite or solid state, the precipitate normally forms within
ferrite becomes saturated with carbon. It is grains. The Ti(CN) appears as a cube of gold
mainly chromium carbide, but iron can substitute TiN surrounded by gray TiC. The Nb(C,N) is
for chromium up to about 50%. Other elements, less regularly shaped. They affect mechanical
such as tungsten, vanadium, and molybdenum, properties in ferrite both by their influence on
can also dissolve in this carbide. The ratio of recrystallization and by their ability to act as nu-
chromium to iron in the carbide increases with cleation sites for brittle fracture
time and temperature, as chromium diffusion Nitrides. At low levels, nitrogen can substi-
permits, up to a maximum of 4 or 5 to 1. tute for carbon in M23C6. At higher nitrogen
levels, Cr2N can form. This can occur in duplex tal as inclusions have been shown to be the initi-
alloys if they are heated to a solution annealing ation sites for corrosion pits, which have been
temperature at which the alloy has high solubil- linked to both their sulfur ions disrupting the
ity for nitrogen. Cooling from these tempera- passive layer and their chromium content caus-
tures can cause the excess nitrogen to precipi- ing slight local chromium depletion.
tate as needles of Cr2N. Another nitride CrN
can form in HAZs of welds.
Precipitation-Hardening Phases. Phases
that have a very similar lattice match to the par- Properties of Stainless Steels
ent phase can precipitate coherently, that is,
without changing the continuity of the crystal Physical and mechanical properties of repre-
lattice. In these cases, the slight mismatch sentative stainless steel alloys are summarized
causes a strain that can significantly restrict dis- in Appendix 2. Properties are also discussed in
location movement and thereby strengthen the chapters specific to each alloy family. The
matrix. One such precipitate is gamma prime, reader is referred to primary sources, such as
an intermetallic, ordered, fcc phase with the company web sites, such as Ref 4 and 5.
composition Ni3(AlTi). Copper forms the ep-
silon phase, essentially pure copper, which
causes precipitation hardening. The secondary REFERENCES
hardening of martensite due to the precipitation
of molybdenum nitride or carbide is also a pre- 1. D.J. Kotecki, Welding of Stainless Steels,
cipitation-hardening reaction. Welding, Brazing, and Soldering, Vol 6,
Inclusions. Inclusions are principally oxides ASM Handbook, ASM International, 1993,
and sulfides that form in the melt (type I), at the p 677–707
end of solidification (type II), or in the solid 2. G.E. Totten, M. Narazaki, R.R. Black-
(type III). Type I inclusions are the largest and wood, and L.M. Jarvis, Failures Related to
are globular. Except when they are deliberately Heat Treating Operations, Vol 11 ASM
kept to improve machinability, they are physi- Handbook, ASM International, 2002, p
cally removed by various steelmaking practices. 192–223
Type II inclusions form in interdendritic spaces 3. High Performance Stainless Steels, Refer-
as the solubility of oxygen and sulfur drop on ence Book Series 11 021, Nickel Develop-
solidification. Type III inclusions precipitate the ment Institute, p 16
remaining oxygen and sulfur, up to 100 ppm for 4. ASM Handbook, Vol 1, Properties and
normal manganese-silicon killed stainless Selection, ASM International, 1990
steels, in the solid state either on preexisting in- 5. ASM Speciality Handbook, Stainless Steels,
clusions or as micron-size particles. Inclusions ASM International, 1996
are mainly oxides and sulfides of silicon and
manganese. If more reactive elements, such as
aluminum or titanium, are present, their oxides SELECTED REFERENCES
and sulfides can also be present.
Sulfides and oxysulfides can be beneficial for • D.J. Kotecki and T.A. Siewert, WRC 1992
machining as solid-state lubricants and chip Constitution Diagram, Welding Journal, Vol
breakers. Otherwise, their presence is detrimen- 5, 1992, p 171s–178s
CHAPTER 2
Corrosion Theory
Summary Electrochemical Reactions
THIS CHAPTER INTRODUCES THE funda- In electrochemical reactions, charge is trans-

mentals of electrochemical theory as it pertains ferred across interfaces of species of different
to corrosion. Topics covered include an overview chemistry. Consider, for example, the reaction:
of electrochemical reactions, Faraday’s law, the
Nernst equation, galvanic versus electrochemical 2Fe (s) + O 2 (g) + 2H 2 O → 2Fe 2+ + 4 OH − (Eq 1)
cells, and Pourbaix diagrams. The examples pro-
vided relate these fundamentals to the corrosion
resistance of stainless steels. An inspection of this reaction suggests that
three phases must be present for the reaction to
proceed: an ion-conducting phase (water-based
solution), a metallic phase (iron), and a gas
Introduction phase O2(g). Second, electrons have been trans-
ferred from the metallic phase, iron to O2 +
Corrosion—the environmental degradation H2O. Figure 1(a) shows the arrangement of an
of materials through electrochemical reac- experimental setup in which Reaction 1 could
tions—is a key subject for more or less all proceed.
classes of alloys that fall within the broad defi- On the left, iron is allowed to dissolve
nition of stainless steels because these alloys according to:
were developed with the intention of prevent-
ing corrosion. This chapter aims first to provide 2Fe (s) → 2Fe 2+ + 4 e − (Eq 2)
an introduction to the fundamentals of electro-
chemical theory as it pertains to corrosion. resulting in Fe2+ ions that dissolve in the
Thermodynamics are presented in light of elec- water-based solution and electrons that are car-
trochemical potentials as opposed to purely ried to the right side, where they participate in
chemical ones. Chapter 3 introduces the formal the reaction:
terms needed to describe electrode reaction ki-
netics. Chapter 4 describes the various forms of O 2 + 2H 2 O + 4 e − → 4 OH − (Eq 3)
corrosion and how they are related to alloy
metallurgy, chemistry, and structure. Chapter 5 Inside the water-based solution, ions (Fe2+,
focuses on oxidation. For an in-depth study of OH⫺, H+, or any others) migrate, thereby con-
electrochemical kinetics and electroanalytical stituting a so-called ionic current. This current
methods, Ref 1 is recommended. For a broader together with Reactions 2 and 3 and the trans-
study of corrosion, the reader is referred to port of electrons from left to right form a closed
texts by Jones (Ref 2), Uhlig and Revie (Ref 3), circuit called an electrochemical cell. The cell is
and Fontana (Ref 4) and to ASM Handbook, made up of four parts: the two electrodes where
Volume 13A (Ref 5). the charge transfer Reactions 2 and 3 take place
often are described as half cells, for example,

Fe / O 2 / OH − and Fe / Fe 2+ .
Faraday’s Law
If the cell in Fig. 1(a) was allowed to proceed

and thermodynamics favored to proceed ac-
cording to the direction in Reaction 1, then a
current i will flow from the anode to the cath-
ode, and the amount of charge passed per unit
time as a result of this current will be linked to
the amount of iron dissolved per unit time or the
amount of oxygen reacted per unit time by
virtue of Eq 2 and 3. This is given by Faraday’s
law:
nNF = it (Eq 4)
Here, i * t is the charge passed (in coulombs);

N is the moles of consumed/produced specie
(e.g., moles consumed iron in Reaction 2); n is
the ratio of electrons to consumed/produced
species, which in the case of Reaction 2 will be
2; and F is Faraday’s constant, which is essen-
tially the charge in coulombs corresponding to
1 mole of electrons.
Fig. 1 Schematic illustration of (a) a differential aeration cell

The Nernst Equation
involving iron dissolution and (b) the same cell with a
variable resistor and voltmeter
Electrochemical reactions require a transfer
(the anode and cathode, respectively), an elec- of charge; hence, there is a coupling between
trolyte, and an electron pathway. It should be chemical and electrical energy. Consider the hy-
noted that electrodes are interfaces that require pothetical setup in Fig. 1(a) with the addition of
several phases to be in contact. Oxidation, Re- a variable resistor and a voltmeter, resulting in
action 2, occurs at the anode and reduction, Re- the arrangement shown in Fig. 1(b). Thermody-
action 3, occurs at the cathode. The electrolyte namically, the Gibbs free energy of the cell is
is the medium through which the ions migrate; that of Reaction 1:
in the case of corrosion reactions, this is most
commonly a water-based solution, but at high ΔG = ΔH − T ΔS
(a ) (a )
4 2
temperatures it could be a solid oxide. The final (Eq 5)
OH − Fe 2+
constituent of the electrochemical cell is a path- = ΔG 0
+ RT ln
way through which electrons can migrate from aH O PO
2 2
the anode to the cathode.

As a shorthand notation, electrochemical cells where ⌬G is the Gibbs free energy, H is the en-
are written by separating components within a thalpy, S is the entropy, R is the gas constant,
phase by a comma and separating phases by a and T is the absolute temperature. If this is neg-
slash; gaseous species are written next to their ative, the reaction would be expected to proceed
conducting electrode. For example, the cell de- spontaneously as written in Reaction 1. Let us
scribed in Fig. 1(a) would be recorded as assume that this is the case. The thermal heat
Fe / O 2 / OH − , Fe 2+ / Fe . This cell is an example produced by the system can be divided into two
of a differential aeration corrosion cell, which is parts: the thermal heat produced by the cell Qt
discussed later. Processes at a single electrode and the heat produced at the resistor QRes. QRes
Chapter 2: Corrosion Theory / 13
in this case is heat, but in essence it represents Here, pi and ri are the concentrations of reac-
the available energy or work, which in the case tant and products, respectively, and αi and βi are
of a resistance is given by the product of charge the numbers that are needed to balance the reac-
passed times potential difference. If the resist- tion stoichiometrically. In the case of Reaction
ance approached infinity ( R→ ), Reaction 1 1, Eq 10 would be:
would proceed through infinitesimal steps and
⎛
( ) ( ) ⎞⎟
4 2
can be considered thermodynamically reversible.
RT ⎜ aOH− aFe 2+
In this case, the thermal heat produced by the Erxn = Erxn
0
− ln (Eq 11)
cell is minimized and according to thermody- 4F ⎜ aH O PO ⎟
⎝ 2 2 ⎠
namics is given as Qt = Qrev = T⌬S1. On the
other hand, the net work gained QRes is maxi- If the emf according to Eq 11 is positive, this
mized and constitutes the rest of the free energy: means that the free energy is negative (according
to the Nernst equation); hence, the net reaction is
QRes = ΔG = ΔH − T ΔS (Eq 6) thermodynamically favored as it is written in
Reaction 1. By inspection of Eq 11, it can be
As mentioned, the energy dissipated through seen that it is the difference between two hypo-
the resistance is charge passed times potential thetical half reactions, ( Erxn = EO /OH− − EFe 2+ /Fe )
difference, and in this case the potential differ- defined as:
2
ence is the reversible potential difference E;

thus, in an absolute sense: ⎛
( )
⎞
4
RT ⎜ aOH− ⎟
EO = EO 0
− ln (Eq 12)
ΔG = nFE (Eq 7) 2 /OH − 2 /OH −
4 F ⎜ aH O PO ⎟
⎝ 2 2⎠
Here, n is the number of electrons passed per which corresponds to the reduction Reaction 3
atom of iron reacted, and F = 96,485 C per mole and:
electrons, is Faraday’s constant. The reversible
potential difference E represents the potential
difference between the two electrode reactions EFe 2+ /Fe = EFe 2+ /Fe −
0 RT ⎛ ( aFe ⎞
ln ⎜
4
⎟
) (Eq 13)
(cathode and anode), and as such they are asso- 4 F ⎜⎝ aFe 2+ ⎟⎠
ciated with Reaction 1 rather than a physical
cell. The potential difference is referred to as which corresponds to the reverse of Reaction
the electromotive force (emf) of the cell. It is 2, that is, if it was a reduction reaction. The
also referred to as the open circuit potential be- potentials as written in Eq 12 and 13 are
cause it is the potential measured by the volt- called reduction potentials, and because
meter in Fig. 1(b) when a negligible current Erxn = EO /OH− − EFe 2+ /Fe has to be positive for the
flows. It is defined here as Erxn. By convention, 2
reaction to be thermodynamically favored as
this potential is positive for a spontaneous reac- written in Eq 1, the reduction potential E O 2 /OH−
tion (as opposed to the chemical free energy, has to be larger than E Fe 2+ /Fe. If it was not, then
which is negative); hence, Eq 7 becomes: Reaction 1 would proceed in the reverse direc-
tion, which means that the electrode Reactions
ΔG = − nFErxn (Eq 8) 2 and 3 would be reversed and thus so would
the anode and cathode of the cell.
and if all elements have unit activities: It is useful to list reduction potentials for half-
cell reactions, just as it is useful to list free en-
ΔG 0 = − nFErxn
0
(Eq 9) ergy data. However, half-cell potentials (like
any electrical potentials) cannot be measured in
an absolute sense; only potential differences can
Equation 8 is the Nernst equation. By virtue be measured. ( Erxn = EO /OH – − EFe 2+ /Fe can be
of Eq 8 and 9 and the expression for Gibbs free 2
measured because it is a difference.) Therefore,
energy of a reaction (e.g., Eq 5), an expression
half-cell potentials are measured with respect to
for Erxn is obtained:
a reference electrode. Reference electrodes are
constructed such that they have a stable potential;
RT ⎛ p1 1 p2 2 p3 3 ... ⎞
α α α
Erxn = Erxn
0
− ln ⎜ ⎟ (Eq 10) this is discussed further in Chapter 3. A com-
nF ⎝ r1β1 r2β2 r3β3 ... ⎠ mon reference electrode in aqueous solutions is
the normal hydrogen electrode (NHE), also Galvanic versus Electrochemical Cells
known as the standard hydrogen electrode
(SHE), with a potential set (arbitrarily) as zero When reactions in a cell occur spontaneously
at all temperatures. The NHE is schematically in the direction dictated by the open-circuit po-
shown in Fig. 2. In shorthand notation, it is: tential of a cell that is positive ( Erxn > 0) , a cur-
Pt / H 2 (a = 1)/ H + (a = 1) , and the half-cell reac- rent flows as shown in Fig. 3(a). This is the case
tion is: in environmentally caused electrochemical cor-
rosion reactions. It also is the case in fuel cells
2 H + + 2 e− = H 2 (Eq 14) and batteries (under discharge), in which the
Table 1 (Ref 6) lists half-cell reduction stan- current is used as electricity. These types of
0
dard potentials ( EOx/Re) versus NHE that are a cells are called galvanic cells, in which chemi-
result of the emf of the following types of cells cal energy is converted to electrical energy.
(for Reaction 2, as an example): Most of the discussion in the following chapters
concerns these types of cells. In electrolytic
Pt / H 2 (a = 1)/ H + (a = 1), Fe 2+ (a = 1) / Fe cells (Fig. 3b), an imposed electrical potential
counters the “natural” cell potential to drive a
reaction in a desired direction. These types of
cells are used for many metallurgical processes,
such as electroplating, electrorefining and elec-
troextraction (e.g., the Hall-Heroult aluminum
smelting cell), and for other applications, such
as charging batteries. In the case of corrosion,
the principle is used for protection against cor-
rosion. In electrolytic cells, electrical energy is
converted to chemical energy.
Table 1 Standard half-cell reduction potentials

versus the normal hydrogen electrode
Standard half-cell reduction
Reaction potential vs. NHE(a) (V)
Fe 3+ + e− = Fe 2 + 0.771
O 2 + 2H 2 O + 4e − = 4 OH – (pH = 14) 0.401
2 H + + 2 e− = H 2 0.000
Ni 2 + + 2 e− = Ni –0.250
Fe 2+ + 2e− = Fe –0.447
Cr 3+ + 3e− = Cr –0.744
2 H 2 O + 2 e− = H 2 + 2 OH − (pH = 14 ) –0.828
Fig. 2 The normal hydrogen electrode (NHE) (a) NHE, normal hydrogen electrode. Source: Ref 6
Fig. 3 Schematic of (a) galvanic cell and (b) electrolysis cell

Corrosion Tendency ing to Eq 13, which, assuming a Fe2+ activity of

10−6 (this is an arbitrary value but is usually
The tendency to corrode, that is, whether a taken to represent a low ion concentration), be-
system consisting of anode, cathode, and elec- comes at room temperature (using Table 1 for
trolyte can react thermodynamically, is deter- the standard potential):
mined by evaluating Erxn. If this is positive, then
there is thermodynamically a possibility for cor- EFe 2+ /Fe = −0.447 − 0.0295 log(aFe 2+)
rosion. The rate of corrosion, which is in most
cases determined by corrosion kinetics, is dis- = –0.624 V (vs. NHE) (Eq 16)
cussed in Chapter 3. Consider, for example, a
case of iron in aerated water. Figure 1 (with Figure 4(a) shows a schematic plot of the two
electrode Reactions 2 and 3) can be viewed as reduction potentials (Eq 15 and 16) versus pH.
an idealized equivalent cell for this situation. It Because a spontaneous reaction requires Erxn
should be noted, however, that the locations of to be positive, if the only pertinent reactions
anode(s) and cathode(s) on the iron surface can- were Eq 2 and 3, this means that corrosion (due
not be identified with ease. At room temperature to iron dissolution to Fe2+ and oxygen reduc-
(298 K), 1 atm oxygen partial pressure, and tion) is possible when the line representing
using Table 1, Eq 12 can be written by assuming EO2/OH− (Eq 15) lies above the line representing
unit activity for water and unit activity coeffi- EFe / Fe (Eq 16). This is indicated by the region
2+
cient for OH−: shaded in gray in Fig. 4(a). Hydrogen reduction

is another possible cathode reaction in water:
EO – = 0.401 + 0.059 pOH
2 / OH
2 H + + 2 e− → H 2 (Eq 17)
= 0.401 + 0.059(14 − pH )
= 1..227 − 0.059 pH V (vs. NHE) (Eq 15) and its reduction potential is (using the defini-
tion of pH):
Here, the following definition of pH has been
RT PH2
used: pH = –log CH+, pOH = –log COH− and pH EH+ / H = EH0 + / H − ln
+ pOH = 14. Similarly, the iron dissolution Re-
2 2 nF aH2 +
action 2 will have a reduction potential accord- = 0 − 0.059 pH V vs. NHE (Eq 18)
Fig. 4 Reduction potential versus pH for iron and (a) oxygen gas reduction and (b) hydrogen ion reduction
Figure 4(b) shows the condition in which corro- Fe(OH)2 = HFeO 2− +H + ,

sion under deaerated conditions (due to iron dis-
solution to Fe2+ and hydrogen ion reduction) is (
pH = 14.30 + log aHFeO−
2
) (Eq 19b)
possible as a gray shaded region. In Fig. 4(a)
and (b), the regions where iron is stable are
denoted as immunity (corresponding to immu-
nity from corrosion). When comparing these Fe 3+ + 3H 2 O = Fe(OH)3 +3H + ,
two figures, it is noteworthy that hydrogen ions pH = 1.613 – (1/3) log(a Fe3+ ) (Eq 19c)
are able to cause corrosion only under relatively
low pH conditions, whereas oxygen gas is able
to corrode iron in the entire pH range. Since these are independent of potential, they
will appear as vertical lines (see lines 19a to 19c
in Fig. 5a).
The following pH-independent electrochemi-
The Construction of Pourbaix Diagrams cal reactions need to be considered, and they
will result in horizontal lines (Fig. 5a):
Figures 4(a) and (b) are types of phase dia-
grams that show the stable phases in an area
bounded by pH and potential. In reality, several Fe 2+ + 2 e− → Fe,
electrochemical and chemical reactions need to
be considered when constructing these types of EFe 2+ /Fe = −0.447 + 0.0295 log aFe 2+( ) (Eq 20a)
diagrams. Each reaction is represented by a line.
In the case of iron, the following chemical reac-
tions will have to be considered (the pH de- )
HFeO 2− + H 2 O = Fe ( OH 3 + 2e− ,
)
pendency of these reactions is listed next to
them [Ref 7] and since they are not electro-
chemical, they are evaluated from the equilib-
EFe(OH) / HFeO2– = −0.810 − 0.0591 log aHFeO–
3
( 2
rium constants): (Eq 20b)
The following electrochemical reactions will

Fe 2 + + 2 H 2 O = Fe(OH )2 + 2 H + ,
depend on pH and thus will be sloped depend-
(
pH = 6.65 − 0.5 log aFe 2 + ) (Eq 19a) ing on this dependence (Fig. 5a).
Fig. 5 Pourbaix diagram for iron. (a) Schematic matching Eq 19 to 21 in text to lines. (b) Actual complete diagram. Source: Ref 7
Fe + 2H 2 O = Fe ( OH 2 + 2H + + 2e− ,)
EFe(OH) = −0.0470 − 0.0591 pH (Eq 21a)
2 / Fe
Fe + 2H 2 O = HFeO −2 + 3H + + 2e− ,
EHFeO- /Fe = 0.495 − 0.0886 pH
2
(
+ 0.0295 log aHFeO–
2
) (Eq 21b)
Fe 2+ + 3H 2 O = Fe ( OH 3 + 3H + + e− , )
EFe(OH) 2+ = 1.057 − 0.1773 pH
3 / Fe
( )
− 0.0591 log aFe 2+ (Eq 21c)
Fe(OH)2 + H 2 O = Fe(OH)3 + H + + e− , Fig. 6 Pourbaix diagram for chromium in water. Source: Ref 8
EFe(OH) = 0.271 − 0.0591 pH (Eq 21d)
3 /Fe(OH)2
For the pH-dependent reactions (chemical reduction are able to cause corrosion through
and electrochemical), one can readily label the the entire pH region. Unfortunately, Fe-OH cor-
regions depending on what iron species increas- rosion products are generally not passivating.
ing pH favors. If iron would be an anode and Iron or carbon steel alloys are therefore not par-
the tendency to corrode were to be evaluated, ticularly corrosion resistant in water solutions.
then the reduction potential for a possible cath- Figure 6 shows the Pourbaix diagram for
ode reaction would be placed on this diagram. If chromium (Ref 8). While chromium oxidizes
this point were to be, for example, in A in Fig. even more readily than iron, it forms Cr2O3 over
5(a), this means that the reduction potential for a significantly large region that is of relevance
this assumed cathode lies below any reduction to pH values in water solutions. Since Cr2O3 is
potential of iron, and hence under these condi- protective, it prevents further corrosion. When
tions iron is immune (since Erxn is negative). In chromium is added to iron as an alloying ele-
fact any Fe2+ ions present could plate as iron. ment, it corrodes selectively due to its low re-
On the other hand, if the reduction potential of duction potential, but this means that it also pro-
the assumed cathode reaction were to lie in tects the iron alloy due to the properties of
point B, then there is a tendency to dissolve iron Cr2O3. This is the basic design principle behind
to Fe2+ since Erxn is positive. Finally, if the re- iron-chromium-based stainless steels.
duction potential of the assumed cathode was at
point C, corrosion would occur, resulting in
Fe(OH)3, but when oxides or hydroxides are REFERENCES
formed there is a possibility that this product
could form a solid protective layer that kineti- 1. A.J. Bard and L.R. Faulkner, Electrochemi-
cally hinders further corrosion. These types of cal Methods: Fundamentals and Applica-
diagrams are called Pourbaix diagrams. Figure tions, 2nd ed., Wiley, 2001
5(b) shows the Pourbaix diagram for iron over- 2. D.A. Jones, Principles and Prevention of
laid with the common cathode reactions in Corrosion, 2nd ed., Prentice Hall, 1996
water, Eq 15 and 18 (Ref 8). The ionic activity 3. H.H. Uhlig and R.W. Revie, Corrosion and
was previously arbitrarily set at 10–6, but from Corrosion Control: An Introduction to Cor-
the Pourbaix diagram it can be seen that changes rosion Science and Engineering, 3rd ed.,
in ion activity do not have dramatic effects on Wiley, 1985
the boundaries. It can be seen that both the 4. M.G. Fontana, Corrosion Engineering, 3rd
oxygen gas reduction reaction and hydrogen ion ed., McGraw-Hill, 1986
5. ASM Handbook, Vol 13A, Corrosion: Fun- 8. S.A. Bradford, Corrosion Control, 2nd ed.,
damentals, Testing, and Protection, S.D. CASTI Publishing, Inc., 2001, p 41
Cramer and B.S. Covino Jr., Ed., ASM
International, 2003
6. Handbook of Chemistry and Physics, 71st SELECTED REFERENCE
ed., CRC Press, 1991
7. D.A. Jones, Principles and Prevention of • M. Pourbaix, Atlas of Electrochemical Equi-
Corrosion, 2nd ed., Prentice Hall, 1996, p. 59 libria in Aqueous Solutions, NACE, 1974
CHAPTER 3
Corrosion Kinetics
Summary the penetration due to dissolution of element i

becomes:
CORROSION INVOLVES chemical reac-
tions with equilibrium that is known through
t ⋅ Mi
thermodynamics. In practice, the rate at which r= j (Eq 1)
corrosion reactions occur is the most important ni ⋅ F ⋅ ρ
consideration. This chapter deals with corro-
sion kinetics, which allows us to understand the The penetration rates for iron and various
rates of corrosion. stainless steels are listed in Table 1 (Ref 1) in
units of mils (0.001 in.) per year, or mpy. In the
case of alloys, the ratio of Mi /ni is computed as
Introduction an equivalent weight (EW) according to:
Consider the differential aeration cell dis-

cussed in the Chapter 2 on corrosion theory, Fe/ 1
EW =
O2/OH–, Fe2+/Fe. If the thermodynamic condi- fn (Eq 2)
tions favor electrochemical corrosion of iron, ∑ Mi i
i
that is, Erxn = EO / OH – EFe / Fe is positive, then
– 2+
2
a net corrosion current i will flow, resulting in

iron dissolution and consumption of oxygen gas where fi, ni, and Mi are the weight fraction, va-
according to the net reaction, 2Fe (S) + O2 + lence, and molar mass of element i,
–
2H2O → 2Fe2+ + 4OH . The magnitude of this respectively.
current will determine the rate or iron dissolu- As stated, the amount of corroded (dis-
tion according to Faraday’s law, which was solved) iron is determined by the current i, and
introduced in Chapter 2: nNF = it. Because n = the magnitude of this current is determined by
2 and F = 95,485 C per mole electrons, the
moles of dissolved iron are given as a function
of time as N = i* t/(2* 95,485). Practically, this
Table 1 Penetration rates for a current of 1
can be readily converted to lost mass m, which
µA/cm2 (mpy)
in the case of iron loss becomes m = N* MFe =
MFe* i * t/(2* 95,485), or thickness reduction r, Density, Penetration
Alloy Element/oxidation state g/cm3 EW(a) rate, mpy
which in the case of iron becomes r = MFe* i*
t/(2* 95,485* A* ␳Fe). Here, MFe and ␳Fe are Fe
304
Fe/2
Fe/2,Cr/3,Ni/2
7.87
7.9
27.92
25.12
0.46
0.41
molar mass and density of iron, respectively. It 321 Fe/2,Cr/3,Ni/2 7.9 25.13 0.41
is often the thickness loss (referred to as 309 Fe/2,Cr/3,Ni/2 7.9 24.62 0.41
316 Fe/2,Cr/3,Ni/2,Mo/3 8.0 25.50 0.41
penetration per unit time) that is useful; 430 Fe/2,Cr/3 7.7 25.30 0.42
therefore, i/A is often replaced by j, which is 446 Fe/2,Cr/3 7.6 24.22 0.41
20Cb3 Fe/2,Cr/3,Mo/3,Cu/1 7.97 23.98 0.39
defined as current density and has the units
(a) Equivalent weight. Source: Ref 1
amperes/square meters. A general equation of
RT ⎛ ( aFe ) ⎞
4
EFe 2+ /Fe = EFe0 2+ /Fe − ln ⎜ ⎟

4 F ⎝ aFe 2+ ⎠
When a cell is not under open circuit (i.e., a

net current passes through it), the cathode and
anode potentials deviate from the half-cell po-
tentials, and the electrode states are then defined
as being polarized. The polarization is quantified
as overpotentials ␩, which are defined by the de-
viation from the equilibrium half-cell potentials,
that is, for the cathode, ηc = Ecathode – EO2 / OH and
for the anode, ηa = Eanode – EFe 2 + Fe. Effectively, the
Fig. 1 Schematic illustration of a differential aeration cell overpotential reduces the activation energy for
involving iron dissolution. Kinetic steps: (1) electrode the electrode Reactions 3 and 4. In the case of a
reactions, (2) ion conduction, (3) electron conduction
reduction reaction at a cathode, such as Reac-
tion 3, the overpotential is negative, and driving
the corrosion potential. The corrosion potential an electrode toward a lower potential drives
is determined by the reaction potential (which electrons from the electrode into the solution,
was discussed in Chapter 2) and the kinetics of resulting in a net cathodic current ic at this elec-
the various steps involved in completing the trode. Similarly, at the anode the overpotential
electrochemical circuit depicted in Fig. 1. is positive, which results in electrons that are fa-
These involve: (a) electrode reactions at the vored to be removed from the solution and
cathode and anode, (b) conduction of ions in transferred into the electrode, thus producing a
the electrolyte, and (c) conduction of electrons net anodic current ia.
from the anode to the cathode. The conduction If the magnitudes of cathode and anode polar-
of electrons is generally not a problem in stain- ization are large, as would be expected in a gal-
less steels because the corroding metal (iron) vanic cell, the relation between each electrode
and scale (Cr2O3) provide an easy path for elec- current/current density and overpotential is
trons. The other two kinetic processes are dis- given by the following equations (for a thor-
cussed briefly in this chapter. ough derivation of the current overpotential
equation, Ref 2 is recommended):
The Butler-Volmer Equation
ic CO (0, t ) ⎛ αnF ηc ⎞
jc = = j 0 ,c 2 exp ⎜ − (Eq 5)
For the case study 2Fe (S) + O2 + 2H2O Ac C O* ⎝ RT ⎟⎠
→ 2Fe2+ + 4OH–, the cathode and anode reac- 2
tions are:
and
O 2 + 2H 2 O + 4e − → 4 OH − (Eq 3)
ia ⎛ (1 − α )nF ηa ⎞
ja = = j0 ,a exp ⎜ ⎟⎠ (Eq 6)
Aa ⎝ RT
and
2 Fe ( s ) → 2 Fe 2 + + 4 e− (Eq 4) where the jo,i terms are the exchange current

densities and represent the equally large cath-
The Nernst equation predicts an open circuit ode and anode currents at equilibrium (zero
potential of Erxn = EO / OH – EFe / Fe overpotential) at the electrodes. The exchange
– 2+
2 current densities are a measure of the electrocat-
where alytic ability of the surface to promote/demote
⎛
( )
⎞
4
the electrode charge transfer reactions; as such,
RT ⎜ aOH – ⎟
EO = EO
0
− ln they can vary over many orders of magnitude
2 /OH – 2 /OH –
4 F ⎜ aH O PO ⎟ depending on the surface chemistry and struc-
⎝ 2 2⎠
ture and on electrode reaction. The α-terms are
and fractions that define the amount to which the
Chapter 3: Corrosion Kinetics / 21
activation energies are lowered. They do not electrode, and the rate of cathode reaction will
have to be the same for the anode and the cath- depend on how rapidly oxygen molecules dif-
ode, but due to the uncertainty in evaluating fuse to the electrode/electrolyte interface. As a
them, they are often taken as 0.5. The concen- limiting case, when the oxygen concentration is
tration terms represent the ratios between the re- actually zero at the interface, the corrosion cur-
actant concentration at the electrode/electrolyte rent can, through Faraday’s law, be coupled to
interface and bulk, which could deviate from the steady-state flux of diffusive oxygen supply
unity as a result of consumption/production of through a boundary layer δ. This limiting case
species at the interface. In an iron-based alloy, current is called the limiting current (iL or jL)
this ratio for the anode would be close to unity and can be expressed as:
because the reactant is iron itself, and no
concentration gradient would be expected as a DO nFCO*
result of the corrosion reactions. When a net jcorr = jl = 2 2 (Eq 11)
corrosion current flows, icorr = ia = ic. If the cath- δ
ode and anode areas are assumed to be equal,
then jcorr = ja = jc and Eq 5 and 6 can be rewrit- In a nonlimiting case, the corresponding
ten (using βc = 2.3RT/(αnF) and βa = 2.3RT/ equation would be:
((1 ⫺ α)nF) as:
DO nF (CO* − CO (0, t ))
j 0 ,c CO (0, t ) jcorr = 2 2 2
(Eq 12)
ηc = β c log + β c log 2 δ
(Eq 7)
jcorr C O*
2
Combining Eq 11 and 12, one obtains:
and
CO (0, t ) jcorr
j
2
= 1− (Eq 13)
ηa = β a log corr (Eq 8) CO * jl
2
j0 ,a
Thus, Eq 7 can be written:
Tafel Regime: Electrode-Kinetics Control.
If the electrode charge transfer reactions are
rate limiting, the supply of oxygen to the reac- j 0 ,c ⎛ j ⎞
ηc = β c log + β c log ⎜ 1 − corr ⎟ (Eq 14)
tion site would be rapid enough to maintain a jcorr ⎝ ji ⎠
concentration at the electrode close to that of
the bulk. In this case, Eq 7 and 8 would both re- The slower the diffusion (small d), the lower
sult in a linear dependence of the overpotentials the limiting current and thus a larger contribu-
versus log jcorr: tion from the mass-transfer-dependent second
term on the overpotential.
ηc = β c log j0 ,c − β c log jcorr (Eq 9)
Migration and Ionic Diffusion
and
The ionic transport in the electrolyte phase,
ηa = β a log jcorr − β a log j0 ,a (Eq 10) the flux of an ion i under an electric field φ
across a distance L, can be shown to be:
Mass Transfer Control. In Eq 7, the term:
∂ci zi F Δφ
CO (0, t ) J i = − Di − Dc (Eq 15)
2 ∂x RT i i L
CO*
2
In an electrolyte with many different ions, an
stands for the ratio of oxygen gas concentration ion current through an area A can be computed
at the electrode/electrolyte interface and the by multiplying Eq 15 with zi * A and summing
concentration in the bulk, sufficiently far away the contribution from all ions:
from the interface. If the electrode reaction ki-
netics are very fast, the depletion of oxygen will ∂Ci + ( x ) F2A
i = FA∑ zi Di + +
lead toward a zero oxygen concentration at the ∂x RT
and the respective equations describing the

× ∑ zi 2Ci Di Δφ / L (Eq 16) overpotentials will be:
j 0 ,c ⎛ j ⎞
Because the first term is important only at the ηc = β c log + β c log ⎜ 1 − corr ⎟ (Eq 21)
jcorr ⎝ ji ⎠
regions near the electrodes (where consump-
tion/creation of species occur), the current in
the majority region of the electrolyte can be es- and
timated as:
jcorr
2
ηa = β a log (Eq 22)
F A j0 ,a
i≅
RT
∑ zi 2Ci Di Δφ / L (Eq 17)
Now, the potentials of anode and cathode
Using Ohm’s law (R = U/i), the electrolyte re- when current is flowing are in each case the
sistance can be computed as: equilibrium potential plus overpotential, that is:
⎛ F2A ⎞ RT PH2
Relectrolyte = L / ⎜
⎝ RT
∑ zi 2Ci Di ⎠⎟ (Eq 18) Ecathode = EH0 + / H −
2
ln
nF aH2 +
+ ηc
The resistivities of some test solutions are j0 ,c

= 0 − 0.059pH + β c log
shown in Table 2. jcorr
⎛ j ⎞
Mixed Potential Theory and + β c log ⎜ 1 − corr ⎟ (Eq 23)
⎝ ji ⎠
Polarization Diagrams
Viewing the electrochemical cell as an elec-

trical circuit, Kirchoff’s law can be used to
design a so-called polarization diagram. Con-
Eanode = −0.447 − 0.0295 log aFe 2+ + ηa ( )
sider, as a case study, a steel corroding under jcorr
deaerated conditions, in a water solution, as = −0.624 + β a log (Eq 24)
j0 ,a
shown in Fig. 2. Assume that the pH is such that
a passive layer does not form (see the discus-
sion of Pourbaix diagrams in Chapter 2). The A polarization diagram is now constructed by
cathode and anode reactions, respectively, are: plotting the anode and cathode potentials versus
log jcorr. Strictly speaking, to close the circuit,
2 H + + 2 e− → H 2 (Eq 19)
Fe ( s ) → Fe 2 + + 2 e− (Eq 20)
Table 2 Test solution resistivity

Ratio by Resistivity,
Test solution volume ohm-cm
Natural Seawater ... 25
Fresh (tap) water adjusted 28:1 500
with seawater
Fresh (tap) water adjusted 68:l 1,000
with seawater
Fresh (tap) water adjusted 950:1 3,000
with seawater
Deionized water adjusted 21:10 10,000
with fresh (tap) water
Fig. 2 Schematic polarization diagram
the potential drop across the electrolyte needs to Passivation

be included, which simply equals icorr * Relectrolyte
(the electrolyte resistance is evaluated from Eq Theory. In Chapter 2, it was identified
18); however, in many cases, this term can be through the Pourbaix diagrams that there were
neglected. A schematic polarization diagram is conditions under which an alloy could be
shown in Fig. 3. The anode polarization is linear passive. In the case of stainless steels, the range
with decade current as predicted by Eq 24 be- of pH and other conditions under which this
cause the overpotential has only a Tafel regime would occur has been increased thanks to the
and no mass transfer dependence. On the other chromium content, which readily forms a Cr2O3
hand, the cathode polarization deviates from the scale. In general, a passive layer constituted of
Tafel behavior as a result of the effect of the adsorbed molecules or thin oxide/hydroxide
mass transfer (hydrogen ion supply), dependent layers decreases the corrosion current. Re-
on the limiting current in Eq 23. It is notewor- searchers (Ref 3) have reported that the con-
thy that the cell shown in Fig. 1 does not have a stituents of the passive film are alpha Cr2O3 and
macroscopic anode and cathode. Different mi- Cr(OH)3nH2O. The structure is reported to be a
croscopic regions on the surface are assumed to nanocrystalline spinel, epitaxial to the surface.
act as cathodes and anodes, and in the lack of The grain size may decrease with increasing
more detailed knowledge, the cathode and chromium content. This protection by
anode areas are assumed to be equal. The over- chromium requires a threshold level of 11 to
all mixed potential of the surface would be at a 12% chromium.
corrosion potential Ecorr, defined in Fig. 2. Effect on Polarization Diagrams. The polar-
In effect, the corrosion current resulting from ization diagram for a passive alloy is quite dif-
the cell depends on the equilibrium half-cell po- ferent from those discussed for active alloys. A
tentials (Ecathode and Eanode), the Tafel slopes (βc schematic of a typical polarization curve is
and βa), the exchange current densities (jo,c and shown in Fig. 5. When a passive alloy is anodi-
jo,a), and any limiting current density (jl). Figure cally polarized, it initially behaves like an active
3 shows schematically how decreasing any of the alloy (i.e., with a Tafel slope, etc., as the pas-
Tafel slopes and increasing an exchange current sive layer is building up). The building up is
density increases the corrosion rate. The effect of actually a selective dissolution of iron, which
the electrolyte resistance has been ignored; that causes a greater remaining surface concentra-
is, corrosion current is where the two polarization of chromium and other alloying elements.
tion curves intersect. Figure 4 shows the effect of Once the passive layer is formed and offers
increased mass transfer, which would result in an protection against further dissolution, the po-
increase in the limiting current. In an active (non- tential-decade current relation drops to lower
passive) alloy, this results in an increased corro- currents. This happens at potentials beyond the
sion current up to a point. passivation potential Epp. At some high enough
Fig. 3 Corrosion rate and the effect of (a) Tafel slope and (b) exchange current density
polarization level, the passive layer breaks j0 ,c

down, and the metal becomes active again; this = 0 − 0.059pH + β c log (Eq 25)
jcorr
region is called the transpassive regime. The
design of a structure involving a passive metal
should aim at forming a corrosion cell in which This will result in a straight line as shown in
the cathode polarization curve intersects the an- Fig. 6, which will be shifted vertically depend-
odic one in the passive regime. ing on the pH. The dashed circles indicate the
Consider, for example, an alloy that exhibits intersection between anode and cathode polar-
the behavior shown in Fig. 6 in deaerated ization curves that would yield the corrosion
acidic solutions with different pH. If mass current. At a sufficiently high pH (= pH1), the
transfer limitations due to hydrogen ion supply alloy is clearly not optimal because intersection
are neglected, then the cathode polarization is occurs in the active regime, and the passive
given by: properties are not utilized. This is what occurs
when a reducing acid is too strong for a given
stainless steel, such as with concentrated hy-
RT PH2 drochloric acid. At pH = pH2, on the other
Ecathode = EH0 + / H − ln + ηc hand, a low-corrosion current is obtained as a
2 nF aH2 +
result of intersection at the passive regime.
This is the benevolent case when stainless steel
is correctly matched to the environment, and
low rates of uniform corrosion occur. Finally, at
pH = pH3, the resulting corrosion current is
again high as a result of intersection occurring
at the transpassive region. This could occur
with some stainless steels exposed to a very
strong alkali solution. Using a similar argu-
ment, the readers can themselves deduce the
effects of cathode exchange current density and
Tafel slopes. In the discussion of active anode
polarization, it was found that increasing the
transport rate of cathode reactants through
agitation, for example, would increase the cor-
rosion rate up to a point but beyond that have
no further effect (see Fig. 4). In the case of a
passive/active behavior, the effect of mass
Fig. 4 Effect of increasing the limiting current by, for example,
increased agitation in the electrolyte. Beyond the
dashed line, increasing the limiting current would have no fur-
ther effect
Fig. 5 Schematic of a passive anode polarization curve Fig. 6 Effect of cathode polarization
is the actual passivation; the iron removal is

really a chemical cleaning operation, which hap-
pens to be called passivation. During the produc-
tion of stainless steel, after a final anneal another
version of passivation is carried out. The oxide
from annealing in air is dissolved by a strong
mixture of nitric and hydrofluoric acids, which
does not allow passivation. This treatment,
called pickling, removes by dissolution both the
oxide layer and the chromium-depleted layer
below the oxide formed during annealing. The
depleted layer can extend a number of microns
in depth and would seriously degrade corrosion
resistance if not removed (Ref 4). This is then
followed by a straight nitric acid immersion,
Fig. 7 Effect of mass transport which ensures complete passivity. This is the
procedure that should be performed on the ox-
transport is somewhat different, as shown ides formed during welding if full corrosion re-
schematically in Fig. 7. sistance is to be restored. Simply removing the
Increasing mass transport, such that the limit- oxide through mechanical means leaves a
ing current increases, results initially in an in- chromium-depleted layer that corrodes more
creased corrosion current (e.g., increasing jl readily than is expected of the alloy.
from 1 to 2). It should be noted that there are
several intercepts possible (both in the active
and passive regime), but assuming there are REFERENCES
defects present, it is likely that there will be cor-
rosion corresponding to the higher current. In- 1. D.A. Jones, Principles and Prevention of
creasing the limiting current beyond the knee Corrosion, 2nd ed., Prentice Hall, 1996
corresponding to Epp, however, results in a drop 2. A.J. Bard and L.R. Faulkner, Electrochemi-
in the current because now the only corrosion cal Methods: Fundamentals and Applica-
potential possible is at the intersection in the tions, 2nd ed., Wiley, 2001
passive regime. This is the case for jL3. 3. M.P. Ryan et al., Critical Factors in Focal-
In the normal use of stainless steel, achieving ized Corrosion, Proc. Electrochem. Soc.,
passivity takes on several forms. What is often Vol 150, 2003, p 583–594
called passivation is actually a cleaning process 4. J. Grubb and J. Maurer, “Corrosion of the
in which contaminants, such as tramp iron, are Microstructure of a 6% Molybdenum Stain-
removed from the surface. Dilute nitric acid is less Steel with Performance in a Highly Ag-
an excellent vehicle to achieve this. This gressive Test Medium,” paper 300 pre-
medium has the additional benefit of forming a sented at Corrosion 95, NACE
passive film on an active stainless surface. This International, 1995
CHAPTER 4
Corrosion Types
Summary plex in their behavior because the influence of

processing and alloying variables changes the
STAINLESS STEEL is unusual among alloy ability of this layer to form and remain stable in
systems in that its corrosion resistance derives the face of aggressive environments. The be-
from the passivating ability of a minor con- havior of stainless steel is further affected by its
stituent, chromium. Thus, while stainless steels microstructural complexity. Stainless steel al-
can be made to be essentially immune to corro- loys may have many constituent elements and
sion in many environments, it can also experi- many thermodynamically possible phases, and
ence various debilitating forms of localized cor- none of these are necessarily uniform in their
rosion, which stem from the failure of this composition. Yet, it is the composition of the
passive film. This chapter explores the behavior alloy in contact with the specific environment at
of stainless steel in media that promote uniform any microscopic point that determines the cor-
corrosion and the various mechanisms of local- rosion resistance of that particular point.
ized corrosion, such as pitting and crevice cor-
rosion.
Uniform Corrosion
Introduction When all parts of a corroding surface have
equal access to the corroding atmosphere and
To most designers, the most recognized char- the structure of the corroding metal is relatively
acteristic of stainless steel is corrosion resist- uniform, a uniform thinning of the material is
ance. Stainless, unlike noble metals such as expected. Stainless steels are materials of choice
gold, does not obtain its excellent corrosion re- because, by virtue of their passive behavior, they
sistance from inertness. Instead, it is the reactiv- show very low rates of uniform corrosion in
ity of chromium that allows the surface layer of many environments. The metallurgy and pro-
corrosion product to become sufficiently adher- cessing of a particular grade are designed to pro-
ent and impenetrable, which effectively stops vide passivity in a given environment. The envi-
further corrosion by isolating the base material ronment can be too aggressive to allow passivity
from the environment. This resistance to corro- to be maintained either by being too reducing, as
sion is called passive behavior or passivity. with some acid media, so that passivating
Other metals, such as aluminum and titanium, species cannot form or by being too oxidizing so
form similar layers and also exhibit passivity. that the oxidized species that normally affect
The important difference in the case of stainless passivity are no longer stable. The former is
steel is that chromium is still a minor con- called dissolution in the active state, while the
stituent, never more than 30% by weight, some- latter is termed transpassive dissolution.
times little more than 10%. How much Intelligent design and knowledge of the envi-
chromium there is and how uniformly it is dis- ronmental variables for a stainless steel compo-
tributed have a profound effect on corrosion re- nent ensure that the alloy is used in the passive
sistance by virtue of its ability to concentrate state, at which uniform corrosion occurs at a
into the surface film. Stainless steels are com- very low rate.
Among the important media with which we sive, or transpassive; Fig. 1 illustrates the effect
encounter uniform, but acceptably controlled, of redox potential on a solution.
corrosion in stainless steel are atmospheric and Certain anions have strong effects in media
marine environments and chemical environ- through their well-known, if not well under-
ments such as sulfuric acid, phosphoric acid, ni- stood, disruption of the passive film. Halides are
tric acid, strong bases, and organic acids, such well known for this effect, but sulfides are
as acetic and formic. active. These anions seem to intervene in the
Pickling is an example of controlled, acceler- adsorption of the hydroxyl ions. In acid media,
ated uniform corrosion. This is typically done these anions accelerate uniform corrosion,
with 10 to 20% hot sulfuric acid or a mixture of while in neutral media they may result in local-
hydrofluoric and nitric acids. ized corrosion. Anions that form soluble com-
plexes with elements in stainless, such as
Environmental Variables Influencing amines, formates, or acetates, can also disrupt
the stability of the passive film and thus pro-
Uniform Corrosion
mote active corrosion.
The corrosion of stainless steels is usually the Of the physical variables of the environment,
result of contact with an electrolyte, allowing a it should be obvious that temperature is para-
complex set of partial electrochemical reac- mount since all the reactions are thermally acti-
tions, which may occur sequentially or concur- vated. Increasing temperature may speed the
rently. The corrosion rate depends on the formation of the passive film when thermody-
current exchanged between the negative and namic conditions are favorable, but in general
positive electrode (anode and cathode). These one expects increasing temperature to increase
may be on a macroscopic or microscopic level. corrosive attack. Access to passivating species,
The main consideration is normally ionic trans- such as oxygen, is important in establishing and
por tthrough the passive film, which after all maintaining passivity.
is what makes stainless so effective against Increased diffusion of reacting species in the
corrosion. liquid will normally accelerate the partial reac-
The chemical parameters that influence the tions, but if passivity is stable, the rate-limiting
media with respect to uniform corrosion rate are transport through the passive film will not be af-
the acidity and the oxidation-reduction (redox) fected. Therefore, increasing the flow rate of a
potential of the electrolytic medium, both of corrosive fluid does not automatically acceler-
which act through their influence on the stabil- ate corrosion. The reduction of concentration
ity of the passive film, rendering it active, pas- gradients can be beneficial against localized
Fig. 1 Reduction potential versus pH for iron and (a) oxygen gas reduction and (b) hydrogen ionreduction
Chapter 4: Corrosion Types / 29
corrosion, and flow can bring to the surface an stituent of sigma phase, which it promotes,
increased supply of passivating species. however, it combines with chromium. If this
Increased flow rate in a fluid medium is dele- happens at relatively low temperatures, the sur-
terious if it induces mechanical damage to the rounding matrix is depleted of both chromium
passive film by erosion, abrasion, or cavitation. and molybdenum, and the corrosion resistance
These are complex mechanisms, but it should in that region is diminished. Nitrogen also is ef-
be apparent that the success of a stainless steel fective when in solid solution in austenite but
to a given flow condition will depend mainly on can precipitate as a chromium nitride under cer-
its ability to form and re-form its passive film. tain conditions and cause depletion of the re-
Somewhat counterintuitively, thinner passive maining matrix. Local structure and composi-
films are more protective than thicker films tion are paramount. This becomes more
among stainless alloys. The tenacity of the thin important to localized corrosion, as discussed
passive films on stainless (and titanium) make later, but it should be remembered in examining
these alloys quite resistant to flow-accelerated uniform corrosion because corrosion will cease
corrosion, as contrasted to copper and alu- to be uniform when composition becomes
minum alloys, which have soft, thick corrosion nonuniform.
product films. The compositional material variables that in-
fluence uniform corrosion are not exactly the
same as those that will be seen to influence lo-
Material Variables
calized corrosion. The foremost element is, of
Stainless steels have a great variety of alloy- course, chromium. Researchers (Ref 1) have re-
ing elements and microstructures. As a general- ported that the constituents of the passive film
ization, we can say that corrosion resistance is a are alpha Cr2O3 and Cr(OH)3nH2O. The struc-
function of composition rather than structure. ture is reported to be a nanocrystalline and epi-
Then, we must quickly add the qualifiers to this taxial to the surface. The grain size may de-
statement. On an undisrupted, stress-free sur- crease with increasing chromium content. This
face, local composition does quite precisely de- protection by chromium requires a threshold
termine corrosion resistance. But, stainless level of 11 to 12% chromium. This threshold
steels are seldom homogeneous or at thermody- has been attributed most convincingly to the
namic equilibrium. Impurities such as oxygen minimum chromium content that permits
and sulfur are usually present, mainly as inclu- chromium atoms on surface sites to be linked by
sions since they have diminishingly small solu- adsorbed oxygen atoms (Ref 2). In any event,
bility at room temperature. At high tempera- the mechanism by which this thin, several-
tures after solidification, as in welds, they can nanometer-thick, film forms is the subject of on-
be present in supersaturation, ready to precipi- going debate, but we do know that it is enriched
tate as inclusions that alter local composition. in chromium, and that it is thinner for higher
The tendency of carbon and nitrogen to form chromium alloys. The critical current density j,
precipitates is controlled by diffusion rates, as measured during polarization, is also smaller
which if elevated by increasing temperature can as chromium content increases (Fig. 2). This is
cause debilitating, composition-altering precipi- consistent with the lower dissolution of noncon-
tation. The even more slowly diffusing substitu- tributing elements required to achieve a critical
tional alloying elements, such as chromium, surface chromium concentration. Increases in
molybdenum, and nickel, have strong tenden- chromium can also be seen to lower the current
cies to form phases that disturb their uniformity density in the passive region. This is manifest in
in the austenite or ferrite matrix in which they alloy performance as a reduction in the uniform
are intended to work. So, any discussion of the corrosion rate in a given medium. From an elec-
influence of alloying element on corrosion re- trochemical point of view, this is explained as a
sistance of a phase like austenite or ferrite must manifestation of the stability of the Cr(OH)3
recognize that alloying elements exert their ef- nH2O.
fect when they are in solution in that phase. The The role of molybdenum is less clear. The ob-
same element may under some conditions not served action of molybdenum is to greatly re-
be in solution and have a contrary effect. An ex- duce the critical current density required for
ample is molybdenum, which is obviously a passivation. This is also seen as accelerating the
great enhancer of corrosion resistance when in formation of the passive films and as increasing
solid solution. When it precipitates as a con- the resistance of the alloy to depassivation at
Fig. 3 Influence of alloying element on corrosion rate as ex-

plained by the effect on polarization.Source: Ref 6
Fig. 2 Schematic illustration of polarization behavior for a

passive alloy with and without pitting occurring
lower pH. The role of molybdenum is not to en-

rich in the passive film itself, although it can be
found in the film. Its potency is far more than its
presence can take into account. Pure molybde-
num is itself not passive. Its action does not ap-
pear to be via a product of reaction. Instead, it
seems to reduce the dissolution rate of elements
other than iron, which would promote a surface
richer in chromium (Ref 3). The action of
molybdenum as an alloying element is compli-
cated by the fact that molybdate ions are known
to impede pit growth as a separate effect from
Fig. 4 Influence of alloying elements on uniform corrosion
their action within the alloy matrix (Ref 4). rate in 20% sodium chloride solution with carbon
Copper has a similarly complicated effect, with dioxide pressure of 20 MPa. Source: Ref 7
copper ions gettering sulfide ions and redeposit-
ing as metallic copper (Ref 5). chromium in the passive layer and to decrease
Nickel also lowers the critical current density active dissolution of noniron alloying elements,
for passivation without contributing directly to thereby promoting both the formation and sta-
the passive film’s stability. This also may be the bility of the passive film. A summary of the
result of the stronger bond between nickel and known major alloying effects in acidic chloride
chromium reducing the anodic dissolution rate media is shown in Fig. 3 in acidic chlorides. Al-
of the alloy by permitting the anodic enriching loying elements provide benefits in the part of
of the surface by selection iron dissolution. the chart where they appear (Ref 6). From this,
Nickel does not actively help passive film for- it can be seen that chromium, molybdenum,
mation and can actually hinder film stability in nickel, copper, and nitrogen all assist in the ac-
highly acidic/oxidizing environments. tive region, while chromium, molybdenum, and
Nitrogen, however, appears to be more like nitrogen expand the region of passivity and di-
molybdenum in its effect. While nickel and cop- minish the corrosion current.
per provide no benefit to the stability of the pas- An example of the influence of these alloying
sive film once it is formed, both nitrogen and elements on the uniform corrosion rate of stain-
molybdenum do, and to a degree that cannot be less steels in a sodium chloride/carbon dioxide
explained by their presence in the film. This environment is shown in Fig. 4 (Ref 7). Note
may then relate to their thermodynamic action the alloying composition is measured by a
within the alloy itself. Molybdenum and nitro- crevice corrosion index (CCI), which is dis-
gen act both to enhance the enrichment of cussed in the section Localized Corrosion.
Fig. 5 Corrosion table for stainless steels and titanium in sulfuric acid plus copper sulfate. Corrosion rate legend: 0, < 0.1 mm/yr
(corrosion resistant); 1, 0.1–1 mm/yr (useful in certain circumstances); 2, > 1.0 mm/yr (material not recommended). Source:
Ref 8; see source for interpretation of data. Courtesy of Outukumpu Stainless
Unfortunately, it cannot be assumed that this less steels in a great number of environments
relationship is true for other environments, al- can be obtained. Figure 5 shows an example of
though other empirical relationships exist or can one such table. Many of the isocorrosion charts
be generated. Because the influence of alloying in this book are reprinted from this source,
element varies with environment, we need to http://www.outokumpu.com/applications/docu-
discuss some of the more commonly encoun- ments/start.asp (Ref 8).
tered severe environments. These tables are supplemented by isocorro-
sion diagrams such as that shown in Fig. 6.
Corrosion in Acids and Bases These diagrams show constant corrosion behav-
ior under varying environmental conditions such
The examples discussed in Chapter 3, “Cor- as temperature and solution composition. This
rosion Kinetics,” refer mostly to corrosion in information is available to guide the designer in
these aqueous solutions, in which the slow thin- selecting appropriate steels for various environ-
ning rate of the chosen alloy can be determined ments, and it is highly recommended that it be
through the mixed potential theory and polar- used. Free sites tend to promote proprietary al-
ization diagrams. In the case of stainless steels, loys, as these charts suggest. The serious engi-
the alloy chemistry is chosen such that the neer will consult multiple sources and unbiased
passive-active behavior favors corrosion in the sources before making alloy decisions.
passive regime. The corrosion rate of the vari- The influence of alloying element is by no
ous stainless steels in the myriad possible envi- means the same in all environments. So, while it
ronments has been measured in probably all is useful and necessary to have these experi-
practical cases. These data can be obtained from mental data, it is also helpful to understand the
a number of sources, such as the National Asso- peculiarities of some of the major alloy-envi-
ciation of Corrosion Engineers (NACE) and ronment pairings.
ASM Handbook volumes. None is more accessi- Sulfuric Acid. Stainless steels require more
ble than the Web site of Outukumpu, which than a minimum amount of alloying to resist sul-
contains a “Steel Professional Tool,” a lookup furic acid. Straight 16% chromium grades such
table in which the corrosion rate of many stain- as 430 fare poorly, while the nickel-containing
Fig. 6 Isocorrosion curves for 17-12-2.5 stainless steel and titanium in sulfuric acid plus copper sulfate.Source: Ref 8. Courtesy of
Outukumpu Stainless
304 has more than an order of magnitude better

corrosion rate in either dilute or concentrated
sulfuric acid at ambient temperatures. Figure 7
(Ref 9) shows the isocorrosion rate curves for
several common alloys. Alloying with molyb-
denum is also very effective, as is alloying with
copper. If passivity cannot be established, in-
creasing chromium content actually increases
corrosion rate.
The corrosion behavior of sulfuric acid varies
greatly with concentration. At low concentra-
tions, sulfuric is a classic reducing acid. It dis-
sociates in water to create hydrated hydrogen
ions (H3O+) that release hydrogen gas bubbles
during the corrosion reaction. As the acid con-
centration increases, the solutions become more
corrosive, and progressively more highly al-
loyed stainless steels are required to provide ad-
equate corrosion resistance. At about 50% acid,
only the most highly alloyed stainless alloys
(alloy 20, AL-6XN, C-276, etc.) can provide ac-
ceptable corrosion rates, and even these alloys
are restricted to use at near ambient tempera-
tures. As acid concentration increases beyond
50%, the solution begins to show oxidizing be-
havior. At acid concentrations above 80%,
nickel-molybdenum-copper-bearing stainless
steels begin to exhibit useful corrosion resist-
ance. In the 93 to 98% sulfuric acid concentra-
tion range, carbon steel can be used to hold sul- Fig. 7 Isocorrosion rates of various stainless steels in sulfuric
furic acid at ambient temperatures, although acid. Source: Ref 9
stainless steels provide better performance at el-
evated temperatures or if flow-erosion can occur. resistant high-chromium (type 310S) and high-
In the 96 to 100% sulfuric acid concentration silicon (MECS ZeCor UNS S38815 and Sand-
range, at elevated temperatures, the oxidizing vik SX S32615) stainless steels are frequently
character is quite pronounced, and oxidation- used, especially in sulfuric acid-manufacturing
Fig. 8 Influence of alloying element on corrosion rate in contaminated sulfuric acid. Source: Ref 11
non-molybdenum-bearing alloys. Their superi-

ority in deaerated solutions is much less
marked. Studies (Ref 10) have shown that in
sulfuric acid molybdenum is highly enriched in
the passive film, and when molybdenum is an
alloy, chromium also is enriched. This is a man-
ifestation of selective dissolution of other ele-
ments in the matrix.
Oxidizing impurities, such as ferrous ions, act
like aeration to diminish the corrosive attack,
but reducing impurities such as halides have an
extremely negative effect, as the corrosion ta-
bles will show. These effects are not linear and
underscore the value of these tables.
The uniform corrosion rate in contaminated
Fig. 9 Isocorrosion curves for various alloys in sulfuric acid sulfuric acid may be more important than in
pure acid since this represents a potentially
equipment. Sulfuric acid-containing dissolved likely failure mode because contamination is a
sulfur trioxide is called oleum, and such solu- constant hazard.
tions are often identified as sulfuric acid of Figure 8 shows the corrosion rate of various
greater than 100% concentration. High- alloys in sulfuric acid contaminated with chlo-
chromium stainless steels (i.e., type 310S) are rides and iron. These researchers (Ref 11) found
among the very few materials that exhibit corro- that the resistance to attack correlated to the
sion resistance in oleum. (See MTI Materials alloy content by the formula shown.
Selector Volume 3—Sulfuric Acid at www.mti- Figures 9 and 10 show how isocorrosion rates
global.org for more information.) vary with alloy and contamination level.
Aeration has a major influence on corrosion Hydrochloric acid is very destructive of the
rates because oxygen stabilizes the passive film. passive film on stainless. An alloy like 304 is
Molybdenum-alloyed stainless has dramatically not suitable even in a deaerated 1% HCl solu-
lower corrosion rates in aerated solutions than tion at room temperature. Chromium additions
Fig. 12 Isocorrosion curves for austenitic AL-6XN (UNS

N08367) and 904L (UNS N08394) stainless steels
Fig. 10 Isocorrosion curves for various alloys in sulfuric acid in hydrochloric acid. Source: Ref 12
with chlorides
Nitric acid is strongly oxidizing. This actu-

ally promotes the passive film formation; con-
sequently, even low-chromium alloys remain
passive at all concentrations at ambient temper-
ature (see Fig. 14) (Ref 8).
The addition of molybdenum, which is so
generally helpful, is deleterious in this case be-
cause it forms soluble compounds. It is useful
to keep carbon, silicon, and phosphorus as low
as possible.
Silicon is unusual in that normal levels (0.4 to
1.0 %) are worst, with very low (0.05%) or very
high levels (4.0%) beneficial. The low levels of
silicon contents of these alloys are useful for
their action in minimizing grain boundary seg-
Fig. 11 Isocorrosion curves for various stainless steels in
regation, which is the usual locus of attack.
hydrochloric acid. Source: Ref 8. Courtesy of Out-
okumpu Stainless High silicon levels contribute to a general pro-
tective silica surface layer in concentrated acid,
which augments the true passive layer. This su-
periority appears above the azeotropic composi-
are only modestly helpful, while nickel, copper, tion of about 67%, which is a common com-
and molybdenum are more beneficial. Stainless mercial concentration, as shown in Fig. 15 for
steels are not good materials for contact with high-silicon austenitic stainless steels (Ref 13).
hydrochloric acid. Figures 11 (Ref 8) and 12 Phosphoric Acid. This oxidizing acid be-
(Ref 12) show how even the most highly al- haves more like sulfuric acid in that simple
loyed grades can withstand only dilute concen- iron-chromium alloys have only moderate re-
trations and low temperatures. sistance to uniform corrosion in them, while al-
While a stainless steel vessel may not be in- loying with molybdenum and copper produces
tended to be used for hydrochloric acid, resist- major improvements. This can be seen in Fig. 16,
ance to lesser amounts of chlorides is important in which alloys with increasing nickel (18-10)
because of the possibility that an acidic environ- show clear benefits over a chromium-molybde-
ment may be contaminated with chlorides. When num alloy (18-2), and the added molybdenum
this is a possibility, then proper alloy selection in 317 (17-14-4) is better, while 904L with
must guard against it. Figure 13 shows a correla- nickel, molybdenum, and copper is even better
tion between alloy content and resistance to the (Ref 8).
general corrosion (GI) by sulfuric acid contami- In the commercial production of phosphoric
nated with hydrochloric acid (Ref 11). acid, halide impurities may be present, in which
Fig. 13 Influence of alloy content on corrosion rate in hydrochloric acid
case alloys with higher molybdenum, chromium,

copper, and nitrogen may be required.
Organic Acids. The weakly dissociating or-
ganic acids are normally not aggressive against
stainless steels. The exceptional dangerous en-
vironments are those that include high tempera-
ture and the presence of chloride contamination.
It should be noted that in formic acid, which
does dissociate more strongly, nickel is detri-
mental. This phenomenon is also seen in the
production of urea via the intermediary ammo-
nium carbamate. The difficulty lies in the high-
temperature solubility of nickel complexes and
is best addressed by the use of ferritic or duplex
alloys.
Fig. 14 Isocorrosion curve for nitric acid. Courtesy of
Outokumpu Stainless Alloying with molybdenum seems to provide
the greatest resistance to uniform corrosion in
strong organic acids, as illustrated in Fig. 17
(Ref 8). If halides are present in organic acids
and liberated by contact with water, then pH
and chloride concentration will govern the cor-
rosive attack, which could then become nonuni-
form.
Strong Bases. In strong bases, the stainless
steels are generally quite resistant to uniform
corrosion. Straight chromium (17%) alloys are
usable at any concentration up to 50 °C. Adding
molybdenum and nickel does little to further
improve performance as the underlying resist-
ance is due to chromium. Increasing chromium
levels provide increased resistance. Attack
when it does occur can be manifested as grain
Fig. 15 Corrosion behavior of high-silicon alloys in concen- boundary attack. Figure 18 shows isocorrosion
trated nitric acid. Courtesy of Outokumpu Stainless curves for sodium hydroxide (Ref 8).
Fig. 16 Isocorrosion curves in phosphoric acid: (a) 0.1 mm/yr for various stainless steels; (b) 0.1 mm/yr for titanium and 17-12-2.5
stainless steel. Courtesy of Outokumpu Stainless
Fig. 17 Isocorrosion curves in organic acids: (a) acetic acid; (b) formic acid. Source: Ref 8. Courtesy of Outokumpu Stainless
In the pulp-and-paper industry, chemical

pulping is called the kraft or sulfate process. In
the presence of sulfur, nickel can be quite detri-
mental, and ferritic or duplex alloys are
preferred. This again is caused by the solubility
of nickel complexes formed in the presence of
sulfur-containing compounds. This can be seen
in Fig. 19, which shows a 26-1 (chromium-
molybdenum) alloy significantly outperforming
higher alloys that contain nickel and molybde-
num (Ref 12).
Atmospheric Corrosion
Atmospheric corrosion is an example of uni-
Fig. 18 Isocorrosion curves for various materials in sodium
form corrosion that occurs when a thin layer of
hydroxide. SCC, stress corrosion cracking. Courtesy
of Outokumpu Stainless water condenses on a metal surface and as such
2) that atmospheric corrosion can in effect be

prevented.
The most deleterious impurity in the atmos-
phere for stainless is the chloride ion. Chlorides
are pervasive. Borne from oceans by normal cli-
matological processes, they are found far inland.
In many colder climates, they are also seen in
high concentrations from road salts. Without
washing or the natural rinsing by rain, surface
chloride concentrations can become very high.
Thus, the rules of corrosion of aerated aqueous
solutions are followed by stainless with respect
to atmospheric corrosion. The difficulty is accu-
rately estimating the solution that constitutes the
aqueous solution. Much experience has shown
that if coastal and road salt effects are minimal,
then 18% chromium alloys such as 304 experi-
Fig. 19 Corrosion rates of various alloys in simulated evapo-
rator liquid. Source: Ref 12 ence such negligible visible corrosion that they
can be used for exposed, unrinsed architectural
depends on humidity, temperature, and other at- purposes. If the same alloy is used in an unrinsed
mospheric conditions. The rate of corrosion coastal environment, red rust stain will occur.
measured as defined in Chapter 3, “Corrosion This is the corrosion product from metastable
Kinetics,” as dissolution r of element i as and possibly stable pitting. In Japan, where
coastal conditions prevail throughout, much re-
t ⋅ Mi search has been done that has shown that a pit-
r= j ting resistance equivalent number (PREN) of 25
ni ⋅ F ⋅ρ
is necessary for freedom from corrosion (i.e.,
zero pitting) (Ref 14). This contrasts to a re-
where j is current density, t is time in seconds, quirement of about PREN 35 to resist pitting in
M is molar mass, n is valence, F is the faraday seawater. Pitting is a form of localized corro-
constant, and ρ is the density), and therefore has sion, and PREN is an index to pitting resistance.
two contradicting effects of temperature. In These concepts are examined in the next section.
general, temperature increases the exchange
current density and transport properties and thus
Localized Corrosion
the kinetic rates involved in corrosion. On the
other hand, increasing temperatures may re- Localized corrosion is in general more
duce the concentration of dissolved oxygen in damaging from a structural integrity point of
the electrolyte and eventually will dry the sur- view than uniform corrosion since the corro-
face and thus limit the electrochemical corrosion current is limited to a small area and the
sion due to the access to an electrolyte. In steels, penetration distance is large. Often, localized
the corrosion products are (a) an outermost corrosion involves a large-area cathode and a
layer of porous rust (FeOOH) characterized by small-area anode, which means that for a given
low water content but easy access to oxygen corrosion current, the corrosion current density
and (b) an inner layer of magnetite (Fe2O3) in at the anode is very large. In localized corro-
which pores are filled with water. The access of sion, unlike uniform corrosion, the anode and
oxygen to the bare metal limits the cathode re- cathode are clearly identifiable locations, and
duction reaction rate, and in relatively pure at- the reason that certain structural features as-
mospheres, the corrosion rate is negligible due sume the roles of cathode and anode can be
the protective nature of the oxide. However, sul- used to categorize and exemplify different
fur dioxide impurities in the atmosphere react cases. Interestingly, the cathode and anodes,
with water to form sulfuric acid, which tends to while identifiable, can vary across scales, that
dissolve the protective oxide. In the case of is, from distinct macroscopic components or
stainless steels, the passive region is extended parts to microstructural features.
due to chromium, to a wide enough region in In Chapter 2, the tendency for corrosion
terms of pH (see Pourbaix diagrams in Chapter was introduced as a positive value for an
electrochemical cell potential (Erxn) correspon- arises in which access of oxygen is not the same
ding to a spontaneous electrochemical reaction to different areas of a sample. In effect, this re-
forming a galvanic cell. Erxn is obtained as: sults in that the cathode reaction:
RT ⎛ p1 1 p2 2 p3 3 ... ⎞
α α α
O 2 + 2H 2 O + 4e − → 4 OH − (Eq 3)
Erxn = Erxn
0
− ln ⎜ β β β ⎟ (Eq 1)
nF ⎝ r1 1 r2 2 r3 3 ... ⎠ is limited from proceeding in some areas but not
others. This gives rise to a differential aeration
Here, pi and ri are the concentrations of reac- cell. For example, consider Fig. 20(a), in which
tant and products, respectively, and αi and βi are a metal is partially immersed in water. Trans-
the numbers that are needed to balance the reac- port distance of oxygen increases with depth;
tion stoichiometrically. Any time that Erxn is the limiting current would then vary with depth,
positive, there is thermodynamically a tendency such as at locations 1 and 2, and cathode polar-
for an electrochemical reaction, in our case a ization curves as a result of this are schemati-
corrosion reaction. The rate of corrosion, as dis- cally plotted in Fig. 20(b). Near the surface of
cussed in Chapter 3, is dependent on the polar- the water, where oxygen is readily replenished,
ization behavior. passivation is likely to be fast, and thus anodic
iron dissolution is slow. This region assumes the
role of the cathode, and reaction 1 occurs. Suffi-
Dissimilar Metals and Differential ciently far away from the surface, if there are
Aeration Cells regions where passivation is incomplete (e.g.,
surface defects or scratches) or has broken
The case of dissimilar metals and differential down as a result of, for example, Cl− (see sec-
aeration cells is perhaps more important in action on pitting), repassivation does not readily
tive alloys than for stainless steels, which are occur since oxygen transport is too slow. These
generally passive, and occurs when two regions become anodes where the following re-
metals/alloys are in contact that have elements action occurs:
in them that are dissimilar in the electromotive
force (emf) series (see Chapter 2) and there is an 2Fe (s) → 2Fe 2+ + 4e − (Eq 4)
electrolyte present. For example, if nickel and
iron pipes are connected and water flows though
them containing some traces of Ni2+ ions, then: The distance at which this occurs is balanced
by being large enough to limit the rate of
RT ⎛ aFe 2 + ⎞
oxygen transport but not too long to be
Erxn = ENi
0
2+
/ Ni
− EFe0 2 +/ Fe − ln ⎜ ⎟ (Eq 2) strongly influenced by ion transport that is
2 F ⎝ aNi 2 + ⎠
needed to complete the electrochemical cell.
Resulting corrosion currents are shown in Fig.
In this case, the corrosion tendency is prima- 20b. This type of degradation is called water-
rily caused by the first two terms on the right line corrosion.
side of Eq 2, the dissimilarity in the standard Crevice Corrosion. In stainless, the more
half-cell reduction potentials: significant occurrence of this type of cell occurs
when a crevice, from whatever cause, exists,
0
ENi 2+
/ Ni
− EFe0 2 + / Fe = − 0.250 + 0.447 = 0.197 V and reactions within the crevice or pit cause the
accumulation of iron ions by:
This tendency is caused by the galvanic dis-
similarity between the metals. This is normally 2Fe (s) → 2Fe 2+ + 4e − (Eq 5)
important for alloys joined to stainless that are
themselves less noble. Less-noble alloys, such The regions adjacent to the drop that main-
as carbon steel, can fail rapidly if coupled to tained their passive layer and have access to
stainless. A classic example is the use of carbon oxygen act as cathodes where the oxygen reduc-
steel fasteners for joining stainless sheets. Dif- tion reaction takes place:
ferent stainless steel alloys have minor differ-
ences when passive, but if the environment is O 2 + 2H 2 O + 4e − → 4 OH − (Eq 6)
such that one alloy is active while another is
passive, then the galvanic differential could be This reaction maintains an alkali solution. As
harmfully large. In many cases, a situation a result of the geometry, Fe2+ ions remain and
Fig. 20 Schematic illustration of (a) sample partially immersed in water; (b) resulting polarization behavior for two different passi-
vating alloys (A and B polarization curves)
enrich in the water-filled pit; to maintain charge Pitting corrosion is important to designers be-
neutrality, Cl– migrates into the pit. This causes cause it is corrosion under conditions at which
the following reaction: corrosion may not have been anticipated. Thus,
it is both a materials selection and an environ-
)
Fe 2+ + 2H 2 O + 2Cl − → Fe ( OH 2 + 2HCl (Eq 7) mental control problem. Its consequences may
be only cosmetic, such as on a building or appli-
which has several consequences: (a) Hy- ance facade, or potentially catastrophic, such as
drochloric acid further acidifies the pit and if leaks of toxic materials were to result from
increases the rate of iron dissolution since de- perforation. Stainless steels are designed to be
creasing pH increases cathode half-cell poten- passive, and localized corrosion is the local loss
tial, which increases corrosion rate (see polar- of passivity. Whether the consequences are
ization diagram construction in Chapter 3, major or not, it is always undesirable, and good
“Corrosion Kinetics.”) (b) The formation of design allows it to be avoided.
porous Fe(OH)2 further helps to isolate the pit, What do we know for certain about pitting?
thereby separating anode and cathode regions in We know quite a lot, really. Experts now con-
the differential aeration cell. (c) The presence of clude that since the early 1970s the local chem-
Cl– prevents repassivation. istry of pitting has been understood (Ref15).
As a result of increased acidification, the dis- The greatest contributions to this field have
solution rate becomes autocatalytic, and as a re- been electrochemical studies. The tools of elec-
sult the pit grows in depth. trochemistry have been especially successful in
At the outside, the reaction: elucidating the mechanism involved in pit
growth and pit stability (Ref 16). The local en-
)
2Fe ( OH 2 + O 2 + H 2 O → 2Fe ( OH )3
(Eq 8) vironment within pits has been sufficiently
measured and correlated with cavity geometry
further consolidates the isolation of the pit and that some experts can say, “In a sense, all pitting
impedes the ingress of oxygen. is crevice corrosion” (Ref 15). This is to say that
the electrochemistry of cavities such as pits and
crevices is quite similar and has been well mod-
Pitting Corrosion eled. These same tools, however, have been
Pitting corrosion is the most intensely studied much less successful in clarifying the mecha-
and debated form of corrosion of stainless steel. nism of pit initiation, which is still the subject
of debate, possibly indicating that the root Inclusions. The question of what causes the
causes are more metallurgical than electro- initial dissolution that causes both stable and
chemical. metastable pits focuses on inclusions, which
Figure 21 depicts a polarization curve for most authorities (Ref 18) have concluded are
stainless steel in a chloride-containing solution. associated in some way with pit initiation. In
Pitting occurs in the zone in which passivity is the absence of inclusions, metastable pitting
expected. As potential increases, small spikes in events are not noted, and the potential at which
corrosion current occur. These spikes measure pitting occurs is the beginning of the transpas-
local dissolution, called metastable pitting. sive regime.
Some such sites complete their dissolution and What are the typical inclusions in stainless
repassivate, while others continue to grow as steel? Inclusions in steel are normally the
stable pits. The potential at which stable pitting residue of normal deoxidation and desulfuriza-
occurs is the pitting potential, while metastable tion taken during steel refining usually done in
pitting can occur at much lower potentials. Pit- an argon oxygen decarburization (AOD). After
ting events, stable or not, cause the generation removal of the carbon, the subsequent objective
of iron ions and local pH reduction. To the ex- is to remove or render less harmful the dis-
tent these remain concentrated in a small vol- solved oxygen and sulfur, which if left in solu-
ume, they will affect subsequent events. The tion would later precipitate as low-melting-
dissolution during metastable pitting is located point iron compounds that would make the steel
at the matrix-inclusion interface. Different re- fragile and unworkable at high temperatures.
searchers assume dissolution of the inclusion, Inclusions in stainless steel are typically oxides
while others assume dissolution of the matrix. and sulfides. A key point to understand when
The dissolution parameters, as measured by considering inclusions as initiation sites for pit-
current transients, depend on variables not of ting is that inclusions are not simply inert debris
the inclusion chemistry but of the matrix com- but precipitates that are seeking thermodynamic
position, notably molybdenum and nitrogen equilibrium with the steel in which they have
levels (Ref 17), which is in keeping with the re- previously been dissolved. The reactions in
duction in dissolution of the matrix that these stainless steel differ thermodynamically from
alloying elements confer. those in carbon steel because of the presence of
high chromium concentrations. This lowers the
activity of oxygen and sulfur, making them
more soluble, as Table 1 indicates (Ref 19). It
also alters the efficiency of deoxidizing ele-
ments. Aluminum is a powerful deoxidant in
carbon steel but is less effective in stainless,
while titanium becomes a stronger deoxidizer in
stainless. Their effect on sulfur is similar to that
on oxygen.
The bottom line is that oxygen and sulfur are
generally removed by silicon/manganese deoxi-
dation, but that this process occurs in both the
liquid and solid states. That it carries over sig-
nificantly into the solid state means that diffu-
Fig. 21 Schematic of a passive anode polarization curve sion has a major role in determining if equilib-
Table 1 Typical values of activities and activity coefficients in liquid steels: activities in the 1 mass %
solution: ai = fi . %i
Metal Al C Mn P S Si Ti H N O Cr Ni
Carbon steel, %i ... 0.05 0.45 0.02 0.01 0.3 0.05 ... ... ... ... ...
1600 °C fi 1.05 1.06 1.0 1.1 1.0 1.15 0.93 1.0 0.97 0.85 ... ...
ai ... 0.053 0.45 0.022 0.01 0.345 0.046 ... ... ... ... ...
Stainless steel, %i ... 0.05 0.45 0.02 0.01 0.3 0.05 ... ... ... 18 8
1600 °C fi 3.6 0.49 1.0 0.32 0.66 1.24 9.4 0.93 0.17 0.21 0.97 1.0
ai ... 0.025 0.45 0.006 0.007 0.372 0.47 ... ... ... 17.5 8.0
rium reactions occur and whether they go to the liquid. At high sulfur and manganese concen-
completion. We will see that they do not. trations, some manganese sulfides can precipitate
Oxide inclusions also are common. They are during solidification interdendritically, while nor-
formed as the products of the reactions of sili- mal alloys with less than 100 ppm of sulfur form
con and manganese with dissolved oxygen. The their inclusions after solidification. The distinc-
thermodynamics of the reactions determine at tion is important because precipitation in the liq-
any time how much oxygen can be dissolved in uid state permits rapid diffusion, which results in
the steel at equilibrium. That equilibrium is eas- the most thermodynamically favorable species,
ily achieved in the molten state, in which diffu- manganese sulfide, to form. It may, and often
sion is very rapid, but achieved more slowly does, nucleate on a preexisting inclusion, such as
once the material has solidified. The inclusions silicate present from the deoxidation process. In
in the solid state grow by the diffusion of oxy- austenitic steels, manganese is generally present
gen to inclusion sites, where it precipitates as an at a level of around 1.5% as a deoxidant and as a
oxide of silicon or manganese to the extent that substitute for some nickel. An inclusion formed
these are locally present or of chromium when in the molten metal does not cause alloy deple-
its local concentration (or more properly, its action around it. One that forms or grows in the
tivity) makes it more favorable. These oxides solid state does cause depletion of the elements
are often the nucleation sites for manganese sul- that are precipitating, causing its growth.
fide inclusions. If manganese is lowered to very low levels,
Sulfur is a very surface active impurity that as- the supersaturation of sulfides is pushed to a
sists in weld penetration in stainless by virtue of lower temperature, at which lower diffusion
its effect on weld pool circulation. Otherwise, it rates hinder or prevent the precipitation. Thus,
is a detrimental impurity, forming low-melting low-manganese alloys can be free of manganese
oxysulfides that diminish hot workability. Man- sulfide inclusions even at somewhat high sulfur
ganese is a strong sulfide former, and it is the levels. Such alloys have elevated resistance to
main agent used to tie up sulfur. Manganese sul- pit initiation. Lower manganese levels also ther-
fide precipitates as an inclusion as a function of modynamically reduce the chromium sulfide co-
manganese and sulfur concentrations and tem- precipitation in inclusions, lowering chromium
perature. Inclusions form not only in the molten depletion around manganese sulfide/chromium
metal but also in the solidified metal. The solubil- sulfide inclusions.
ity, which is high in the liquid state, decreases on Elements more effective than silicon and
solidification, as seen in Fig. 22. Only resulfur- manganese are now in use for deoxidation and
ized free-machining stainless steels have suffi- desulfurization. These include aluminum, cal-
cient sulfur to precipitate manganese sulfide in cium, cerium, and other rare earth metals
(REMs), and titanium. The action of calcium is
notable. In a well-deoxidized and well-stirred
melt and with a basic slag, calcium dissolved in
the metal will react with dissolved sulfur to
form calcium sulfide, which will be incorpo-
rated into the slag phase. Aluminum, while a
potent deoxidizer, is less effective directly in
desulfurization, but it can act indirectly by re-
ducing a small amount of Ca2+ in the slag, al-
lowing the formation of calcium sulfide. Tita-
nium can sequester some sulfur as titanium
carbosulfide precipitates. The greatest amount
of sulfur removal is obtained by the addition of
cerium or other REMs, usually in the form of
the alloy mischmetal. These reactive elements
typically form oxysulfide particles in the melt
that may be trapped in the slag before metal
solidification.
Oxygen is normally dissolved in solidifying
Fig. 22 Pseudo-binary-phase diagram for iron and sulfur at stainless steel, also at amounts in the neighbor-
1.8% manganese and 18% chromium hood of 100 ppm depending on deoxidation
methods. Inclusions based on oxygen and sulfur solidification as opposed to austenite first, AF),
formed in the liquid or during solidification are as is almost always the case with commercial
relatively large, greater than 1μ. As the alloy alloys, more sulfide precipitation happens in the
cools after solidification, precipitation continues solid state, pitting resistance is lowered propor-
since sulfur and oxygen are decreasingly solu- tionately to the sulfur level (Ref 22), and there
ble with temperature, to virtually nil at room is little negative effect from solidification segre-
temperature. This causes existing inclusions to gation, as is shown in Fig. 23 and 24 (Ref 23).
grow and new ones to nucleate. This precipita- Solidification can also occur in a mixed ferritic-
tion is similar to that which carbon undergoes in austenitic mode, in which case each microstruc-
stainless, except carbon is generally not super- tural component behaves according to the chart.
saturated until below 1200 °C at the highest in The ratio of chromium and chromium-like ele-
most alloys, whereas sulfur and oxygen are nor- ments molybdenum and silicon to nickel and
mally near saturation even at freezing or almost nickel-like elements carbon, nitrogen, man-
always when the solidifying ferrite transforms ganese determines the mode of solidification. It
to austenite. Thus, inclusions grow via diffusion
of oxygen and sulfur, which, as interstitials, dif-
fuse much more rapidly than the silicon or man-
ganese with which they have the greatest ther-
modynamic affinity. But precipitate they must,
even if the silicon and manganese in the vicinity
of their inclusion are exhausted. Thus, inclu-
sions can grow with chromium substituting for
either silicon or manganese as the precipitating
partner for oxygen and sulfur. The inclusion
growth necessarily depletes the surrounding re-
gion of reactants, silicon, manganese, and
chromium (Ref 20). Inclusions thus formed are
nonequilibrium in nature, and thermal cycles of
steel production are rarely sufficient for the
equilibrium to be attained. The chromium en-
richment of such inclusions and corresponding
chromium depletion of surrounding regions has
been measured (Ref 21) and corresponds to the Fig. 23 Influence of sulfur level on pitting resistance of
unannealed welds for different solidification modes.
depletion seen next to chromium carbide pre- Source: Ref 23
cipitates at grain boundaries in sensitized al-
loys. These zones are altered in size and shape
by thermomechanical processing in wrought al-
loys but exist fairly undistorted in welds. Hot
rolling and cold rolling followed by annealing
elongate manganese sulfide inclusions and flat-
ten them, allowing depleted zones around the
inclusion in the reduced dimension to be more
rapidly homogenized during annealing. Thus,
wrought material has better pitting resistance
than cast or welded material. Inclusions that
precipitate from the liquid, as is more the case
for alloys solidifying in an austenitic mode, are
at equilibrium with the surrounding matrix by
virtue of the faster diffusion in liquids, do little
to diminish the chromium content around them,
and have a small effect on lowering pitting re-
sistance. Pitting resistance is still affected to a
degree by alloy depletion due to solidification Fig. 24 Influence of sulfur level on pitting resistance of welds
without homogenizing anneal. FA, ferrite forming
segregation. However, if the alloy solidifies in a first on solidification as opposed to austenite first, AF. Source:
ferritic mode (FA, i.e., ferrite forming first on Ref 23
can also be altered by freezing rate. Faster cool- stress corrosion cracking (SCC). There have
ing favors austenitic solidification. been numerous proposed mechanisms for the
Long-term annealing of welds has shown breakdown of a passive film in chloride-con-
that sufficient time and temperature to achieve taining media; these have been summarized in
some rehomogenization the alloy result in bet- other publications (Ref 23). These hypotheses
ter pitting resistance (Ref 24), approaching that deal with how a passive film on a homogeneous
of the wrought alloy. Examination of the de- surface could break down. They include:
creasing solubility of sulfur in stainless in Fig. • Adsorption of chloride ions
22 indicates that the precipitation of sulfides • Penetration of the passive film by chloride
that cause chromium depletion occurs in delta
ferrite on freezing when sulfur exceeds 0.007% ions
• Film breakdown by electrostriction
and in austenite when sulfur exceeds 0.003%. • Formation of stable metallic chlorides
Oxygen behaves in a parallel manner and is • Coalescence of cationic vacancies
usually present in sufficient quantities, about • Random localized thinning of the passive
0.01% in manganese/silicon deoxidized steels,
to cause the same phenomenon. This funda- film
• Local variations in the composition of the
mentally is due to the high ratios of the diffu-
sivities of oxygen and sulfur to chromium, corrosive medium
which are about 10,000 and 680, respectively. By and large, these mechanisms presuppose
Whenever fast-diffusing elements such as oxy- a stainless steel surface that is homogeneously
gen, sulfur, carbon, and nitrogen, which have a passive and try to explain the observed inho-
strong affinity for chromium and a solubility mogeneous behavior of the passive film. How-
that decreases strongly with temperature, are ever, since it is clear that the surface is not ho-
present in steel, their precipitation will result in mogeneous, especially with regard to the
some degree of chromium depletion around the passive film, these hypotheses are not neces-
precipitation site because chromium diffuses sary to explain the behavior of everyday stain-
too slowly to be replenished. less steels, which unfortunately have abundant
The low chromium around inclusions is a suf- inclusions and chemical inhomogeneities capa-
ficient condition for the local dissolution meas- ble of locally diminishing the integrity of the
ured as metastable pitting, and if the depletion passive film. More research in understanding
zone shape and size are favorable, then stable the exact nature of the inhomogeneity of stain-
pitting would ensue. less steel surfaces is necessary for a complete
Certain other types of inclusions/precipitates understanding of pit nucleation and therefore
are less harmful in this regard. Titanium, for in- prevention.
stance, which is often added to form carbides Pitting Resistance. Pitting has been exten-
and nitrides, also forms sulfides and oxides sively correlated with environment and compo-
more strongly than manganese and therefore sitional variables. The most well-known and
does so at higher temperatures. Such precipi- useful correlations are between the PREN and
tates have a much lower tendency to allow the critical pitting temperature (CPT) and by ex-
chromium to join in the precipitation since the tension to the pitting potential.
higher the temperature of precipitation the more For austenitic alloys:
that diffusion allows the more favorable reac-
tion to occur. Rare earths also behave the same PREN = % Cr + 3.3 % Mo + 30 % N (Eq 9)
way. Metastable pitting is diminished by the
presence of these elements. For ferritic alloys, which hold no nitrogen in
The initiation of pitting is also affected by solution:
stress and inclusion orientation (Ref 25), which PREN = % Cr + 3.3 % Mo (Eq 10)
the researchers correlated to the dimensions of
the inclusion-derived cavity being able to sus- For duplex alloys, which have two phases,
tain a sufficiently low pH due to iron dissolu- neither of which matches the bulk composition:
tion to maintain stable pitting. The influence of
PREN = % Cr + 3.3 % Mo + 16 % N (Eq 11)
stress was to cause cracking at otherwise unfa-
vorably shaped inclusions, which then provided These equations are useful, if approximate,
a crevice capable of sustaining stable pitting. and their correlation is shown in Fig. 25 (Ref
This will be relevant to later discussions of 26) They do not include tungsten, which, if
alloy behavior predictions. It is noteworthy that

the elements copper and nickel, which are bene-
ficial against uniform corrosion and which slow
the growth of pits by this same action, do not
contribute to increasing the resistance to the
onset of pitting. This is another manifestation of
pitting initiated by the local stability of the pas-
sive film, which is primarily a function of local
chromium content. Molybdenum and nickel
thus seem to bolster local chromium content in
the passive film. Nitrogen seems to act by con-
centrating at the passive film-alloy interface
rather than by buffering the solution by ammo-
nia formation, which has been proposed (Ref
36). Research on very pure sputtered films of
iron-chromium alloys have demonstrated that
both titanium and niobium in solution diminish
active dissolution, assist repassivation, and im-
prove pitting resistance (Ref 29). In most practi-
cal cases, these elements are not found in solu-
tion because of their affinity for oxygen, sulfur,
Fig. 25 Variation of critical pitting temperature with pit- carbon, and nitrogen, with which they form
ting resistance equivalent number (PREN) of
austenitic steels in water plus 6% FeCle. Source: Ref 26 compounds.
It should also be noted that the critical PREN
values vary with crystallographic structure. Fer-
ritic alloys require somewhat lower PREN val-
ues to exhibit similar pitting resistance as
austenitic alloys of somewhat higher PREN.
While pitting is of great theoretical and prac-
tical interest, there are significant problems in
actually conducting good tests. Monitoring of
the electrochemical potential during the test is
considered mandatory by most researchers.
How is a metallic sample suspended in a solu-
tion without creating any crevices and without
exposure at the liquid-gas interface? The devel-
opment of the flooded gasket technique (used in
ASTM G150) was a milestone, but it also has
Fig. 26 Differential variation of critical pitting temperature some problems—most notably the potential for
of several stainless steel alloys for unwelded dilution of the test solution, especially during
wrought and welded material. Source: Ref 13
prolonged testing. FeCl3 testing benefits from
the fact that the solution creates a reproducible
present, has half the effectiveness of molybde- positive potential.
num. They neglect carbon, which seldom varies While the PREN approximates the pitting re-
enough to have a visible effect but has been sistance of an alloy, there is a standard test by
shown when in colossal supersaturation to have which the CPT is measured. Pitting in a given
a factor of about 10, not unlike nitrogen, an- medium capable of causing pitting does not
other interstitial that it resembles in solution occur below a temperature that is characteristic
thermodynamically (Ref 27). It also does not in- of the medium and the material, with the myriad
clude the negative influence of elements such as exceptions of stress state, surface finish, mi-
sulfur. Likewise, the equations cannot deal with crostructure, etc. The most commonly used test
inhomogeneity issues, so welded alloys have media are the unacidified 10% FeCl3, which is
different CPTs for the same PREN (Fig. 26) used in the ASTM G 48 practice B, and the 3.5%
(Ref 13). These equations are all-other-things- NaCl solution of the ASTM G 150. The latter, if
being-equal equations and are useful for gross modified to 0.1N NaCl, allows the ECPT, the
electrochemical pitting potential, of lower alloys Cl = %Cr + 4.1%Mo + 27%N (Eq 12)
such as 304 to be measured (Ref 30).
Since a crevice has a preexisting favorable
geometry for pit growth, any pitting event,
Crevice Corrosion metastable or stable, can initiate ongoing
In the case of pitting, the geometry that crevice corrosion. Crevices are thus incubators
makes up the pit is essential in creating the dif- for corrosion triggered by metastable pitting
ferential aeration cell and to cause the autocat- events. The dissolution of iron during passiva-
alytic dissolution process. In many cases, a tion itself as well as the differential oxygen cell
geometry that retains and acidifies water is al- created by the crevice contribute to the process.
ready present in crevices in different types of It is logical to think that alloying the elements
structures such as gaskets, under faulted coat- that contribute to lowering the critical current
ings, under bolt or screw heads, etc. Crevice density for passivation and the uniform corro-
corrosion occurs because zones have restricted sion rate, such as nickel, would reduce the cre-
access of reactants and restricted exit of corro- ation of the reactants that start the crevice corro-
sion products. It is especially the inhibition of sion process, but this presumed effect is not
the cathodic reaction inside the crevice by the strong enough to be reflected in this actual be-
dearth of oxygen, which sets up a more aggres- havior Eq 12 represents, although it is generally
sive environment within the crevice than with- acknowledged that austenitic steels perform bet-
out. The interior reactions become increasingly ter than ferritic steels in the absence of molyb-
anodic, and the aggressiveness of the environ- denum. Materials are characterized as having a
ment can reach a threshold at which active cor- critical depassivation pH. If crevice conditions
rosion occurs, while the situation exterior to the are such that the reactions over time allow the
crevice is safely passive. Crevice corrosion oc- pH to be reduced to this level, then active corro-
curs at lower temperature than pitting in the sion will begin within the crevice. Thus, passive
same environments, so it is a greater danger in film stability seems to be the critical factor
that sense. rather than corrosion rate after initiation.
The relationship between the alloy content, Preventing Crevice Corrosion. The coun-
given as the crevice corrosion resistance equiv- termeasures against crevice corrosion are ca-
alent number (CCREN), and critical crevice thodic protection, design, maintenance, and, of
corrosion temperature (CCT), shown in Fig. 27 course, alloy selection. Designing to avoid
(Ref 11), is similar to that of PREN (PI) to CPT crevices should include maximizing the volume
except for the molybdenum factor being more of unavoidable crevices, engineering flow to en-
important: hance transport in and out of crevices, and
Fig. 27 Variation of critical crevice corrosion temperature with alloy content

avoiding stagnation. Any maintenance or design chromium. In the various grades of stainless
procedure that prevents formation of deposits is steels, there are many intermetallic phases that are
beneficial. Welds are particularly vulnerable thermodynamically stable but kinetically slow to
surface sites, so any combination of welds and precipitate that are enriched in chromium. An ex-
crevices or crevices caused by poor weld geom- ample of such a phase is chromium carbide (Fe,
etry must be avoided. S32205 is a benchmark Cr)23C6. These phases tend to form at grain
alloy of sorts. It has just sufficient alloying to boundaries where nucleation is favored, resulting
resist pitting in seawater, but it is susceptible to in a depletion of chromium in the adjacent re-
crevice corrosion. gions, as shown in Fig. 28. Thus, the chromium-
As a practical matter, crevices are almost im- depleted regions near the grain boundaries are
possible to eliminate. Threaded fasteners and sensitized in that they behave as active anodes
joints represent severe crevices and should be compared to the larger interior of the grains that
avoided in aggressive environments if possible. are still passive. In an aerated corrosive environ-
Gasketed joints are another severe crevice loca- ment, the smaller chromium-depleted nonpassive
tion, and their usage should be curtailed to the anodes dissolve, whereas the larger cathodes re-
minimum practical extent. In these situations, duce oxygen, resulting in a localized corrosion
judicious use of very expensive, highly corrosion along grain boundaries. Any heat-treating or
resistant materials is justified. The use of smooth welding procedure of stainless steels should thus
welded joints is thus generally preferred. In a be tailored to avoid sensitization.
more general consideration, deposition and When a stainless steel is heat treated, there is
fouling create crevice sites, and design and op- a risk that the unwanted phases may form, de-
erational controls to preclude the formation of pending on the time-temperature history and
deposits and the prompt removal of sludge and precipitation kinetics of the unwanted phase.
the like are necessary. But in some situations, Figure 29 shows schematically the temperature
such as marine exposures, biofouling will create versus time due to welding and the resulting
crevice sites. This fouling may be macroscopic, sensitization. Figure 29 shows a TTT (time-
such as from shellfish and barnacles, or it may temperature-transformation) curve for precipi-
be microscopic. Microscopic biofouling causes tation of the unwanted phase. Near the weld
the special form of crevice corrosion called mi- (A), the time spent in the temperature region
crobiologically influenced corrosion (MIC) dis- where precipitation occurs is too short, whereas
cussed in a separate section). far away from the weld (C) the temperature ex-
perienced is too low. At location B, there is,
Sensitization/Grain Boundary Corrosion however, a risk for sensitization.
The maintenance of a passive layer in a wide Austenitic. Sensitization can occur at any
range of pH conditions in stainless steels is de- temperature at which carbon is supersaturated
pendent on the alloying elements, primarily in an alloy. Current austenitic stainless steels
have carbon levels of under 0.10% normally
Fig. 28 Schematic illustration of sensitization due to Fig. 29 Schematic illustration of how a heat treatment re-
chromium-rich precipitates that deplete adjacent lates to sensitization due to precipitation kinetics.
regions of chromium. GB, grain boundary TTT, time-temperature-transformation
and under 0.03% for low-carbon L grades. activity of carbon and make alloys more suscep-
Thus, normal grades sensitize below around tible. Nitrogen lowers the tendency for carbide
800 °C. The supersaturation increases with formation and slows sensitization. Nonthermo-
decreasing temperature, but below about 500 °C dynamic effects are those of austenite grain size
diffusion of carbon is too slow for carbon to and prior cold work. Decreasing grain size and
move to grain boundaries and cause the damag- therefore increasing grain boundary surface
ing combination with chromium that causes area decreases the amount of precipitate per unit
sensitization. Low-carbon grades avoid sensiti- area of grain boundary and therefore the amount
zation because they are not sufficiently super- of chromium depletion per unit area. Cold work
saturated at temperatures at which carbon is accelerates diffusion and makes precipitation
mobile enough to diffuse to grain boundaries. more rapid, thus aggravating sensitization.
Ferritic. Another situation exists in ferritic The thermodynamic affinity tool can be used
stainless steels, in which carbon is much less to prevent chromium carbide formation in an-
soluble but is much more mobile. Annealing other way. Introducing alloying elements that
over 900 °C can put enough carbon in solution to combine with carbon more strongly and rapidly
cause sensitization even at the lowest carbon lev- than chromium can exhaust the supply of car-
els attainable in an AOD and even at the fastest bon available to precipitate as chromium car-
possible quench rates. The damaging chromium bide. There are a number of candidate elements,
depletion caused by this very rapid precipitation zirconium, vanadium, tantalum, niobium, and
can be undone by a simple rehomogenization titanium, most prominently. Of these, the diffu-
anneal of the remaining chromium. This is theo- sivity and affinity for carbon of niobium and ti-
retically possible with austenitic alloys also, but tanium make them the best for this purpose.
the diffusion rates of chromium in austenite as Each forms stable carbides at much higher tem-
so slow that it is impractical in most real cases. peratures than chromium, starving chromium of
Duplex steels have a subtle near immunity to sufficient carbon to form damaging precipitates.
carbide sensitization. While they are typically The caveat with titanium is that it forms oxides,
low carbon anyway, the carbides that do form sulfides, and nitrides preferentially to carbides.
do so at ferrite-austenite grain boundaries. Here, Therefore, sufficient quantities must be used to
chromium is consumed from the chromium-rich accommodate the prior formation of these
ferrite phase, leaving the austenite intact. Their phases. Niobium tends more toward carbide
large grain boundary area keeps carbide con- than nitride formation but is a weaker carbide
centration per unit area low, and the fast diffu- former than titanium. The solubility products of
sion in the ferrite keeps austenite from becom- these precipitation reactions are:
ing depleted. However, the rapid formation of
intermetallic phases at the ferrite-austenite in- 6780
log [Ti ][C ] = 2.97 − (Eq 13)
terfaces can lead to a rapid loss of corrosion re- T
sistance and a severe loss of toughness if expo-
sure to temperatures within the intermetallic
precipitation range is not controlled. 9350
Martensitic steels are quenched as austenite log[ Nb][C ] = 4.55 − (Eq 14)
T
to and through the Ms temperature without time
for carbon to precipitate in austenite. The car- These equations follow the form of the gen-
bon in the martensite can precipitate and cause eral equation for precipitation reactions:
sensitization if reheated to the 300 to 700 °C re-
gion. Fortunately, heating to above 700 °C re- log[ M ][ X ] = A − H / RT (Eq 15)
homogenizes the chromium and eliminates sen-
sitization. in which A is a constant, H is the heat of disso-
Effect of Alloying. Besides determining lution, R is the gas constant, and T is the ab-
basic phase structure, alloying plays a role in solute temperature. If the amount of titanium or
susceptibility to sensitization. Those elements niobium is stoichiometrically sufficient, no car-
that reduce the tendency of chromium carbides bon will form chromium carbides under equilib-
to form also reduce the susceptibility to sensiti- rium conditions. It is possible to defeat the sta-
zation. This is a purely thermodynamic effect. bilization reactions by quenching the alloys
Molybdenum, silicon, and nickel promote car- from temperatures at which titanium carbide or
bide formation by increasing the thermodynamic niobium carbide is dissociated. If free carbon is
left free in the matrix by quenching, then on re- layer occur, and region II, where the protective
heating it may form carbides with the most lo- layer is not fully developed, suggesting an ap-
cally accessible favorable element, such as preciable electrochemical effect. The latter is a
chromium, rather than the most thermodynami- zone that exists in alloys that have zones of
cally favorable element, which would be tita- chromium depletion.
nium or niobium. This can occur when a stabi- Stress corrosion cracking has always been
lized alloy such as 321 is welded. A zone away among the most controversial subjects among
from the weld may experience a high enough metallurgists and electrochemists. The debate
temperature to put carbon into solution and then centers on whether the critical mechanism is
cool just rapidly enough to not form only the dissolution or fracture, and if a fracture, by what
equilibrium titanium carbide but also Cr23C6 at mechanism. Is the cracking zone locally soft-
grain boundaries, causing the type of sensitiza- ened, locally hardened, transformed, to a more
tion called knife-line attack. This problem has brittle phase or embrittled by hydrogen? As of
nearly ceased to exist as modern 321 has low this writing, there is no general agreement on
levels of carbon and nitrogen for economic rea- which type mechanism is the fundamental
sons; this effectively precludes this chromium cause, but there is room for convergence. Obvi-
carbide precipitation in most cases. ously, elements of many may come into play. It
Welding. Many of the most severe problems is likely, as in most prolonged arguments, that
of sensitization arise when stainless steels are no hypothesis is completely correct. We will try
welded to carbon or low-alloy steels. In these to fairly set out what is known and agreed on as
situations, construction code rules usually re- fact and then present researchers’ views in an
quire that the carbon steel component be given a unbiased manner, but since we concern our-
stress relief annealing (SRA) treatment. Such selves only with stainless steel, no attempt is
SRA treatments are typically in the sensitization made to address an all-encompassing theory.
temperature range for austenitic stainless steels. Crack Initiation. In stainless steels, cracks
Use of low-carbon or stabilized grades is neces- can be seen to initiate at surface defects and ir-
sary in such cases. Even then, use of the lowest regularities. In stainless steel, it must be agreed
allowable temperature SRA treatment for the by all that the preponderant initiation site is a
shortest allowable time is preferred. corrosion pit or, in some cases, a crevice. Inter-
granular corrosion sites, as are seen in sensi-
tized material, can also provide the conditions
Corrosion Combined with for SCC initiation. The interrelationship be-
tween pits and SCC cracks has been studied
Fatigue or Fracture
(Ref 25). Stress lowers the anodic potential at
which pitting occurs and permits metastable pits
Environmentally induced failure occurs when
brittle failure under tensile mechanical loading
occurs at a lower stress when a material is sub-
jected to a corrosive environment than what
would happen in a noncorrosive environment.
This introduces us to what is perhaps the most
controversial technical subject in all of stainless
steel research, SCC.
Stress Corrosion Cracking

The key cause for SCC is the cooperating ef-
fects of tensile stress and a corrosive environ-
ment. Such cases can be identified in most alloy
systems, and even pure metals, which were
thought to be more or less immune, also have
had cases of SCC reported.
In passive metals, two sensitive potential re-
gions for the occurrence of SCC have been
identified and are shown in Fig. 30: region I,
where pitting and breakdown of the passive Fig. 30 Zones of susceptibility to stress corrosion cracking
to become stable via the generation of cracks. environmental variables. The most common ex-
Cracks, once formed, presumably have favor- ample is that of sensitized 304 in high-tempera-
able geometry to duplicate pit internal chemical ture water or caustic media. The relevance of
reactions and must be considered to be de- this to the normal case of stainless steels must
scribed by the models that apply to pits and be questioned since, by definition, the sensitized
crevices. grain boundaries themselves can be depleted of
The stress at which SCC initiates has a chromium to a degree they are not stainless and
threshold, which has been reported as between have a much less stable austenitic structure,
25 and 50% of the yield strength in austenitic having their martensite start temperature Ms,
stainless steel. raised by the loss of chromium.
The temperature at which SCC is initiated Material Variables. Martensitic stainless
ranges from ambient to under 100 °C for steels and martensitic precipitation hardened
martensitic materials, while austenitic alloys stainless steels are quite susceptible to SCC.
begin their sensitivity above room temperature This susceptibility increases with hardness,
and increase in susceptibility with increasing yield strength, and embrittling heat treatments.
temperature. The ferritic steels, while consid- They will crack at threshold stresses equal to
ered nearly immune to SCC, have their maxi- 50% of yield strength. These alloys can be tem-
mum susceptibility in the same range as marten- pered at sufficiently high temperatures that they
sitic steels. In environments of mixed chlorides become soft and tough enough to have very
and sulfides, however, SCC can occur in all good resistance.
types of stainless at room temperature. This has Ferritic stainless steels of low and medium
been seen in the SCC of austenitic stainless chromium are generally not susceptible to SCC.
steel in swimming pool environments, in which Ferritic alloys, which can have a martensitic
chloride ions can condense on the stressed steel structure, should be considered martensitic for
and cause pitting and SCC. SCC purposes. If purely ferritic alloys are al-
Cracks propagate very slowly below spe- loyed with copper, molybdenum, and nickel,
cific certain stress intensity levels, but once that they can become susceptible. The presence of
intensity is reached, they have a plateau rate α'( or high-temperature embrittlement also in-
that is fairly constant until the stress level at creases susceptibility, as does cold work.
which catastrophic failure occurs at very high Despite the controversy surrounding the
propagation rates. Rates of crack propagation mechanism of SCC in austenitic stainless steels,
are exponentially increased by increasing tem- there is almost complete agreement that SCC of
perature. The crack propagation rate has been body-centered cubic (bcc) stainless steels,
seen across a range of alloys to be linearly pro- martensitic, ferritic, and pH is simply a mani-
portional to the average current density that festation of HE, with hydrogen provided by ei-
alloy experiences when its surface is strained, ther anodic (e.g., active corrosion within a pit)
indicating that reactions at the crack tip are or cathodic reactions.
strain sensitive, and overall rate limiting, but Duplex stainless steels have low susceptibil-
not necessarily the mechanism of cracking. ity to SCC. Their dual-phase microstructure en-
Crack growth is discontinuous with individual sures that under conditions that crack austenite,
steps of growth many times the average rate, ferrite remains as a crack-arresting phase, while
which is similar to that seen with gaseous hy- under conditions that cause SCC in highly al-
drogen embrittlement (HE). The crack growth loyed ferrite, the austenite is a crack arrester.
gives off acoustic emissions as cracking steps A second explanation of the resistance of du-
occur. These steps of growth are brittle and are plex alloys to SCC is that their two phases have
seen as facets on fractographs with cleavages different corrosion potentials, and that the
corresponding to crystallographic planes. The mixed potential that arises because they are in
crack facets match with high perfection, show- intimate contact is outside the potential range
ing almost no evidence of plastic deformation for SCC on either phase. This fits with the re-
or dissolution. sistance to SCC of wrought alloys with a lamel-
The propagation path may be intergranular or lar structure and the lesser resistance of cast al-
transgranular. Grain boundary propagation in loys that lack that structure.
stainless steels usually corresponds to condi- Austenitic stainless steels are the type of stain-
tions under which grain boundaries are less cor- less steel generally associated with SCC, and
rosion resistant because of either material or they vary in their degree of susceptibility to
SCC. All other things being equal, alloying which are probably the greater cause of resist-
elements that delay or prevent localized corro- ance to SCC.
sion do the same to delay SCC. This is simply Environmental Variables. There are three
the delay of initiation. However, if pitting can be key types of environments in which SCC occurs
delayed indefinitely, then SCC can also, assum- in stainless:
ing, of course, more harmful localized corrosion, • Chloride-containing solutions
such as that due to intergranular chromium de- • Caustic solutions
pletion, is not occurring. Molybdenum, which • Polythionate and thiosulfate solutions
we already know helps prevent pitting and
crevice corrosion, also increases the threshold The cases of polythionate and thiosulfate solu-
stress for SCC, as shown in Fig. 31 (Ref 31). tions are industrially important but can be ade-
But, if metastable or stable pitting is occurring, quately explained as simply the stress-assisted
the threshold stress has been reached, and the intergranular corrosion of sensitized material.
temperature is sufficient, then SCC will proceed. Hot caustic solutions are aggressive against
It is mitigated by material variables such as cold stainless steels. Certain combinations of concen-
work and by alloying elements that increase trations, temperature, impurity, and dissolved
austenite stability. Many publications cite nickel oxygen can cause SCC as well as other undesir-
as beneficial in enhancing resistance to SCC, able corrosive attack. Resistance to general cor-
often referring to the data from Fig. 32. However, rosion is proportional to nickel content, but fer-
its role seems mainly to be as an austenite stabi- ritics and duplex alloys are less prone to SCC.
lizer and as a retarder of active corrosion. The 304 has been reported to have no meaningful
minimum in the curve corresponds to the nickel threshold stress for SCC in hot caustic solutions,
level at which the structure is entirely austenitic, leading one to question whether such a failure
but least stably so. Lower nickel levels produce should even be classified with SCC of the typi-
better immunity through the duplex structure, cal chloride-induced type or belong with the pre-
while higher levels promote austenite stability vious polythionate and thiosulfate solutions.
and correspond to alloys having more alloying Chloride containing environments are the
elements, such as chromium and molybdenum, main ones that induce SCC. Water can cause
Fig. 31 Influence of molybdenum on resistance to stress Fig. 32 Variation of resistance to stress corrosion cracking
corrosion cracking (SCC) in austenitic steels with nickel (and other) content and structure
SCC at sufficiently high temperatures (i.e., does advance discontinuously, and after each
above 100 °C) if there are even very low com- advance there is fresh surface, which comes
bined concentrations of chloride (greater than into equilibrium with the solution within the
0.1 ppm) and oxygen (greater than 0.1 ppm) crack. So, any experiment, such as that shown
dissolved (see Fig. 33) (Ref 32). in Fig. 34 (Ref 33), that tests crack propagation
Failure times decrease exponentially with de- against electrochemical events will absolutely
creasing chloride content. Crack growth rate in- support this model.
creases by a factor of ten with each 30 °C rise in It is axiomatic that films must rupture and re-
temperature. Decreasing pH lowers the temper- form as cracks advance discontinuously. The
ature at which SCC occurs in a given time. weakness of this model is that it does not pro-
Mechanisms. There have been many mecha- vide a mechanism for brittle fracture, and the
nisms proposed for SCC in stainless steel. We very brittle features of transgranular SCC frac-
focus only on those that address the failure in ture surfaces do not show any supporting evi-
chloride-containing media, the main concern for dence of dissolution. Research (Ref 34) show-
users of stainless steel. ing that metal dissolution at the crack tip is
The models that have found some support isotropic rather than crystallographically ori-
are: ented make dissolution models incapable of
• Slip dissolution being reconciled with the crystallographic frac-
• Adsorption-enhanced plasticity ture surface facets.
• Adsorption-induced brittleness Adsorption-induced brittleness, also known
• Hydrogen embrittlement as stress-sorption, looks to the parallels be-
tween liquid metal embrittlement and SCC to
Slip dissolution (anodic dissolution) was the explain the mechanism of SCC as the action of
earliest proposed model for SCC. It simply pro- adsorbed species weakening atomic bonds on
poses that at a crack tip a passive film forms, and the crack tip surface. If the action is on the sur-
after time it fractures by an unspecified mecha- face, however, the mechanism cannot produce
nism. The fresh active surface may or may not the observed discontinuous, brittle cracks which
repassivate, after which the process repeats itself. characterize SCC. Only in alloys such as Fe-3Si
The strength of this model is that it actually are steps small enough to make this mechanism
does describe what is happening. The crack plausible.
Fig. 33 Variation of susceptibility to stress corrosion cracking (SCC) with media oxygen and chloride content for 304 stainless steel.
Source: Ref 32
Fig. 34 Crack propagation rates of various metals plotted versus current density. Source: Ref 33
Adsorption-enhanced plasticity/hydrogen em- actually take place, but none is specific enough
brittlement encompasses a number of models to have been tested by critical experiments to
that observe that adsorbed species enter the lat- prove or disprove it.
tice in the vicinity of the crack tip and then It has been demonstrated that hydrogen is ab-
cause failure by one of several mechanisms: sorbed into the material at the crack tip. The
• Dealloying and porosity main question is whether it causes damage by
• Adsorption-induced brittleness creating porosity, altering dislocation mobility,
• Coalescence of voids formed by cross slip or causing lattice decohesion. There is support
for each.
enhanced by the adsorbed species It has been observed that where SCC occurs
Since hydrogen is the only species that is pro- there is a large concentration of vacancies. This
duced in quantity and is capable of diffusing into has led to speculation that porosity is a weaken-
the lattice, HE is implicit in all these models. ing mechanism responsible for SCC (Ref 35). It
All of the models have support in that they has been proposed and supported by calcula-
have some experimental observations that show tions that hydrogen lowers the energy required
that the phenomena they propose as causal or vacancy formation. The lowest energy
configuration is calculated as two hydrogen

atoms per vacancy. This pairing of hydrogen
solute atoms to dislocations is very reasonable
given the major distortion the interstitial hydro-
gen causes to the lattice, so there is no basis to
challenge the enhanced vacancy formation.
Whether the effect is large enough to cause fail-
ures has not been demonstrated. The largest
measurable effect of hydrogen has been a slight
acceleration of stress relaxation in martensite.
The relevance of hydrogen-induced vacancy ag-
glomeration as the principal cause of failure must
be considered questionable until some further
critical experiments link the vacancies to the ob-
served instances of failure quantitatively, and Fig. 35 Stress-strain curve for single crystals of stable
austenitic stainless steel with and without hydrogen.
more important, to show how this mechanism Source: Ref 36
could account for the temperature and stress de-
pendence observed. The major influence of va- in Fig. 35 (Ref 36). All these interstitials strain
cancy formation due to hydrogen may be to en- the lattice and therefore harden in proportion to
hance the volume expansion due to hydrogen. their atomic size. Hydrogen, as the smallest of
Adsorption-induced brittleness, also known them, has about half the distorting effect and
as stress sorption, looks to the parallels be- half the hardening effect. But, its small size
tween liquid metal embrittlement and SCC to makes it mobile at ambient temperatures, so it
explain the mechanism of SCC as the action of can diffuse to sites where it can alter mechani-
adsorbed species weakening atomic bonds on cal properties.
the crack tip surface. If the action is on the sur- But, while hydrogen causes dislocation mo-
face, however, the mechanism cannot produce tion and lower work hardening, it does not
the observed discontinuous, brittle cracks that weaken austenite, so this theory by itself cannot
characterize SCC. Only in alloys such as Fe- account for the role of hydrogen in SCC and, by
3Si are steps small enough to make this mech- inference, in HE in the more general case.
anism plausible. In stainless steels, there The quandary of hydrogen finally having
seems to be nothing to support this proposed been shown to have a clear effect on mechanical
mechanism. properties but having that not account for either
Adsorption-enhanced plasticity has become SCC or HE may be put to rest by the additional
known recently as HELP or hydrogen-enhanced observations of hydrogen’s role as a lattice dis-
localized plasticity. The underlying mechanism torter (Ref 37). While not formalized as a pro-
at work in this model is the hydrogen-induced posed hypothesis for SCC, the role of hydrogen
shielding between microstructural defects. This as a generator of very high stresses has been
has been observed distinctly in single crystals of pointed out as a factor that cannot be neglected
austenitic stainless alloys. The Cottrell atmos- when evaluating other proposed mechanisms.
phere of hydrogen around dislocations causes Hydrogen has been shown to distort the lattice
mutual repulsion, causing strain to be localized in proportion to its concentration. The effect is
on certain slip systems. This has been observed not small, accounting for about 1% strain per
to occur and has caused deformation to become 0.1% concentration by weight, as shown in
concentrated in Luders bands in austenitic al- Fig. 36 (Ref 38). At hydrogen levels of over
loys, which of course do not show such behav- 1000 ppm, which are thought to exist around
ior without hydrogen (Ref 36). This also pro- growing SCC crack tips, there could therefore
duces ε-martensite in austenitic alloys, which be hydrogen concentration gradients capable of
would be considered stable without hydrogen producing additional tri- or biaxial stresses on
and deformation. the matrix ahead of the crack tip that may ap-
This theory encounters a problem, however, proach the yield stress and account for some or
with the fact that the same studies showed that all of the difference between the normal fracture
hydrogen actually strengthens the matrix by toughness KI and the KISCC, that for SCC. This
solid solution hardening. It acts in much the also precludes the necessity of hypothesizing
same way as carbon and nitrogen do, as shown hydrogen-induced phase changes, although
Fig. 36 Dilation of austenite due to hydrogen in solution. Source: Ref 38
were they to exist, they would result in the same volved distinguishing among the same mecha-
lattice expansion. In both cases, the failure nisms, namely:
would occur at a region ahead of the crack tip • Decohesion
and beyond the highest hydrogen concentration, • Enhanced local plasticity
which is what is observed to occur. The growth • Adsorption embrittlement
of this stress over time with increasing hydro- • Void coalescence
gen-producing corrosion would account for the
observed kinetics, locus, and stress dependence The identification of the operative mechanism
of SCC. for HE involves again distinguishing what role
If nothing else, the main models for SCC and each of the above contributes to HE in a given
the experimental results on which they are situation since all are known to be real metallur-
based should be reexamined in view of the fact gical phenomena.
that the stresses induced by hydrogen are not The main difference between HE and SCC in
negligible and, in fact, may account for much of stainless steel is that HE is limited to ferrite,
the observed SCC behavior of stainless steels. which is hardened by cold work or alloying, and
The next few years may finally see the resolu- martensite. Austenite is somewhat diminished
tion of the lengthy debate over the causes of in ductility by hydrogen, but not subject to the
SCC. If it comes, it will be from critical experi- completely brittle, discontinuous cracking of
ments, which can quantitatively differentiate bcc stainless. The observations that make a
among the above effects and measure the con- given model plausible as a mechanism for SCC
tribution of each. lack traction for the same materials in HE. It
hard to envision enhanced plasticity involved in
the completely brittle fracture of high-strength
Hydrogen Embrittlement
martensitic stainless steels, whereas void coa-
Like SCC, there has been debate about HE lescence by vacancy creation seems more likely
that has produced more heat than light. This in- to account for the observed behavior.
The resolution of mechanism here also must Biocorrosion and Microbiologically

account for the contribution of hydrogen-in- Induced Corrosion
duced stress as well as hydrogen effects on me-
chanical processes, especially since the ob- There are many cases for which biological or-
served susceptibility to HE is proportional ganisms contribute to initiating or enhancing
hardness, therefore to the amount of hydrogen a rates of corrosion. This can occur in natural en-
given material can hold both in normal intersti- vironments such as ground or seawater as well
tial solution and the amount it can trap at lattice as domestic and industrial environments such as
defects (Ref 39), especially the dislocations the nuclear and chemical processing industries,
within the plastic zone at the crack tip, which for example. This is called biocorrosion or
provide enhanced hydrogen solubility where it MIC, microbiologically induced corrosion.
can aggravate the applied crack opening with a The bacteria that are known to influence cor-
wedge effect from hydrogen dilation. rosion can be sorted as aerobic bacteria that lie
in aerated water and anaerobic bacteria. Among
Corrosion Fatigue the anaerobic bacteria that are known to [16]
Just like SCC, corrosion fatigue causes brittle (Ref 40) affect stainless steels can be counted:
failure under a combined environment of corro- Desulfibrio and Desulfotomaculum. Both of
sion and a tensile stress component. The stress, these are so-called sulfate-reducing bacteria
however, is cyclic and in a test of stress versus (SRB), which means that they promote the reac-
number of cycles (S vs. N), failure will occur at a tion:
lower N under the corrosive environment. The SO 24− → S2− + 4 O (Eq 16)
cracks are transgranular, and the collaborative ef-
fect of corrosion and fatigue is that corrosion ac-
celerates the plastic deformation that accompa- which in turn accelerates the cathode reaction:
nies the evolution of extrusions and intrusions.
In corrosion fatigue, an obvious pit corrosion 2H = + 2e − → 2H (Eq 17)
site may not be necessary because of the com-
bined action of cyclic stresses and the environ- Aerobic bacteria flourish under oxygen (Ref
ment. However, an initiation site that is the 40). Examples are the iron-oxidizing Gal-
weakest link in a combined mechanical and lionella and Sphaerotilus, which increase the
metallurgical sense will be the initiation point anode dissolution reaction:
after which conditions that may not cause SCC
can help propagate fatigue cracking at lower Fe → Fe 2+ = 2e − (Eq 18)
stresses than would be expected in more benign
environments and in environments that may not by converting the ferrous iron-ion product
cause SCC or pitting under static loads. (Fe2+) to less soluble ferric (Fe3+). Due to this,
The importance of the environmental interac- macroscopic so-called tubercules form that can
tion is reflected in the sensitivity to frequency of cause crevice-type shelters where differential
stress application. High-frequency loading aeration and pit initiation can occur.
gives less time for corrosive attack and brings Countering MIC with biocides can cause
crack propagation rates down closer to those in problems in manganese-containing waters. Oxi-
air. In some materials, crack propagation rates dizing biocides, such as ozone, chlorine, or per-
are elevated above those in air at all stress lev- oxide, can cause manganese to be oxidized to
els, while in others a threshold stress intensity manganese dioxide. The precipitated deposits of
must be reached before an acceleration is noted. manganese dioxide can accelerate pitting corro-
Some materials show a combination of both. sion even in low-chloride waters in which al-
The first case seems to be merely fatigue as- loys such as 316 would otherwise be safe from
sisted by corrosion, while the last two seem to pitting attack.
indicate an SCC–type behavior. Biocorrosion is most commonly encountered
The same uncertainties that cloud our under- in ambient aqueous environments, which are the
standing of SCC necessarily disguise the pre- environments in which most microorganisms
cise mechanism of corrosion fatigue, which have evolved to thrive. So, it tends to be a prob-
must be viewed as a combination of SCC and lem for the medium-alloyed steels such as 304
fatigue. and 316, which are used in these environments.
The more chemically or thermally hostile envi- 15. R.C. Newman, Corrosion, Dec 2001,
ronments in which higher alloyed grades are p 1030–1041
used are also hostile to bioorganisms and thus 16. N.J. Laycock and R.C. Newman, Corros.
minimize the problem. Sci., Vol 39, 1997, p 1771
The development of microbiological consortia 17. Y. Kobyashi, S.Virtanen, and H. Bohni,
allow anaerobes to flourish under biofilms that Proc. Electrochem. Soc., 1999, p 533–540
form in an aerated environment. These represent 18. Z. Szlarska-Smialowska, Pitting Corrosion
a differential aeration cell that acts just like a se- of Metals, NACE, Houston, TX, 1986
vere crevice. Also, the action of microbes in 19. E.T. Turkdogan, Fundamentals of Steelmak-
raising the corrosion potential is key to under- ing, Institute of Materials, 1996
standing why natural seawater is so much more 20. H.S. Kim and H. Lee, Met. Trans. A,
corrosive than sterile sodium chloride or syn- Vol 32A, June 2001, p 1519
thetic seawater solutions. And, macrofouling or- 21. M.P. Ryan, D.E. Williams, et al., Nature,
ganisms are important. They create crevices and Vol 415, Feb 2002, p 770–777
sites where microfouling can start early. At the 22. A.J. Grekula et al., Corrosion, 40, 1984,
same time, they are sources of turbulence in p 569
flowing systems, and this turbulence can cause 23. Stainless Steels Les Editions de physique, 1993
flow erosion in copper materials, making use of 24. N. Suutala and M.Kurkela, Stainless Steel
stainless steels more attractive. ‘84, Metals Institute, 1985, p 240–247
25. T.Suter, E.G. Webb, H. Bohni, and
REFERENCES R.C. Alkire, J. Electrochem. Soc., Vol 148
(No. 5), 2001, B174–B185
1. M.P. Ryan et al., Critical Factors in Local- 26. M.O. Spiedel, Stainless Steels ‘87, Institute
ized Corrosion, Proc. Electrochem Soc., Vol of Metals, London, 1988, p 247–252
150, 2003, p 284–294 27. Y. Cao, F. Ernst, and G.M. Michal, Acta
2. K. Sieradski and R.C. Newman, J. Elec- Mater., Vol 51, 2003, p 4171.
trochem. Soc., Vol 133, 1986, p 1980 28. G. Lothongkum et al., Corros. Sci., Vol 48,
3. L. Brewer, Science, Vol 161, 1968, p 115 2006, p 137–153
4. W.J. Tobler and S.Virtanen, Critical Factors 29. S. Fujimoto , Sci. Technol. Adv. Mater., Vol
in Localized Corrosion, Proc. Electrochem 5, 2004, p 195–200
Soc., 2003, p 583–594 30. J.D. Fritz, J.F. Grubb, B.W. Parks, and
5. B. Baroux et al., Corros. Sci., Vol 47 (No. C.P. Stinner, Stainless Steel World, KCI,
5), 2005, p 1097–1117 P01488, 2001
6. http://www.alleghenyludlum.com/pages/ 31. M.O. Spiedel, Met. Trans. A, Vol 12A,
products/xq/asp/T.1/qx/productCategory. 1981, p 779
html 32. A.J. Sedricks, Corrosion of Stainless Steels,
7. K. Kimura et al., High Cr Stainless OCTG Wiley, 1979, p 158
with High Strength and Superior Corrosion 33. R.N. Parkins, Br. Corros. J., Vol 14, 1979, p 5
Resistance, JFE Technical Report 7, Jan 2006 34. S. Tahtinen, H. Hahhinen, and T. Hakkarainen,
8. http://www.outokumpu.com/applications/ Stainless ‘84, Metals Institute, 1985, p 143–148
documents/start.asp 35. M. Nagumo et al., Met. Trans. A, Vol 32A,
9. J.E. Truman, Corrosion: Metal/Environ- Feb 2001, p 332
ment Interaction, Vol 1, Newness-Butter- 36. H. Hanninen et al., Hydrogen Effects on
worths, 1976, p 352 Materials Behavior, TMS, 2003, p 201–210
10. J.P. Audouard, Stainless Steels, Les Editions 37. V.J. Gadgil, Scr. Metal., Vol 28, 1993,
de physique, 1993, p 268 p 1489–1494
11. H. Mimura et al., Nippon Steel Tech. Report 38. M. Hoelzel et al., Mater. Sci. Eng. A,
90, July 2004, p 94–99 Vol 384, 2004, p 255–261
12. http://www.alleghenyludlum.com/ludlum/ 39. B.G. Pound, Hydrogen Effects on Materials
Documents/AL-6XN_sourcebook.pdf Behavior, TMS, 2003, p 93–103
13. http://www.alleghenyludlum.com/ludlum/ 40. S.C. Dexter, Microbiologically Influenced
Documents/al610_611.pdf Corrosion, Corrosion: Fundamentals, Test-
14. F. Tagashi et al., Kawasaki Technical Re- ing, and Protection, Vol 13A, ASM Hand-
port 31, 1994 books, 2003, p 398–413
CHAPTER 5
Oxidation
Summary is negative. In Eq 1, the free energy G is decreased

by a lower nobility of the metal (or a higher activ-
STAINLESS STEEL, often considered mainly ity a of a metallic alloying element), a lower tem-
as a corrosion-resisting material, plays an impor- perature T, and a higher partial pressure P of the
tant role as a heat-resisting material. This is oxidizing gas according to:
partly due to its ability to retain strength at higher
temperatures at which many otherwise useful al- ⎛ aMeX ⎞
loying systems, such as aluminum, copper, and ΔG = ΔG 0 + RT ln ⎜ y
⎟ (Eq 2)
⎜⎝ aM PX0.5 y ⎟⎠
even titanium, soften. Stainless steel retains 2
strength and has excellent oxidation resistance

from room temperature to nearly 1000 °C, at In the case of alloy oxidation, for which tem-
which other economical alternatives are lacking. peratures are high enough to form mixed oxides
or spinels, the activities of the oxide species
also need to be considered.
The standard Gibbs free energy ΔG0 is often
Introduction
presented in Richardson-Jeffes (Gibbs free
energy-temperature) diagrams such as the one
High-temperature oxidation is a form of envi-
shown in Fig. 1 (Ref 4).
ronmental degradation of metals and alloys that
It is evident from Fig. 1 that the major alloying
results from the following chemical reaction in
element in stainless steels, chromium, forms a
which metal atoms M react with gaseous oxidants:
thermodynamically significantly more stable
M (s) + 0.5 yX 2 ( g) → MX y (Eq 1) oxide, Cr2O3, than those of the base alloy iron
(FeO, Fe3O4, and Fe2O3) or the major ternary el-
Due to the high temperatures involved, these ement nickel (NiO), and to a great extent, the
reactions are generally rapid and thus are a con- chromium content determines the oxidation be-
cern for high-temperature applications such as havior of stainless steels.
components for power generation. The elec- The Effect of Chromium. The oxidation of
tronegative gaseous oxidant X could be sulfur, multicomponent alloys is a complex process
chlorine, etc., but the discussion here mainly is from both thermodynamic and kinetic points of
limited to oxidation by oxygen or water vapor view. A range of oxides may form with various
(in the latter case, hydrogen would be added as degrees of thermodynamic stabilities and stoi-
a product in Eq 1. For a thorough study of oxi- chiometries (including complex ones with dif-
dation, referred to Ref 1 to 3). ferent cations), and there might be degrees of
solubilities of oxides in one another. Kinetics of
their growth is complex because metal solute
diffusion in the metal phases varies, as do metal
Thermodynamics of Oxidation and oxygen ion mobilities in the different oxide
phases.
As discussed in Chapter 2, Corrosion Theory, a Birks, Meier, and Pettit distinguished between
reaction will be possible when the net free energy two basic types of behavior: (a) a noble matrix
Fig. 1 Standard Gibbs free energy of formation of some metal oxides as a function of temperature. Source: Ref 4
metal with less-noble alloying elements and (b) spinels Fe1.5Cr1.5O4 (with a solid solubility with
both matrix element and alloying elements are Fe3O4) and FeCr2O4 form. The progressive
nonnoble. The concept of nobility is decided by change in oxidation behavior as chromium is
the thermodynamic conditions of Eq 2; that is, added to iron has been described in the litera-
an element for which the free energy defined by ture (Ref 1).
Eq 2 is negative is nonnoble. We first discuss the At lower chromium contents and above a min-
more common case (b), in which oxidation takes imum temperature, an iron-chromium alloy
place under significantly oxidizing conditions, would behave as pure iron, where FeO would
such as air. In such a situation, it can be seen form next to the metal, then gradually Fe3O4 and
from Fig. 1, that the matrix iron is nonnoble and Fe2O3 would form toward the gas as oxygen po-
so are many of the solutes (chromium, molybde- tential increases. Isolated pockets of spinel may
num, aluminum, silicon, manganese, etc.). form within the FeO layer. The oxidation of iron
A high-temperature Fe-Cr-O phase diagram is proceeds predominantly due to the rapid ionic
shown in Fig. 2. It can be seen here that Fe2O3 diffusion of Fe2+ cations on the FeO layer,
and Cr2O3 are soluble in each other, and that the which leads to growth of this layer. If chromium
Chapter 5: Oxidation / 59
Fig. 3 Parabolic rate constants for the growth of several oxides.

Source: Ref 6
Here, m is the added mass, A is the area

exposed to the oxidizing atmosphere, t is the
time exposed, and km is the parabolic rate con-
stant. The subscript “m” is added here to denote
that the reaction is measured as added mass (it
can also be defined for oxide thickness X). As
Fig. 2 The iron-chromium-oxygen phase diagram at 1300 °C. shown in Fig. 3, the parabolic rate constant for
Source: Ref 5 the oxides of chromium, silicon, and aluminum
are low compared to others, and this is the rea-
in the base alloy is increased, the spinel pockets son that these elements are used as alloying ele-
increase, and the mobility of Fe2+ decreases. As ments to reduce oxidation rates for alloys in
chromium content is increased further, a mixed- high-temperature applications.
spinel scale is formed. Iron diffusion through the The solubility of Fe3O4 in the spinel will
mixed spinel is significant, and thus the scale is eventually result in continued iron oxide for-
not yet protective. As chromium content is in- mation, as the iron-chromium system is not an
creased further, an outer layer of Cr2O3 is optimal basis for high-temperature oxidation
formed, and the oxidation behavior becomes resistance, although it might be an option from
similar to that of chromium. A chromium limit an economical standpoint compared to other
of roughly 20% is needed to achieve a perma- alloy systems such as superalloys. The high
nent Cr2O3 scale. This amount decreases if mobility of both iron and manganese in the
nickel is added. While the thermodynamic driv- spinel structure is also an important factor.
ing force is important, the chromium content of The major stainless steels used for oxidation
the alloy can override because the supply of resistance fall into two categories: the ferritic
chromium to the interface becomes dominant. stainless steels and the austenitic. Table 1 lists
At chromium contents less than about 16 wt%, some of the more significant alloys commonly
the oxidation rate is influenced by the rate of encountered in applications for which oxida-
supply of chromium from the alloys beneath the tion resistance is paramount. The value of fer-
oxide. Above 16%, the supply of chromium is ritic alloys (such as 409, 439, and 446) is that
fast enough that chromium gradients are low they are relatively inexpensive, and that they
enough that instead transport in the oxide layer have a thermal expansion coefficient that is
controls the rate. The rate of oxidation then fol- closer to that of the oxide than do austenitic al-
lows a so-called parabolic law (this is explained loys (such as 302B, 309, and 310). This gives
in the next section), by which the mass change them an advantage in cyclic oxidation applica-
per unit area due to oxidation (incorporation of tions even though their strength at high temper-
oxygen) is given by: atures does not rival that of austenitic alloys.
The ferritic stainless steels are the most widely
( m / A)
2
= km t (Eq 3) used alloys based on their low cost, which has
Table 1 Oxidation-resisting grades of stainless steel in common use

Composition, %
UNS Name C N Cr Ni Mn Si Ti Nb Other

S40900 409 0.08 10.5–11.75 0.5 1 1 6X(C + N) to ... ...
1.10
11Cr-Cb(a) 0.01 0.015 11.35 0.2 0.25 1.3 ... 0.35 ...
12SR(a) 0.02 0.015 12 ... ... ... 0.3 0.6 1.2 Al
S43935 439 0.07 0.04 17.0–19.0 0.5 1 1 0.20 + 4X(C + ... ...
N) to 0.75
18Cr-Cb(a) 0.02 ... 18 ... 0.3 0.45 0.25 0.55 ...
18SR(a) 0.015 ... 17.3 0.25 0.3 ... 0.25 ... 1.7 Al
4742(a) 0.08 ... 18 ... 0.7 ... ... ... 1.0 Al
S44600 446 0.2 0.25 23.0–27.0 0.6 1.5 1 ... ... ...
S30215 302B 0.15 ... 17.0–19.0 8.0–10.0 2 2.0–3.0 ... ... ...
S30415 153MA 0.04–0.06 0.12–0.18 18.0–19.0 9.0–10.0 0.8 1.0–2.0 ... ... 0.04 Ce
S38150 253MA 0.05–0.10 0.14–0.20 20.0–22.0 10.0–12.0 0.8 1.4–2.0 ... ... 0.04 Ce
S30900 309 0.2 ... 22.0–24.0 12.0–15.0 2 0.75 ... ... ...
S31000 310 0.25 ... 24.0–26.0 19.0–22.0 2 1 ... ... ...
Note: All compositions include Fe as balance. Single values are maximum, unless otherwise specific (a) Indicates typical analysis
made them the standard alloys for automotive scale-gas surface. Similarly, the electric charge
exhaust systems. can be carried by either n-type (electrons) or p-
type (electron holes) electronic defects. The
case will be determined by the equilibrium de-
Transient Oxidation fect structure of the oxide, which depends on
temperature and oxygen partial pressure. In the
The oxidation of a clean metal surface on ex- case of Cr2O3, chromium cations are the pre-
posure to an oxidizing environment will ini- dominantly mobile defects (Fig. 4b) as a result
tially lead to all the nonnoble components of the of a very small degree of deviation from stoi-
alloy being oxidized together, forming mixed chiometry in the cation lattice, that is, Cr2-5O3,
oxides having composition similar to the base leading to metal deficiency. The contribution of
alloy. In stainless steels, these initial oxides are chromia grain boundary diffusion is large and
typically Fe-Cr-Ni-Mn mixed oxides. This is probably dominates the process at temperatures
called transient oxidation. As these oxides of interest. The defect can be described as an in-
thicken, the partial pressure of oxygen at the teraction with oxygen, at high oxygen poten-
scale-metal interface falls until only the most tials, through Kroger-Vink notations as:
reactive element present in high concentration
can be oxidized. For stainless steels, this means 3 3
O ( g) = VCr''' + OOx + 3h •
that a layer of Cr2O3 is eventually established in 2 2 2
contact with the alloy.
−3 3
⎡VCr''' ⎤ p3 = K 2 pO3/ 4 → ⎡VCr''' ⎤ = K12 K 2 Po4 (Eq 4)
⎣ ⎦ ⎣ ⎦
The Electrochemical Nature of 2 2
Oxidation
The electron holes that form as charge-com-
pensating defects serve as the “electron lead” in
Once an inner scale of Cr2O3 is formed, as
the electrochemical cell in Fig. 4b. The free en-
shown in Fig. 4(a), the oxidizing gas is reduced
ergy of Reaction 4 determines the concentration
at the gas-scale interface, and the chromium is
of mobile defects and thus the diffusion coeffi-
oxidized at the metal-scale interface. The Cr2O3
cient and electrochemical mobility (Be) of the
scale serves as both electrolyte, through which
cation according to:
ions are transported, and electron lead, through
which electronic defects are transported. In
principle, either or both metal or oxygen ions ⎛ ΔGm ⎞
DCr 3+ = ⎡⎣VCr 3+ ⎤⎦ γΛ 2 υ ⋅ exp ⎜ −
can migrate. If oxygen ion mobility dominates, ⎝ RT ⎟⎠
then the scale would continue to grow at the
⎛ ΔH m ⎞
oxide-metal interface, whereas if chromium ion = ⎡⎣VCr 3+ ⎤⎦ ⋅ cons tan t ⋅ exp ⎜ − (Eq 5)
mobility dominates, the oxide will grow at the ⎝ RT ⎟⎠
Fig. 4 Metal with oxide scale. (a) A protective scale that prevents gas access. (b) Schematic of electrochemical oxidation through a
protective oxide scale that serves as electrolyte and electron lead. The case is for mobile cations
3FDCr 3+
effects are ignored (i.e., scales are relatively
3+ =
e
BCr (Eq 6) thick compared to range of space charge ef-
RT fects). As a case study, let us assume that the
mobile ion defect is cations due to metal vacan-
where zi is the ion charge, F is Faraday’s con- cies in the scale.
stant (96,457 C.eq–1). Compared to wustite The molar flux J (moles/m.s) of a particle i in
(FeO), the equilibrium constant of Eq 4 is quite an electrolyte subjected to an electrochemical
low, resulting in a low degree of nonstoichiome- potential gradient was shown to be:
try in Cr2–δO3 compared to Fe1–δO (where δ can
be as large as 0.05), and thus the transport of
Ji = − ci Bi
∂ ( μ i + zi Fφ ) (Eq 7)
Cr3+ through its scale is much slower than the ∂x
transport of Fe2+ through FeO and thus the dif-
ference in parabolic rate constants in Fig. 3.
where zi is the ion charge, F is Faraday’s con-
stant (96,457 C/gram equivalent), and φ is the
electric field (V). The electronic or ionic con-
Kinetics and Oxidation Rates: Wagner’s ductivity κ in an electrolyte can be computed
Theory through:
κ = F 2 ∑ zi Bi ci
2
The parabolic oxidation rate was introduced (Eq 8)
without explanation in Eq 3. It was first de-
scribed in terms of oxide defect structure and re- The contribution of a given ion specie type or
sulting transport properties by Wagner (Ref 6), electron defect type to this conductivity is de-
and the theory is explained in most of the mono- noted as the partial conductivity and computed
graphs on oxidation, such as Chapter 3 in Ref 3 as:
and Chapter 4 in Ref 1. This treatment follows
the derivation in Ref 1. Consider a general case, 2
as shown in Fig. 4 under the assumptions that κ i = F 2 zi Bi ci (Eq 9)
(a) the scale is compact and adherent, (b) elec-
trode reactions are rapid enough to be in equi- Inserting Eq 9 into Eq 7 yields:
librium at the interface and surface, (c) nonstoi-
chiometry is small and uniform throughout the
Ji = −
κi ∂ ( μ i + zi Fφ )
scale (i.e., defects are in thermal equilibrium (Eq 10)
2
F zi
2
∂x
throughout the scale), and (d) double-layer
Now, if the mobile particles are a single type assumed, that is, a linear drop across X, at all
of metal cations (e.g., Cr3+) and electrons, then times:
two fluxes are present:
dx k
Jc = −
κc ∂ ( μ c + zc Fφ ) =
dt X (t ) (Eq 18)
2 2
∂x (Eq 11)
F zc
The constant k is the parabolic rate constant.
and A mass balance can be written where the flux of
cations for a period of time dt is equated to the
Je = −
κe ∂ ( μ e + ze Fφ ) amount of metal being accumulated as cations
2 2
∂x (E q 12) inside the scale of thickness dx:
F ze
Jc ⋅ dt = CC dx (Eq 19)
Electrical neutrality requires that:
Jc Z c + Je Z e = 0 By combining Eq 18 and 19 and inserting Eq

(Eq 13)
16 for the flux, an expression for the parabolic
rate constant is obtained:
And at the oxide-scale/metal interface, the
anode reaction is in equilibrium, that is: μ 'M
z + 1 κ cκ e
M = M + zce − and therefore,
c
k= 2 2 ∫κ dμ (Eq 20)
zc F CC μ ''M c
+ κe M
μ M = μ c + zc μ e (Eq 14) If the mobility and thus partial conductivity

of electrons is significantly higher than that of
Combining Eq 11 to 14, the potential gradient the ions (a reasonable assumption), then Eq 20
is eliminated, and the chemical potential gradi- can be simplified as:
ents can be replaced by the metal (M) potential
gradient, and the following equation results: μ 'M
1
κ cκ e ∂μ M
k= 2 2
zc F CC ∫ κ dμ
μM
c M (Eq 21)
Jc = −
zc2 F 2 ( κ c + κ e ) ∂x (Eq 15)
Since from diffusion theory we know that
To obtain an explicit function for the cation Dc = BcRT, and inserting this in Eq 9, one
flux, Eq 15 needs to be integrated after variable obtains
separation, keeping in mind that conductivities
and metal chemical potential may vary within 2 Dc
κ c = F 2 zc c
the scale. Integrating from the gas-scale surface RT c
(x = 0, μM = μ′′M) to the scale-metal interface
(x = X, μM = μ′′M) one obtains. and inserting this into Eq 21 results in:
μ" μ 'M
1 M
κκ 1
Jc = − 2 2 ∫ c e dμ M (Eq 16) k= ∫ D dμ (Eq 22)
zc F x μ ' c + κ e
κ RT μ ''M
c M
M
Now, if the growth of the oxide scale is con- It is often more convenient to express Eq 22
trolled by the flux of cations: in terms of oxygen potentials rather than the
metal potentials. It was assumed at the onset of
dx ∂C this analysis that the deviation from stoichiom-
∝ JC = − DC C (Eq 17)
dt ∂x etry is small and constant throughout the scale.
Therefore, the oxide potential is constant and:
The concentration drop across the scale is
constant since interface and surface reactions zc
μ + μ M = μ MO = constant (Eq 23)
are at equilibrium. If quasi steady state is 4 O2 zC / 2
And thus, dμ O = dRT ln PO ∝ − dμ Me. Therefore, 2

thickness loss due to evaporation, which is de-
2
scribed by a first-order reaction kinetics expres-
Eq 22 can be written:
sion with rate constant ke. Thus, the thickness
ln PO'' PO'' change becomes:
2 2
Dc
k∝ ∫ Dc d ln PO =
2 ∫ dP
PO O2 (Eq 24) dx k
ln PO'
2
PO'
2
2 = −k (Eq 27)
dt X (t ) e
In the case of Cr2O3, if bulk diffusion is dom-

inating, Eq 4 and 5 inserted into the diffusion This results in a so-called paralinear (as op-
coefficient in Eq 24 for C = Cr3+, results in: posed to parabolic) rate for the oxide thickening
(Fig. 6), and at a critical oxide scale thickness
X( the rate of thinning due to evaporation equals
PO'' PO''
2
P 3/ 4 2 the rate of thickening due to oxidation. In Eq
k∝ ∫ PO
dPO = ∫ P −1/ 4 dPO
2 2
27, this means that dx/dt = 0; consequently, X =
PO'
2
2 P' O2 k/ke. While at first this seems to suggest that it
( ) − (P )
3 ⎡ '' 3/ 4 3/ 4
⎤ does not affect the oxidation process in that the
= P '
(Eq 25)
4 ⎢⎣ O2 O2 ⎥⎦ rate of oxidation does not increase, the forma-
tion and evaporation of chromium oxides re-
sults in greater chromium consumption in the
Thus, the parabolic rate constant would be pre- alloy compared to what would be the case if
dicted to vary with the power of three-quarters of evaporation did not occur. As a result of evapo-
the external oxygen partial pressure. ration losses, stainless steels that depend on a
Grain boundary diffusion has however been protective chromium oxide layer are limited in
identified to be important in the case of Cr3+ use to temperatures up to 900 to 1000 °C.
transport (Ref 8). The observed growth rate of The presence of water vapor promotes the
Cr2O3 polycrystalline films is far too fast to be formation of even more volatile oxyhydroxides
accounted for by bulk diffusion of chromium (e.g., CrO2(OH)2 ) (Ref 11, 12).
ions; instead, grain boundary diffusion would
be expected to dominate (Ref 9).
Spalling and Cracking of the Scale
The Volatile Nature of Cr2O3 At elevated temperatures or during tempera-

ture cycling, there are multiple ways in which
At high enough temperatures and high stresses can develop that may crack and blister
enough oxygen partial pressures, the formation the scale, rendering it nonprotective. The differ-
of a gaseous hexavalent chromium oxide CrO3* ent causes of stress generation are described in
could lead to thinning of the Cr2O3 scale ac- Chapter 5 in Ref 1. So-called growth stresses
cording to the following reaction: arise due to changes caused by the oxidation
process itself. These include differences in lat-
3 tice mismatch, alloy depletion in the metal,
Cr2 O3 (s) + O 2 ( g) = 2CrO3 ( g) (Eq 26) point-defect gradients in scales containing ox-
2
ides such as FeO, with large deviation from sto-
Figure 5 shows the vapor pressure of the su- ichiometry, recrystallization, and volume differ-
peroxide as a function of temperature and par- ences between the oxide and metal. The last is
tial pressure of O2. perhaps the most commonly mentioned and is
The effect of this reaction on the oxidation ki- characterized by the Pilling-Bedworth ratio, ab-
netics can be described as follows: The thick- breviated as PBR (Ref 13).
ness change described through the parabolic
VOxide VmOxide
rate constant in Eq 18 is corrected for by the PBR = = Metal (Eq 28)
VMetal Vm * ν
* Hexavalent chromium is now considered a human car-

cinogen and is rigorously regulated by both the Occupa- Here, the subscript m stands for molar volume,
tional Safety and Health Administration (OSHA) and the and υ is the number of metal atoms needed to
U.S. Environmental Protection Agency (EPA). form a stoichiometric unit of the oxide (in the
Fig. 5 Chromium-oxygen system species volatility as a function of temperature and oxygen pressure. Source: Ref 10
case of Cr2O3, υ is 2, and in the case of FeO, it

is 1). When PBR is greater than 1, then the
oxide is expected to be in compression and is
likely to be protective, whereas if it is less than
1, the oxide is in tension and thus nonprotective.
There are, however, many exceptions to this,
partly because the stress state often depends
more on the mechanisms and conditions of the
oxidation process rather than the properties of
metal and oxides.
Thermal, stresses are caused by differences in
thermal expansion between the oxide and metal,
and the stresses generated in oxide scales can be
estimated: Fig. 6 Schematic of paralinear oxidation as a result of evapo-
ration of chromium superoxide
σ Oxide =
)
− EOxide ( α Oxide − α Metal ΔT
ity, t is thickness, and ΔT is the temperature
(Eq 29)
⎛ ⎞
( ) t E
1 − ν p ⎜ 1 + 2 Oxide Oxide ⎟
⎝ t Metal E Metal ⎠
change. In general, α is larger for the metal than
the oxide; thus, during cooling the stresses are
expected to be compressive and during heating
The equation is written for a case shown tensile. Thermal stresses can cause spalling of
schematically in Fig. 7, where both sides on a the protective oxide layer, and it is most severe
metal undergo oxidation. Here, σ is the stress, under cyclic conditions.
υp is Poisson’s ratio (it has been assumed that Effect of Silicon, Aluminum, and Molyb-
there is no mismatch), α is the coefficient of denum. Due to concerns about the cost of
thermal expansion, E is the modulus of elastic- chromium and its (former) classification as a
internal Cr2O3 layer forms and thickens. Even-

tually, the surface aluminum-oxide layer flakes
off. Also, due to the low oxygen potential
needed to form Al2O3, internal oxidation may
result below the metal-scale interface in alloys
tox tm tox in which formation of a continuous alumina
scale film does not occur.
Molybdenum is suggested to strain the lat-
tice due to its larger size and consequently in-
Oxide Metal Oxide crease the rate of bulk diffusion of elements
(Ref 11), which can enhance the rate of initial
Cr2O3 formation. Molybdenum is usually con-
sidered detrimental for oxidation resistance.
Molybdenum normally forms MoO2 oxide, but
Fig. 7 Schematic of a cross section of oxidized sample indi- this can oxidize further to form the low-melt-
cating dimensions in Eq 29 for predicting thermal ing and volatile MoO3. If the MoO3 evapo-
stresses
rates, there is little problem, but if its
volatilization is inhibited by low atmosphere
circulation, liquid MoO3 can accumulate and
strategic material, there were efforts to try to dissolve the protective Cr2O3 scale, leading to
substitute less-expensive elements such as alu- catastrophic oxidation.
minum and silicon that also are known to form Effect of Rare Earth Additions. Cerium, lan-
protective layers (Ref 14), even though they thanum, and yttrium additions are known to im-
have significant metallurgical and mechanical prove oxidation resistance of high temperature
drawbacks. nickel- and iron-based alloys (Ref 6). Rare earth
Silicon additions of 1.5 wt% or more have the additions have been suggested to have a multi-
effect of forming a continuous amorphous sub- tude of beneficial effects, such as reducing the
surface layer in iron-silicon and Fe-Cr-Si alloys growth kinetics of Cr2O3 scales, stabilizing
that is relatively impervious to transport of ions. Cr2O3 scales at lower chromium levels, increas-
The mechanism for the evolution of such a layer ing adhesion, and preventing spalling of the
is as follows: (1) The more readily available oxide scale during thermal cycling. The expla-
iron or chromium first forms a surface layer, nation for any of this does not seem clear, but
and this causes an enrichment of silicon at the some hypotheses have been suggested. The ef-
oxide-metal interface. (2) As sufficient silicon is fect on the growth kinetics could be because the
enriched, the SiO2 layer is formed. It has been reactive element ions collect at grain boundaries
reported that alloys with chromium content as and block fast path diffusion.
low as 6 wt% and silicon content of 1.5 wt% The improved adherence could be because
perform in terms of oxidation as well as com- these elements getter tramp elements such as
mercial stainless steels. Also, an addition of 4 sulfur and suppress void formation at the inter-
wt% Si to a Fe14wt%Cr14wt%Ni alloy resulted face. Furthermore, they might form so-called
in a 200-fold reduction in weight gain at 900 °C oxide pegs at the interface (Ref 16). The pre-
However, this SiO2 layer seems to promote cise role of the rare earth additions to Cr2O3
oxide spalling, especially in cyclic service. oxide protection and the mechanism by which
Aluminum forms a very stable thin outer they are incorporated into the scale during the
layer of Al2O3 that initially reduces the oxida- surface treatment processes remain unknown.
tion rate. Alumina is among the most stable and An understanding of these fundamental issues
defect-free oxides, giving it an extremely low would help to develop optimum alloy
diffusion rate. In consequence, if sufficient alu- chemistries for selected high-temperature and -
minum is present to maintain the protective alu- pressure applications and to further develop the
mina scale, the aluminum-bearing alloys pro- surface infusion process. The lack of funda-
vide the greatest oxidation resistance attainable mental understanding of how rare earths im-
in engineering alloys. If, however, the alu- prove oxidation resistance has not stopped the
minum content is not sufficient to force alumina development of several alloys that benefit from
scale formation at the scale-metal interface, an the effect.
Oxidation Under Less-Oxidizing quasi-steady-state situation as shown in Fig. 9,

Atmospheres the flux can be written:
dm N X − N OS NS
The two types of alloy oxidation behaviors, JO = = − DO O = DO O (Eq 30)
(a) a noble matrix metal with less-noble alloy- dt Vm X Vm X
ing elements and (b) both matrix element and
alloying elements are nonnoble, were men- Within the depth X, the oxygen solubility is
tioned. When designing against oxidizing envi- such that Eq 2 is negative enough that Cr2O3
ronments, case b is perhaps the most relevant, forms. Beyond X, it is not. At the distance X, the
and most of the discussion has been devoted to oxygen concentration is negligibly low com-
this. However, during annealing for microstruc- pared to the surface composition, which is in
tural control, steels are exposed to furnace gases equilibrium with the gas phase. The molar vol-
at high temperatures that have relatively low ume Vm is used to obtain the flux in units of
oxygen or steam contents, for which case a will moles per square meter. Within the layer 0 < x <
apply, that is the atmosphere does not cause iron X, all the chromium is assumed to be oxidized;
(or nickel) to oxidize but chromium (and alu- therefore, the accumulated mass due to oxygen
minum, silicon, molybdenum, etc.) does. addition:
For simplicity, assume a binary system A-B
of “noble” iron and “reactive” chromium. In o 3
this case, depending on the concentration of N B0 υX N Cr 2 X

m= = (Eq 31)
the reactive element and atmosphere, the Vm Vm
oxide of the reactive element can, in principle,
form either on the surface (as has been dis- Differentiating Eq 31 with time, equating to
cussed so far) or internally as discrete oxide Eq 3, and separating variables results in:
particles in a metal matrix through oxygen
diffusion into the metal. Both cases are shown N OS DO N OS DO
schematically in Fig. 8. Let us discuss the XdX = dt = (Eq 32)
υN BO 3 0
conditions that promote one or the other of N
2 Cr
these by starting with a situation in which (a)
no surface oxide exists and (b) the oxygen at-
mosphere is such that the solubility of oxygen Integrating Eq 32 from x = 0 to x = X results in
within a distance X is enough to thermodynam- an expression for the internal oxidation depth X:
ically render Cr2O3 stable according to Eq 2 1/ 2
but none of the iron oxides. The derivation is 1/ 2
⎛ ⎞
⎛ 2N S D ⎞ ⎜ 2N S D ⎟
done in terms of both a generic system A-B X =⎜ O0O t⎟ =⎜ O O t⎟ (Eq 33)
causing an oxide BOν and for iron-chromium ⎝ υN B ⎠ ⎜ 3 N Cr
0
⎟
causing Cr2O3. Assume for a start that the DO ⎝ 2 ⎠
>> DCr, and thus while oxygen diffuses into
the alloy, chromium does not counterdiffuse.
The flux of oxygen inward into the metal is
then the cause of increased mass. Assuming a
Fig. 8 Schematic of two cases in a less-oxidizing atmosphere. Fig. 9 Quasi-steady-state approximation of the moving
(a) Adsorption of oxygen leading to internal oxidation boundary problem of internal oxidation. Counterdif-
and (b) external oxidation as the B element migrates. fusion of B is assumed to be negligible
This expression predicts a parabolic depend-

ence of X with time, just as Eq 3 did for the ex-
ternal oxidation. For a more rigorous derivation
(without assuming quasi steady state), refer to
Appendix B in Ref 1. Now, let us see what
causes this to transition into an external scale. It
was assumed in the derivation of Eq 33 that
counterdiffusion of chromium does not occur.
When considering Eq 33, is clear that the rate of
penetration of the internal oxidation front will
decrease with (a) increasing NB0, (b) decreasing
NOS, and (c) decreasing DO. If DCr was not negli-
gible, there would be a gradual change in oxida-
tion morphology if the ratio (NB0DB)/(NOSDO).
Gradually, if the ratio were increased there
would be a slowing of the inward penetration of
the internal oxide front and an enrichment of Fig. 10 Temperature dependence of metal dusting of iron.
BOυ in the internally oxidized zone. Wagner de- Source: Ref 18
veloped a model (Ref 16), based on that at some
point, when the volume fraction of BOυ versus trogen absorption, but if the oxygen is depleted
volume metal in the internally oxidized zone before all surfaces are oxidized, the remaining
reaches a critical value g*, there is a transition material can be rapidly nitrided by the residual,
from internal to external oxidation; specifically, essentially pure, nitrogen atmosphere.
this happens when:
1/ 2
⎡ πg* S DOVM ⎤ REFERENCES
N >⎢
0
NO ⎥ (Eq 34)
⎣ 2ν
B
DBVOx ⎦ 1. N. Birks, G.H. Meier, and F.S. Pettit, Intro-
duction to the High-Temperature Oxidation
If more than one reactive elements were pres- of Metals, 2nd ed., 2006, Cambridge Uni-
ent (such as is the case in stainless steels in versity Press, New York
which aluminum, silicon, molybdenum, nio- 2. P. Kofstad, High Temperature Corrosion,
bium, etc., may be present), this will decrease 1988, Elsevier Applied Science, London
the inward flux of oxygen, and thus the transi- 3. K. Hauffe, Oxidation of Metals, Plenum
tion to external may occur at a lower solute Press, New York, 1965
(CCr0) concentration than what is predicted by 4. F.D. Richardson and J.H.E. Jeffes, J. Iron
Eq 34. Steel Inst., Vol 160, 1948, p 261
Metal Dusting. Under reducing conditions, 5. C. Wagner and K. Grünewald, Z. Phys.
in products of combustion atmospheres, oxida- Chem., Vol 40B, 1938, p 455
tion and carburization may occur simultane- 6. J.H. Park, W.E. King, N.L. Peterson, and
ously and at a higher rate than exhibited in pure S.J. Rothman, The Effect of Reactive Ele-
oxidation. Under even more reducing condi- ment on Self-Diffusion in Cr2O3, Norman
tions, the condition called metal dusting may L. Peterson Memorial Symposium, Oxida-
occur. Metal dusting is often characterized by tion of Metals and Associated Mass Trans-
the generation of large, smooth pits that look as port, edited by M.A. Dayananda, S.I. Roth-
if metal had been scooped from the surface. The man, and W.E. King, AIME, Warrendale,
underlying phenomenon is the formation of PA, 1998, p 103–107
metal carbides, which manifests itself as the 7. C. Wagner and K. Grünewald, Z. Phys.
breakup of bulk metal to metal powder. This oc- Chem., Vol 40B, 1938, p 455
curs at temperatures at which the carbide is 8. D. Caplan and G.I. Sproule, Effect of Oxide
most stable (Fig. 10). Grain Structure on the High Temperature
During oxidation in air, if large surface area is Oxidation of Cr, Oxid. Met., Vol 9, 1975,
present under conditions of restricted air supply, p 459–472
oxygen can be depleted to the point that oxida- 9. B.B. Ebbinghaus, Combust. Flame, Vol 93,
tion essentially ceases. Oxide films inhibit ni- 1993, p 119–137
10. K. Hilpert et al., JECS, Vol 143/11, 1996, p of Metals and Associated Mass Transport,
3642–3647 edited by M.A. Dayananda, S.I. Rothman,
11. N.B. Pilling and R.E. Bedworth, J. Inst. and W.E. King, AIME, Warrendale, PA,
Met., Vol 29, 1923, p 529 1998, p 323–340
12. J.K. Tien and J.M. Davidson, Oxide Spalla- 14. E.J. Felten, J. Electrochem. Soc., Vol 108,
tion Mechanisms, Stress Effects and the Ox- 1961, p 490
idation of Metals, ed. J.V. Cathcart, AIME, 15. C. Wagner, J. Electrochem. Soc., Vol 103,
New York, 1975, p 200 1956, p 571
13. J. Rawers, Understanding the Oxidation 16. C.M. Chun, J.D. Mumford, and T.A. Rama-
Protection of Fe-Cr-Si Alloys, Norman L. narayanan, J. Electrochem. Soc., Vol 147,
Peterson Memorial Symposium, Oxidation 2000, p 3680
CHAPTER 6
Austenitic Stainless Steels
Summary because their greater thermal expansion co-

efficient tends to cause the protective oxide
AUSTENITIC STAINLESS STEELS are the coating to spall.
most common and familiar types of stainless 2. They can experience stress corrosion crack-
steel. They are most easily recognized as non- ing (SCC) if used in an environment to which
magnetic. They are extremely formable and they have insufficient corrosion resistance.
weldable, and they can be successfully used 3. The fatigue endurance limit is only about
from cryogenic temperatures to the red-hot tem- 30% of the tensile strength (vs. ~50 to 60%
peratures of furnaces and jet engines. They con- for ferritic stainless steels). This, combined
tain between about 16 and 25% chromium, and with their high thermal expansion coeffi-
they can also contain nitrogen in solution, both cients, makes them especially susceptible to
of which contribute to their high corrosion re- thermal fatigue.
sistance. Were it not for the cost of the nickel However, the risks of these limitations can be
that helps stabilize their austenitic structure, avoidable by taking proper precautions.
these alloys would be used even more widely.
Introduction Alloy Families in Perspective

Austenitic stainless steels have many advan- The fundamental criterion in the selection of
tages from a metallurgical point of view. They a stainless steel is generally that it can survive
can be made soft enough (i.e., with a yield with virtually no corrosion in the environment
strength about 200 MPa) to be easily formed by in which it is to be used. Good engineering
the same tools that work with carbon steel, but practice sometimes requires that materials be
they can also be made incredibly strong by cold selected for sufficient, but finite, service life.
work, up to yield strengths of over 2000 MPa This is especially true for high-temperature
(290 ksi). Their austenitic (fcc, face-centered service, for which creep and oxidation lead to
cubic) structure is very tough and ductile down limited life for all materials. The choice among
to absolute zero. They also do not lose their the stainless steels that can be used in that envi-
strength at elevated temperatures as rapidly as ronment is then based on the alloy from which
ferritic (bcc, body-centered cubic) iron base al- the component can be produced at the lowest
loys. The least corrosion-resistant versions can cost, including maintenance, over the intended
withstand the normal corrosive attack of the service life. The ferritic stainless steels are less
everyday environment that people experience, expensive for the same corrosion resistance but
while the most corrosion-resistant grades can sometimes are found lacking because of:
even withstand boiling seawater.
• Lack of toughness, as is the case at subambi-
If these alloys were to have any relative
weaknesses, they would be: ent temperatures or in thicknesses greater
than about 1.5 mm
1. Austenitic stainless steels are less resistant • Lack of great ductility, specifically if more
to cyclic oxidation than are ferritic grades than about 30% elongation is needed
• Susceptibility to high-temperature embrit- The austenitic alloys can have compositions

tling phases when moderately alloyed anywhere in the portion of the Delong diagram
labeled austenite shown in Fig. 1 (Ref 1). This
The less-expensive martensitic grades are diagram was designed to show which phases are
used instead of austenitic when high strength present in alloys in the as-solidified condition,
and hardness are better achieved by heat treat- such as found in welds. Thus it also applies to
ing rather than by cold work, and mechanical castings and continuously cast products. As a
properties are more important than corrosion re- practical matter of castability, the composition
sistance. This is also the case for the more ex- of most commercial alloys falls along the zone
pensive PH grades, which can achieve corrosion of several percent ferrite as cast. The salient fea-
resistance only matching the least corrosion re- ture of austenitic alloys is that as chromium and
sistant of the austenitic alloys. molybdenum are increased to increase specific
Duplex grades match austenitic grades in cor- properties, usually corrosion resistance, nickel
rosion resistance and have higher strength in the or other austenite stabilizers must be added if
annealed condition but present the designer the austenitic structure is to be preserved.
with challenges with regard to embrittling The traditional way of displaying the
phases that can form with prolonged exposure austenitic stainless steels is to present 302 as a
to elevated temperatures and only moderate base. Figure 2 shows one such diagram. Dia-
ductility like the ferritic alloys. grams such as these treat alloys as an evolution-
So, the austenitic grades are the most com- ary family tree and subtly mislead. Many alloys
monly used grades of stainless mainly because, were pushed toward obsolescence because of
in many instances, they provide very predictable advances in processing. For instance, 321 was
levels of corrosion resistance with excellent me- developed as an alloy in which the detrimental
chanical properties. Using them wisely can save effects of carbon were negated by addition of ti-
the design engineer significant costs in his or her tanium. The widespread adoption of the argon
product. They are a user-friendly metal alloy oxygen decarburization (AOD) in the 1970s
with life-cycle cost of fully manufactured prod- made this alloy unnecessary, except for special
ucts lower than many other materials. circumstances, since carbon could be cheaply
Fig. 1 Schaeffler-Delong stainless steels constitution diagram. Adapted from Ref 1, 2

Chapter 6: Austenitic Stainless Steels / 71
Fig. 2 The austenitic stainless family. Source: Ref 3
removed routinely. Likewise, 302 gave way to nickel equivalents (manganese, nitrogen, carbon,
the lower-carbon 304, for which the even lower- etc.) must also be added in matching amounts,
carbon 304L is commonly substituted and du- austenite stability is also increased. If molybde-
ally certified to qualify as either grade. While num, a chromium equivalent, is added, corro-
low carbon prevents sensitization, stabilized sion resistance but not oxidation resistance is
grades may still be preferred for special applica- enhanced. And, if nitrogen is the austenite stabi-
tions such as type 321 in aerospace and type lizer added to balance increases chromium or
347 in refinery service. Similar inertia keeps the molybdenum, then corrosion resistance is also
higher-nickel 300 series as the de facto standard increased. With small exceptions, that is the
when the more cost-efficient high-manganese rationale of austenitic grade design. Silicon
200 series is the logical basic grade. The rele- is used as an alloy to promote oxidation resist-
vant types of austenitic alloys can nonetheless ance and resistance to corrosion by oxidizing
be rationalized with this diagram. acids. Copper is used to promote resistance to
As chromium is added, oxidation resistance sulfuric acid. Rare earths make a more stably
and corrosion resistance increase. Because oxidation-resisting scale. Niobium increases
creep resistance. Sulfur and selenium increase Lean Alloys

machinability.
In this chapter, austenitic alloys are classified Lean austenitic alloys constitute the largest
into three groups: portion of all stainless steel produced. These are
• Lean alloys, such as 201 and 301, are gener- principally 201, 301, and 304. Alloys with less
ally used when high strength or high forma- than 20% chromium and 14% nickel fall into
bility is the main objective since the lower, this unofficial category. Since they are stainless,
yet tailorable, austenite stability of these al- it is generally taken for granted that these alloys
loys gives a great range of work-hardening will not corrode, and these alloys have suffi-
rates and great ductility. Richer alloys, such cient corrosion resistance to be used in any in-
as 305, with minimal work hardening are the door or outdoor environment, excluding coastal.
high-alloy, lowest work-hardening rate These grades are easily weldable and formable
grade for this purpose. The general-purpose and can be given many attractive and useful sur-
alloy 304 is within this group. face finishes, so they are very much general-
• Chromium nickel alloys when the objective purpose alloys. Table 1 lists some typical com-
is high temperature oxidation resistance. positions of the most commonly used lean
This can be enhanced by silicon and rare austenitic alloys. These typical compositions
earths. If the application requires high-tem- vary with end use, raw material cost factors, and
perature strength, carbon, nitrogen, niobium, the preference of a given manufacturer. The
and molybdenum can be added. 302B, 309, compositions of standard alloys are often fine-
310, 347, and various proprietary alloys are tuned to the intended end use. In this table, the
found in this group. word drawing indicates higher nickel for lower
• Chromium, molybdenum, nickel, and nitro- work hardening, while tubing indicates alloys
gen alloys when corrosion resistance is the with higher sulfur to facilitate gas tungsten arc
main objective. Alloys such as silicon and welding (GTAW) penetration. Tensile indicates
copper are added for resistance to specific lower alloy levels to increase the work-harden-
environments. This group includes 316L, ing rate for material that is intended to be used
317L, 904L, and many proprietary grades. in the cold-worked, high-strength condition.
316L is included in its most common tubing end
Wrought alloys generally have cast counter- use chemistry even though it is a corrosion-re-
parts that differ primarily in silicon content. sisting alloy because it is so pervasively used as
Versions that require enhanced machinability a service center sheet item.
have a high content of controlled inclusions, The main difference among the lean
sulfides, or oxysulfides, which improve machin- austenitic alloys lies in their work-hardening
ability at the expense of corrosion resistance. rate: the leaner the alloy, the lower the austenite
Carbon is kept below 0.03% and designated an stability. As unstable alloys are deformed, they
L grade when prolonged heating due to multi- transform from austenite to the much harder
pass welding of heavy section (greater than martensite. This increases the work-hardening
about 2 mm) or when welds requiring a post- rate and enhances ductility since it delays the
weld stress relief are anticipated. onset of necking since greater localized
Table 1 Typical compositions of the most commonly used lean austenitic alloys
Alloy Designation C N Cr Ni Mo Mn Si Other Other Other
201 S20100 0.08 0.07 16.3 4.5 0.2 7.1 0.45 0.001 S 0.03 P 0.2 Cu
201 drawing S220100 0.08 0.07 16.9 5.4 0.02 7.1 0.5 0.001 S 0.30 P 0.6 Cu
201LN S20153 0.02 0.13 16.3 4.5 0.2 7.1 0.45 0.001 S 0.03 P 0.5 Cu
301 tensile S30100 0.08 0.4 16.6 6.8 0.2 1.0 0.45 0.001 S 0.03 P 0.3 Cu
301 drawing S30100 0.08 0.04 17.4 7.4 0.02 1.7 0.45 0.007 S 0.03 P 0.6 Cu
303 S30300 ... ... ... ... ... ... ... ... ... ...
304 S30400 0.05 0.05 18.3 8.1 0.3 1.8 0.45 0.001 S 0.03 P 0.3 Cu
304 drawing S30400 0.05 0.04 18.4 8.6 0.3 1.8 0.45 0.001 S 0.03 P 0.3 Cu
304 extra drawing S30400 0.06 0.04 18.3 9.1 0.3 1.8 0.45 0.001 S 0.030 P 0.4 Cu
304L tubing S30403 0.02 0.09 18.3 8.1 0.3 1.8 0.45 0.013 S 0.030 P 0.4 Ci
305 S30500 0.05 0.02 18.8 12.1 0.2 0.8 0.60 0.001 S 0.02 P 0.2 Cu
321 S32100 0.05 0.01 17.7 9.1 0.03 1.0 0.45 0.001 S 0.03 P 0.4 Ti
316L S31603 0.02 0.0 16.4 10.5 2.1 1.8 0.50 0.010 S 0.03 P 0.4 Cu
deformation is more than offset by greater lo- such as occurs when they are sensitized or when
calized strain hardening. solute segregation occurs, as from welding, then
These grades are best viewed as a continuum the equation applies on a microscopic scale.
with a lower boundary at 16%Cr-6%Ni and an Sensitized zones (i.e., the regions near grain
upper boundary at 19%Cr-12%Ni. This repre- boundaries where chromium carbides have pre-
sents the range from minimum to maximum cipitated) will have a much higher tendency to
austenite stability. Since that is the main distinc- transform to martensite. Figures 3(a) and (b)
tion within this grade family, let us examine its show the changes in phase structure as a func-
basis. tion of composition over ranges that encompass
Martensite and Austenite. Stability. The these alloys.
formation of martensite at room temperature Martensite can be present in two different
may be thermodynamically possible, but the forms. The α′-form is the bcc magnetic form,
driving force for its formation may be insuffi- while ε is a nonmagnetic, hcp (hexagonal close-
cient for it to form spontaneously. However, packed) version. The formation of ε versus α′ is
since martensite forms from unstable austenite related to the stacking fault energy of the alloy,
by a diffusionless shear mechanism, it can occur which is given by (Ref 6):
if that shear is provided mechanically by exter-
nal forces. This happens during deformation, Y300SF (mJ m-2) = Y0SF + 1.59Ni – 1.34Mn
and the degree to which it occurs varies with + 0.06Mn2 – 1.75Cr + 0.01Cr2
composition according to (Ref 4): + 15.21Mo – 5.59Si
Md30 (°C) = 551 – 462(%C + %N) –60.69(C + 1.2N)1/2
– 9.2(%Si) – 8.1(%Mn) – 13.7(%Cr) + 26.27(C + 1.2N)
– 29(%Ni + Cu) – 18.5(%Mo) (Cr + Mn + Mo)1/2
– 68(%Nb) – 1.42 (GS – 8) (Eq 1) + 0.61[Ni•(Cr + Mn)]1/2 (Eq 2)
This is the temperature at which 50% of the Epsilon martensite formation is favored in
austenite transforms to martensite with 30% alloys of lower stacking fault energy. The fcc
true strain (Ref 5). It should be noted that even structures deform by slip between (111)
elements that are chromium equivalents in pro- planes. Viewed from these planes, the structure
moting ferrite are austenite stabilizers in that is a series of ABCABC atom arrangements.
they impede martensite formation. This temper- Slip between planes can result in an
ature is the common index of austenite stability. ABCA/CAB structure. This so-called stacking
This regression analysis was generated for ho- fault generates an hcp structure. With lower
mogeneous alloys. If alloys are inhomogeneous, stacking fault energies, these are more readily
Fig. 3 (a) Iron-chromium phase diagram at 8% nickel; (b) iron-nickel phase diagram at 18% chromium
Fig. 4 Variation of martensite formation with temperature and true strain for 304. Source: Ref 7
formed, and ε predominates. The stacking fault and tensile strength, respectively, are reported
can also be viewed as two partial dislocations (Ref 10) to follow the equations:
with the material between them faulted. These
partial dislocations, when generated in abun- YS( MPa) = 15.4[4.4 + 23(%C) + 32(%N)
dance, cannot readily slip past one another + 0.24(%Cr ) + 0.94(%Mo)
and thus pile up, increasing work-hardening + 1.3(%Si ) + 1.2(%V)
rates. + 0.29(%W ) + 2.6(%Nb)
As in carbon and alloy steels, the martensite
+ 1.7(%Ti ) + 0.82(%Al )
transformation can take place simply by cool-
ing, but in the lean austenitic alloys the temper- + 0.16(%Ferrite )
atures are well below ambient. The more stable + 0.46(d −1/1/ 2 )
(Eq 3)
alloys do not transform even with cryogenic
treatment. Figure 4 shows the variation of TS ( MPa) = 15.4[29 + 35(%C) + 55(%N)
martensite formation with temperature and true
+ 2.4(%Si ) + 0.11((%Ni ) + 1.2(%Mo)
strain for 304. Martensite formed in these alloys
is quite stable and does not revert until heated + 5.0(%Nb) + 3.0(%Ti ) + 1.2(%Al )
well above the temperatures (Fig. 5) at which it + 0.14(%Ferrite ) + 0.82(d −1/ 2 ) (Eq 4)
was formed. The carbon levels of austenitic
stainless steels are always relatively low, so In each case, d is the grain diameter in mil-
strain-induced martensite is self-tempering and limeters.
not brittle. Another researcher (Ref 11) gave the rela-
Martensite has been found to form in unstable tionships as:
austenite due to the electrochemically induced
supersaturation by hydrogen (Ref 9). Under YS ( MPa ) = 120 + 210 N + 0.02 + 2 Mn + 2Cr
conditions of cathodic charging, superficial lay- + 14 Mo + 10 Cu + (6.15 − 0.054δ)δ
ers were found to transform to ε under condi- + (7 + 35(N + 0.2))d −1/ 2 (Eq 5)
tions of intense hydrostatic compression. Dur-
ing subsequent outgassing, α′ was found to
form due to reversals in the stress state. Marten- TS = 470 + 600( N + 0.2)
site thus formed is, of course, susceptible to + 14 Mo + 1.5δ + 8d −1/ 2 (Eq 6)
hydrogen embrittlement.
Mechanical Properties. The tensile proper- Again, d is grain diameter in millimeters, and
ties in the annealed state not surprisingly relate δ is percent ferrite. The claimed accuracy for
well to composition. The 0.2% yield strength the latter set of equations is 20 MPa and is said
Fig. 6 Variation of impact strength with temperature for (a)

austenitic, (b) duplex, and (c) ferritic stainless steels
form more stable dislocation arrays that break

loose at a higher and distinct yield point.
The tensile properties of austenitic stainless
steels with unstable austenite, that is, those with
Md30 temperatures (Eq 1) near room tempera-
ture, are very strain rate dependent. This is sim-
Fig. 5 Reversion of martensite formed by cold work. Source:
ply due to the influence of adiabatic heating
Ref 8
during testing increasing the stability of the
to apply to both austenitics and duplex stainless austenite. Tests run under constant temperature
steels, but clearly the tensile strength relation- conditions, either by slow strain rates or use of
ship must break down for leaner alloys, such as heat sinks, produce lower tensile strengths.
301, in which tensile strength increases with de- Thus, reported tensile strengths should not be
creasing alloy content because of the effect of taken as an absolute value but a result that can
increasing alloying causing less transformation be significantly changed by changes in testing
to martensite, which inarguably produces higher procedure, even with accepted norms and
tensile strengths in austenitic stainless steels. standards.
Equation 3 must also be favored over Eq 5 in Highly cold-worked austenitic stainless
that it accounts for carbon explicitly. steels are often used for their robust mechanical
One other hardening mechanism is possible properties. Few metallic materials can match
in austenitic stainless steels, and that is precipi- the very high strengths they can achieve. Very
tation hardening. Most precipitation-hardening lean 301 can be cold worked to yield strengths
stainless steels are unstable austenite, which is on the order of 2000 MPa because of its unsta-
transformed to martensite before the precipita- ble austenite transforming to martensite. When
tion hardening takes place. One commercial cold worked to lower degrees, it can provide
alloy, A-286, is entirely austenitic and employs very high strength while keeping impressive
the precipitation within the austenite matrix of ductility.
Ni3 (titanium, aluminum) for strengthening. Austenitic stainless steels have exceptional
This is dealt with in a separate section. toughness. The ambient temperature impact
Austenitic stainless steels do not have a clear strength of austenitic stainless steels is quite
yield point but can begin to deform at as little as high. This is not surprising in view of their high
40% of the yield strength. As a rule of thumb, tensile strengths and high elongations. What is
behavior at less than half the yield strength is most remarkable is the absence of a transition
considered fully elastic and stresses below two- temperature, which characterizes ferritic and
thirds of the yield strength produce negligible martensitic materials. Figure 6 shows impact
plastic deformation. This quasi-elastic behavior strength of the various stainless steel types ver-
is a consequence of the many active slip sys- sus temperature. This again is due to the multi-
tems in the fcc structure. Even highly cold- plicity of slip systems in the fcc structure and
worked material exhibits this phenomenon, al- the fact that they do not require thermal activa-
though stress-relieving cold-worked material tion. This makes the austenitic stainless steels,
will cause dislocations to “lock in place” and especially the 200 series, the optimal cryogenic
material, surpassing the 9% nickel martensitic tion in austenite, and diffusion rates are suffi-
steels in cost, toughness, and, of course, corro- cient for carbon and chromium to segregate into
sion resistance. precipitates. The solubility of carbon in austen-
Precipitation of Carbides and Nitrides. ite is over 0.4% at solidification but decreases
Carbon is normally considered as an undesir- greatly with decreasing temperature. The solu-
able impurity in austenitic stainless steel. While bility is given by (Ref 12):
it stabilizes the austenite structure, it has a great
thermodynamic affinity for chromium. Because 6272
log (C ppm ) = 7771 − (Eq 7)
of this affinity, chromium carbides, M23C6, form T (°K )
whenever carbon reaches levels of supersatura-
The equilibrium diagram for carbon in a basic
18%Cr10%Ni alloy is shown in Fig. 7. At room
temperature, very little carbon is soluble in
austenite; even the 0.03% of L grades is mostly
in a supersaturated solution. The absence of car-
bides in austenitic stainless is due to the slow
diffusion of carbon and the even slower diffu-
sion of chromium in austenite. At a carbon level
of 0.06%, which is found in most 304, supersat-
uration is reached below about 850 °C. Below
this temperature, supersaturation increases ex-
ponentially, while diffusion decreases exponen-
tially. This results in precipitation rates that vary
with temperature and carbon level as shown in
Fig. 8. At these temperatures, grain boundary
diffusion is much more rapid than bulk diffu-
sion, and grain boundaries provide excellent nu-
cleation sites, so precipitation occurs along
grain boundaries. And, because carbon diffuses
several orders of magnitude more rapidly than
chromium, carbon diffuses to and combines
Fig. 7 Carbon solubility in 18–10 austenitic stainless. Source: with chromium essentially in situ, depleting the
Ref 13 grain boundaries of chromium in solution.
Fig. 8 The precipitation rates for Cr23C6 as a function of carbon content

Fig. 10 Variation of carbide precipitation locus with time.

Source: Ref 16
Much longer term heat treatment is required to

eliminate these depleted zones by rehomoge-
nization of slowly diffusing chromium than the
short time required to form them. This is very
evident for carbides, but also true for oxides.
Underneath chromium-rich oxide scales is a
layer depleted in chromium and lower in corro-
sion resistance. This is why not only scale from
welding must be removed, but also the underly-
ing chromium-depleted zone.
Other precipitation processes that give rise to
chromium depletion are α and χ and the solid-
state precipitation of oxides, nitrides, and sul-
fides. Chromium precipitates that form in the
liquid alloy do not cause depletion of chromium
Fig. 9 Depletion of chromium from the austenite near grain
locally because no chromium gradients are set
boundaries due to chromium carbide precipitation.
Source: Ref 14 up around them during precipitation as diffu-
sion in the liquid is very rapid. Thus, primary
Figure 9 shows that the local chromium de- carbides, oxides, and sulfides are not per se
pletion is such that the chromium level can be- harmful to corrosion resistance. But, if the same
come low enough that it has not even enough to compounds form and grow in the solid state,
be stainless and certainly much lower corrosion chromium depletion occurs (Ref 15).
resistance than the surrounding area. This zone, Alloying elements can have a major influence
because it is lower in chromium, also has very on carbide precipitation by their influence on the
unstable austenite and is quite prone to marten- solubility of carbon in austenite. Molybdenum
site formation. Figure 10 shows how the locus and nickel accelerate the precipitation by dimin-
of precipitation changes with time and tempera- ishing the solubility of carbon. Chromium and
ture. Carbon relatively far from grain bound- nitrogen increase the solubility of carbon and
aries in the interior of grains remains in super- thus retard and diminish precipitation. Nitrogen
saturation until much longer times and much is especially useful in this regard (Fig. 11).
greater supersaturation since bulk diffusion is Increasing austenite grain size accelerates
required for the nucleation and growth of these precipitation, as does cold work, especially in
precipitates. the interior of grains, where diffusion is en-
The key observation is that any solid-state hanced by increased defect density.
precipitation of a chromium-rich precipitate Nitrogen is much more soluble than carbon
will necessarily cause local chromium depletion and does not give rise to sensitization phenom-
and a resulting loss of corrosion resistance. ena as does carbon even though Cr2N can be a
Fig. 11 Delay in carbide precipitation induced by nitrogen

level. Source: Ref 17
stable phase when the solubility limit is ex-

ceeded. The solubility is over 0.15% in austen-
ite, so its precipitation seldom has the possibil-
ity of occurring, but it does become an issue in Fig. 12 Variation of hardness with depth and therefore car-
ferritic stainless steels in this regard, for which bon content in colossal supersaturation
solubility is much lower. Manganese and
chromium increase the solubility of nitrogen in to ignore the titanium-consuming capacity of
austenite. oxygen and sulfur unless they have been mini-
Stabilization. Before carbon was easily low- mized by refining, which can be done quite
ered to harmless levels, it was found that adding readily.
more powerful carbide formers than chromium Even if sufficient titanium or niobium is pres-
could preclude the precipitation of chromium ent to combine with all carbon, kinetic consid-
carbides. Titanium and niobium are the most erations may result in that not occurring. High
useful elements in this regard. They form car- temperatures, such as encountered in welding,
bides with solubility that follows the following dissociate carbides. If quenched from this state,
equation type: carbon can be free to form Cr23C6 if it is re-
heated to temperatures above 500 °C.
log [ M] [ X] = + A − H (Eq 8) Carbon has always been considered totally
RT undesirable from a corrosion point of view be-
cause of its tendency to form chromium car-
For titanium carbide and niobium carbide, the bides. Recently, however, new processes have
respective solubilities are: been developed to supersaturate carbon in
6780 austenite below the temperatures at which it has
log [Ti ] [C] = 2.97 − (Eq 9) sufficient mobility to form carbides. This so-
T
called colossally supersaturated austenite re-
sults in very high hardness (Fig. 12) and corro-
9350 sion resistance over limited depths. From this,
log [ Nb] [C] = 4.55 − (Eq 10)
T however, we can see that carbon, like nitrogen,
is actually beneficial to corrosion resistance in
Oxides and sulfides are more energetically fa- solid solution, although this is not observed at
vorable than are carbides and nitrides of these normal concentrations. It is possible to see that
metals. Thus, any additions made to form car- if it could be kept in solution it would be appro-
bides must be sufficient to account for the prior priate to give it a factor of around 10 in the pit-
formation of these compounds. Nitrogen also ting resistance equivalent number (PREN)
competes with carbon for available titanium or equation:
niobium. Thus, for successful gettering of all
carbon, there must be sufficient titanium or nio- PREN = %Cr + 3.3(%Mo)
bium to combine stoichiometrically with all
these species present. + 30(%N) + 10(%C) (Eq 11)
This requires in rough terms that titanium ex-
ceed four times the carbon plus nitrogen, or that This is consistent with the similar thermody-
niobium exceed eight times, after accounting namic interaction coefficients that carbon and
for the oxygen and sulfur. It would be a mistake nitrogen share with regard to chromium.
High-Temperature Alloys
The austenitic stainless steels can have an

exceptional combination of strength and corro-
sion resistance at temperatures above 500 °C.
They are often called on to resist attack by
oxygen, sulfur, carburizing, nitriding, halogens,
and molten salts. Austenitic stainless steels are
the most creep resistant of the stainless steels.
Alloying with carbon, nitrogen, and niobium
produces the greatest strength at elevated tem-
peratures. Refer to the properties database for
comparisons among the grades. This discus-
sion concentrates on their resistance to high-
temperature environments, which is their salient
characteristic.
Oxidation Resistance. Resistance to oxida-
tion comes from the protective Cr2O3 scale that
forms on the surface of the material. Above
about 18% chromium, a continuous scale forms.
The scale acts as a barrier to oxygen and greatly
slows further oxidation of metal below the
scale. Below the composition at which com-
plete Cr2O3 coverage occurs, the film will also Fig. 13 Variation of parabolic oxidation rate with chromium
level and temperature. Source: Ref 18
contain the less-protective spinel FeCr2O4. The
Cr2O3 scale is more protective because it better the less-protective FeCr2O4, and the scale
restricts the diffusion of oxygen to the interface growth rate will increase beyond the parabolic
between the scale and the base metal, which is relationship, leading to breakaway oxidation.
where the oxidation reaction occurs. As the The breakaway temperature increases with in-
oxide grows, the path to the interface lengthens, creasing chromium level.
and the rate of oxidation slows. This generates The austenitic alloys benefit over the ferritic
the parabolic-type oxide growth rate that char- alloys from the presence of nickel. For a given
acterizes these alloys. The rate of oxide growth chromium level, oxidation rates decrease with
is expressed simply as: increasing nickel content. Figures 14, 15, and
16 display this relationship. The optimal range
R = kt (Eq 12) for the iron base stainless steels, shown in Fig.
14, is reached by the commercial alloy 310,
in which the rate is in units of mass gained per with 25Cr-20Ni composition.
unit of surface area and time. This rate is a Alloying can alter the oxidation-resisting per-
strong function primarily of chromium level, as formance of the austenitic stainless steels. Some
can be seen in Fig. 13. elements form more protective oxide layers
The rate increases exponentially with temper- than Cr2O3. Aluminum and silicon are most use-
ature since diffusion governs this phenomenon. ful in this regard. Aluminum forms a layer of
The rate drops dramatically as chromium Al2O3 that is more restrictive to oxygen diffu-
reaches the concentration necessary to generate sion than is Cr2O3, as does silicon through the
the protective Cr2O3 layer. Above this suffi- formation of SiO2. The alloys 302B, 153MA,
ciency level, further increases in chromium are and 253MA all use elevated silicon levels. Alu-
not as beneficial; they mainly provide a reser- minum’s strong ferrite-promoting tendency re-
voir of chromium to re-form the Cr2O3. stricts its utility in austenitic grades, however.
As long as chromium can diffuse to the inter- While the gains from using under 3% silicon
face at a sufficient rate to satisfy the incoming are impressive, rare earths can yield even
flux of oxygen, the parabolic rate is maintained. greater benefits from mere trace additions.
If there is insufficient chromium flux, then the 153MA (UNS S30415) is a variation on 304
oxygen penetrates beyond the interface and using silicon and cerium. Cerium appears to act
forms Cr2O3 in situ. The oxide will change to at the metal-scale interface such that the oxides
Fig. 14 Influence of nickel on oxidation of iron-chromium

alloys. Source: Ref 19 Fig. 16 Corrosion rates for various stainless steels and nickel
base alloys. Source: Ref 20
Fig. 15 Isooxidation curves. Source: Ref 20
formed are thinner, tougher, more adherent, and ritic alloys, they stress their scale more during
more protective, although there is no consensus thermal cycling. This can fracture the scale,
on the mechanism. Figure 17 shows the im- causing spalling and rapid subsequent oxidation
provement quantitatively (Ref 21). of the underlying metal. This serious perform-
Because austenitic stainless steels have a ance flaw also is remedied by rare earths, as
greater thermal expansion coefficient than fer- shown in Fig. 18 (Ref 21).
Fig. 17 Comparison of rare earth-alloyed stainless alloys to conventional stainless alloys, 4833 = 309S, 4845 = 310S.
Source: Ref 21
The formation of an oxide on stainless steel

should be understood to imply de facto the de-
pletion of chromium from the underlying
metal surface. Whether the scale is formed in
service, during heat treating, or during welding
the surface, once the oxide is removed, there
will be less chromium than the bulk alloy,
often by a very significant amount, and there-
fore the corrosion resistant will be less. This is
why welds must have not only their heat tint
removed, but also the underlying metal which
is depleted in chromium, to a depth on the
order of 10 μ (Ref 24).
Fig. 18 310S (4845) versus rare earth-modified 253MA in Microstructure can also affect oxidation re-
cyclic oxidation. Source: Ref 21
sistance. As a generalization, it can be said that
changes that promote the diffusion of chromium
While the mechanism by which rare earths assist in the formation of a protective scale and
make the scale tougher and more adherent are improve oxidation resistance. Thus, cold work
vague, their effect in making austenitic alloys and finer grain size are positive factors via their
much better at resisting high-temperature oxida- enhancement of chromium diffusion.
tion, especially cyclic oxidation, are undeniable. At the very highest temperatures, Cr2O3 can be
Alloying elements can also be detrimental to further oxidized to CrO3, which has significant
oxidation resistance. Manganese, an even more vapor pressure above about 1100 °C. The com-
potent oxide former than chromium, forms a positions of some of the main high-temperature
manganese-chromium spinel that is less protec- austenitic alloys mentioned here are given in
tive than the Cr2O3. Molybdenum and tungsten, Table 2.
which are refractory metals and are beneficial to Other Environments. The most common ad-
aqueous corrosion resistance, form volatile, low- dition species that aggravates high-temperature
melting oxides (MoO3 and WO3) that promote oxidation is water vapor. At 10%, water vapor
catastrophic oxidation (Ref 22, 23). Vanadium will increase oxidation by a factor of ten. It acts
also forms an oxide, V2O5, which melts at 660 by increasing the porosity of the oxide scale and
°C and can also cause catastrophic oxidation. by promoting formation of the volatile CrO2
(OH)2 species. As a rule of thumb, maximum In lean austenitic alloys used in high-temper-
service temperatures should be reduced by 50 to ature, 600 to 1000 °C service, formation times
100 °C in the presence of steam. are relatively long, on the order of 100 h or
Halogens can attack oxide scales and cause more. In richer alloys, such as 310, times can be
their degradation or volatilization. as short as several hours. If the temperature at
Carburization and nitriding are best prevented which the alloy is to be used is one in this tem-
by an oxide layer that forms at very low oxygen perature range, then some σ is a foregone con-
partial pressures as chromium and silicon con- clusion, and while σ will have little detrimental
tents are increased. Austenitic alloys have no effect on short-term properties at these tempera-
advantage over ferritic in this regard. tures, long-term properties such as creep, stress
Intermetallic Phases. Transition elements rupture, and especially rupture ductility are de-
may combine to form intermetallic phases in graded by σ. For alloys, σ is an even greater
which the formula can vary from B4A to A4B. concern as these are prone to its formation and
Table 3 lists the most common secondary can inadvertently form some during processing.
phases encountered in austenitic stainless steels, If such alloys are intended for use near room
i.e., apart from austenite and ferrite. Sigma for- temperature, then their toughness will be seri-
mation is retarded by nitrogen, so alloys such as ously reduced by the brittle σ, which will form
153MA are less prone to it. Lower chromium first at triple points and then throughout grain
and higher nickel are beneficial. Silicon and boundaries. Because of this morphology, just a
aluminum are detrimental, as is molybdenum. few percent intermetallic phase can cause
The most relevant is σ. It can contain as little toughness to decrease by an order of magnitude.
as four iron to one chromium or molybdenum in High-Temperature Mechanical Properties.
a tetragonal structure. Thus, it can exist in many Above about 500 °C, yield strength is less ap-
conventional austenitic alloys. Other relevant propriate than creep strength in assessing the
phases are χ and Laves. The greatest risk from adequacy of an austenitic stainless steel for
these phases is the loss of room temperature structural purposes. The resistance of a material
toughness, followed by some loss of corrosion to creep is generally measured by the creep rup-
resistance. ture strength, which is the stress that causes
Table 2 Notable high-temperature austenitic alloys

Alloy Designation C N Cr Ni Mo Mn Si Other Max temp, °C
302B S30215 0.15 0.07 17.8 8.1 ... 1.8 2.5 ... 950
304H S30409 0.08 0.08 18.8 8.1 ... 1.8 0.50 ... 820
321H S32109 0.06 0.03 17.8 9.1 ... 1.8 0.50 0.6 Ti 820
153MA S30415 0.05 0.15 18.5 9.5 ... 0.6 1.3 0.05 Ce 1000
309S S30909 0.08 0.07 23.0 12.2 ... 1.7 0.50 ... 1040
309Si DIN 1.4828 0.08 0.07 19.8 11.1 ... 1.8 2.0 ... 1040
253MA S30815 0.08 0.17 21.0 10.5 ... 0.6 1.5 1.0 Al, 0.05 Ce 1100
310S S31008 0.05 0.03 24.6 19.2 ... 1.6 0.60 ... 1090
353MA S35315 0.05 0.15 25.0 35.0 ... 1.5 ... 0.05 Ce 1200+
330 S33000 0.06 ... 18.0 35.0 ... 1.7 0.90 ... 1200
332Mo S35125 0.04 0.04 21.0 34.5 2.4 1.1 0.40 0.40 Nb 1200
Table 3 Secondary phases in austenitic stainless steel

Precipitate Structure Parameter, (Å) Composition
NbC fcc(a) a = 4.47 NbC
NbN fcc a = 4.40 NbN
TiC fcc a = 4.33 TiC
TiN fcc a = 4.24 TiN
Z-phase Tetragonal a = 3.037, c = 7.391 CrNbN
M23C6 fcc a = 10.57–10.68 Cr16Fe5 Mo2C (e.g.)
M6C Diamond cubic a = 10.62–11.28 (FeCr)21Mo3 C; Fe3Nb3C; M5SiC
σ-phase Tetragonal a = 8.80, c = 4.54 Fe,Ni,Cr,Mo
Laves phase Hexagonal a = 4.73, c = 7.72 Fe2Mo, Fe2Nb
χ-phase bcc(b) a = 8.807–8.878 Fe36Cr12 Mo10
G-phase fcc a = 11.2 Ni16Nb6 Si7, Ni16Ti6 Si7
(a) fcc, face-centered cubic; (b) body-centered cubic
rupture after 10,000 or 100,000 h. If deforma- preferred. Substitutional elements have limited
tion is a greater concern, however, the creep de- effect, but interstitial solid solution elements,
formation strength, that is, the stress that results such as carbon and nitrogen, are quite useful.
in a strain of 1% after 10,000 or 100,000 h, can Nitrogen is the better addition in this regard,
be used as a basis for design. plus it has the collateral benefit of strongly
Cold work and precipitates tend to be ineffec- retarding intermetallic phase precipitation.
tive strengtheners at temperatures that produce Figures 19 to 22 compare mechanical properties
solution annealing and precipitate coarsening of the major high-temperature austenitic alloys
(overaging). Thus, solid solution hardening is (Ref 25).
Fig. 19 Charpy V toughness after 200 hr aging
Fig. 20 Relative 100,000-h creep strength

Fig. 21 100,000-h creep rupture strength
Fig. 22 High-temperature, short-time yield strength
Corrosion-Resistant Austenitic Alloys temperature, thus giving the common perception

that they are superior in corrosion resistance.
Stainless steels are almost always chosen at The main advantage austenitic alloys have is
least in part for their corrosion resistance. In their ability to utilize the powerful and inexpen-
normal atmospheric conditions, alloys with sive alloying element nitrogen. That is the key
more than 10.5% chromium do not rust. aspect of the more modern austenitic stainless
Austenitic alloys require higher levels than steels that have come into use in the last two
this to stabilize the austenitic structure at room decades.
The ion that is most aggressive against stain- The main factors in the resistance of austenitic
less steel is one of the most pervasive in our alloys to pitting attack are generally given by:
environment. The chloride ion is found in
abundance over the entire earth. It is, of PREN = %Cr + 3.3(%Mo) + 30(%N) (Eq 13)
course, in seawater, but also in the rain, on
roads, in food, and even in our bodies. Chlo- Pitting resistance is measured by ASTM G 48
rides destabilize the passive film. If the condi- (practice C) in which the lowest temperature at
tions of chloride concentration, temperature, which pitting occurs in a 6% FeCl3 is measured.
and acidity are sufficiently aggressive to break This is the critical pitting temperature, CPT. The
down the film, then active corrosion ensues. If relationship between PREN and CPT is shown
this is highly localized because of a local in Fig. 23 (Ref 26).
weakness in the passive film, then pitting oc- The function of chromium in the passive film
curs. Such a pit may be unstable and repassi- is intuitively clear. As the chromium content of
vate, or it may grow without limit. Other an alloy increases, the ready reservoir of
halides have the same effect, but they are less chromium to form the chromium-rich layer is fa-
ubiquitous. cilitated. The roles of molybdenum and nitrogen
Because of the specific virulence of the chlo- are subtler and are still subject to controversy. It
ride ion and because of its universal presence, is the subject of much research, which has been
corrosion-resistant austenitic stainless steels all summarized in reviews. The obvious paradox is
look like they were designed to resist chloride- how can elements that are not active in the pas-
pitting attack. Pitting in stainless steels is in sive film be so effective in maintaining its in-
most instances the threshold level of corrosion. tegrity. We do know that the essential chromium
Crevice corrosion is, however, more severe and in the matrix of stainless steels is quite reactive
usually design limiting vis-à-vis corrosion. and will form compounds with carbon, oxygen,
Crevices can exist not only in deliberate joints sulfur, and other transition elements. When it
but also under environmental deposits, paint does, it is no longer effective as a passive film
films, weld splatter, etc. There are other envi- former. The regions from which the chromium
ronments in which the resistance follows differ- diffused to form the chromium-rich phase will
ent rules, such as oxidizing acids, bases, and or- be poor in chromium unless subjected to a
ganic acids, but these are best regarded as lengthy homogenizing anneal. Most theories of
exceptions. pitting founder at the start because they assume
Fig. 23 Critical pitting temperature versus pitting resistance equivalent number (PREN); SUS 329J4L = S31260, YUS 270 = S31254.
Source: Ref 26
a homogeneous passive film, which is an impos- The discovery that nitrogen was beneficial
sible goal in reality. The chapter on corrosion against corrosion permitted a breakthrough in
deals with this topic in more depth. alloy development by the 1980s. Nitrogen was
The passive layer is extremely thin compared increased to around 0.20% from nominal levels
to oxide layers. It is on the order of 1 to 10 nm of 0.05%. This was found to increase PREN by
thick. Its formation does not cause chromium another five units, but more importantly, also
depletion beneath it, as oxide layers do. As the suppressed σ formation to times that permitted
alloy content of chromium and molybdenum in- thicker sections to be welded without embrittle-
crease, the film is thinner, and the current den- ment. Research into the thermodynamics of
sity required to form the film is correspondingly nitrogen in austenite showed that manganese in-
reduced. creased the solubility of nitrogen appreciably.
The corrosion-resistant austenitic stainless This permitted even higher levels of total alloy-
steel grades range from 316 to the various high- ing to be achieved. This was exploited in the
molybdenum, high-nitrogen alloys commercial- alloys UNS S34565 and S32654, which contain
ized in the last ten years, the most notable of 3 to 6% manganese and about 0.50% nitrogen.
which are listed in Table 4 with their typical The PRENs of these alloys are over 50, which
analyses. gives them a critical pitting temperature around
Early grades were based on alloying with 100 °C. Table 5 lists the performance of the var-
chromium and molybdenum with sufficient ious popular corrosion-resistant austenitic stain-
nickel added to preserve the austenitic structure. less steels.
Each of these elements facilitates the formation The advances are quite appreciable and made
of the passive film and reduces the corrosion stainless steel a viable material for many appli-
rate in the active state. Further experimentation cations for which previously nickel base or tita-
showed that molybdenum was not beneficial nium alloys had been required.
under highly oxidizing conditions, but that sili- The obvious question in view of the success
con was helpful under such conditions. Copper of the use of high manganese levels in conjunc-
was beneficial against sulfuric acid. tion with high nitrogen levels in the most highly
Alloy development came in stages. First, 317 alloyed austenitic stainless steels is when this
was the most significant corrosion-resistant approach will be used for the medium-alloyed
alloy. Then, more chromium and molybdenum austenitics to make alloys superior to 316, 317,
were added, and the class of alloys known col- and 904 without the high levels of nickel and
lectively as the 6%Mo alloys was commercial- molybdenum that render these alloys so expen-
ized. Allegheny’s AL-6X™ is representative of sive. It does not take much imagination to envi-
this group. With PRENs of around 40, these al- sion alloys equal to 904L in PREN with less
loys were resistant to seawater at ambient but nickel and molybdenum than 316L that would
not at elevated temperatures. This left a great be almost totally resistant to intermetallic phase
deal wanting in corrosion resistance. These al- precipitation and have much greater resistance
loys were also very difficult to process, at least to SCC because of higher austenite stability.
in part because they rapidly formed brittle grain The same case could be made for a 317-equiva-
boundary σ-phases in as little as several minutes lent alloy in corrosion resistance with less than
at some temperatures. This limited chromium 7% nickel and 0.5% molybdenum, essentially a
and molybdenum levels to a total of about 30%. 301 in alloy cost. In view of these trends in
Table 4 Typical compositions of corrosion-resistant austenitic stainless steels

Alloy Designation C N Cr Ni Mo Mn Si Other Other
316L S31603 0.02 0.03 16.4 10.5 2.1 1.8 0.5 ... ...
316Ti S31635 0.02 0.03 16.4 10.5 2.1 1.8 0.5 0.40 ...
317L S31703 0.02 0.06 18.4 12.5 3.1 1.7 0.5 ... ...
317LM S31725 0.02 0.06 18.4 13.7 4.1 1.7 0.5 ... ...
904L N80904 0.02 0.06 19.5 24.0 4.1 1.7 0.5 1.3 Cu ...
JS700 N08700 0.02 0.06 19.5 25.0 4.4 1.7 0.5 0.4 Cu 0.3 Nb
254SMO S31254 0.02 0.20 20.0 18.0 6.1 0.8 0.4 0.8 Cu ...
4565 S34565 0.01 0.45 24.0 18.0 4.5 6.0 ... ... ...
654SMO S32654 0.01 0.50 24.0 22.0 7.2 3.0 ... 0.5 Cu ...
AL6-XN N08367 0.02 0.22 20.5 24.0 6.2 0.4 0.4 0.2 Cu ...
Al6-XN Plus N08367 0.02 0.24 21.8 25.3 6.7 0.3 0.4 0.2 Cu ...
alloy development, the future use of 316 and thermomechanical history. As a rule of thumb,
317 should be numbered. There is no justifica- the environment to initiate SCC must be suffi-
tion for the use of scarce and expensive re- ciently severe to cause localized corrosive at-
sources such as nickel and molybdenum when tack. The most dangerous situation is one in
cheap, abundant replacements like manganese which the expectation of pitting is marginal.
and nitrogen are available. The use of 316 as a The mechanism of SCC has been a subject of
standard alloy should in the future be eroded by intense academic controversy for many years.
more cost-effective alloys such as the lean du- Because of this, much of the literature has fo-
plex alloys like 2003. cused on arguing a case rather than clarifying
The same environments that cause pitting the phenomenon. What can be said about SCC
corrosion also cause crevice corrosion. A in austenitic stainless steels with consensus?
crevice is a volume in and out of which diffu-
• Risk of SCC is low at room temperature and
sion is restricted to a degree that corrosion prod-
increases exponentially with temperature.
ucts accumulate and cause the contained envi- • SCC is preceded by localized corrosive at-
ronment to become increasingly aggressive in
tack, which has an incubation time, and then
pH and [Cl–]. Conditions that are below the
threshold for pitting can cause crevice corro- proceeds in a discontinuous manner.
• Fracture may be transgranular, intergranular,
sion. The critical temperature for crevice corro-
sion is also measured in 6% FeCl3 (ASTM G-48 or both. It is almost entirely lacking in plas-
B or D). It is the lowest temperature at which tic deformation with little, if any, metal loss.
crevice corrosion occurs. This temperature, the • Alloying or treatments that delay localized
CCT (critical crevice temperature), is lower attack or stabilize austenite can delay SCC
than the CPT. GTAW a wrought alloy also low- up to the point of virtual immunity.
ers the CPT to about the level of the CCT. The • SCC is aggravated by increased chloride
reason for this lowering of resistance to local- concentration and acidity, but also exists in
ized attack has been thought to be related to caustic environments.
alloy depletion caused either by dendrite coring • Stress must exceed a threshold for a given
during solidification or chromium depletion set of conditions for SCC to occur.
around inclusions. The relation to crevices • Anodic or cathodic polarization may prevent
would thus seem to be that surfaces contain nu- SCC under conditions at which it would oth-
merous flaws with respect to corrosion resist- erwise be expected, or it may cause SCC in
ance, which, while not capable of sustaining pit- environments in which it would not other-
ting, can in a crevice dissolve and alter the wise occur.
environment sufficiently that the new harsher Austenitic stainless steels are not alone in their
environment can generally destabilize the pas- susceptibility to SCC. All stainless steels suffer
sive film and proceed autocatalytically. from SCC under some set of conditions of envi-
Stress corrosion cracking is the bane of ronment and material thermomechanical history.
austenitic stainless steels. SCC occurs when The key is to choose an alloy that is resistant
there is both a tensile stress of a sufficient mag- under the conditions of use. To a first approxima-
nitude and a sufficiently aggressive environ- tion, this means using an alloy that will not pit
ment. The threshold stress varies with alloy and under the conditions of use, then designing its
Table 5 Corrosion resistance ratings of various austenitic stainless steels, using 30 factor for
nitrogen
Alloy Designation PREN(a) CPT oC CCT oC
316L S31603 24 15 –3
316Ti S31635 23 15 –3
317L S31703 30 25 0
317LMN S31726 34 30 4
904L N80904 35 40 15
JS700 N08700 36 43 15
254SMO S31254 46 73 38
4565 S34565 53 90 50
654SMO S32654 63 105 75
AL6-XN N08367 50 78 43
AL-6XN Plus 50 min 95 60
(a) PREN, pitting resistance equivalent number.
use to be below the threshold stress completes a Nitric acid is not particularly aggressive
sound design approach if residual stresses can be against stainless steels. Resistance to it is pro-
accurately known. Otherwise, assuming that the portional to chromium content. So, attack,
metal will have residual stresses equal to 100% should it take place, is preferably at grain
of the yield strength is the prudent engineering boundaries, where segregation of elements such
approach. Figure 24 shows threshold stress for a as carbon, phosphorus, and silicon can lower
number of alloys. chromium locally. These elements are kept low
Special Corrosive Environments. Knowl- for nitric acid service. Pitting is not a risk. Stan-
edge of the ability of the various stainless steels dard usage is:
to resist specific environments is essential to the
design process. This information is extensive • Below 50% concentrations and below 100
since it must correlate many environments and °C, 304L and 17% Cr ferritics are used.
temperatures for many materials. Hence, refer • Around the 65% aziotrope, 310 is most re-
to the publications of organizations such as the sistant, especially a low carbon version, but
National Association of Corrosion Engineers 304 NAG with low carbon, phosphorus, and
(NACE) or to the Web sites of companies such silicon is more often used.
as Allegheny Ludlum or Outukumpu, where • For 98% solutions or for lower concentra-
such information is available freely. The more tions that contain other stronger oxidizers,
reputable producers will give assistance on spe- alloys with 4% Si, 18% Cr, and 15% Ni or
cific questions. Engineering forums on the In- 5% Si, 17% Cr, and 17% Ni (UNS S30600
ternet, such as www.eng-tips.com, should also and S30601) have been developed.
be considered a resource. The following discus- Hydrochloric acid, not surprisingly, is quite
sion presents just the principles of the resistance aggressive against stainless steel. It is very ef-
of austenitic stainless steels to specific, more fective in destabilizing the passive film. Thus,
common environments. resistance to hydrochloric acid is simply an ex-
Sulfuric acid is common, aggressive, and must treme case of resistance to pitting in chlorides
be contained. Figure 25 shows the isocorrosion with resistance given by Eq 12. Only the most
curves for several alloys in pure sulfuric acid. highly alloyed austenitic alloys, such as AL-
Alloy 20, 904L, and alloy 825 were devel- 6XN®, should be considered and then under
oped specifically for sulfuric acid service. Each conditions that are tolerable, such as those
contains several percent copper that, while not shown in Fig. 26.
beneficial against pitting, concentrates in the Strong bases such as NaOH and KOH are not
passive film and diminishes general corrosion. especially aggressive against stainless. The 17%
Molybdenum and tungsten are also very benefi- chromium alloys can be used up to 50 °C, while
cial for resistance to sulfuric acid. Phosphoric 304L can be used to 90 °C. As is the case with
acid is similar to sulfuric acid in its effect on nitric acid, chromium and nickel are beneficial,
austenitics but somewhat less aggressive. while molybdenum is counterproductive. The
25% chromium alloys such as 310 or an equal
Fig. 24 Threshold stress for stress corrosion cracking (SCC)

for various alloys. Source: Ref 27 Fig. 25 Isocorrosion in pure sulfuric acid. Source: Ref 28
contributing causes. Powder injection-molded

stainless components often have porosity that is
generally spherical. When exposed to the sur-
face, such pores act as crevices and lower the
pitting potential also. All of these factors are op-
erative and can act in unison.
Very fine abrasive polishing causes little
residual stress and has very minimal crevice
creation. Thus, mirror-type polished finishes do
not degrade corrosion resistance, but they do
not enhance corrosion resistance as does elec-
tropolished mirror finishes, which remove
chromium-depleted sites, which can initiate
Fig. 26 Resistance to hydrochloric acid. Source: Ref 28 pitting.
chromium duplex can be used to 150 °C, above REFERENCES

which temperature nickel base alloys are re-
quired. High-chromium ferritic stainless steels 1. A.L. Schaeffler, Constitution Diagram for
are also very good choices. Stainless Steel Weld Metal, Met. Prog.,
Organic acids are generally less aggressive Vol 56, Nov 1949, p 680–688
against stainless than are mineral acids since 2. W.T. Delong, A Modified Phases Diagram
they are less dissociated in solution. They be- for Stainless Steel Weld Metals, Met. Prog.,
come hazardous when they contain chloride Vol 77, Feb 1960, p 98
ions, at high temperatures, or when they disso- 3. Design Guidelines for Selection and Use of
ciate strongly, such as with formic acid. Stainless Steel, SSINA,1998, p 3
Because of the large number of organic com- 4. K.-J. Blom, “Press Formability of
pounds that may be considered, refer to the var- Stainless Steels,” paper presented at Stain-
ious corrosion tables. less steels ‘77
Surface Finish. The corrosion resistance of 5. F.B. Pickering, “Physical Metallurgical De-
austenitic stainless steels is quite dependent on velopments in Stainless Steel,” paper pre-
surface condition, as are other stainless steels. sented at Stainless Steel ‘84, Goteborg
Treatments that enhance surface concentrations 6. Q.-X. Dai et al., Chin. Phys., Vol 11, 2002, p
of beneficial elements or remove detrimental 596–600, doi:10.1088/1009-1963/11/6/315
constituents can greatly alter performance. 7. Aciers Inoxidables, Les Editions de
Oxide formation depletes surface chromium, so Physique, Les Ulis, Paris, 1993, p 564
strong pickling or electropolishing of the 8. Aciers Inoxidables, Les Editions de
descaled surface is especially important. Studies Physique, Les Ulis, Paris, 1993, p 565
have shown chromium depletion of a maximum 9. P. Marshall, Austenitic Stainless Steels, Mi-
of 6% extending 10 μ before reaching bulk crostructure and Mechanical Properties,
chromium levels. This is equivalent to the de- Elsevier, 1984
pletion seen in sensitization. The increase in at- 10. Aciers Inoxidables, Les Editions de
tack rate from this depletion is huge. A 1000- Physique, Les Ulis, Paris, 1993, p 579
fold increase in weight loss in the ASTM G 48 11. H. Nordberg, Mechanical Properties of
B test has been seen by a superficial loss of 6% Austenitic and Duplex Stainless Steels, In-
chromium. novation in Stainless Steels ‘93 (Firenze),
Likewise, surface abrasion, especially coarse 1993, p 2.217
abrasion, has a major detrimental effect. The 12. Aciers Inoxidables, Les Editions de
120 grit #3 finish often seen on stainless reduces Physique, Les Ulis, Paris, 1993, p 566
pitting resistance by as much as the equivalent 13. S.J. Rosenberg and C.R. Irish, J Res. Nat.
of 5 PREN, that is, equal to a reduction in Bar. Stand., Vol 48, 1952, p 40
chromium content of 5%. Rolled finishes are 14. Aciers Inoxidables, Les Editions de
much preferred. The mechanism for this has not Physique, Les Ulis, Paris, 1993, p 410
been clearly established; exposure of MnS in- 15. M. McGuire, “A Diffusion Model for the
clusions, the microcrevices abrasion produces, Influence of Oxygen and Sulfur on the Non-
and residual stress have been cited as possible Equilibrium Distribution of Chromium in
Austenitic Stainless Steel Welds and Slabs,” 23. N. J. Grant, Accelerated Oxidation of Met-
paper presented at Proceedings MS&T als at High Temperature, Trans. ASM, Vol
‘04,2004 44, 1961, p 128–137
16. R. Stickler and A. Vinckier, Trans. ASM, 24. J.F. Grubb, paper 04291 presented at Corro-
Vol 54, 1961, p 362 sion 2004, NACE, 2004, p 1–15
17. Aciers Inoxidables, Les Editions de Phy- 25. ACOM Files, High Temperature Stainless
sique, Les Ulis, Paris, 1993, p 568 Steels, www.outukumpu.com
18. Aciers Inoxidables, Les Editions de 26. J. Okamoto et al., A Super-Austenitic Stain-
Physique, Les Ulis, Paris, 1993, p 448 less Steel for Tubing and Piping Applica-
19. Aciers Inoxidables, Les Editions de Phy- tions, Nippon Steel Technical Report 90,
sique, Les Ulis, Paris, 1993., p 453 July 2004
20. Aciers Inoxidables, Les Editions de 27. ACOM Files, High Temperature Stainless
Physique, Les Ulis, Paris, 1993, p 454 Steels, www.outukumpu.com
21. www.outukumpu.com 28. Allegheny Technologies, “AL6-XN®
22. W.C. Leslie, Mechanism of Rapid Oxida- Alloy”
tion at High Temperature, Trans. ASM, Vol
41, 1958, p 1213–1219
CHAPTER 7
Duplex Stainless Steels
Summary they are exceptional materials. Whether these

duplex alloys will grow to the full extent of
THE NEWEST FAMILY of stainless steels is their potential depends on several factors:
the duplex alloys. The mixture of ferrite and
austenite in their structure gives them higher • Will high nickel and molybdenum prices be
strength than either phase by itself. Duplex al- sufficient motivation to drive designers to
loys have at least 20% chromium, so they are explore alternatives to traditional austenitic
considered as highly corrosion-resistant alloys grades?
but not high-temperature alloys because of em- • Will producers overcome their inhibition to
brittling phases. Their low nickel content makes aggressively market these grades through
them more economical than austenitic alloys of their cost-saving potential?
the same level of corrosion resistance, espe- • Will producers perfect the techniques to pro-
cially when their greater strength can be utilized duce these grades reliably so that their avail-
to reduce the amount of material required. They ability is unquestioned?
should largely replace alloys such as 316L • Will design codes change to correctly reflect
and317L in the future. the duplex materials’ higher ratio of yield
strength to tensile strength?
Why are there such issues with a family of al-
Introduction loys that has been successfully used for 20 years?
The concept of duplex stainless steels is simple:
Duplex stainless steels are the newest and islands of austenite in a continuous matrix of
fastest-growing alloy group in the stainless steel highly alloyed ferrite, as seen in Fig. 1. This
family. They are called duplex because at room
temperature they consist of two phases, ferrite
and austenite. Discovered in the 1920s, they
languished in a suboptimized and underutilized
state until recently. They possess excellent
strength, toughness, and corrosion resistance.
They also display exceptional resistance to
stress corrosion cracking (SCC) and corrosion
fatigue. The leaner grades, such as 2304, corre-
spond to 316L in corrosion resistance but have
double the yield strength, while the higher alloy
grades like 2507 compete with the 6% molyb-
denum superaustenitics in corrosion resistance
while still possessing much greater strength.
Their limitations lie in their lack of cryogenic
toughness and their inability to withstand tem-
peratures much above 300 °C without forming
embrittling phases. But between –100 and 300 °C Fig. 1 Wrought 2205 duplex microstructure
combination in principle offers high strength be- in the ferrite matrix, and each phase would have
cause of the possibility of refining the dual-phase equal corrosion resistance despite having differ-
grain structure and thereby raising yield strength ent compositions. It took a long time for that to
according to the Hall-Petch relationship as well be accomplished.
as by solid solution hardening, especially with Figure 2 shows a simple Fe-Cr-Ni constitu-
nitrogen. In addition, the absence of a continuous tional diagram (Ref 1). The salient points are
austenite phase provides relief from SCC by hav- that the typical successful alloys nearly bisect
ing any propagation of cracks in austenite ar- the two-phase field for austenite and ferrite. It is
rested by the ferrite phase. also obvious that the composition of the austen-
The optimization of the alloy system had to ite and the ferrite must be quite different.
wait for two events, both related to nitrogen. Ferrite contains a great deal more chromium
First, the control of nitrogen in the refining by than austenite; hence, its pitting corrosion re-
the argon oxygen decarburization (AOD) process sistance contribution from chromium is much
allowed the control nitrogen content up to the greater than the resistance of the austenite be-
solubility limit. Second, the understanding of cause in duplex grades:
the thermodynamics of the alloy system became
understood and reduced to a computer model. PREN = %Cr + 3.3 × %Mo + 16 × %N (Eq 1)
At this point, the alloys developed over the first
50 years of development became obsolete, and If one were to add molybdenum to increase
new grades with higher nitrogen vastly improved pitting resistance, it would preferentially parti-
the performance and user friendliness. Why this tion to the ferrite, further exacerbating the dif-
was so important can be seen by studying the ferential between the two phases.
structure of these alloys. This is where nitrogen saves the day. Addi-
tions of nitrogen concentrate nearly entirely in
the austenite. This lowers the activity of
Structure and Alloy Design chromium and thereby effectively attracts more
chromium to the austenite phase than would
The ideal structure of a duplex grade would otherwise be present. This stabilizes the austen-
be a stable 50-to-50 ratio of austenite to ferrite ite, keeping the ratio of ferrite to austenite more
at all temperatures at which it is to be used with- nearly constant with temperature. The pitting
out other phases. The austenite would be islands resistance of the austenite increases signifi-
Fig. 2 The Fe-Cr-Ni phase diagrams. The shaded area results from nitrogen additions
Chapter 7: Duplex Stainless Steels / 93
cantly to approximately that of the ferrite. In ad- (PREN) of about 35 and fills a niche in corrosion
dition, the nitrogen solid solution strengthens resistance where austenitics and ferritics are
the austenite and retards the formation of inter- lacking, greater than 317L stainless, PREN = 30,
metallic phases, which is not bad for an element and below the 6% molybdenum grades, such as
that costs nothing. AL-6XN alloy, with PRENs of around 45. Fer-
ThermoCalc, developed by the Swedish Royal ritics have a gap between 442 (18Cr-2Mo) and
Academy, has been an especially valuable tool in the super ferritics (28Cr-4Mo). By varying the
helping us understand and design better duplex chromium, nickel, and molybdenum, leaner al-
stainless steels. Without being able to computer loys can be devised that save cost based on re-
model the thermodynamics of the system, it duced molybdenum and nickel. Conversely,
would be impossible to project the partitioning of more corrosion-resistant alloys with higher
the various potential alloying elements. Figure 3 PRENs can also be mapped, such as Fig. 4, with
shows isopleth diagrams for a basic 2205 compo- the same diagrams varying nickel for a higher
sition in which nickel is varied. molybdenum level. This composition includes
The 2205 is the workhorse grade of duplex. It the important 2507 alloy.
has a pitting resistance equivalent number Partitioning of elements (Fig. 5) between
austenite and ferrite is an important issue. The
partitioning tendency is a strong function of
temperature. Figure 6 shows that as temperature
Fig. 3 The iron-nickel diagram for 22% Cr, 3% Mo, 0.15% N Fig. 5 Partitioning tendencies of various elements between
ferrite and austenite. Source: Ref 2
Fig. 4 The iron-nickel diagram for 25% Cr, 4% Mo, and

0.25% N: N is a nitride, χ is chi, σ is sigma, α is fer- Fig. 6 Variation of partitioning ratio with temperature.
rite, and γ is austenite Source: Ref 2
increases, the partitioning diminishes until at cleation and diffusion. The areas around the
just above 1300 °C it approaches unity for all newly formed σ are naturally somewhat di-
normal substitution-alloying elements (Ref 2). minished in chromium and molybdenum, so
For nitrogen, however, the tendency is to in- the alloy’s resistance to localized corrosion is
creasingly segregate to austenite as temperature compromised also.
increases (Ref 2). Figure 9 shows the TTT (time-temperature-
A danger in these alloys is that austenite transformation) diagram for various high-alloy
formed from ferrite on heating, such as during stainless steel, including austenitic, ferritic, and
welding or annealing, will contain only the low duplex. Alloys of all structures, ferritic,
amount of nitrogen that was in the ferrite from austenitic, and duplex, with high chromium and
which it was formed, until diffusion can restore molybdenum encounter the σ problem fairly
equilibrium. If the heating time does not permit equally and in proportion to their alloy content
this, this so-called secondary austenite will (Ref 2). This is the reason that the use of nitro-
have low nitrogen and therefore low pitting gen instead of molybdenum is so beneficial to
corrosion resistance, as shown in Fig. 7(e). Ni- the leaner alloys, not just in cost, but for the
trogen alters the phase stability, making austen- major reduction in rate of formation of sigma.
ite stable to higher temperatures. This helps Figure 10 shows the large reduction in σ forma-
keep welds from becoming excessively ferritic tion enjoyed by the lean alloy AL 2003™ mate-
and disturbing the desirable 50-to-50 ratio of rial compared to the higher molybdenum 2205
austenite to ferrite. Secondary austenite with alloy (Allegheny Ludlum).
low nitrogen is remedied by diffusion if the There are other intermetallic phases in addi-
phase forms at higher temperatures at which tion to σ. They include χ, R, π, and τ. These are
diffusion of nitrogen can rehomogenize the ni- of more research than practical interest because
trogen level. σ, with its bad consequences, forms sooner and
A crucial aspect of alloy design in the duplex in greater quantity under the same conditions
alloys involves the avoidance of unwanted compared to the others.
phases. The duplex stainless steels have all the Carbides and nitrides can also form in duplex
potential problems with embrittling phases of alloys. The nitride Cr2N can form when satu-
the ferritic and austenitic stainless steels com- rated ferrite is quenched from a high tempera-
bined since they contain both as phases. Ferrite ture, as can occur in the welding process. It is
forms two main embrittling phases, α′ and σ. possible that this would result in nearby
The α′ is generally believed to be a result of chromium depletion and a decrease in corrosion
the miscibility gap that exists in the iron- resistance. Carbide formation does not as easily
chromium system, by which ferrite undergoes cause chromium depletion in duplex alloys be-
spinodal decomposition into the iron-rich α, cause the precipitation at the ferrite-austenite
normal ferrite, and the chromium-rich α′, grain boundary does not deplete the austenite as
which is a brittle ordered alloy. Higher levels greatly in chromium locally because of the
of chromium or the presence of copper or neighboring ferrite having a much higher diffu-
molybdenum exacerbate this reaction, which sivity for chromium. The point is generally
has a formation that follows an Arrhenius-type moot since all modern duplex grades contain
curve with a maximum at around 400 °C. Fig- less than 0.030% carbon.
ure 8 shows the α′ formation kinetics for five Table 1 lists the duplex grades currently
duplex alloys (Ref 2). While duplex grades available commercially. Figure 7 shows a series
have good oxidation resistance and high-tem- of duplex photomicrographs.
perature strength, the α′ problem restricts their
use to below about 315 °C. Ferrite and austen-
ite both form intermetallic phases, of which
the most prominent and dangerous is σ, a Mechanical Properties
tetragonal phase richer in chromium and
molybdenum than the ferrite from which it In many ways, the duplex stainless alloys
forms. It is brittle and forms at grain bound- represent a best of both worlds in combining
aries, so its precipitation has the immediate ef- traits from the austenitic and ferritic alloys.
fect of lowering toughness. Cold work acceler- They offer high as-annealed strength with
ates the precipitation process by up to an order good toughness and ductility. Table 1 lists
of magnitude by virtue of its dual effect on nu- the major grades of duplex stainless steels;
Fig. 7 (a) As-cast duplex structure, austenite in a ferrite matrix. (b) 2205 annealed; austenite phase contains twins.
(c) 2507 as-welded; weld is highly ferritic because of rapid cooling rate. (d) Same weld as (c) after homogenization anneal.
(e) 7-Mo Plus with ( (dark areas) that has induced the formation of secondary austenite (arrows)
Table 2 lists typical and minimum properties The most striking and unexpected character-
for the major duplex alloys and those of some istic of the duplex grades is their high yield
comparable ferritic and austenitic alloys for strength, more than double that of comparable
comparison. austenitic grades.
Fig. 8 Kinetics of ( formation
The strength of the duplex grades is driven by

the strength of the continuous ferrite phase. It
owes its strengthening primarily to:
• Solid solution hardening by nickel, molyb-
denum, chromium, copper, and manganese
• Interstitial solid solution hardening by car-
bon and nitrogen
• Strengthening by grain refinement
These components have been related to the
mechanical properties by the following equa-
tions (Ref 3):
Fig. 9 Sigma formation kinetics at various alloy levels Rp0.2 = 120 + 210 N + 0.02
+2( Mn + Cr ) + 14 Mo + 10 Cu
+ (6.15 − 0.54δ)δ
+ (7 + 35( N + 0.02))d −1/ 2 (Eq 2)
Rp1.0 = Rp0.2 + 40 ± 9 (Eq 3)
Rm = 470 + 600( N + 0.02)

+14 Mo + 1.5δ + 8d −1/ 2 (Eq 4)
where δ is the ferrite content in percent, d is the

lamellar spacing, and results are in megapas-
cals.
The influence of nitrogen is interesting in that
at lower levels (e.g., below 0.1% nitrogen)
Fig. 10 Delay in ( precipitation in lean duplex 2003 austenite is the weaker phase, but additional ni-
Table 1 Duplex compositions

UNS Name C N Cr Ni Mo Mn Si Cu W P S
S32900 329 0.08 ... 23.0–28.0 2.5–5.0 1.0–2.0 1.0 0.75 ... ... 0.040 0.030
S31200 44LN 0.03 0.14–0.20 24.0–26.0 5.5–6.0 1.2–2.0 2.0 1.0 ... ... 0.040 0.030
S31260 DP3 0.03 0.10–0.30 24.0–26.0 5.5–7.5 2.5–3.5 1.0 0.75 0.2–0.8 0.1–0.5 0.030 0.030
S31500 3RE60 0.30 0.05–0.10 18.0–19.0 4.25–5.25 2.5–3.0 1.2–2.0 1.4–2.0 ... ... 0.030 0.030
S31830 2205(old) 0.03 0.08–0.20 21.0–23.0 2.5–3.5 2.5–3.5 2.0 1.0 ... ... 0.030 0.020
S32001 19 D 0.03 0.05–0.17 19.5–21.5 1.0–3.0 4.0–6.0 1.0 ... ... 0.040 0.030
S32003 2003 0.03 0.14–0.20 19.5–21.0 3.0–4.0 1.5–2.0 2.0 1.0 ... ... 0.040 0.030
S32101 2101 0.04 0.20–0.25 21.0–22.0 1.35–1.70 0.1–0.8 4.0–6.0 1.0 0.1–0.8 ... 0.040 0.030
S32205 2205 0.03 0.14–0.20 22.0–23.0 4.5–6.5 3.0–3.5 1.0 2.0 ... ... 0.030 0.020
S32304 2304 0.03 0.05–0.20 21.5–23.5 3.0–5.0 2.5 1.0 0.05–0.6 ... 0.040 0.040
S32520 Uranus 0.03 0.20–0.35 24.0–26.0 5.5–8.0 3.0–5.0 1.5 0.8 0.5–3.0 ... 0.035 0.020
52N+
S32550 255 0.04 0.10–0.25 24.0–27.0 6.0–8.0 2.9–3.9 1.5 1.0 1.5–3.0 . . . 0.040 0.030
S32750 2507 0.03 0.20–0.30 24.0–26.0 6.0–8.0 3.0–5.0 1.2 0.8 0.5 ... 0.035 0.020
S32760 Zeron 100 0.03 0.20–0.30 24.0–26.0 6.0–8.0 3.0–5.0 1.0 1.0 0.5–1.0 0.5–1.0 0.030 0.010
S32906 2906 0.03 0.30–0.40 28.0–30.0 5.8–7.5 1.5–2.6 0.8–1.5 0.5 0.8 ... 0.030 0.030
S32950 7–Mo Plus 0.03 0.15–0.35 26.0–29.0 3.5–5.2 1.0–2.5 2.0 0.6 ... ... 0.035 0.010
S39274 DP3W 0.03 0.24–0.32 24.0–26.0 6.0–8.0– 2.5–3.5 1.0 0.8 0.2–0.8 1.5–2.5 0.030 0.020
S39277 AF 918 0.025 0.23–0.33 24.0–26.0 6.5–8.0 3.0–4.0 0.8 0.8 1.2–2.0 0.8–1.2 0.030 0.020
Table 2 Duplex mechanical properties

Charpy-V
Grade Name Rp0.2 Rm A5 HB RC –40 °C, J
S31200 44LN 450 690 25 293 31 ...
S31260 DP3 485 690 20 290 31 ...
S31830 2205(old) 450 62 25 293 31 ...
S32003 2003 450 620 25 290 30 40
S32001 19D 450 640 25 290 25 ...
S32101 2101 450 650 25 290 32 40
S32205 2205(new) 460 640 25 290 32 40
S32304 2304 400 600 25 290 31 40
S32520 Uranus 52N+ 550 770 25 ... 28 ...
S32550 Ferralium 550 760 15 302 32 ...
S32750 2507 550 795 15 310 32 40
S32760 Zeron 100 550 750 25 270 ... ...
S32960 7-Mo Plus 485 690 15 293 32 ...
trogen strengthens the austenite so that above above 300 °C is not recommended, no higher-
0.2% nitrogen, the austenite becomes the temperature properties are shown. Since they
stronger phase. have a ductile-to-brittle transition, they also are
The two phases are elongated parallel to the not well suited to cryogenic use.
major strain axis of working such as from hot or Impact Strength. Toughness is a significant
cold rolling. As working increases, the mi- consideration when using duplex alloys to re-
crostructure and properties become increasingly place the extremely tough austenitic alloys.
anisotropic, with the austenite taking on a (110) Duplex alloy low-temperature toughness is in-
[223] texture and the ferrite (100) [011] to (211) termediate to that of ferritic and austenitic al-
[011] (Ref 2). loys. This having been said, it should be noted
Because the ferrite phase controls mechani- that the duplex alloys can have excellent tough-
cal properties, the dependence of these proper- ness levels, such as 100 J at –100 °C in the so-
ties on temperature is significant since flow in lution-annealed condition. As would be ex-
body-centered cubic (bcc) structures is ther- pected, toughness improves with decreasing
mally activated. Figure 11 shows the variation grain size and deteriorates with cold work. The
of yield and tensile strengths of various grades most deleterious effect on toughness comes
along with that of austenite and ferrite of simi- from the precipitation of intermetallic phases,
lar composition. Because use of these alloys such as α′ and σ, which cause a sharp decrease
Fig. 11 Variation of ferrite, austenite, and duplex with temperature. Source: Ref 4
in toughness level and a concurrent increase in

transition temperature. The combined effect of
cold work and α′ can be seen in Fig. 12. Lean
alloys such as 2001, 2003, and 2101 have a
much slower rate of formation of α′ and are
much less at risk for loss of toughness from ex-
posure in the 300 to 600 °C range, as was
shown in Fig. 11.
Fatigue. Fatigue tests on duplex stainless
steels indicate that they possess a fatigue limit
of about 50% of the yield strength when tested
in air (Ref 4). The ratio of the fatigue strength in
a hostile environment to that in air is a useful
measure of the complementary strong points of
the duplex grades (i.e., strength and corrosion
resistance). Figure 13 shows that ratio for vari-
ous alloys plotted versus their PREN. As an
alloy’s resistance to corrosive attack increases,
its fatigue limit in a given environment ap-
proaches that in air, indicating simply that cor-
rosion plays an increasingly small role in fa-
tigue crack propagation as corrosion resistance
increases. While this is intuitively reasonable, it
is not diminished because the duplex reward the
user with a higher level of yield strength and fa-
tigue strength in air, so the net useful strength
Fig. 12 Increase in transition temperature with α′ formation
under cyclic loading is much greater than that of with aging for (a) annealed 2705 and (b) cold-
equivalent-PREN austenitic alloys. worked 2205. Source: Ref 4
Fig. 13 Influence of pitting resistance equivalent number (PREN) to fatigue strength in NaCl solution versus in air. Source: Ref 2
Forming and Machining Heavily formed sections should be fully

annealed, not just stress relieved, whenever
The higher strength and lower ductility of the there is a potential for SCC in the service
duplex grades compared to austenitics gives environment.
them correspondingly less ability to be cold
formed. Duplex alloys have sufficient ductility
to be cold drawn; they behave like ferritics or Corrosion Resistance
austenitics of similar alloy level. This, however,
is an alloy level at which excellent formability Because duplex alloys are made up of two
is seldom expected. Nevertheless, duplex alloys phases, ferrite and austenite, each must carry its
can be cold formed like austenitic alloys. Oper- own weight in resisting corrosion. Early alloys
ations such as bending, drawing, and pressing that were lacking in nitrogen generally had a fer-
can readily be performed. Bend radii should be rite phase that, because of the greater partition-
at least twice sheet thickness. Tubing can be ex- ing of the chromium and molybdenum to the fer-
panded into tube sheets, but care must be taken rite, had higher corrosion resistance than the
to produce tight roller-expanded joints. Tubing austenite. As nitrogen is added, it enriches the
bend radii should be at least twice tubing out- austenite phase preferentially until the corrosion
side diameter (OD). resistance of the austenite phase reaches that of
the ferrite. This approach is common to all more the aggressiveness of these media is enhanced
recently developed alloys starting with the revi- by contamination.
sion of 2205 from UNS S31803 to S32205, Sulfuric Acid. Figure 14 shows the behavior
which has primarily higher nitrogen. The net re- of S32304 compared to 304 and 316 in sulfuric
sult is a type of alloy that has most of the highly acid. Figure 15 shows additional, more highly
desirable corrosion resistance characteristics of alloyed duplex grades. The use of copper as an
superferritic grades without their limiting lack of alloying element in S32550 (1.5%) and S32760
mechanical properties, mainly toughness. (0.5%) gives them much better performance
than the otherwise similar S32750.
In real-life situations, such as seen in flue gas
General Corrosion
desulfurization, sulfuric acid can be contami-
The duplex alloys offer important advantages nated with chlorides. While this contamination
in performance over the austenitic grades in a is deadly to 316 and 317, it has only a minor ef-
number of significant aggressive media, includ- fect on the copper-alloyed duplexes (Fig. 16).
ing sulfuric acid, hydrochloric acid, sodium hy- Hydrochloric Acid. Historically stainless
droxide, phosphoric acid, and organic acids. steels have had their poorest performance when
This performance extends to situations in which confronted by hydrochloric acid. Here again, the
Fig. 14 The 0.1 mm isocorrosion curves. Source: Ref 5 Fig. 15 The 0.1 mm isocorrosion curves. Source: Ref 5
Fig. 16 Isocorrosion (0.1 mm/yr) performances of several austenitic and duplex alloys. Source: Ref 6
copper/tungsten-alloyed duplexes show excep- considered for use with nitric acid, and even
tionally good performance, as seen in Fig. 17. then no advantage can be claimed.
This extends the usefulness of stainless steels to Sodium Hydroxide. Much of the older pub-
an environment that had previously been off lished data on the behavior of stainless steel has
limits. Indeed, the duplex stainless steels in gen- seemed to promote the notion that higher nickel
eral can be said to be relatively indifferent to the levels were beneficial in strong bases. There
pH of chloride solutions and are affected rather seems now to be little to support that notion.
by the chloride concentration and temperature. Figures 18 and 19 clearly indicate, respectively,
Nitric Acid. It is fairly well known and ac- that the duplex alloys with their relatively low
cepted that resistance to nitric acid, which was nickel levels significantly outperform the higher
one of the first uses of stainless steel, depends nickel 304L and 316L, with performance im-
almost entirely on the chromium content. proving with increasing chromium content. The
Molybdenum, in all other instances a very bene- advantage is magnified when the environment is
ficial alloying element, has a strongly negative contaminated with chlorides, as is the case of
influence on resistance to this highly oxidizing the white liquors of kraft digesters.
acid. Consequently, only the leanest-molybde- Phosphoric Acid. While pure phosphoric
num duplex alloys, such as S32304, should be acid is not a very corrosive medium for stainless
Fig. 17 Isocorrosion (0.1 mm/yr) performance of duplex in HCl compared to 316L. Source: Ref 6
Fig. 19 Corrosion rates in white liquors plus chlorides.

Fig. 18 Corrosion rates in boiling NaOH. Source: Ref 7 Source: Ref 8
Fig. 20 Minimum temperatures for wet phosphoric acid (WPA) with an isocorrosion rate of 0.127 mm/yr. Source: Ref 9
In combinations of acetic and formic acid, the

superiority of duplex alloys is quite evident, as
seen in Fig. 22. S32750 shows virtual immu-
nity, while in mixtures contaminated with
halides its performance ranks very closely to
expensive nickel-based superalloys such as
N06625 and N06455. Even the lower alloyed
S32205 can offer an order of magnitude im-
provement over S31703 in hot contaminated
acetic acid.
Pitting Corrosion
The different analysis of the two main phases
in duplex alloys means that each has its own pit-
ting resistance equivalent number, PREN. The
Fig. 21 Isocorrosion (0.1 mm/yr) performances of various
ferrite phase has the relationship common to
alloys. Source: Ref 9
ferritic grades:
steel, contaminants again can render it so. PREN + %Cr + 3.3%Mo (Eq 5)
Halides are particularly common and aggressive
contaminants. Figure 20 shows the substantial while the austenite obeys the more familiar:
improvement in performance of the duplex al- PREN + %Cr + 3.3%Mo + 30%N (Eq 6)
loys over 316L when contaminants are present.
The duplex grades partition these critical ele-
Performance again improves with increasing
ments in such a way that the overall PREN of
chromium, molybdenum, and nitrogen levels.
most alloys comes out to be approximately Eq 1.
Organic Acids. Duplex alloys perform partic-
If one has the actual analysis of each phase,
ularly well in organic acids and have an excellent
then the proper relationship to use is Eq 2. These
record in industrial plants. In acetic acid, 304L
relationships are incomplete in that they only ad-
handles lower temperatures and concentrations.
dress the major alloying elements. Tungsten has
Alloys such as S32205 perform well. In
half the value of molybdenum and is frequently
formic acid, the most aggressive organic acid,
included:
S32750 is resistant at all concentrations almost
to the boiling point, outperforming even tita- PREN = %Cr + 3.3(%Mo + 0.5 × %W) + 16%N
nium (see Fig. 21). (Eq 7)
Fig. 22 Corrosion rates for various alloys in 50% acetic plus formic acid, boiling. Source: Ref 10
If nonwrought material is involved, as in as- temperatures. This austenite, which forms from
cast and welded alloys, these relationships ferrite, has very little nitrogen, which clearly
greatly overstate PREN. This is because non- lowers its pitting corrosion resistance.
equilibrium-diminished concentrations of The duplex alloys stand up very well in com-
chromium are often found around precipitates, parison to corresponding superaustenitic alloys.
especially (manganese, chromium) S inclusions Figure 23 shows how CPT varies with PREN.
(Ref 11, 12) and because of lower alloy content This ranking is not always linear, as Fig. 24
locally due to solidification segregation, princi- shows, with pitting potential dropping fairly
pally of molybdenum. This is most significant rapidly with temperature and at different rates
in welded tubing, which can have higher sulfur for different alloys. Figures 25 and 26 show the
levels to increase weld penetration. Tube welds influence of pH and chloride concentration, re-
can be reequilibrated by high-temperature an- spectively. In 3% NaCl (Fig. 26), the rankings
nealing, but field girth welds will show dimin- show a minor variation with pH and a rational
ished corrosion resistance if unannealed. So, relationship to alloy content. The influence of
untreated welds can have PREN’s 5 to 15 lower chloride concentration is strong over a wide
than the parent alloy, which equates to the local- range of concentrations.
ized lowering of chromium levels. The critical These tests are best for judging relative per-
pitting temperature (CPT) of welds often de- formance of alloys and must be used cautiously
creases to near the critical crevice corrosion when extrapolating lab results to service per-
temperature (CCT) of the parent metal. formance. The degree to which short-term tests,
The precipitation of chromium- or molybde- whether potentiostatic or strictly immersion, re-
num-rich second (third, in this case) phases, flect long-term performance has not been well
such as σ or α′ inevitably results in diminish- established.
ment of these key alloying elements in the re-
gion surrounding the precipitate, which will
make it more prone to localized corrosion. Crevice Corrosion
This can also occur when secondary austen- Crevices exist both by design and inadver-
ite forms during the heating of alloys to high tently. Crevices are occluded volumes of liquid
Fig. 23 Critical pitting temperature in seawater measured potentiostatically versus pitting resistance equivalent number (PREN).
Source: Ref 13
Fig. 24 Variation of pitting potential with temperature. SCE, saturated calomel electrode. Source: Ref 14
in which oxygen and corrosion products reach The difference increases with total alloy con-
levels quite different from the exterior environ- tent, as can be seen in Fig. 27. Interestingly, the
ment and become highly corrosive. Thus, the difference is approximately the same as is the
tighter the crevice is, the greater the restriction difference in CPT between the wrought alloy
of diffusion between the crevice and the bulk and the welded alloy.
and therefore the greater the chance of crevice
corrosion occurring. An alloy’s susceptibility to
Stress Corrosion Cracking
crevice corrosion is proportional to its resist-
ance to pitting corrosion under the same condi- Stress corrosion cracking (SCC) has long
tions. The CCT is lower than the CPT by about been the Achilles’ heel of stainless steels. Only
10 to 30 °C. soft ferritic stainless steels are immune to it. It
occurs at temperatures and in environments

where stainless would be a perfect material if
only it did not stress corrosion crack. The
arrival of duplex stainless steels has to a very
large degree ameliorated, if not solved, that
problem.
SCC is unfortunately poorly understood. Like
pitting, whose initiation mechanism has not
been identified, SCC has both its initiation and
propagation mechanisms still open to debate.
But the duplex alloys have good strength
mainly through fine grain size and solid solution
hardening, which seems to avoid the hydrogen-
trapping dislocation types that seem to be asso-
ciated with hydrogen failures. So, while we can-
Fig. 25 Variation of critical pitting temperature (CPT) with not state the mechanism for SCC, we can map
pH. Source: Ref 14 out the conditions under which duplex alloys
are susceptible to SCC. The major environmen-
tal factors that affect SCC are chloride concen-
tration and temperature. Figure 28 shows the
remarkable advantage the duplex alloys have
over the comparable austenitic alloys with re-
gard to the temperatures at which they may be
used without SCC. The duplex alloys in this re-
gard are governed in their behavior by their fer-
rite matrix, through which cracks must propa-
gate (Ref 16). Ferritic stainless steels are known
for their resistance to SCC in the annealed con-
dition. The advantage of the duplex lies in their
composite-type microstructure with the crack-
arresting austenite phase and the toughening
fine grain structure. The duplex alloys show a
higher threshold stress for SCC as a percentage
of their yield strength (Fig. 29) than austenitic
alloys. This is in spite of their higher yield
strength, again giving these alloys more usable
strength.
Fig. 26 Critical pitting temperature (CPT) as a function
In ferrite, SCC susceptibility is a maximum
of NaCl concentration. SCE, saturated calomel
electrode. Source: Ref 5 below 100 °C, while in austenite susceptibility
appears to begin around 50 °C and increase
monotonically with temperature. The tempera-
ture at which SCC occurs at the fastest rate in-
creases with nickel content. This is also char-
acteristic of ferritic and martensitic materials
and mirrors their hydrogen embrittlement be-
havior. H2S also accelerates failure in chloride
environments (Fig. 30), and cold work acceler-
ates failure and lowers threshold stress values.
While duplex alloys behave in many regards
like ferritic alloys in their SCC or hydrogen
embrittlement response, they do not have the
same relationship between susceptibility and
bulk hardness. Other ferritic and martensitic
Fig. 27 Critical crevice temperature (CCT) and critical pit- alloys display pronounced susceptibility to
ting temperature (CPT). Source: Ref 15 these failures modes when their hardness
Fig. 28 Stress corrosion cracking (SCC) in neutral aerated NaCl. Testing duration 1000 hr. Source: Ref 5
Fig. 30 Suggested chloride and pH limits for cold-worked

duplex alloys. Source: Ref 17
exceeds Rc 22. The duplex alloys have an-

nealed hardness over Rc 30 without being in
danger. This probably simply indicates that
hardness as a measure of susceptibility is valid
only insofar as it reflects a certain yield
strength threshold as it does in tempered
martensite and is not valid for ferrite/austenite
composite structures. Thus, it is very important
to understand duplex SCC behavior as a sepa-
Fig. 29 Constant load stress corrosion cracking (SCC) tests in rate study and not interpret it in terms of
aerated MgCl2 at 150 °C. Source: Ref 5 austenitic or martensitic SCC.
REFERENCES 9. Avesta Sheffield, Corrosion Handbook for

Stainless Steels, 1994
1. P. Lacombe, B. Baroux, and G. Beranger, 10. B. Walden et al., Stainless Steel ’93,
Stainless Steels, Les Editions de Physique, Florence, AIM, 1993, p 3.47
2003 11. M.F. McGuire, MS&T Conf. Proc., 2004, p
2. R.N. Gunn, Duplex Stainless Steels, Abing- 831–846
ton Publishing, 1997, p 28 12. M. Ryan, D. Williams, R. Chater, B.
3. H. Nordberg H, Innovation of Stainless Hutton, and D. McPhail, Why Stainless
Steel, Conf. Proc., AIM, Florence, 1994, p Steel Corrodes, Nature, Vol 412, p 770
2.217–2.229 13. C.V. Roscoe et al., Duplex Stainless Steels
4. Charles, Duplex Stainless Steels ’91, Vol 1, ’86, The Hague, Nederlands Instituut voor
Beaune, Les Editions de Physique, 1991, p Lasteckniek, 1986, p 126–135
3–48 14. J.M. Drugli et al., Paper 270 presented at
5. S. Bernhardsson, Duplex Stainless Steels Corrosion ’90, Las Vegas, NACE, 1990
’91, Vol 1, Beaune, Les Editions de 15. S. Bernhardsson, Paper 164 presented at
Physique, 1991, p 137–150 Corrosion ’90, Las Vegas, NACE, 1990
6. J. Nichols J, 12th International Corrosion 16. T. Kudo, H. Tsuge, and A. Seki, Stainless
Congress, Houston, NACE, p 1237 Steel ’87, The Institute of Metals, 1988, p
7. E.-M. Horn, Werkstoffe und Korrosion, Vol 168–175
42, 1991, p 511–519 17. R. Francis, Duplex Stainless Steels ’94, Vol
8. J.P. Audouard, Stainless Steel Europe, April 3, Glasgow, TWI, 1994, paper KIV
1992, p 45
CHAPTER 8
Ferritic Stainless Steels
Summary technology was documented long before AOD

was invented (Ref 1). It was not until carbon
THE FERRITIC STAINLESS STEELS are and nitrogen levels were brought down to AOD
the lowest-cost highly corrosion- and oxidation- levels that it became truly practical for ferritic
resisting alloys in existence. They are useful alloys. The level of carbon plus nitrogen was
mainly as light-gauge sheet since their tough- lowered from around 0.10% to around 0.04%,
ness drops off rapidly for heavier sections. Even and less-expensive high-carbon ferrochromium
as they have grown in use more than any other could be used instead of expensive low-carbon
type of stainless, they could still economically versions. Thus, there exist two types of ferritics:
displace the popular but expensive 304 for the early high-carbon types such as 430, 434,
many routine applications. 436, and 446 and the more modern stabilized al-
loys led by 409 and 439.
The older, unstabilized grades are not always
Introduction fully ferritic. Their carbon levels cause them to
form some high-temperature austenite, which
Ferritic stainless steels are simplest, lowest- transforms to martensite if quenched. This
cost stainless steels. In their minimal form, they makes their welds brittle. To be used, they are
contain simply enough chromium to overcome normally in the annealed condition, which re-
their inherent level of carbon impurity and hit quires a lengthy subcritical anneal to avoid
the 11% chromium in solution required for martensite and to evenly distribute chromium
“stainlessness.” Early in the 20th century, 430 after all carbides have stably formed. The newer
came into being, and the attainable levels of stabilized alloys behave as if they are interstitial
carbon removal required 16% chromium for free. They are ferritic at all temperatures (ex-
this to occur. So much extra chromium was re- cluding for the moment the possibility of extra-
quired because during annealing, to develop the neous phases such as (α' and σ) and can be eas-
fully ferritic structure, carbon combines with ily welded without fear of unwanted phases.
chromium, rendering it useless as a corrosion Stabilization does not preclude excessive grain
fighter. In October 1967, the first commercial growth in the fusion or heat-affected zone
use of argon oxygen decarburization (AOD) (HAZ) of welds, which can render them brittle.
changed the world for ferritic stainless steel. The mechanical properties of ferritic stainless
This process, in which argon and oxygen are steels appear similar to austenitics strengthwise,
blown through the molten metal to selectively but they lack the ductility of austenitics, and
remove carbon without removing chromium they are limited at low temperatures by brittle-
(described in detail elsewhere in this book), re- ness and at high temperatures by softness.
duced the carbon plus nitrogen levels suffi- The lower thermal expansion coefficient of
ciently that their effect could be nearly negated ferritics makes their scale more compatible with
by small additions of titanium or niobium, the base alloy and provides them with a lesser
which combine strongly with carbon and nitro- tendency to spall. This makes them excellent
gen and effectively remove them from solution. for high-temperature applications with thermal
This process is called stabilization, and the cycles, provided their strength is adequate.
The corrosion resistance of ferritics is ham- A similar predecessor was 410S, a low-carbon
pered by their inability to utilize nitrogen. The version of 410 to which some understabilizing
absence of nickel, which characterizes these al- amount of titanium is added but that still re-
loys, is not a problem since nickel adds little to quires annealing for full ferritic properties. The
corrosion resistance. The titanium stabilization key issue of the 11% chromium ferritics is how
of the modern alloys has quite a beneficial effect to deal with carbon and nitrogen. The 405 and
since titanium is a powerful deoxidizer and 410S take the approach of minimizing it and live
desulfurizer, both of which can cause local with annealing. The 409 uses full titanium stabi-
chromium depletion and pitting. Ferritics, more- lization. The hidden problem with using only ti-
over, are essentially free from stress-corrosion tanium is that unless nitrogen levels are made
cracking (SCC) since they are below the thresh- very low, the amount of titanium required to
old hardness for hydrogen embrittlement in combine with it can reach levels at which the
body-centered cubic (bcc) ferrous alloys. There first TiN precipitates in the molten metal. This
are a few exceptions. slaggy precipitate agglomerates, causing casting
The main attraction of ferritic stainless steels problems and surface defects. This gave 409 a
over austenitics is their cost. The old compari- reputation as a grade unsuitable for applications
son of 430 versus 304 is a bit unfair since 304 is that required good appearance because the tita-
richer in chromium. A fair comparison might be nium streaks were difficult to avoid and greatly
between 439 and 304. The corrosion resistance highlighted by polishing. This has largely been
of these two alloys is barely distinguishable overcome by better refining techniques to reduce
under normal ambient conditions. They are both carbon plus nitrogen to levels below 0.02% and
very formable and weldable. The vast majority the use of dual stabilization by titanium and nio-
of the objects made commercially from 304 bium; 468 (UNS S40930) is such an alloy.
could be switched to 439 with no adverse con- The historical archetype of ferritic stainless
sequence. But, if nickel is selling for $7 per steels was 430, which has existed since the
pound, then the total cost of 304 versus 439 is 1920s and is still widely used. Its drawbacks are
doubled by its presence. No design engineer can lack of weldability, relatively poor corrosion re-
afford to ignore this level of incentive to learn sistance because so much of its chromium is
to use ferritic stainless steels. tied up as carbides, and modest formability. The
new archetype for this medium-chromium level
is 439. With 17% chromium and single (439) or
Ferritic Stainless Alloys dual stabilization (468), this alloy overcomes
the problems of 430 and can readily replace 304
The ferritic stainless alloys generally group in in most applications with significant cost sav-
low (10.5 to 12.0%), medium (16 to 19%), and ings. In North America, 439 is mainly used as a
high (greater than 25%) chromium. They can be higher-temperature automotive exhaust alloy,
stabilized or not. These distinctions are some- but in Europe 430Ti is used extensively in more
what imposed after the fact. Rather than giving visually challenging applications, such as appli-
them an order that they truly do not possess, the ances. There, it is generally used instead of 439
most significant alloys are all listed in Table 1 whenever the part can be designed to be formed
with their compositions. from it.
The low-chromium ferritic stainless steels Now, 434 and 436 are little used as their his-
began with the development of MF-1, the prede- torical application in automotive trim finds little
cessor of 409, in the 1960s. Its excellent corro- place in today’s automotive styling. A modern
sion resistance, compared to carbon steel; rela- offshoot of these alloys, which are basically
tively low cost; good welding; and formability molybdenum enhanced 430, is 444. This alloy
permitted it to replace aluminized carbon steel has roughly the corrosion resistance of 316L but
and cast iron in automotive exhaust systems, is fully resistant to SCC in the welded or an-
opening up what eventually became the largest nealed condition. This makes it especially useful
single market for stainless steel. It was made for applications such as hot water heaters, heat
possible by the very low carbon plus nitrogen exchangers, and food- and beverage-processing
levels the AOD process provided and the use of equipment.
stabilization. Thus, 409 was an improvement on Both the nominally 11 and 18% chromium
405 in which aluminum performed a quasi stabi- alloys are sometimes modified to enhance
lization, and low carbon suppressed martensite. their high-temperature strength or oxidation
Table 1 Ferritic stainless compositions
Alloy Designation C N Cr Ni Mn Si Mo P S Ti Nb Other
405 S40500 0.08 ... 11.5–14.5 0.6 1 1 ... 0.04 0.03 ... ... 0.10–0.30 Al
400 AK alloy 0.05 ... 12.0–13.0 ... 1 1 ... 0.03 0.03 ... ... 0.25 Al
409 S40900 0.08 ... 10.5–11.75 0.5 1 1 ... 0.45 0.45 6x(C + N) to 0.75 ... ...
409 S40910 0.03 0.03 10.5–11.7 0.5 1 1 ... 0.04 0.02 6x(C + N) to 0.5 0.17 ...
409 S40920 0.03 0.03 10.5–11.75 0.5 1 1 ... 0.04 0.02 8x(C + N) to 0.15–0.50 ... ...
409 ultraform AK alloy 0.02 0.02 10.5–11.7 0.5 0.75 1 ... 0.04 0.01 8x(C + N) ... ...
466 S40930 0.02 0.02 10.5–11.75 0.05 1 1 ... 0.04 0.01 0.8 + 8x(C+N) Ti + Nb ... ...
409Cb S40940 0.06 ... 10.5–11.7 0.5 1 1 ... 0.04 0.04 10xC to 0.75 Nb ... ...
409Ni S40975 0.03 0.03 10.5–11.7 0.5–1 1 1 ... 0.04 0.03 ... ... ...
11 Cr-Cb AK alloy 0.01 0.015 11.35 0.2 0.25 1.3 ... ... ... ... 0.35 ...
typical
12 SR AK alloy 0.02 0.015 12 ... ... ... ... ... ... 0.3 0.6 1.2 Al
typical
Alfa I ATI alloy 0.025 ... 13 ... 0.035 0.03 ... ... ... 0.4 ... 3 Al
typical
Alfa II ATI alloy 0.025 ... 13 ... 0.035 0.03 ... ... ... 0.4 ... 4 Al
typical
4724 Outukumpu 0.08 ... 13.5 ... 0.7 1 ... ... ... ... ... 1 Al
typical
429 S42900 0.12 ... 14.0–16.0 0.75 1 1 ... 0.04 0.03 ... ... ...
430 S43000 0.12 ... 16.0–18.0 0.75 1 1 ... 0.04 0.3 ... ... ...
430F S43020 0.12 ... 16.0–18.0 ... 1.25 1 ... 0.06 0.15 ... ... ...
min
430Se S43023 0.12 ... 16.0–18.0 ... 1.25 1 ... 0.06 0.06 ... ... 0.15 Se
430Ti S43036 0.1 0.04 16.0–19.5 1 1 1 ... 0.04 0.03 0.20 + 4x(C+N) to 1.10 ... 0.15 Al
439 S43035 0.07 0.04 17.0–19.0 0.5 1 1 ... 0.04 0.03 0.20 + 4x(C+N) to 1.10 ... ...
439LT S43932 0.03 0.03 17.0–19 0.5 1 1 ... 0.04 0.03 0.20 + 4x(C+N) to 0.75 Ti+Nb ... 0.15 Al
439 HP ATI, AK 0.01 0.01 17.5 0.2 0.35 0.45 ... 0.02 0.001 0.35 ... ...
439 ultraform alloys
468 S46800 0.03 ... 18.0–20.0 0.5 1 1 ... 0.04 0.03 Ti + Nb: 0.20 + 4x(C + N) to 1.10 ... ...
18 Cr-Cb AK alloy 0.02 ... 18 ... 0.3 0.45 ... ... ... 0.25 0.55 ...
typical
18SR AK alloy 0.015 ... 17.3 0.25 0.3 ... ... ... ... 0.25 ... 1.7
typical Al
4742 Outukumpu 0.08 ... 18 ... 0.7 1.3 ... ... ... ... ... 1
typical Al
(continued)
Chapter 8: Ferritic Stainless Steels / 111
Table 1 Ferritic stainless compositions
(continued)
Alloy Designation C N Cr Ni Mn Si Mo P S Ti Nb Other
434 S43400 0.12 ... 16.0–18.0 ... 1 1 0.75–1.25 0.04 0.03 ... ... ...
436 S43600 0.12 ... 16.0–18.8 ... 1 1 0.75–1.25 0.04 0.03 ... Nb + Ta 5xC 0.7 . . .
441, S44100 0.03 ... 17.5–18.5 ... 1 1 ... 0.04 0.015 0.1-0.6 9xC 0.3–1.0 . . .
4509,
430J1L
442 S44200 0.2 ... 18.0–23.0 0.6 1 1 ... 0.04 0.04 ... ... ...
436S ATI alloy 0.01 0.015 17.3 0.3 0.2 0.4 1.2 0.02 0.001 8x(C + N) min ... ...
typical
444, YUS S44400 0.025 0.035 17.5–19.5 1 1 1 0.75–1.25 0.04 0.03 Ti + Nb: 0.20 + 4x(C + N) to 0.80 ... ...
190-EM
433 ATI alloy 0.01 ... 20 0.25 0.3 0.4 ... 0.02 0.001 ... 10x(C + N) . . .
typical
4762 Outukumpu 0.08 ... 24 ... 0.7 1.4 ... ... ... ... ... 1.5
typical Al
453 ATI alloy 0.03 ... 22 0.3 0.3 0.3 ... 0.02 0.03 0.02 ... 0.60 Al
typical 0.1
REM
E-Brite, S44627 0.01 0.015 25.0–27.5 0.5 0.4 0.4 0.75–1.25 0.02 0.02 ... 0.5–0.20 0.2 Cu
26-1 0.5
Cu + Ni
Monit S44635 0.025 0.035 24.5–26.0 3.5–4.5 1 0.75 3.5–4.5 0.04 0.03 Ti + Nb: 0.20 + 4x(C + N) to 0.80 ... ...
Sea-cure S44660 0.025 0.035 25.0–27.0 1.5–3.5 1 1 2.5–3.5 0.04 0.03 Ti + Nb: 0.20 + 4x(C + N) to 0.80 ... ...
29-4C S44735 0.025 ... 28.0–30 0.5 1 0.75 3.5–4.5 0.04 0.03 Ti + Nb: 0.20 + 4x(C + N) to 0.80 ... ...
446 S44600 0.2 0.25 23.0–27.0 0.6 1.5 1 ... 0.04 0.03 ... ... ...
CC-50 Cast alloy 0.5 ... 26.0–30 4 1 1.5 ... ... ... ... ... ...
resistance. Again, the driving force has been changers and extensively in condensing por-
the requirements of the hot end of exhaust tions of high-efficiency residential furnaces.
systems (e.g., exhaust manifolds). Alloying The lower-alloyed Seacure had a slight tough-
with niobium and molybdenum adds to the ness advantage that permitted it to be used at
high-temperature strength, while additions of wall thicknesses of 1/16 in. when AL29-4C®
chromium, silicon, and aluminum increase ox- was too brittle. As with other ferritics, these al-
idation resistance. There exists an array of loys are generally only suitably tough when
proprietary alloys as shown in Table 1; these used in thin section size (i.e., less than several
are usually developed for specific automotive millimeters).
needs and employ all or some of these alloy- It is difficult to say ferritic stainless steels are
ing variations. The use of silicon and alu- underutilized since they account for about half
minum decreases formability and can acc- the world’s production of stainless, but there are
elerate (formation, so their use involves many applications in which more expensive
trade-offs. austenitic stainless steels are used needlessly.
Alloys with more than 20% chromium are Ferritic stainless steels are a viable alternative
used specifically for high-oxidation or corrosion to nickel-bearing austenitics when thickness is 2
resistance. Despite the relative lack of high- mm or less and drawing and bending instead of
temperature strength, these alloys are particu- stretch forming is permitted. There are many
larly useful because of their high-oxidation re- applications where the longer corrosion life of
sistance, which they derive from the tight low-chromium ferritics should economically re-
adherence of their oxide scale. The close match place carbon steel, as they have in automotive
between the thermal expansion coefficient of exhaust systems. There are no technical barriers
the scale and the alloy prevents spallation of the to obtaining these savings; design engineers
oxide, which would lead to breakaway oxida- need to learn how to use these alloys.
tion. This was the purpose of the earliest high-
chromium ferritic stainless, 446. The perform-
ance of 446 has been exceeded by lower alloyed Metallurgy of Ferritic Stainless Steels
grades, such as the aluminum-alloyed ferritics.
A prime example of the state of the art is 453, Chromium stabilizes the ferritic structure at
which has not only 22% chromium and 0.6% high temperatures. Thus, above about 11%
aluminum but also rare earths in trace amounts chromium, austenite does not exist at any tem-
(i.e., 0.1%). perature in pure iron chromium alloys, as seen
As in austenitic alloys, rare earths act as very in Fig. 1.
powerful oxide and sulfide formers that concen- However, iron-chromium alloys devoid of
trate at the metal-oxide interface and stabilize it, carbon are not practical, so early metallurgists
again preventing spallation. This type of alloys saw the diagram shown in Fig. 2 with the level
finds use in high-temperature applications such
as planar oxide fuel cells.
The high-chromium alloys, when used for
corrosion resistance, are usually called super-
ferritics. In the 1960s, E-Brite® was developed.
To obtain high toughness, it was vacuum refined
to very low carbon plus nitrogen levels. It was
followed by the more capable 29-4®. Later, this
alloy was stabilized and became the still-popu-
lar AL 29-4C®. (E-Brite now has a new life as a
fuel cell material based on its oxidation resist-
ance and very low thermal expansion coeffi-
cient.) These alloys saw success as replace-
ments for 316L when SCC was a problem. This
alloy and its close neighbor SeaCure® are used
primarily in tubing where corrosion resistance
is most important. It was developed for welded
condenser tubing where seawater or brackish
water is involved. It is also used in heat ex- Fig. 1 Iron-chromium phase diagram from Thermocalc
of carbon at 0.20%, which represented the pu- sufficiently long homogenization anneal at a
rity level attainable in arc furnace refining. Car- low enough temperature that carbon and nitro-
bon is essentially insoluble in ferrite at ambient gen have very little solubility. This is standard
temperatures, and carbides of chromium and in the processing of unstabilized ferritic stain-
iron will form to the extent carbon is available. less steels, such as 430. Rapid cooling of unsta-
Since carbon diffuses interstitially much more bilized alloys causes carbon and nitrogen to pre-
rapidly than chromium can substitutionally, cipitate within grains. This severely embrittles
chromium is combined in situ, especially along the material and does not avoid sensitization.
grain boundaries, which are fast-diffusion This is called high-temperature embrittlement
paths. This locally depletes chromium, and the because it comes from putting carbon and nitro-
alloy is sensitized. This can be eliminated by a gen into solution at a high temperature and then
causing it to precipitate in a harmful manner.
These alloys were only ferritic at room tem-
perature if they were given a subcritical anneal
to transform austenite to ferrite. Otherwise, at
room temperature they would be ferrite plus
martensite. There are alloys that are intended to
use a mixed ferrite/martensite structure, but
they are treated later as a variation from the nor-
mal ferritic alloys.
The introduction of AOD refining permitted
much lower levels of carbon, as seen in Fig. 3,
opening the door for fully ferritic stainless steels.
Carbon and nitrogen added together produce
about the same effect as carbon alone. So, un-
stabilized fully ferritic alloys are not feasible
below 20% chromium without extreme refining
techniques, such as electron beam refining,
which are not commercially viable for low-cost
Fig. 2 Iron-chromium phase diagram at 0.20% carbon alloys. Thus, nearly all modern ferritic alloys
Fig. 3 Iron-chromium diagram at low carbon levels Source: Ref 2

are “stabilized.” This means that a strong car- factors prevented TiC formation. The latter ef-
bide former such as titanium or niobium is fect was real in early austenitic alloys, such as
added in sufficient quantity to combine with all 321, leading to knife-line corrosion attack after
the carbon plus nitrogen, removing them from welding, but does not exist in low interstitial
solution. ferritic alloys, which have much greater diffu-
These reactions are simply: sion rates than austenitic alloys. But, since car-
bon mobility is quite high, it is not practical to
Ti + C = TiC (Eq 1) quench alloys quickly enough to prevent car-
Ti + N + TiN (Eq 2)
bide precipitation as is possible in austenitics
(detailed in the Chapter 6, “Austenitic Stainless
Nb + C = NbC (Eq 3) Steels”). Figure 4 shows the time-temperature-
transformation (TTT) curve for an unstabilized
Nb + N = NbN (Eq 4) 430-type alloy with carbon plus nitrogen of
Titanium is the stronger getter for carbon and 0.08% (Ref 3).
nitrogen. The thermodynamic driving force for Stabilization causes nonchromium carbides to
carbide and nitride formation is given by form at high temperatures, precluding chromium
carbide precipitation. The net effect is that
−7700 modern stabilized ferritic alloys behave as inter-
(Ti )(C) = + 2.75 (Eq 5)
stitial free and can be mapped using the pure
T
iron-chromium diagram shown in Fig. 1.
The rate of diffusion of carbon in ferrite is
−15790 around 100 times greater than that of carbon in
(Ti )( N) = + 5.40 (Eq 6)
T austenite. The solubility of carbon in ferrite is
vastly lower than it is in austenite. Because of
It must be noted that titanium has an even these factors, the heat treatments to avoid sensi-
higher affinity for oxygen and sulfur than for tization are essentially reversed. Carbon in
carbon, so that the removal of carbon from so- austenite can be retained in supersaturation for
lution is preceded by the removal of oxygen, extended periods of time. This is why austenitic
nitrogen, and sulfur in that order. This will be L grades do not sensitize even though they are
seen to have a major influence on corrosion re- slightly supersaturated. Sensitization occurs
sistance as the MnS inclusions generally asso- at higher levels of carbon by prolonged heating
ciated with the initiation of pitting are not at 600 to 850 °C. In ferritics, carbon cannot
found in titanium-stabilized grades of nor- be kept in supersaturation even by the most
mally low sulfur. In practice, the removal of rapid quenching, and sensitization is alleviated
oxygen begins in the molten state with the for- by prolonged heating in the 600 to 850 °C
mation of titanium sulfide and nitride and next range to allow chromium to equalize where
in the molten or solid state, depending on con- carbide precipitation has previously made it
centrations. It is desirable to keep sulfur and inhomogeneous.
nitrogen low enough that precipitation is in the
solid state so that precipitates do not agglom-
erate and cause large primary inclusions that
become unsightly surface defects. TiCS forms
in the solid state if sulfur is present; if not, TiC
forms. Essentially all carbon is removed from
solution below 1250 °C if carbon and nitrogen
are kept as low as possible and a stoichiomet-
ric amount of titanium is available (i.e., greater
than about four times the carbon plus nitrogen).
The stabilization formula in various specifi-
cations is more than four times the carbon plus
nitrogen because experimentally it has been
found that sometimes understabilization oc-
curs. This is due to the influence of oxygen and
sulfur having prior compound formation with
Fig. 4 430 time-temperature-transformation (TTT) curve. K,
the titanium and less importantly that kinetic carbideSource: Ref 3
Ferritic alloys, like austenitic alloys, can form temperature with the same structure as ferrite
intermetallic phases. The most prominent is σ, but with the chromium and iron atoms in an or-
which can be seen to form in higher-chromium dered bcc matrix in which iron and chromium
stainless steels (i.e., those with chromium plus occupy sites equivalent to two interlocking sim-
molybdenum of 20% or more). Formation of σ ple cubic matrices. Because the lattice so
occurs when such alloys are held between 500 closely matches that of ferrite, the precipitate is
and 800 °C; it is a hard, brittle tetragonal phase coherent and causes hardening. The α' embrit-
with equal parts iron and chromium. Thus, its tlement causes an extreme loss of toughness as
formation causes chromium depletion of the ad- well as hardening. It also causes a loss in corro-
joining ferrite. Formation requires substitutional sion resistance via the chromium depletion of
diffusion of chromium so is slower to form than that part of the matrix that surrenders chromium
carbides, minutes rather than seconds. Since to the α'-phase.
cold work enhances substitutional diffusion, it Figures 5 and 6 show the hardening effect of
accelerates σ formation. The σ forms preferen- α' and the resulting loss of toughness, respec-
tially along grain boundaries for diffusion rea- tively (Ref 6).
sons, and this causes it to have a major embrit-
tling effect. The σ may be redissolved by
solution annealing, but regaining full homo- Mechanical Behavior
geneity is not immediate.
Another embrittling phenomenon is the for- Ferritic stainless steels are quite similar in
mation of α'. This was named 885 °F or 475 °C their mechanical behavior to carbon steel. The
embrittlement before its cause was understood. main influence of chromium is to produce some
Before the nature of α' was known, it was con- solid solution hardening. Let us review the
fused with temper embrittlement, which occurs strengthening mechanisms of bcc iron. Pure
in martensitic alloys at the same temperature. iron is an extremely soft material with a yield
Temper embrittlement is the segregation of strength well under 10,000 psi. This softness is
phosphorus to prior austenitic grain boundaries not seen in practice because steel is never pure.
and does not occur in fully ferritic alloys. The α' Carbon has an extremely powerful effect on
is the ordered equiatomic chromium iron phase hardening, as does nitrogen.
that forms by spinodal decomposition; it has the The influence of substitutional alloying ele-
same composition as σ but exists at a lower ments is also quite significant. According to
Fig. 5 Influence of α' formation on hardnessSource: Ref 4

Fig. 6 Influence of α' formation on toughness Source: Ref 5
Paxton (Ref 7), the misfit of solute atoms causes

lattice strains proportional to the amount dis-
solved and provides strengthening through the
lattice friction term. This mechanism also in-
creases the impact transition temperature unfa-
vorably. Elements that produce a refining of
grain size are the exception to this general rule in
carbon steel, but the lack of an austenite-to-fer-
rite transformation in stabilized ferritic stainless
steels negates this benefit for them. Figure 7
shows that fairly common ingredients and impu-
rities have strong hardening effects (Ref 6). Man-
ganese and silicon are normally deoxidizers, but
in titanium-stabilized alloys, titanium takes over
the deoxidizing role so their presence can be lim- Fig. 7 Influence of substitutional elements on hardness of
iron alloys
ited. Phosphorus is virtually impossible to refine
from stainless steel, so its presence at around
0.02% is normally a given unless low-phospho- The softest ferritic stainless alloys are the
rus raw materials are used as a starting point. 409 variations made for highly formed exhaust
The worst toughness-inhibiting effects come system components. They contain as little
from interstitial elements to grain boundaries: manganese, silicon, nickel, and other substitu-
oxygen, carbon, and nitrogen. The effect of car- tional elements as possible and have a mini-
bon plus nitrogen on transition temperature is mum of carbon plus nitrogen, so that the re-
profound, as seen in Fig. 8 (Ref 8). sulting precipitate fraction after titanium
Stabilizing removes the interstitial carbon and addition is as low as possible. To maximize
nitrogen, along with oxygen and sulfur, from softness and formability, titanium and niobium
solution. This does not produce a major soften- in excess of that required for stabilization must
ing, however, because the precipitate itself has a also be minimized as they will cause solid so-
hardening effect. lution hardening.
Fig. 9 Corrosion of titanium-stabilized 29% Cr plus 4% Mo

alloys in ASTM A 763 Y test. Source: Ref 11
Fig. 8 Influence of interstitial carbon and nitrogen on tough-

ness transition temperature Source: Ref 8
Stabilization
Stabilization is essential to ferritic stainless

steels to avoid the precipitation of grain bound- Fig. 10 Corrosion of niobium-stabilized 29% Cr plus 4% Mo
alloys in ASTM A 763 Y test. Source: Ref 11
ary carbides. Combined carbon plus nitrogen
levels below 100 ppm are necessary to avoid
both sensitization and embrittlement, but with- As titanium and niobium are added to alloys,
out proper heat treatment even alloys of this pu- their corrosion resistance is improved (Figs. 9
rity can incur debilitating loss of toughness due and 10) (Ref 11). Maximum improvement in
to carbide and nitride precipitates (Ref 9). These corrosion resistance levels off once full stabi-
levels are not economically attainable for com- lization is reached. Excess amounts of the stabi-
mercial alloys, so stabilization is the correct en- lizing elements have negligible effect, but tita-
gineering answer. nium-stabilized alloys have a lower rate of
Stabilization is generally considered as the corrosion than niobium-stabilized alloys. This is
simple gettering of carbon and nitrogen by a probably due to titanium’s ability to eliminate
suitable carbide and nitride former. It was not sulfur and oxygen from solution.
known until about 1980 just what the mecha- Toughness improves for niobium-stabilized
nisms of embrittlement were in the ferritic alloys up through full stabilization and then be-
stainless steels, however. The distinguishing of gins to decline. This is a result of excess stabi-
α' from those related to interstitials and their lizing alloy acting as a solid solution hardener
stabilizers (Ref 10) permitted stabilizing ele- and therefore a toughness reducer. This tough-
ments to be optimized. ness reduction is more pronounced with tita-
Titanium combines with carbon and nitrogen nium, which is a stronger solid solution hard-
stoichiometrically by: ener (Figs. 11 and 12) (Ref 11).
The upshot of this understanding was the in-
Ti = 4 × C + 3.4 × N (Eq 7) troduction of dual stabilization, through which
both weld and base metal toughness and corro-
Niobium requires a greater weight percentage: sion resistance are optimized. The same study
recommended that dual stabilization follow the
Nb = 7.7 × C + 6.6 × N (Eq 8) following formula:
(Ti + Nb) ≥ 6 × (C + N) (Eq 9)
The toughness of these alloys has a broad op-

timum that takes advantage of the corrosion-re-
sisting benefits of titanium (Fig. 13) (Allegheny
Ludlum).
Other strong carbide formers such as zirco-
nium and vanadium are ineffective stabilizers
because their mobility at the temperatures at
which they are thermodynamically capable of
forming sufficiently large percentages of car-
Fig. 11 Charpy V-notch impact ductile to brittle transition bides and nitrides is too low to rid the matrix of
temperature (DBTT) of titanium-stabilized 29%Cr these elements. They also have too great a ten-
plus 4%Mo alloys test. Source: Ref 11
dency to form intermetallic compounds.
Toughness in ferritic stainless steels is a
major consideration. If ferritic alloys enjoyed
the same toughness as austenitic alloys, there
would be few instances when the use of the
much more expensive nickel-bearing grades
would be justified. Because stabilized alloys
are ferritic at all temperatures, there is no
automatic grain-refining transformation as ex-
ists in carbon steel. If grains grow large from
annealing at high temperatures or welding, then
the transition temperature increases. Section
size also has an effect. Stabilized ferritic stain-
less steels are seldom used in thicknesses of
Fig. 12 Charpy V-notch impact ductile to brittle transition over several millimeters because of decreasing
temperature (DBTT) of niobium-stabilized 29%Cr toughness. Figure 14 shows how transition
plus 4%Mo alloys test. Source: Ref 11
Fig. 13 Toughness of dual-stabilized low-alloy ferritic stainless. AL 466 is recognized as S40930

tion. But, the anisotropy does result in remark-

able drawing characteristics, with ferritic
stainless steels with elongations in tensile tests
in the mid-30% range being nearly equal to
austenitic stainless steels with over 50% elon-
gation.
The measure of anisotropy is the Lankford
ratio. It is expressed as:
r0 + 2r45 + r90
R= (Eq 10)
4
When this expression equals 1, then a mate-

rial is isotropic. As the value increases from 1,
the drawability increases, as measured by the
Fig. 14 Change in transition temperature with thickness for
limiting drawing ratio (LDR), the ratio of the
29Cr-4Mo-2Ni alloy. Source: Ref 12
diameter of a disk to that of the deepest cylinder
temperature can increase with thickness (Ref into which it can be drawn. The ferritic stainless
12). This effect is due simply to stress states steels in sheet form have LDRs of around 2.2
transitioning from biaxial to the more embrit- compared to 2.0 for 304.
tling triaxial with increasing thickness. The good formability of ferritic stainless
steels has some drawbacks. They are subject to
ridging, which is the formation of visible ridges
Texture and Anisotropy parallel to the direction of elongation. This is an
artifact of texture in the material. A combination
The deformation of ferritic bcc materials is of careful chemistry design and thermomechan-
characterized by limited slip systems, high ical processing is required to keep it under con-
stacking fault energy, and lattice anisotropy. trol. The approach centers on variables that in-
So, when ferritic stainless are deformed, dis- crease stored energy from deformation to
locations tend not to dissociate as they do in promote recrystallization over recovery during
austenitic stainless steels. The lack of dissoci- annealing.
ation of dislocations encourages cross slip. The ferritic stainless steels even carry for-
This minimizes dislocation tangles and work ward some of the preferred grain orientation
hardening. that come from initial solidification when
When ferritic stainless steels are deformed, growth of dendrites is along preferred crystallo-
certain crystallographic slip systems predomi- graphic directions. Hot working merely reori-
nate, so that large deformations mechanically ents these similarly oriented grains en masse.
bring different grains via rotation into closer Without phase changes or enough stored energy
crystallographic alignment. This preferred de- to provoke full recrystallization, randomness of
formation along easier slip planes results grain orientation is never achieved.
macroscopically in overall mechanical proper- Titanium-stabilized steels show more texture
ties varying with direction with respect to the and recovery versus recrystallization than do
prior deformation. Thus, ferritic stainless niobium-stabilized alloys. This is because tita-
steels, like low-carbon steels, have pronounced nium carbides and nitrides form at higher tem-
mechanical anisotropy. This is manifest in perature and are therefore coarser. They thus
their deep drawing characteristics. Heavily present less obstruction to dislocation motion
cold-rolled and annealed ferritic stainless than finer niobium precipitates. Furthermore,
steels draw quite well. They resist thinning. niobium precipitates tend to dissociate to a
When elongated, they contract in the width di- greater degree than those of titanium. This puts
rection while keeping virtually the same thick- niobium in solution during hot working where it
ness. This same phenomenon means that they can interact with dislocations. Thus, alloys at
cannot be stretch formed since plain strain least partially stabilized with niobium can
quickly results in fracture because of the re- achieve greater recrystallization, which can
sistance to deformation in the thickness direc- translate to finer grain size and less anisotropy.
Boron additions to ferritic stainless steels chromium and molybdenum. Likewise, other
result in the formation of grain boundary car- elements can have a negative effect. Any ele-
bides, M23(C, B)6. If added to titanium-stabi- ment that can combine with chromium or
lized steels, the carbides form on preexisting molybdenum can detract from corrosion resist-
TiN particles and result in coarser overall pre- ance by their removal of these essential ele-
cipitate arrays since finer, lower-temperature ments from solution. The most notorious of
precipitating TiC or TiCS precipitates are at these is carbon, whose tendency to form
least partially precluded. The net result is chromium carbides causes areas around such
coarser grain size and no major improvement carbides to be partially depleted of chromium.
in mechanical properties over the use of tita- However, nitrogen, oxygen, and sulfur can also
nium alone (Ref 13). Additions of boron to form chromium compounds and cause localized
niobium-stabilized steels does cause finer pre- loss of corrosion resistance. Manganese sul-
cipitates and grain size than would niobium fides, for instance, are almost always seen to be
alone (Ref 14). the locus of pitting corrosion (Ref 15). More
careful examination has shown that such sul-
fides grow in the solid state as chromium/man-
High-Temperature Properties ganese sulfides and deplete their very close sur-
roundings of chromium, inviting corrosion to
begin at the inclusion-matrix interface, where
High-temperature mechanical properties of
chromium levels in solution are reduced (Ref
ferritic stainless steels are often important to
16).
their successful use because their oxidation re-
Other factors that lead to loss of localized cor-
sistance is excellent and better than austenitics,
rosion resistance are the formation of
but their high-temperature strength is lower
chromium-rich phases such as α' and σ. Either
than that of austenitics. This has led to consider-
of these with about 50% chromium will cause
able development of high-temperature proper-
adjoining ferrite to have lower chromium levels.
ties, primarily for the automotive market. Re-
Because ferrite has a non-close-packed struc-
search has determined that high-temperature
ture, diffusion rates, both substitutional and in-
strength and creep resistance are best served by
terstitial, are about two orders of magnitude
stabilizing grain size and having niobium in
higher than in austenite. That means that any
solid solution. Adding titanium to niobium-sta-
deleterious chromium-depleting reaction can
bilized steels stabilizes the type of carbide, es-
happen more rapidly. Alloys cannot be
pecially preventing the formation of the coarse
quenched rapidly enough to forestall sensitiza-
M6C, whose growth decreases strength. The
tion, the precipitation of chromium carbides
relatively high insolubility of TiC causes this.
that depletes grain boundary regions of
Niobium is concurrently made available for
chromium. Instead, carbon must be neutralized
high-temperature solid solution strengthening.
by stabilization, or the chromium depletion
must be removed by homogenization in long-
box anneals. Note that the latter technique is
Corrosion and Oxidation Resistance also possible in austenitics but would require
annealing for excessively long times, 102 h or
Corrosion resistance is chemistry dependent so.
rather than structure dependent, so ferritic stain- The ferritic stainless steels are valued for
less steels behave just as do other stainless their resistance to SCC. Even in environments
steels of the same crucial alloy content. that cause pitting, the normal initiation step for
The main alloying elements that provide re- SCC, annealed ferritic stainless steels do not
sistance to localized corrosion, general corro- undergo SCC as long as alloying elements such
sion, and crevice corrosion are chromium, as nickel, copper, and cobalt are kept below
molybdenum, and nitrogen. Since nitrogen is 0.5% in aggregate. Cold work sufficient to raise
essentially insoluble in ferrite, it cannot con- their hardness above Rc 20 to 22 can make them
tribute to the corrosion resistance of ferritic susceptible to both SCC and its cousin, hydro-
stainless steels as it can in austenite. Other al- gen embrittlement. The more highly alloyed su-
loying elements, such as copper and nickel, can perferritic alloys are even susceptible to hydro-
add to corrosion resistance in special cases, but gen embrittlement in the annealed condition
they are of secondary importance compared to (Ref 17). As with martensitic stainless steels,
this susceptibility is a maximum near room tem- 10. J. Grubb, R. Wright, and P. Farrar, “Micro-
perature and declines with increasing tempera- mechanisms of Brittle Fracture in Titanium-
ture, as opposed to austenitics, which see their Stabilized Stainless Steels,” Special Publi-
maximum susceptibility above room tempera- cation 706, ASTM, 1980
ture. This limits these alloys’ ability to employ 11. J. Grubb, Stabilization of High-Chromium
cathodic protection safely to –0.80 Vsce, at Ferritic Stainless Steels, Proc. Int. Conf.
which point corrosion in seawater is, if not Stainless Steels, ISIJ, Chiba, 1991
eliminated, reduced to very low levels (Ref 18). 12. M.A. Streicher, Stainless Steel ‘77, p 27
13. E. El-Kashif, K. Asakura, T. Koseki, and
K. Shibata, ISIJ Int., Vol 44, 2004, p
REFERENCES 1568–1575
14. N. Fujita, K. Ohmura, E. Sato, and A.
1. F.M. Beckett and R. Franks, Trans AIME, Yamamoto, Nippon Technical Report 71,
Vol 113, 1934, p 126–143 Oct 1996
2. Stainless Steel, Les Editions de Physiques, 15. T. Suter, E. Webb, H. Bohni, and R. Alkire,
1992, p 483 Pit Initiation in I M NaCl With and Without
3. Stainless Steels, Les Editions de Physique, Mechanical Stress, J. Electrochem. Soc.,
2003 Vol 148 (No. 5), 2001, B174
4. H.D. Newell, High Chromium Irons, Met. 16. M. Ryan, D. Williams, R. Chater, B. Hutton,
Prog., April 1947, p 617–626 and D. McPhail, Why Stainless Steel Cor-
5. P.J. Grobner, The 885 °C (475 °C) Embrit- rodes, Nature, Vol 412, 2002, p 770
tlement of Ferritic Stainless Steels, Metall. 17. J. Grubb, “Hydrogen Embrittlement of Su-
Trans., Vol 4, 1973, p 251–260 perferritic Stainless Steels,” paper presented
6. Handbook of Stainless Steels, Peckner and at 1984 ASM Int’l Conference on New De-
Bernstein, McGraw Hill, 1977, p 5–9, 5–12 velopments in Stainless Steel Technology,
7. H.W. Paxton, Alloying, ASM, 1998, p 213 Detroit, September 1984
8. H. Abo et al., Stainless Steel ‘77 18. J. Grubb and J. Maurer, “Use of Cathodic
9. J. Grubb and R. Wright, The Role of C and Protection With Superferritic Stainless
N in the Brittle Fracture of Fe-26Cr, Met. Steels in Seawater,” paper presented at Cor-
Trans. A, Vol 10A, Sept 1979, p 1247–1255 rosion 84, New Orleans, April 1984
CHAPTER 9
Martensitic Stainless Steels
Summary which restricts the temperature and composition

ranges over which it is possible to obtain a fully
THE SMALLEST CATEGORY of stainless austenitic structure from which to form marten-
steels in usage volume is the martensitic stain- site. The presence of ferrite in a martensitic
less steels. This is mainly because these alloys structure is detrimental to strength, hardness,
are limited in corrosion resistance because of and toughness. Ferrite can appear in the as-cast
the necessity of keeping alloy levels low to pro- structure and be formed during austenitizing or
duce the martensite structure. Even so, they fill tempering. All the usual concerns inherent in
an important niche as a strong, hard, and tough any martensitic alloys are still present; temper
alloy of fairly good corrosion resistance and as embrittlement, retained austenite, etc.
a strong, stable, high-temperature alloy. Martensitic stainless steels are the most mar-
ginally corrosion resistant of all the stainless al-
loys. The requirement that they be fully austeni-
Introduction tizable limits the amount of corrosion-resisting
chromium and molybdenum they can contain.
Nearly 100 years ago cutlery was first sold in Much of the carbon in them detracts from the ef-
Great Britain with a composition of 13% fective chromium content by forming chromium
chromium and 0.25% carbon. This was the first carbides. In addition they are always susceptible
commercial use of stainless steel and cutlery to stress corrosion cracking (SCC) when their
with the same basic analysis is still sold today. hardness exceeds about Rc 22. These limitations
The useful alloys of martensitic stainless steel combine to make their excellent properties us-
contain from roughly 11 to 18% chromium and able in only mild environments compared to
up to 1.0% carbon. Relatively small amounts of other stainless steels. Their high strength and
nickel, molybdenum, tungsten, vanadium, and hardness for their relatively low cost ensure their
niobium are also added at times for specific pur- place as a very useful engineering material.
poses explained in this chapter. Those marten- Table 1 lists the most significant of the
sitic stainless steels in which elements such as martensitic stainless steel alloys. The reader
copper and titanium are added to produce addi- should be aware that some alloys which are
tional hardening through precipitation are dis- quite similar are discussed primarily in other
cussed in Chapter 4, “Corrosion Types.” chapters dealing with specifically PH stainless
The designers and engineers already familiar steels or primarily ferritic stainless steels. The
with martensitic carbon and alloy steels will distinction between martensitic stainless steels
find nothing confusing about martensitic stain- and some other stainless alloy families is some-
less steels. There is no aspect of martensitic times vague. Nearly all the precipitation-hard-
steels that does not apply directly to stainless ening stainless steels are used in the martensitic
martensitic steels. The additional concerns one state, but their special hardening mechanism of
must have with stainless martensite relate precipitation within a martensitic matrix causes
mainly to those that are due to the strong ferri- them to be categorized separately somewhat ar-
tizing influence of chromium. Chromium bitrarily. By this conventional logic, some of the
strongly promotes the formation of ferrite, martensitic alloys containing molybdenum or
Table 1 Compositions (wt%) of martensitic stainless steels
Alloy UNS Form C Mn S Si Cr Mo Ni Other Other
403 S40300 Wrought 0.15 max 1 0.03 0.5 11.5-13.5 ... ... ... ...
410 S41000 Wrought 0.15 max 1 0.03 1 11.5-13.5 ... ... ... ...
410S S41003 Wrought 0.03 1 0.03 ... 10.5- ... 1.5 0.03 ...
12.5 N
410 S41008 Wrought 0.08 1.5 0.03 1 11.5- 0.6 ... ... ...
13.5
410Cb S41040 Wrought 0.18 max 1 0.03 1 11.5- ... ... 0.05-0.30 ...
13.5 Nb
412 S41003 Wrought 0.030 max 1.5 0.03 1 10.5- ... 1.5 0.030 max ...
12.5 N
414 S41400 Wrought 0.15 max 1 0.03 1 11.5-13.5 ... 1.25-2.50 ... ...
414 mod S41425 Wrought 0.05 0.5-1.0 0.005 0.6 12.0- 1.5- 4.0- 0.06-0.12 0.30 Cu
15 2 7 N
415 S41500 Wrought 0.05 max 0.50-1.00 0.03 0.6 11.5-14.0 0.50-1.00 3.50-5.50 ... ...
416 S41600 Wrought 0.15 max 1.25 0.15-0.30 1 12.0-14.0 ... ... ... ...
416Se S41623 Wrought 0.15 max 1.25 0.06 1 12.0- ... ... 0.15 min ...
14 Se
418 S41800 Wrought 0.15-0.20 0.5 0.03 0.5 12.0-14.0 ... 1.80-2.20 2.50-3.50 W ...
420 S42000 Wrought 0.15 min 1 0.03 1 12.0-14.0 ... ... ... ...
4116 DIN Wrought 0.5 ... ... ... 14.5 0.65 ... ... 0.15 V
1.4116 Nominal
420F S42020 Wrought 0.15 min 1.25 0.15 min 1 12.0-14.0 0.6 ... ... ...
420FSe S42023 Wrought 0.15 min 1.25 0.06 1 12.0-14.0 0.6 ... 0.15 min Se ...
422 S42200 Wrought 0.20-0.25 1 0.03 0.75 11.0-13.5 0.75-1.25 0.50-1.00 0.75-1.25 W 0.15-0.30V
424 S42400 Wrought 0.06 max 0.50-1.00 0.03 0.30-0.60 12.0-14.0 0.30-0.70 3.50-4.50 ... ...
425 S42500 Wrought 0.08-0.20 1 0.01 1 14.0-16.0 0.30-0.70 1.00-2.00 ... ...
425mod ... Wrought 0.50-0.55 1 0.03 1 13.0-14.0 0.80-1.20 0.5 ... ...
Trinamet ... Wrought 0.30 max 1 0.03 1 12.0 1.00 ... 2.00 ...
-14 -3 -3.00 Cu
HP13Cr-1 JFE Wrought 0.025 0.45 ... ... 13 1 4 ... ...
Nominal
HP13Cr-2 JFE Wrought 0.025 0.45 ... ... 13 2 5 ... ...
Nominal
NT-CRS Nippon Wrought 0.03 1.45 ... ... 12.7 1.4 4.5 1.5 Cu 0.040 N
Nominal
(continued)
Table 1 (continued)
Compositions (wt%) of martensitic stainless steels
NT-CRSS Nippon Wrought 0.02 2 ... ... 12.3 2 5.8 1.5 Cu 0.015 N
Nominal
KL-12Cr JFE Wrought 0.01 ... ... ... 11 ... 2.4 0.5 Cu 0.010 N
Nominal
KL-HP JFE Wrought 0.01 ... ... ... 12 2 5.5 ... 0.010 N
12Cr Nominal
431 S43100 Wrought 0.20 max 1 0.03 1 15.0-17.0 ... 1.25-2.50 ... ...
440A S44002 Wrought 0.60-0.75 1 0.03 1 16.0-18.0 0.75 ... ... ...
440B S44003 Wrought 0.75-0.95 1 0.03 1 16.0-18.0 0.75 ... ... ...
440C S44004 Wrought 0.95-1.20 1 0.03 1 16.0-18.0 0.75 ... ... ...
440F S44020 Wrought 0.95-1.20 1.25 0.10-0.35 1 16.0-18.0 0.40-0.60 0.75 ... ...
440FSe S44023 Wrought 0.95-1.20 1.25 0.03 1 16.0-18.0 0.6 0.75 0.15 min Se
BG-42 Nominal Wrought 1.15 ... ... 0.3 14.5 4 ... 1.2 V ...
ATS-34 Nominal Wrought 1.05 0.4 ... 0.35 14 4 ... ... ...
14-4 Nominal Wrought 1.05 0.5 ... 0.3 14 4 ... ... ...
CrMo
154 CM Nominal Wrought 1.05 0.45 ... 0.3 14 4 ... ... ...
CPM Nominal PM 1.45 ... ... ... 14 2 ... 4.0 V ...
S30V
CPM Nominal PM 2.15 0.4 ... ... 17 0.4 ... 5.5 V ...
S60V
CPM Nominal PM 2.2 ... ... ... 13 1 ... 9.0 V ...
S90V
CA-15 J91150 Cast 0.15 max 1 0.04 1.5 11.5-14.0 0.5 1 ... ...
CA 15M J91151 Cast 0.15 max 1 0.04 0.65 11.5-14.0 0.15-1.00 1 ... ...
CA-40 J91153 Cast 0.20-0.40 1 0.04 1.5 11.5-14.0 0.5 1 ... ...
CA-40F J91154 Cast 0.20-0.40 1 0.020-0.040 1.5 11.5-14.0 0.5 1 ... ...
CB-6N J91650 Cast 0.06 max 0.5 0.02 1 10.50-12.50 1 ... ...
CB-6MN J91540 Cast 0.06 max 1 0.03 1 11.50-14.00 0.40-1.00 3.50-4.50 ... ...
CA-28MVW J91422 Cast 0.20-0.28 0.50-1.00 0.03 1 11.00-12.50 0.90-1.25 0.50-1.00 0.90-1.25 W 0.20-0.30 V
Chapter 9: Martensitic Stainless Steels / 125
tungsten should also be considered precipitation- on temperature. The amount is given by the
hardening alloys, but they customarily are not Koistinen and Marburger equation (Ref 1):
and will not be in this work.
The ferritic alloys often have compositions 1− Vα ′ = exp{β (Ms − T )}
(Eq 1)
that allow them to be partially martensitic under
some conditions. 430 (UNS S43000) and The martensite is coherent with the parent
3CR12 (UNS S41003) can contain some austenite and resembles the passage of slip dis-
martensite if their heat treatment is such that locations through the crystal. The sum of many
austenite is allowed to form and is followed by such dislocations is shear, and this can be
rapid cooling. Even 409 (UNS S409XX) can macroscopically visible as in Fig. 1.
form some austenite if chromium is at the high The formation of martensite is essentially me-
end of its possible range and nickel and man- chanical (i.e., via deformation, not diffusion).
ganese residual levels are high. The martensitic The shear and volume expansion, about 4%,
alloys themselves can be made to be partially which accompanies the transformation, in-
ferritic by forcing their carbon contents to low volves a great deal of strain energy that must be
levels as is customarily done with 410S (UNS taken into account. This is shown diagrammati-
S41003). Not understanding these alloys can cally in Fig. 2 (Ref 2).
lead to unexpected consequences in mechanical
properties or corrosion performance.
Martensite Formation
Martensite as a phenomenon deserves a brief

review. Martensite forms as result of the diffu-
sionless transformation of austenite. The
austenite may be supersaturated with carbon or
nitrogen, but that is not necessary for the trans-
formation. The driving force for the transforma-
tion is simply the much lower free energy of the
ferrite phase over the austenite phase, which
can be attributed largely to large mutual repul-
sion between iron atoms that possess unpaired
outer electrons with the same quantum number
and magnetic polarity. This free-energy differ-
ential increases with decreasing temperature. At
a certain temperature, the martensite start tem- Fig. 1 Martensite platelets emerging from the surface.
Source: Ref 2
perature Ms, the transformation occurs sponta-
neously via the coordinated movement of atoms
in a shearing-type mode at very high speeds ap-
proaching the speed of sound in the material.
The composition of the martensite is identical to
that of the parent austenite.
There is regularity to the relationship between
the parent austenite and the martensite.
Greninger and Troiano determined that the
close-packed planes of the austenite {111} var-
ied from the {011} of the martensite by only
0.2°. Further, the direction of the <101 bar> of
the austenite was only 2.7° from the <1 bar 11
bar> of the martensite. These relationships de-
fine the habit plane that constitutes the austenite
martensite boundary.
Martensite forms essentially independent of Fig. 2 The martensite reaction ab contrasted to the nucleation
time and the fraction transformed depends only and growth-type transformation of austenite to ferrite, ac
This energy differential between ferrite and stitial sites and forms various carbides, leaving
martensite is stored in the high-strain energy the parent martensite less strained, softer, and
matrix. Applied strains affect the transforma- tougher. Figure 4 shows that the large strain en-
tion. Indeed, metastable austenite can readily be ergy in martensite varies with the carbon con-
transformed to martensite by deformation. tent, and Fig. 5 (Ref 4) shows how hardness
However, the untransformed austenite is hin- varies with carbon content.
dered from transforming by the compression it Nitrogen behaves similarly to carbon in both
receives from the already-formed martensite. austenite and martensite, but its solubility is
Thus, some residual austenite is commonly lower, and it is less significant as an alloying el-
found between lathes of martensite. ement accordingly. Hydrogen and boron, as in-
At the Ms temperature, the body-centered terstitials, also raise hardness.
cubic (bcc) phase becomes preferable energeti-
cally, but this temperature is too low for diffu-
sion transformation, and a slight shear in the Phase Structure
austenite lattice causes a rearrangement of the
atoms from a face-centered cubic (fcc) to a dis- Figure 6(a) to (h) shows a series of photomi-
torted bcc structure. The amount of distortion is crographs of various martensitic alloys (Ref 5).
proportional to the amount of carbon in the in- A stainless martensitic alloy should have the
terstices of the structure. These interstices are following characteristics:
considerably smaller in the bcc structure even
though it is expanded from the fcc. The octahe- • It must have at least 10.5% chromium to
dral sites change from 2.86 by 3.56 A to 2.86 by qualify as stainless and even more for better
2.86 A, as shown in Fig. 3 (Ref 3). The distor- corrosion resistance.
tion is accommodated by accommodation from • It should be fully austenitic at some temper-
site to site at low carbon levels, but above about ature.
0.018% carbon this can no longer be accommo- • The temperature at which austenite forms on
dated and a tetragonal distortion occurs (Ref 3). heating should be sufficiently high to permit
The carbon is in a state of supersaturation in the tempering above the temper embrittlement
as-formed martensite. When the martensite is range.
tempered, the carbon diffuses from these inter- These criteria are somewhat challenging. Fig-
ure 7(a) shows that at low-carbon (0.05%) lev-
els austenite is stable up to about 12%
chromium, above which some δ-ferrite tends to
be stable at all temperatures below the melting
point. Increasing carbon slightly expands the
chromium level at which full austenitization can
occur (Fig. 7b) (Ref 3).
Fig. 3 Change in size of the octahedral interstitial site with

the change from face-centered cubic (fcc) to body- Fig. 4 Strain energy of martensite dependence on carbon
centered cubic (bcc). Source: Ref 2 content. Source: Ref 2
Fig. 5 Variation in martensite hardness with carbon content
The interplay between chromium and carbon The variety of martensitic stainless steels
is further explained in Fig. 8(a) and (b), in would be very limited if only chromium and
which it becomes clear that for higher- carbon were available as alloying elements, but
chromium alloys the range over which full fortunately nickel again can make an important
austenitization can occur is further restricted. contribution. Nickel greatly expands the
Fig. 6 (a) Annealed 410 showing carbides within an equiaxed ferrite matrix. (b) 410 quenched and tem-
pered. (c) 416 quenched and tempered: white ferrite and gray sulfides in a martensite matrix. (d) 420
quenched and tempered showing fine carbides in a martensite matrix. (e) 420 quenched and tempered showing
surface decarburization. (f) 440A annealed displaying primary and smaller secondary carbides in a ferrite ma-
trix. (g) 440B quenched and tempered displaying both primary and secondary carbides. (h) 440C quenched and
tempered displaying significant primary carbides plus finer secondary carbides in a martensite matrix
Fig. 7 Iron-chromium phase diagrams at two low-carbon levels
1800 1800
1600 Liquid 1600 Liquid
L+α
1400 1400
Temperature, °C
Temperature, °C
σ
α+γ
γ γ
1200 1200 α+γ
γ + carbide
1000 1000
γ + carbide
800 800
α + carbide α + carbide
600 600
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Mass, %C Mass, %C
Fig. 8 (a) Iron-chromium phase diagram at 12% chromium; (b) iron-chromium diagram at 17% chromium
chromium levels and temperatures at which amount of total alloy that can be used and in the
austenite is stable as is shown in Fig. 9 (Ref 3). end puts an upper limit on the ability of marten-
Table 2 quantifies the influences of the vari- sitic stainless steels to achieve high corrosion re-
ous possible alloying elements on the key prop- sistance. This is because as the main corrosion
erties of martensitic stainless steels. fighters, chromium and molybdenum, which are
It can be seen that the elements that promote ferritizers, are increased, so must austenitizers
austenite, with the exception of cobalt, all de- such as nickel. The coordinated increase in these
press the Ms temperature. This puts a limit on the elements lowers the Ms to such a degree that the
alloys become stably austenitic before much Austenitizing is complicated in martensitic

higher corrosion resistance is obtained. stainless steels because many grades contain
The ability to temper without austenite rever- carbon at levels intended to produce carbides
sion is an important trait. Obviously, if transfor wear resistance purposes. Since carbon solu-
forming martensite to austenite during temper- bility varies strongly with temperature at
ing caused subsequent untempered martensite austenitizing temperatures (Fig. 7a and b), con-
or other undesirable phases, this would limit trol of temperature is vital to have the correct
one’s ability to temper at a high enough temper- balance of carbon in solution versus carbon as
ature to achieve desired toughness. This limits carbide since carbon in solution has such a
the use of nickel while encouraging the use of strong influence on ferrite content, Ms, and me-
elements like molybdenum. chanical properties.
Copper has become an important alloying el- Austenitizing temperature also determines
ement in martensitic stainless steels because it austenite grain size. This affects Ms, but more
greatly improves corrosion resistance in certain importantly it influences subsequent toughness.
environments without diminishing an alloy’s Phosphorus precipitates at prior austenite grain
ability to be tempered. boundaries during tempering with a maximum
effect at 475 °C. This is the infamous temper
embrittlement. Figure 10 (Ref 3) shows the sig-
Thermal Processing nificant toughness change that occurs as in-
creasing austenitizing temperature increases
The main concerns with processing marten- austenite grain size and permits greater phos-
sitic stainless steels are austenitizing, quench- phorus concentrations at grain boundaries.
ing, tempering/stress relieving, and annealing. Refining phosphorus from any chromium-
containing steel is quite challenging thermody-
namically, so achieving low phosphorus levels
depends mainly on restrictions on raw materials
for melting. Because this is difficult or costly,
grain size control is the main tool for control-
ling temper embrittlement.
The higher-carbon grades, those above 0.20%
carbon, should be heated gradually through
stage heating to avoid cracking due to thermal
stresses. Soaking at 800 °C until uniform tem-
perature is achieved minimizes this risk.
Another concern during austenitizing is su-
perficial carbon loss, an example of which is
shown in Fig. 6(e). Heating in air to 1050 °C
can cause surface carbon to decrease by approx-
imately 0.10% per hour, resulting in much
lower surface hardness. This loss increases with
base carbon level and austenitizing temperature.
Carbon or nitrogen pickup could also occur if
the atmosphere was rich in these elements. The
carbon potential of the furnace atmosphere must
be controlled to avoid potentially serious prob-
Fig. 9 The expansion of the range of austenite stability with lems. If hydrogen atmospheres are used the
nickel content
Table 2 Influence of alloying elements on ferrite, Ms, and austenite start

Element N C Ni Co Cu Mn Si Mo Cr V Al
Lowering of % –220 –210 –20 –7 –7 –6 6 5 14 18 54
ferrite per % element
Lowering of MS –475 –475 –17 0 to 10 –17 –30 –11 –21 –17 –46 ...
per % element
Change of AC 0 to 280 0 to 250 –30 to –115 0 0 –25 to –66 25 to 73 25 to 60 0 to 35 50 to 290 30 to 750
per % element
+120
P: 0.047%
[ ]
P: 0.035%
+80
[ ]
P: 0.021%
Transition temperature, °C
+40
P: 0.003-0.004%
[ ]
0 [ ]
[ ]
−40 [ ]
−80
5 10 20 50 100 200
Austenite grain size, μm
Fig. 10 Influence of austenite grain size and phosphorus level on toughness
danger of embrittlement after quenching must as the transformation of M3C into M7C3 can re-
be recognized. Stress relief without delay would sult in a secondary hardening, a true precipita-
be mandatory. tion-hardening effect. In the presence of strong
The high chromium content of these alloys carbide-forming alloying elements such as
renders them very deep hardening. Air harden- molybdenum, vanadium, and tungsten, the M2X
ing is generally sufficient. Oil quenching which carbide can become the more stable species and
is faster may be slowed by heating the oil. be responsible for the secondary hardening. At
Avoiding quench cracking and excessive 500 °C, coarser M23C6 and M7C3 begin to grow
warpage is almost always a greater concern than at grain boundaries. This is accompanied by a
depth of hardening so air quenching is standard. pronounced softening. The hardening reduction
Because the quenching and the transforma- with stress relief and tempering for a 12% Cr
tion it causes are inevitably accompanied by alloy is shown in Fig. 11 (Ref 6).
residual stresses in a brittle material, stress re- Separately at the 475 °C range, the previously
lieving should be immediate to avoid cracking. mentioned phosphorus segregation to prior
Higher-carbon grades should not even be al- austenitic grain boundaries occurs. This effect
lowed below room temperature before stress begins to disappear above 550 °C. Thus true
relief. Pickling should never be done on as- tempering is conducted above this temperature.
quenched material because this could easily re- The microstructural changes at these tempera-
sult in hydrogen uptake and delayed cracking tures are the above-mentioned loss of carbon
by hydrogen embrittlement. from solid solution, carbide precipitation and
Heating as-quenched material to between 150 coarsening, and, of course, stress relief. The re-
and 400 °C produces stress relieving. Besides sult is a pronounced softening and toughening.
the normal flow on a microscopic scale, which If the material contains retained austenite, it
we understand as stress relieving, there is a may decompose to ferrite and carbide with a
slight growth in the number of fine cementite negative effect on toughness.
particles and a corresponding decrease in the The molybdenum, vanadium, and tungsten-
amount of carbon in solid solution. This results alloyed grades will resist softening during tem-
in a slight decrease in hardness. At 400 °C, a pering because of the strength of the secondary
further precipitation of M2X and M7C3 as well hardening they undergo due to precipitation
500
Initial hardness
450 300°C
300°C 350°C
350°C
400°C
400°C
400 450°C 450°C
500°C
550°C
350
600°C
Hardness
650°C
300 700°C
750°C
500°C
550°C
250
600°C
650°C
200 700°C
750°C
150
11 12 13 14 15 16 17 18 19 20 21 22 23
T (20 + LOG t) × 10−3
Fig. 11 Influence of tempering on hardness
hardening of carbides and nitrides. Nickel most sophisticated martensitic stainless alloys.
seems to amplify this action by its influence on While 420 is the common alloy and is quite
diminishing the solubility of carbon in the ma- serviceable, much more wear- and corrosion-re-
trix. Thus, the tempering of the higher-alloy sistant alloys exist. At one time, 440C was the
martensitic stainless steels can truly be consid- maximum step up from 420; however, further
ered a precipitation-hardening reaction. alloying with molybdenum for corrosion resist-
The higher-carbon, higher-chromium grades ance and vanadium for hardness of the carbide
are typically only stress relieved because the re- phase has led to improvements. The wear resist-
moval of chromium from solution by carbide ance of a blade is largely determined by the
formation at higher temperatures causes an un- hardness and amount of carbides while the
acceptable loss of corrosion resistance. toughness is governed by the matrix properties.
These alloys are used at very high hardness
levels, so cleanliness is very important to tough-
Applications ness, which measures the ability to withstand
chipping in use. Electroslag remelting (ESR) or
High-Temperature Use. The basic 12 % Cr vacuum induction melting-vacuum arc remelt-
martensitic alloy has been the basis of alloying ing (VIM-VAR) provides the cleanliness re-
improvements that were done to produce better quired, while powder metallurgy is optimal for
high-temperature performance, especially for obtaining very fine carbide size and uniformity.
turbines. The nominal analyses of some prominent
The addition of vanadium and niobium, both grades are shown in Table 3.
of which form much more stable carbides than The martensitic alloys have a tendency to-
chromium, results in alloys that have vastly im- ward centerline segregation during solidifica-
proved creep resistance in the 550 °C range, as tion as well as toward the formation of primary
shown in Fig. 12 (Ref 3). carbides. This has produced limitations in the
Tool and Cutlery Alloys. A high-profile use amount of highly wear-resistant constituents
of martensitic stainless steels is in cutlery. Hunt- such as vanadium carbide (hardness Rc 75),
ing knives, sport knives, and chefs’ cutting tools which can be introduced into the matrix in con-
are highly valued items and contain some of the ventional production. Powder metal techniques
4
50
2
45
40
Applied stress, kg/mm2
35
3
30
1
25
20
1 (0.2C-10.5Cr)
2 (0.2C-10.5Cr-0.1Nb)
3 (0.2C-10.5Cr-0.1V)
16
4 (0.2C-10.5Cr-0.1V-0.1Nb)
3 10 30 100 300 1000 3000 10,000
Rupture life, h
Fig. 12 Influence of vanadium and niobium on high-temperature properties
Table 3 Tool and cutlery martensitic stainless steels alloy compositions

420 S42000 Wrought 0.15 min 1 0.03 1 12.0-14.0 ... ... ... ...
4116 DIN 1.4116 Wrought 0.5 ... ... ... 14.5 0.65 ... ... 0.15 V
Nominal
440A S44002 Wrought 0.60-0.75 1 0.03 1 16.0-18.0 0.75 ... ... ...
440C S44004 Wrought 0.95-1.20 1 0.03 1 16.0-18.0 0.75 ... ... ...
BG-42 Nominal Wrought 1.15 ... ... 0.3 14.5 4 ... 1.2 V ...
ATS-34 Nominal Wrought 1.05 0.4 ... 0.35 14 4 ... ... ...
14-4 Nominal Wrought 1.05 0.5 ... 0.3 14 4 ... ... ...
CrMo
154 CM Nominal Wrought 1.05 0.45 ... 0.3 14 4 ... ... ...
CPM Nominal PM 1.45 ... ... ... 14 2 ... 4.0 V ...
S30V
CPM Nominal PM 2.15 0.4 ... ... 17 0.4 ... 5.5 V ...
S60V
CPM Nominal PM 2.2 ... ... ... 13 1 ... 9.0 V ...
S90V
are not subject to the same limitations as contin- good toughness without tempering and mini-
uous casters and have alloys the production of mize the loss of chromium to carbides, main-
alloys with high volume content of VC. One taining it in solution for corrosion resistance.
such alloy is Crucible CPM 90V with 14% Cr, The addition of nickel and molybdenum yields
9% V, 1% Mo, and 2.3% C. This alloy has equal full austenite and martensite transformation and
or better toughness and corrosion resistance as improves corrosion resistance. Table 4 lists sev-
440C but has ten times the wear resistance at eral such alloys by JFE: the first two can be
the same macrohardness. made to meet L80 specifications and produced
Oil Country Tubular Good and Line Pipe. as seamless. JFE reports production of over
The need for corrosion resistance in oil produc- 100,000 tons per year of this product (Ref 7).
tion tubulars has grown as the quality of petro- The third alloy is a near match for the precipita-
leum deposits has become less optimal. Use of tion-hardening stainless Custom 450 (UNS
stainless can eliminate for corrosion inhibitors S45000) (Ref 8). Like other precipitation-hard-
in H2S and CO2 environments. This has led to enable steels, it shows excellent resistance to
the use of low-carbon martensitic stainless SCC at high strength levels. Figures 13 and 14
steels. Low carbon and nitrogen levels give show the improvements in corrosion resistance
Table 4 JFE Steel/Nippon Steel oil country tubular goods and line pipe alloys
HP13Cr-! JFE Wrought 0.025 0.45 ... ... 13 1 4 ... ...
Nominal
HP13Cr-2 JFE Wrought 0.025 0.45 ... ... 13 2 5 ... ...
Nominal
NT-CRS Nippon Wrought 0.03 1.45 ... ... 12.7 1.4 4.5 1.5 Cu 0.040 N
Nominal
NT-CRSS Nippon Wrought 0.02 2 ... ... 12.3 2 5.8 1.5 Cu 0.015 N
Nominal
KL-12Cr JFE Wrought 0.01 ... ... ... 11 ... 2.4 0.5 Cu 0.010 N
Nominal
KL-HP JFE Wrought 0.01 ... ... ... 12 2 5.5 ... 0.010 N
12Cr Nominal
ronments to about Rc 22 to avoid SCC by the hy-

drogen embrittlement mechanism. The stainless
can resist this failure mode at higher strengths. As
specifying bodies such as the American Petroleum
Institute (API) approve the use of stainless tubu-
lars at higher strength levels than carbon steel
tubulars are safely capable of handling, then the
strength improvement, coupled with the orders of
magnitude improvement in corrosion resistance,
will cause a great increase in their use.
Lower carbon levels permit the use of field
welds without tempering so that similar alloys
can be used for line pipe. These are corrosion
Fig. 13 Corrosion rates of stainless versus carbon steel resistant and yet meet X70 and X80 class speci-
fications. These alloys are the last two in Table
2. These uses of martensitic stainless steels for
oil production represent possibly the greatest
growth area for any kind of stainless steel in the
first decade of the 21st century.
REFERENCES
1. D.R. Koistinen, R.E. Marburger, “A General

Equation Prescribing the Extent of the
Austenite/Martensite Transformation in
Pure Iron,” Acta Met, Vol 7, 1959, p 59
2. http://www.msm.cam.ac.uk/phase-trans/
2002/martensite.html
3. Bletton, Aciers Inoxidables, Les Editions de
Physique les Ulis, Paris, 1993, p 481
4. ASM Handbook Desk Edition, 1985, p 28–9
5. http://products.asminternational.org/mgo/
Fig. 14 Corrosion rates for stainless oil country tubular goods 6. K.J. Irvine et al., JISI, Vol 195, ISIJ Interna-
(OCTG) alloys under severe operating conditions
tional 1960, p 386–405
over carbon steel L80 oil country tubular goods 7. S. Deshimaru et al., “Steels for Production,
(OCTG) under test conditions representative of Transportation and Storage of Energy, JFE
difficult real-use environments (Ref 7). Technical Report (No. 2), March 2004
The improvements in martensitic steels for p 55–67
these applications are hardly more than a thor- 8. M. Kimura et al., “High CR Stainless
ough revisiting of the developments of the OCTG with High Strength and Superior
1950s and 1960s. This does not diminish their Corrosion Resistance,” JFE Technical Re-
importance. Carbon steels are limited in sour envi- port (No. 7), Jan 2006, p 7–13
CHAPTER 10
Precipitation-Hardening
Stainless Steels
Summary Introduction
THE PRECIPITATION-HARDENABLE (PH) The PH stainless steels exploit the low austen-
grades are a highly specialized family of stain- ite stability possible in the chromium/nickel
less steels whose existence derives from the stainless steels by making the alloys so lean in
need for very high-strength materials with good composition that they can be made to trans-
corrosion resistance. The workhorse alloys are form nearly entirely to martensite by thermal
the martensitic PH grades, which are used in or mechanical treatment. This martensite can
many forms. Primarily used as forgings, bar, then be further hardened by the coherent pre-
and other hot-worked forms, they can also be cipitation of intermetallic compounds, ele-
obtained in cold-rolled sheet and strip, although mental copper, nitrides, or even phosphides.
not with the flatness expected from non-PH This precipitation hardening can also be made
stainless. The semiaustenitic alloys are more to occur in a fully austenitic matrix, and this
amenable to production as sheet, strip, and wire also provides a commercial PH alloy. But, the
and are designed for applications that require martensitic PH grades are by far the more
extensive forming before hardening. The fully common. The border between the more highly
austenitic PH alloys fill a small niche where alloyed martensitic stainless steels, which un-
high mechanical properties are required at tem- dergo secondary hardening during tempering,
peratures above or below which the other PH and the PH alloys is indeed vague. Some au-
grades are found lacking, when a nonmagnetic thors have astutely treated them as a single
material is required, or when the higher thermal group. Here, we treat them separately because
expansive coefficient of an austenitic material is they are traditionally considered as separate
desired. In no case is corrosion resistance better alloys.
than that of normal 304 found in PH stainless The advantage of the PH alloys over the
steels. If enhanced strength and very high corro- strictly martensitic stainless steels is that they
sion resistance are required, then the designer attain great strength with higher toughness and
should look to duplex stainless steels for the op- corrosion resistance than can be obtained
timal material. If cost is a greater concern than through the hardening of martensite through
corrosion resistance or toughness, then marten- carbon. In addition, they can be fabricated in a
sitic stainless steels should be considered for relatively soft state and then hardened with very
applications where strength and hardness over little dimensional change.
that of annealed ferritic and austenitic stainless The PH grades were developed at the begin-
is required. ning of World War II, with Stainless W (UNS
The increased use of titanium alloys and ad- S17600) by U.S. Steel generally acknowledged
vanced composite materials may occur at the as the first. The later-developed grades are dis-
expense of the stainless PH alloys and at the tinguished from the first by their more uniform,
same time may create some new niche applica- and therefore tougher, microstructure through
tions for them. the elimination of residual δ-ferrite and retained
austenite and by more astute alloy design and movement. As time and temperature of precipi-
chemistry control. tation increase, the zones can grow to sizes that
The mechanism of precipitation hardening is cannot accommodate the small size differential;
parallel to that used to strengthen aluminum al- coherency is lost, and with it the hardening ef-
loys in which the precipitation of a coherent sec- fect diminishes. The precipitation has the dual
ond phase from a supersaturated solid solution is function of stress relieving the martensite while
produced by an aging heat treatment. The coher- further hardening the matrix through the precipi-
ent precipitate strains the lattice and impedes the tation of the coherent precipitate. The mechani-
motion of dislocations, producing strengthening. cal properties of the final microstructure depend
Overaging causes the precipitates to lose co- on the initial strength of the matrix before aging,
herency, and softening follows. The precipitate the amount of precipitate, and the coherency of
that causes the hardening is normally nickel the precipitate. The ideal microstructure for the
(aluminum/titanium) (Ref 1). Figure 1 shows the initial matrix is 100% martensite. To the extent
compounds that can form from the precipitation there is δ-ferrite or retained austenite, properties,
of supersaturated aluminum and titanium in an especially yield strength and toughness in the
iron alloy matrix. transverse direction, are compromised. The
It is also possible to produce a hardening reac- aging temperatures can also be high enough that
tion by the precipitation of elemental copper reversion of martensite to austenite occurs,
(Ref 2). In nitrogen-bearing alloys, a hardening which also lowers subsequent tensile properties.
may be produced by the precipitation of Cr2N While the presence of persistent, large bands
(Ref 3). The precipitation begins with the diffu- of either δ-ferrite or γ-austenite is undesirable,
sion of the precipitating species to sites on the but both also have benefits. The presence of
existing matrix. These enriched zones are called some fine bands of δ-ferrite promotes easier and
Guinier-Preston (GP) zones. Close dimensional more reproducible precipitation of chrome car-
matchup between the precipitating species and bides at the δ/γ interface during the “austenite
the parent matrix is required. The differential conditioning” or “trigger anneal” heat treatment
should be on the order of a percent. This allows step for semiaustenitic alloys (17-7, AM350,
not only coherency but also strain. The coherent etc.). Although bands of stable austenite are
precipitate is a effective barrier to dislocation undesirable, it is the presence of residual interlath
γ + Ni(AlTi) + Ni3AlTi Ni(AlTi)

Ni(AlTi) + Ni2AlTi
4 Ni2AlTi
Ni2(AlTi) + Ni3AlTi
Ni2AlTi
γ + Ni2AlTi Ni3Ti
γ + Ni(AlTi)
σ
Aluminum, wt%
Limit of austenite
ferrite region in
solution treated
2 conditions
γ + Ni2(AlTi) + Ni3AlTi
γ + Ni3(AlTi)
1
γ γ + Ni3Ti + Ni3(AlTi)
Cellular
γ + Ni3Ti
precipitation
0 1 2 3 4
Titanium, wt%
Fig. 1 Possible aluminum/titanium precipitates

Chapter 10: Precipitation-Hardening Stainless Steels / 139
austenite that provides the work-hardening abil- strength than that of which the martensitic or
ity in many of these PH alloys. It is this work semiaustenitic alloys are capable. The austenitic
hardening that gives the PH alloys, especially the PH strength is better above 750 °F.
semiaustenitic ones, their unusual combination
of high strength plus ductility and toughness in
the fully hardened state. Martensitic Precipitation-Hardenable
The complexity of PH steels comes from the Stainless Steels
processing involved in producing the martensitic
structure in which the precipitation will occur. The martensitic PH alloys are, as stated, fully
The most straightforward alloys are the marten- martensitic at room temperature. Their marten-
sitic, also called the martensitic PH alloys. These site is a relatively soft, low-carbon (less than
steels are supplied in the fully martensitic condi- 0.05%) martensite as opposed to the higher car-
tion with hardness in the low Rc 30s. This is con- bon found in the martensitic stainless steels. The
fusingly called the annealed condition, or condi- early alloys of this type, 17–7 PH and 17-4 PH,
tion A, even though the matrix is untempered contained up to 10% δ-ferrite stringers, which
martensite. After the material is fabricated, it is caused poor through-thickness toughness. This
subjected to an aging treatment designated by would be expected from the Schaeffler-Delong
the aging temperature in Fahrenheit (e.g., diagram, but this is asking too much of the Scha-
H-950). These aging temperatures range from effler-Delong diagram, which was developed for
950 °F (510 °C) to 1150 °F (620 °C). welds, to predict the phase composition of alloys
A second major group of PH grades is the that have been homogenized by hot working.
semiaustenitic. These grades in the normally The inaccuracy of the diagram for more com-
furnished condition A are fully austenitic. This plex systems was overcome, and alloys were de-
is accomplished by adding elements that lower signed that had minimal δ-ferrite and still trans-
the martensite start temperature, such as more formed entirely to martensite, if not at room
chromium, molybdenum, and nickel. The temperature, at least at a reasonably attainable
austenite is more formable than martensite, and subzero temperature. This was done first by trial
it has the possibility of superior corrosion resist- and error and more recently by using thermody-
ance because of higher chromium content. This namic computer models, such as ThermoCalc, to
is balanced by the need to use either cold work, predict equilibrium phase composition. This de-
cryogenic treatment, or a destabilizing anneal to velopment was very significant for making the
cause the matrix to become martensitic before alloy family useful as a high-strength/high-
its precipitation aging treatment. toughness material for demanding applications
Last, if the austenite is made very stable by requiring high mechanical properties and corro-
further alloying additions, a precipitation reac- sion resistance. The most advanced PH alloys
tion can still be made to occur by the same type are martensitic PH grades by Cartech, Custom
of aging treatment without martensite ever form- 465 and 475.
ing. The precipitation takes place in austenite Table 1 shows the more significant of these al-
and therefore results in lower room temperature loys compared on a strength basis; Table 2 shows
Table 1 Mechanical properties of martensitic precipitation-hardenable alloys
Toughness,
Alloy UNS Condition Yield, MPa Tensile, MPa Elongation, % HRC CVN ft-lb
Stainless W S17600 H-950 ( 510) 1240 1340 14 42 ...
17-4 PH S17400 H-925 (495) 1210 1310 14 41 40
15-5 PH S15500 H-925 (495) 1210 1300 15 41 20
H1100 (595) 930 1025 17.50 34 70
13-8 PH S13800 H-950 (510) 1450 1550 12 47 25
H1050 (565) 70
Custom 450 S45000 H-900 (480) 1270 1350 14 42 60
H1100 (595) 460 970 23 ... 180
Custom 455 S45500 H-950 (510) 1515 1585 10 48 8
H1050 (565) 1205 1310 14 40 25
Custom 465 S46500 H-950 (510) 1650 1765 11 49 13
H1000 (535) 1500 1600 13 48 28
Custom 475 ... H-975 (525) 1855 2005 5 54 ...
H 1100 (595) 1315 1572 13 48
Ferrium S53 ... ... 1565 1985 14–16 54 18
Table 2 Composition of martensitic precipitation-hardenable alloys
Alloy Designation C Mn Si Cr Ni Mo Al Cu Ti Other

Stainless W S17600 0.1 0.5 0.5 17 6.3 ... 0.2 ... 1 P 0.3
17-4 PH S17400 0 0.6 0.6 16 4.3 ... ... 3.2 ... ...
15-5 PH S15500 0 0.6 0.6 15 4.3 ... ... 3.2 ... ...
13-8 PH S13800 0 0.1 0.1 13 8.5 2 1.1 ... ... ...
Custom 450 S45000 0 0.3 0.3 15 6 0.8 ... 1.5 ... 0.3 Nb
Custom 455 S45500 0 0.3 0.3 12 8.5 ... ... 2.5 1 0.3 Nb
Custom 465 S46500 0 0.2 0.2 12 11 1 ... ... 2 ...
Custom 465 (275) ... 0 0.2 0.2 12 11 1 ... ... 2 0.2 Nb
Custom 475 ... 0 0.4 0.4 11 8 5 1.2 ... ... 8.0 Co
Ferrium S53 ... 0.2 0.1 0.1 10 5.5 2 0 ... ... 1W
... 0.3 V
... 14 Co
Fig. 2 Typical microstructures of precipitation-hardenable (PH) stainless steels: (a) 15-5PH as-quenched martensite; (b) 13-8 PH so-
lution treated and aged displaying fine martensite; (c) 17-7 PH displaying ferrite stringers in a martensite matrix; (d) 17-7 PH
showing residual ferrite stringers and inclusions
their nominal compositions. Figure 2 shows a se- less W to the latest in Custom 475 besides the
ries of photomicrographs of PH alloys. elimination of δ-ferrite is in the volume fraction
The main advancement metallurgically in of the precipitating phase and the elimination of
these alloys from the top, and earliest, in Stain- retained austenite. Alloy designers found that to
reduce δ-ferrite they also tend to stabilize • Molybdenum to offset loss of corrosion re-
austenite, which also happens to reduce the sistance by minimization of chromium, to
temperature at which martensite forms, MS. increase the temperature at which austenite
Last, the higher levels of molybdenum reduce forms, and to form another hardening pre-
the tendency to form secondary austenite during cipitate in the presence of cobalt
aging. To minimize δ-ferrite requires reducing • Cobalt to stabilize austenite while raising
chromium or molybdenum, which also reduces MS
corrosion resistance. As a result, as alloy • Aluminum or titanium to form intermetallic
strength increases, corrosion resistance is com- precipitates with nickel or copper to precipi-
promised. The alloys with the greatest strength tate as elemental copper
potential, Custom 465 and 475, barely qualify
It is possible to quantify these various influ-
as stainless with around 11% chromium. But,
ences on phases. This is summarized in Table 3
from a utility point of view, these alloys are de-
in terms of the influence of the element on dif-
signed to have maximum mechanical properties
ferent factors measured in degrees Centigrade
with adequate corrosion resistance, so this is
for a 12% chromium alloy.
viewed as an acceptable compromise.
Rapid quenching of these alloys is not re-
The newest alloy is Ferrium S53, one of the
quired. They are air hardenable. But, the cool-
recent alloys designed by computer-assisted
ing of these alloys must be completed expedi-
thermodynamic calculations. It was designed to
tiously through the final stages of martensite
replace 300M, 4340, and AerMet 100 on an
formation with minimal delay. During delays
equal mechanical properties basis but also pro-
after the start of martensite transformation has
vide the corrosion resistance necessary to be
occurred, the remaining austenite tends to stabi-
used in aircraft components without cadmium
lize, and full transformation to martensite does
plating. Its composition superficially seems de-
not occur. When this happens, the higher levels
ficient in chromium to provide “stainlessness,”
of austenite reduce subsequent mechanical
but the cobalt level raises the thermodynamic
properties after aging.
activity of chromium sufficiently that the equiv-
As with any alloy used at such high strength
alent of 12% chromium in a non-cobalt-contain-
levels, microstructural cleanliness is essential,
ing alloy is achieved. The precipitation harden-
but air melting and argon oxygen decarburiza-
ing mechanism is the precipitation of Mo2C. It
has been established that this hardening mecha- tion (AOD) refining are quite adequate.
nism optimizes resistance to stress corrosion Corrosion Resistance. The martensitic PH
cracking (SCC) for a given strength level. stainless steels obey the same rules as other
The alloying characteristics of these grades stainless steels with regard to corrosion resist-
are: ance. The martensite carries no nitrogen in solu-
tion, so the resistance to pitting is given by:
• Low carbon, nitrogen, silicon, and man-
ganese because these elements lower MS PREN = %Cr + 3.3%Mo (Eq 1)
without contributing to age hardening
• Low chromium to suppress δ-ferrite In the martensitic PH alloys, no chromium is
• Sufficient nickel to suppress δ-ferrite and rendered ineffective by the formation of Cr23C6
provide for precipitates without excessively since carbon is either held low or stabilized by
depressing MS titanium or niobium. Thus, the corrosion
Table 3 Influence of alloying elements on key transformations

Element N C Ni Co Cu Mn Si Mo Cr V Al
Lowering –220 –210 –20 –7 –7 –6 6 5 14 18 54
of % ferrite
per % element
Lowering of MS –475 –475 –17 0 to 10 –17 –30 –11 –21 –17 –46
per % element
Change of AC 0 to 280 0 to 250 –30 to –115 0 0 –25 to –66 25 to 73 25 to 60 0 to 35 50 to 290 30 to 750
per % element
resistance will equal that of stabilized ferritic al- ments, the austenite could be made stable at
loys of the same pitting resistance equivalent room temperature. This would make the alloy
number (PREN) for which voluminous data are softer and more fabricable and, most impor-
available. tantly, permit them to be manufactured as cold-
A greater concern is the risk of SCC in these rolled sheet and strip. If the austenite could then
alloys. While the mechanism of SCC in be transformed to martensite by cryogenic treat-
austenitic alloys is still debatable, it has long ment, cold work, or special heat treatment, then
been clear that, for martensitic alloys, SCC is it could be age hardened just like the martensitic
simply a manifestation of hydrogen embrittle- PH grades. This has been accomplished for a
ment in which the hydrogen is provided by local group of alloys called the semiaustenitic PH
corrosion. The existence of pitting is a suffi- grades. The “semi” signifies that the austenite
cient, if not necessary, condition for SCC to in these alloys is metastable rather than stable at
occur if the temperature is within the range of ambient temperatures. Also, it should be noted
susceptibility and the material is inherently sus- that these semiaustenitic alloys usually contain
ceptible. The material susceptibility is largely a some δ-ferrite in their predominantly austenitic
function of resistance to crack propagation in microstructure after annealing.
any given alloy that is measured by fracture These alloys are complex metallurgically be-
toughness. The martensitic PH grades have ex- cause of the technique used to make the austenite
cellent toughness and low rates of crack propa- stable at room temperature after a full solution
gation, but none should be considered immune anneal. The austenite is rendered stable by fairly
to SCC since their hardness is never below the high levels of carbon, a powerful austenite stabi-
Rc 22 level, which is considered to be the lizer, in solution. The amount of carbon that can
threshold hardness for susceptibility to SCC in be held in solution is a function of annealing tem-
body-centered cubic (bcc) ferrous alloys. perature. The 1050 °C anneal of the condition A
The suitability of high-strength alloys for use mill anneal puts all the carbon in solution, giving
in potential SCC-provoking environments con- the austenite the stability of a normal 301-type
taining H2S is regulated in many locales by Na- alloy. This permits extensive forming. The key is
tional Association of Corrosion Engineers to apply a subsequent lower-temperature anneal
(NACE) International Standard MR01-75. In it, so that less carbon goes into solution. Some of it
the use of S17400 is permitted if it is double will thus form M23C6. This is in a sense deliber-
tempered at 620 °C and its hardness is 33 HRC ately sensitizing the alloy, but the sensitization
or less, while S45000 can be used if it has been takes place at such a high temperature that
aged at 620 °C for 4 h and its hardness is 31 chromium deficits around precipitated carbides
HRC or less. These permissible hardness levels are minimized by diffusion. This causes a higher
are significantly higher than allowed in non-PH Ms temperature because of the lower amount of
martensitic alloys, 22 HRC, which reflects the carbon, and chromium, in solution in the austen-
fact that the martensitic matrix has the tough- ite. Depending on the temperature at which the
ness of a lower-hardness martensite. anneal is done, the Ms temperature can be con-
In marine environments, the PH alloys are trolled so that a transformation to martensite can
susceptible to SCC if used at a high strength be raised to either room temperature or some at-
level. S17400 aged at 480 °C with a yield tainable cryogenic temperature. Figure 3 shows a
strength of 1240 MPa is susceptible to SCC, chart of these heat treatment options.
while higher aging temperatures (above 540 °C) The lower strength levels achieved by the
producing lower strengths renders the materials condition T route in Fig. 3 reflect the lower car-
immune at stresses near the yield strength, ap- bon content of the martensite, while the highest
proximately 1170 MPa. The threshold strength strength of the condition C route reflects the
for ordinary martensitic stainless steels would compound influence of cold work and marten-
be 1030 MPa (Ref 4). site hardness with similar subsequent contribu-
tion from age hardening.
The main alloys of this group are listed in
Semiaustenitic Precipitation- Table 4. Examination of the chemistries in this
Hardenable Stainless Steels table shows that the first two alloys rely on the
precipitation of Ni3Al for the hardening, while
If a martensitic stainless steel were alloyed the last two have no apparent precipitating
more strongly with austenite-stabilizing ele- components. Their hardening is a more subtle
Fig. 3 Processing routes for S15700 Source: Ref 5
Table 4 Compositions of semiaustenitic precipitation-hardenable alloys

Alloy Designation C Mn Si Cr Ni Mo Al N
17-7 PH S17700 0.1 0.5 0.3 17 7.1 ... 1 0
15-7 PH S15700 0.1 0.5 0.3 15 7.1 2.2 1 0
AM-350 S35000 0.1 0.8 0.4 17 4.3 2.8 ... 0.1
AM-355 S35500 0.1 0.9 0.4 16 4.3 2.8 ... 0.1
secondary hardening from the tempering of titanium would preferentially deplete the alloy
martensite rather than the classic precipitation of carbon and nitrogen, precluding the action of
hardening via precipitation of intermetallic the conditioning heat treatment, which relies on
compounds. In 17-7 and 15-7, aluminum rather manipulating the amount of carbon in solution.
than titanium is the precipitating agent because In AM350 and AM355, it is the precipitation of
Table 5 Mechanical properties of semiaustenitic precipitation-hardenable alloys

Yield, Tensile,
Alloy UNS Condition MPa MPa Elongation, % HRC
17–7 PH ... TH 1050 (565) 1100 1310 10 42
RH 950 (510) 1380 1520 9 46
CH 900 (480) 1585 1655 2 49
15–7 PH ... TH 1050 (565) 1380 1450 7 45
RH 950 (510) 1550 1650 6 48
CH 900 (510) 1720 1790 2 50
AM-350 ... SCT 850 (450) 1210 1420 12 46
SCT1000 (540) 1020 1165 15 40
AM-355 ... SCT 850 (450) 1250 1510 13 48
SCT 1000(540) 1035 1124 22 38
(Cr,Fe)2N within the martensite phase that is that all carbon is present as chromium carbide,
responsible for age hardening. In addition, and that the chromium content is diminished by
molybdenum produces a secondary hardening that amount before applying Eq 1. In addition,
in carbon-bearing martensite. These two early carbon can remove some molybdenum in the
alloys did not possess the hardening potential form of carbides, and nitrogen can remove
that alloys employing copper-, titanium-, or chromium as a nitride. The effective corrosion
aluminum-based precipitates enjoy. resistance of these alloys thus is similar to fer-
The mechanical properties of these alloys ritic 430. The designer is thus advised to consult
are not greatly different from the martensitic al- with producers about corrosion resistance de-
loys, as can be seen in Table 5. Their separate pending on the thermal processing that will be
existence is due to the need for alloys that are used, especially if double aging is performed,
more fabricable at room temperature than are the which can cause some degree of chromium de-
alloys that are martensitic at room temperature. pletion at grain boundaries.
This benefit is offset by the necessity to condi- While the general and pitting corrosion re-
tion anneal before age hardening. If corrosion sistance of the semiaustenitic PH alloys are
resistance is a high concern, then the alloy and never quite as good as most austenitic stainless,
heat treatment that yields the greatest amount of they have very good resistance to SCC com-
chromium in solution should be chosen. Thus pared to ordinary martensitic stainless steels.
condition CH is better than RH, which is better The δ-ferrite and the generally well-tempered
than TH, the order of ascending solution anneal- martensitic matrix provide a crack-arresting
ing temperature and ascending Ms. feature and good inherent toughness that resist
The martensitic PH grades have somewhat SCC at higher strength levels than in straight
better strength because they have a more uni- martensitic stainless steels. AM-355 in the SCC
formly martensitic structure, and they employ a (850 °F) condition can withstand stresses of
tougher, lower-carbon martensite. The fact that 75% of 0.2% offset yield strength in salt spray
the semiaustenitic alloys are typically sheet without SCC failure.
products generally makes their service tough-
ness requirements less onerous, so that their re-
tained δ-ferrite is not a crippling drawback be-
cause of its detrimental effect on through Austenitic Precipitation-Hardenable
thickness toughness. Stainless Steels
Corrosion Resistance. The semiaustenitic
PH alloys tend to higher values of PREN than The austinitic PH class consists of just one
the martensitic alloys inherently since they are important commercial alloy, A-286. The impor-
alloyed to have lower Ms temperatures. The tance of the alloy is that it is entirely stable
thermal processing of these alloys causes a sig- austenite in both the solution-annealed and the
nificant portion of the chromium to be removed age-hardened condition. This means it is very
from solution as chromium carbide. This lowers formable and nonmagnetic. And, because the
the corrosion resistance from what would be ex- precipitation takes place in an austenite matrix,
pected based on the bulk composition. From an the precipitation takes place at a higher temper-
engineering point of view, it is best to assume ature, around 700 °C. This gives the alloy the
Table 6 Austenitic precipitation-hardenable composition

Alloy UNS C Mn Si Cr Ni Mo Al V Ti
A-286 S66286 0.05 1.5 0.5 15 25.5 1.3 0.15 0.3 2.15
Discalloy S66220 0.04 1.6 0.5 14 26 3 ... ... 1.7
Fig. 4 A-286 properties as a function of test temperature. Source: Ref 5
potential to be used to near 700 °C without fear PH alloys, makes it attractive for a variety of
of overaging. Thus, austenitic PH stainless rep- aerospace and nonaerospace uses.
resents a way to strengthen the austenite matrix, The hardening mechanism of A-286 is the
which has the following advantages: precipitation of Ni3 (aluminum, titanium). Dif-
fusion, even at the higher temperatures, is
• High ductility and therefore high formability slower in austenite, so aging treatments are typ-
in the soft, unaged condition ically 16 h. Table 6 gives the typical composi-
• High toughness at all temperatures and tion of A-286, and Fig. 4 shows some typical
strength levels properties as a function of temperature for a
• Excellent creep and stress rupture properties standard 980 °C solution treatment followed by
• Excellent oxidation, corrosion, and SCC re- a 720 °C, 16 h aging.
sistance The mechanical properties can be greatly en-
This alloy is rightly considered an iron- hanced by cold working prior to aging, as is
based superalloy and is the root of the group shown in Fig. 5.
that succeeds it in properties, the nickel- and The toughness of the austenitic matrix is
cobalt-based superalloys. These last alloys can abundant and quite temperature insensitive.
attain greater strength and creep resistance Charpy V-notch values of over 60 J are typical
than A-286, effectively ending further devel- from –200 to 800 °C.
opment of austenitic PH alloys. The lower cost The corrosion resistance of A-286 is compa-
of the A-286 alloy, compared to nickel-base rable to that of 304 and 316. It has slightly
Fig. 5 The influence of cold work on aging response in A-286. DPH, diamond pyramid hardness. Source: Ref 5
better resistance to SCC despite its higher allurgical and Materials Transactions A,
strength level. Vol 30A, Feb 1999.
3. G. Aggen, Ph.D. thesis, Carnegie Mellon
REFERENCES University
4. E.E. Denhard, “Stress Corrosion Cracking
1. F.B. Pickering, Physical Metallurgical De- of High Strength Stainless Steels in Atmos-
velopments of Stainless Steel, Stainless ’84, pheric Environments”, paper presented at
Goteborg, Sept 3–4, 1984, p 2–28. the Twenty-fourth Meeting of the AGARD
2. M. Murayama, Y. Katayama, and K. Hono, Structures and Materials Panel (Turin,
Microstructural Evolution in a 17-4 PH Italy), April 17–20, 1967.
Stainless Steel After Aging at 400 °C, Met- 5. Allegheny Technology Blue Sheets
CHAPTER 11
Casting Alloys
Summary essential element. Thus, the casting alloys listed

in Tables 1 and 2 (Ref 1) are recognizable as ap-
WITH TYPICAL ALLOY SYSTEMS, cast- proximate counterparts of the co-listed wrought
ing is often the most convenient method by alloys (AISI grade). This cross reference to
which to produce components. This is true for wrought equivalents is helpful when looking for
stainless steels—both for corrosion-resisting data about an alloy that may be more easily
and for heat-resisting applications. This chapter found for wrought alloys than for cast.
discusses primarily the alloys used for stainless The High Alloy Product Group of the Steel
steel castings and their metallurgy. Foundry Founder’s Society of America employs a nam-
methods are discussed to the degree they are ing system (ACI, the Alloy Casting Institute)
specific to the stainless alloys. for cast alloys that is significant; these designa-
tions are currently assigned by ASTM as grades
and are added to ASTM specifications. The first
Stainless Steel Casting Alloys letter, “C” or “H,” indicates corrosion resisting.
The second letter indicates the relative amount
Essentially any wrought stainless alloy compo- of nickel, from a minimum of 0 to 1% for “A”
sition can be modified to be made as a cast alloy. up to 30% nickel for “N” alloys. The number
The systemic difference between cast alloys and following the hyphen for “C” alloys designates
their wrought equivalents is that cast alloys gen- the maximum carbon in hundredths of a percent.
erally contain between 1.0 and 2.5% silicon. As The suffix letters designate additional alloying
with other ferrous alloys, this is done to increase elements, such as Cu for copper, M for molyb-
the fluidity of the melt to make it cast more effec- denum, N for nickel or nitrogen, F for free ma-
tively. Silicon has strong metallurgical effects, chining, and C for columbium (niobium). The
both beneficial and detrimental, which should be heat-resisting, “H,” alloys have generally only a
understood by the user of cast stainless steels. second letter designating relative nickel level on
These are explained. A second general observa- the same scale as “C” alloys but going past stain-
tion is that the stabilized ferritic stainless steel al- less steels all the way to nickel-based alloys. The
loys, which constitute almost half the tonnage of inclusion of a number after the first two letters
all stainless steel used, are notably absent from indicates the center of the carbon range ex-
the cast alloys. This is because these alloys are pressed in hundredths of a percent by weight.
single phase at all temperatures in the solid state To learn more about the influence of alloying
and because they have large as-cast grain sizes elements, refer to the chapters on the individual
that can only be refined by heavy cold work fol- alloy families; see Section 3. Here, we briefly
lowed by annealing. This makes them quite lack- summarize:
ing in toughness as cast. Since heavy cold work • Pitting and crevice corrosion resistance, as
defeats the purpose of casting to achieve a near- well as general corrosion resistance, are en-
net shape, stabilized ferritic stainless steels are hanced by chromium, molybdenum, tung-
seldom used as castings. Also, the standard stabi- sten, and nitrogen and carbon in solution.
lizing alloy, titanium, is too readily oxidized for • Localized corrosion is caused by chromium
normal foundry practice to avoid the loss of this depletion, which occurs when precipitates
Table 1 Compositions of cast stainless corrosion resisting alloys

Composition(a), wt%—maximum or range
Nearest
ACI designation AISI grade UNS %C %Mn %Si %Cr %Ni %Mo %Other
Chromium alloys
CA-15 410 J91150 0.15 1.00 1.50 11.5–14.0 1.0 0.50
CA-15M J91151 0.15 1.00 0.65 11.5–14.0 1.0 0.15–1.00
CA-40 420 J91153 0.40 1.00 1.50 11.5–14.0 1.0 0.50
CA-40F 420F J91154 0.2–0.4 1.00 1.50 11.5–14.0 1.0 0.20–0.40 Ss
CB-30 431,442 J91803 0.30 1.00 1.50 18.0–22.0 2.0
CC-50 446 J92613 0.30 1.00 1.50 26.0–30.0 4.0
Chromium-nickel alloys
CA-6N J91650 0.06 0.50 1.00 10.5–12.5 6.0–8.0
CA-6NM S41500 J91540 0.06 1.00 1.00 11.5–14.0 3.5–4.5 0.4–1.0
CA-28MWV 422 J91422 0.20–0.28 0.50–1.00 1.00 11.0–12.5 0.5–1.0 0.9–1.25 0.9–1.25 W,
0.2–0.3 V
CB-7Cu-1 17-4PH (AISI 630) J92180 0.07 0.70 1.00 15.5–17.7 3.6–4.6 2.5–3.2 Cu,
0.2–0.35 Nb,
0.05 N max
CB-7Cu-2 15-5 PH (XM-12) J92110 0.07 0.70 1.00 14.0–15.5 4.5–5.5 2.5–3.2 Cu,
0.2–0.35 Nb,
0.05 N max
CD-3MN 2205 (S32205) J92205 0.03 1.50 1.00 21.0–23.5 4.5–6.5 2.5–3.5 1.0 max Cu,
0.10–0.30 N
CD-3MCuN 255 (S32550) J93373 0.03 1.20 1.10 24.0-26.7 5.6–6.7 2.9–3.8 1.4–1.9 Cu,
0.22–0.33 N
CD-3MWCuN (S32760) J93380 0.03 1.00 1.00 24.0–26.0 6.5–8.5 3.0–4.0 0.5–1.0 Cu,
0.5–1.0 W,
0.20–0.30 N
CD-4MCu J93370 0.04 1.00 1.00 24.5–26.5 4.75–6.0 1.75–2.25 2.75–3.25 Cu
CD-4MCuN J93372 0.04 1.00 1.00 24.5–26.5 4.7–6.0 1.75–2.25 2.75–3.25
Cu, 0.10–0.25 N
CD-6MN J93371 0.06 1.00 1.00 24.0–27.0 4.0–6.0 1.75–2.25 1.75–2.5 Cu,
0.15–0.25 N
CE-3MN 2507 (S32750) J93404 0.03 1.50 1.00 24.0–26.0 6.0–8.0 4.0–5.0 0.10–0.30 N
CE-8MN J93345 0.08 1.00 1.50 22.5–25.5 8.0–11.0 3.0–4.5 0.10–0.30 N
CE-30 312 J93423 0.30 1.50 2.00 26.0–30.0 8.0–11.0
CF-3 304L J92500 0.03 1.50 2.00 17.0–21.0 8.0–12.0
CF-3M 316L J92800 0.03 1.50 2.00 17.0–21.0 8.0–12.0 2.0–3.0
CF-3MN 316LN J92700 0.03 1.50 1.50 17.0–21.0 9.0–13.0 2.0–3.0 0.10–0.20 N
CF-8 304 J92600 0.08 1.50 2.00 18.0–21.0 8.0–11.0
CF-8C 347 J92710 0.08 1.50 2.00 18.0–21.0 9.0–12.0 Nb
CF-8M 316 J92900 0.08 1.50 2.00 18.0–21.0 9.0–12.0 2.0–3.0
CF-10 304H J92590 0.04–0.10 1.50 2.00 18.0–21.0 8.0–11.0
CF-10M 316H J92901 0.04–0.10 1.50 1.50 18.0–21.0 9.0–12.0 2.0–3.0
CF-10MC 316H J92971 0.10 1.50 1.50 15.0–18.0 13.0–16.0 1.75–2.25 (10xC)–1.2 Nb
CF-10SMnN NITRONIC™ 60 J92972 0.10 7.00–9.00 3.50–4.50 16.0–18.0 8.0–9.0 0.08–0.18 N
CF-12M 316 0.12 1.50 2.00 18.0–21.0 9.0–12.0 2.0–3.0
CF-16F 303 J92701 0.16 1.50 2.00 18.0–21.0 9.0–12.0 1.5 max 0.2–0.35 Se
CF-20 302 J92602 0.20 1.50 2.00 18.0–21.0 8.0–11.0
CG-6MMN NITRONIC™ 50 J93790 0.06 4.00–6.00 1.00 20.5–23.5 11.5–13.5 1.5–3.0 0.1–0.3 Nb,
0.1–0.3 V,
0.2–0.40 N
CG-8M 317 J93000 0.08 1.50 1.50 18.0–21.0 9.0–13.0
CG-12 308 J93001 0.12 1.50 2.00 20.0–23.0 10.0–13.0
CH-8 309S J93400 0.08 1.50 1.50 22.0–26.0 12.0–15.0
CH-10 309H J93401 0.04–0.10 1.50 2.00 22.0–26.0 12.0–15.0
CH-20 309 J93402 0.20 1.50 2.00 22.0–26.0 12.0–15.0
CK-3MCuN 254SMO™ J94653 0.025 1.20 1.00 19.5–20.5 17.5–19.5 6.0–7.0 0.5–1.0 Cu,
0.18–0.24 N
CK-20 310 J94202 0.20 2.00 2.00 23.0-27.0 19.0–22.0
CN-3M 904L J94652 0.03 2.00 1.00 20.0–22.0 23.0–27.0 4.5–5.5
CN-3MN AL-6XN® J94651 0.03 2.00 1.00 20.0–22.0 23.0–27.0 6.0–7.0 0.18–0.24 N
CN-7M 320 N08007 0.07 1.50 1.50 19.0-22.0 27.5–30.0 2.0–3.0 3.0–4.0 Cu
CN-7MS J94650 0.07 1.50 3.50 18.0–20.0 22.0–25.0 2.5–3.0 1.5–2.0 Cu
CT-15C N08151 0.05–0.15 0.15–1.50 0.50–1.50 19.0–21.0 31.0–34.0 0.5–1.5 Nb
(a) Balance Fe for all compositions. Source: Ref 1
Chapter 11: Casting Alloys / 149
Table 2 Compositions of cast heat-resistant stainless and nickel base alloys

Composition(a), wt%—maximum or range
ACI designation Nearest AISI grade UNS %C %Cr %Ni %Si max
HA 504 J82090 0.20 max 8–10 1.00
HC 446 J92605 0.50 max 26–30 4–max 2.00
HD 327 J93005 0.50 max 26–30 4–7 2.00
HE 312 J93403 0.20–0.50 26–30 8–11 2.00
HF 302B J92603 0.20–0.40 19–23 9–12 2.00
HH 309 J93505 0.20–0.50 24–28 11–14 2.00
HI J94003 0.20–0.50 26–30 14–18 2.00
HK 310 J94224 0.20–0.60 24–38 18–22 2.00
HK-30 J94203 0.25–0.35 23–27 19–22 1.75
HK-40 J94204 0.35–0.45 23–27 19–22 1.75
HL N08604 0.20–0.60 28–32 18–22 2.00
HN J94213 0.20–0.50 19–23 23–27 2.00
HP N08705 0.35–0.75 24–28 33–37 2.00
HP-50WZ 0.45–0.55 24–28 33–37 2.50
HT 330 N08605 0.35–0.75 13–17 33–37 2.50
HT-30 N08603 0.25–0.35 13–17 33–37 2.50
HU N08005 0.35–0.75 17–21 37–41 2.50
HW N08006 0.35–0.75 10–14 58–62 2.50
HX N06050 0.35–0.75 15–19 64–68 2.50
(a) Balance Fe for all compositions. Manganese content: 0.35–0.65% for HA, 1% for HC, 1.5% for HD, 2% for the other alloys. Phosphorus and sulfur contents: 0.04
(max) for all but HP-50WZ. Molybdenum is intentionally added only to HA: 0.90–1.2%. Maximum molybdenum for other alloys is 0.5%. HH contains 0.2% N
(max). HP-50WZ also contains 4–6% W, 0.1–1.0% Zr, and 0.035% S (max) and P (max). Source: Ref 1
form in the solid state. These precipitates are minimize solidification hot cracking or to allow
carbides, oxides, and sulfides as well as inter- weld repair of cracks that do form. It has been
metallic phases richer in chromium, molyb- shown that the existence of ferrite can increase
denum, or nitrogen than the matrix. the resistance to stress corrosion cracking.
• General corrosion resistance follows the
above guidelines but is also helped by cop-
per and nickel, which do not assist in pitting
resistance. Metallurgy of “C” Alloys
• High-temperature oxidation resistance is en-
hanced by increasing chromium and silicon. The corrosion-resistant “C” series have
Wrought alloys employ aluminum and rare wrought counterparts from which they differ es-
earths to help oxidation resistance, but the sentially only in silicon content. This silicon has
difficulty of keeping these elements from no significant influence on corrosion resistance
being oxidized requires special techniques or mechanical or physical properties, so an un-
such as vacuum induction melting and inert derstanding of these alloys by approximating
refractories for molds. them to their wrought counterparts is justified.
• Iron-chromium (ferritic) alloys have better The main difference between the cast and
thermal fatigue resistance but poorer creep wrought product forms of these alloys is the
resistance than iron-chromium-nickel grain structure. In wrought grades, the grain
(austenitic) alloys. structure can be manipulated by deformation
and heat treatment. The use of deformation is
The alloy designation system largely ignores not an option in cast alloys; consequently, the
the wrought alloy distinctions by microstructure opportunities for grain refinement in cast alloys
(i.e., ferritic, austenitic, duplex, PH [precipita- are limited. To counteract the problems of lower
tion hardening], and martensitic). One reason is corrosion resistance of cast grades, a homoge-
that the most widely used wrought-stabilized nizing solution heat treatment is necessary to
ferritics (e.g., 409, 439) do not exist as common counteract the chromium depletion that occurs
casting alloys, and nominally austenitic alloys due to solidification segregation and precipitating
in the cast form contain enough ferrite to be sig- phases. Representative mechanical properties
nificantly magnetic. Thus, the distinctions based for “C” alloys are listed in Table 3 (Ref 1).
on phase are not as well defined for casting al- Martensitic. CA alloys are martensitic. The
loys. The high ferrite content in the nominally metallurgy is straightforward and equivalent to
austenitic casting alloys is to avoid or at least their wrought counterparts. The mechanical
Table 3 Room temperature mechanical properties of corrosion resisting cast stainless alloys
Tensile strength Yield strength, 0.2% offset Charpy toughness

Elongation % Reduction Hardness,
Alloy Heat treatment(a) MPa ksi MPa ksi in 50 mm, 2 in. in area, % HB J ft-lb Specimen
CA-6NM >955 oC (1750 oF), AC, T 827 120 689 100 24 60 269 94.9 70 V-notch
CA-15 980 oC (1800 oF), AC, T 793 115 689 100 22 55 225 27.1 20 Keyhole notch
CA-40 980 oC (1800 oF), AC, T 1034 150 862 125 10 30 310 2.7 2 Keyhole notch
CB-7Cu 1040 oC (1900 oF), OQ, A 1310 190 1172 170 14 54 400 33.9 25 V-notch
CB-30 790 oC (1450 oF), AC 655 95 414 60 15 ... 195 2.7 2 Keyhole notch
CC-50 1040 oC (1900 oF), AC 669 97 448 65 18 ... 210 ... ... ...
CD-4MCu 1120 oC (2050 oF), FC to 745 108 558 81 25 ... 253 74.6 55 V-notch
1040 oC (1900 oF), WQ
1120 oC (2050 oF), FC to 896 130 634 92 20 ... 305 35.3 26 V-notch
1040 oC (1900 oF), A
CE-30 1095 oC (2000 oF), WQ 669 97 434 63 18 ... 190 9.5 7 Keyhole notch
CF-3 >1040 oC (1900 oF), WQ 531 77 248 36 60 ... 140 149.2 110 V-notch
CF-3A >1040 oC (1900 oF), WQ 600 87 290 42 50 ... 160 135.6 100 V-notch
CF-8 >1040 oC (1900 oF), WQ 531 77 255 37 55 ... 140 100.3 74 Keyhole notch
CF-8A >1040 oC (1900 oF), WQ 586 85 310 45 50 ... 156 94.9 70 Keyhole notch
CF-20 >1095 oC (2000 oF), WQ 531 77 248 36 50 ... 163 81.4 60 Keyhole notch
CF-3M >1040 oC (1900 oF), WQ 552 80 262 38 55 ... 150 162.7 120 V-notch
CF-3MA >1040 oC (1900 oF), WQ 621 90 310 45 45 ... 170 135.6 100 V-notch
CF-8M >1065 oC (1950 oF), WQ 552 80 290 42 50 ... 170 94.9 70 Keyhole notch
CF-8C >1065 oC (1950 oF), WQ 531 77 262 38 39 ... 149 40.7 30 Keyhole notch
CF-16F >1095 oC (2000 oF), WQ 531 77 276 40 52 ... 150 101.7 75 Keyhole notch
CG-8M >1040 oC (1900 oF), WQ 565 82 303 44 45 ... 176 108.5 80 V-notch
CH-20 >1095 oC (2000 oF), WQ 607 88 345 50 38 ... 190 40.7 30 Keyhole notch
CK-20 1150 oC (2100 oF), WQ 524 76 262 38 37 ... 144 67.8 50 Izod V-notch
CN-3MN 1150 oC (2100 oF), WQ 770 112 365 53 50 185 190 140 V-notch
CN-7M 1120 oC (2050 oF), WQ 476 69 214 31 48 ... 130 94.9 70 Keyhole notch
(a) AC, air cool; FC, furnace cool; OQ, oil quench; WQ, water quench; T, temper; A, age. Source: Ref 1
properties are governed by the thermal pro- high level of nickel negates much of the potential
cessing, and strength, hardness, and toughness cost savings duplex alloys offer. All castings are
can be varied over a wide range. The CB 30 solution annealed and quenched to eliminate em-
and CC 50 alloys are ferritic and, as such, have brittling intermetallic phases.
negligible toughness but effectively deliver The cast duplex alloys may offer a better engi-
corrosion resistance. The toughness of CB 30 neering approach than the equivalent austenitic
can be improved by balancing the chromium cast alloys because they have greater strength
and silicon to a lower part of the range and the and lower alloy cost for the same level of corro-
carbon and nickel to the higher end to render sion resistance. They do not have the same prob-
the microstructure partly martensitic. lems of hot cracking that make casting austenitic
Precipitation Hardening. The cast PH alloys steels difficult. The poor hot workability of
include CB-7Cu-1, which behaves in a similar duplex steel is not an issue for castings. It is im-
way to 17-4PH, which has an overlapping com- portant for designers to understand that cast du-
position range. Note that most other major plex steels are totally compatible galvanically
wrought PH grades rely on titanium and alu- with wrought or cast austenitic alloys of the
minum to form coherent strengthening precipi- same corrosion resistance. Mixing components
tates and so do not have cast counterparts. Cop- with different microstructures does not create a
per, which can harden ferrite but not austenite, galvanic differential when corrosion resistance
is thus the only strengthener available. There is levels are similar. Reluctance to mix alloys for
one cast PH alloy that has no wrought counter- galvanic reasons can be an expensive error when
part. It is CD-4MCu; however, it is rarely used their similar corrosion resistances makes them
in the precipitation-hardened condition and is compatible, even if they are quite different
most commonly classified as a duplex stainless microstructurally.
steel in which the nitrogen level is closely con- Austenitic-Ferritic. The typical CF alloys,
trolled. This is a highly alloyed duplex grade which make up about two-thirds of U.S. stain-
that contains copper to precipitation harden the less steel castings, are nominally austenitic but
ferrite phase. Oil field CO2 corrosion is resisted always contain ferrite. This is not detrimental
by alloys that resemble the martensitic PH and improves resistance to stress corrosion
grades. These alloys are discussed in Chapter cracking and sensitization. Homogenization an-
22, “Petroleum Industry Applications” and can nealing can reduce the amount of ferrite and re-
be considered castable alloys. sult in lower yield and tensile strength and
Duplex. The cast equivalents of alloys 2205 higher elongation and toughness. The composi-
and 2507, J 92205, and J 93380 have similar tion balance is the main determinant of ferrite
properties and corrosion resistance. Modern level. Increasing the nickel, nitrogen, manganese,
wrought duplex alloys rely on nitrogen to parti- or carbon content decreases ferrite. Increasing
tion the alloy with uniform corrosion resistance chromium, silicon, or molybdenum content in-
in each phase and to suppress intermetallic phase creases ferrite. Increasing the solidification rate
formation. Cast alloys are effectively limited to will increase the ratio of austenite to ferrite in
0.25% nitrogen before gas porosity becomes ex- duplex or austenitic-ferritic alloys. The predom-
cessive. Porosity can be reduced by replacing inantly austenitic matrix has a very high tough-
some nickel with manganese, which increases ni- ness even at cryogenic temperatures. Ferrite, if
trogen solubility. Doing so would expand the continuous, decreases toughness. Fortunately, it
most promising area of stainless steel develop- is seldom present as a continuous phase. The
ment, lean duplex alloys such as 2101 and 2003, loss of toughness associated with high ferrite
to the cast grades. Alloy 2101 with 4 to 6% man- content can be aggravated by heating the ferrite
ganese provides the corrosion resistance of CF- above 475 oC (885 oF) for a sufficient time for
8M or 316L with total nickel plus molybdenum the ferrite to decompose to the brittle α and α'.
of only 2% versus the 12% required for the At higher temperatures, development of the σ
austenitic alloy. The duplex alloys also have phase would have a similar embrittling effect.
greater strength and are nearly immune to stress These phases thus formed are quickly redis-
corrosion cracking. These alloys represent signif- solved and removed by annealing. Note that
icant cost-savings potential for the foundry and sometimes copper is added to austenitic alloys
for its customers. CE-30 is duplex steel, which is to improve corrosion resistance in sulfuric acid
fairly simple metallurgically and uses only environments. It has no precipitation hardening
chromium for corrosion resistance. However, its effect in austenite, as it does in ferrite. When
used as a precipitating hardening agent, copper has higher thermal expansion and lower thermal
does not increase corrosion resistance. conductivity than ferrite, which aggravates ther-
Virtually any non-titanium-bearing, corro- mal fatigue and oxide spalling. Nevertheless,
sion-resistant, austenitic, wrought alloy can the better high-temperature strength of austenite
have a cast counterpart. Curiously, the 2xx low- generally is the predominant consideration, and
nickel alloys are not found in most cast alloys most “H” alloys are austenitic. Tables 4 and 5
lists. If a specific wrought alloy cannot be found list properties of “H” alloys (Ref 1 to 3).
to have a published cast counterpart, the de- Ferritic HA, HC, HD. Of the ferritic alloys
signer should not avoid requesting a producer HA, HC, and HD, HA with less than 10%
to supply a version that the foundry is confident chromium is not quite stainless but is useful to
of making. The designer must thoroughly un- 650 oC (1200 oF) for petroleum refinery applica-
derstand the design of the alloy desired so that tions. HC and HD are very high chromium
any alterations to its composition necessary to ferritics that have very low toughness and creep
allow castability will not compromise expected resistance but are quite oxidation and sulfida-
performance. tion resistant. They can be cost-effective materi-
als when high-temperature strength is not an
overriding concern.
Metallurgy of “H” Alloys Austenitic HE-HP. The predominant high-
temperature grades are the austenitic HE
The heat-resisting “H” alloys are principally through HP, after which come the nickel alloys,
austenitic. Alloying elements and impurities which are not generally classified as stainless
diffuse more slowly through the face-centered steels because they contain less than 50% iron.
cubic (fcc) austenitic structure than the bcc The high material cost of the nickel base alloys
ferrite structure, making the austenite more re- restricts their use to those specific environments
sistant to diffusion-controlled creep. Austenite where maximum carburization or nitriding
Table 4 Mechanical properties of heat-resistant cast stainless alloys at room temperature

Tensile strength Yield strength
Alloy Condition MPa ksi MPa ksi Elongation, % Hardness, HB

Standard grades
HA N + T(a) 738 107 558 81 21 220
HC As-cast 760 110 515 75 19 223
Aged(b) 790 115 550 80 18 ...
HD As-cast 585 85 330 48 16 90
HE As-cast 655 95 310 45 20 200
Aged(b) 620 90 380 55 10 270
HF As-cast 635 92 310 45 38 165
Aged(b) 690 100 345 50 25 190
HH, type 1 As-cast 585 85 345 50 25 185
Aged(b) 595 86 380 55 11 200
HH, type 2 As-cast 550 80 275 40 15 180
Aged(b) 635 92 310 45 8 200
HI As-cast 550 80 310 45 12 180
Aged(b) 620 90 450 65 6 200
HK As-cast 515 75 345 50 17 170
Aged(c) 585 85 345 50 10 190
HL As-cast 565 82 360 52 19 192
HN As-cast 470 68 260 38 13 160
HP As-cast 490 71 275 40 11 170
HPNb(d) As-cast 450 220 8 ...
HPNbTi(e) As-cast 450 220 8 ...
HT As-cast 485 70 275 40 10 180
Aged(c) 515 75 310 45 5 200
HU As-cast 485 70 275 40 9 170
Aged(f) 505 73 295 43 5 190
HW As-cast 470 68 250 36 4 185
Aged(g) 580 84 360 52 4 205
HX As-cast 450 65 250 36 9 176
Aged(f) 505 73 305 44 9 185
(a) Normalized and tempered at 675 °C (1250 °F). (b) Aging treatment: 24 h at 760 °C (1400 °F), furnace cool. (c) Aging treatment: 24 h at 760 °C (1400 °F), air cool.
(d) ISO 13583-2 specification minima. (e) ISO 13583-2 specification limits for microalloyed grade. (f) Aging treatment: 48 h at 980 °C (1800 °F), air cool. (g) Aging
treatment: 48 h at 980 °C (1800 °F), furnace cool. Source: Ref 1
Table 5 High-temperature mechanical properties of “H” alloys

Tensile Creep rate Stress to Stress to
Yield strength strength 0.0001%/h 1% in 100,000 rupture rupture
Alloy Temp ksi, MPa ksi, MPa psi, MPa h psi, MPa in 1000 h in 10,000 h
HA 1400 oF 16 27
1800 oF
HC 1400 oF 8.7 10.5 1.3 1.3
1800 oF 2.1 2.5 3.6 0.6
HD 1400 oF 36 3.5 7.0
1800 oF 15 1.0 2.5
HE 1400 oF 3.5 11.0
1800 oF 1.0 2.5
HF 1400 oF 35 6.0 4.4 9.1
1800 oF 1.6 (est)
HH 1400 oF 17 33 3.0
TYPE 1 1800 oF 6.3 9 1.1
HH 1400 oF 18 35 7.0 2.0 8.0
TYPE 2 1800 oF 7 11 2.1 1.6 0.9
HI 1400 oF 6.6 8.5
1800 oF 1.9 2.6
HK 1400 oF 6.8 6.3 12.0
1800 oF 2.7 0.9 2.8 1.7
HL 1400 oF 50 7.0 15
1800 oF 18.7 2.8 (est) 5.2
HN 1400 oF
1800 oF 2.4 2.1
HP 1400 oF
1800 oF 2.1
HPNb(a) 800 oC ... ... 51 MPa ... 55 MPa
HPNbTi(a) 800 oC ... ... 54 MPa ... 64 MPa
HT 1400 oF 26 35 8.0 12
1800 oF 8 11 2.0 2.7 1.7
HU 1400 oF 40 8.5
1800 oF 6.2 10 2.2 2.9 1.8
HW 1400 oF 23 32 6.0 7.8
1800 oF 8 10 1.4 2.6
HX 1400 oF 19.5 42 6.4
1800 oF 6.9 10.7 1.6 2.2
(a) Data from ISO 13583-2. Source: Ref 2, 3
resistance is mandatory. For oxidation and sulfi- alloying also are discussed in the oxidation sec-
dation resistance, the iron base alloys are pre- tion of Chapter 6.
ferred. These cast stainless steels derive their The major problem that all producers of stain-
oxidation resistance from their chromium level. less steels face is that of transferring molten
The chromium near the surface acts as a reser- metal from the furnace to the mold cavity. This
voir to replenish the protective iron/chromium problem is heightened when the foundry makes
oxide scale as explained in Chapter 6 in the sec- complex shapes. Methods developed to protect
tion on oxidation. Silicon, another stable oxide the molten stream from exposure to air to prevent
former, assists in forming this protective scale reoxidation have shown great promise and have
and resistance to carburization. Other typical been demonstrated by the wrought alloy produc-
alloying elements do not aid in oxidation resist- ers who tend to produce much simpler shapes
ance. If it were possible to cast these alloys with than the foundry. Protection of the molten stream
aluminum or rare earth additions without them could result in castings with much better high-
being lost to oxidation before solidification, temperature performance that could be used in-
there could be some impressive benefits. Such stead of some use of higher nickel alloys.
alloys exist in wrought form, for example, High-temperature strength is modestly im-
153MA and 253MA. The metallurgical basis of proved by higher levels of chromium and
the benefits from aluminum and rare earth nickel. Molybdenum improves high-temperature
strength, but its detrimental effect on oxidation higher-performance grades. It is also possible to
resistance and its promotion of intermetallic use AOD-refined master melt stock to achieve
precipitation limits its use. Carbon is very effec- the same benefits as AOD refining while using
tive for promoting high-temperature strength induction melting.
and suppression of intermetallic phase forma- Welding of cast stainless alloys is a common
tion. All “H” alloys, therefore, employ much practice and does not present problems when
higher carbon levels than the “C” alloys. This using approved weld procedures and qualified
does, however, directly imply that the corrosion welders. Chapter 17 describes joining methods
resistance of “H” alloys, should it be an issue, is in detail. The same precautions about sensitiza-
significantly degraded over otherwise similar tion apply to castings. Welding of non-niobium-
“C” alloys. stabilized “C” alloys with carbon levels above
The HP grades have undergone significant 0.03% will require postweld annealing to redis-
development over the last 30 years. This devel- solve chromium carbides, which will otherwise
opment has come about through the addition of make the alloy susceptible to corrosive attack in
niobium to increase creep and rupture proper- the chromium-depleted regions of the heat-
ties. The use of microalloying additions has de- affected zone.
livered creep and rupture properties some 30% Iron and nickel base “H” alloys that are fully
higher than the HP grade without niobium mi- austenitic can suffer from hot shortness due to
croalloying. It is unfortunate that these HP sulfide films that precipitate along grain bound-
grades have not at this time found their way into aries even at low bulk sulfur levels. This makes
ASTM standards; however, work is under way them susceptible to hot cracking of welds. Al-
to remedy this omission. Currently, the most up- loys with some ferrite are less susceptible to hot
to-date collection of these grades can be found cracking, so most “C” alloys are highly resistant
in ISO 13583-2. to this problem.
Foundry Practice REFERENCES
1. Cast Stainless Steels, Metals Handbook,

While the scope of this book does not extend
desk ed., J.R. Davis, Ed., ASM Interna-
to the production of castings, certain aspects are
tional, 1988, p 386–390
important to the user of castings. For the last 50
2. International Organization for Standardiza-
years, virtually all stainless has been refined in
tion, www.iso.org, ISO 13583-2
argon oxidation decarburization (AOD) vessels
3. Steel Founders Society of America, online
or versions thereof. For “C” alloys, this refining
documents: http://www.sfsa.org/sfsa/pubs/
method should be considered a basic require-
index.html
ment for good quality where the carbon levels
are restricted to low levels (e.g., CF3M). “H”
alloys are less refined inherently and can be SELECTED REFERENCE
simply arc or induction melted; however it may
be necessary to use refining techniques for the • http://www.sfsa.org
CHAPTER 12
Melting, Casting, and Hot Processing
Summary austenitic alloys, impossible to produce. The

advent of AOD, continuous casting, ladle metal-
THE PRIMARY PRODUCTION PROC- lurgy, and powerful hot rolling mills has led to
ESSES of melting, casting, and hot processing stainless steels of much higher quality produced
are invisible to the end user. The vast majority of at lower cost. Ironically, the low processing cost
stainless steel is made by arc furnace melting fol- of stainless steel has spurred demand and made
lowed by argon oxygen decarburization (AOD) some of its ingredients, such as molybdenum
refining and continuous casting. It is not normal, and nickel, which are relatively scarce and ex-
and it is seldom beneficial for the end user to pensive commodities, even more costly, forcing
specify processing paths. The end user should, the cost of many alloys to spike even higher
however, be knowledgeable and require the pro- than in earlier years.
ducer to document the process and the producer’s
control of it.
Melting and Refining
Introduction The arc furnace is nearly universally used for

the first step in the production of stainless steel.
The manner in which stainless steel is made The arc furnace is quite flexible in the types of
at the producing mill can have a great impact on charge materials it can accept. Since the charge
its final properties. These production methods materials for stainless steel are typically carbon
have undergone a major evolution over the last steel and stainless steel scrap, this flexibility al-
50 years and are mainly responsible for stain- lows scrap of all types to be used. The necessary
less steels becoming the practical, widespread chromium is added as ferrochromium, whose
engineering materials they are today. Traditional cost is inversely related to its carbon content.
carbon and alloy steel-making methods are not The carbon content of the heat of steel is roughly
suitable for stainless steels. The fundamental 1.5 to 2.5% when it is melted and ready to
difference is that the basic decarburization step, charge into the separate refining vessel.
which is common to all steel making, is thermo- It is this carbon whose removal is the primary
dynamically very difficult in stainless steel be- focus of refining. In the 1960s, Union Carbide
cause the essential element, chromium, reacts engineers perfected a method, the previously
more strongly with the purifying agent, oxygen, mentioned AOD process, of removing nearly all
than does carbon. Thus, early stainless steel mak- the carbon from molten stainless steel without
ing, done in an arc furnace, was a lengthy process significant loss of chromium. This process is
that necessarily involved high chromium losses based on the following chemical reaction:
to the slag as carbon was removed. This process
Cr3O4 (Solid) + yC = yCO (gas) + Cr (Eq 1)
was not only very expensive, the carbon levels
that could be achieved were not much below The equilibrium for this reaction is:
0.10%, making most of today’s stainless steels,
whose carbon levels range from 0.010% in sta- −ΔG
Ln (K ) = (Eq 2)
bilized ferritic alloys to about 0.07% in normal 4575T
where K is the equilibrium constant, and G is steel combine to reduce the oxygen content of
the Gibbs free energy. the steel to around 100 ppm. This could be fur-
Working through the thermodynamics yields ther reduced by aluminum, but aluminum-based
the relationship that summarizes the important inclusions are generally undesirable. The ther-
relationship among carbon, chromium, and CO modynamic activity of aluminum is consider-
(Ref 1): ably reduced in iron as chromium levels in-
crease, so its role as a deoxidizer is less valuable
in stainless steels. Titanium, on the other hand, is
%Cr 13,800
Log = =− + 8.76 − 0.925 log p CO enhanced as a deoxidizer in chromium-iron al-
%C T loys, and consequently small amounts of it are
(Eq 3) sometimes used as a supplementary deoxidant in
alloys even though an alloy specification may
Thus, increasing the temperature works to in- not call for any. Titanium is believed to reduce
crease the elimination of carbon as CO, which hot working defects. More active deoxidants,
evolves from the melt. This is similar in principle such as calcium and magnesium, can be used
to the basic oxygen furnace (BOF) process for when required. Also note that even if no inten-
carbon steel in which oxygen is injected into tional addition of metallic calcium is made,
molten steel to remove carbon by oxidizing it. strong deoxidation with aluminum or titanium
The key to the AOD process, though, is the in- can reduce small amounts of calcium from the
jection of oxygen and argon into the bath to keep CaO in the slag, producing measurable calcium
the partial pressure of CO (pCO) very low. This is content in the metal.
done at a temperature consistent with economic Besides carbon and oxygen, other impurities
refractory life. The injection is done through can be removed from the molten stainless. Once
tubes called tuyeres in the bottom of the barrel- the steel has been deoxidized, sulfur can be
shaped vessel. The injection and the reaction readily removed by contact with a basic slag.
cause extremely thorough mixing, which would Sulfur can be reduced to less than 0.001% in
never happen in the flat, stagnant, arc furnace the AOD, and this excellent purity level is com-
bath. This mixing not only allows the CO-pro- mercially furnished without additional price
ducing reaction to reach equilibrium, but also premium. Sulfur, although a harmful impurity
the mixing of the slag and metal also permits from a corrosion standpoint, is often deliber-
desulfurization. By increasing the ratio of argon ately kept at moderate levels (0.008 to 0.015%)
to oxygen in the injected gas as the refining pro- for tungsten inert gas (TIG) welding penetration
ceeds, the carbon is selectively oxidized with- (see Chapter 17) and at high levels (0.15%+) for
out concurrent chromium oxidation. A typical machinability (see Chapter 15). These trade-offs,
starting ratio is 3 to 1 oxygen to argon/nitrogen which are beneficial to processors, should be
by volume. The ending ratio can be as low as 1 viewed with skepticism by end users, whose
to 9, oxygen to argon/nitrogen. The choice of product integrity is compromised. There are
which inert gas to use, argon or nitrogen, is based processing methods for which higher levels of
on cost and final nitrogen content desired. Stabi- sulfur are not necessary that are preferable to
lized stainless steels require low carbon and ni- the end user while not compromising welding
trogen levels, for instance, so the more expensive or machining costs. For example, machinability
argon must be used. can be improved by calcium additions that pro-
It is possible to use a vacuum system to keep duce malleable oxides to replace the deleterious
the partial pressure of CO low when refining sulfides (see Chapter 15), and welding methods,
with injected oxygen. This is the vacuum oxy- such as laser welding, can be used in many
gen decarburization (VOD) process. The VOD cases to eliminate the need for the weld penetra-
process can achieve slightly lower carbon levels tion enhancement of sulfur while increasing
but does not achieve cleaner steel as some welding speeds.
believe. Phosphorus is an impurity for which no prac-
In both processes, after final carbon content tical removal technology exists in stainless
has been achieved ferrosilicon is added to reduce steel. Any known process to remove it first re-
the chromium in the slag and have it return to the moves chromium. Thus, it exists in almost all
molten steel. The excellent mixing of the slag stainless steel at levels close to its normal speci-
and metal in the AOD permits this to be done fication limit, about 0.030% in austenitic alloys
efficiently. The silicon plus the manganese in the and 0.020% or less in ferritic alloys, which are
Chapter 12: Melting, Casting, and Hot Processing / 157
made from a higher percentage of low-phospho- on interstitial solubility. The higher solubility
rus carbon steel scrap. The deleterious effects of of carbon, nitrogen, and oxygen in stainless
phosphorus on corrosion are not avoided unless steels is significant. A manganese/silicon deox-
much lower levels are achieved. Consequently, idized stainless steel will still have about 100
its presence is tolerated since it has no differen- ppm of dissolved oxygen at the freezing tem-
tial effect over the range in which it is found. perature as opposed to the less than 10 ppm of
Heavy metals are eliminated by high-temper- oxygen found in aluminum-killed carbon steel.
ature AOD blowing, as is hydrogen. Care must This oxygen precipitates as oxides in the solid
be taken not to reintroduce such impurities after state.
refining, which is a risk when using damp or Vacuum induction melting (VIM) is another
contaminated scrap for coolant. method of melting stainless steels. This is a
Alloy adjustment can be done in the AOD or nearly slag-free process, and little refining is
preferably in a treatment-and-transfer ladle. The possible. Melt purity is largely controlled by the
tapped molten steel generally has excess heat purity of the starting material, and use of AOD
from the highly exothermic refining process. master melt stock for VIM remelting is com-
This allows the composition to be measured and mon. Limited decarburization is possible via in-
adjusted before it must be cast. This can be done jection of oxides such as Fe3O4 or SiO2 to create
very precisely by wire feeding of alloying ele- CO evolution inside the vessel. Using this tech-
ments through the slag into the heat, which can nique, very low carbon levels (less than 50
be stirred by argon bubbling via porous plugs. ppm) are achievable commercially. Use of VIM
This technique is very effective for the fine- is generally limited to high-value, high-purity,
tuning of reactive elements such as titanium. or low-tonnage melts.
The refining treatments used for carbon steel
and stainless steel are very similar, but there are
subtle differences because of the difference in Remelting
the thermodynamics of dilute solutions like car-
bon steel and highly alloyed, nondilute solutions Some stainless steels and related alloys are
like stainless steel. Table 1 shows the factors by remelted to refine composition or ingot struc-
which additions of various elements to stainless ture. There are two principal remelt processes:
steel (j) alter the thermodynamic activity of vacuum arc remelting (VAR) and electroslag
other alloying elements (i). remelting (ESR).
Equation 4 is used to calculate the activity of In VAR, the material to be remelted is cast
elements in steel. The activity coefficient γ into a cylindrical electrode and placed inside a
varies with the concentration of alloying ele- cylindrical water-cooled vacuum chamber. A
ment x by: high-current direct current (dc) arc is estab-
lished between the electrode and a starter plate
n
δ ln γ i at the bottom of the chamber. The end of the
RT lnγ i = RT lnγ i0 + ∑
j =1... n
RT
δxj (Eq 4) electrode is melted, and the molten drops fall
through the intervening vacuum. Volatile con-
stituents escape from the molten drops, and the
This calculation is best left to computer pro- purified drops collect to form a molten pool on
grams such as Thermo-Calc that have been per- top of the starter plate. VAR parameters are ad-
fected for these lengthy procedures. It should justed to maintain a shallow pool, which solidi-
be noted that chromium, which is always pres- fies in a bottom-up fashion. The shallowness of
ent in nondilute quantities, has a powerful effect the molten pool produces a refined grain
Table 1 Influence of alloying elements on the thermodynamic activity of carbon, nitrogen,

sulfur, and oxygen
J
i Al C Cr Mn Mo N Ni O S Si Ti W
O .04 .14 –.02 –.01 –.01 .11 .01 –.34 .05 .08 ... –.005
N –.03 .13 –.05 –.02 –.01 0.0 .01 .05 .01 .05 –.53 –.001
S ... .11 –.01 –.03 .003 .01 0.0 –.27 –.03 .06 –.07 .01
0 –.39 –.45 –.04 –.02 .003 .06 .006 –.20 –.13 –.13 –.6 –.01
structure with less solidification segregation ceramic tube, called the submerged entry
than found in typical cast product. nozzle, into the mold, which is covered with a
In ESR, the material to be remelted is cast consumable protective and lubricating slag
into an electrode of similar shape, but slightly cover, called a mold powder. The mold powder,
smaller than the water-cooled mold. A gap be- which melts in the mold as it is added, contains
tween the electrode and a starter plate at the bot- ceramics, fluxes, and carbon. The level of the
tom of the mold is filled with a prepared slag. molten metal should be carefully controlled by
Typically, this slag is calcium fluoride-based ultrasonic measurement, or other methods, to
with high lime (CaO) content. Additional ingre- prevent fluctuations in level that may entrap
dients control the basicity, fluidity, oxidizing slag in the slab surface. The entire water-cooled,
potential, and other properties of the slag. A copper alloy mold oscillates in a precise pattern
high current is used to melt the slag, which in as the solidifying strand of steel is withdrawn
turn melts the end of the electrode, and the from the mold bottom by pinch rolls and
molten drops fall through the slag. Reaction of sprayed with water to cool it. The pinch rolls
the molten drops with the slag removes sulfur apply enough pressure to slightly deform the
and some other impurities, and the purified slab. This deformation has a crucial, seldom-rec-
drops collect to form a molten pool on top of the ognized effect. It causes a beneficial recrystal-
starter plate. ESR melting typically is done at a lization that improves hot working characteris-
higher rate than VAR, and the molten pool is tics of austenitic and duplex alloys. In ferritic
deeper. This deeper pool produces a grain struc- alloys, it can cause excessive grain growth,
ture between that of VAR and typical cast prod- which detracts from hot workability. The initial
uct, with commensurate intermediate segrega- portion of slab cast in a sequence is seldom
tion patterns. of adequate quality to be used because of
exogenous inclusions, entrapped mold powder,
and non-steady-state solidification structure.
Casting The defective portion must be identified and
scrapped or diverted to low-quality requirement
Continuous slab, billet, and bloom casting end uses.
have become the standard methods of making The strand is bent from an initial slightly
stainless steel primary products, replacing the curved shape to flat and cut into slabs. More
obsolete ingot method. There are some alloys than one heat of steel may be cast sequentially
that cannot be continuously cast, but these repre- without restarting the process. This is ideal eco-
sent a miniscule percentage of stainless produc- nomically and for quality reasons since initial
tion. Continuous casting produces slabs directly, and final segments of a casting can contain
thus removing the costly soaking and slab- more inclusions and aberrant structure. Some
rolling processes. In a well-executed continuous end users stipulate that no first slabs be applied
casting operation, slabs are of sufficient quality to their orders. Producers generally apply first
that they require no surface conditioning before slabs to less-critical uses or discard suspect sec-
being hot rolled. Slabs range in thickness from tions of them. If casting conditions are not opti-
13 to 63 cm (5 to 15 in.). The segregation in con- mal, the result can be slabs with poor surface
tinuous casters is less than in ingots because of quality that must be surface ground.
the smaller section size. It is not eliminated, Slabs are sometimes quenched to avoid pre-
however, and certain alloying elements concen- cipitation of phases; however, they may be held
trate at the centerline, where they defy homoge- at high enough temperatures prior to hot rolling
nization. Carbon and molybdenum are examples to stay above the temperature range in which
of alloying elements with this tendency. embrittlement can occur or to stay above the
In properly executed continuous casting, the temperature at which an embrittled slab can
ladle feeds by a slide gate, or preferably a stop- fracture. Ferritic and martensitic alloys are es-
per rod gate, into a ceramic tube into the large pecially prone to these problems.
tundish situated over the caster mold. The metal There has been great interest for decades in
in the tundish is covered with a protective slag producing stainless steel coils directly from
cover, and flow patterns within the tundish are the melt in so-called strip casters. Elimination
designed to minimize dead spots and encourage of hot rolling could be quite valuable in stain-
removal of inclusions by impingement with the less steel, whose hot rolling from slab can be
slag cover. The metal feeds through another both expensive and problematic. There are a
Chapter 12: Melting, Casting, and Hot Processing / 159
number of such machines in pilot or limited mills that routinely roll carbon steel to 1.5 mm
production. They have not had sufficient com- (0.06 in.) can struggle to attain 4.5-mm (0.18-in.)
mercial or technical success to have become a thickness for 316 stainless.
factor in the industry. Since their development The high separating forces on the hot rolling
is only being undertaken by those large stain- mill stands also cause greater roll deflection and
less steel producers who already have the hot compression, which if not countered by roll
rolling assets that strip casting would replace, bending or roll shifting schemes can lead to sig-
it seems unlikely that strip casting will soon nificant variation in thickness across the sheet,
become a major factor even if it is perfected as much as 0.25 mm (0.01 in.). This variation as
technically. a percentage of thickness is not reduced by cold
Another method of shortcutting the casting/ rolling and is a major cause of tolerance loss in
ingot step has been perfected: the powder met- sheet and strip. Hot-rolled bands vary in thick-
allurgy approach. In powder metallurgy, the re- ness along the length of the coil because the tail
fined molten metal is atomized by gas or liquid end of the slab is colder and harder to roll. Coil
and made to freeze into small particles. These boxes (on reversing mills) address this problem
particles, having been quenched extremely to a degree by permitting the semirolled coil to
rapidly, are quite homogeneous. Powder tech- equalize in temperature.
nology methods allow for the design of alloys Hot strip tandem mills powerful enough to
that would otherwise freeze with too much successfully roll high-quality stainless steel hot-
segregation and too coarse a structure with rolled bands are massively expensive and are
conventional production methods. Traditional seldom justified for the tonnage of stainless steel
powder metallurgy production methods are rolling a given melt shop produces, although
used to make small near-net shape compo- rolling stainless on hot tandem mills used pri-
nents, avoiding most of the costly machining marily for carbon steel can be an excellent pro-
steps. More impressively, powder technology duction method.
is also used to produce massive components. Hot Steckel mills have become the favored
For example, very high carbon/vanadium method of hot rolling stainless steel because
stainless tool steel components can be made by their throughput better matches stainless steel
encapsulating powder in an evacuated canister melt shop production outputs. This permits the
in which it can be sintered and hot worked to melt shop and caster to be adjacent to the hot
100% density and virtually complete homo- mill, which permits energy-saving hot charging
geneity. Chapter 9 on martensitic alloys dis- of slabs. In hot Steckel mills, typically a four-
cusses these materials. high reversing rougher rolls slabs to about 3-cm
(1.2-in.) thick. Then, the transfer band is rolled
to final gauge on a separate reversing four-high
Hot Rolling finishing mill with coil boxes to preserve tem-
perature. The economy of having only two mill
Hot rolling remains an essential process for stands makes these mills ideal for typical stain-
the vast majority of stainless steel used. Hot less production quantities and permits the cost
rolling characteristics of stainless steels vary of sophisticated mill capabilities, such as roll
greatly. Ferritic stainless steels are extremely shifting, roll crossing, or roll bending, not to
easy to hot roll since they have a soft, single- have to be duplicated among many stands. This
phase structure at hot rolling temperatures. is the same justification for using Sendzimir
Martensitic stainless steels roll like their carbon mills to cold roll stainless. In both cases, the
and alloy steel counterparts since their mi- logic applies more to austenitic alloys than to
crostructure during hot rolling is a moderately the easily rolled ferritic stainless alloys.
alloyed austenite similar to alloy steels. The mi- In either case, the hot-rolled band carries a
crostructure during hot rolling is the crucial fac- heavy, embedded scale that must be removed
tor. Austenitic stainless steels have high strength from the surface before further processing in
at hot rolling temperatures. Furthermore, the most cases. Some alloys can be cold rolled in the
low diffusion rates in austenite slow recrystal- “black band” state at a cost of coarser surface
lization so that the steel does not always soften finish and greater rolling mill roll wear. If nor-
between stands in tandem mills. This increases mal cold rolling or use as hot-rolled coil is fore-
mill loads, and lower reductions must be taken seen, the hot-rolled band must then be annealed
than for alloy steels. Powerful hot strip tandem and pickled since the as-rolled hot-rolled band
has poor corrosion resistance, poor mechanical rily as edge cracks and slivers. Edge cracks are
properties, residual cold work and hardness vari- simply a lack of ductility at the colder strip edge.
ations, as well as a heavy oxide layer. Stainless hot ductility often has a narrow tem-
perature window, and many factors can affect
the size of that window depending on alloy type.
Defects The most inherently challenging alloys for hot
working are the duplex alloys and the alloys that
Stainless steel hot-rolled bands can contain solidify in the fully austenitic state. The former
many types of defects. These are seldom seen has a mixed-phase structure, and the phases can
by the end user because they are removed exhibit mechanical incompatibility at certain
when they are not prevented. They do have temperatures. The latter alloys reject sulfur and
repercussions on delivery. The major cate- oxygen during solidification and slab reheating
gories are: to the grain boundaries, where they form very
• Hot mill defects weak films. But, even alloys such as 304 and
• Inclusion-related defects 316 can have very poor hot ductility if they con-
• Hot ductility-related defects tain much sulfur and oxygen or if they are re-
heated for long times or at temperatures above
Stainless steel is less forgiving of hot mill 1250 °C (2280 °F), which facilitates diffusion of
faults because its surface is not removed by oxi- sulfur and oxygen to the grain boundaries and
dation to the degree carbon steel’s surface is. also encourages very large grains. This poor hot
Thus, a skid mark from a slab-heating furnace ductility manifests itself as “slivers,” which can
will remain through the hot rolling, annealing, require grinding of the entire hot band surface.
and cold rolling processes. This is true of all hot These tendencies are fought by low oxygen and
mill scratches, gouges, digs, etc. Rolling stain- sulfur levels and minimal slab-reheating temper-
less requires a different mindset than rolling atures and times, as well as slab surface working
carbon steel, which argues against the benefits in the caster pinch rolls. Sometimes, very poor
of rolling stainless on a mill built and primarily hot working alloys are given a single hot reduc-
used for carbon steel. tion pass on a hot mill to produce a full recrystal-
Inclusion-related defects are all essentially lization that disperses grain boundary-weakening
avoidable by using state-of-the-art technology. elements on subsequent reheat.
Protecting metal from reoxidation and keeping
precise mold-level control in the continuous
caster prevents all inclusions of a size that can REFERENCE
produce a defect.
Hot ductility defects are more subtle. They 1. D. Peckner and I.M. Bernstein, Handbook of
arise from many causes and are manifest prima- Stainless Steels, McGraw-Hill, 1977, p 3–13
CHAPTER 13
Thermal Processing
Summary desirable, and some are potentially very detri-

mental. Readers are encouraged to review the
THE THERMAL PROCESSING of stainless earlier chapters on phases in stainless steel
steel is a topic the end user should approach (Chap. 6–10) to familiarize themselves with
with great respect. It is not simple in concept or these phases.
in practice. Before attempting to carry out any Each of the alloy groups of stainless steels has
thermal processing on stainless steel, the practi- radically different thermal processing objectives
tioner must understand the alloy design, compo- and requirements; therefore, each is discussed
sition, and processing history of the material in separately.
question. The thermal processing then must be
designed and executed in a planned, controlled
manner. The consequences of failure in thermal Austenitic Stainless Steels
processing can become catastrophic to mechan-
ical properties and corrosion resistance. Thermal processes applied to austenitic stain-
less steels include:
• Soaking for homogenization and preparation
Introduction for hot working
• Annealing to remove the effects of cold
The thermal processing of stainless steels can work and to put alloying elements into solid
have many purposes. Normally, the objectives solution (solution annealing)
are simple: heating for hot working, annealing to • Stress relieving
soften after cold working, solution annealing
The temperatures at which these processes
to homogenize, heating to temper martensite, or
are carried out are shown in Table 1 for typical
to stress relieve. However, even if the objective
austenite compositions.
is simple, the processes that occur are anything
but simple. Variations in temperature, times at
temperature, heating and cooling rates, and at- Soaking
mosphere can have complex and easily unin- Because virtually all stainless steel is continu-
tended consequences. There is no substitute for ously cast, the older soaking process of holding
understanding the processes that are occurring ingots in soaking pits for many hours is rarely
when stainless steels are heated for successful used. The soaking had two functions. The obvi-
heat treating to be achieved. ous one was to equilibrate at the right tempera-
Stainless steels have many alloying elements ture for hot working. The second, less-obvious,
in large amounts. Many of these elements are one was to achieve greater chemical homogene-
highly reactive thermodynamically. The practi- ity. The lack of homogeneity comes from the
cal consequence of this is that many phases are solute segregation that occurs as solute ele-
thermodynamically possible at different temper- ments are rejected from the material that was
atures. Stainless also reacts with its environment first to freeze. Solute segregation was exagger-
at high temperatures, causing changes in surface ated by the slow solidification of ingots, and
alloy content. Some of the resulting phases are continuous casting helped make the stainless
Table 1 Recommended thermal processing temperatures for austenitic alloys

Annealing Annealing ASTM A480(a) Stress Stress
Alloy temperature, °C temperature, °F 2006, °F relieving, °F relieving, oC
Standard alloys ... ... ... 1500–1600 non-L, 815–870 non-L,
1000–1600 L grades 540–870 L grades
201, 202, 201LN 1010–1120 1850–2050 1900 min ... ...
301, 301LN, all versions 1010–1120 1850–2050 1900 min ... ...
304, 304L, 305, all versions 1010–1120 1850–2050 1900 min ... ...
316, 316L, 316N, 317, 317L 1040–1175 1900–2150 1900 min ... ...
308, 309, 309S, 310, 310S 1040–1175 1900–2150 1900 min ... ...
Stabilized alloys ... ... ... 1000–1600 540–870
321 955–1065 1750–1950 1900 min ... ...
347, 348 980–1025 1800–1950 1900 min ... ...
20Cb-3 925–955 1700–1750 ... 925–1010 ...
Moderately alloyed, Creq<26 1120–1175 2050–2150 Various 1500–1600 815–870
S31725, N08028, JS700
Highly alloyed, lower sigma 1120–1175 2050–2250 Various Not recommended ...
alloys, Creq>30, high N
AL6-XN, 4565, 654SMO,
254SMO
Highly alloyed, sigma-prone 1205–1230 2200–2250 Various Not recommended ...
alloys, Creq>30, low N
AL6-X
(a) Standard specification for general requirements for flat-rolled stainless and heat-resisting steel plate, sheet, and strip.
steel more homogeneous. Nevertheless, cast rounding matrix by sufficient soaking. Welds that
slabs and blooms must be soaked to eliminate are unannealed have such precipitated inclusions
as-cast segregations. This process, to the extent in an unequilibrated state, and the result is dimin-
it is done, occurs as they are reheated to the ap- ished chromium concentration and poorer pitting
propriate temperature for hot working. Soaking resistance.
dissolves the few percent of residual delta fer-
rite that are present on slab solidification. It is
important to soak at the highest temperature at Annealing
which delta ferrite is not a stable phase so that it Annealing serves two main functions in
will dissolve, about 1250 °C (2280 °F) for most stainless steel: It removes the effect of cold
austenitic stainless steels. work by replacing strained microstructure with
Soaking at higher temperatures causes ferrite new strain-free grains, that is, recrystallization.
levels to increase, negating the homogenization New grains nucleate and grow. If stored strain
and causing very poor hot workability. Longer energy is insufficient, as happens often with
times at temperature than the minimum required ferritic stainless steels, true recrystallization is
for thermal uniformity also cause problems as difficult to achieve, and the annealing process
any sulfur and oxygen impurities are rejected may only produce recovery without recrystal-
from austenite and can diffuse to grain bound- lization, leaving the same grains relieved of
aries, where they form weak, plastic films that stress. This leaves the surviving grains with the
also degrade hot workability. Grain growth, by same crystallographic orientation that deforma-
reducing grain boundary area, exacerbates this tion produced and may or may not be the de-
effect. Thus, soaking times are best minimized sired outcome. Second, annealing returns into
and closely controlled. Alloys are therefore de- solution solute that has been precipitated as un-
signed to have only a slight amount of delta fer- wanted phases, principally carbides, but also
rite to be redissolved during soaking. Ferrite is intermetallic phases. Annealing also may help
useful because it has a high solubility for oxygen to reduce solute segregation remaining from
and sulfur. Having none would result in impurity the casting process, making the composition
rejection of these elements to grain boundaries more homogeneous. The homogenization
during solidification, which results in the worst- process is accelerated by the reduction in di-
possible hot working characteristics. The oxygen mensions from hot and cold working. A reduc-
and sulfur trapped in the ferrite during solidification in dimension by a factor of two reduces the
tion precipitates in the solid state as inclusions, time to achieve a given degree of homogeniza-
which also must be equilibrated with the sur- tion by a factor of four.
Chapter 13: Thermal Processing / 163
Annealing to recrystallize is fairly rapid. To a carbides can be redissolved and carbon dif-
first approximation, it is instantaneous, and the fused away from the carbide, but this does not
results are merely a function of the maximum mean that all composition gradients have been
temperature attained. This may not be the case reduced to zero. This means that precipitates
for continuous annealing lines, in which transit may re-form more rapidly in such a material
time can be short enough, less than a minute at than they would in a completely homogeneous
temperature, to limit the grain size attained. The alloy.
driving force for recrystallization is the strain The annealing temperature for a given alloy is
energy stored in the lattice from deformation. chosen based on the temperature required to put
The strain energy in a given material is propor- all alloying elements into solution. Higher car-
tional to the square of the material’s flow stress. bon levels, for instance, require higher tempera-
As the material is heated, recovery occurs first. tures to dissolve all the carbon. This is one of
This is the change in physical and mechanical the principle values of accurate phase diagrams.
properties associated with dislocation annihila- If it were simply a consideration of recrystal-
tion and polygonalization that occurs before the lization, all alloys could be annealed at similar
nucleation and growth of new grains. temperatures at the low end of the recom-
The nucleation of new grains occurs at high- mended range. Within the recommended range,
angle grain boundaries and proceeds by the the temperature selected should be determined
movement of roughly hemispherical growth by the desired grain size. End use determines
fronts into strained areas. The percentage re- whether a fine or coarse grain size is preferable.
crystallized, once a sufficient temperature is Table 1 lists recommended annealing tempera-
reached, grows sigmoidally. It is normal for tures for austenitic stainless steels.
the time to fully recrystallize to be rather less The overall interplay between prior cold work
than the time to attain that temperature. Even and annealing temperature on mechanical prop-
at the lower range of annealing temperatures, erties of annealed material can be summarized
times are generally less than 1 min. Recrystal- as (Ref 3):
lization will not occur if the lattice contains in-
sufficient strain energy. Thirty percent cold • Grain size of a given alloy is the most im-
work should be used as a rough threshold for portant parameter in characterizing mechan-
the required amount. Annealing after lower ical properties.
amounts of cold work is characterized by • Yield and tensile strength are essentially
scarce nucleation sites and can result in very constant for a given grain size regardless of
large and nonuniform grain size. Hot-worked the amount of prior cold work; however, the
material often has a composite structure that elongation depends on the prior reduction.
may have already had some recrystallization • Yield strength, tensile strength, and hardness
depending on the final reduction temperature. are essentially linear functions of grain size.
Annealing may not produce a clear recrystal- • Elongation decreases with finer grain size
lized structure in this case. and at an increasing rate as grain size be-
The relative rapidity of recrystallization an- comes finer as long as cross-section size of
nealing is due to the fact that it is rate con- the test specimen is not extremely small.
trolled by short-range diffusion. Solution an- This is not true for very coarse-grained ma-
nealing requires longer-range diffusion and terial.
thus can require much longer times. Some stud- • Maximizing elongation comes from maxi-
ies have shown that welds, for instance, do not mum prior cold work and medium annealing
recover completely from their loss of corrosion temperatures
properties that arise from local alloy depletion • Anisotropy coefficients, or plastic strain ra-
unless they have been annealed for times on the tios r are constant up to about 40% reduction
order of 1 h (Ref 1). Others have seen homoge- after, which r45 and rn increase sharply, while
nization in as little as 10 min (Ref 2). Wrought rt decreases. This leads to earing during
materials can require much shorter times be- drawing.
cause reductions during hot working have re- • The increase in properties for a one ASTM
duced diffusion distances. It should be noted grain size increment is:
that precipitates can be redissolved and not ap- a. 13 MPa (2 ksi) for tensile strength
parent in the annealed microstructure without b. 20 MPa (3 ksi) for yield strength
full homogeneity being achieved. For example, c. 2 HRB for hardness
Atmospheres for annealing are important. metallic phases. Use of high-chromium and-
Austenitic stainless steels heated in air, of molybdenum alloys without enhanced nitrogen
course, form oxide scales. Beneath this oxide, is no longer recommended, and the use of lower-
the metal matrix becomes significantly depleted nitrogen alloys should be reexamined and ques-
of chromium (Ref 4), often more than 5% lower tioned if specified.
in chromium and to a depth of as much as 10 µ Last, stainless surfaces should be scrupu-
(395 µin.). So, not only must any oxide be re- lously clean before annealing. Even hard water
moved, so must the chromium-depleted layer. deposits can cause differential oxide growth,
This requires aggressive pickling, which while which can cause etched spots on the surface,
done commonly, may not be practical for many where the postanneal pickling attacks the differ-
stainless users. The chromium-depleted zone, ent oxide more strongly. Carbonaceous materi-
however, does pickle rapidly precisely because als left on the surface are even more objection-
it does have less chromium. To avoid oxide able because they can cause carburization and
scale formation, vacuum, hydrogen, or inert gas subsequent loss of corrosion resistance.
atmospheres may be used. Stabilizing anneals are sometimes conducted
If vacuum is used, it should be less than on stabilized alloys such as 321 and 347. This is
2 × 10–3 torr (0.3 Pa). If an inert gas or hydrogen useful when carbon levels are sufficiently high
is used, the key consideration is moisture con- that significant dissociation of carbides occurs
tent. The dew point must be –40 °C (–40 °F) or at annealing temperatures. A second anneal at
lower. More stringent levels may be required if lower temperature, about 900 °C (1650 °F),
mirror finishes are desired. Cool down must be then is done to permit the carbon to combine
rapid as oxidation potential increases as tempera- with the stabilizing element rather than leaving
ture decreases. Vacuum or inert gas is preferable it available to form chromium carbides. Current
to hydrogen for alloys containing stable oxide preferred practice for these alloys is to maintain
formers such as aluminum or titanium or for al- carbon and nitrogen below 0.03% for corrosion-
loys containing boron. resistant service, which renders this stabilizing
Austenitic alloys that are subject to sensitiza- unnecessary. Alloys used for high-temperature
tion must be cooled rapidly enough from anneal- service benefit from the creep-resisting contri-
ing temperatures to avoid carbide precipitation butions of higher carbon levels.
during cooling. If forced air or water quenching
are impractical or if section size prohibits rapid
cooling, then using stabilized or low-carbon Stress Relieving
grades is indicated. Austenitic stainless steel weldments often con-
Superaustenitic stainless steels, and even al- tain residual stresses, which can cause distortion
loys like 317, present a special problem because or lead to stress corrosion cracking in service.
these alloys have significant sigma-forming ten- They are commonly stress relieved at tempera-
dencies. Sigma forms initially because solidifi- tures slightly below the annealing temperature,
cation segregation causes local enrichment of so that residual stresses may be relieved by
sigma-promoting elements, such as molybde- creep. One hour at 900 °C (1650 °F) reduces
num. It can also form from slow cooling of slabs residual stress by about 85%. Lower tempera-
or hot bands. This latter sigma forms at grain tures require exponentially longer times for the
boundaries and will cause embrittlement and re- same stress relief, with times doubling for each
duced corrosion resistance, so it must not only 100 °C (180 °F) decrement as decreasing diffu-
be redissolved, but also the alloy must be ho- sion rates, which govern creep, are encountered.
mogenized to remove the residual concentration Cold-worked austenitic stainless steels have a
gradients from the sigma. If this is not done, markedly diminished proportional limit, partic-
chromium- and molybdenum-depleted regions ularly in compression. This Bauschinger effect,
will still exist, and sigma will re-form much which arises from the easy mobility of disloca-
more rapidly during subsequent exposure to tions, can be eliminated by stress relieving at
high temperatures. For this reason, the higher around 350 °C (660 °F) for 2 h, which provides
ends of the annealing ranges are recommended, the thermal energy for dislocation interactions
and annealing times should be generous. Newer to lock into place. This produces a sharp yield
alloys have higher nitrogen contents to suppress point without premature nonproportional elastic
formation of sigma and other deleterious inter- deformation.
Ferritic Stainless Steels phases that require dissolution. Alloys with high
chromium and molybdenum contents can form
Ferritic stainless steels, from an annealing σ and/or α', the brittle, ordered body-centered
point of view, must be discussed in two cate- cubic (bcc) phase, at temperatures below an-
gories. First are the modern, stabilized alloys, nealing temperatures, so rapid cooling is pru-
which are ferritic at all temperatures. These al- dent when chromium plus molybdenum ex-
loys behave as interstitial-free (IF) alloys be- ceeds 20%.
cause the interstitial carbon and nitrogen are re- The driving force for recrystallization in these
moved from solution as a stable precipitate. In alloys is limited by the lower stored energy
the second category are the older ferritic steels, from deformation inherent in the bcc structure.
which have enough austenitizing elements, usu- In addition, the pronounced deformation texture
ally carbon, in solution to cause them to form leads to annealing responses that are more accu-
austenite at what would otherwise be a good an- rately characterized as recovery and grain
nealing temperature. This makes them truly growth with diminished recrystallization. These
quasi-martensitic alloys, and they must be alloys retain this texture after annealing, and
treated accordingly. Table 2, which lists heat- this characteristic anisotropy is exploited for
treating temperatures for ferritic stainless alloys, good drawability. The major concern is to avoid
also shows which grades fit into which category. excessive annealed grain size, which greatly re-
duces toughness. Anneal at the higher end of the
Soaking range only if the loss of toughness associated
Heating of ferritic stainless for hot working is with large grain size is not a concern. Stabiliz-
straightforward. Whether stabilized or not, ing anneals are normally unnecessary for stabi-
these alloys are heated to the 1000 to 1100 °C lized ferritics as their high diffusion rates ensure
(1830 to 2010 °F) range for hot working. The freedom from knife-edge attack due to sensiti-
superferritics can be heated to up to 1300 °C zation from free unbonded carbon combining
(2370 °F). At this temperature, no debilitating with chromium at grain boundaries. The stabi-
phases occur, and ductility is good. The high lizing additions of titanium and/or niobium tie
diffusion rate inherent to the ferritic structure up the carbon as stable TiC or NbC, which does
makes homogenization easy. As long as hot not redissolve during annealing.
working is completed at temperatures above The interstitial-bearing ferritic stainless
that at which austenite forms, good hot ductility steels must be annealed subcritically, or the for-
is expected. This is not a concern with IF alloys, mation of austenite at higher temperatures
which do not form austenite. would make martensite formation on cooling
virtually unavoidable. Thus, a typical primary
Annealing anneal cycle for a typical alloy such as 430
would be nearly 24 h at 750 °C (1380 °F), the
The IF ferritics do not undergo any phase majority of which is thermal equilibration of
change with temperature during the course of the large coil mass. The actual time at tempera-
properly executed heat treatment. The objective ture required is less than 1 h. Continuous an-
of annealing is generally simply to remove the nealing is not practical because the diffusion of
effects of cold work. This is because they do not carbon is too slow to occur in the dwell time at
need to have carbon put into solution and, ex- temperature typical in continuous annealing
cept in rare cases, do not have intermetallic lines. This cycle also precipitates essentially all
the carbon and nitrogen as mixed Cr/Fe car-
Table 2 Recommended annealing temperatures bides and nitrides and homogenizes chromium
for ferritic alloys content. This necessarily slow process permits
Annealing Annealing
Alloy temperature, oC temperature, oF
subsequent subcritical annealing for mechani-
Stabilized, Cr+Mo<20
cal properties (to alleviate the effects of cold
409, 439,18 SR 870–925 1600–1700 work) to be done in a few minutes since carbon
Unstabilized, Cr+Mo<20 has been eliminated from solution by the for-
405, 430, 434, 436 705–790 1300–1450
Stabilized, Cr+Mo>20 mation of fairly stable carbides. Since the ma-
29-4C, Monit, Seacure, 444 1010–1065 1850–1950 terial is generally purchased in the annealed
Unstabilized, Cr+Mo>20, condition, the user need never be concerned
446 760–830 1400–1525
with such lengthy anneals.
Stress relieving is rarely a concern for any martensite cannot be avoided by furnace cool-
type of ferritic stainless. Unstabilized grades ing from austenitic temperatures, then only sub-
should not be welded, and if they are, full subcritical annealing is feasible. But, even for
critical annealing is required. Stabilized grades nickel-free alloys the hardenability is so great
have no need for postweld heat treatment. Low- that annealing by slow cooling is quite difficult.
temperature heat treatment runs the risk of α' Martensitic alloys are put into the annealed con-
formation and is best avoided. dition for processing before they are quenched
and tempered for their final use. Thus, the more
economic subcritical anneal is the predominant
Martensitic Stainless Steels annealing heat treatment.
The nickel-bearing alloys have such high
The martensitic stainless steels resemble the hardenability that annealing in the critical range
unstabilized ferritic stainless steels described. cannot produce softening by any practical cool-
The martensitic stainless steels form essentially ing rate, so subcritical annealing is always rec-
100% austenite on heating and have very high ommended for these alloys. Nickel reduces the
hardenability, so their ability to be softened by temperature at which austenite is stable as
annealing is limited. The traditional martensitic shown in Chapter 9, Fig. 9. Other additions like
stainless steels are iron/chromium/carbon al- vanadium, molybdenum, and tungsten promote
loys, sometimes with a small amount of nickel secondary hardening and tempering resistance,
and/or molybdenum. More recently, alloys have and subcritical annealing of these alloys be-
been developed for petroleum applications that comes a slow, difficult process. This is a charac-
contain high copper, nickel, and/or molybde- teristic of the so-called super 12Cr alloys.
num and low carbon. The principles of heat Martensitic alloys have lower corrosion re-
treatment of the two alloy categories are the sistance in the annealed condition than in the
same. The more highly alloyed newer alloys hardened condition because in this state they
are, in fact, simpler to heat treat because their have the maximum amount of chromium tied up
low carbon and nitrogen levels alleviate the as chromium carbide.
need to temper.
Austenitizing
Soaking Table 3 lists the austenitizing and tempering
Hot working should be carried out in the ranges for martensitic stainless steels. Full
austenitic range. Temperatures for this are austenitizing is crucial to producing a micro-
listed in Table 3. Forging and hot working structure that is fully martensitic. Only austenite
should always be followed by annealing to transforms to martensite. If other constituents,
avoid stress cracking due to the deep hardening such as δ ferrite or carbides, exist during the
of these alloys. austenitizing heat treatment before quenching,
they will not transform to martensite. Some
alloys, such as the 440 group, have enough car-
Annealing
bon that the austenitizing temperature deter-
Martensitic stainless steels can be annealed mines how much carbon is put into solution.
by subcritical anneal and sometimes by full an- The carbon in solution in the austenite will be-
neal depending on alloy level. If the alloy level come the carbon level in the martensite, which
is such, as in the nickel-containing grades, that directly determines strength and corrosion
Table 3 Recommended annealing, austenitizing, and tempering temperatures for martensitic alloys
Alloy Subcritical anneal, Full anneal, Austenitizing, Tempering, low Tempering, high
o
C (oF) o
C (oF) o
C (oF) range, oC (oF) range, oC (oF)
Straight 650–760 830–885 925–1010 205–370 400–700
Cr, C<0.20, 410, (1200–1400) (1525–1625) (1700–1850) (565–605) (1050–1125)
416,403
Ni/Mo, C<0.20, 620–705 Not recommended 980–1065 205–370 400–700
414, 431, 415, (1150–1300) (1800–1950) (565–605) (1050–1125)
425, C>0.20
440A,B,C F, 420 675–760 845–900 1010–1065 150–300 Not recommended
(1245–1400) (1555–1650) (1850–1950) (300–700)
resistance because undissolved carbides contain precipitate with the abundant iron atoms first. At
chromium, which diminishes that available for higher temperatures and longer times, more ther-
corrosion resistance. Austenitizing temperature modynamically stable carbides, such as Cr23C6,
and holding time become most significant when form. Carbide formation is a complex function
carbon exceeds 0.20%, where its solubility is a of temperature, time, and composition. The
steep function of temperature. growth of carbides reduces strain and hardness.
The ␦ ferrite is a generally undesirable phase There are exceptions, such as the precipitation of
that can be produced by temperature excursions Mo2C, whose morphology produces a precipita-
or composition variations. Excessive austenitiz- tion hardening (PH), called secondary harden-
ing temperatures can cause its formation, as can ing. Niobium and vanadium also form carbides
low levels of carbon, which may be originally that result in higher hardness at all tempering
present in the alloy or arise from decarburiza- temperatures. Had this been understood early
tion. It will cause lower hardness and toughness on, these steels could have been correctly in-
if present. cluded in the PH stainless group. The PH steels,
Heating rates should be such that a uniform AM-350 and AM-355, both derive their PH
temperature is attained before the allotropic from the precipitation of Mo2C and Mo2N. In all
transformation from bcc to face-centered cubic other cases, higher tempering temperatures lead
(Fcc), which involves a more than 1% linear di- to lower hardness.
mension change and can cause distortion or The nickel-bearing alloys have a restricted
cracking. Oxidation during austenitizing can upper tempering temperature because of the
cause serious carbon loss on the surface, which danger of re-forming austenite, which would
will result in serious loss of surface hardness. then transform to untempered martensite during
Heating 410 in air for 10 min at 1100 °C (2010 cooling, requiring a second tempering operation.
°F) can cause surface carbon to decrease by Intermediate temperatures can lead to the
one-half, lowering hardness from HRC 45 to phenomenon of temper embrittlement, which is
under 20. caused by the precipitation of phosphorus and
Quenching rate is not a significant issue for other species, such as, but not limited to, car-
the martensitic stainless steels since they have bides, at prior austenitic grain boundaries. This
such high hardenability, but some, especially phenomenon is distinct from the precipitation of
those with higher carbon levels, may have re- α', which causes the so-called 475 °C embrittle-
tained austenite, which can lower hardness and ment, which occurs more severely in alloys
cause problems with dimensional stability. The with higher chromium levels. Because the for-
quenching rate must be sufficient, however, to mation of martensite is diffusionless, the
avoid precipitation of carbides in the austenite austenite boundaries maintain the microcompo-
during quenching since the sensitization would sitions they have at high temperatures. Auste-
persist in the final microstructure. If this oc- nite has low solubility for impurities such as
curs, a subzero treatment at below –75 °C phosphorous, so phosphorous is more highly
(–100 °F) should be undertaken immediately concentrated in the grain boundary regions.
after quenching. This compositional inhomogeneity can be elim-
inated by higher tempering temperatures or by
the addition of molybdenum, which combines
Tempering with the phosphorus and prevents the embrittle-
Untempered martensite has insufficient tough- ment. The existence of temper embrittlement is
ness to be a useful engineering material. During the reason that Table 3 recommends avoiding
tempering, carbon is precipitated from the super- certain temperature ranges for tempering.
saturated state it is in when it is quenched into Lower tempering temperatures and higher
the bcc martensite structure during the diffusion- austenitizing temperatures are best for corrosion
less transformation. The strain energy associated resistance because both minimize the amount of
with the lattice strain of the bcc martensite chromium tied up as carbide. Quenching from
caused by the poor fit of the carbon in the tetrag- higher temperatures also enhances toughness.
onal interstices is very large. Heating to even
low temperatures allows carbon enough mobil-
Stress Relieving
ity to diffuse and precipitate as carbide. Since
carbon diffuses 106 times as fast as iron, If quenched martensitic steels are not imme-
chromium, or other carbide formers, it tends to diately tempered, then they should be promptly
stress relieved. Otherwise, the residual stress eliminating retained austenite, enhances dimen-
from quenching could result in stress corrosion sional stability but diminishes toughness. Table 4
cracking even in seemingly benign environ- lists the solution treatments for all PH grades.
ments. Stress relieving is simply a low-tempera- The as-quenched state is called condition A.
ture tempering operation, but some elimination This is the normal condition in which the mate-
of residual stress does occur. Higher tempera- rial is supplied from the mill and is intended to
tures and longer times produce greater stress be soft enough for machining and some forming.
relief and maximize elastic properties, but opti- If softer material is required, the H-1150M con-
mal toughness is obtained at intermediate times dition exists in which the material is first highly
and temperatures. overaged at 760 °C (1400 °F), allowing some
austenite to re-form. The subsequent aging then
overages that martensite while retaining some
Precipitation Hardening Stainless Steels stable austenite. The result is a very tough
microstructure.
There are three classes of PH steels, and each Aging. The time and temperatures required
requires totally different heat treatment. The to produce this precipitation are also given in
classes are martensitic, semiaustenitic, and Table 4. The condition code itself tells the tem-
austenitic. The most straightforward alloys are perature at which the aging is conducted in
the martensitic PH grades. Like the plain marten- that the code numbers are based on thermal
sitic alloys, the martensitic PH alloys are care- processing temperatures expressed in degrees
fully designed to produce a nearly fully austenitic Fahrenheit (oF), for example, TH 900 means
structure at high temperature that quenches to a transformed to martensite (T) and aged at 900
nearly fully martensitic structure on cooling. The degrees Fahrenheit. The final properties are a
martensite is low in carbon, so it is relatively soft function of both aging time and temperature.
and not prone to brittleness. So, the hardness and Lower temperatures result in higher possible
strength of these alloys is derived from a subse- hardness but lower toughness. The precipi-
quent tempering-type heat treatment during tates, as mentioned, are optically invisible and
which various constituent elements form ex- cause very little dimensional change. Contrac-
tremely fine coherent precipitates that greatly tion on the order of 0.0005 in./in. from aging is
strain and therefore strengthen the matrix. There typical, often permitting machining to final di-
are numerous precipitates that can provoke this mensions in condition A. All aging treatments
effect, and they are described in detail in the PH are above the temper embrittlement range to
chapter (see Chapter 10). All require the short- which these alloys are susceptible. Service
range diffusion of substitution elements to form temperatures in this range would result in em-
these optically invisible precipitates. brittlement, so use above 350 °C (660 °F)
should be avoided. Molybdenum-bearing
grades should be selected to minimize this
Martensitic PH Grades phenomenon if high-temperature use is con-
Solution treatment of these alloys is con- templated.
ducted to achieve a fully austenitic structure. Solution Heat Treatment and Conditioning.
The constituent elements are easily dissolved, The semiaustenitic grades are more complicated
so excessive temperature or time is unnecessary than the martensitic PH alloys. These alloys are
and could be counterproductive if it were to re- designed so that they are austenitic when
sult in ferrite formation or surface oxidation, quenched from the solution heat treatment tem-
which would be detrimental to final mechanical perature. This is also called condition A, and it
properties. The presence of retained ferrite is permits them to be highly formable. This stabi-
mainly a function of alloy and composition lization of the austenite comes mainly from
within the allowed range. Earlier grades such as higher chromium and carbon levels. These al-
17–4 and the obsolete stainless W intrinsically loys essentially resemble a lean 301 austenitic
contained some ferrite. The subsequent alloys alloy, many with some molybdenum substitut-
are substantially ferrite free. Most alloys may ing for part of the chromium. The key to these
retain some austenite after quenching to room grades is making them behave as a martensitic
temperature, in which case subzero treatment alloy. This is done by precipitating some of the
should be done within 24 h to avoid further sta- carbon as chromium carbide at a temperature at
bilizing the austenite. Subzero treatment, by the high end of what would normally be consid-
Table 4 Recommended annealing and stress-relieving temperatures for martensitic PH grades

Alloy Condition code Solution anneal Conditioning Aging
Martensitics
13-8 A 925 °C 15 min, oil or air cool below 15 °C –75 °C 8 h
RHxxx 925 °C 15 min, oil or air cool below 15 °C –75 °C 8 h xxx °F for 4 h
Hxxx 925 °C 15 min, oil or air cool below 15 °C ... xxx °F for 4 h
15-5 A 1035 °C 30 min, oil or air cool below 15 °C ...
17-4 A 1035 °C 30 min, oil or air cool below 15 °C ...
450 A 1035 °C for 1 h, water quench ...
Hxxx 1035 °C for 1 h, water quench ... xxx °F for 4 h
455 A 830 °C for 1 h, water quench ...
Hxxx 830 °C for 1 h, water quench ... xxx °F for 4 h
465, 275 A 980 °C for 1 h, cool rapidly –75 °C 8 h
Hxxx 980 °C for 1 h, cool rapidly –75 °C 8 h xxx °F for 4 h
475 A 925 °C 1 h, air cool –75 °C 8 h
Hxxx 925 °C 1 h, air cool –75 °C 8 h xxx °F for 4 h
Semiaustenitics
17-7, 15-7 A 1065 °C for 30 min, air cool ... ...
T 1065 °C for 30 min, air cool 760 °C (1400 °F), 90 min, ...
air cool to RT for 30 min.
C 1065 °C for 30 min, air cool Cold reduce ...
R 1065 °C for 30 min, air cool 955 °C (1750 °F) 10 min, air ...
cool, chill to –75 °C for 8 h
THxxx 1065 °C for 30 min, air cool 760 °C (1400 °F) xxx °F for 4 h
90 min, air cool to RT for 30 min
CHxxx 1065 °C for 30 min, air cool Cold reduce xxx °F for 4 h
RHxxx 1065 °C for 30 min, air cool 955 °C (1750 °F) 10 min, air xxx °F for 4 h
cool, chill to –75 °C for 8 h
AM-350 A 1010–1065 °C ...
L (equivalent to T) 1010–1065 °C 930 °C for 90 min, air cool ...
SC (equivalent to R) 1010–1065 °C 930 °C for 90 min, air cool, ...
180 min at –75 °C
SCTxxx 850 °F or 1010–1065 °C 930 °C for 90 min, air cool, xxx °F for 180 min
1000 °F 180 min at –75 °C
DA (double aged) ... 930 °C for 90 min, air cool, 450–540 °C 180 min
730–760 °C 180 min
Am-355 A 1025–1040 °C ... ...
L (equivalent to T) 930 °C for 90 min, air cool ...
SC (equivalent to R) 930 °C for 90 min, air cool. ...
180 min at –75 °C
SCTxxx 850 °F or 930 °C for 90 min, air cool, xxx °F for 180 min
1000 °F 180 min at –75 °C
DA (double aged) 930 °C for 90 min, air cool. 440–470 °C
730–760 °C 180 min for 180 min
Equalized and 930 °C for 90 min, 540–590 °C for
overtempered air cool, 730–760 °C 180 min 180 min
Austenitic
A-286 ST1650 900 °C 120 min, oil/water quench ... ...
ST1650A 900 °C 120 min, oil/water quench 730 °C 16 h
ST1650DA 900 °C 120 min, oil/water quench 730 °C 16 h,
650 °C 8 h
ST1800 980 °C 120 min, oil/water quench
ST1800A 980 °C 120 min, oil/water quench 730 °C 16 h
PH, precipitation-hardenable; RT, room temperature.
ered sensitization, 760 °C (1400 °F). The car- finishes (Mf) near room temperature. This
bide precipitation occurs at the interfaces of the process is called austenite conditioning. The
small amount of residual ferrite these alloys heat treatment scheme just described would be
have and also at grain boundaries. The deletion condition A-1400. This material, after quench-
of carbon and chromium from the matrix ing to room temperature, would be said to be
changes the matrix composition sufficiently that condition T.
the temperature for the start of the martensitic The higher the temperature of the condition-
transformation (Ms) of the depleted austenite in- ing, the less carbon is precipitated and the lower
creases from below zero to about 65 to 100 °C the resulting Ms. The highest conditioning tem-
(18 to 212 °F). The martensitic transformation peratures of 955 °C (1750 °F) cause a sufficiently
low Ms that subzero treatment is required to for a normal solution anneal of non-PH austen-
obtain the fully martensitic structure required for ite. The subsequent precipitation aging requires
age hardening. This would be called A-1750. higher temperatures and longer times because
After subzero treatment at –73 °C (–100 °F), it diffusion is much slower in austenite.
would be called condition R-100.
It is even possible to obtain the fully marten-
sitic structure by cold work from full condition Duplex Stainless Steels
A, which is quenched from 1065 °C (1950 °F).
This is called condition C and requires heavy Duplex stainless steels are both ferritic and
cold rolling to accomplish. Since there is no heat austenitic, so their heat treatment combines the
treatment to precipitate carbon from the austen- same elements and principles as their principle
ite matrix, the resulting martensite is the hardest. phases.
The aging treatments of the semiaustenitic
alloys are identical to those for the martensitic
Soaking
alloys because the treatments are standardized.
The resulting mechanical properties vary in a Duplex alloys are multiphase at all useful
complex fashion with the alloy composition and working temperatures, making their hot worka-
thermomechanical treatment history before bility quite poor. It is extremely important to
aging. The principles at work that determine the drive sulfur to the lowest possible levels, less
mechanical properties, besides the strain in- than 0.001%, to achieve satisfactory hot ductil-
duced by the precipitate phase, are: ity. Otherwise, soaking should proceed the same
• Carbon level of the quenched martensite as for ferritic stainless alloys since ferrite con-
• Amount of cold work of austenite prior to stitutes the continuous phase to be worked.
Because these alloys always contain high
aging chromium and generally high molybdenum,
These factors do not come into play with the they should be cooled as rapidly as possible
martensitic PH grades because all of the carbon from high temperatures to avoid formation of
is in solution, and they are not significantly cold sigma or other intermetallic phases.
worked prior to aging. The semiaustenitic PH
alloys, because of the variable amount of car- Annealing
bide precipitation, have an equally variable
chromium content, so corrosion resistance will The function of annealing in the duplex alloys
vary accordingly. The highest-temperature solu- is generally to:
tion and carbide precipitation treatments pro- • Remove the effects of cold work
vide the best corrosion resistance, as do the • Restore the balance between the volume
lowest aging temperatures. fraction of ferrite and austenite
• Achieve equilibrium composition within
Austenitic PH Alloys both the austenite and ferrite
• Dissolve unwanted intermetallic phases
In contrast to the martensitic and semi-
austenitic PH alloys, the austenitic PH alloys The annealing range of duplex alloys is some-
are austenitic under all conditions. Their what restricted, approximating the overlap be-
strengthening reaction comes from precipitation tween what each of the two constituent phases
of titanium- and aluminum-nickel intermetallic would be annealed at separately. Table 5 lists
phases within the austenite matrix as occurs in the normal annealing temperatures for these al-
superalloys. Solution treatment is very typical loys. The use of nitrogen as a key alloying
Table 5 Recommended annealing and stress-relieving temperatures for duplex alloys

Annealing Annealing
Alloy temperature, °C temperature, °F ASTM A480 2006 Stress relieving, °F Stress relieving, °C
Lean duplex, Cr+Mo<23, 1010–1100 1850–2010 Various Not recommended
2003-2101, 19-D-2304
Medium alloy, Cr+Mo<26, 1040–1100 1900–2010 1040 min Not recommended
2205
Cr+Mo>26, 2507, 52N+, 1050–1150 1925–2100 Various Not recommended
Zeron 100, 255
element has improved the annealing behavior of tions, either from ␣' beginning at 350 °C (660
these alloys since its diffusion is quite rapid, °F) or from ␣/␹, which takes over at 600 °C
causing ferrite-austenite equilibrium to be at- (1110 °F), would occur and so is not indicated
tained very rapidly. Nitrogen also hinders (for- for these alloys.
mation and facilitates the dissolution of second-
ary austenite, which can form after quenching
from welding temperatures and cause regions of REFERENCES
poor corrosion resistance.
These alloys are not very susceptible to car- 1. A. Garner, The Effects of Autogenous
bide sensitization and normally have very low Welding on Pitting Corrosion in Austenitic
carbon content. Thus, the guiding principle in Stainless Steel, Corrosion, Vol 35 (No. 3),
annealing is simply to achieve phase balance 1979, p. 108–114
and avoid cooling so slowly that intermetallic 2. J.F. Grubb, personal communication, June
phases may form. 4, 2006
The strengthening of duplex is normally 3. Data courtesy J&L Specialty Steel, October
achieved by the strong grain refinement and 2, 1998
solid solution hardening. No strengthening heat 4. J.F. Grubb, Proceedings of the International
treatments are used. Stress relief would have to Conference on Stainless Steels, 1991,
occur at temperatures at which embrittling reac- Chiba, ISIJ
CHAPTER 14
Forming
Summary shown that it is important to understand and

exploit these characteristics to optimize forming
STAINLESS STEELS RANGE in formability of stainless steels.
from the extremely formable austenitic alloys to
the hard-to-form martensitic alloys. For steels
with equivalent corrosion resistances, formabil- Flat, Rolled Stainless Steel
ity increases with the level of austenitizing al-
loying elements. The ferritic alloys are the least The vast majority of carbon steel and, espe-
alloyed, least expensive, and least formable; cially, stainless products are flat products. These
duplex steels are next, and austenitic steels are steels are formed by bending, roll forming, spin-
the most formable but most expensive. How- ning, hydroforming, and deep drawing. Roll
ever, if the specific structural anisotropy of forming is most commonly used to produce
ferritic alloys, which gives particularly good welded pipe and tubing and is simply bending
deep-drawing characteristics for a given level of done in a continuous manner. Bending is a sim-
ductility, can be exploited, the best forming ple operation, and there is no meaningful change
economies can be gained. in thickness of the sheet during the operation.
The higher yield strength and work-hardening
rates of most stainless steels will result in greater
Introduction springback than would be experienced in carbon
steel. Tooling must be adjusted to compensate
The technology for forming stainless steel is for this. In neither of these forming methods is
quite similar to that for forming carbon steels. there large motion of the formed material across
The primary difference is the higher strength of the die surface, so lubricant is not normally
all types of stainless steels compared to drawing- used. The reader is referred to the Forming and
type carbon steels. This higher strength requires Forging in Volume 14 of the ASM Handbook for
greater sophistication in tooling and lubricants, detailed charts on bend radii and springback re-
and it requires more powerful forming equip- lated to bending.
ment. The higher initial strength is also Other forming techniques employ more com-
accompanied by a higher work hardening rate in plex deformation processes. Deep drawing is
austenitic stainless steels, which further distin- the foremost of these. Figure 1 (Ref 1) shows
guishes them from carbon steel. Galling also schematically what occurs during drawing. A
must be recognized as a danger and prevented. round blank is held between dies over a cavity,
Stainless steel has lower thermal conductivity and a punch pushes the material into the cavity
than carbon steels, which can cause it to retain to produce the part. If the dies pinch the blank
heat from deformation and friction, thereby to only a small degree, the process is normal, or
decreasing lubricity. Last, stabilized stainless al- ordinary, drawing. If the dies significantly
loys contain abrasive carbide microconstituents. restrain the periphery of the blank so it cannot
The various types of stainless steel have very move, stretch forming occurs. Material proper-
different deformation characteristics in terms of ties determine whether a material is most suc-
strain hardening and anisotropy. It will be cessfully formed by stretch forming or drawing.
rate. If it work hardens faster than it becomes

thinner, the strain is distributed, and local fail-
ure is prevented. Austenitic materials have the
face-centered cubic (fcc) crystal structure with
many slip systems and low stacking fault ener-
gies. This means that they can generate many
complex arrays of tangled dislocations, which
cause strain hardening. They can also transform
during stretching to the much harder martensite,
for which the deformation is greatest, again
redistributing the deformation away from the
potentially thinning area. This makes austenitic
stainless steels particularly suitable for stretch
forming.
Deep Drawing. Deep drawing, referred to
simply as drawing, without stretching requires a
different material characteristic. For drawing, a
low work-hardening rate is desirable so that the
material can compress in the circumferential
direction while elongating in the radial direction.
Obviously, a high ability to elongate is always
useful regardless of any other characteristic.
But, ferritic material has one other advantage:
Body-centered cubic (bcc) alloys have more
slip systems than fcc alloys. When bcc alloys
are rolled to become flat stock, they may retain
a preferred crystallographic orientation, called
texture, as a result of the deformation. This non-
random crystal structure can cause the material
to have higher strength in the through-thickness
direction. This directional variation in proper-
ties is called anisotropy.
When a material with desirable texture is
stretched, it flows in the stretching direction and
contracts laterally at lower stresses than are
required to initiate plastic flow in the through-
thickness direction. As long as the work-hard-
ening rate keeps the flow stress below the
through-thickness yield strength, there will be
no thinning. The geometry of deep drawing
with constraint fits such materials’ capabilities.
This is why carbon steels and ferritic stainless
steels deep draw well. If the material were con-
strained from contracting while being stretched,
the tensile strength would be exceeded before
through-thickness flow occurred and the mate-
rial would fracture with little deformation. So,
Fig. 1 Deep drawing schematic. Source: Ref 1 these materials cannot be stretch formed. Even
without stretching, some hold-down pressure is
Stretch Forming. In stretch forming, the required to prevent wrinkling of the blank
material is constrained from moving wholly before it is pulled into the die. This is more pro-
into the die. Thus, the section that enters the die nounced for thinner blanks and for material
is stretched more and must become thinner. with higher work-hardening rates. In many
Whether the material becomes so thin locally forming situations, adjustments to the drawing
that it fails is governed by its work-hardening process (i.e., hold-down force adjustment, draw
Chapter 14: Forming / 175
bead contour, blank size, lubrication, die radii, and 2.0 are more common. Figure 2 shows the
etc.) may be more important than material prop- LDR as a function of the Lankford ratio. It
erties in determining whether the desired part should noted that the very ductile 304 fares no
can be made successfully. better than carbon steel because of the advantage
The material properties that are important to of a beneficial anisotropy, which the ferrous bcc
formability are ductility, as measured by tensile structure has. Indeed, the best-performing deep-
test elongation; the work hardening rate, which drawing stainless steels are low interstitial ferritic
is the instantaneous slope of the true stress, true steels with boron added. A number of stainless
strain curve and is called n; and the anisotropy. steels are compared in Table 1 (Ref 4).
The measure of anisotropy is the Lankford ratio It shows that when comparing different types
(Ref 2), expressed as: of materials, some tests are not good predictors
of deep-drawing performance. Optimizing the
material/drawn component combination is far
r0 + 2r45 + r90 from simple, especially when other considera-
R= (Eq 1)
4 tions, such as cost and material performance in
service must be factored in. End users are
where R is the average strain ratio, r0 is the strain encouraged to deal directly with the producing
ratio in the longitudinal direction, r45 is the strain mill early in the design stage of any new high-
ratio measured at 45° to the rolling direction (of production, deep-drawn component. The pro-
the sheet metal-forming operation), and r90 is the ducing mills, while not necessarily exhaustive
strain ratio in the transverse direction. R deter- sources of information, are certainly reservoirs
mines the average depth (that is, the wall height) of knowledge of current practice.
of the deepest draw possible. When this expres- The most widely used summary of a mater-
sion equals 1, then a material may be considered ial’s formability is contained in its forming limit
isotropic, that is, the material properties are the diagram (FLD), developed by Keeler and Back-
same for all crystallographic orientations. As a ofen (Ref 5). This diagram shows the locus of
first approximation, the Lankford ratio equals failure under varying strain states. Figure 3
the ratio of the lateral strain to the through-thick- shows a comparison of the FLDs for austenitic
ness strain during the tensile test of a sheet spec- stainless steel and carbon steel. The FLD tells
imen. As the value increases from 1, the drawa-
bility increases because the material tends to
maintain a constant thickness while changing
shape from a flat blank to a cup shape. The abil-
ity to be deep drawn is measured by the limiting
drawing ratio (LDR), the ratio of the diameter of
a disc to that of the deepest cylinder into which
it can be drawn.
The ferritic stainless steels in sheet form have
LDRs of around 2.2 compared to 2.0 for 304. For
austenitic steel, the ratio is about 1.0, while for
flat-rolled carbon steel and ferritic stainless steel,
Fig. 2 Limiting drawing ratio variations with Lankford ratio.
it can be greater than 2.0, but values between 1.5 Source: Ref 3
Table 1 Deep-drawing materials comparison

0.2% proof Tensile Erichsen Conical cup
Lankford value
stress, strength, Elongation, Hardness value, value,
Steel N/mm1 N/mm2 % HV n value ro r45 r90 r̄ mm mm
YUS 190 343 497 33.8 173 0.20 1.60 1.47 2.10 1.66 9.5 26.7
YUS 436S 275 459 34.8 135 0.21 1.67 1.63 2.12 1.76 9.8 26.9
(B-added)
YUS 436S 284 483 34.5 137 0.22 1.49 1.90 2.01 1.83 9.8 26.8
(B-free)
YUS 4O9D 239 424 37.2 116 0.24 1.51 1.77 2.11 1.79 9.7 26.7
SUS 430 308 472 31.8 159 0.21 0.94 0.92 l.50 1.07 8.9 28.5
SUS304 281 705 64.0 172 0.44 0.91 1.19 0.83 1.03 12.5 27.0
Source: Ref 2
Fig. 4 Forming limit diagrams for categories of stainless

steels. A, austenitic stainless steel; F, ferritic stainless
Fig. 3 Forming limit diagram of carbon steel compared to steel; HAS, high-strength austenitic stainless steel; HSF, high-
austenitic stainless steel. Source: Ref 3 strength ferritic stainless steel; FA(50), ferritic-austenitic stainless
steel. Source: Ref 3
the point of failure for a given sheet material

with a given thermomechanical history over a
full range of combined strain states. These dia- total interstitial content (i.e. carbon plus nitro-
grams are generated by examining circle grids gen) and by thermomechanical working to give
printed on material that is deformed to failure in a fine-grained, fully recrystallized, yet benefi-
various modes. The single most important value cially anisotropic, microstructure. Figure 5
on the curve is the intersection of the curve with shows how the FLD of an enhanced 409 ferritic
the major strain axis at zero minor strain. This stainless steel, 409 Ultra Form, compares to the
can be used as an index of formability and is the already highly evolved 409.
elongation possible for plane strain conditions. For austenitic stainless steel, maximum
This value within a given class of materials is drawability is obtained by low work-hardening
proportional to the strain-hardening exponent rates coupled with maximum elongation, as
because a higher work-hardening rate causes exemplified by 305 and high-nickel 304. This
higher localized resistance to thinning, which is comes with a cost penalty as the easiest way to
the precursor of failure. improve formability is to increase the nickel
There is much more variety within the fami- level, although replacing the expensive nickel
lies of stainless steel than within carbon steel. with copper or manganese, as in 204Cu, has
Figure 4 shows generalized FLDs for austenitic, been shown to be effective. Many austenitic
ferritic, high-strength ferritic, high-strength components are made by stretched deep draw-
austenitic, and duplex stainless steels. ing. In this case, the preferable alloys are the
Deep drawing of components is seen as a way leaner austenitics, 201 and 301. These alloys
to obtain near-net shape. Since tooling is costly, it form martensite more rapidly than do 304 and
is necessarily a high-volume application. Quite 305 during stretching. Martensite has a 4%
often, designers push component design to the greater volume than the austenite from which it
limit of a material’s ability to be formed. There forms and a much greater strength. This gives
are various drivers that cause this. One is to elim- 201 and 301 the ability to redistribute deforma-
inate extra operations or components by consoli- tion from thinning areas elsewhere and stretch
dating them into one more complex deep-drawn extensively, making them an optimal material
part. Another is to use the least-expensive alloy. for objects such as sinks.
In some industries, such as the household appli- The specific alloy composition is often finely
ance industry, as many components as possible tuned for a given part and tooling design, and
are deep drawn from ferritic stainless alloys and small deviati ons can dramatically increase
the more costly austenitic alloys are used only breakage rates. Even such minor process
when the part cannot be made from a ferritic. changes as blank temperature variations due to
There has been much research to develop fer- ambient temperature can alter work-hardening
ritic stainless alloys with improved formability. rates enough to cause breakage problems. Cer-
This has been accomplished by reducing the tainly, this can occur when designs push the
Fig. 5 Optimized 409 for forming versus normal 409. Source: Ref 5
envelope of a material’s capability, but cost A last key variable in which practice and ma-
pressures generally drive designers to this terial interact is strain rate. Ferritic steels flow
extreme. more easily at lower strain rates and are thus
On very severe forming, intermediate more formable. Austenitic steels experience the
annealing may be required to either enhance opposite effect if they are susceptible to marten-
ductility or reduce required pressure. Stainless site formation. Adiabatic heating can retard the
steel in the as-drawn condition will have resid- martensitic transformation and reduce the work-
ual stress and may have sufficient hardness to hardening rate, changing their forming charac-
be susceptible to delayed failure if placed in a teristics. When drawing, this is good, but for
corrosive environment. Bright-annealed alloys stretching it may not be.
with high martensite levels from forming can Hydroforming, a variation on deep drawing
fail by hydrogen embrittlement with just the in which hydrostatic pressure forces a blank
residual hydrogen from annealing. Therefore, into the die cavity, can improve the degree to
the use of bright-annealed lean alloys such as which stainless steels can be deep drawn. The
301 is not recommended. hydroforming process avoids friction between
the blank and the tool. Deformation is spread forming difficulties and lack of compliance with
more evenly across the blank, and the material mechanical property requirements. They also
forms close to its theoretical best. Productivity have significant rolling anisotropy, which causes
using this technique is relatively low, so its use the yield strength transverse to the rolling direc-
is justified mainly when it is not otherwise pos- tion to be consistently higher than it is in the
sible to make a certain design in one drawn rolling direction. This behavior is contrary to the
component. An example of this may be the pro- general behavior of single-phase alloys. The dif-
duction of a complex exhaust manifold that reference in yield strength is sufficient, reportedly
quires a higher-alloyed ferritic stainless with up to 15% lower in the longitudinal direction
relatively low formability. than in the transverse direction required for ten-
Besides failure by breakage, there are other sile tests (Ref 7), that it is both serious design
less-severe defects found on deeply drawn and forming considerations.
parts. Austenitic steels can develop a surface Ferritics undergo a more specialized surface
condition known as orange peel, the result of relief because their anisotropic grain structure
slip planes within a grain disrupting the sur- can yield in a more concerted fashion and give
face. Orange peel is prevented by keeping even greater surface relief, called ridging and
grain size fine so that the surface relief is too roping. This is minimized by refining grain
small to be seen. Austenitic stainless steels can size, achieving full recrystallization versus just
also develop anisotropy, which while less recovery during annealing, and temper passing
severe than ferritic steels, can cause “earing,” (i.e., elongation of about 1% on a cold-rolling
in which round blanks deform nonuniformly in mill) to suppress the yield point phenomenon
a four-, six-, or eight-fold symmetry, causing that is characteristic of ferrous bcc materials.
excess ear-shaped material to extend beyond End users should always make their use of the
the intended dimensions of the component. material known to the producing mill so that
Material is wasted because larger blanks have the correct thermal processing path can be
to be used. The steel producer can minimize employed for the manufacturing process the
the phenomenon by keeping cold roll reduc- material will undergo.
tions above about 60%. One measure of the One of the most important material consider-
earing tendency is derived from the Lankford r ations for deep drawing is surface finish. Flat-
measurements: rolled stainless should be fully annealed and
pickled so that the surface holds lubricant well
and yields as readily as possible. Temper pass-
r0 + r90 + 2 r45 ing will drastically reduce the drawability of
Δr = (Eq 2)
2 stainless by smoothing the surface (rather than
increasing the yield strength). Temper passing
The left side of Eq 2, ⌬r, is a measure of the with roughened rolls does not significantly
variation of plastic strain ratio r with direction harm drawability. The surface finishes that are
in the plane of a sheet. Values of ⌬r near zero produced by temper rolls with special finishes,
generally indicate minimal tendency toward such as Koolline, retain lubricant well and can
earing, while ⌬ values significantly above or be drawn with minimal distortion.
below zero indicate increased tendency toward Tooling for stainless must be strong and wear
earing. A combination of a high R value from resistant. Traditional tooling materials are D2
Eq 1 and a low ⌬r value provides optimal tool steel and high-strength aluminum bronze.
drawability. D2 tool steel must be hardened to HRC 60 to
It should be noted that deformation always 62 and must have smooth surfaces. The use of
produces some surface relief, so highly reflec- powder metal techniques to produce tool steels
tive surfaces become spectrally diffuse, or for dies has permitted much higher volume frac-
cloudy, after plastic deformation. This has been tions of ultrahard microconstituents such as
an issue for items such as automotive bright vanadium carbide to be introduced, thereby
trim. Mechanical buffing can restore the luster, vastly improving wear resistance without harm-
but the time and expense of buffing increase ing toughness or even raising overall hardness.
dramatically if orange peel or roping (a similar The benefits of cast aluminum bronzes are low
surface defect) is excessive. friction, high thermal conductivity, and low ten-
Duplex stainless steel flat products exhibit dency to gall. They are preferred when finished
significant in-plane anisotropy that can cause part surface appearance is more important than
Table 2 Suitability of various lubricants for use in forming of stainless steel

Blanking Press- Drop Contour
and brake Press Multiple-slide Deep hammer roll
Lubricant piercing forming forming forming drawing Spinning forming forming Embossing
Fatty oils and C B C A C A C B B
blends(a)
Soap-fat NR NR C A B B C B C
pastes(b)
Wax-base B B B A B B C B A
pastes(b)
Heavy-duty B NR B A B B NR A B
emulsions(c)
Dry film (wax B B B NR B A B NR A
or soap
plus borax)
Pigmented B NR A B A C NR NR NR
pastes(b)(d)
Sulfurized or A A B+ A C NR A B A
sulfochlorinated
oils(e)
Chlorinated oils or A(h) NR A NR A NR A(i) A NR
waxes(f) high-
viscosity types(g)
Chlorinated oils or B+ A A A B NR A(i) A A
waxes(f) low-
viscosity types(j)
Graphite or NR (l) (l) NR (l) NR (l) NR NR
molybdenum
disulfide(k)
A, excellent; B, good; C, acceptable; NR, not recommended; SUS, Say-bolt universal second. Ratings consider effectiveness, cleanliness, ease of removal, and other
suitability factors. (a) Vegetable or animal types; mineral oil is used for blending. (b) May be diluted with water. (c) Water emulsions of soluble oils; contain a high con-
centration of extreme pressure (EP) sulfur or chlorine compounds. (d) Chalk (whiting) is most common pigment; others sometimes used. (e) EP types; may contain
some mineral or fatty oil. (f) EP chlorinated mineral oils or waxes; may contain emulsifiers for ease of removal in water-base cleaners. (g) Viscosity of 4,000 to 20,000
SUS. (h) For heavy plate. (i) For cold forming only. (j) Viscosity (200 to 1000 SUS) is influenced by base oil or wax, degree of chlorination, and additions of mineral oil.
(k) Solid lubricant applied from dispersions in oil, solvent, or water. (l) For hot forming applications only.
absolute die life and forming pressures are medium-sized hardware items, such as screws,
moderate. bolts, nuts, rivets, and specialized fasteners.
Lubricants for stainless steel forming must be As with flat-rolled forming operations, the
able to prevent metal-to-metal contact under primary difference between carbon/alloy steels
higher pressures than those seen with carbon and stainless steels comes from the higher yield
steel. The ASM Metals Handbook, Desk Edi- strength and higher work-hardening rates of
tion, lists common lubricants as shown in Table stainless. Anisotropy is not a significant consid-
2 (Ref 1). Not listed in the table are the newer eration for long products. The ferritic stainless
thermoplastic acrylic polymers that, when ap- steels are the most easily cold headed. The use
plied to the surface at a density of around 1 g/m2 of the most formable stainless alloys for flat
(0.004 oz/ft2), provide a dry film with lubricat- products, the stabilized ferritic alloys, is limited
ing properties surpassing any of those listed in for long products because of the severe lack of
Table 2. toughness these alloys show for cross sections
greater than about 2 mm (0.08 in.). But, the
nonstabilized ferritics, the martensitic, precipi-
tation-hardenable (PH), austenitic, and duplex
Stainless Long Products grades are all cold formable.
In cold-heading terminology, the maximum
Cold heading, one of the most important possible deformation an alloy can tolerate is ex-
forming operations conducted on stainless long pressed in terms of the length of long product
products, is a forming process that increases the exposed beyond the die that can be successfully
cross-sectional area of a room temperature forged into the upset. This is measured in the
blank at one or more points along its length. number of diameters of initial stock. So, an op-
Cold heading is typically a high-speed process timal ferritic such as 430 can tolerate upsets up
in which the blank is progressively moved to about 2.25 diameters, while a very low work-
through a multistation machine. The process is hardening austenitic, such as 384, can tolerate
widely used to produce a variety of small- and 3.0. The martensitic, PH, and richer duplex
their work hardening is not instantly removed

by annealing. For example, the initial pressure
requirements for a given deformation range for
304 can be three times what is required to de-
form carbon steel and up to five times for higher
molybdenum alloys. For greater deformation,
this multiple increases (see Fig. 6a and b).
Austenitic stainless steel loses ductility when
heated above 1280 °C (2335 (F) because of low
melting phases in the grain boundaries. As
much as possible, all alloys should be hot
worked in a single phase field of the phase dia-
gram to avoid mixtures of ferrite and austenite
since the great difference in their great strengths
can cause failures. Duplex stainless steels and
other alloys (e.g., 17-7 PH) that have high lev-
els of ferrite in austenite or austenite in ferrite at
the hot-working temperature exhibit reduced
hot ductility compared to either fully austenitic
or fully ferritic stainless steels and are more dif-
ficult to hot work.
REFERENCES
1. Forming of Sheet, Strip, and Plate: Deep

Drawing, in Metals Handbook, desk ed.,
J.R. Davis, Ed., ASM International, 1998, p
782–829
2. U.F. Kocks, C. Tomé, H.-R. Wenk, Ed., Tex-
ture and Anisotropy, Cambridge University
Press, Cambridge, UK, 1998
3. E. Schedin, “Forming Stainless Steel,”
Fig. 6 Forces required for hot working. Source: Ref 1 ACOM Technical Paper, www.outokumpu
.com
4. H. Sumitomo and T. Tanoue, Nippon Steel
alloys are in the 1.5 to 2.0 range of formability. Technical Report 71, October 1996
The lean duplex, when they find their way into 5. S.P. Keeler and W.A.A. Backofen, ASM
wider use, should resemble 430 with 2.25 diam- Trans. Q, Vol. 56 (No. 163), 1963, p 25–48
eters maximum. 6. “409 Ultra Form Stainless Steel,” Product
Hot forming of stainless steel is done as an in- Data Bulletin, www.AKSteel.com
tegral part of their production; therefore, all 7. R. Cordewener et al., “Duplex Stainless
stainless alloys can be forged. The main issue is Steels,” Paper 109, TWI conference, Glas-
that the high hot strength of stainless requires gow, 1994
much more force than would be required for
carbon steels. Martensitic stainless steels re- SELECTED REFERENCE
quire 10 to 100% more force than 4340 alloy
steel, while the austenitics require much more • ASM Handbook, Vol. 14, Forming and
because of their high hot hardness and because Forging, ASM International, 1988
CHAPTER 15
Machining
Summary From a more focused viewpoint, the machin-

ability of a material is further described by:
MACHINING STAINLESS STEELS is a
complex operation. Not only does a shop need 1. Consistency: Does the material machinabil-
the correct supporting equipment and supplies, ity stay the same when bundles are changed?
a better understanding of the metal itself is ad- 2. Tool life/wear: How long does the tool last
vantageous. Technology in the production of a in the machining operation? This could be
more machinable stainless steel is advancing. minutes, hours, shifts, or days.
The incorporation of complex oxides has led to 3. Productivity: How many parts were made
the development of materials that allow higher in an hour, shift, or day?
machining speeds and increased productivities, 4. Cost per part: What is the cost of the final
both of which are reducing machining costs and geometry?
keeping shops competitive. 5. Cycle time: How fast can a part be com-
pleted?
6. Surface finish: How smooth or shiny is the
part?
Introduction 7. Chip control: Are the chips manageable?
8. Maximum cutting speed: How fast can the
Stainless steel forgings, castings, plate, and part be cut without affecting tool life?
long products all are frequently machined. This 9. Maintaining tolerances: How long can the
fundamentally involves the removal of a layer machining operation continue before ad-
of material from the workpiece with a cutting justments are made?
tool one or multiple times until a finished or 10. Minimal operator intervention: Does the
semifinished part is produced. Machining, in it- operator need to constantly adjust setup?
self, is a complex topic with many variables.
Rather than attempt to understand all aspects of
machining, it is helpful to consider a material’s
machinability, that is, its ability to be machined
and the factors that affect its ability to be ma-
chined. In Fig. 1, a macroview shows how the
machinability of a material is influenced by the
interaction of humans, machine, methods, mate-
rial, and management. Some of the variables
can affect the appearance of the material, while
others affect the performance of the piece, mak-
ing machining an art as well as a science. Opti-
mum machinability is obtained when each of
these sectors come together, providing the best
possible conditions for efficient machining. Any
change in one of these sectors can change the
behavior or efficiency of a machining job. Fig. 1 The 5 M’s of machinability
This list is somewhat empirical or job re- ƒ(chemistry, cleanlinesss,

lated, but it provides guidelines for defining a Machinability of
job since cutting conditions can be very differ- stainless steels = structure, processing,
cross section)
ent for each material and part. For example, if
the surface finish of a part is very important, it
may be necessary for chip control and tool life Each variable contributes uniquely to machin-
to be sacrificed. Clearly, machining involves ability. Machine shops and users generally have
much more than simply cutting a piece of very little influence on these material variables.
metal. Because no two mills are exactly identical, there
Machining is a very empirically mature sub- will be differences in machinability of a steel
ject. The recommended feed rate, depth of cut, grade provided by different mill suppliers. How-
tool material, and cutting fluid for a given mate- ever, having an understanding of how these vari-
rial/material condition (thermomechanical his- ables contribute to machinability is invaluable.
tory) can be found in readily available pub- Armed with an understanding of the material and
lished tables. Books such as the ASM how it is made, one can determine the tooling,
Handbooks; Machinery’s Handbook, published coolant, and setup of the machining job.
by Industrial Press; Marks’ Standard Hand- Let us take a closer look at these variables.
book for Mechanical Engineers, published by
McGraw-Hill Book Company; or the Machin-
Chemistry
ing Data Handbook, 3rd edition, by the
Machinability Data center at the Institute of Ad- The role of chemistry is to define not only
vanced Manufacturing Services (IAMS; for- the different grades of stainless steel (ferritic,
merly known as Metcut Research Associates martensitic, etc.), but also how the grade
Inc.) in Cincinnati, OH; include much of the is chemically balanced within the specific
data used in industry today. Material manufac- grade; for example, the amount of carbon in a
turers are also a source of valuable machining martensitic stainless can change tool wear char-
data. A typical guide from ASM is shown as acteristics, or a change in nickel content within
Table 1 (Ref 1). specification limits can alter the stringiness of a
Rather than simply reproducing data, the chip. Combined, both will be the basis the mate-
focus of this chapter is the metallurgical factors rial’s machinability.
governing the machinability of stainless steels. Each of the elements used to produce stain-
Most of the information regards machining less steels will contribute some general machin-
stainless bar products; however, many of the ing attributes. The effects of the elements as de-
concepts could be applied to forgings as well as scribed next are general, and slight deviations
castings. may be encountered depending on the stainless
grade. However, for the more common stainless
grades used today, these effects of these alloy-
ing elements are fairly accurate.
Physical and Mechanical Properties Iron is the base element in a stainless steel. It
is a soft, gummy material that has high work-
The machinability of stainless steels is very hardening characteristics. Iron is characterized
difficult to characterize in definitive terms be- by surface finishes that are difficult to obtain
cause of the broad nature of these materials. A and chips that are stringy, and it has a high ten-
ferritic stainless steel, such as type 430, will dency toward tool built-up edge (BUE).
machine very differently from the martensitic. Chromium strengthens and reduces ductility
In some sense, this is like comparing brass to of stainless steel. Machine and tool setup re-
carbon steel. Both type 410 and type 430 are quire more rigidity. Chromium allows chips to
stainless steels, but the chemistry and structural begin breaking.
differences create diversity in machining char- Carbon content increases strengthen stainless
acteristics. steels and promote carbide formation. Low car-
The machinability of stainless steels can be bon levels, typical in ferritic stainless steels, do
thought of as a function of the steel’s chemistry, not help machinability much. Increasing
cleanliness, structure, processing history, and amounts of carbon to greater than 0.08% will aid
the cross-section size of the stock, with no one in chip breakability and reduced BUE in these
factor more important than another: grades. However, as carbon content increases,
Chapter 15: Machining / 183
Table 1 Machining setup recommendations for turning wrought stainless steels

Uncoated Coated
High-speed steel tool Tool Tool
Hardness, Depth of Speed, Feed, Tool material Speed, fpm Feed, material Speed, Feed, material
Material HB Condition cut(a), in. fpm ipr AISI Brazed Indexable ipr grade fpm ipr grade
Ferritic steels
405, 409, 429, 135-185 Annealed 0.040 150(235) 0.007 M2, M3 575 650 0.007 C-7 850 0.007 CC-7
430, 434, 0.150 120(190) 0.015 M2, M3 450 500 0.015 C-6 650 0.015 CC-6
436, 442, 0.300 95(150) 0.020 M2, M3 350 400 0.030 C-6 525 0.020 CC-6
446(c) 0.625 75(115) 0.030 M2, M3 275 310 0.040 C-6 ... ... ...
Austenitic and duplex steels
201, 202, 301, 135-185 Annealed 0.040 95 0.007 M2, M3 325 375 0.007 C-3 500 0.007 CC-3
302, 302B, 0.150 75 0.015 M2, M3 300 325 0.015 C-3 425 0.015 CC-3
304, 304L, 0.300 60 0.020 M2, M3 225 250 0.020 C-2 325 0.020 CC-2
305, 308, 0.625 45 0.030 M2, M3 175 200 0.030 C-2 ... ... ...
309, 309S, 225-275 Cold drawn 0.040 80 0.007 T15, M42(b) 300 325 0.007 C-3 425 0.007 CC-3
310, 310S, or 0.150 65 0.015 T15, M42(b) 250 275 0.015 C-3 350 0.015 CC-3
314, 316, duplex 0.300 50 0.020 T15, M42(b) 290 215 0.020 C-2 275 0.020 CC-2
316L, 317, annealed 0.625 40 0.030 T15, M42(b) 140 165 0.030 C-2 ... ... ...
321, 330,
347, 348,
384, 385(c)
2205, 2507 295-310 Annealed
Martensitic and PH steels

403, 410, 420, 135-175 Annealed 0.040 155 0.007 M2, M3 475 620 0.007 C-7 800 0.007 CC-7
422, 501, 0.150 125 0.015 M2, M3 400 480 0.015 C-6 625 0.015 CC-6
502(c) 0.300 100 0.020 M2, M3 320 380 0.030 C-6 500 0.020 CC-6
0.625 80 0.030 M2, M3 240 300 0.040 C-6 ... ... ...
175-225 Annealed 0.040 145 0.007 M2, M3 460 570 0.007 C-7 850 0.007 CC-7
0.150 115 0.015 M2, M3 385 450 0.015 C-6 550 0.015 CC-6
0.300 90 0.020 M2, M3 300 350 0.030 C-6 450 0.020 CC-6
0.625 70 0.030 M2, M3 235 265 0.040 C-6 ... ... ...
275-325 Quenched 0.040 95 0.007 T15, M42(b) 360 465 0.007 C-7 700 0.007 CC-7
and 0.150 75 0.015 T15, M42(b) 280 360 0.015 C-6 450 0.015 CC-6
tempered 0.300 60 0.020 T15, M42(b) 225 280 0.020 C-6 375 0.020 CC-6
375-425 Quenched 0.040 65 0.007 T15, M42(b) 290 320 0.007 C-7 475 0.007 CC-7
and 0.150 50 0.015 T15, M42(b) 225 250 0.015 C-6 300 0.015 CC-6
tempered 0.300 40 0.020 T15, M42(b) 180 200 0.020 C-6 250 0.020 CC-6
PH, precipitation-hardenable. Source: Ref 1
(a) Caution: check horsepower requirements on heavier depths of cut. (b) Any premium high-speed steel (T15, M33, M41–M47). (c) Free machining versions.
the amount of carbide increases, the structure as duplex and 200 series alloys, manganese has
changes to martensitic, and the wear on tools the same relative effect as nickel when used in
increases. greater amounts, as, for instance, in the 200 se-
Nickel increases the toughness and ductility ries stainless steels.
of stainless and reduces the work hardening Molybdenum increases the strength and ele-
rate. Nickel also increases elevated temperature vated temperature mechanical properties. This
mechanical properties. This causes chips to be increase in hot hardness and strength means
more difficult to break. Nickel will have a ten- more energy will be needed to cut the material,
dency toward increased BUE; however, better thus creating hotter cutting conditions. While
tool life will generally result. the molybdenum helps in chip breakability, it
Sulfur reduces mechanical and corrosion will require more rigid setups and will reduce
properties and can be a cause of hot cracking in tool life.
the resulfurized grades. It is best known as a Copper improves ductility and reduces the
free-machining contributor that promotes better strain-hardening or work-hardening rate (with
tool life and greater machining speeds. the exception of participation-hardening alloys,
Manganese is generally added to combine for which copper is used as the precipitant).
with sulfur to form manganese sulfide (MnS), Chips can be more difficult to break, which in-
which acts as a self-lubricant and improves creases the tendency of BUE and promotes bet-
machinability. In high-manganese grades, such ter tool life.
Nitrogen strengthens stainless steels. It aids measure and control it to varying degrees. The
in chip breakability and reduces BUE but in- material’s grain size results from the thermal
creases tool wear. and mechanical history during manufacturing
Titanium promotes carbide formation and in- and from the mill’s equipment capability and
creases tool wear. practices.
Niobium promotes carbide formation and in- The grain size of a particular product can dra-
creases tool wear. matically change its machinability. It is entirely
The production of stainless steels is identi- possible for the grain size difference between
fied by industry specifications, such as AISI, two lots of material to be large enough to pre-
UNS, EN, JIS, etc. These specifications are all vent both lots from being effectively machined
defined with fairly broad elemental chemical with the same setup, requiring adjustments in
compositions. For example, an AISI 304 has a the machining setup to remedy the situation.
2 wt% window for the nickel content; that is, Finer grain sizes strengthen the stainless steel,
this grade can have a nickel level of 8 to 10%. cause hotter cutting conditions, and have a
A type 304 with 8% nickel can have different higher tendency of BUE. On the brighter side,
machinability characteristics from a type 304 finer grain sizes yield better surface finishes and
with 10% nickel. This 2% difference alters chip smoother roll thread crests.
morphology and surface finish capability. Since
today’s mill technology can meet very tight ele-
Process
mental targets within the grade specification,
how the mill balances the grade’s chemistry The type of equipment used by the stainless
will provide the foundation of its machinability manufacturer, the manufacturing sequence, and
characteristics. the practices employed by the mill can affect
machinability as well as mechanical properties,
but more important, processing affects how
Cleanliness consistently the material can be machined. The
The cleanliness of steel is determined by the melt type, hot rolling parameters, thermal treat-
amount and type of inclusions it contains. Vac- ments, cold-finishing parameters, and sequence
uum and argon oxygen decarburization (AOD) of these operations can affect how consistently
melting and refining along with proper steel- a material machines. Many times, the culprit is
making techniques can reduce the inclusions to equipment operational procedures or practices
negligible levels. It is beneficial to machinabil- that can vary one day to the next. Equipment
ity to avoid hard inclusions. However, certain types can also play a role in machinability. For
inclusions are plastic and act as solid-state lu- example, machinability can vary when the
bricants and chip breakers and prevent adhesion same-size material is drawn across two different
of the material to the tool. The beneficial effect draw benches using different pulling mecha-
of controlled inclusions is discussed in this nisms and two different straightening mecha-
chapter. nisms. Whether the material is continuously an-
nealed or batch annealed can cause different
strain distributions across the material cross
Structure section as well as material strength differences.
Material structure consists of both the phases Various annealing lines vary in time/tempera-
that are present and the microstructure of those ture profiles and therefore result in different
phases. Each type of stainless steel belongs to a grain size and mechanical properties.
larger family, which is characterized by a single With all this in mind, manufacturing consis-
predominant phase or a combination of two. tency can be a great asset in machinability. A
These are ferritic, austenitic, martensitic, pre- machine shop can adjust when a material is con-
cipitation hardening, and duplex (see the chap- sistently bad, but it is very difficult when one lot
ters on stainless steels, Chapters 6 to 10, in this is easy to machine followed by a lot that is
Volume). Their machining characteristics are tough to machine, while a third bundle performs
described in the next section. The microstruc- differently from the first two. Mills that promote
ture of a given alloy is independent of the grade machining consistency pride themselves by
type and composition and is mainly influenced practicing manufacturing consistency. Toler-
by grain size. Grain size is not normally speci- ance variation will be tighter and machining
fied or reported on certifications; however, mills costs will be lower with their products.
Cross-Section Size will tend be stringy but can be broken through

aggressive chip breaking, and surface finishes
Mill processing equipment dictates a manu-
will be somewhat of a challenge. Ferritics are
facturing route based on size. Smaller diameters
the easiest of the stainless steels to machine, do
are cold drawn, while larger diameters are
not require much horsepower, have a low work-
straightened/cut/turned, yielding a softer prod-
hardening rate and better tool wear, and will
uct. This can have an impact on machining per-
generally have higher speed and feed capabili-
formance despite all other factors being the
ties than other stainless families.
same. Cold finishing of stainless steels can be
accomplished via a couple of general manufac-
turing routes. Martensitic
The first is by cold drawing to bar, and the sec- Martensitic stainless are also very basic
ond is simply a straightening-turning operation. straight chromium stainless steels, 400 series
The mechanical properties of the straightened- stainless grades, and are similar to the ferritic
turned bars will be softer than the bars made by grades. The difference is that the martensitic
cold drawing. Typically, sizes greater than 1 in. grades have much higher carbon levels, which
(25 mm) are annealed/turned and straightened, further strengthen the materials and allow these
with virtually no strain in the product. materials to be hardenable by heat treatment.
These grades will have higher carbide levels,
which will lead to higher tool wear. This is es-
Machinability of the Stainless pecially true if the material is being machined in
Steel Families the hardened condition. The higher strengths
will require more horsepower to cut and will
Comparing the machinability of stainless need more rigid setup than ferritic steels. The
steels with other materials such as carbon steels, work-hardening rate of martensitic stainless is
brass, or aluminum, there are some striking dif- lower than for ferritic stainless. Martensitic
ferences. In general, stainless steels have: stainless also has a small yield-to-tensile ratio,
making chips easier to break.
1. Low thermal conductivity
2. High work-hardening rates Austenitic
3. High tensile strengths
4. High toughness The austenitic grades, the 300 series stainless
5. High ductility grades, are more difficult to machine than the
6. Large spreads between the yield and tensile ferritic and martensitic families. Austenitic
strengths stainless steels are more highly alloyed and are
more prone to higher work-hardening rates. This
Each stainless steel family (ferritic, marten- leads to the need for higher horsepower and
sitic, etc.) brings its own general set of machin- more rigid setups. These grades are very prone
ing rules. This is mainly due to the chemistry of to BUE and hence are prone to poorer surface
these families and its resultant effect on the finishes and tend to tear. The yield-to-tensile ra-
physical and mechanical properties. A general tios of austenitic stainless steel is very large,
description of the machining behavior is pro- making chips hard to break. Chips in this family
vided next. One must keep in mind that these are of alloys tend to be long and stringy. The higher
general characteristics. Further alloying of these strength and higher ductility of these grades also
families, such as with a sulfur addition, can re- tend to increase cutting temperatures, necessitat-
sult in a radical change in machining behavior. ing tooling with higher heat resistance.
Ferritic Precipitation Hardening

Ferritic stainless steels are the most basic Precipitation hardening stainless steels are
stainless steels and are part of the 400 series characterized by higher strength and toughness.
grades. Their basic chemical composition is The solution-annealed hardness of AISI 630, for
iron and chromium. These grades generally ex- instance, is HRC 36 versus HRC 23 for a 304.
hibit lower strengths, more ductility and soft- Higher horsepower requirements, high tendency
ness, and a close yield-to-tensile ratio. These to BUE, higher tool wear, and difficulty break-
grades will have a high tendency to BUE, chips ing chips are familiar scenarios for this class of
stainless. Except for alloy A-286, the precipita- machinability of a material. It was once a lead-
tion-hardenable (PH) grades are all martensitic ing addition to carbon steels, but its reported en-
alloys and can be treated as such for machining vironmental toxicity has diminished its role.
purposes. Many carbon mills are looking at other machin-
ing agents to replace lead. Lead, however, has
Duplex not been a large factor for stainless steels be-
cause of the extremely negative effect it has on
The duplex is a unique class of stainless char- hot workability, always a serious consideration
acterized by a dual-phase structure. Duplex al- in stainless steel design.
loys have a structure that is roughly a 50% mix Selenium and tellurium have similar charac-
of austenite and ferrite; thus, two hardness ma- teristics to lead as additives but also are non-
terials with different hardnesses coexist side by competitive due to cost, toxicity, and incompati-
side. The tool will alternate cutting between soft bility with stainless.
and hard grains of the duplex structure, leading
to an automatic tendency to initiate chatter in
the cutting system. Strength levels of duplex al- Sulfur
loys are quite a bit higher than austenitic grades. It became obvious very early to metallurgists
Between the duplex structure and high-strength that higher sulfur levels correlated to better
levels, high horsepower is necessary, and highly machinability, and sulfur remains the popular
rigid setups are required. Some work at Ugitech additive choice. Sulfur is cost-effective as a
found that to effectively machine these grades, machinability additive and can be easily re-
highly alloyed carbide tooling with high hard- moved with modern refining methods. Sulfur is
ness and high heat resistance, such as the a natural impurity and has negative effects on
C7/C8-type carbides, should be used. mechanical and corrosion properties, discussed
separately here. The role of sulfur as a machin-
Super Stainless Steels ability agent in stainless steels is very complex
Super stainless steels are today’s highly spe- and not necessarily straightforward, but in gen-
cialized stainless grades. These grades, like the eral sulfur has been extremely beneficial in in-
duplex alloys, are being developed to increase creasing the machinability of stainless steels.
corrosion performance parameters to meet some Generally, as the molten metal cools to solid
of today’s increasing performance require- form, sulfur combines with manganese to form
ments. These alloys are more highly alloyed manganese sulfide inclusions. Manganese be-
than the duplex materials. Strength levels are comes a very important variable during this re-
higher and toughness is greater, driving machin- action. Two basic sulfide forms are found in
ability downward. stainless steels—manganese sulfides and
chromium sulfides—and the form the sulfide
will take depends on the manganese content.
Role of Inclusions When manganese levels are less than 0.4%,
chromium sulfides and chromium-rich sulfides
Metallurgists have long known that the pres- will be present. As manganese levels reach 0.4
ence of a soft second phase dispersed in the ma- to 1.8%, chromium-rich manganese sulfides are
trix of a parent metal can improve its machin- present. For manganese levels beyond 1.8%,
ability. These particles provide a solid-state comparatively pure manganese sulfide will be
lubricant between the chip and tool or a discon- found. The manganese-to-sulfur ratio is also im-
tinuity in the material to aid in chip breaking. portant. For highest machinability levels, a high
The challenge to the alloy designer has been to manganese-to-sulfur ratio is desired. However,
develop second phases that produce these bene- if corrosion resistance is desired, a low man-
ficial effects with a minimal of degradation to ganese level is preferred to encourage the for-
the material itself. mation of chromium sulfides or chromium-rich
manganese sulfides since these sulfide forms
have superior corrosion resistance.
Lead, Selenium, Tellurium
Sulfides form initially as spherical inclusions
The range of additions possible to stainless within the cast structure. Hot working, as well as
steel is the same as for carbon steel. Lead addi- cold working, elongates these inclusions as
tions probably are the best source for improving shown in Fig. 2. This elongation increases the
machining speeds can be increased, improving

machining productivity.
Much has been written about the benefits of
sulfides in a machining operation. A metallurgi-
cal perspective shows the more complex nature
of sulfides. And, there is more to this than just
adding sulfur. The discussion of the combina-
tion of manganese and sulfur revealed that there
is a particular balance of manganese to sulfur to
achieve desired needs.
In addition to chemistry (manganese-sulfur
balance), the size and shape (relative elongation
or globular nature) of sulfides contributes to the
machinability of stainless steels. Sulfides are
defined into four categories based on morphol-
ogy: type I to IV sulfides. Type I sulfides form
from the melt as large globular sulfides and are
assumed to be best for machinability. Type II
sulfides generally form as a eutectic-like distri-
bution of finer rod-like sulfides in interdendritic
regions. Type III sulfides form as angular-
shaped particles. Type IV sulfides form from the
melt as plate-like sulfides in a ribbon-shaped
pattern. It has been shown that the larger, more
Fig. 2 Typical AISI 303. Source: Ref 2
globular (less-elongated) type I, homoge-
neously dispersed sulfides are better than the
other types for enhancing machinability in both
carbon and stainless steels. However, it seems
that sulfides that are too large or too small can
be detrimental to machinability. The coarsest
type I sulfides, once thought to be best for
machinability, are difficult to attain with today’s
stainless manufacturing equipment. Ingot cast-
ing, with the slow cooling and solidification
rates, is beneficial to the creation of the coarser
type I sulfides. With the transition to continu-
ously cast blooms, solidification rates are much
faster, effectively creating a finer type I sulfide.
The effect of sulfur on the machinability of
stainless steels is more effective in smaller
Fig. 3 AISI 303 chip breaking at the sulfides. Courtesy of amounts than in larger amounts. Figure 4 shows
Ugitech a graph of drill penetration results in varying
amounts of sulfur in an 18-9 stainless. It is eas-
surface area of the sulfides and will form weak ily seen that small additions of sulfur have the
planes that mechanically weaken the steel, per- greatest effect at sulfur levels <0.10%. This also
mitting chip breakage. The deformation caused can be proven with a 304 or 316 stainless steel,
by the severe bending of the chip during cutting for which sulfur levels are inherently lower than
causes the chip to break at the sulfide striations, 0.030%. Machinists readily can see differences
shown in Fig. 3. An additional benefit of these in tool wear between a 316 with 0.023% and
manganese sulfide stringers is that when these one with 0.028% sulfur. That small addition of
stringers touch and pass the cutting tool, a small sulfur has a dramatic effect. On the opposite
amount of manganese sulfide is deposited on the side of the scale, with sulfur levels greater than
tool surface, providing a lubricating layer be- 0.20%, the curve flattens out.
tween the chip and tool. This reduces friction, From a machining perspective, stainless
thus reducing heat to the tool. Consequently, steels can be classified into three groups based
Sulfur hurts corrosion resistance by locally de-

pleting chromium from the matrix to precipitate
as a sulfide. These manganese sulfide inclusions
become exposed on the surface of the bar and be-
come initiation sites for pitting corrosion. Passi-
vation of the components will help, but the hole
left behind by the sulfide becomes a collection
site for contaminants. Other sulfide inclusions
that form on the grain boundaries contribute to
intergranular corrosion. Hot workability is also
hurt by sulfides. Sulfur increases the hot short-
ness of materials during hot-forming operations.
Fig. 4 Effect of sulfur on stainless machinability. Source: Ref 2
Manganese sulfides form stress risers within
the material, which lead to reduced mechanical
on the sulfur additions. In the first group, sulfur properties in notch-sensitive alloys, especially in
content is limited to 0.010%. For these stainless the transverse direction. As the amount of sul-
steels, corrosion resistance, weldability, or me- fides increases in stainless steels, a susceptibility
chanical properties are more important than to longitudinal cracking can become an issue.
machinability. Grades in this group are With these stress risers in place, any cold de-
316LVM and other remelted stainless grades, formation can lead to cracking along the sul-
many duplex and super stainless grades. The fide stringers. As a general rule, the smaller the
second group has sulfur contents ranging from bar, the higher the sulfur, and the higher the
0.010 to 0.030%. This group of stainless steels strain produce a high probability for a crack to
still holds corrosion resistance, welding, and initiate.
mechanical properties as critical but has the ad- Despite these deficiencies, the sulfides found
vantage of increased machinability. Grades in in stainless steels are very effective in improving
this group include 316/316L, 304/304L, 321, machinability, especially in austenitics, where
347, 410, and 430, among others. The third contact forces are very high. The benefits of sul-
group with sulfur levels ⭓0.15% are considered fur to improve machinability outweigh the
resulfurized grades. Sulfur levels are generally losses due to defective parts, at least from the
0.25 to 0.35%, with some grades reaching machinist’s viewpoint. The question of perform-
0.45%. The grades in the third group include ance of the finished component is another issue,
303, 420F, 430F, 430FR, 1.4570, and others. A which concerns the end user; for the end user,
fourth group of grades can be included in this sulfur is a major negative factor for corrosion re-
classification, but the amount of sulfur is not the sistance. This has led to the development of al-
grading criterion. This is where mills will rebal- ternative methods of improving machinability.
ance chemistries and processing to enhance
machinability beyond what the first three groups
Oxides
offer to create enhanced machining grades.
Product offerings include Ugitech’s Ugima and The basic machinability-enhancing agent dis-
Ugima XL (Ugima 2) materials, Sandvik’s San- cussed thus far has been sulfur. The beneficial
mac materials, Outokumpu’s Prodec materials, effects of sulfur are undeniable, but the detri-
and Carpenter’s Project 70+ materials. mental effect is equally evident. This has led
Sulfides in stainless steels have a dark side, steel producers to look at other inclusion sys-
especially when sulfur levels exceed 0.15%. It tems for a viscoplastic inclusion without the
is well documented that sulfides negatively af- negative effects of sulfur. Since the 1900s,
fect corrosion resistance and mechanical prop- steelmakers have known that injecting calcium
erties of stainless steels. Sulfur tends to form into the melt converts refractory inclusions into
segregated films with low melting points during soft, malleable, complex oxides that act as free-
solidification. And, since these films are low machining agents with high-temperature lubri-
strength, they may induce the formation of mi- cating capabilities. The oxide inclusion chem-
crocracks brought on by solidification shrinkage istry is based on the CaO-Al2O3-SiO2 system.
stresses. Further processing of these materials Figure 5 shows that small amounts of calcium
can induce cracking even further, leading to can greatly increase tool life. These calcium-
poor processing yields. based oxide formulations have been commer-
Fig. 6 Complex Ugima oxides populating the 303 matrix.

Courtesy of Ugitech
Fig. 5 Effect of calcium on machinability of 303. Source:

Ref 2
Fig. 7 X-ray examination showing manganese sulfides and Ugima oxides coating the tool surface. EDAX, energy dispersive analysis
by x-ray. Courtesy of Ugitech
cialized for stainless steels, but the machinabil- The Ugima oxide works similarly to sulfur by
ity agent is, at this point, not standardized as coating the cutting tool and acting as a lubricant.
this process is difficult to reproduce consis- Figure 7 shows EDAX (energy dispersive analy-
tently from heat to heat. However, Ugitech SA sis by x-ray) spectra of the surface of a carbide
(formerly Ugine Savoie and now part of tool, proving the existence of coatings of man-
Schmolz and Bittenbach) developed a propri- ganese sulfide and Ugima oxide. The Ugima
etary and patented process sold under the trade oxide performs synergistically with sulfur. Since
name Ugima and Ugima XL (Ugima and Ugima the oxide alone has limited lubricating abilities
2 in Europe and Asia). Figure 6 shows the at slow machining speeds, the manganese sul-
Ugima oxide coexisting with sulfur in AISI. fides in the stainless are the dominant machin-
ing agent. As machining speeds increase, man-

ganese sulfide eventually loses its lubricity, and
the oxide acts as a high-temperature lubricant
that will allow faster machining speeds without
the need to use specialized tooling. Data gener-
ated at Ugitech as well as data generated by
other producers have shown that once sulfur
content reaches 0.30%, the contribution toward
machinability flattens and possibly decreases,
which seems to contradict the experience of
many machine shops. However, the combina-
tion of sulfur and the Ugima oxide extends the
machinability range for sulfur levels beyond
0.30%, as seen in Fig. 8.
Tool lubrication is only one function the com-
plex oxide performs. Like sulfur, the Ugima Fig. 8 Comparison of machinability of AISI 303 at different
sulfur levels with and without the Ugima oxide.
oxide is a discontinuity and will aid in chip The vertical axis, VB30/0.3, represents 0.3 mm of tool wear in
30 min.
breakage. Figure 9 exhibits cross-sectional pho-
tomicrographs of 304L chips with a 0.025% sul-
fur level. As shown, the chip is shearing along Tooling and Coolants
complex oxide stringers, helping the chip break.
Ugitech has seen synergistic effects with The machining of stainless steels can be
copper additions as well. Machining tests at much more complicated than for other materi-
Ugitech and many field experiences have shown als. Machine and tool rigidity, machine power
increases in machining performance of 30% requirements, sharp cutting tools, and optimum
with the addition of 1.4 to 1.8% copper to a lubrication type and amounts are very impor-
303 stainless steel with the Ugima oxide. The tant. Feed rates need to be sufficiently more ag-
grade chemistry meets EN 1.4570 and is sold gressive than with carbon steels to prevent work
under the trade name 303 Ugima UX (4570 hardening and glazing of the material as well as
Ugima in Europe). Further additions of copper to avoid reducing tool life.
up to 4% have exhibited even better machining High-Speed Tool Steels. The cutting tools
performance. themselves are the main variable other than the
Fig. 9 Comparison of 304L chips with and without the Ugima oxide. Courtesy of Ugitech
material in determining the quality of the ma- hard layer on the tool surface that will provide
chined part. The earliest cutting tools were additional lubricity between the tool and chip as
those made of the high-speed steels (HSSs). The well as potentially providing heat resistance for
wrought HSS tooling was very versatile, al- the tool. Like the carbide, there are certain coat-
lowed resharpening of the tools many times, ings that are more beneficial than others. Tita-
and is still a preferred choice in many shops. nium nitride (TiN) was a great coating when it
With time, mills perfected their ability to pro- was first introduced, but further development in
duce more highly alloyed tool steels to meet the this area has created other coatings that work
increasing demands of the machining industry. even better. Grades like titanium-aluminum-
Today, tools are being made with a powder met- nitride (TiAlN) and aluminum-titanium-nitride
allurgy process, by which ingots of compacted (AlTiN) are great coatings for turning, form cut-
high-speed tool steel powders have more struc- ting, cutoff, grooving, drills, reamers, and
tural homogeneity and thus better wear and heat milling. These tools are usually subjected to
resistance. Powder metallurgy techniques pro- higher temperatures from high speeds, deep
duce even more highly alloyed high-speed tool cuts, and limited coolant. The aluminum in the
steels with properties approaching those of car- coating breaks down and combines with the
bide tooling, allowing better machining per- surrounding oxygen to form aluminum oxide
formance. Although high-speed tool technology (Al2O3), a material with higher hardness and
has improved, its limiting factors are heat and better thermal resistance. When tooling will not
wear resistance. Use of HSS tooling still limits be subjected to much heat, such as with taps,
machining speeds and performance. high wear resistance coatings such as titanium
Carbides. Carbide tooling is the next genera- carbonitride (TiCN) coatings are good.
tion of tool materials after HSS, although it Coolants. Stainless cutting is hot, has high
tends to be hard and brittle. In the past, unless it frictional forces, and has tendencies of the
was possible to feed hard, maintain fast speeds, metal to stick to the tools. Coolants need to re-
and have uninterrupted cuts, carbide tooling move this heat and provide lubrication to reduce
was a good choice. However, carbide tool tech- friction and minimize BUE. Coolants available
nology has come a long way in grades and are petroleum based, semisynthetic, synthetic,
technology. Micrograin and nanograin carbides water soluble, and the new type: vegetable
are providing carbide tooling with increased based. All of these coolants need to be highly
toughness. Single-point tooling is now able to fortified for use in stainless machining. For ex-
withstand the punishment of an interrupted cut, ample, petroleum-based oils need high sulfur,
and drills are able to withstand some flexing. chlorine, and fat levels for lubricity at higher
Generally, the carbide grades to be used when temperatures. The other coolants need high-
machining stainless steels are the C5- to C8- pressure additives, again to help lubricity. It is
type carbides. These classes of carbides are always a good idea to discuss the options with
harder and have more heat resistance, properties the coolant supplier.
that are needed when cutting stainless. The dis-
cussion in this chapter has stated that the cutting
of stainless is more difficult and generates more
REFERENCES
heat. Cutting materials that withstand these cir-
cumstances are needed. The C2- to C4-type car- 1. Machining Data Recommendations, in Met-
bides are not well suited for stainless steels be- als Handbook, desk ed., 2nd ed., J.R. Davis,
cause they do not have the heat resistance Ed., ASM International, 1998, p 917–950
needed. There are always exceptions. For exam- 2. T. Kosa and R.P. Ney, Sr., Machining of
ple, it is possible to use the C2- to C4-type car- Stainless Steels, in ASM Handbook, Vol
bides when using the older cam-operated multi- 16, Machining, ASM International, 1989,
spindle and Swiss machines, for which machine p 681–707
speeds are limited. Tooling manufacturers such
as Kennametal, Sandvik, Iscar, etc. have their
own proprietary grade designations, but many
correspond to the C5 to C8 types. SELECTED REFERENCE
Coatings. Tool coatings have contributed to
machining improvements. Coatings add a very • www.ugitech.com
CHAPTER 16
Surface Finishing
Summary Beyond such considerations, one can also

create surfaces on the stainless that enhance its
SURFACE TREATMENTS are extremely beauty and performance. The surface can be
important to the end user, and they are totally made reflective or matte, ground or mechanically
within the end user’s control and specification. patterned, coated, painted, plated, or oxidized.
They include the proper cleaning of stainless, It can be treated chemically or electrolytically.
the various means of descaling after thermal The surface can be altered on an atomic basis,
treatment, and the choice and application of sur- sometimes producing profoundly different me-
face finishes. The cost ramifications of improper chanical and corrosion-resistant properties. All
or suboptimal surface treatments are immense of these are discussed in this chapter.
because of the possibility of them compromis-
ing corrosion performance, which is a charac-
teristic for which the end user pays dearly. Function of Surface Treatments
Removal of Oxide Scale

Introduction Oxide scales form on stainless steel during
annealing, hot-forming, and joining operations.
Surface finishing is usually very important for Removal of this scale is important to proper cor-
stainless steel. The underlying economic justifi- rosion resistance. This is because the chromium
cation for using stainless steel is that it does not in the steel oxidizes much more readily than
corrode if properly specified for the environment other elements, so the surface of the steel under
it faces. Thus, its surface appearance remains the oxide is chromium depleted and therefore
normally intact throughout its life. This appear- has lost possibly a significant amount of corro-
ance should therefore be aesthetically pleasing, sion resistance. An oxide scale is quite different
even in an industrial setting, and the surface fin- from a passive film. Scales deplete chromium;
ish should not detract from its performance. passive films enrich the surface in chromium by
Raw stainless surfaces resulting from rolling selective loss of iron.
and annealing operations are not considered at- Oxide scales are arguably best removed by
tractive and are used only for functions in which pickling. Pickling is the chemical dissolution of
aesthetics are a negligible consideration. Even the oxide scale. The acids commonly used are ni-
so, there are surface treatments required for tric (HNO3), hydrofluoric (HF), and sulfuric
stainless steel intended for such uses. The stain- (H2SO4). HCL, which is commonly used to
less surface must be freed of: pickle carbon steel, is not recommended for stain-
• Oxides resulting from annealing, joining, or less because it locally attacks (i.e., pits) the sur-
hot forming face. HNO3 is an oxidizing acid that by itself does
• Accumulated ambient foreign material not pickle stainless. It is used in combination with
• Applied process materials, such as forming HF to modulate the attack by the strongly reduc-
lubricants, fluxes, etc ing action of the HF. This combination allows
• Contamination from other materials, espe- good control of pickling rates by varying the ratio
cially iron. of the two acids. H2SO4 is used alone. It is often
used as a preliminary pickling because of its ag- The HNO3 provides:

gressive, and less-controllable, action.
• a source of H+ ions
Pickling involves both scale dissolution and
• stabilization of the passive film
metal dissolution. Thus, pickling can be consid-
• elevation of the redox potential
ered a deliberately imposed corrosion process.
• an oxidizing agent of the base metal
The scale formed during hot rolling may be an
• a dissolving agent for the scale
embedded scale with a minimal chromium-
depleted layer beneath the scale. Annealing and The nitrates carried off from this pickling
shotblasting are required before pickling so that process are an environmental problem. It has
the acid can penetrate to the chromium-depleted been shown that the HNO3 can be eliminated
region under the scale and selectively dissolve and replaced by hydrogen peroxide (H2O2).
this layer. Nearly all the action of pickling de- This is the proprietary, patented UG3P process
rives from undercutting of the oxide scale and developed by Ugine (Ref 1) or Henckel’s
dissolution of the chromium-depleted layer. It is “CleanOx” process.
also vital that this layer be removed as it dimin- Electrolytic pickling is commonly used in the
ishes the corrosion resistance of the surface. production of cold-rolled stainless. This process
Simply removing the scale mechanically may uses alternating positive and negative polariza-
be insufficient if only the oxide is removed and tion in baths using sulfates or nitrates. The H+
the compromised (depleted) metal surface is ions are produced by the polarization instead of
left. Metal removal rates are proportional to the by an acid, so neutral solutions can be used.
chromium level. Thus, sensitized material will The low solubility of trivalent chromium
undergo grain boundary attack if pickled. All (Cr+3), typically found in oxide scale, makes
stainless has slightly greater oxidation in the such scale hard to remove. Use of strong acids
grain boundaries when it is oxidized, and this and complexing agents is required to get pick-
leads to “ditching” of the grain boundaries dur- ling to occur at an economically acceptable rate.
ing pickling, resulting in the matte appearance Hexavalent chromium (Cr+6) is much more
of a pickled surface, as seen on two-dimensional readily soluble in aqueous solutions, for exam-
flat-rolled material. ple, oxidizing treatments such as Kolene,* a
Typical concentrations for H2SO4 pickling are molten mixture of nitrate salt and hydroxide, or
8 to 15% by weight. Bath temperatures are gen- electrolytic conversion treatments such as elec-
erally 150 °F (65 °C). The rate of pickling de- trolytic sodium sulfate. Both treatments convert
creases rapidly as iron builds up in the bath, and the trivalent chromium to hexavalent form, al-
the bath must be replaced for efficient pickling. lowing for easy scale dissolution. Their use
The attack of H2SO4 on the base metal can be may, however, lead to environmental problems
severe, and undissolved constituents can remain as hexavalent chromium compounds are toxic
on the surface as “smut.” This smut must be and heavily regulated.
physically removed or dissolved by subsequent Pickling may be assisted by prior treatment of
HF/HNO3 pickling. the scale in molten 85% sodium hydroxide
The HF/HNO3 pickling is carried out between (NaOH), 14% sodium nitrate, and 1% sodium
120 and 140 °F (50 and 60 °C). Higher tempera- chloride bath. The strong oxidizing action of
tures cause excessive HF evaporation and can such a bath chemically alters the chromium
also lead to visible emission of nitrogen oxides. oxide in the scale to a more soluble oxide with-
The NO2 (nitrogen dioxide) produced during out attacking the metal, making it more easily
the reduction of the HNO3 (oxidation of the removed by subsequent pickling.
metal) is visibly brown-red. The rate of NO2
formation increases at higher temperatures. Cleaning
HNO3 concentrations are normally from 10 to
25%. HF concentrations can vary from 1% for Stainless steel is very resistant to chemicals
light scales to 8% for maximum aggressiveness of many kinds, permitting it to be cleaned by
and difficult-to-pickle alloys. many aggressive agents. Because contamination
The HF provides: is “on” the surface of stainless rather than “in”
the surface as is the case with many materials,
• a complexing agent for iron and chromium simply using the cleaning agent appropriate for
• destabilization of the passive film
• stabilization of the redox potential * Kolene is a registered trademark of Kolene Corporation.
Chapter 16: Surface Finishing / 195
the contaminant will permit very thorough carbonate or brush material that is sufficiently
cleaning of the stainless surface. The stainless soft not to mar the stainless surface. The user is
surface will not be harmed by cleaning as long encouraged to test whether a product meets this
as strong halides, iron utensils, and abrasives requirement by testing it on a small, preferably
that alter the surface finish are avoided. Table 1 unexposed, area. Blasting the surface with car-
summarizes some cleaning recommendations bon dioxide pellets is a rapidly growing process
based on the contaminant to be cleaned. for removing paint and other adherent, soft
The greatest controversy regarding cleaning coatings and deposits without damaging the sur-
stainless steels involves the use of sodium face of stainless steel.
hypochlorite. Many highly qualified experts These cleaning recommendations apply to in-
flatly disapprove of its use because it can easily dustrial, architectural, and domestic uses of
leave a residue of chloride on the stainless sur- stainless. For special levels of cleanliness re-
face, leading to localized corrosion. But, in real- quired for medical, pharmaceutical, or semicon-
ity, health considerations make this position un- ductor applications, refer to the chapters dealing
tenable. Sanitation concerns in the food service with those applications.
industry take precedence over this prohibition. Passivation is a very commonly used surface
In fact, every stainless steel, as explained in the treatment to remove surface contamination, no-
“Corrosion and Oxidation” section of this Vol- tably iron, and to form a passive film. The film
ume, is resistant to some level of chloride con- forms of its own accord when a clean surface
centration at a given temperature and pH. Thus, encounters moist air, but film formation can be
short-term, room temperature use of hypochlo- accelerated by controlling the environment.
rite bleaches will generally not harm most types The surface to be passivated should first be
of stainless steel if well rinsed after application. cleaned by one of the methods discussed. This
Field experience has shown that stainless steels allows uniform passivation and avoids contami-
with less than 16% chromium can be harmed by nation of the passivating solution, especially by
hypochlorite bleaches. Damage occurs mainly chlorides, which can cause a rapid attack. Those
on abrasively polished surfaces in alloys with stainless steels with more than about 17%
16% chromium, such as 430, but does not occur chromium can be pickled in 20% HNO3 at 50 to
on roll-finished or bright-annealed surfaces. 60 °C (140 °F). Precipitation-hardenable (PH),
Stainless steels with slightly higher chromium martensitic, and straight chromium grades with
and nitrogen, such as 201, 301, and 304, are not less than 17% chromium should have 2.2 g/L
attacked unless concentrated chloride solutions sodium dichromate added to that solution to
are permitted to stay on the surface, particularly avoid attack. The free-machining grades are
in crevices. the most easily attacked of all types. They
The recommendation not to use abrasives should be immersed in 5% NaOH for 30 min at
does not include soft abrasives such as calcium 75 °C (165 °F) followed by a rinse before the
Table 1 Recommended cleaning methods

Contaminant Cleaning recommendation Comments
Exterior soiling Soap, detergent, or dilute ammonia Use a soft cloth or sponge, clean water; dry with
forced air or a dry cloth.
Fingerprints Detergent and warm water or a hydrocarbon Wax and oil polishes minimize fingerprinting. Glass
solvent cleaner is appropriate for mirror finishes.
Grease, oil Hydrocarbon solvent Alkaline cleaners may also be used in severe cases
but may require cleaning the entire surface to
maintain visual uniformity.
More severe stains, discolorations, Nonscratching creams or polishes Do not use HCl-containing products.
and rust stains Hypochlorite bleaches must be well rinsed to
avoid pitting.
Hard water scale, mortar 10–15% phosphoric acid, sulfamic acid- Neutralize with ammonia, rinse, and dry.
containing, or oxalic-containing cleansers Do not use HCl-containing products.
Oxides, heat tint If severe, treat by pickling; Scotchbrite©, If abrasives must be used, they should blend
stainless scouring pad, or nonscratching with existing surface finish in size and direction.
cream or polish
Paint Alkaline, trisodium phosphate, or hydrocarbon Follow manufacturer’s directions
solvent
nitric/dichromate passivation is performed. This Electrochemical methods permit more uni-

should be followed by a rinse, another 30-min form and reproducible color and tougher films.
NaOH treatment, and a final rinse. The most prominent is the International Nickel
In recent years, citric acid has become a pop- Company (INCO) process. In this two-step
ular replacement for HNO3 because it avoids process, the steel is first immersed in a mixture
the problem of toxic nitrates and hexavalent of one part chromic acid, two parts H2SO4, and
chromium. When citric acid is used, a 10% one part water at 30 °C (85 °F). This is followed
solution is applied for the same time and tem- by an electrochemical treatment with the same
perature as with HNO3. The use of NaOH be- bath or with H3PO4 substituted for H2SO4.
fore and after is still recommended for low- The colors produced are correlated to thick-
chromium and free-machining alloys. Refer to ness and treatment parameters (Ref 2) as shown
ASTM documents A 967-01, A 380-99, B 912- in Table 2. Rocha-Fila et al. confirmed that
00, as well as Federal Specification QQ-P-35. these coloring treatments did not degrade the
pitting corrosion resistance since their forma-
Brightening tion mechanism more closely resembles passive
film formation than oxidation.
Stainless steels can be brightened by chemi-
cal or electrolytic action, which selectively dis-
solves the surface in such a way that it becomes
microscopically smoother. Electropolishing is Aesthetic Surface Finishes
the most effective means of accomplishing this.
It causes the surface roughness to decrease by When steel is used in other than the as-
approximately one-half in Ra (the arithmetical annealed and pickled state, it is often for aes-
average of surface peaks and valleys as meas- thetic rather than functional reasons. The major
ured over a straight line). Electropolishing, like exception to this is temper-rolled strip, which
pickling, also selectively removes any exposed has a bright surface from cold rolling, but
chromium-depleted regions, leaving only the whose normal use is strictly functional. The hot-
bulk alloy with the intended corrosion resist- rolled or cold-rolled annealed and pickled sur-
ance on the surface. Apparent pitting resistance face finish is nonuniformly dull and unattractive
is thus increased. The resulting surface is bright to most observers. When stainless steel first ap-
and cleanable and provides the optimal corro- peared on the market, a highly polished surface
sion resistance that a given alloy can achieve. was the paragon, as showcased on the Chrysler
The reflectivity of the surface is a function of Building in New York City (Fig. 1).
the preelectropolished surface. This surface finish was very expensive to pro-
The most commonly used electropolishing duce since it had to be done by polishing and
solution is 40% H2SO4 and 45% phosphoric buffing. Soon, the more economical abrasive
acid (H3PO4), balance water, used at 90 °C polishing became the standard, and it had the
(194 °F) with a current density of between 1.0 benefit of removing the many cosmetic defects
and 3.0 amp/m2. Other baths using perchloric that then were common to the manufacture of
acid are technically good but carry the risk of stainless. Much higher levels of surface quality
explosion. were made possible with the development of the
Sendzimir mill to the bright anneal process per-
Coloring mitted much higher levels of surface quality.
Stainless steel passive films are so thin that
they are quite invisible. Oxide films are thicker Table 2 Parameters for oxide film coloring of
and through optical interference can cause differ- stainless steel
ent colors. The films are formed thermally in a Oxide film
process known as heat tint. The color depends on Heat treatment time Color thickness, nm
oxide film thickness, which is a function of the 10 Gold brown 70

time at a given temperature, well metal composi- 15 Brown/red/blue 100
20 Brown/blue 120
tion, and oxygen partial pressure. The colors typ- 25 Green/blue 140
ically range from light yellow formed at 300 °C 30 Green/gold/blue 165
(570 °F) through violet formed at 420 °C (790 35 Gold/green 185
°F) and dark blue formed at 600 °C (1110 °F). 40 Gold/green/brown 210
Reproducibility is sometimes a problem. 50 Red/brown 250
Stainless steel surface finishing is governed by:

• High strength and high work-hardening rate
require more power for metal removal.
• High surface hardness influences which
abrasive will be effective.
• Low thermal conductivity can cause exces-
sive temperatures during processing and
distortion.
• Oxidation (heat tinting) that occurs over
about 250 °C (480 °F)
• Residual stress due to surface working, es-
pecially in austenitic alloys
Grinding is a relatively coarse procedure with
usefulness that is largely confined to defect re-
moval. Polishing for aesthetic purposes is gen-
erally done with abrasive coated belts and is
done to both coils and sheets. There are no offi-
cial roughness values for the various surface
finishes. ASTM merely describes the finish by a
grit with which it is typically made. Table 3
shows the varying surface roughnesses that are
typical of polished stainless.
Neither the producer nor the customer is pro-
tected by a clear specification as of this writing.
Thus, the producer is encouraged to publish
standards, and the customer should specify min-
Fig. 1 View of the Chrysler Building in New York. imum and maximum Ra values for his purpose.
©iStockphoto.com/stevenallen
The typical number 4 polish varies in roughness
from the beginning of the coil to the end be-
Mirror finishes could be achieved with cold cause of the wear of the abrasive particles on
rolling followed by bright annealing. This has the belts. The difference can be more than 10
not entirely replaced sheet buffing to obtain μin. (0.25 μm) in Ra, is quite visually apparent,
mirror finishes, but sheet buffing has been rele- and will make adjacent pieces of stainless from
gated to those few applications where a high the same coil look different. This may be objec-
degree of perfection is demanded. The main tionable for certain products, especially appli-
methods of producing aesthetic surface finishes ances and architectural panels.
on stainless steel are abrasive finishes and rolled Polish appearance varies also with the pres-
finishes. The latter are superior in uniformity ence or absence of lubricant. Lubrication during
and corrosion resistance, but the polished fin- polishing (i.e., wet polishing) gives sharper cuts
ishes are still more common. and less heat tint from frictional heating, result-
ing in more surface brightness. Polish appear-
Polished Finishes ance also depends on the length of the grit line
caused by an individual particle of abrasive. The
Polishing is carried out with coarse- to length of the grit line varies with the speed of the
medium-sized abrasives that are bonded to a material as it passes the rotating abrasive belt,
flexible backing. This is distinct from grinding,
in which abrasives are bonded rigidly to each
other on a rigid backing. Buffing is done by Table 3 Polished finish designations based on
grit sizes to achieve target surface roughnesses
very fine abrasives, which are loose and do their
work by being forced along the surface by a soft Finish number Grit number Ra max, µm Ra max, µin.
material. The standard rule is that materials 3 60–120 1.0 40
used in finishing must not permit iron or iron 4 120–180 0.75 30
oxide particles to come in contact with the 7 240–320 0.30 8
stainless surface or passivation layer. 8 500 0.15 4
and the rotational speed and diameter of the and rolling by which the surface is made flaw-
backup roll for the abrasive belt. This combina- less and very smooth. Final brightening occurs
tion of variables, while seemingly controllable, with temper passing after bright annealing.
usually varies enough so that no two polished
sheets look identical to the trained eye. This Rolled Finishes
variability also causes problems in field repair.
Repair of a polished surface damaged, for The bright-annealed and temper-passed sur-
example, by welding or scratching is quite face may be used in its mirror-like condition,
challenging. A perfect match and blend to the or it can be used as a basis for rolled finishes,
surrounding original surface is an art, and no which can take on any appearance and can be
practitioner of that art can accomplish it per- engraved onto a temper mill roll. All polished
fectly. If the repaired area is small, then it can finishes can be duplicated as rolled finishes.
be acceptable, if not undetectable, but for large But, in addition, many other designs such as
areas, repolishing of the entire surface is often replicating fabric or leather, geometric de-
the best remedy. signs, or matte finishes can also be made. The
A solution to the problem of matching pol- pattern on the roll is impressed into the stain-
ished surfaces is to use very long grit lines for less surface, and elastic flattening of the roll
finishes of the same roughness. By eliminating and the stainless cause about 50% less Ra on
the variable of grit length, only correct pressure the stainless than on the roll. This effect in-
and grit size are required to achieve good visual creases with the yield strength of the stainless
matching. This can be done with a belt sander. being processed. The benefits of rolled finishes
These finishes are called “hairline” in Japan and are:
“grainline” and other names in the United • They can be made identically from coil to
States. Lack of uniformity, difficulty of repair, coil since roll engraving is quite precise.
and a decrease in corrosion resistance are the • They retain the enhanced corrosion resist-
chief drawbacks of polished finishes. ance of bright-annealed material.
• They are less expensive to produce, so the
Bright Annealing final product is generally priced lower than
Aesthetic finishes that do not depend on abra- polished material.
sion are derived from bright annealing. Bright In theory, such finishes can be made from air-
annealing is annealing in a very low oxygen annealed material, but finishes applied by roll to
atmosphere, either dissociated ammonia* or two-dimensional surfaces that have been dulled
hydrogen/argon. This process was originally by pickling are somewhat gray and indistinct. If
developed as a means of producing bright fer- the pickling is kept mild enough to retain surface
ritic trim for automobiles. That use has largely brightness, then it is possible insufficient prick-
passed out of favor, but the process itself is in- ling has occurred to remove the chromium-
trinsically superior to annealing in air because depleted layer caused by the air anneal. Some of
no oxide is formed only to be later removed, the finishes produced by rolling are shown in the
usually at significant expense. Because the orig- architectural chapter (see Chapter 18). These fin-
inal product was strip, many older bright-an- ishes can be applied to all the normal sheet
nealing facilities are narrower than normal sheet alloys, including the lean duplex alloys. This has
width. The wider, more modern lines are high permitted rolled finishes to be used on UNS
speed and wide enough to produce bright sheet. 32003 for building exteriors in Doha, Qatar,
Bright-annealed sheet will only be as bright and where corrosive conditions require a PREN
flawless as the cold-rolled sheet that is an- (percentage chromium equivalent) of 25.
nealed. Therefore, a quality product, generally The most critical applications from a surface
as mirror-like as possible, must be produced perfection point of view use rolled finishes
with great care. The better producers have pro- rather than abrasively finished number 4 pol-
prietary methods of prior pickling, annealing, ishes. These finishes also permit type 430 stain-
less steel to be used successfully in restaurant
* At one time, “dissociated ammonia” referred to a 3:1 applications without corrosion issues where
(mole ratio) mixture of hydrogen and nitrogen produced by
the catalytic decomposition of anhydrous ammonia. Today, polished type 430 material previously had
this composition is prepared by mixing gases. For many corroded. There are also instances for which
stainless steels, the 3:1 H2:N2 can be varied without difficulty. polished UNS S31600 was corroding in coastal
architectural applications in the United States. exposing an activated stainless surface to a

When the abrasively polished surface was high carbon fugacity at 470 °C (880 °F) for
replaced with rolled finish UNS S30400 around 200 h a 50 nm thick layer with 12 at.%
(Koolline), there was no further corrosion. Be- carbon in supersaturation can be achieved with-
cause alloying is such a high component of cost out carbide formation. The properties of this
in stainless, it makes sense to employ rolled fin- layer are phenomenal; the hardness is 1000
ishes whenever possible. However, it must be Vickers 25 compared to 200 for the base alloy
noted that when manual polishing is used to re- (Ref 3). In addition, the corrosion resistance in-
pair scratches or other damage in rolled finish creases significantly. Only by such supersatura-
material, its corrosion resistance may be reduced. tion with carbon could it be determined that
carbon, like nitrogen, is a powerful antipitting
Surface Alteration alloying element when kept in solid solution.
These processes are at the very initial stages of
In carbon steels, surface chemistry can be commercial use.
changed to affect certain properties. Carburiz-
ing and nitriding are examples of such
processes. Simply using these processes on REFERENCES
stainless cannot be done because these ele-
ments combine too strongly with chromium as 1. Stainless Steels, Les Editions de Physiques,
carbides or nitrides, dramatically reducing the 1989
corrosion resistance. There have been modifi- 2. C. Rocha-Filo et al., J. Braz. Chem. Soc.,
cations to carburizing and nitriding that permit Vol. 15 (No. 4), 2004, p 472–480
austenitic stainless steels to have very high car- 3. Y. Cao, F. Ernst, and G. Michal, Acta
bon levels implanted to a thin surface layer. By Mater., Vol 51, 2003, p 4171–4181
CHAPTER 17
Welding
NEARLY ALL WELDING of stainless steel characteristics of austenitic stainless steels that
is done by end users or processors. Like thermal distinguish them from ordinary carbon steels in
processing, it is complex in theory and practice. welding are:
This chapter gives a basis for understanding the • Austenitic stainless steels have lower thermal
influence of alloy composition and metallurgy
conductivity and higher thermal expansion
on the welding process, which must be re-
than carbon steels or ferritic stainless steels,
spected as a process that combines melting, re-
which can localize the heating, thus increas-
fining, and thermal processing. Knowledge of
ing the potential for residual stress and
each aspect is required for the process to be de-
therefore hot cracking.
signed and executed properly. • Stainless steels contain readily oxidized
The welding and joining of stainless steels re-
chromium, which must be protected.
quires knowledge of both the technology of the • Surface oxidation during welding depletes
welding or joining process and the response of
chromium in all types of stainless steel from
the steel to the thermal and mechanical effects
the underlying surface, resulting in reduced
of the process. The welding process must, of
corrosion resistance unless this layer is re-
course, produce a sound joint, but it must also
moved.
result in the weld and its surrounding affected • The possible formation of chromium car-
metal having correct strength, toughness, corro-
bides in the heat-affected zone (HAZ) can
sion resistance, etc. for the intended service
cause susceptibility to grain boundary corro-
conditions. This chapter does not attempt to
sion (sensitization).
teach welding. The main objective is to show • The possible precipitation of intermetallic
how standard welding technology is correctly
phases in the HAZ can lower toughness and
applied to stainless steels.
corrosion resistance.
The foremost special consideration of weld- • There is increased microsegregation in the
ing stainless steel as opposed to carbon steel is
fusion zone with increasing alloy content.
that the chromium in stainless steel, which is • There are thermodynamically metastable
what makes it stainless, must be protected from
conditions due to the low diffusion rates in
oxidation, so that:
the face-centered cubic (fcc) matrix.
1. It stays in solution as a corrosion-resisting
The influence of carbon has been well ad-
element.
dressed using low-carbon versions of all grades
2. It does not form refractory oxides that would
whenever welding involves significant time be-
diminish weld soundness.
tween 600 and 900 °C (1110 and 1650 °F). This
prevents rapid precipitation by reducing the su-
persaturation of carbon. The older method of
Welding Characteristics of preventing sensitization is to stabilize the alloys
Stainless Steels with titanium, as in type 321, or with niobium,
as in 347. This is foolproof only if carbon levels
Austenitic stainless steels are readily are low, less than 0.04%, since TiC can dissoci-
welded by nearly all welding techniques. The ate at elevated temperatures and not be able to
recombine successfully with titanium during some highly alloyed grades with compositions
cooling, permitting a thin zone of sensitization that do not permit a ferritic solidification mode.
called knife-line attack. Fortunately, most 321 In such alloys, sulfur and other contaminants,
and 347 are produced with carbon levels below such as phosphorus, oxygen, zinc, and copper,
0.03%. The higher carbon-stabilized alloys and must be excluded from the weld zone. Welds of
the high-carbon (>0.03%) unstabilized alloys less highly alloyed austenitics, generally those
must be annealed after welding to redissolve with less than 20% chromium, which are bal-
chromium carbides if the cooling was suffi- anced to freeze in a ferritic mode, retain some
ciently slow for the carbides to have formed. ferrite at room temperature, normally between
This is avoided only in thin-gauge (>1.5 mm, 3 and 10%. This is not harmful since the ferrite
0.06 in.) material or when the HAZ is drasti- is richer in chromium and in molybdenum, if
cally reduced, as in laser welding. present.
The high thermal expansion of austenitic The amount of ferrite expected can be meas-
stainless steel can cause high residual stress ured by magnetic devices and estimated from
around welds, which may require annealing to the Schaeffler diagram, a useful empirical map-
eliminate. Another serious threat posed by ther- ping of weld metal phase composition shown in
mal stresses is hot cracking. This can occur to Fig. 1. This diagram has an arbitrary cooling
material that has just solidified when geometric rate resembling that of tungsten inert gas (TIG;
constraints to contraction imposed by the sur- described in a separate section of this chapter)
rounding material imposed act on weak grain welds. Faster or slower cooling will change the
boundaries. This weakness occurs when the relative amounts of ferrite and austenite because
steel solidifies in an austenitic mode. When of the need for diffusion to achieve the most sta-
austenite freezes, it strongly rejects sulfur to the ble phase balance. Very rapid cooling, as with
intergranular areas, where it forms weak films. laser welding, tends to make austenitic welds
This is solved by balancing the composition so less ferritic and has the opposite effect in duplex
that alloys solidify first as ferrite, which does alloys.
not reject the sulfur, forcing it to precipitate as The Schaeffler diagram has been improved
sulfide inclusions within the grains. This ap- by the Welding Research Council’s adoption of
proach is highly effective but cannot be used for the modification shown in Fig. 2, which super-
Fig. 1 The Schaeffler diagram. Source: Ref 1

Chapter 17: Welding / 203
Fig. 2 Welding Research Council’s (WRC’s) 1992 constitution diagram
imposes the solidification mode as a function of Restricting heat input to under 16 kJ/mm
the composition. The crucial line on this dia- (400 kJ/in.) and interpass temperature to under
gram is dotted-dashed line AF, which delineates 150 °C (300 °F) helps to minimize each of
those compositions that solidify in a primary these risk factors inherent to the more highly
ferrite mode, precluding the problem of inter- alloyed austenitic grades. Note that the influ-
granular solidification cracking. ence of microsegregation of alloying elements
Another problem particular to the more is separate from and in addition to the negative
highly alloyed grades is the formation of inter- influence of sulfur on the corrosion resistance
metallic phases from long cumulative exposure of welds. Austenitics at the alloy level of 316
to temperatures in the 600 and 900 °C (1110 and and above should not have sulfur above
1650 °F) range (coincidentally, the same as for 0.001% for these alloys to deliver the expected
carbide precipitation). The slow diffusion of al- corrosion resistance.
loying elements in austenitics makes this a The austenitic stainless steel weld metal com-
lesser problem than in ferritics or duplex. This position can be altered by the gases to which the
adverse precipitation is largely prevented in the molten base metal is exposed. Lack of shielding
modern, nitrogen-alloyed grades, so these al- can lead to oxygen combining with chromium
loys are recommended if extensive welding is and other elements, creating slag and depleting
planned. The more highly alloyed grades also the alloy of needed elements. Thus, oxygen-free
suffer from greater microsegregation during so- gas mixtures are used to exclude the ambient at-
lidification. This causes austenitic dendritic mosphere from the molten pool during electric
cores to have lower chromium and molybde- arc welding. Inert gases provide the barrier,
num content and consequently lower corrosion while the addition of 3 to 5% by volume of ni-
resistance. Thus, the welds have lower resist- trogen gives the necessary partial pressure to
ance to localized corrosion. This is addressed by ensure that welds will not be depleted of vital
using more highly alloyed filler metal or by so- nitrogen content. Figure 3 (Ref 1) shows the in-
lution annealing the welds. fluence of nitrogen content of the shielding gas
• Duplex alloys are more sensitive to prob-

lems in the HAZ because their generally
high chromium and molybdenum content
plus their ferritic content make the precipita-
tion of embrittling intermetallic phases more
rapid than in austenitics, so minimizing the
total time at high temperature is the overrid-
ing concern.
• While carbide sensitization is not an issue
with the duplex alloys, the formation of in-
termetallic phases can cause loss of corro-
sion resistance.
• Duplex, like all stainless types, must be pro-
tected from oxidation by shielding gas, and
Fig. 3 Effect of weld shielding gas composition on crevice
since nitrogen is a crucial alloying element,
corrosion resistance of autogenous welds in AL-6XN
alloy tested per American Society for Testing and Materials especially in duplex alloys, it must be a
(ASTM) G-48B at 35 °C (95 °F)
component of the gas mixture.
• Cleaning before and after welding is equally
on corrosion resistance of a highly alloyed important in duplex as in austenitics.
austenitic grade. Excess nitrogen in the shield- Modern duplex alloys derive their impressive
ing gas (e.g., more than 10%) can cause poros- strength, toughness, and corrosion resistance
ity in the weld, and greater than 5% is detrimen- from their nearly equal percentage of ferrite and
tal to the life of the tungsten electrode. austenite. The nitrogen content of the austenite
The heat from welding can produce a surface brings its corrosion resistance up to that of the
oxide composed mainly of iron and chromium. ferrite phase, which is richer in chromium and
The underlying surface can be significantly de- molybdenum. Nitrogen additions partition to
pleted of chromium because of the loss of the austenite and thus both strengthens it and in-
chromium to this scale and therefore signifi- creases its corrosion resistance to close to that
cantly lower in corrosion resistance. Pits can of the ferrite. The early duplex alloys had a ten-
start in this thin layer and propagate into sound dency to form excessive ferrite when welded
metal beneath. For heat-tinted surfaces, the and formed embrittling intermetallic phases
darker the tint, the stronger will be the effect. To rather rapidly. The additions of larger amounts
fully restore corrosion resistance, the area must of nitrogen stabilized the austenite to higher
be ground to remove the oxide and any depleted temperatures, so welds did not become so fer-
base metal. This should be followed by acid ritic. The nitrogen also decreased the speed at
pickling, which completes the removal of the which intermetallic phases form, enlarging the
oxide and depleted zone. time window for welding without their precipi-
Duplex stainless steels differ from austenitic tation. And, by promoting greater austenite for-
stainless steels in their metallurgical response to mation at high temperature, the addition of high
welding mainly because their approximately (>0.12%) nitrogen actually reduces the ten-
50% ferrite causes greater thermal conductivity dency for chromium nitride precipitation. De-
at lower temperatures, and ferrite has greater spite these advances, the key precaution in
diffusion rates. These alloys solidify in a com- welding duplex alloys is to prevent the forma-
pletely ferritic mode, and since ferrite rejects tion of embrittling phases while preserving as
little sulfur on solidification, hot shortness is not close to a 50/50 austenite/ferrite structure as
a problem. So, compared to austenitic stainless possible. Minimizing time at red heat tempera-
steels, duplex stainless steels have the following tures (500 to 900 °C, 930 to 1650 °F) is the ob-
distinguishing factors: jective. But, sufficient time must be spent above
about 1000 °C (1830 °F) to promote the forma-
• The ferritic solidification mode of duplex tion of sufficient austenite. If the weld cannot be
stainless steels provides very good hot annealed, increased nickel filler metal (e.g.,
cracking resistance. The rapid cooling of 2209 with 2205 base metal) should be used.
welds produces welds and HAZ with more Thus, joint preparation must be done correctly
ferrite than the parent metal by quenching in and not left to the welder to correct using time-
the high-temperature ferrite. consuming remedial procedures.
Duplex stainless steels, because of their mod- • The stabilized group can lose toughness via
erate thermal expansion and higher thermal excessive grain growth.
conductivity, can tolerate relatively high heat • The grades more highly alloyed with
inputs since these factors determine the stress chromium and molybdenum can form α' and
intensity that will be generated by thermal gra- σ, leading to embrittlement.
dients. However, excessively low heat inputs
can result in fusion zones that are predomi- The semiferritic alloys such as 430, 434, and
nantly ferritic, with a resultant loss of toughness 436 are seldom welded and often called un-
and corrosion resistance. At the other extreme, weldable. The reason is that the welds are in-
heat inputs that are too high lead to the forma- variably partially martensitic and thus normally
tion of embrittling intermetallic phases. This brittle. Only very specially controlled composi-
issue concerns the HAZ, which must dwell in tions of 430 can be welded successfully, and
σ-forming temperatures for some period of these are not generally available commercially.
time. The key is to limit the time at those tem- While the technical remedy for this is simply
peratures by not permitting interpass tempera- annealing, it is seldom economically viable. It is
tures to exceed 150 °C (300 °F) because work- rare to see any welding more extensive than
piece temperature has the greatest influence on spot welding of unexposed surfaces with these
time at σ-forming temperatures. It is prudent to alloys. If for some reason they must be used and
impose this limitation when qualifying the weld welded, then the techniques for welding
procedure and then monitoring the production martensitic stainless steels should be employed.
welding interpass temperature electronically to The stabilized ferritic stainless steels are
ensure qualifying procedures are not more le- commonly welded. The levels of stabilizing ele-
nient than are those of production. ments required to prevent austenite formation
Postweld stress relief is not needed for duplex and sensitization are well known and are re-
weldments and indeed could be harmful be- flected in the alloys’ chemistry specifications.
cause of the danger of embrittlement. Full an- For 409, the required titanium level is Ti > 0.08
nealing can be done and can restore the original + 8(C + N), while the requirement for the higher
phase balance and composition that gives the chromium 439 is 0.20 + 4(C + N). These are
optimal toughness and corrosion resistance empirical relationships that take into account
found in wrought material. that some titanium oxidizes before it can stabi-
Ferritic stainless steels can be split into two lize carbon and nitrogen. Niobium can replace
groups for purposes of welding: the older semi- some titanium. This is discussed in detail in
ferritic group and the more prevalent stabilized Chapter 8 on ferritic stainless steels. Because of
ferritic group. The first group, in which the low toughness these alloys have in large
chromium is between 16 and 18% with carbon cross sections, these alloys are only rarely seen
up to 0.08%, is exemplified by the alloy 430. with minimum section size of more than 3 mm
These alloys form appreciable amounts of (0.11 in.) and normally have sections less than 2
austenite when heated above 800 °C (1470 °F). mm (0.08 in.). Thus, successful welding is sim-
Unless they are cooled extremely slowly, more plified to making a sound, well-shielded weld
slowly than can be done in welds, the austenite without producing excessive grain growth in the
transforms to martensite, which is very brittle. HAZ. In practice, this can be achieved by limit-
The stabilized grades commonly use titanium ing heat input to less than 6 kJ/cm. An empirical
or niobium to combine with the carbon and ni- relationship between grain diameter D and heat
trogen, which otherwise would cause the forma- input E (kJ/cm) has been reported (Ref 2). In
tion of the high-temperature austenite, render- the fusion zone, the relationship is:
ing the alloys ferritic at all temperatures. D = 206 × E – 585.6 (Eq 1)
The salient metallurgical characteristics for
welding of the two groups are: In the HAZ, it is:
• Both groups offer good thermal conductivity D = 29.6 × E – 50.6 for up to 6.6 kJ/cm (Eq 2)
and low thermal expansion.
• Both groups require protection from oxida- and
tion by shielding gases. The stabilized group D = 75 × E – 350 above 6.6 kJ/cm (Eq 3)
should not be exposed to nitrogen.
• The semiferritic group will form martensite, The light gauges ensure sufficiently short
which requires annealing to eliminate. times at high temperature that precipitation of
intermetallic phases should not be a concern, cleanliness and shielding. If mechanical re-
even though they can form, especially in super- quirements permit, the use of austenitic (309L)
ferritic alloys. weld filler metal should be considered. The soft
The impact properties of ferritic stainless joint may deform to accommodate thermal
steels are always a concern because their transi- strains and thus minimize weld cracking.
tion temperature can become elevated to ambi- Precipitation-Hardening Stainless Steels.
ent levels. It has been determined that there ex- Last, precipitation-hardening (PH) stainless
ists an optimum level of titanium around 0.10%, steels, while very complex metallurgically, are
which ensures this minimum transition temper- straightforward from a welding perspective.
ature (Ref 3). Because it is difficult to have low Obviously, any heat treatment to achieve the
enough carbon plus nitrogen to stabilize at this properties of which these alloys are capable
titanium level, dual stabilization with titanium must be a final step. The considerations in weld-
and niobium as well as not having excessive ing them are:
heat input are the best way to ensure weld • Shielding must be sufficient to prevent loss
toughness.
Especially in the superferritics, maintaining of oxidizable alloying elements such as tita-
the benefits of having the fairly precise balance nium, aluminum, and, of course, chromium.
• Filler metal must match the base metal if
of carbon plus nitrogen to the stabilizing ele-
ments titanium and niobium requires that nei- like properties are required.
• Postweld heat treatment solution annealing
ther carbon nor nitrogen come into contact with
the weld pool. Likewise, oxygen must be rigor- must be adequate to homogenize weld solid-
ously avoided because it will quickly deplete ification segregation.
• Austenitic PH grades are fully austenitic and
the essential titanium, which is even more read-
ily oxidized than chromium. Extraordinary sur- subject to hot short cracking.
• The high aluminum or titanium contents of
face cleaning at and near the weld will pay divi-
dends in final quality. many PH alloys cause their welds to be
Martensitic stainless steels vary little in alloy “slaggy,” and these slaggy welds have are ir-
content, ranging from 11 to 18% chromium with regular with objectionable recesses,
small amounts of nickel and molybdenum. crevices, or prominences.
Their carbon content ranges from 0.10 to over These alloys are easily welded and not prone
0.30%. Thus, the major challenge they present to cracking or developing embrittling phases.
is avoiding the potential cracking, which is But, because these alloys are designed for ex-
most likely to occur in the HAZ from stresses treme mechanical performance, it is essential to
caused by the austenite-to-martensite transfor- preserve their correct chemistry by shielding
mation on cooling. Since this transformation with a fully inert gas mixture. If mechanical
cannot be avoided, the desired approach is to properties equal to that of the base metal are not
start with a well-tempered or annealed material required in the weld, then austenitic filler, such
and then preheat and maintain high interpass as 309L, can be used.
temperatures. For low carbon levels, below Table 1 summarizes the major metallurgically
0.10%, preheat can be omitted, but between important parameters for the various types of
0.10 and 0.20% carbon, preheating to 250 °C stainless alloys. It is prudent to consult with the
(480 °F) is advised and for higher carbon levels, manufacturer’s data sheets for specific recom-
300 °C (570 °F). The problem becomes more mendations on alloys that they produce as they
severe with increasing carbon level because the are often privy to test data and user experience
transformation takes place at lower tempera- that cannot be found elsewhere in the literature.
tures in more brittle material. Even with pre-
heating, distortion may be encountered. For all
normal uses of martensitic stainless steels, a Material Selection and Performance
final heat treatment is required to achieve the
quenched and tempered properties for which Stainless alloys that are prone to precipitation
these alloys are designed. of intermetallic phases require special preweld-
Aside from the cracking consideration, ing consideration. Such alloys include duplex,
martensitic welding considerations are similar superferritic, and superaustenitic alloys. Any
to, but less stringent than, those of low-alloy amount of time for which these alloys have
stabilized ferritic stainless steels with regard to been exposed to temperatures at which inter-
Table 1 Welding parameters for various stainless steels

Heat input Shielding Interpass Postweld
Alloy group Filler kJ/cm(max) gas Preheat max heat treat
Austenitic ... 20–40 Ar+2% O2, 150 oC 150 oC None or full anneal
Ar/3% CO2/2% H2
He+7.5%Ar+2.5 CO2
301, 302, 304 308, 308L Same Same ... ... ...
304L 308L Same Same ... ... ...
309 309, 310 Same Same ... ... ...
310 310 Same Same ... ... ...
316L, 316Ti 316L, 317L Same Same ... ... ...
321, 347 347, 308L Same Same ... ... ...
Superaustenitic 22, 675, 276 16 Argon/helium or 50 oC 100 oC None or full anneal
argon + 3–5%
N2:no O2
PH grades Same as base alloy 20–40 Argon/helium no ... Full solution anneal
Martensitic
410 410, 308, 309L 20–40 Ar+2% O2, 250oC 250 oC min Slow cool
He+7.5%Ar+2.5 CO2
420 420, 308, 309L, 310 20–40 Ar+2% O2, 250 oC 250 oC min Anneal
He+7.5%Ar+2.5 CO2
440 440, 308, 309L, 310 20–40 Ar+2% O2, 250 oC ... ...
He+7.5%Ar+2.5 CO2
Supermartensiic Same as base metal 20–40 Argon/helium no ... Full solution anneal
Ferritic
430 430, 309L 20–40 Ar+2% O2, no ... Subcritical anneal
He+7.5%Ar+2.5 CO2
434 309 Mo L 20–40 Ar+2% O2, no ... Subcritical anneal
He+7.5%Ar+2.5 CO2
409 410L, 308, 309L 6.0 Ar+2% O2, no n.a. none
He+7.5%Ar+2.5 CO2
439 439L, 309L, 316L 6.0 Ar+2% O2, no n.a. none
He+7.5%Ar+2.5 CO2
Superferritic 29-4C 6.0 Argon/helium no n.a. None or full anneal
2003, 2101, 2304, 2209 5–25 Argon + 3% N2 no 150 oC None or full anneal
19-D
2205 2209 5–25 Argon + 3% N2 no 150 oC None or full anneal
25 Cr duplex 25Cr-10Ni-4Mo-N 5–25 Argon + 3% N2 no 150 oC None or full anneal

2507 25Cr-10Ni-4Mo-N 5-25 Argon + 3% N2 no 150 oC None or full anneal
superduplex
PH, precipitation hardenable
metallic phases form without full subsequent This effect exploits the temperature-dependent
homogenization anneal is time that the welder surface concentration of sulfur in the weld pool,
cannot use to complete a satisfactory weld be- which causes a decreased surface tension to-
fore precipitation occurs. Thus, accurate knowl- ward the hotter center of the pool, causing the
edge of material history is vital. Likewise, vari- molten pool to flow toward the center on the
ations within specification of nitrogen content surface and then flow downward, shooting the
influence the time it takes intermetallic phases hottest metal to the bottom of the weld pool, as
to form. Once a welding procedure is qualified shown in Fig. 4. This speeds welding and mini-
for an alloy with given nitrogen content, use of mizes weld and HAZ width, which is a good
lower nitrogen alloys would not be prudent en- thing. The effect on corrosion resistance is less
gineering practice. desirable since the abundant MnS inclusions
Austenitic stainless steels that are intended that result from the higher sulfur levels decrease
for autogenous welding are often specified with pitting resistance. This decrease in corrosion re-
elevated sulfur levels, on the order of 0.005 to sistance can only be eliminated by a long an-
0.015%. This is done to improve weld penetra- neal. Unfortunately, the pipe purchaser cannot
tion through the so-called Marangoni effect. know if the pipe has had a sufficient anneal.
90 194
85 185
Critical pitting temperature in 6% FeCI3, °C

80 Unwelded 176
Critical pitting temperature in 6% FeCI3, °F

75 Welded 167
70 158
65 149
60 140
55 131
50 122
Fig. 4 Metal flow directions in a weld pool with (left) and
45 113
without (right) sulfur. Source: Adapted from Ref 4
40 104
35 95
30 86
25 77
In-line induction annealing is insufficient for 68
20
this purpose. Furnace anneals of about an hour 15 59
are required. For alloys like 304L and 316L, the 10 50
user should always require material chemistry 5 41
0 32
certifications and assume that any sulfur levels −5 23
above 0.003% are going to result in decreased 1 2 3 4 5 6 7
pitting resistance of 1 to 5 PREN (pitting resist- Molybdenum, wt%
ance equivalent number), which means up to 10
°C (18 °F) decrease in critical pitting tempera- Fig. 5 The influence of molybdenum on critical pitting tem-
perature. Source: Adapted from Ref 5
ture, roughly the difference between 304 and
316 in performance. This also applies to girth
alter the composition of the welded joint, which
welds done by the pipe user.
in turn can alter corrosion and mechanical prop-
Welds are essentially a casting in the midst of
erties and compromise the entire structure.
wrought material. In addition to inclusions de-
Moisture, paint, dirt or grease, oil, and oxides
creasing weld corrosion resistance as men-
all can negate good material, good welding
tioned, solidification segregation can also cause
technique, and good procedural qualification.
microscopic regions to be poorer in corrosion-
Cutting fluids, especially sulfurized oils, are es-
resisting alloying elements chromium, molyb-
pecially detrimental and should be removed
denum, and nitrogen. This effect is minimal for
completely prior to welding. Preheating is never
low-alloy material, but for highly alloyed
strictly forbidden since it is required to elimi-
austenitic grades, it is a major effect, as shown
nate moisture.
in Fig. 5. Eliminating this effect requires a thor-
Joint design does not differ in principle from
ough homogenization anneal.
that of other steel weldments. There is, however
The use of filler metal with higher corrosion
an increased need for dimensional uniformity
resistance does not totally offset the influence of
for the alloys susceptible to intermetallic precip-
welding on corrosion resistance because some
itation since minimizing time at temperature is a
of the base metal melts and is not altered in
priority, and variations in joint geometry impede
composition by the filler metal. This is called
the swift completion of the weld. This is also
the unmixed zone. It is essentially a zone with
true for alloys that are susceptible to excessive
properties equal to that which would occur in an
grain growth, such as the stabilized ferritics, or
autogenous weld, that is, the corrosion resist-
to sensitization.
ance is lower depending on total alloy level and
Figure 6 shows some joint designs appropri-
sulfur content.
ate to stainless steels, including the more sensi-
tive alloys. These, like all joint designs, aim to
ensure full penetration without burn through.
Welding Processes Gas tungsten arc welding (GTAW)/tung-
sten inert gas (TIG) is commonly used for the
All stainless steels should be very clean prior automated production of stainless steel pipe and
to welding. The chemistries of both base metals tube, as well as manual short runs. It is versatile
and filler metals are carefully formulated to pro- and generally used when thicknesses are less
duce the mechanical and corrosion properties than 6 mm (0.2 in.). It can produce very high-
that these alloys have been designed to produce. quality welds. A constant-current power supply is
Virtually any contaminant can either interfere preferred. It is best performed with the DCSP (di-
with the welding procedure or detrimentally rect current straight polarity) electrode negative
Thickness Gap d, Root K,

Groove Process th, mm (in.) mm (in.) mm (in.) Bevel α(°)
GTAW 3–5 1–3 ... ...
t
GMAW 3–5 1–3 ... ...
d
SMAW 3–4 1–3 ... ...
a SMAW 4–15 1–3 1–2 55–65
GTAW 3–8 1–3 1–2 60–70
GMAW 5–12 1–3 1–2 60–70

d
SAW 9–12 0 5 60
a SMAW >10 1.5–3 1–3 55–65
k GMAW >10 1.5–3 1–3 60–70
d SAW >10 0 3–5 80
a
SMAW >25 1–3 1–3 10–15
k GMAW >25 1–3 1–3 10–15
d SAW >25 0 3–5 10–15

r = 6-8mm
GTAW >3 0–2 ... ...

t
GMAW >3 0–2 ... ...
d
SMAW >3 0–2 ... ...
a SMAW 3–15 2–3 1–2 60–70

GTAW 25–8 2–3 1–2 60–70
GMAW 3–12 2–3 1–2 60–70
d SAW 4–12 2–3 1–2 70–80
a SMAW 12–50 1–2 2–3 10–15

GTAW >8 1–2 1–2 10–15
GMAW >12 1–2 2–3 10–15
d
r = 6-8mm SAW >10 1–2 1–2 10–15
GMAW, gas metal arc welding; GTAW, gas tungsten arc welding; SAW, submerged arc welding; SMAW, shielded metal arc welding
Fig. 6 Joint designs. Courtesy Ugine S.A.
technique. It is helpful to incorporate a high- The shielding gas must replicate the con-
frequency circuit to aid in establishing the arc. trolled gas mixtures used to refine stainless steel
Thoriated electrodes containing 1.7 to 2.2% tho- and establish the original composition. The
ria are recommended because they have better weld pool exposes a great deal of surface area to
emissive properties and provide better arc stabil- the atmosphere in a very turbulent manner. Gas
ity at higher currents. If consumable electrodes flows, usually 12 to 18 L/min, must be adequate
are used, the shielding gas precludes the need for to prevent air infiltration by aspiration or turbu-
coatings. The weld metal alloys are not necessar- lence before arc contact, ideally until tempera-
ily the same as the parent alloys but are chosen tures cool to below oxidation temperatures.
based on their ability as weld metals to provide For manual GTAW using a filler wire, the
the most acceptable corrosion and mechanical wire should be fed continuously into the weld
properties. This sometimes means using pool. Intermittent wire addition can lead to cre-
austenitic filler with a ferritic base or higher ation of zones of essentially autogenous weld,
nickel content in an austenitic or duplex base to negating many of the benefits of filler metal ad-
compensate for the solidification rate or inher- dition. Moving the tip of the wire in and out of
ently lower corrosion resistance of the weld. the protection of the gas shield is especially
bad. The hot tip can carry oxides and nitrides for GMAW, but the consumable electrode, the
into the weld, defeating the action of the shield FCW filler metal, has a flux core that supple-
gas and impairing weld quality. ments the shielding gas. Because of the flux, the
Gas metal arc welding (GMAW) is arc shielding requirements are reduced; gases can
welding in which a consumable electrode pro- be argon/25% carbon dioxide for horizontal
vides larger amounts of filler weld metal than welding with current and voltages from 150 to
practical in GTAW. There are three GMAW 200 amp and 22 to 38 V, respectively. Vertical
techniques: welds can use 100% carbon dioxide with am-
• Pulsed arc transfer perage of 60 to 110 amp and voltage of 20 to 24
• Spray transfer V. Flow rates of gas are 20 to 25 L/min. It is
• Short-circuiting transfer possible to get high-carbon welds, which may
not resist corrosion as well as desired, so as al-
Pulsed arc transfer employs a power source ways, weld qualification, including corrosion
that is switched rapidly to provide transfer of evaluation, is critical.
weld metal droplets at regular intervals. Spray Oxyfuel gas welding (OFW), “torch” weld-
transfer uses a high current to form a stream of ing, uses oxygen to accelerate fuel (typically
fine drops from the end of the electrode. This is acetylene) combustion to produce temperatures
done with high power, resulting in a large fluid that can melt steels. By controlling the fuel-air
weld pool, and therefore limits the technique to mixture, the flame can be made nonoxidizing
horizontal orientations and thick material. for low-alloy steels. However, these “neutral”
Short-circuiting transfer uses arc contact with flames can simultaneously oxidize and carbur-
the workpiece at low power to melt the elec- ize stainless steels. Thus, the OFW process is
trode, after which the short circuit is broken, not suitable for use with stainless steels.
and material transfer ceases. The technique cre- Laser welding has become a major produc-
ates a minimal weld pool and is viable in many tion method when it can be automated, as for
orientations. It is a low-heat process suitable for pipe and tube or high-production manufactured
thin material but may cause lack of penetration items, such as air-bag canisters. Metallurgically,
defects if used for thick-section welding. it resembles resistance welding in that both
For all GMAW processes, excessive protru- have minimal HAZ and very high quenching
sion of the wire should be avoided; otherwise, rates, both of which can have a pronounced ef-
the full benefit of the inert gas shielding may be fect on some types of stainless steel. The effect
lost. is to undercool the molten metal and suppress
Submerged arc welding (SAW) employs a the transformation that would normally occur.
consumable electrode immersed in a conductive So, an austenitic alloy that normally solidifies in
flux that acts as a protective shield from the at- a ferritic mode before transforming to austenite
mosphere. The arc is struck through the flux, freezes directly as austenite. The freezing is so
and gravity deposits the molten metal to the rapid that the normal hot shortness of austenitic
workpiece. The large weld pool has high heat solidification is avoided, so quality is not com-
input and can deposit large amounts of metal promised. In fact, laser welds quench the mate-
relatively quickly. Thus, SAW may be prefer- rial so rapidly that corrosion resistance is en-
able to multipass techniques for alloys such as hanced since inclusions cannot nucleate and
duplex for which time at temperature is limited. grow. Duplex alloys, on the other hand, freeze
It is restricted to horizontal orientations and rein their high-temperature ferrite structure be-
quires postweld slag (flux) removal. cause the fast quench prevents the nucleation
Shielded metal arc welding (SMAW) is and growth of austenite. Unless this ferrite is
done manually with short lengths (“sticks”) of heated to permit austenite to form, lower-tough-
coated electrodes. This method has great versa- ness welds will result. Ferritic, martensitic, and
tility with some trade-off in cost and quality. PH alloys are not harmed by the rapid quench.
This last aspect is arguable, but the lack of Resistance welding is readily done on most
shielding gas may introduce oxygen to the weld types of stainless steel. Allowance must be made
metal, which can be detrimental to toughness. for the lower thermal and electrical conductivity
Flux cored wire (FCW) welding is a method of stainless steels compared to other common
that is able to accommodate a large range of materials. Most resistance welds, including both
thickness and orientations while providing high seam and spot welds, have deep, tight crevices
deposition rates. The equipment is the same as adjacent to the welds. The possibility of crevice
corrosion in these regions should be considered (Chapter 13) of this book. Brazes and solders
when contemplating the use of spot welds in rarely match the corrosion resistance of stain-
stainless materials. The possibility of entrap- less steels, and careful attention should be given
ment of foreign material and the difficulty of re- to the potential for galvanic and other forms of
moving it from such crevices should also be corrosion when considering the use of soldered
considered, especially in equipment for food or brazed joints with stainless steels.
handling, pharmaceutical production, etc.
High-frequency induction welding of
stainless steel is more difficult than for low-
alloy steel because of the refractory nature of Welding Practices
chromium oxide, which has a higher melting
temperature than does the stainless base metal. Safety must always be considered when
This is opposite from the situation in low-alloy welding. In addition to the normal hazards
steels, for which the iron oxide melts at a lower (which are not discussed here) associated with
temperature than does the iron base metal. The welding, welding of stainless steels presents a
presence of this refractory oxide on the surfaces special hazard: hexavalent chromium. The fume
to be joined makes it more difficult to obtain a created by welding stainless steel contains sig-
defect-free weld. nificant concentrations of chromium trioxide
Thermal cutting of stainless steels is rou- and other forms of hexavalent (Cr+6) chromium.
tinely practiced, but the processes and parame- Hexavalent chromium is a carcinogen and regu-
ters used are determined by the refractory na- lated by the Occupational Safety and Health
ture of the chromium oxides that form on Administration (OSHA). Exposure to and in-
stainless steels. The high temperatures attain- halation of stainless steel welding fumes must
able with lasers or plasma arc torches provide be avoided. The product exposure limit for
good cutting action, and these processes are fre- hexavalent chromium is 5 μg/m3 as of Decem-
quently used. To expand the range of thick- ber 31, 2008. Refer to OSHA for further updates
nesses that can be cut or to increase cutting on this limit. Use of fume extraction equipment
speed, supplemental oxygen or nitrogen blast is generally the preferred method of minimizing
jets may be used. Stainless steels may also be hexavalent chromium exposures. Positioning
cut using oxyfuel equipment if supplemental and operation of the fume extraction device
iron powder is used. Combustion of the iron in- must be done precisely to ensure effective fume
creases the temperature, while the iron oxide removal while avoiding excess turbulence,
helps flux the refractory chromium oxide. Ther- which can cause loss of effective inert gas
mally cut edges of stainless steel usually require shielding of the weld pool. Thermal cutting of
subsequent cleaning, typically by grinding or stainless steels also generates hexavalent
milling. Chemical cleaning of all surfaces of cut chromium, and similar procedures are required
pieces to remove heat tint, fume deposits, and to minimize exposure during such operations.
other contaminants is advisable. Nondestructive Evaluation (NDE) is used
Soldering and brazing are possible with all almost universally to ensure weld quality. All of
stainless steels. Soldering is done below 450 °C the standard NDE techniques used with other
(840 °F), while brazing is done above 450 °C materials are applicable to stainless steel weld-
(840 °F). Solders are generally alloys of tin and ments. Allowance must be made for the differ-
bismuth, lead, silver, or antimony or combina- ing physical properties of stainless steels, and
tions of several of these. Brazes are normally ei- appropriate reference defect standards must be
ther silver based or nickel based. The provided. However, one technique—magnetic
chromium-rich oxide coating must be removed particle inspection—is problematic. The pres-
by a suitable flux for bonding to occur. Fluxes ence of bands of persistent austenite in marten-
are typically acid type with chlorides. Thus, sitic or PH stainless steels can lead to spurious
after the soldering or brazing, the flux must be defect indications. For this reason, magnetic
thoroughly removed to prevent subsequent pit- particle examination of stainless steel welds is
ting corrosion. Brazing temperatures must be best avoided.
chosen to avoid ranges at which unfavorable Recent developments in stainless steel have
phases form. The best range can be determined been made with weldability as a major consid-
from examining temperature ranges to be eration. Highly alloyed, low-carbon martensitic
avoided in the thermal processing chapter alloys for line pipe have been developed with
the express purpose of use in the as-welded con- generation of weld fume (see the discussion of
dition. The low carbon makes welds of this ma- safety). Other new welding processes, such as
terial that are tough and do not require temper- multiple (GTA or GMA) torch welding, laser-
ing, so girth welds in the field are possible. assisted GMA or GTA welding, etc. promise
Likewise, the lean duplex alloys have very greater productivity.
delayed precipitation of intermetallic phases be-
cause of their higher nitrogen and lower
chromium and molybdenum contents. This REFERENCES
makes welding of these alloys much more fool-
proof than with the early duplex alloys, such as 1. D.J. Kotecki, Welding of Stainless Steels,
S31803. The dual-stabilized ferritic alloys have Welding, Brazing, and Soldering, Vol 6,
tougher welds than those stabilized with only ti- ASM Handbook, ASM International, 1993,
tanium or niobium. p 677–707
New developments in welding also have an 2. B. Aziez and R. Feen, Sheet Metal Ind., 1,
impact on stainless steels. The friction stir weld- 1983, p 28–34
ing (FSW) process offers the promise of reliable 3. S.D. Washko and J.F. Grubb, Proc. Int’l
solid-state joining. By avoiding melting and Conf on Stainless Steel, 1991, Chiba, ISIJ
resolidification, issues associated with solute 4. Stainless Steels, Les Editions de Physiques,
redistribution are eliminated. The relatively low 1992, p 786
temperatures involved essentially eliminate 5. A. Garner, Corrosion, 37, 1981, p 178
CHAPTER 18
Architecture and Construction
Summary can choose from a variety of stainless alloys

with sufficient corrosion resistance to withstand
STAINLESS STEEL IS OFTEN EMPLOYED any environment.
as an architectural material; the material can it- The ability of stainless steel to resist corro-
self be viewed as a metaphor for architecture, a sion resides in its chromium-rich superficial
discipline that must balance aesthetics, economy, passive layer. Stainless steel by definition must
and structural integrity. Stainless steel’s unique contain slightly more than 10% Cr. The passive
combination of beauty, strength, and economy layer forms spontaneously in air or water, and
makes it a remarkably appropriate material for if it is removed, say by abrasion, it re-forms by
uses as diverse as sculpture and concrete rein- itself. This is explained in greater technical
forcing bar. However, stainless steels are com-
plex; they come in many different grades (chemi-
cal analyses), and these grades have varying
strengths, appearance, resistance to corrosion,
availability, and costs. The success of a building
project involves careful planning for the
appropriate use of materials. This chapter deals
with the technology of stainless steel as it per-
tains to its proper use in architecture, art, and
construction.
Corrosion Resistance
Corrosion is the life-limiting factor for archi-

tectural metals. Steel, copper, aluminum, lead,
bronze, and other alloys react with the environ-
ment and degrade over time, as does wood,
stone, plastic, paint, and even glass. With stain-
less steel, it is possible to choose a material that
can withstand attack from the environment in-
definitely. One of the first major uses of stain-
less steel in architecture was in New York’s
Chrysler building, completed in 1930. Despite
the rather crude methods of early production,
limited alloy options, and lack of application
experience involved in its construction, the
domed top of the Chrysler building still shines
undiminished by the harsh coastal and urban Fig. 1 The Chrysler building with its famous bright stainless
details. Copyright © iStockphoto.com/Steven Allen.
climate (Fig. 1). With today’s technology, one Used with permission
detail in this book in the chapters on corrosion, Balancing Corrosion Resistance,

but the key aspects are that the chromium Processing Characteristics, and
atoms on the surface of the metal react with
Economy
oxygen in air and water to form with neighbor-
ing iron atoms into a tight, ionically noncon-
The rule of thumb for grade selection is usu-
ductive layer that prevents any further oxygen
ally somewhat oversimplified to recommend the
penetration. This layer is mere atoms thick and
use of types 430 and 304 on interior applica-
completely invisible.
tions, type 304 on exteriors where salt is not a
The strength of the layer in resisting corrosive
problem, and 316 where road salts or seacoast
attack is proportional to several key alloying
effects make a more corrosion-resistant grade
elements: chromium, molybdenum, and nitro-
necessary. A leading architectural metals com-
gen. They contribute according to the following
pany (Ref 1) makes the following recommenda-
formula:
tions, which mirror these traditional views:
PREN = %Cr + 3.3(%Mo) + 16(%N) (Eq 1) • Type 304 should be used for most exterior
applications.
PREN stands for pitting resistance equivalent • Type 316 should be used within ten miles of
number. This number can be related to resist- saltwater bodies. However, if the building is
ance to the mildest form of corrosive attack that subject to saltwater spray, a nobler grade of
stainless steel undergoes, pitting corrosion. stainless steel, such as 2101 or 2003, should
Pitting is a “weakest link” phenomenon in be specified.
which corrosion begins in small, micron-size • In close proximity to deicing salt use, even
parts of the surface and then grows by virtue of on nearby roadways where vehicle traffic
the more aggressive media that form within can create airborne particles, type 316
them because of the corrosion reaction’s prod- should be used. If periodic rinsing will not
ucts. Pitting occurs in environments that contain occur on all exterior surfaces, these areas
chlorides. Chloride ions compete with oxygen must be washed each spring. If dependable
and disrupt the integrity of the protective pas- maintenance is not predicted, a nobler grade
sive layer. As alloys become richer in the alloy- of stainless steel, such as 2101 or 2003,
ing elements mentioned, their ability to maintain should be used.
the passive layer can overcome the chlorides’ • Specify types with low carbon, less than
ability to destroy it. The key is to choose an 0.030%, if welding will be employed.
alloy rich enough in chromium, molybdenum, • Any grade, including type 430, may be used
and nitrogen to withstand any environment the in interior applications.
structure will experience. Pitting corrosion, and • In the most severe environments—high heat
another similar form of corrosion called crevice and humidity, low rainfall, and high salinity,
corrosion, can be prevented by proper choice of such as are found in Middle Eastern coun-
alloy, finish, and design. Crevice corrosion oc- tries—a grade with a PREN of 25 or above
curs when recessed spaces are small enough to is recommended.
act like a corrosion pit. The acidity within the
crevice increases because of restricted diffusion These guidelines are based on the admittedly
in and out of the crevice, just as happens within easy availability of these alloys and a lack of
a pit. The buildup of iron and chloride ions concern for cost during times of peak raw mate-
makes a very corrosive medium that disables rial prices. For projects where quick availability
the passive film formation. A design that avoids is not more important than cost, 439 and 201
crevices is the best defense. should be considered as viable replacements for
The decision criterion for material selection 304. Stainless steel 2003 (UNS S32003) or an
with stainless steel should then be, Which grade equivalent lean-duplex grade can replace 316 at
and finish will exclude the possibility of pitting a cost advantage during times of high alloy cost
corrosion at the lowest cost? Then, proper de-
sign should be used to exclude the possibility of Table 1 Ranking of common stainless steels by
crevice corrosion. Other factors involving pitting resistance equivalent number (PREN)
strength and fabrication should also be consid- Alloy 430 439 201 304 316 2101 2003 2205 2507
PREN 15 17 17 19 24 26 28 35 38
ered. Table 1 ranks a number of stainless alloys
Higher PREN values indicate greater pitting resistance.
by pitting resistance.
Chapter 18: Architecture and Construction / 215
as have been experienced on occasion, such as resistance in abraded stainless steel surfaces is
during the period of 2004 to 2007. It should be not seen on finishes that are produced by pat-
noted that the leaner alloys suggested (439, 201, terns imprinted by hard-rolling mill rolls that
and lean-duplex alloys such as 2003) can be have been engraved with the desired pattern in
somewhat more difficult to form. If panel de- reverse; for this reason alone, this method of
signs call for 90° bends, this is not an issue. surface finishing is recommended. These archi-
However, for applications requiring severe tecturally useful surface finishes are produced
forming, as in the case of a double-lock seam on by the preferred rolled-on, or embossed,
a standing seam roof, these grades can provide a method. In addition to their advantage in corro-
challenge to the fabricator/installer. Further, sion resistance, they are extremely uniform
these leaner grades can pose challenges with from batch to batch, unlike finishes produced by
certain finishing methods, such as abrasive pol- abrasive-coated belts, which change in grit
ishing and embossing. To avoid unwanted com- coarseness with use.
plications related to grade selection, the speci-
fier should consult a competent architectural
metals supplier. This effort will ensure a viable Balancing Service Environment,
specification is written that will balance cost Design Requirements, and
with the necessary performance attributes to
Maintenance Considerations
make the part as well as resist corrosion once
installed.
An expert system has been developed that en-
It is valuable to know, in times of high nickel
ables designers and specifiers to analyze the
prices, that both the low-nickel 201 and the no-
trade-offs of climate, design requirements, and
nickel 439 can be used in place of 304, while
maintenance on grade selection (Ref 3). An-
2101 (UNS S32101) and 2003 (lean duplexes)
swering the questions in Fig. 3 for a particular
can replace 316. To obtain these grades usually
application yields a score that can be used to
involves working with a producing mill since
identify an appropriate alloy according to the
they are not typically stocked in service center
scale shown Fig. 4. The Nickel Institute, for-
inventories. However, any competent architec-
merly the Nickel Development Institute, also
tural metals supplier will not shy away from the
offers excellent publications on topics related to
use of specialty grades where appropriate.
alloy selection for specific service environments
and design requirements (Ref 4).
Reviewing a map of the salinity of rainwater
Surface Finish and Corrosion Resistance in the United States is instructive of the degree
to which geography influences corrosion sever-
Surface finish is usually an aesthetic choice, ity. The average atmospheric chloride levels
but it has a significant influence on corrosion re- collected in rainwater are shown in Fig. 5 (Ref
sistance and must be factored into grade selec- 5). The highest levels occur along the coastlines
tion. Mill finishes such as 2B and 2D are incon- of the Atlantic and Pacific Oceans and the Gulf
sistent because they are annealed and pickled to
remove oxides. These unattractive surfaces,
however, have correct corrosion resistance for
their alloy content. Welding or abrading the sur-
face degrades the corrosion resistance by a sig-
nificant amount. An un-heat-treated weld has
lower resistance to corrosion in proportion to
the alloy content of the grade. Type 316 welds
have the corrosion resistance of wrought 304.
Abrasion has a similar effect. Type 316 with a
No. 4 polish behaves like 304 with a 2B mill
finish. Because welds are abraded, this com-
pounds the effect.
Very smooth abrasively polished finishes mit-
igate this reduced corrosion resistance, as Fig. 2 The decrease in corrosion resistance with increasing
shown in Fig. 2. The effect of reduced corrosion surface roughness by abrasion. Source: Ref 2
Fig. 3 Stainless steel selection expert system. Source: International Molybdenum Association (Ref 3)
Fig. 4 Grades recommended based on the expert system. Source: International Molybdenum Association (Ref 3)
Fig. 5 Average chloride concentration (mg/L) in rainwater in the United States. Source: Ref 5
of Mexico. The maximum corrosion rate is re- erly maintained, will stay new looking indefi-
lated to the maximum chloride in the atmos- nitely.
phere. This will be related to the distance in- Surface finish aesthetics are arguably more
land, the height above sea level, and the important architecturally than the influence sur-
prevailing winds (Ref 6). face finishes have on corrosion. Numerous fin-
ishes have been developed to try to meet vari-
ous objectives. Finishes vary in reflectivity,
Aesthetic Considerations directionality, and subtlety. Figure 6 shows
some of the finishes that go beyond the familiar
A correctly chosen grade of stainless steel brushed look (Ref 7), while Fig. 7 shows special
will have no degradation over time and, if prop- finishes created by one manufacturer.
Fig. 6 Various rolled-on stainless steel finishes. Source: Ref 7. Courtesy of Outokumpu
Fig. 7 Special finishes for 304/304L and 316/316L stainless steels available from one manufacturer. (a) Rolled-in low-glare finish
(InvariMatte). (b) Rolled-in no. 4 finish (InvariBlend). (c) Rolled-in moderate-glare finish (InvariLux). Source: Contrarian Metal
Resources (Ref 8)
The classic abrasively produced finishes are “Standard Test Method for Specular Gloss”).
No. 3, 4, and 8. These are American Society for The standards for appearance do not exist
Testing and Materials (ASTM) designations for within specifications, only the method of pro-
abrasively produced finishes, which are tradi- ducing them. There is considerable difference
tionally produced by abrading the surface with in appearance from sheet to sheet, coil to coil,
different grit size abrasives. Finish No. 3 calls and manufacturer to manufacturer. The greatest
for 80 to 100 grit abrasive; No. 4 calls for 120 to consistency of appearance comes from specify-
150 grit abrasive. Finish No. 8 is a mirror finish ing a brand of rolled-on finish from a given
obtained by final polishing with 800 grit abra- manufacturer. Any of the traditional finishes
sive. Finishes No. 3 and 4 are directional, with can be replicated by a rolled-on finish with
grit lines typically 1 cm (0.4 in.) in length. Fin- greater uniformity, with the possible exception
ish No. 3 has a surface roughness average (Ra) of bright annealed having a difficult time match-
of 0.4 to 0.8 µm (15 to 30 µin.). Specular gloss ing the mirror quality of a No. 8 finish. This is
at 85° is typically 40 to 60 (per ASTM D 523, crucial in architecture, where the discovery of
unacceptable visual nonuniformity on large Because stainless in sheet form is usually re-
areas can be disastrous, especially when this ap- flective, small deviations from flatness can be
pears late in the construction process, as is nor- very visible. A good standard for flatness that
mally the case with exterior components. precludes visible distortion is five I units. Steel
Reflectivity or gloss can be a major consider- producers have various means to produce this
ation in the choice of a surface finish. Mirror level of flatness, the most extreme of which is
finishes are often used for high impact, but actually stretching the steel sheet or coil until
more diffusely reflecting surfaces are more all distortions are eliminated. Sometimes,
common. Patterned surfaces provide consistent rather than aiming for high flatness a controlled
reflectivity from a moderately reflective 40 to deviation from flatness is used, such as slightly
60 specular gloss at 85° to a dull matte of less concave panels or panels with a die-pressed
than 20, the latter having been developed for design. Another option to ensure flatness is to
airport roofing, such as at Reagan Airport in back light-gauge stainless steel with a stiff
Washington, D.C., or the Pittsburgh Convention material.
Center (Fig. 8). Deviations of sheets from squareness and
Flatness is a special consideration for panels straightness (camber) are also objectionable be-
where lack of flatness, such as by “oil-canning,” cause such deviations can cause gaps between
can cause a very shoddy appearance. Flatness is panels. The degree to which this is objection-
measured in I units. able is a function of design, and tolerances can
be held tightly at a cost. Width tolerance is
Flatness (Iunits) = 2 (πH / 2L) × 105 (Eq 2) normally +1/16 in./–0 in 48 in., while length is
where H is the height of the deviation from flat- held to +1/8 in./–0 in 10 ft or less. Maximum
ness, and L is the distance between peaks of de- camber is 3/32 in. in 8 ft. Closer tolerances can
viations, assuming a sinusoidal wave. be negotiated.
Fig. 8 The Pittsburgh Convention Center with low-gloss finish stainless steel roof
Maintenance and Repair polymer coatings are applied to stainless steel at

some producing mills to permanently provide a
Maintenance is a significant cost of any struc- film to which additional fingerprint oil cannot
ture. One of the great values of stainless steel is add a noticeable discoloration.
its low cost of ownership. Stainless steel can be The greatest ally of stainless on building exte-
abused, however, and it does benefit from riors is the cleansing action of rain. Rain does
proper maintenance. The main objective of the not completely clean the surface, but it does di-
maintenance of stainless is keeping it clean. lute any harmful contaminants and forestall cor-
There are two reasons for this. The obvious first rosion from accumulated chlorides. Without the
reason is that whatever is soiling the surface is benefit of cleansing by rain, stainless exteriors
probably not attractive. The second reason is should be washed during routine window wash-
that it may harm the surface by allowing corro- ing operations.
sion agents to concentrate. Table 2 provides rec- Given the importance of cleaning, the follow-
ommended practices for removing various sub- ing design considerations are recommended:
stances from stainless steel surfaces (Ref 9).
One of the most common complaints about • Designs that can collect dirt, such as hori-
maintenance of stainless is the work involved in zontal surfaces and recesses, should be
removing fingerprints. The oil from fingerprints avoided.
makes an easily visible interference film on the • Designs that create uneven flow or drainage
reflective stainless surface. The typical remedy patterns producing uncleansed areas should
is to clean stainless steel with a solution con- be avoided.
taining light oil and a detergent. If the oil from a • Sheltered areas and areas subject to splatter,
hand contacts the uniformly thin film of clean- especially roadside spatter, should be de-
ing oil, no visible mark is left. Alternatively, signed so that they are easily cleanable.
Table 2 Cleaning methods for uncoated stainless steel

Requirement Suggested method(a) Comments
Routine cleaning of light soiling Soap, detergent, or dilute (1%) ammonia Satisfactory on most surfaces
solution in warm clean water. Apply with a
clean sponge, soft cloth, or soft-fiber brush,
then rinse in clean water and dry.
Fingerprints Detergent and warm water; alternatively, Proprietary spray-applied polishes available to
hydrocarbon solvent clean and minimize re-marking
Oil and grease marks Hydrocarbon solvent Alkaline formulations are also available with
surfactant additions.
Stubborn spots, stains, and light Mild, nonscratching creams and polishes. Avoid cleaning pastes with abrasive additions.
discoloration; water marking; Apply with soft cloth or soft sponge; rinse Cream cleaners are available with soft
light rust staining off residues with clean water and dry. calcium carbonate additions. Avoid chloride-
containing solutions.
Localized rust stains caused by carbon Proprietary gels or 10% phosphoric acid Small areas may be treated with a rubbing
steel contamination solution (followed by ammonia and water block comprising fine abrasive in a hard rub-
rinses) or oxalic acid solution (followed by ber or plastic filler. Carbon steel wool and
water rinses) pads that have previously been used on
carbon steel should not be used. A test
should be carried out to ensure that the
original surface finish is not damaged.
Adherent hard water scales and mortar/ 10–15 vol% solution of phosphoric acid. Proprietary formulations available with
cement splashes Use warm, neutralize with diluted ammonia surfactant additions. Avoid the use of
solution, rinse with clean water and dry hydrochloric acid-based mortar removers.
Heat tinting or heavy discoloration Nonscratching cream or polish. Apply Suitable for most finishes.
with soft cloth or soft sponge. Rinse off
residues with clear water and dry.
Nylon-type pad Use on brushed and polished finishes along the
grain.
Badly neglected surfaces with hardened A fine abrasive paste as used for car body May brighten dull finishes. To avoid a patchy
accumulated grime deposits refinishing. Rinse clean to remove all paste appearance, the whole surface may need to
material and dry. be treated.
Paint, graffiti Proprietary solutions or solvent paint stripper Apply as directed by manufacturer
depending on paint type. Use soft, nylon or
bristle brush on pretreated material.
(a) Cleaning agents should be approved for use under the relevant national environmental regulations and should be prepared and used in accordance with the com-
pany’s or supplier’s health and safety instructions. Source: Adapted from Ref 9
• Contamination by rust from carbon steel is • Separation of tools and work areas between
corrosive and must be removed. those used for stainless steel and carbon
• Designs should facilitate easy access for steel is prudent. Contamination of stainless
cleaning. surfaces with carbon steel from welding,
grinding, and cutting can stain the surface of
While any structure would benefit from these stainless steel and result in corrosion. This
guidelines, they are especially valuable in maxi- can be remedied by passivation, but it is
mizing the benefits of stainless steel. much better to avoid it in the first place.
Repair of more severe damage done to sur- • Welding is better done in the shop than in
faces, such as scratching, is difficult. If a deco- the field. Correct filler metals must be used,
rative surface pattern has been damaged, the and proper weld finishing is essential. The
challenge is in trying to replicate it in the field. ability of contactors to produce sound, at-
Very few surface finishes are wholly repairable tractive welds is an indicator of their overall
in the sense that they can be repaired in a spot competence with stainless steel.
so that the repair is invisible. The reason for • Fasteners used with stainless steel should
that is mainly that abrasive finishes are applied also be stainless steel. Galvanized steel, car-
by rotating belts, and it is virtually impossible bon steel, and aluminum will corrode more
to match the pressure, grit size, and arc of con- readily than the stainless, and this corrosion
tact that created the original surface. It can be is aggravated by galvanic contact. The re-
more nearly done to a surface with a rolled-on sulting corrosion products are also harmful
finish, which has a very consistent grit length as well as unsightly. Fasteners should not be
and depth. The exception to this rule are abra- permitted to cause distortion of flat panels.
sively applied or rolled-on long-grain finishes.
These have very long grit lines, so grit length
is easily duplicated with a belt sander, the Additional Service Considerations
usual tool available for field repairs. Even
welds can be removed and reblended to be in- Fire resistance is an important consideration
distinguishable from the surrounding original in buildings. Stainless steel is the only com-
surface. The ability to be repaired should be a mon building material that remains strong and
top criterion in the choice of a surface finish tough at temperatures encountered in fires. Or-
whose appearance is critical and that may be dinary carbon steel undergoes a phase change
subject to damage. at about 760 °C (1400 °F). This change in
atomic structure results in a sudden shrinkage
of more than 1 linear percent. This can literally
pull a building apart. When this occurs to a
Fabrication Considerations structure already weakened by heat, as carbon
steel is, catastrophic failure ensues. Austenitic
Fabrication and joining of stainless steel em- stainless steel keeps the same atomic structure
ploy the same techniques as for carbon steel and and remains much stronger than carbon steel at
other metals. The specifics of cutting, forming, elevated temperatures. Thus, austenitic stain-
joining, soldering, and welding are described in less steel has great value as a material for
the processing section of this book and are not structures that must retain structural integrity
repeated here. The main distinction in the use of in a fire.
stainless steel in this regard is that its higher Tests have been conducted on glass-rein-
strength and corrosion resistance permit the use forced plastic, aluminum, galvanized steel, and
of lighter gauges. This in turn permits designs in austenitic stainless steel ladders under load and
hollow or rolled-formed sections, which have exposed to flame temperatures of more than
higher stiffness and low weight and potentially 1000 °C (1830 °F). The plastic and aluminum
lower overall cost than using less-expensive ladders failed in less than a minute. The galva-
metals. A second aspect of higher strength and nized carbon steel lasted 5 min, while the stain-
lighter gauge is greater spring back in forming less steel remained intact (Ref 10). If the need
operations, such as press braking. for fire resistance is serious, stainless steel
Certain processing principles related to archi- becomes the material of choice. It is used on
tectural and building applications of stainless offshore oil platforms for stairways, ladders,
steels should be emphasized: walkways, handrails, gratings, floor systems,
Fig. 9 Graphic depicting low release of metal ions from two grades of stainless steel (304 and 316) to the environment, based on a
4-yr multidisciplinary research project involving both field research and laboratory studies. Source: Ref 11
firewalls, blast walls, living modules, and so In bridges, parking garages, and other concrete
forth. structures, saltwater can penetrate the cement
Ecological considerations are never trivial over time. If the internal rebar corrodes, the
when considering a construction material. expansion of the corrosion products spalls
Many materials used in buildings degrade envi- the concrete, leading to the failure of the
ronmentally, usually by corrosion, and enter the structure.
general environment. Asbestos, lead-based This can be delayed by treating the concrete
paints, lead coatings, and others are once-ac- to repel the incursion of water or by coating
cepted materials whose long-term effects have the carbon steel rebar with epoxy. These are
been dangerous and costly. less than 100% effective. A more certain ap-
Stainless steel, because it does not corrode proach is to use stainless steel rebar. The stain-
when properly used, does not enter the environ- less steel for this duty need not resist pitting
ment. While this seems obvious, it has been the corrosion, which affects a tiny percentage of
object of interdisciplinary studies that have the steel volume. Therefore, an inexpensive,
demonstrated its innocuousness even under low-nickel grade, such as 409, 430, or 201, can
conditions of heavy acid rain on freshly abraded be used. Most of the work to date has been
surfaces (Fig. 9) (Ref 11). with more expensive grades, such as 316 and
Stainless is a material that will never come 2205. The lean duplexes are ideal for this ap-
back to haunt an architect years later. Its intrin- plication because of their high strength, resist-
sic raw material content value ensures that even ance to corrosion and SCC, and moderate cost.
with the end of the life of a structure, the stain- The use of even these alloys reduces the long-
less in the structure will be recycled. term cost of these structures, so the future
adoption of less-expensive stainless steels
holds great promise. The more enlightened
Concrete Reinforcing Bar transportation departments in the United King-
dom; Ontario, Canada; and Michigan, New
Concrete reinforcing bar is one of the least Jersey, and Oregon in the United States have
glamorous uses of stainless steel in structures. led this development.
REFERENCES dustries, Vol 13C, ASM Handbook, ASM

International, 2006, p 42–60
1. Stainless Steel Selection Criteria, Contrar- 7. Guide to Stainless Steel Finishes, 3rd ed.,
ian Metal Resources, www.metalresources. Euro Inox, 2005, www.euro-inox.org, ac-
net, accessed June 2008. cessed June 2008
2. Bulletin of the National Dairy Federation 8. Contrarian Metal Resources, www.metalre-
189, 1985, p 3–12 sources.net, accessed June 2008
3. C. Houska, “Which Stainless Steel Should 9. Care and Maintenance of Stainless Steels,
Be Specified for Exterior Applications?” In- Leda-Vannaclip, www.l-v.com.au, accessed
ternational Molybdenum Association, www. June 2008
imoa.info, accessed June 2008 10. The Nickel Institute, www.nickelinstitute.
4. The Nickel Institute, www.nickelinstitute. org, accessed June 2008
org, accessed June 2008 11. D. Berggren et al., Release of Chromium,
5. H. Guttman, Atmospheric and Weathering Nickel and Iron from Stainless Steel Ex-
Factors in Corrosion Testing, Atmospheric posed Under Atmospheric Conditions and
Corrosion, W.H. Ailor, Ed., John Wiley and the Environmental Interaction of These
Sons, 1982, p 51 Metals, European Confederation of Iron and
6. R.B. Griffin, Corrosion in Marine Atmos- Steel Industries, Oct 2004, www.eurofer.
pheres, Corrosion: Environments and In- org, accessed June 2008
CHAPTER 19
Automotive and Transportation

Applications
Summary the auto manufacturers have become the

largest users of stainless steel.
THE ADEQUATE DURABILITY and life Exhaust systems constitute the largest use of
span of cars, trucks, or any transport system stainless steel in the automotive market, but
requires freedom from corrosion. This has re- there are other important applications that can-
quired subsystems, such as those for exhaust not be ignored: valves and gaskets, hose
and fuel, to resist more corrosive environments clamps, seat belt and air bag components, tub-
for longer periods of time. The main result has ing, hardware, and filters. And, there will be
been a strong growth in the use of the leaner new applications that respond to new socioeco-
ferritic stainless steels in many components. nomic needs, such as for greater crash worthi-
As more exotic propulsion systems and fuel, ness, lighter weight, or resistance to the corro-
such as fuel cells and ethanol, emerge, stain- sion of new fuels. But, since exhaust systems
less steels may be required to endure the corro- currently predominate, they are covered first.
sive environments.
Exhaust Systems
Introduction
Laws enacted in the United States in the
The use of stainless steel in automobiles 1970s mandated automotive emission stan-
used to be mainly a story of decorative appli- dards that could be met only with catalytic
cations: wheel covers and trim with a minor converters. The only practical materials that
amount used for valves and hose clamps. How- could withstand the temperatures of the hot
ever, as automobiles became more sophisti- end of an exhaust system using a catalytic con-
cated technically and as durability and envi- verter were stainless steels. From Allegheny-
ronmental demands grew, the role of stainless Ludlum’s MF-1 evolved a succession of fer-
became increasingly functional and less orna- ritic alloys that grew in sophistication to meet
mental. Stainless alloys in common automo- the increasing needs of corrosion resistance,
tive use now are generally highly engineered oxidation resistance, creep, thermal fatigue re-
for their specific application and represent sistance, and formability. Soon, entire exhaust
some of the most highly evolved applications systems were made of stainless; often, they
engineering in any use of stainless. An exami- would last the life of the vehicle, rendering
nation of the preferred practices in materials obsolete an entire muffler replacement indus-
selection in automotive systems is an excellent try. The compelling need of the automotive in-
example of the rule of using the simplest and dustry for economy drove the widespread
lowest alloy content grade that can do the job. adoption of argon oxygen decarburization
Because automotive and steel-producing engi- (AOD), the continuous caster, and other high-
neers have collaborated so well, both parties volume methods of the carbon steel industry.
have benefited greatly, as have consumers, and So, even while the traditional automotive uses
of stainless steels—wheel covers and other or- The metallurgy of ferritic stainless is dis-
namental trim—faded to nearly nothing, the cussed in depth in Chapter 8. Rather than reex-
use of stainless in automobiles grew to about plain these concepts here, we revisit only the
30 kg (65 lb) per vehicle by the turn of the cen- main points that are relevant to alloy selection
tury. The stainless steel industry was changed for exhaust systems.
from a boutique industry to a mass production • As chromium level increases, so does re-
industry by its embracing the needs of the au-
tomotive market. sistance to oxidation and corrosion, but
An exhaust system normally consists of a yield strength also increases, and ductility
manifold to collect exhaust gases, a catalytic decreases.
• Alloying with silicon, aluminum, and
converter to reduce NOx and CO emissions, and
a muffler; each of these are connected by pip- molybdenum also increases oxidation resist-
ing. Each component of the system has different ance, but these elements have the same
requirements for formability, resistance to oxi- detrimental effect on ductility while increas-
dation, resistance to external corrosion, resisting hot strength.
• Niobium above that needed for stabilization
ance to internal corrosion, and mechanical
properties. At the outset, it should be noted that is a powerful solid solution hardener and is
ferritic stainless steels, as opposed to austenitic, effective at high temperatures.
• Ferritic stainless steels have very anisotropic
are optimal for oxidation resistance, especially
cyclic oxidation. It is not a difference in the forming properties. They resemble high-
oxide scale. formability carbon steels in that they tend
The reason is that the thermal expansion of not to thin when stretched, which greatly as-
ferritic stainless more closely matches that of sists in formability.
• Ferritic alloys can form a hard, brittle phase
the oxide scale than does that of austenite. This
prevents the fracturing and spalling of the scale. called α’ in a process commonly called 885
The intact scale of ferritic stainless is thus pro- °F (or 475 °C) embrittlement. This is only a
tective up to the temperature at which oxygen factor in alloys with chromium of 18% or
diffusion through the scale becomes great more, especially those containing molybde-
enough so that “breakaway” oxidation occurs. num and aluminum. Cold work accelerates
At the breakaway temperature level, scale the formation of this phase.
• The σ phase does not readily occur in alloys
growth is no longer parabolic with time but be-
comes linear and therefore no longer protective. containing less than 20% Cr, so it is not a
The temperature of this breakaway increases as consideration for exhaust systems unless sil-
chromium content increases. We will see that icon or molybdenum are also elevated.
• Coating ferritic steel with aluminum is effec-
other alloying elements can also improve this
performance. tive in preventing oxidation and corrosion.
Not only are the steels in exhaust systems al- All these factors come into play in the design
most exclusively ferritic, they are also stabi- of exhaust systems. Because the alloys have
lized by titanium or niobium. This prevents sen- evolved so well to fit the individual require-
sitization and makes all the chromium content ments for each component, we discuss them seg-
useful as alloy. Titanium stabilization greatly ment by segment through the exhaust system.
improves corrosion resistance by removing not The exhaust manifold collects the hot, com-
just the carbon and nitrogen, but also the oxy- busted gases from the engine and delivers
gen and sulfur from solution. This sharply im- them to the front pipe. The exhaust manifold
proves resistance to pitting corrosion. Niobium must possess good high-temperature strength
is used to costabilize and fight creep. These al- and resistance to thermal fatigue. It must also
loys are therefore essentially interstitial free and be able to resist oxidation at the exhaust tem-
have excellent formability, which the designs of perature, which can reach 950 (C (1740 (F).
exhaust system components require. Exhaust manifolds had previously been heavy
Their formability is further enhanced, when castings but are generally now formed from
necessary, by low additions of tramp substitu- stamped sheet stainless steel or formed from
tional alloying elements such as manganese, welded tubing that may have a double-wall
nickel, and copper. Special thermomechanical structure to insulate the gases from heat loss,
processing is also used to optimize texture and which could preclude successful catalytic con-
grain size. version downstream.
Chapter 19: Automotive and Transportation Applications / 227
As the highest-temperature component of the loys for spalling resistance. All approaches in-
exhaust system, the exhaust manifold must pos- volve raising chromium content but use differ-
sess the greatest resistance to high-temperature ent techniques to enhance the effect of
oxidation damage. Risk of such damage is due chromium. Table 1 lists a number of the grades
to the intermittent use of vehicles, which causes of stainless steel commonly used in exhaust sys-
cyclic oxidation and the ensuing spalling of the tems and where they are used. The alloys are
oxide scale. There are numerous alloying ap- listed in order of increasing severity of the re-
proaches for optimizing the ferritic stainless al- quirements for each major system component.
Table 1 Alloys normally used for the major elements of automotive exhaust systems
Service temperature
Alloys currently used, common name
Component ºC ºF Requirements (related designation)
Exhaust manifold 750–950 1380–1740 High-temperature • T439HP (UNS S43035, dual-stabilized 439)
strength, thermal • 18CrCb (DIN 1.4509, 18CrCb)
fatigue strength, • 441 (DIN 1.4509)
oxidation resistance, • 304/304L/304H (UNS S30400, S30403,
formability S30409)
• 321 (UNS S32100)
• 309S (UNS S30908)
• 310S (UNS S31008)
• 332Mo (S35125)
• 600 (N06600)
• 601 (N06601)
• 625 (N06625)
Front pipe 600–800 1110–1470 High-temperature • 409 ALMZ (aluminized 409)
strength, thermal • T439HP (UNS S43035, dual-stabilized 439)
fatigue strength, • 18CrCb (DIN 1.4509, 18CrCb)
oxidation resistance, • 441 (DIN 1.4509)
formability • 436S (type 436S)
• 444 (UNS S44400, T441)
• 433 (T443)
Flexible pipe 600–800 1110–1470 High-temperature • 304/304L (UNS S30400, S30403)
strength, thermal • T321 (S32100)
fatigue strength, • 316/316L (S31600/S31603)
oxidation resistance, • 316Ti (S31635)
formability, salt • 332Mo (S35125)
attack resistance • 625 (N06625)
Catalytic converter shell 600–800 1110–1470 High-temperature • 409HP (UNS S40930, dual-stabilized 409)
strength, salt • T439HP (UNS S43035, dual-stabilized 439)
attack resistance, • 441 (DIN 1.4509)
formability • 18CrCb (DIN 1.4509, 18CrCb)
• 444 (UNS S44400, T441)
• 433 (T443)
Catalytic converter 1000–1200 1830–2190 Oxidation • ALFA-IV (FeCrAl)
substrate resistance,
thermal shock
resistance
Center pipe 400–600 750–1110 Salt damage • 409HP (UNS S40930, dual-stabilized T409)
resistance • 409 ALMZ (aluminized 409)
• T439HP (UNS S43035, dual-stabilized T439)
• 441 (DIN 1.4509)
• 18CrCb (DIN 1.4509, 18CrCb)
• 444 (UNS S44400, T441)
• 433 (T443)
Muffler 100–400 210–750 Corrosion • 409HP (UNS S40930, dual-stabilized T409)
resistance, • 409 ALMZ (aluminized 409)
from inner and • T439HP (UNS S43035, dual-stabilized T439)
outer surface • 436S (T436S)
• 441 (DIN 1.4509)
• 18CrCb (DIN 1.4509, 18CrCb)
• Type 304/304L (UNS S30400, S30403)
Tailpipe 100–400 210–750 Corrosion • 409HP (UNS S40930, dual-stabilized T409)
resistance, • 409 ALMZ (aluminized 409)
from inner
and outer surface
Source: Adapted from Allegheny Technologies Inc.
The best choice for a given design is not obvi- some regions, so it must resist hot salt corro-
ous. We attempt to simplify the choices. sion. This may force the use of 316L versus the
Thus, the basic alternatives for exhaust sys- normal choice of 304L.
tem alloys are: The catalytic converter is the next compo-
• Straight chromium alloying at 11 to 12% nent of the exhaust system. It exposes the ex-
haust gases to noble metal catalysts, which
with stabilization by titanium or niobium,
complete the combustion of the gases to form
the basic type 409 (UNS S40920)
• Straight chromium alloying at 17 to 18% less-noxious compounds. This is an exothermic
reaction at temperatures equal to those in the
with stabilization by titanium or niobium,
exhaust manifold. Thus, the housing, while not
the basic type 439 (UNS S43036)
requiring great hot strength, must resist oxida-
Either of these basic alloys can enjoy en- tion. The catalyst itself is supported by a ce-
hanced oxidation resistance by additional alloy- ramic and ferritic stainless steel carrier that
ing with molybdenum, aluminum, or silicon. In must resist thermal shock and possess low heat
addition, they can be coated with hot-dipped capacity for rapid heating. Exotic alloys of
aluminum-silicon alloy to increase oxidation re- 20% Cr with 5% Al are used for the carriers.
sistance. The housing is generally made of a 17% Cr fer-
Chromium or molybdenum alloy additions ritic stainless. The converter is usually directly
increase corrosion resistance, whereas alu- beneath the passenger compartment, so a heat
minum or silicon additions do not improve that shield of type 409 is used to separate it from
trait. Aluminum coating is a powerful corrosion the floor.
fighter, and it has the aesthetic benefit of not The center pipe conveys the converted gases
showing red rust. to the muffler. The cooling exhaust gases no
Use of molybdenum or niobium enhances longer present a major oxidation threat, but the
high-temperature strength. Alloys with these condensing water vapor creates an internal cor-
additions are thus useful for manifolds with a rosion risk, and road salt presents an external
design that constrains expansion and contrac- one. However, a simple grade such as 409
tion, making thermal fatigue a problem. should generally provide sufficient resistance to
All alloying additions detract from formabil- this environment.
ity and toughness, as well adding to basic mate- The muffler, next in line, presents only a cor-
rial costs. Thus, the objective must be to use rosion issue. The muffler must withstand corro-
only those alloying elements that are indispen- sion from the outside, the worst of which comes
sable to performance. from road salt or coastal salt sources. Internal
The front pipe connects the exhaust manifold corrosion is also a major consideration because
to the flexible joint and experiences nearly the condensing exhaust gases create a hostile,
same temperatures as the exhaust manifold, but acidic environment. After startup, the heating of
not the same risk of thermal fatigue. To reduce the muffler to temperatures above 100 °C (212
exhaust noise, a double-wall pipe is sometimes °C) evaporates these condensates, and internal
used for this component. corrosion ebbs. On short runs, this may not
The flexible joint is the one segment of the occur. This represents a worst case for internal
exhaust system for which austenitic stainless corrosion. The dual internal and external corro-
steels are preferred. The function of the flexi- sive attacks require the use of aluminized stain-
ble joint is to prevent vibration from the en- less for best performance.
gine from being transmitted to the rest of the The tailpipe is exposed to view in most vehi-
exhaust system. It consists of a double-wall cles, and its appearance is therefore important.
pipe in a bellows configuration with an outer For this reason, an austenitic such as 304 can be
covering of braided stainless steel wire. It must used, as can chromium plating or aluminizing.
have very good high-temperature fatigue The object here is to avoid visible corrosion.
strength to withstand the cyclic stress of the vi- Truck exhaust systems are beginning to re-
bration it absorbs. The material used must quire similar technical sophistication as their
have exceptional formability to be formed into emissions come under increased regulation.
a bellows. The greater hot strength and forma- However, they do not present any challenges
bility of austenitic steels thus prevails. The not already confronted and solved for passenger
flexible joint is also exposed to road salt in vehicles.
Structural Components but extraordinary resistance to localized thin-

ning, necking, and therefore fracture. When
The driving forces of durability, safety, and crash worthiness becomes a prime considera-
weight reduction have spawned other, more var- tion, then this characteristic makes 301, or its
ied applications for stainless in automotive en- low-nickel counterpart 201, an ideal material for
gineering. Across the board, the main distin- structural, energy-absorbing components since
guishing trait of stainless that qualifies it as the austenitic stainless can be rivaled for such appli-
optimal material is its corrosion resistance, but cations only by heat-treated alloy steel, titanium,
this characteristic would be insufficient in many or aircraft aluminum alloys, all of which are
cases without considering mechanical proper- more expensive, less durable, or less formable.
ties. Indeed, even if stainless steel were not cor- Tables 2 and 3 show the properties of specific
rosion resistant, its superior strength and tough- variations on basic 301 developed by Out-
ness would qualify it for many automotive okumpu and how they stack up against the
applications. most competitive carbon steels, dual-phase
Austenitic stainless steels are the toughest and steels, and TRIP steels (Ref 1). The value of a
stiffest practical materials available to the auto- material as an energy-absorbing structure (i.e.,
motive engineers. Common 301 can be cold one that enhances crash worthiness) is meas-
worked to yield strengths anywhere from its an- ured by the energy it can absorb per unit of
nealed level of about 300 MPa (44 ksi) up to mass. The kinetic energy of a collision that a
2000 MPa (290 ksi). In this higher-strength con- structure can absorb in deformation is propor-
dition, it has become the standard material for tional to its strength multiplied by the amount it
seat belt anchors and hose clamps. Type 301 in can deform before fracturing. The superiority
the annealed condition is actually the original of metastable stainless steels (i.e., 201 and 301,
transformation induced plasticity (TRIP) steel as those that most easily transform to martensite
it can be tailored to have a controlled level of during deformation) is shown in Fig. 1. Even
austenite stability. This allows it to transform at with its lower density, aluminum falls far short
a known rate to martensite during deformation, of austenitic stainless in energy absorption per
giving not only a very high work-hardening rate unit weight.
Table 2 Comparison of tensile properties of carbon steels and stainless steels for automobile
structural components
True stress at
0.2% proof Ultimate tensile ultimate tensile Uniform
Type Thickness, mm strength, MPa strength, MPa strength, MPa elongation, % Total elongation, %
Carbon steels
TRIP 700 1.58 473 703 818 16.4 17
DP 750 1.48 513 811 920 13.4 18.8
DP 800 1.44 573 896 976 8.9 9.9
Austenitic stainless steels
HyTens X 1.16 306 937 1429 52.5 59.3
HT 1000 1.55 639 1068 1377 28.9 38.6
Source: Ref 1
Table 3 Comparison of resilience and toughness of carbon steels and stainless steels for
automobile structural components
Type Resilience, J/m3 Toughness, j/m3
Carbon steels
TRIP 700 0.996 105
DP 750 1.131 101
DP 800 1.32 74
Austenitic stainless steels
HyTens X 0.536 364
HyTens 1000 1.726 269
Source: Ref 1
Fig. 1 True stress-true strain curves for 301 variants (HyTens X and HyTens 1000) versus two duplex carbon steels (DP750 and
DP800) and a transformation induced plasticity (TRIP) steel (TRIP700). Source: Ref 1
These exceptional strength-to-weight and en- on individual body components is thus essen-
ergy absorption-to-weight characteristics permit tially nil. However, if the entire system is stain-
automotive engineers to reduce weight and in- less and the investment is avoided, then the ini-
crease crash worthiness while designing vehi- tial cost of a stainless body actually can be lower
cles with greater life span—because corrosion than one in coated carbon steel. This is the expe-
resistance “comes along for the ride,” as it were. rience of Italian bus manufacturers, who began
Some components in which these virtues are in the 1980s using 304 stainless steel in buses.
most readily exploited are bumper systems (Ref Now, buses are 80% stainless.
2). Porsche uses austenitic stainless steel for Designers began the conversion to gain the
front and rear side members, internal push rods normal advantages the stronger stainless gives:
on front and rear axles, and lower rear wish- over 10% lighter weight and over 10% im-
bones in its Carrera GT. provement in crash worthiness of the passenger
Another manufacturer, Audi, engineered vari- compartment, the accompanying savings in fuel
ous components of austenitic stainless steel into consumption, and the virtual elimination of
its otherwise aluminum-intensive A6 series. The body maintenance. With essentially the entire
use of stainless steel in strategic components body now in stainless, coating and painting
enables greater weight reduction than that could be eliminated. A stainless bus body is
which the vehicle would have in all aluminum. shown in Fig. 2.
Volvo and Saab have designed austenitic stain- This swung the economic pendulum to stain-
less steel bumper systems that also serve to re- less in a major way. Now, not only was the
duce overall vehicle weight. long-term cost of operating the bus lower, but
While it is probably apparent to the reader that the initial cost of the bus was lower. The eco-
essentially any body component can be made in nomic analysis is shown in Table 4 (Ref 2).
stainless and be made better in stainless, the The design key was to use rectangular 304
question of when doing so is a better engineer- stainless structural tubing, which allowed strong,
ing decision involves economic considerations. stiff sections to be welded into space frames. It is
Large automotive companies generally have only a matter of time until this is improved on by
large fixed investments in painting and coating the use of 201 (with 3 to 4% Ni instead of 8 to
systems to protect entire bodies from corrosion. 9% Ni) to lower cost and cold working of the
The incremental savings of eliminating coatings tubing to achieve higher strength levels.
Microcars are now a familiar sight in Europe. few manufacturers, such as Volkswagen, have
These vehicles are prized for their ability to be installed stainless steel fuel tanks in their vehi-
driven and parked in very small or congested lo- cles (Ref 2).
cations. Their economy of operation is also a
major attraction. These considerations combine
to make stainless the best material for many of Trucks
their components. Figure 3 shows a stainless
steel microcar frame. Over-the-road trailers are an excellent exam-
The design by the famous design house Pinin- ple of stainless steel being used for utilitarian
farina employs a stainless frame to give maxi- purposes. Trailers used for hauling foodstuffs or
mum torsional stiffness and crashworthiness corrosive materials are now constructed almost
while eliminating painting entirely. entirely of stainless steel; lined carbon steel
tanks are now largely obsolete. The engineering
basis for this is the same as for buses: high
Other Automotive Components strength, no coating costs, and a product with
long life and low maintenance costs. Structural
members in trailers are typically 304, while
Stringent emissions controls regulations, led
tanks may also be 316L for corrosion resistance
in the United States by the state of California,
when the transported material requires it. Tank
have made manufacturers reexamine the suit-
wrappers are often made of bright annealed and
ability of polymeric fuel tanks. These tanks con-
buffed 304. Manufacturers of trailers would be
tribute more to the required maximum 2 g/day
well advised to consider upgrading to duplex
of hydrocarbon emissions than is tolerable, so a
grades such as 2003 or 2205 or to cold-worked
austenitic stainless, which would permit major
weight reduction. This weight reduction would
directly translate into greater load-carrying ca-
pacity because the payload of liquid-carrying
trailers is limited by total gross weight. The
ability to add a few thousand more pounds of
payload would quickly pay back a small pre-
mium in material cost.
Normal cargo-carrying trailers also use some
stainless where corrosion is problematic, such
as in doors and door frames. Weight reduction is
less important in these trailers, which reach
maximum load at a volume limit rather than a
Fig. 2 Stainless steel bus bodies. Source: Ref 2 weight maximum.
Table 4 Life-cycle cost calculation (LCC) for stainless steel versus carbon steel
for a bus application
Cost of capital 10.00%
Inflation rate 5.00%
Real interest rate 4.76%
Desired LCC duration 20.0 years
Downtime per maintenance/replacement event 1.0 day
Monetary unit U.S. $
Value of lost production 101 U.S. $/day
Stainless steel Carbon steel
Material costs 3.331 1.391
Fabrication costs 25.322 28.582
Other installation costs 2.185 4.050
Total initial costs 30.838 32.023
Maintenance costs 0 1.448
Replacement costs 0 2.897
Lost Production 0 57
Material-related costs 0 0
Total operating cost 0 4.402
Total LCC cost 30.838 36.425
Source: Ref 2
Fig. 3 Microcar frame fabricated from stainless steel. Source: Ref 2
Rail Transport been successful for decades, has not been carried
over into trucks and buses even though it is tech-
Passenger trains have exploited the high nically feasible to economically produce struc-
strength-to-weight and toughness qualities of tural sections in very high-strength stainless.
the 301 family of stainless steels for many
years. The corrosion resistance of these alloys
makes them corrosion free in long use, obviat- REFERENCES
ing the need for painting and lowering mainte-
nance costs. As with any other major use of type 1. R. Andersson, E. Schedin, C. Magnusson,
301, a 5 to 10% increase in economy could be J. Ocklund, and A. Persson, The Applicabil-
achieved if type 201 were used instead of 301. ity of Stainless Steels for Crash Absorbing
No loss in performance would occur. The transi- Components, ACOM, No. 3–4, AvestaPo-
tion to 201 has not occurred simply because of larit AB, 2002
inertia and resistance to change on the part of 2. F. Capelli, V. Boneschi, and P. Viganò,
designers and producers. “Stainless Steel: A New Structural Automo-
Hopper cars made of 12% Cr martensitic tive Material, Vehicle Architectures: Evolu-
stainless steels, typically 409Ni and 3Crl2, have tion Towards Improved Safety, Low-Weight,
excellent abrasion and corrosion resistance as Ergonomics, and Flexibility,” paper
well as high strength and therefore greater load- presented at Florence ATA 2005, 9th Interna-
carrying capacity. Curiously, the use of cold- tional Conference (Florence), May 2005,
worked austenitic stainless in railcars, which has www.centroinox.it, accessed June 2008
CHAPTER 20
Commercial and Residential

Applications
Summary Stainless is without rival for ruggedness and

durability. Steel and aluminum corrode. Glass,
STAINLESS STEEL HAS BECOME the es- stone, and ceramics break. Plastic is weak. The
sential material for products related to food, second, even more important, reason is that
health care, and laundry because it combines stainless steel is essentially benign from a hy-
strength and durability with an unexcelled abil- gienic viewpoint. Stainless steel itself is inert,
ity to be cleaned, disinfected, and sterilized. both chemically and biologically, with respect
These qualities have long been apparent to com- to food. Further, it provides minimal harbor for
mercial food, laundry, and health care profes- unwanted biologic growth as do more porous
sionals and have increasingly carried over into materials. Stainless also competes quite well es-
equivalent domestic areas as consumers have thetically with other materials, offering the de-
become more aware of the benefits of stainless. signer numerous surface finishes. Last, stainless
is very amenable to nearly all manufacturing
techniques. Its lack of need of coatings often
Introduction makes components made from stainless less ex-
pensive to produce than equivalent designs that
The last 20 years have seen the long-standing must be coated with paint, porcelain, or metal.
pervasive commercial use of stainless steel for
food preparation and serving; laundry; heating,
ventilation, and air conditioning (HVAC); and Food Contact Qualifications
other appliances penetrate the domestic market
for the same types of goods. Whether this is a Setting aside cost, esthetics, and manufactur-
fad of an increasingly affluent consumer or a re- ing considerations, a food contact material must
flection of more design engineers and con- first meet three criteria: It must be chemically
sumers being more interested in lasting value inert, biologically inert, and cleanable and able
than they were in the “throwaway” society that to be disinfected.
preceded that period remains to be seen. Stain- Chemical neutrality is achieved by a mate-
less has been increasingly identified with high- rial when the material does not enter into the
quality, high-end products. But, the case for food with which it comes in contact. This has
value rather than fad seems to be stronger if the become an increasing concern as the effects of
lessons of the harshly pragmatic automotive in- ions or chemicals released from food prepara-
dustry, in which decorative use of stainless has tion materials have been viewed as potential
virtually disappeared while utilitarian uses have toxic or disease agents. Medical knowledge is
mushroomed, are any indication. not sufficiently advanced to convince con-
The case for using stainless in appliances of sumers of the harmlessness of such contami-
all types, whether they are commercial or resi- nants, so it is preferable to demonstrate the ab-
dential, relates to stainless being able to provide sence of contamination if one is to win the
the best value over the intended service life. public confidence in a food contact material.
Stainless steel contains many constituent ele- rally occurring level of up to 0.3 μg/g in cereals
ments. Were they to enter the food with which to 1.1 μg/g in meat and fish.
the stainless came in contact, then stainless These negligible levels of leaching simply in-
would be a poor food contact material. The dis- dicate that foodstuffs are a benign chemical to
tinguishing characteristic of stainless, however, stainless steel. Nevertheless, it is necessary to
is the spontaneous passive film, which is so sta- apply the correct assessment of the corrosivity
ble chemically. This film acts as a barrier to cor- of the foodstuff in question. In food production,
rosion, which would result in metal release. as opposed to preparation for serving, more ex-
Stainless therefore is effectively inert. Tests treme levels of acidity and salinity can be en-
have been made of the rates at which metal ions countered. Nippon Steel reported (Ref 3) that
can enter foodstuffs (Ref 1). Table 1 shows the materials used in the manufacture of soy sauce,
vastly lower rates of metal ion release from which can have 15% salt, must withstand pro-
stainless than from aluminum and carbon steel, longed contact at 45 °C (115 °F). Under such
both of which are permissible, if not optimal, conditions, 316 stainless pits in about 1 day,
food contact materials. Aluminum releases alu- while higher alloy grades, the 6Mo alloys, of
minum ions into solution of both cooking oil which their YUS 270 is one (equivalent to UNS
and 3% acetic acid at nearly equal rates of S31254), are projected to last 20 years before
15 mg/cm3 in 30 days. Carbon steel releases pitting.
iron at over 100 mg/cm3 in the same period. This is significant because only pitting corro-
Stainless, however, releases less than 0.010 sion is likely to release metals ions into a food
mg/cm3 of iron. This is complemented by other substance. So, while guidelines exist for the
studies showing that the transfer of ions from minimum alloy content permissible for normal
food contact vessel to food is diminishingly and food contact, such as those promulgated by the
negligibly small. Tests have been conducted on National Science Foundation (NSF), one must
stainless steels, types 304, 439, and 444, that still verify the corrosion due to a particularly
had both industrial finishes (2B and BA) as well aggressive food ingredient. Choosing the proper
as freshly abraded and air-aged finishes. These grade of stainless, based on pH, salinity, and
sample steels were subjected to boiling solu- temperature, is the responsibility of the design
tions of oils, alcohols, water, and 3% acetic engineer. Referred to chapters in this book on
acid. None caused the transfer of either both corrosion and individual alloy families for
chromium or nickel to exceed the statutory 0.1 guidance in choosing an alloy based on resisting
ppm level (Ref 2). Nickel levels of various pitting in a given environment. This having
foods before and after cooking have been scien- been said, no alloy greater in pitting resistance
tifically measured to assess the possibility of than 304 is required in residential or commer-
leaching of that ion from 304 stainless steel cial cooking food contact. The higher alloy
(Ref 2). No increase was noted from the natu- requirements come from the more aggressive
Table 1 Net metal migration into acetic solution (3%)

Metal migration during indicated time, µg/cm3
Material Time Iron Chromium Aluminum Nickel

Austenitic stainless 30 min 2.4 0.12 ≤ 0.19 ≤ 0.12
10 days 4.2 0.22 0.22 ≤ 0.12
20 days 2.7 0.22 0.19 ≤ 0.12
30 days 2.3 0.28 ≤ 0.19 0.31
Ferritic stainless 30 min 3.0 0.43 ≤ 0.19 ≤ 0.12
10 days 7.3 0.40 0.19 ≤ 0.12
20 days 8.6 0.71 ≤ 0.19 ≤ 0.12
30 days 6.6 0.87 ≤ 0.19 ≤ 0.12
Aluminum 30 min 4.9 0.93 930 ≤ 0.12
10 days 18.2 3.42 5,300 ≤ 0.12
20 days 17.9 5.58 7,160 ≤ 0.12
30 days 31.3 12.40 15,350 0.22
Carbon steel 30 min 8,430 0.62 2.7 ≤ 0.12
10 days 57,700 7.40 26.7 ≤ 0.12
20 days 62,900 6.82 24.0 ≤ 0.12
30 days 112,000 14.00 36.9 ≤ 0.12
Source: Ref 1
Chapter 20: Commercial and Residential Applications / 235
concentrations and exposure periods that can be mercialized (Ref 5). Silver ions, like copper
found in food-processing plants. ions, are powerful antimicrobial agents. The
Certain abuses can even damage stainless combination of such a coating with stainless as
cookware. Very high temperatures, such as can a corrosion-proof substrate may represent the
occur when unattended pans have their liquids maximum in hygienic and chemical protection
boiled away, can damage stainless but are more and is already being used in medical applica-
harmful to less-rugged alloys such as copper tions where such concerns exceed those in ordi-
and aluminum. nary food contact situations.
Stainless steel is primarily composed of iron, Cleanliness. A necessary quality in any mate-
chromium, and nickel along with small amounts rial considered for food contact is the ability to
of manganese, silicon, and molybdenum. It con- be cleaned. This includes the removal of both
tains trace amounts of copper, aluminum phos- organic and inorganic substances. The most im-
phorus, and sulfur. Each of these elements is portant objective of cleaning is to remove the
naturally occurring in food. Each can be found visible and invisible materials that can provide a
in a typical multivitamin/multimineral supple- growth medium for microorganisms. This
ment. Stainless is essentially devoid of heavy process is distinguished from disinfection,
metals, such as lead and mercury, which are va- which is the reduction of the microbial popula-
porized at the temperatures at which stainless is tion to a satisfactory level. What this level is de-
refined. Even if toxic metals were somehow to pends on the standards of hygiene in force. And,
be made to contaminate stainless, the passive although cleaning can and does reduce the pop-
film would prevent their release. All these fac- ulation of microorganisms, true bacteriological
tors combine to make stainless the most chemi- cleanliness is obtained only after disinfection.
cally neutral metal found in food contact. Alter- The combination of cleaning and disinfecting is
native alloys, such as copper and aluminum, important. Studies have shown that the efficacy
actively leach into foods. Copper and aluminum of disinfectants is weaker on bacteria that have
have been linked to but not demonstrated to been established in a surface biofilm than on
cause Alzheimer’s disease. bacteria in suspension. The most complete form
Biological Neutrality. Microorganisms ad- of disinfecting is sterilization, whose objective
here to solid surfaces. When a clean surface is the complete removal of all microbial life and
comes in contact with food, a surface deposit is viruses.
formed from the food. The film may also con- The purpose of cleaning stainless steel is to
tain molecules left from previous cleaning and rid it of contamination. Various stainless manu-
disinfecting. The formation of this film is pre- facturers and associations have identified a
sumably influenced by material characteristics number of effective of cleaning products (Ref 2,
such as roughness, although there are no spe- 6, 7):
cific studies on this. However, microorganisms • Alkalines, which dissolve fats and oils
adhere to this film and, as colonies of them • Chelating or sequestering agents, which ag-
grow, form a biofilm. This film consists of lay-
ers of microorganisms that can produce an exo- glomerate contaminants. These are often or-
cellular polymeric matrix, which protects the ganic acids such as citric acid or oxalic acid
colony from cleaning and disinfecting. Geomet- and amine acids such as sulfamic acid and
ric factors also can protect these colonies. ethylene diamine tetraacetic acid (EDTA) or
Rough surfaces are intuitively more difficult to salts of these compounds.
• Hydrocarbon solvents
clean. The ability to maintain a microscopically • Water with soap, detergent, trisodium phos-
smooth surface is an asset in stainless that poly-
meric, enamel, and mineral surfaces lack. Stain- phate, or other surface active agents, which
less steel is much less roughened by abrasion, emulsify
• Dilute oxidizing acids like nitric acid
keeping the surface smooth (Ref 4).This will be • Mild acids such as phosphoric acid
seen to influence its ability to be cleaned and
disinfected. The effectiveness of a cleaner relates mainly
There is some technology to go beyond bio- to the contaminant to be removed. Some trial
logical neutrality in the use of coatings that ac- and error may be required for a given contami-
tively discourage or eliminate growth of mi- nant. Some precautions are worth mentioning.
croorganisms. Polymeric coatings impregnated Abrasive cleaners should be used with caution.
with silver ions have been developed and com- The abrasive size and hardness must be chosen
so that the stainless surface finish is not affected abrasive polishing, permits bacterial colonies to
in an unwanted manner. If the abrasive is harder be removed. The greater roughness of the other
than the stainless or coarser than the stainless materials may serve to protect the bacterial
surface roughness, the underlying finish can be colonies from shear forces and provide greater
disturbed. Care should also be taken to clean specific surface area on which the colonies can
with the polish grain if a polished surface is bond.
being cleaned. Also, cleaners containing chlo- Disinfection. The ability of a surface to be
rides are common. Their use is not recom- disinfected is measured by the concentration of
mended on stainless. Use of hydrochloric a given disinfectant required for a specific re-
(muriatic) acid is especially detrimental. If duction in bacterial population. Numerous
chloride-containing cleaners are used, then thor- studies have been published (e.g., Ref 8–10)
ough rinsing should be conducted to avoid showing that glass and stainless steel have equal
chloride concentration through evaporation, es- aptitude for disinfection, and that polyesters,
pecially in crevices. Steel wool or steel brushes polyurethanes, rubber, and aluminum all re-
should not be used on stainless under any cir- quired about one to two orders of magnitude
cumstances as iron residue interferes with the greater concentrations of disinfectant for the
integrity of the passive film. same result. These results indicate why stainless
The ability of stainless to be cleaned is best is so essential to the food industry. Stainless can
measured by the actual removal of bacteria be disinfected quite readily, which allows the
colonies. This has been done to compare great invisible liability of food-borne diseases
unabraded and abraded (to simulate new vs. to be minimized.
used) stainless steel, enameled steel, mineral The effectiveness of sodium hypochlorite as a
resin, and polycarbonate materials, which can disinfectant is inarguable, also. So, despite its
be used for sinks, counters, food prep tables, potential corrosivity, it will be commonly used.
etc. (Ref 4). Figure 1 shows that the reduction in Taking this into account requires that commer-
bacteria count by the same cleaning technique is cial and residential food equipment be able to
ten times more effective on stainless than on the withstand some chloride level greater than oth-
other material types. Abrasion did not degrade erwise projected. Industry practice in the United
the ability of stainless to be cleaned as it did States has shown that corrosion problems occur
softer materials. The surface of stainless, even at an unsatisfactory level with mechanically
with the seemingly protected recesses due to polished 430 but not with bright-annealed 430.
Fig. 1 Bacterial retention as a function of material and cleaning time. Source: Ref 4
Thus, alloys with less than 16% Cr should not thermal conductivity in a cooking utensil mini-
be used unless corrosion can be accepted. Al- mizes differences in temperature across the sur-
loys containing 16% Cr can be used with opti- face in contact with the food, permitting better
mal surface finish. Alloys as low in carbon as control of the cooking process. The solution to
12% are used for cutlery applications where the problem of thermal conductivity is to make
slight corrosive attack can be accepted. This is a composite materials. Stainless can be bonded to
necessary trade-off required to achieve high copper and aluminum, which allows the stain-
hardness for good cutting edge retention. less to be on both the food contact surface as
Higher chromium grades such as 304 can be well as the exterior, with an inner layer of cop-
used even with mechanically polished surface per or aluminum effectively spreading the heat.
finishes. From a cost-effectiveness point of Aluminum and copper are nearly equally effec-
view, there is no reason to use more expensive tive as inner conductive layers. Premium cook-
alloys than 430 or 201 in the vast majority of ware features them both. The “sandwich” is the
commercial and residential kitchen and laundry optimal design because it optimizes heating uni-
applications from a corrosion standpoint as long formity even more than using aluminum or cop-
as surface finishes that have not been produced per alone would since the high conductivity
by abrasive polishing are specified. Many such inner core functions as an isotherm. The unifor-
finishes are widely used. In North America, the mity is the more important consideration than
rolled-on replicas of No. 4 finish, Koolline, the absolute thermal conductivity or even the
Lustrite, etc., are quite common, while in Eu- thermal diffusivity.
rope the bright-annealed finish has been pre- In a triple layer, the choice of the non-food-
ferred. Both of these are preferable to mechani- contacting stainless is less stringent. Some-
cally polished finishes. times, the exterior is made of a ferritic stainless
The food industry, an immense consumer of steel. The ferromagnetism of ferritic stainless
stainless steel, could do more than any other in- steel makes it ideal for induction heating. Al-
dustry to help conserve nickel by specifying alloys such as 436 have been used for this appli-
loys such as 430, 439, and 201 as their standard cation, while 304 is the pervasive choice for the
alloys as well as by specifying nonabrasive fin- food contact surface. This is despite the fact that
ishes. This can be done with no loss of function- 201 or 301 are quite adequate for this applica-
ality or change of appearance and could save tion. It is also possible to produce a magnetic
23% to 50% in material cost. carbon steel core with stainless bonded to both
sheet surfaces. The exposed edges are rolled to
shield them from corrosion.
Applications Nonstick coatings, such as polytetrafluoreth-
ylene (PTFE), are very popular because of their
Cookware. Any interaction between a food nonstick qualities. Above 350 °C (660 °F) these
contact material and the food is most likely to coatings give off toxic fumes. This is a danger
occur during the cooking process when temper- for certain types of cooking, such as wok cook-
atures are greatest. Only glass and stainless are ing or blackening, but more likely to be encoun-
excellent food contact materials. And, since tered by accidentally high temperatures above
cookware must be flexible enough to handle any those intended. Since they can be scratched and
potential food, the choice of material for cook- are not impermeable, their use does not alter the
ware must be the most conservative. For this choice of the material to which they are applied.
reason and because of the brittleness of glass, Kitchen Appliances. Every type of commer-
stainless is the material of choice. The qualities cial kitchen appliance can be, and usually is,
discussed make aluminum and copper less de- made of stainless steel, as are premium domes-
sirable. Both leach into food. Copper can be tic kitchen appliances. This choice is based on
tinned to combat this. The tin also corrodes over durability and ease of cleaning and disinfecting.
time but has very low toxicity. A larger draw- And, because many commercial appliances are
back is the expense of retinning copper utensils. visible to the customer, aesthetics are also a
Aluminum is known as a toxic metal, with its driving force. Choice of alloy for a given appli-
toxicity causing symptoms similar to those of ance is a crucial cost factor. As was noted that
Alzheimer’s and osteoporosis (Ref 11). These 430 is marginal for kitchen use, because of the
two metals do have one advantage over stain- prevalence of chloride-containing cleaners, un-
less, however: their thermal conductivity. High less it has been bright annealed. All austenitics
are satisfactory under normal use. Designers

seem to generally neglect the possibility that
their equipment may be used in coastal cli-
mates. In the high ambient salinity of coastal
climates, corrosion will occur unless 304 with a
bright-annealed finish or a brushed finish rolled
onto a bright-annealed 304 is used. Mechani-
cally polished 304 stainless is inadequate for
coastal environments. These are the same
guidelines used for architectural applications.
Figure 2 shows how different alloys withstand
coastal conditions. The corrosion on 430 would
be considered excessive, while that on the 201
and 304 is acceptable given that some routine
cleaning would have prevented the corrosion
that is present on these samples, which were ex-
posed to coastal salt and humidity for 10 years
in North Carolina (Ref 12).
In the vast majority of ambient conditions,
coastal salinity is not a problem. This applies to
inland conditions or coastal conditions where
interior environments are protected by air con-
ditioning or adequate cleaning of the stainless is
practiced. This is normally the case for com-
mercial equipment. Under these conditions, 201
is quite adequate, and the use of 304 represents
wasteful overengineering. This choice is sup-
ported by decades of use by the major manufac-
turers of commercial appliances. Many who are
large enough to specify their desired grade on
bills of materials rather than simply buying
from service center inventories have routinely
used 201 and realized an approximately 8%
lower cost before surcharges. Use of 201 versus
304 reduces surcharges by almost 50%, which
can be a much larger savings than the base price
savings. Smaller manufacturers are often pre-
cluded from these savings because of the gen-
eral, if inexplicable, practice of service centers
not stocking 201 despite its being the most cost-
effective general-purpose stainless grade. The
extended nickel price elevation from 2004 on-
ward has a good chance of changing that situa-
tion as end users rebel against surcharges,
which cannot be passed on to their customers. It
has been pointed out that there is an array of
201-type grades, and that this is a drawback to
their wider adoption.
I recommend following American Society for
Testing and Materials (ASTM) A240 and speci-
fying UNS S20100 when substituting for 304 as
this has very similar performance in forming,
welding, and appearance to 304 and can be
most easily interchanged without complications Fig. 2 Stainless steel samples exposed on a North Carolina
in manufacturing and field performance. For beach for 10 yr. Source: Ref 12
parts made by deep drawing, substitution is still concentrations are in weight percent. This is dif-
very possible, but deep-drawn grades are more ficult to achieve for 17% Cr alloys if stabiliza-
finely tuned to specific process paths and must tion is by titanium alone.
be more tightly specified than general-purpose Interior or working parts of appliances, to the
grades. The more commonly used alloys for ap- degree they require high cleanability or contact
pliances are listed in Table 2. food, are also often made of stainless. This is
The greatest savings comes, of course, from especially true of dispensing machines, such as
using ferritic grades, and they should be used for beverages, ice cream, and ice. Stainless inte-
whenever forming requirements permit, which riors are often found in refrigerators and dish-
is the majority of the time, since most appliance washers. In the case of dishwashers, forming re-
components experience little more than cutting, quirements are often severe enough to require
bending, and welding. There are important pre- the use of austenitic stainless. Rolled-on fin-
cautions, however. Mechanical polishing results ishes are generally preferred. Not a small reason
in unacceptable corrosion resistance, and the for this is that this finish requires only a single
low work hardening rate of ferritics causes the temper pass to both flatten and provide the fin-
mechanical polish to take on a different color ish. This yields very consistent forming charac-
shade. This subtle difference can be magnified teristics, meaning much lower breakage during
to objectionable levels when a mechanically press-forming operations. Rolled-on finishes
polished ferritic stainless, such as 430, is put also have very high visual consistency, which is
side by side with an austenitic such as 201 or usually a very important quality criterion for ap-
304. This can be solved by specifying rolled-on pliance manufacturers.
finishes, which look the same on ferritics and Canisters, chafing dishes, serving pans, etc.
austenitics. These finishes also supply the added are generally made from austenitic stainless
corrosion resistance that makes alloys such as steel, which lends itself to the typical deep-
430 acceptable. It is still preferable to use a forming operations used in their manufacture.
dual-stabilized grade such as 468, which can be Coatings are rarely used. If antimicrobial coat-
welded without adverse corrosion effects and ing were to be used in food contact, this would
has high formability and corrosion resistance at be an ideal application since already cooked
as little as half the cost of 304 when alloy sur- food is most often in the intermediate tempera-
charges are factored in. Use of dual stabilization ture danger zone at which bacteria can multi-
permits keeping titanium levels to a minimum, ply. Food preparation tables also fit into this
making it possible to avoid TiN-caused surface category.
defects, which occur if significant TiN precipi- Appliance facades are increasingly using
tation occurs before solidification in the original stainless. These include refrigerators, stoves,
steel production. This occurrence is strictly a microwaves, drawers, etc. Shelves and exhaust
thermodynamic phenomenon related primarily hoods also benefit from being made of stainless.
to the titanium and nitrogen levels, which The drivers here are cleanability, durability,
should be minimized so that the product of tita- and esthetics. There are important visual con-
nium times nitrogen is less than 0.0025 when siderations in these applications. Consistent,
Table 2 Stainless steels commonly used for appliances

Composition, %
Alloy UNS No. C N Cr Ni Mn Si Mo Ti/Nb

201 S20100 0.15 0.25 16.0–18.0 3.5–5.5 5.5–7.5 1.00 ... ...
301 S30100 0.15 ... 16.0–18.0 6.0–8.0 2.00 1.00 ... ...
304 S30400 0.08 0.10 18.0–20.0 8.0–10.5 2.00 1.00 ... ...
316 S31600 0.08 0.10 16.0–18.0 10.0–14.0 2.00 1.00 2.0-3.0 ...
430 S43000 0.12 ... 16.0–18.0 0.75 1.00 1.00 ... ...
439 S43035 0.07 0.04 17.0–19.9 0.50 1.0 1.0 ... 0.20 + 4 × (C + N),
to 1.10
468 S46800 0.030 0.030 18.0–20.0 0.50 1.00 1.00 ... Ti + Nb: 0.20 + 4 ×
(C + N), to 0.80
436 S43600 0.12 ... 16.0–18.0 ... 1.00 1.00 ... Nb + Ta: 5 × C, to 0.70
444, S44400 0.025 0.035 17.5–19.5 1.0 1.0 1.0 1.75–2.50 Ti + Nb: 0.20 + 4 ×
YUS190 (C + N), to 0.80
29-4C S44735 0.025 0.025 28.0–30.0 0.5 1.00 0.75 3.5–4.5 Ti + Nb: 0.20 + 4 ×
(C + N), to 0.80
defect-free surface finishes are paramount. This characteristic, so most cutlery is tempered at
again can really only be achieved by rolled-on low temperatures. The vast majority of require-
finishes since abrasively polished finishes vary ments for high-quality cutlery are satisfied by
excessively in roughness, reflectivity, and color. 420 stainless. If greater cutting edge retention is
Panel flatness is often very important and an- desired, then more or harder carbides are engi-
other benefit from rolled finishes. If visible neered into the martensitic matrix. This is done
welds are required, as is often the case for prod- by adding more carbon and chromium, as is
ucts such as hoods and counters, then special found in 440A and to a greater extent in 440C.
finishes with very long polish grains have a The wear resistance added by carbides is pro-
major advantage in that the weld can be ground portional to their hardness and amount. The
and polished with a belt sander of the appropri- chromium carbides of these straight-chromium
ate grit size so that the weld blends impercepti- martensitic stainless steels are very hard, 1800
bly with the adjoining original surface. This is a HV, versus the 1100 HV hardness of iron car-
practical impossibility with abrasively polished bides. The addition of higher levels of carbon
finishes and very difficult with rolled finishes. ties up chromium so that it cannot add to corro-
Freedom from fingerprinting can be another sion resistance, however, so that it can become
valuable attribute for faÁade applications. This barely rust resistant. Furthermore, at high car-
can be obtained on stainless by the mill applica- bon levels, carbides precipitate in the liquid and
tion of a thin, bonded polymer film. All bare are much coarser. These large carbides can pull
stainless finishes show fingerprints. With un- out during edge honing, making a ragged rather
coated stainless, it is best avoided by using min- than a fine, smooth cutting edge.
eral oil-based cleaners. However, vanadium and tungsten have even
Although very high alloy stainless steels are harder carbides, 2800 and 2100 HV, respec-
used for high-temperature kitchen applications, tively. Through conventional casting and hot
such as heating element sheathing (American working, only a small amount of these carbides
Iron and Steel Institute [AISI] type 334), it is can be introduced into the matrix. The problem
seldom used for oven interiors because it does is that if primary carbides form during solidifica-
take on a heat tint when exposed to tempera- tion, they tend to be coarse and to embrittle the
tures above 300 °C (570 °F). Range tops, which alloy. Hard particles are much more useful for
see lower temperatures, are normally stainless. wear resistance if they are small and widely dis-
Outdoor cooking grills, because they must en- persed. To a degree, this refinement of the pri-
dure exterior environments without corrosion, mary carbides can be achieved by raising nitro-
are almost always 304 or a similar grade. Heat gen levels. These problems can be circumvented
tint does not occur with these to a problematic by the use of powder metallurgy, which permits
degree. Gas burner manifolds are also stainless. the solidification step on a macroscale to be
In this case, ferritics are required because of the skipped. Larger volume fractions of hard car-
need for extreme high-temperature oxidation re- bides such as vanadium carbide and tungsten
sistance and the desirability of a low coefficient carbide can be added and dispersed. Table 3 lists
of thermal expansion. The preferred alloys are the martensitic alloys used for cutlery. It is rea-
those developed for automotive exhaust sys- sonable to say that most of these alloys far ex-
tems, variations on 409 and 439. ceed the requirements of food preparation.
Flatware and cutlery were among the origi- Improved corrosion resistance of these alloys
nal uses of stainless. Stainless filled the gap be- is achieved by adding molybdenum at the ex-
tween carbon steel, which was hard but whose pense of chromium, which would cause exces-
rusting was an obvious problem, and silver, sive δ-ferrite retention if it were raised. This can
which was soft and whose cost prohibited its be seen in alloys above the basic 420.
use to all but the wealthy few. Cutlery is the do- Flatware has no hardness requirement, so
main of martensitic stainless steel. The corro- grade selection is based on the need for per-
sion resistance of martensitic grades cannot be ceived quality. At the high end is 304, which has
improved above modest levels, never reaching all the corrosion resistance that could be needed
that of 304, but this criterion is secondary to for flatware. However, type 301 is commonly
hardness because of the need to keep a sharp used also, as are ferritic steels, such as 430, for
cutting edge. Maximum corrosion resistance is low-cost flatware. Depending on the shape of
achieved in the as-quenched condition. But, the final utensil, material is stamped or forged
some toughness is a valuable but not crucial and then finished.
Table 3 Stainless steels commonly used for cutlery

Composition, %
Alloy Designation Form C Mn S Si Cr Mo Ni Other

420 UNS S42000 Wrought 0.15 min 1.00 0.030 1.00 12.0–14.0 ... ... ...
4116 DIN 1.4116, nominal Wrought 0.50 ... ... ... 14.5 0.65 ... 0.15 V
440A UNS S44002 Wrought 0.60–0.75 1.00 0.030 1.00 16.0–18.0 0.75 ... ...
440C UNS S44004 Wrought 0.95–1.20 1.00 0.030 1.00 16.0–18.0 0.75 ... ...
BG-42 Nominal composition Wrought 1.15 ... ... 0.3 14.5 4.0 ... 1.2 V
ATS-34 Nominal composition Wrought 1.05 0.4 ... 0.35 14.0 4.0 ... ...
14-4 CrMo Nominal composition Wrought 1.05 0.5 ... 0.3 14.0 4.0 ... ...
154 CM Nominal composition Wrought 1.05 0.45 ... 0.3 14.0 4.0 ... ...
CPM S30V Nominal composition PM 1.45 ... ... ... 14.0 2.0 ... 4.0 V
CPM S60V Nominal composition PM 2.15 0.40 ... ... 17.0 0.40 ... 5.5 V
CPM S90V Nominal composition PM 2.20 ... ... ... 13.0 1.0 ... 9.0 V
Many kitchen utensils are also made entirely clothing. Those, along with the implied quality of
or in part with stainless. Type 304 is the alloy stainless, are the main drivers for its use.
most commonly used, but again any of the Heating and Water Heating. With the de-
stainless steels with at least 16% Cr are ade- velopment of high-efficiency, natural-gas-fired,
quate, and grade selection depends on forming forced-air furnaces, stainless has come into do-
and joining requirements. mestic use as a heat exchanger material. These
Laundry appliances have converted signifi- furnaces gain their extra efficiency by condens-
cantly to stainless. This trend began in Europe ing water from combustion gas exhaust. This
with the development of high spin speed, hori- condensate can, depending on incoming air,
zontal axis washing machines. These washers contain corrosion elements, which has led to the
use far less water and energy to achieve higher use of very highly alloyed ferritic stainless steel
levels of cleaning with less damage to clothing. in their construction. Alloy 29-4C (UNS
These features have eroded the share enjoyed by S44735) was the original alloy used nearly uni-
vertical axis, agitator-type washers, whose low versally in the United States. The worst conse-
speed allows them to be constructed of low- quence of perforation by pitting could be the re-
strength materials such as plastic or porcelain- lease into the home of toxic gas, so pitting
coated steel. The stresses induced by the high corrosion must not be allowed.
spin speeds, which are necessary in horizontal The intermediate efficiency furnaces (80 to
axis machines to take water removal from 80% 90%) require the use of corrosion-resistant vent
to 95%, require the strength of stainless steel. pipe to prevent corrosion from condensation in
Porcelain-coated carbon steel obviously can be the flue. High-temperature plastics were tried,
strong, but the coating is cracked by strains that but failed joints in them caused their recall after
the steel itself easily tolerates. An additional ben- several fatalities were reported. High-efficiency
efit to stainless over porcelain is that stainless (90% or higher) furnaces can use low-tempera-
starts smooth and becomes even smoother with ture plastic pipe, but these units require the use
use, while porcelain becomes quite abrasive over of a corrosion-resistant secondary heat ex-
time as wear opens voids with edges that can be changer to recover the latent heat of vaporiza-
quite sharp and cause significant damage to tion of the water from combustion. Alloy 29-4C
clothing. Washer tubs and drums are made of was the original choice for most secondary heat
both ferritic and austenitic stainless. The selec- exchangers, but at least one used alloy 6XN
tion is based on forming requirements rather than (UNS N08367) alloy for formability. Some
corrosion or strength. If components can be made manufacturers have always used lower-alloyed
by bending rather than stretching, then the lower- stainless steels.
cost ferritics can be used. Ferritics should be a The proper handling of combustion products
17% stabilized grade, such as 439 or 468, and is an interesting problem in materials selection.
austenitics can be 201, 301, or 304. Unstabilized The variability of the use environments leads to
ferritic alloys, such as 430, should never be used a huge spread in corrosion conditions and mate-
in welded applications. Dryers are less challeng- rials performance. In the end, one has to balance
ing, and it is difficult to make a strong case for materials selection between cost (fortunately,
the functional value of stainless. Those designs 29-4C alloy is nickel free) and probability of
that use stainless will last longer and be gentler to failure. Given the number of units produced and
the potentially serious consequences of failure, 3. “The Application of High Corrosion

failure rates must be less than 10 – 4, while fail- Resistance Stainless Steel YUS270 in
ure rates much less than 10 – 6 are impossible to Food Processing Facilities and Equipment,”
verify and hard to justify. In any case, the com- Nippon Steel Technical Report 87, Jan
petition is always between various stainless 2003
steels. Galvanized steel will not work. The issue 4. J.T. Holah and R.H. Thorpe, Bacteria Re-
is difficult enough for natural-gas-fired fur- tention on Cleaned Surfaces, J. Appl. Bacte-
naces. Oil fired is a developing situation for riol., Vol 69, 1990, p 599–608
which there is no good consensus. Wood burn- 5. Agion Technologies, www.agion-tech.com,
ers and other unconventional furnaces (such as accessed June 2008
corn burners) present additional challenges, and 6. Removal of Stains and Discolourations, Out-
answers are even less obvious. okumpu, www.outokumpu.com, accessed
Water heaters are sometimes made of stain- June 2008
less steel. It is not uncommon for water to have 7. “The Care and Cleaning of Stainless Steel,”
a sufficient level of chlorides to lead to stress Specialty Steel Industry of North America,
corrosion cracking if an austenitic stainless is www.ssina.com, accessed June 2008
used. Therefore, the recommended alloy for this 8. E.P. Kysinski et al., J. Food Processing, Vol
application is UNS S44400. More recently, lean 55, 1992, p 246–251
duplex alloys have been developed, such as 9. A.A. Mafu et al., J. Dairy Sci., Vol 73,
2101 and 2003, which can perform quite well 1990, p 3428–4332
without corrosion or stress corrosion cracking. 10. P. Gelinas and J. Goulet, Can. J. Microbiol.,
More highly alloyed duplex alloys such as 2205 Vol 29, 1983, p 1715–1730
are more expensive but would work well. 11. R.A. Goyer, Toxicity of Metals, Properties
and Selection: Nonferrous Alloys and Spe-
cial-Purpose Materials, Vol 2, ASM Hand-
REFERENCES book, ASM International, 1990, p 1233–
1269
1. M.J. Julio, M.L. Martin, and J.M. Baena, 12. Allegheny Ludlum research, as presented in
Cation Migration Tests in Metal Containers, D.S. Bergstrom and C.A. Botti, AL
Innovation Stainless Steel (Florence), Oct 201HPTM (UNS S20100) Alloy: A High-
1993, p 1.221–1.226 Performance, Lower-Nickel Alternative to
2. “Stainless Steel in Contact with Food,” 300 Series Alloys, Stainless Steel World,
Document Ugine, June 1996 KCI Publishing, 2005
CHAPTER 21
Marine Systems Applications
Summary restricts their use to items of rather thin gauge,

less than about 1.0 to 2.0 mm, depending on
AS RECENTLY AS THE 1960s AND 1970s, alloy. Thus, their use is limited to tubing. Super-
handbooks on stainless steel were stating that austenitic alloys can be used at any thickness,
“stainless steels are not stainless in seawater,” although they are a costly material. The success
and “successful prolonged corrosion-free serv- story for stainless steel and seawater and there-
ice of stainless steel in seawater requires sophis- fore desalination is that of duplex stainless steel.
ticated corrosion engineering, or enormous good With the same corrosion resistance as any super
fortune” (Ref 1). The advances in stainless steel austenitic or superferritic alloy, it has nearly
made since then have thankfully made these double the strength plus resistance to stress cor-
statements obsolete. Not only have basic corro- rosion cracking. And while duplex stainless steel
sion problems been solved, stress corrosion is not a cheap material, it does contain much less
cracking also can be avoided. More impres- nickel than an equivalently corrosion resistant
sively, this can be done with alloys with austenitic stainless steels, which is a major cost
strengths much higher than those of the alloys, saving factor.
such as 316, that they replace and that have been Desalination technology is relatively new if
only marginally successful in marine environ- one ignores the fact that distillation has been
ments. The inertia in changing from the weaker, around for a very long time. Desalination in com-
less-corrosion-resistant, more expensive aus- mercially viable quantities began with multi-
tenitic stainless steels is large because of less stage flash technology in the 1950s. The underly-
availability of the newer, better alloys, and lack ing principle of this process is the evaporation of
of familiarity with their benefits. Those who water vapor from salt water with the subsequent
understand and use these newer duplex alloys condensation of the salt-free water vapor. In the
will be rewarded. This chapter reviews the major multi-stage flash (MSF) approach feedwater is
marine applications of stainless steels, including heated and the pressure is lowered so that the
desalination equipment, shipping containers, water “flashes” into steam. A variation on this
and heat exchangers that handle seawater. technology is multiple-effect distillation (MED),
another low-temperature distillation process. The
differences in all distillation-based systems
Desalination reduce to the efficiency of the design in minimiz-
ing energy consumed per unit of pure water out-
At one time not long ago stainless steel was put. All distillation processes require heating of
thought to be an inadequate to marginal material the input water and some process power.
for use in seawater. Its use in heated seawater The other basic engineering approach to de-
was therefore all the more suspect. This was first salination is reverse osmosis (RO). The inven-
changed with the development of superferritic tion of polymer membranes that could separate
and superaustenitic alloys. The superferritic al- the salt ions from the water made this technol-
loys such as Seacure (UNS S44660) and 29-4C ogy possible. No thermal energy is required. The
(UNS S44735) are quite resistant to seawater, water is pumped at high pressure through these
even at high temperature. Their low toughness permeable membranes physically separating the
salt from the water. The change in salt concen- for their processes. Type 316 stainless steel has
tration across the membrane is a function of the passed from consideration as a material for han-
pressure and the membrane itself. A second dling brackish water or seawater.
treatment may be used to improve water quality. In distillation systems, the rule of thumb is
The distillation methods require about 5 that 2205 alloy (UNS S32205), with its pitting
kWh/metric ton of water output, while the RO resistance equivalent number (PREN) of 35, is
methods require twice that. The distillation sufficient for seawater up to 20 °C (70 °F); al-
methods require another 20 kWh of thermal en- loys 2507 (UNS S32750) or Zeron 100 (UNS
ergy from some source for feedwater heating, S32760) should be used for seawater at elevated
while the RO method requires none. Thus the temperatures or high salinity. For the output of
ability to find energy from cogeneration or a fresh water, lesser alloying is required. Stainless
source such as solar, etc may determine which steel types 304 (UNS S30400), 316 (UNS
process is preferred. S31600), 2101 (UNS S32101), 2003 (UNS
S32003), or even 439 (UNS S43035) may be
used depending on the combination of salinity
Materials Selection for Desalination
and temperature of the output water.
Materials used for distillation processes have Besides their high corrosion resistance for
evolved from use of type 316 (UNS S31600) lower total alloy cost, the duplex stainless steels
stainless steel, first as lining and then as have higher strength, which is a significant fac-
cladding. The superaustenitic alloys, the 6Mo tor since distillation plants are large. The use of
variations, came next because they truly solved duplex allows wall thickness reductions that
the corrosion problem, but at a price. Then, sep- bring about larger savings than those based
arately the duplex alloys were developed, with solely on their cost per unit weight. Figure 1
the first market the petroleum industry, whose shows the difference among the candidate stain-
demands and research made these alloys possi- less steels in corrosion resistance (Ref 2). The
ble. It was not a stretch to see that high-strength viable materials for seawater are those that can
alloys that could withstand seawater in offshore withstand roughly 20,000 ppm Cl– level at the
applications could do well on land as well. To appropriate temperature.
give full credit, the pulp-and-paper industry was The strengths of the various candidate materi-
also beginning to employ duplex stainless steels als are given in Table 1. These are typical values.
Fig. 1 Corrosion resistance (pitting) as a function of salinity and temperature. 1. 304L (UNS S30403); 2. 316L (UNS S31603);
3. 2205 (UNS S32205); 4. 904L (UNS N08904); 5. 254SMO (UNS S31254). Source: Ref 2
Chapter 21: Marine Systems Applications / 245
Table 1 Typical analyses and properties of major marine alloys

Composition, % Yield strength Tensile strength
Elongation,
Alloy UNS Cr Mo N Ni PREN(a) MPa ksi MPa ksi %
2101 S32101 21.5 0.3 0.22 1.5 26 515 75 650 94 40
2003 S32003 20.5 1.5 0.18 3 29 515 75 725 105 40
2205 S32205 22 3 0.17 5 35 515 75 760 110 35
2507 S32750 25 4 0.27 7 42 550 80 800 116 35
304L S30403 18 0 0.05 8 18 220 32 520 75 50
316L S31603 16 2 0.05 10 24 220 32 520 75 50
317L S31703 18 3 0.05 14 29 230 33 540 78 45
6XN N08367 21 6 0.22 24 45 380 55 760 110 45
254SMO S31254 20 6 0.20 18 43 380 55 750 109 45
Zeron 100 S32760 25 3.5 0.27 7 42 550 80 750 109 35
(+0.75 W)
(a) PREN, pitting resistance equivalent number.
Refer to the appropriate design code for your may be justified based on strength alone; their
particular application to find minimum proper- exceptional corrosion resistance would be simply
ties. The reader is cautioned that duplex longitu- an excellent side benefit. Corrugated stainless
dinal properties are slightly lower than the bulkheads are positioned within the carbon steel
transverse properties that testing requires. hull. The stiff corrugated bulkheads are them-
Pumps for seawater follow the same guide- selves structural strengtheners for the entire ship.
lines as piping, tanks, and all other components. The vertical corrugations also facilitate tank
Cast or wrought duplex are the alloys of choice. cleaning as internal stiffeners are eliminated.
Cryogenic containers are still the bastion of
austenitic stainless steels. As leaner austenitic
Shipping alloys have become less expensive than 9% Ni
alloy steel, a martensitic grade, they have be-
The major uses of stainless steel in shipping come the material of choice. In this case, the
are in bulk storage containment. Cargos range 201 types are preferred to 304 because 201 has
from food and beverages to chemicals and liquid greater strength at the cryogenic operation tem-
natural gas (LNG). Practice in the past has been perature and is, of course, less expensive. The
to use austenitic grades of stainless with cathodic expanding market for LNG has made ocean
protection when necessary to address inadequate transport increasingly important because large
corrosion resistance. However, since 2000 ma- disparities in prices often are due to the diffi-
rine chemical tankers have become the largest culty in transporting it. The two best materials
consumer of duplex stainless steel. The reason are UNS S20153 and S20400, which perform
for this is that cargo tanks ideally have the widest equally well. If higher strength is valuable to a
potential range of cargos possible. This range is design for cryogenic uses, then UNS S21904
defined by corrosion resistance. This factor alone (21-6-9 or Nitronic 40) could be used. This
is reason to choose duplex over austenitic alloys alloy has yield strengths of 460 MPa (65 ksi) at
such as 316L (UNS S31603) or 317L (UNS room temperature and 1200 MPa (175 ksi) at
S31703). An equally decisive factor is strength. –196 °C (–320 °F). It is completely resistant to
With codes permitting the tank’s design to be martensite formation.
based on yield strength, the use of duplex al- Other shipboard systems benefit equally from
loys—with strengths about double those of the use of duplex stainless steel. This extends to
austenitic steels—permits significant weight re- piping, hardware, propellers, shafts, etc.
duction. This is a major economic factor for ship
owners in that dead weight can be replaced by
fee-paying cargo at the same operating cost. Heat Exchangers
These incremental revenues, over the life of the
vessel, are many times the original cost of the Coolers for captive water systems such as for
material. Based on the high value for strength in power plants often need to resist corrosion by
ship economics, it would seem that the highest- brackish water or seawater. To the extent that
strength alloys, such as 2507 (UNS S32750), these are thin-wall tubing, ferritic alloys such as
Seacure (UNS S44660) or 29-4C (UNS S44735) sion cracking to very high temperature and
have been used quite successfully. If thicker salinity.
tubes are required, then the equivalent duplex or
austenitic alloys can be used. This would include
types 2003 (UNS S32003), 2205 (UNS S32205), REFERENCES
or 2507 (UNS S32750) duplex stainless steels,
depending on salinity and temperature; for 1. Peckner and I. Bernstein, Stainless Steel
austenitics, the 6Mo alloys such as 254SMO Handbook, McGraw-Hill, 1966, p 37-1
(UNS S31254) and AL6XN (UNS N08367) may 2. Stainless Steel for Desalination Processes,
be used. The duplex alloys have the advantage Feb 2006, Outokumpu, www.outokumpu.
of lower cost. Both are resistant to stress corro- com, accessed June 2008
CHAPTER 22
Petroleum Industry Applications
Summary ginnings. Demand for steel for drill pipe, cas-

ing, and tubing has led to many developments,
THE PETROLEUM INDUSTRY has had to such as the technology for producing high-
deal with increasingly hostile environments in quality seamless and welded pipe and tubing.
its search for new supplies of oil. And that pe- Pipeline needs have fueled the market for high-
troleum, when found, often contains harmful in- strength, low-alloy plate. Offshore production
gredients. The result is increasing demand for in often-hostile environments has presented se-
steels with greater strength and corrosion resist- vere material challenges. And, as the light sweet
ance. Martensitic and duplex stainless steels crude that was easily found and produced on
have provided the corrosion resistance and land is exhausted, future supplies of hydrocar-
strength to deal with higher levels of hydrogen bons are increasingly likely to contain sulfides,
sulfide, carbon dioxide, chlorides, and acidity. carbon dioxide, and saltwater in sufficient
This chapter reviews the selection of stainless amounts to make corrosion a top priority in se-
steels for petroleum applications, including oil lecting materials. For reference in this chapter,
country tubular goods (OCTGs), line pipe, off- Tables 1 through 5 list the relevant alloys for
shore platforms, and refinery equipment. petroleum industry applications. Many, but not
all, of these alloys are listed in the National As-
sociation of Corrosion Engineers (NACE)
Introduction MR0175, Sulfide Stress Corrosion Cracking Re-
sistant Metallic Materials for Oil Field Equip-
The petroleum industry has driven large segment; the tables in this chapter also include
ments of the steel industry since both their be- some newer alloys not in the NACE document.
Table 1 Ferritic stainless steels for petroleum Table 2 Martensitic stainless steels for
industry applications petroleum industry applications
UNS Common name Hardness,
UNS Common name HRC, max(a)
S40500 405
S40900 409 J91150 CA15 ...
S43000 430 J91151 CA15M ...
S43035 439 J91540 CA6 NM ...
S43400 434 K90941 9Cr 1Mo ...
S43600 436 S14125 S/W 13Cr 28
S44200 442 S41000 410 22
S44400 444 (18-2) S41426 13CRS ...
S44500 … S41427 … 29
S44600 446 S42000 420 22
S44626 26-1 Ti, E-Brite S42400 F6NM 23
S44627 26-1 Cb S42500 15Cr 22
S44635 26-4-4, Monit JFE KL-12G ...
S44660 Seacure, SC-1 JFE KNHP12Cr ...
S44700 29-4 Nippon NT-CRS ...
S44735 29-4C Nippon NT-CRSS ...
S44800 29-4-2 420M ...
S46800 468 L80 13 Cr ...
Note: See Appendix 1 for alloy compositions. Source: Adapted from NACE Note: See Appendix 1 for alloy compositions. (a) As specified in NACE
MR0175, “Sulfide Stress Corrosion Cracking Resistant Metallic Materials for MR0175. Source: Adapted from NACE MR0175, “Sulfide Stress Corrosion
Oil Field Equipment” Cracking Resistant Metallic Materials for Oil Field Equipment”
Table 3 Precipitation hardening stainless steels Table 4 Duplex stainless steels for petroleum
for petroleum industry applications industry applications
Hardness, UNS Common name PREN(a)
UNS Common name HRC, max(a)
J93345 Escoloy 31-47
S13800 13-8 PH ... J93370 CD4MCu 30-34
S15500 15-5 PH 33 J93380 Z100 38-46
S15700 15-7 PH 32 J93404 958 39-47
S17400 17-4 PH 33 S31200 44LN 30-36
S17700 17-7 PH ... S31260 DP3 34-43
S35000 AM-350 ... S31500 3RE60 27-31
S35500 AM-355 ... S31803 2205 (old) 30-36
S45000 Custom 450 31 S32001 19D 20-24
S45500 Custom 455 ... S32003 2003 27-31
S46500 Custom 465 ... S32101 2101 25-29
S66286 A-286 35 S32205 2205 (new) 34-38
Custom 465 (275) ... S32304 2304 23-27
Custom 475 ... S32404 U50 27-32
S32520 52N+ 37-48
Note: See Appendix 1 for alloy compositions. (a) As specified in NACE
MR0175. Source: Adapted from NACE MR0175, “Sulfide Stress Corrosion S32550 255 32-44
Cracking Resistant Metallic Materials for Oil Field Equipment” S32750 2507 38-44
S32760 Zeron 100 40-46
S32803 2803Mo 33-41
The modern dilemma that makes stainless nec- S32900 329 26-35
S32906 2906 36-45
essary is the addition presence of wet carbon S32950 7-Mo Plus 32-43
dioxide, which is extremely corrosive to carbon S32977 AF 918 39-46
S39274 DP3W 39-47
and alloy steel. As if this is not a sufficient ma-
terial problem, sometimes the wetness is from Note: See Appendix 1 for alloy compositions. (a) PREN, pitting resistance
equivalent number. Source: Adapted from NACE MR0175, “Sulfide Stress Cor-
saltwater, which further aggravates corrosivity. rosion Cracking Resistant Metallic Materials for Oil Field Equipment”
This corrosion problem is compounded by the
accelerating influence of high temperature in The chapters on martensitic and precipitation
deeper formations. What is the answer to the hardening stainless steels discuss this in detail.
corrosion problem? Inhibitors, coatings, ca- The martensitic stainless steels used for these
thodic protection, or more corrosion-resistant applications are resistant to carbon dioxide-
materials are the main responses. The first three enhanced corrosion up to partial pressures of
responses are not always practical. They also 100 atm, after which further alloying is neces-
represent an ongoing cost rather than a one-time sary. This cannot be achieved with a martensitic
cost. Each situation must be evaluated regard- structure, but the duplex alloys have the corro-
ing which is the optimal solution. sion resistance and strength to work in this
regime. They have high annealed strength and
can also be cold worked to higher strength levels.
Combating Corrosion in Alloys for If hydrogen sulfide is present, the selection
Petroleum Applications process can become more difficult. High-
strength martensitic steels are susceptible to
Alloying steel with chromium, copper, brittle delayed failure in the presence of hydro-
molybdenum, and nickel can lower the corro- gen sulfide. Being stainless does not by itself
sion rate of steel by a factor of 10,000. Figure 1 provide immunity. If localized corrosion occurs,
shows the influence of chromium alone, which hydrogen uptake ensues, and delayed failure
produces a 100-fold reduction in corrosion of follows. Only keeping hardness below well-
steel in seawater and carbon dioxide (Ref 1). established levels can render a martensitic alloy
Molybdenum is the most powerful alloying immune. If the localized corrosion can be pre-
addition to magnify the benefit of chromium. vented, however, then the stress corrosion
The effects of copper and nickel are also very cracking (SCC) cannot be initiated. Molybde-
significant, as Fig. 2 (Ref 2) indicates. These ad- num alloying expands the pH and chloride
ditions must be made in a very balanced way if range from which an alloy can be free of the pit-
a tough, fully martensitic structure is to be ting corrosion that initiates SCC, as shown in
maintained. Carbon must be kept low to avoid Fig. 3 (Ref 2). Martensitic steels of all types
the formation of chromium carbides during have a maximum in susceptibility to SCC via
tempering, which would counteract the benefit hydrogen embrittlement near room temperature.
of the chromium. Nickel is necessary to prevent Duplex alloys and austenitic alloys become
σ-ferrite formation, which reduces toughness. susceptible at higher temperatures and do not
Chapter 22: Petroleum Industry Applications / 249
Table 5 Austenitic stainless steels for petroleum industry applications

UNS Common name PREN(a) UNS Common name PREN(a)
J92500 CF-3 ... S30100 301 ...
J92600 CF-8 ... S30153 301LN ...
J92602 CF-20 ... S30200 302 ...
J92701 CF-16F ... S30215 302B ...
J92710 CF-8C ... S30300 303 ...
J92800 CF-3M ... S30400 304 ...
CF-12M ... S30403 304L ...
J93000 CG-8M ... S30409 304H ...
J93254 CK3MCuN ... S30415 153MA ...
J93402 CH-20 ... S30453 304LN ...
J94652 CN-3MN ... S30500 305 ...
N06022 AL 22 ... S30800 308 ...
N08007 CN-7M ... S30815 253MA ...
N08020 20Cb-3 28 S30900 309 ...
N08020 AL 20 28 S31000 310 ...
N08024 20Mo-4 38 S31008 310S ...
N08026 20Mo-6 ... S31254 254SMO 46
N08028 Sanicro 28 39 S31266 B66 59
N08031 Nicrofer 3127 hMo 54 S31600 316 ...
N08320 20Mod 38 S31603 316L 25
N08366 AL-6X ... S31609 316H ...
N08367 AL-6XN 49 S31635 316Ti ...
N08700 JS-700 36 S31700 317 ...
N08800 332 ... S31703 317L ...
N08925 25-6Mo 46 S31725 317LM ...
N08926 Cronifer 1925 hMo 47 S31726 317LMN ...
N08932 URSB-8 49 S31753 317LN ...
N80904 904L 39 S32100 321 ...
S20100 201 ... S32109 321H ...
S20153 201LN ... S32200 NIC 25 ...
S20200 202 ... S32654 654SMO 64
S20400 Nitronic 30 (204L) ... S33000 330 ...
S20430 204 ... S33400 334 ...
S20500 205 ... S34565 4565 54
S20910 Nitronic 50 ... S34700 347 ...
S21800 Nitronic 60 ... S34709 347H ...
S21900 Nitronic 40(219) ... S35125 332Mo ...
S21904 21-6-9 LC ... S35315 353MA ...
S24000 Nitronic 33 ... Cronifer 2328 ...
Note: See Appendix 1 for alloy compositions. (a) PREN, pitting resistance equivalent number. Source: Adapted from NACE MR0175, “Sulfide Stress Corrosion
Cracking Resistant Metallic Materials for Oil Field Equipment”
Fig. 1 Influence of chromium on the corrosion rate of steel in

environments experienced by oil country tubular
goods. Test conditions: synthetic sea water; CO2 partial pres-
sure, 0.1 MPa; test temperature, 60 (C °140 °F); test duration, Fig. 2 Influence of copper and nickel on the corrosion rate of
150 h; flow velocity, 2.5 m/s; specific volume, 800 mL/cm2. martensitic stainless alloys used for oil country tubular
SSC, stress corrosion cracking. Source: Ref 1 goods. Source: Ref 2
The NACE recommendations of suitable ma-

terials are defined by MR0175. Table 6 summa-
rizes these recommendations. The reader is en-
couraged to refer to the latest version of this
document for further details.
Oil Country Tubular Goods
Oil country tubular goods (OCTG) include

the drill pipe, casing, and tubing and associ-
ated hardware used to construct oil and gas
wells. Drill pipe is used to twist the drill bit
and convey drilling fluids to the point of con-
tact and flush away debris. Casing is put in
place to stabilize the well walls, while tubing is
placed within the casing to carry oil and gas to
the surface. Each of these components sees sig-
nificant stresses, and high strength-to-weight
materials are needed. Drill pipe is in tension,
torsion, and compression alternately throughout
its life. Casing hangs from the wellhead under
its own weight for distances from hundreds of
meters to 7000 or 8000 m and must withstand
very high collapse as well as burst pressures.
Fig. 3 Influence of molybdenum on susceptibility to stress Well strings, the exact sequence of size and
corrosion cracking in solutions containing (a) 3.5% strength pipe for each level of the well, are opti-
NaCl and (b) 0% NaCl. Source: Ref 2
mized for the conditions of each well. The vari-
ety of strengths and sizes are standardized by
the American Petroleum Institute. The use of
exhibit the same increasing susceptibility with the highest strengths has always been limited by
strength. So, when hydrogen sulfide, which en- hydrogen embrittlement accelerated by hydro-
hances hydrogen uptake, levels exceed about gen sulfide, so that the maximum hardness for a
10–2 atm, the martensitic alloys should no longer given material must be strictly adhered to when
be used, and the duplex alloys are preferred. As hydrogen sulfide is present.
temperatures and hydrogen sulfide partial pres- The terms 13Cr, 22Cr, and 25Cr are com-
sures increase, alloying must also, until at 1 atm monly used in the industry even though this
of hydrogen sulfide nickel base alloys are re- greatly oversimplifies the alloying, and there-
quired. Figure 4 shows this progression with the fore performance, options that exist. The 13Cr
alloy recommendations of Sumitomo. The re- alloys are a family of martensitic stainless
quirements behind this diagram are generic. steels. The 22Cr and 25Cr alloys are duplex
Any producer’s alloys must comply with this grades. The former are used in the quenched
diagram’s regions, which have been defined by and tempered condition, while the duplex alloys
NACE. Stainless steels are required above a are used as annealed or cold worked.
certain carbon dioxide level for all levels of hy- The 13Cr grades began as simply variations
drogen sulfide. Martensitic alloys, commonly on 420, which is a straight-chromium marten-
called “13Cr,” are the first step up from alloy sitic often used for cutlery. This alloy, while far
steels. At higher levels of carbon dioxide and better (about 100 times) than alloy steel in cor-
hydrogen sulfide, duplex alloys are required, rosion resistance, has nearly the least corrosion
with the 22CR alloys such as UNS S32205 used resistance of all stainless steels. To achieve
at temperatures up to 200 °C (390 °F) and the higher corrosion resistance molybdenum is
25CR alloys such as UNS S32507 at tempera- added. Molybdenum at 1% increases resistance
tures up to 250 °C (480 °F). Nickel-base alloys to general corrosion in a sodium chloride/
are required at hydrogen sulfide levels above 1 hydrogen sulfide/carbon dioxide environment by
atm partial pressure. about tenfold. Another 1% increases it another
Fig. 4 Alloy suitability as a function of H2S and CO2 partial pressure. Source: Ref 1
tenfold. The 2% level of molybdenum also Martensitic alloys are susceptible to SCC by a
greatly reduces pitting, which in turn eliminates hydrogen embrittlement mechanism. This sus-
the initiation point of SCC. Simply adding ceptibility is strongly temperature dependent. It
molybdenum would cause the alloy to have ex- decreases with temperature from a maximum at
cessive δ-ferrite, which cannot transform to ambient to none at around 100 °C (210 °F). If the
martensite and would therefore reduce mechan- hydrogen sulfide level exceeds 0.03 atm, then
ical properties. Thus, nickel must be added to 22Cr alloys should be used rather than 13Cr
counter the ferrite stabilizing effect, unfortu- because of this risk. Hydrogen sulfide may be
nately, but necessarily increasing the cost. The contained in the petroleum, or it may come from
nickel does help lower the general corrosion sulfate-reducing bacteria, introduced by flooding,
rate. Carbon and nitrogen in these alloys are for example. This can cause a sulfide-free system
kept at low concentrations. These alloys are oth- to become sulfide rich after the fact and make ini-
erwise almost identical to precipitation harden- tial materials choice wrong after the fact.
ing martensitic stainless steels without the pre- The 22Cr and 25Cr alloys have significantly
cipitating phase. higher resistance to chlorides and wet hydrogen
Table 6 Restrictions in use recommended by NACE MR0175 for selected stainless steels used for
petroleum industry applications
Common Temperature, Hardness, Cl–,
UNS name PREN(a) ºC (ºF) pH, min H2S, kPa HRC mg/L
J91150 CA15 ... ... 3.5 10 22 ...
J91151 CA15M ... ... 3.5 10 22 ...
J91540 CA6NM ... ... 3.5 10 23 ...
J93254 ... ... ... ... ... 100 HRB ...
J95370 ... ... 150 (300) ... 700 94 HRB 90,000
N08926 ... ... 121 (250) 3.5 700 ... 60,700
S15500 15-5 ... ... ... 3.4 33 ...
S15700 15-7 ... ... ... ... 32 ...
S17400 17-4 ... ... ... 3.4 33 ...
S20910 ... ... 66 (150) ... 100 ... 35
S41000 410 ... ... 3.5 10 22 ...
S41425 ... ... ... ... 10 28 ...
S41426 ... ... ... 3.5 10 27 ...
S41427 ... ... ... 3.5 10 29 6,000
S41429 ... ... ... 4.5 10 27 ...
S41500 F6NM ... ... 3.5 10 23 ...
S42000 420 ... ... 3.5 10 22 ...
S42400 ... ... ... 3.5 10 23 ...
S42500 ... ... ... 3.5 10 22 ...
S45000 450 ... ... ... 1 31 ...
S66286 A286 ... 65 (150) ... 100 35 ...
Austenitic A-2 ... ... ... ... 22 ...
Duplex <40 232 (450) ... 10 ... ...
>40 <45 232 (450) ... 20 ... ...
Superaustenitic, (Ni+2Mo) >30 ... ... ... ... ...
type 3a
Superaustenitic, >40 121 (250) ... 700 ... 5,000
type 3b 149 (300) ... 310 ... 5,000
171 (340) ... 100 ... 5,000
Notes: See NACE MR0175 for further use and processing restrictions. See Appendix 1 for alloy compositions. (a) PREN, pitting resistance equivalent number. Source:
Adapted from NACE MR0175, “Sulfide Stress Corrosion Cracking Resistant Metallic Materials for Oil Field Equipment”
sulfide and can resist SCC at ten times higher gen. The more common alloys are UNS S32654
concentrations than the 13Cr alloys. Besides the and N08367.
inherently greater corrosion resistance that de- The recent development of lean duplex alloys
rives from the chromium, molybdenum, and ni- has not yet made its way into OCTGs. These al-
trogen levels of the 22Cr and 25 Cr alloys (see loys offer an inherent alloy savings over the
the chapters on corrosion in this book), the du- 13Cr grades in nickel and molybdenum content
plex alloys have very fine grain size and a while offering better corrosion and SCC resist-
roughly 50/50 mixture of ferrite and austenite. ance. Their strength levels in the annealed con-
This acts as a crack arrestor should one phase be dition, 450 MPa (65 ksi), are lower than those
susceptible to cracking while the other is not. of the martensitic alloys, 600 MPa (87 ksi), so
There have been no reported downhole fail- for most downhole applications they will re-
ures of annealed or cold-worked duplex alloys. quire cold working. It is likely, however, that
There was one instance of very high-strength these alloys will see their first service as line
tubing cracking after cathodic contact with car- pipe, where they will not need to be cold worked
bon steel casing. This was after removal from to higher strength levels to be widely used.
the well and after handling damage had oc-
curred. The affected microstructure was found
to be high (70%) in ferrite (Ref 3). Line Pipe and Flow Lines
For corrosion resistance above that furnished
by superduplex materials such as the 25Cr al- With the awesome cost of corrosion, the
loys, super austenitic alloys fill a gap before case for stainless line pipe is easily made.
nickel base alloys are needed. These alloys Whether to use stainless depends on vulnera-
achieve a tenfold increase in hydrogen sulfide bility of carbon steel. This evaluation is made
resistance and very elevated SCC resistance. based on the carbon dioxide, hydrogen sulfide,
These are the so-called 6Mo grades. The more water, salinity, temperatures, pressures, flow
advanced of them contain high levels of nitro- conditions, and so forth. The normal basis for
these calculations follows that published by C. and nitrogen, both relatively inexpensive alloy-
de Waard of Shell (Ref 3). The competing tech- ing elements. Alloys S32001 and S32101 are
nology when corrosion dangers arise is the use well formulated for medium and high levels of
of corrosion inhibitors, cathodic protection, or corrosion resistance required for wet carbon
to line carbon steel with a protective coating. dioxide, hydrogen sulfide, and trace chlorides.
The use of inhibitors is subject to the risk of ve- The main precaution for duplex alloys is main-
locity limitations, temperature limitations, and taining a nominally 50/50 mixture of ferrite and
simply of the inhibitor working appropriately, austenite with no embrittling intermetallic
not to mention cost. Cathodic protection is phases. The modern alloys have high (greater
costly and complex. Coatings can be damaged than 0.14%) nitrogen, which helps to preserve
by numerous occurrences in acidity, mechani- austenite levels after welding and suppress in-
cal damage, or fluctuations in temperature or termetallic formation. Nevertheless, minimiza-
pressure. Use of a stainless corrosion-resistant tion of time above 350 °C (660 °F) is important.
alloy can have well-defined and controlled This tendency increases with chromium and
costs and performance over the life of an instal- molybdenum content, which is another reason
lation. Line pipe differs from downhole in hav- why the lean duplex alloys are so attractive.
ing strength requirements more in line with that For subsea use, 22Cr duplex generally re-
of annealed duplex alloys. These requirements quires cathodic protection because of the risk of
gave birth to modern duplex alloys, starting crevice corrosion. 25Cr duplexes are used with-
with UNS S31803 and evolving to UNS out cathodic protection. Duplex pipelines have
S32205 as the value of higher nitrogen became been in service in the North Sea since the 1970s.
understood.
Nitrogen not only enhances corrosion resist-
ance, but also suppresses the formation of unde- Umbilical Tubing and Risers
sirable and embrittling intermetallic phases that
might otherwise form at welding temperatures. Increasingly, wells are located undersea. It is
It also keeps the desirable austenite/ferrite ratios standard practice to control and monitor these
in weld metal. wells via bundled umbilical tubing. The tubing
Since the development of the first widely ac- can provide hydraulic and electrical power, con-
cepted duplex alloys, more alloys have trol and adjust pressure, carry communications,
emerged. Superduplexes, such as UNS S32750, and even introduce chemical to the well. The
have become accepted alloys. Then, the need to depth of wellheads can increase collapse pres-
improve costs led in the 1990s to the use of sures to levels beyond the capability of thermo-
martensitic alloys with high levels of nickel and plastics, which has led to the use of duplex
molybdenum, which at the time were lower stainless steel because of its strength and resist-
cost. The emergence in the early 2000s of lean ance to corrosion and SCC. When resistance to
duplex alloys provided strength and more corro- seawater is the main concern, the rule of thumb
sion resistance with lower nickel levels, giving is that a pitting resistance equivalent number
them a cost advantage during periods of high (PREN) of 35 or greater is required, whereas re-
nickel cost. sistance to crevice corrosion requires a PREN
The main attribute required by line pipe that of at least 40. This has made the superduplex
is not as important in OCTGs is weldability. UNS S32750 the standard. Such a critical item
This is not an overwhelming challenge for du- as an umbilical may seem like a poor applica-
plex alloys, but for martensitic alloys, it re- tion on which to economize, but again the lean
quires a very low interstitial level so that the duplexes offer possibilities to do so. By zinc
martensite is self-tempering and ductile in the coating lean duplexes such as alloy 19D (UNS
as-welded condition. This can be achieved by S32001) and 2101 (UNS S32101), very long
stabilizing the alloy with small amounts of tita- service lives can be safely extrapolated. These
nium. It would appear that under current condi- alloys are being promoted on their lower sus-
tions that alloy 2101 (UNS 32101) has a ceptibility to σ formation during welding, and if
cost/performance edge over the martensitic welding thermal cycles cannot be controlled
competition and should for the long term. The that may be an issue, but superduplex seems to
main ingredients required in a duplex for have become a pervasive choice because it is
strength and corrosion resistance are chromium superbly reliable.
Risers are now produced in coiled tubing of Liquefied Natural Gas Vessels
over 100 mm (4 in.) diameter, so that very eco-
nomical long lengths are feasible. Liquefied natural gas (LNG) is becoming an
increasingly important commodity as the value
of stranded gas makes it economically desirable
Platforms to convert it to a transportable state. Converting
natural gas to a cryogenic liquid presents a ma-
Platforms present a special case in which the terial problem. Vessels to contain it must have
costs of maintenance are high, the corrosion en- strength and toughness at temperatures below
vironment is severe, and the penalty for excess –150 °C (–240 °F). The traditional material, 9%
weight is also high. A savings of 1 ton in weight Ni martensitic steel, has become expensive
topside can save over $100,000 in steel in the compared to the lower-nickel austenitic stain-
subsea jacket. This leads to a rapid payback for less steels, such as 201LN (UNS S20153),
the use of materials that are sufficiently resistant which have no transition temperature and
to corrosion such that corrosion loss allowance strengthen with decreasing temperature. Alloy
can be eliminated. Both titanium and stainless 201LN is cheaper, easier to weld and fabricate,
alloy UNS S32750 are equal candidates for this and of course is stainless, which 9% Ni steel is
service, depending on availability and current not. The extreme ductility of 201LN compared
alloy prices. Except in rare cases, stainless steel to martensitic steel gives it a decided advantage
wins the cost battle between these alloy systems. in terms of rupture resistance, which is a major
Almost any structure is a candidate for stain- design and political concern with this poten-
less topside processing: piping, pumps, flanges, tially explosive commodity.
fittings, etc. Hardware of any type and construc-
tion materials benefit from being stainless. Sea-
water systems often employ 22Cr duplex with Refinery Equipment
cathodic protection or unprotected 25Cr duplex.
A wise preventive action is to paint stainless Corrosion resistance is a major factor in the
that is covered by insulation or similar material, choice of materials in refinery operations. As we
which otherwise can result in concentration discussed, crude oil itself is sometimes a very
cells and consequent pitting. corrosive fluid, but in refining the by-products,
Table 7 Stainless steels used in various refinery processes

Process Corrosive agents Applications Alloys Notes
Crude distillation Sulfur-containing acids Preheaters, distillation 405, 409, 410 …
(SCAs) tower
Vacuum fractionalization SCA, chlorides Towers 405, 410, 316 Depending on crude
corrosivity
Condensers S44735, 2205 Depending on chloride level
Coker SCA, H2S Coke drums 409 Depending on crude corro-
sivity
Gas plants H2S, water, Cl–, ammonia Compressor coolers, AL-6XN, 2205, 2507 …
reboiler tubes
Trays 410S, 316L …
Amine plant Ammonia, MEA, DEA Reboilers, trays, filters, 304L, 316L …
condenser tubing
Sulfuric acid alkylation Sulfuric acid Contactor, mixer 20Cb3 Low pH excursions possible
Dilute sulfuric acid Effluent piping 316L …
Hydrotreating H2S, ammonia, PTA(a) Hot sections 321, 347 Long exposure at high tem-
perature
General 410S, 304 …
Catalytic re-forming High-temperature strength Reactor internals 304 HCl catalyst regeneration
needed
HCl residue Heat exchangers 2205 …
Fluid catalytic cracking High temperature Trays 410 Condensers may need 6Mo
Cyclones, vapor lines 304 …
Hydrogen plant Tubing, heat exchangers 304 …
Hydrocracking Sulfides, chlorides Heat exchangers 409, 321, 347, 2205, Depending on temperature,
6Mo risk of chlorides
Sour water stripping Sulfuric acid, ammonium Stripper 304, 20Cb3, 2205 Severity depends on pres-
bisulfide, chlorides ence of sulfuric acid
(a) PTA, polythionic acid. Source: Ref 4
chemicals used in refining and the temperatures using austenitics, which are prone to grain
used may further aggravate that corrosivity. The boundary chromium depletion by sensitization,
aggressive chemical agents that refinery materi- and instead use low-carbon grades and stabi-
als must withstand include wet hydrogen sulfide lized grades.
and carbon dioxide, napthenic acids, polythionic
acids, chlorides, sulfuric acid, and alkalines as
well as simple oxidation. Sometimes, tempera- REFERENCES
tures of use are such that embrittling or sensitiz-
ing phase transformations may occur. Table 7 1. Sumitomo Products for the Oil and Gas In-
lists some major refinery processes and the ma- dustries, www.sumitomometals.co.jp, ac-
terials used in them (Ref 4). cessed June 2008
Most of these situations are discussed else- 2. H. Asahi et al., “Development of High
where in this book in detail. One that is quite Chromium Stainless Line Pipe,” Nippon
specific to refinery applications is polythionic Steel Technical Report 72, January 1997
acid (PTA) attack. These acids usually form ac- 3. C. de Waard and U. Lotz, “Prediction of
cidentally when sulfide corrosion products react CO2 Corrosion of Carbon Steel,” Paper 69,
with moisture and air. The attack is intergranu- presented at Corrosion/93, National Associ-
lar, and materials respond to it much as they do ation of Corrosion Engineers, 1993
to the Strauss test. The remedies are to prevent 4. C.P. Dillon, Corrosion Resistance of Stain-
the inadvertent formation of PTA and to avoid less Steels, Marcel Dekker, 1995
CHAPTER 23
Chemical and Process Industry

Applications
Summary Single- and Dual-Environment Systems
ENGINEERS IN THE PROCESS industries Under ideal conditions, a material may need
must have materials that can contain a huge vari- to resist one single major corrosion threat. If the
ety of chemical species at many temperatures, most potentially damaging species can be
pressures, and flow rates. This is applied corro- clearly identified, then candidate materials can
sion engineering combined with physics and be found by searching published data. These
structural design. It is obvious that this task de- data are available freely online from Web sites
pends on the availability of corrosion data, more (such as Ref 1 and 2) or for a charge from
than can be presented here. This chapter covers sources such as the National Association of
what data are necessary and how they can be Corrosion Engineers (NACE; Ref 3) and ASM
found. International (Ref 4). It is difficult for any pub-
lished data to keep up with the latest develop-
ments. The testing alone of new materials can
Introduction take a long time, and then it must wait for publi-
cation. All materials are not covered, especially
The need to work with hostile chemicals be- when a manufacturer publishes data on propri-
gins with the manufacture of those chemicals. It etary alloys and excludes competitive materials.
was in the production of nitric acid that stainless That having been said, any improvements over
had its first industrial application. These are in- standard alloys will first be reported by the de-
dustries with purely need-driven material chal- velopers of the alloy, and they will logically
lenges. New processes are constantly in devel- tout its strongest points. For this reason, dia-
opment, and they present new environments in logue with the primary steel producers is en-
which materials must perform. The choices are couraged. No one has more exposure to the lat-
highly pragmatic. In an industrial environment, est trends in applications.
the costs of a poorly performing material can be A single-environment system is typically one
well known by its effect on downtime, mainte- in which the aggressive chemical species is the
nance, liability, etc. The essential knowledge is only consideration. This is normally the case for
which materials will work. piping, tanks, or reaction vessels holding the
The selection of materials for the chemical species or materials immersed in the aggressive
and power industries is first a study of corrosion species.
resistance, including resistance to stress corro- A dual-environment system is typically en-
sion cracking (SCC). Strength plays a secondary countered in heat exchangers, but it must also
role but can be an important cost factor. These be extended to single-environment systems in
considerations may occur at very high or very which the exposure of the nonreactant side of
low temperatures, in which case corrosion resist- the material to the ambient environment cannot
ance may become oxidation resistance and be neglected, as in the case of marine ambient
strength may mean creep strength. environments.
The challenges that must be met are primarily

ensuring adequate corrosion resistance and sec-
ondarily having acceptable mechanical proper-
ties. The corrosion issues run the full gamut of
potential forms of corrosion:
• General corrosion
• Pitting corrosion and crevice corrosion
• Intergranular corrosion
• Stress corrosion cracking
• Erosion corrosion
In addition to these forms of corrosion associ-
ated with liquids, there are considerations of gas
phase attack, which may be oxidation, sulfida-
tion, or attack by other gases.
Mechanical design considerations are nor-
mally limited to static stress allowances. Previ- Fig. 1 Isocorrosion chart for sulfuric acid. Source: Ref 1
ously, handbooks dealt very lightly with this
topic because all the normally recommended of carbon steel and titanium gives a valuable
steels had similar strength. The proliferation of frame of reference for the engineer.
duplex stainless steels has changed that. Now, If the forms of localized corrosion discussed
high-strength alloys of high corrosion resistance next can be avoided, the corrosion tables are
and SCC resistance are available and are mak- sufficient to guide the designer to a reasonable
ing traditionally chosen stainless steels less than selection of candidate materials for any process
optimal. in which the chemical species involved have
been identified. If the data have not been devel-
oped for a certain environment, then the tables
Corrosion Types give a first approximation of which materials
may be resistant from examination of similar
A designer wants to deal with general corro- environments, and a final decision can only be
sion. Its rate can be predicted, and thickness reasonably made though direct corrosion testing
can be chosen to allow for its occurrence. Cor- of candidate materials. Refer to the chapters on
rosion data for general corrosion are normally corrosion for a more thorough discussion of
presented in isocorrosion charts. These present uniform corrosion.
the temperatures and concentrations for a
given environment at which various materials Pitting and Crevice Corrosion
will exhibit the same corrosion rate. This rate
is most often 0.1 mm/yr, an amount that can be Stainless steel is unique among metals and al-
thought of as a tolerable level for many uses. loys in that it derives its corrosion resistance
Figure 1 shows an isocorrosion chart for stain- from constituent alloying elements working to-
less steels in sulfuric acid (Ref 1). The data are gether to form a thin passive layer that, when
clear when presented in this fashion. It can fur- intact, is highly resistant to corrosion. The
ther be appreciated that in general reducing strength of the passive layer in resisting attack
alloy performance to a mathematical formula, by halide ions, which are the most disruptive
such as the pitting resistance equivalent num- ions to the layer, is proportional principally to
ber (PREN) equation, would not be reasonable the chromium, nitrogen, and molybdenum con-
since the relative performance of alloys tents of the alloy. This relationship follows the
changes considerably with concentration. formula:
Thus, the design engineer must rely on experi- PREN = %Cr + 3.3%Mo + 30%N (Eq 1)
mentally developed data. Since these data are
available both online and in print, no attempt This formula is one of the commonly used
will be made to reproduce them fully here. Ex- versions, none of which is universally correct.
amples are given in Tables 1 and 2 (Ref 1). Both tungsten and carbon can increase pitting
Such tables are very useful, although the pre- resistance, while sulfur diminishes it. This is
sentation is not visually compact. The inclusion discussed in the corrosion section of this book.
Chapter 23: Chemical and Process Industry Applications / 259
The important consideration is that this formula resistance, which is what causes pitting.
assumes that the key alloying elements are ho- Chromium is very reactive: Its affinity for oxy-
mogeneously distributed in solution. This will gen makes the passive film strong. Pitting has
only be true if correct thermomechanical pro- nearly always been associated with manganese
cessing occurs because, thermodynamically, sulfide inclusions, and although there is still de-
these alloys are not used in an equilibrium con- bate over the precise mechanism, it appears that
dition. Were they to attain equilibrium, say by chromium depletion at the metal-inclusion in-
overheating, alloy segregation by precipitation terface is to blame. Eliminating inclusions by
could occur, causing localized loss of corrosion eliminating either manganese or sulfur improves
Table 1 Corrosion table for sulfuric acid (H2SO4)

Concentration, % 0.1 0.5 0.5 0.5 1 1 1 1 1 2 2 2 3 3 3
Temperature, °C 100 = BP 20 50 100 = BP 20 50 70 85 100 = BP 20 50 60 20 35 50
Carbon steel 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
13% Cr steel 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
18-2 (UNS S44400) 2 0 2 2 0 2 2 2 2 0 2 2 0 2 2
3R12 (UNS S30400) 2 0 1 2 0 1 1 2 2 0 1 1 0 1 1
3R60 (UNS S31600) 1 0 0 1 0 0 0 1 1 0 0 0 0 0 0
18-13-3 1 0 0 1 0 0 0 1 1 0 0 0 0 0 0
17-14-4 1 0 0 1 0 0 0 0 1 0 0 0 0 0 0
2RK65 0 0 0 1 0 0 0 0 1 0 0 0 0 0 0
(UNS N08904)
Sanicro 28 ... 0 0 0 0 0 0 0 0 0 0 0 0 0 0
(UNS N08028)
254SMO ... 0 0 ... 0 0 0 0 1 0 0 0 0 0 0
(UNS S31254)
654 SMO ... 0 0 ... 0 0 0 0 0 0 0 0 0 0 0
(UNS S32654)
SAF 2304 1 0 0 ... 0 0 0 0 1 0 0 0 0 0 0
(UNS S32304)
SAF 2205 ... 0 0 1 0 0 0 0 ... 0 0 0 0 0 0
(UNS S31803)
SAF 2507 ... 0 0 ... 0 0 0 0 0 0 0 0 0 0 0
(UNS S32750)
Titanium 1 0 0 1 0 0 1 1 1 0 0 1 0 0 1
Concentration, % 3 3 5 5 5 5 5 5 10 10 10 10 10 20 20
Temperature, °C 85 100 = BP 20 35 60 75 85 101 = BP 20 50 60 80 102 = BP 20 40
Carbon steel 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
13% Cr steel 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
18-2 (UNS S44400) 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
3R12 (UNS S30400) 2 2 1 1 2 2 2 2 2 2 2 2 2 2 2
3R60 (UNS S31600) 1 2 0 0 1 1 2 2 0 1 1 2 2 0 1
18-13-3 1 2 0 0 0 1 2 2 0 1 1 2 2 0 1
17-14-4 1 2 0 0 0 1 2 2 0 0 1 2 2 0 1
2RK65 (UNS 0 1 0 0 0 0 1 2 0 0 0 1 2 0 0
N08904)
Sanicro 28 ... 1 0 0 0 0 0 2 0 0 0 0 2 0 0
(UNS N08028)
254SMO ... 1 0 0 0 0 1 2 0 0 0 0 2 0 0
(UNS S31254)
654 SMO 0 0 0 0 0 0 0 2 ... ... 0 0 ... 0 0
(UNS S32654)
SAF 2304 ... 1 0 0 0 0 0 2 0 0 0 2 2 1 2
(UNS S32304)
SAF 2205 ... 1 0 0 0 0 0 2 0 0 0 1 2 0 0
(UNS S31803)
SAF 2507 ... 1 0 0 0 0 ... ... 0 0 0 0 2 0 0
(UNS S32750)
Titanium 1 2 0 1 1 2 2 2 1 2 2 2 2 2 2
(continued)
Notes: 0, corrosion rate of less than 0.1 mm/yr. The material is corrosion proof. 1, corrosion rate of 0.1–1.0 mm/yr. The material is not corrosion proof but useful in cer-
tain cases. 2, corrosion rate of more than 1.0 mm/yr. Serious corrosion. The material is not usable. BP, boiling solution. Source: Adapted from Ref 1
Table 1 (continued)
Concentration, % 20 20 20 20 30 30 30 30 40 40 40 40 50 50 50
Temperature, °C 50 60 80 100 20 40 60 80 20 40 60 90 20 40 70
Carbon steel 2 2 ... 2 2 2 2 ... 2 2 2 2 2 2 2
13% Cr steel 2 2 ... 2 2 2 2 ... 2 2 2 2 2 2 2
18-2 (UNS S44400) 2 2 ... 2 2 2 2 ... 2 2 2 2 2 2 2
3R12 (UNS S30400) 2 2 ... 2 2 2 2 ... 2 2 2 2 2 2 2
3R60 (UNS S31600) 1 2 ... 2 1 2 2 ... 2 2 2 2 2 2 2
18-13-3 1 1 ... 2 1 1 2 ... 2 2 2 2 2 2 2
17-14-4 1 1 ... 2 1 1 2 ... 2 2 2 2 2 2 2
2RK65 0 0 1 2 0 0 1 ... 0 0 1 2 0 0 2
(UNS N08904)
Sanicro 28 0 0 ... 2 0 0 1 ... 0 0 1 2 0 0 1
(UNS N08028)
254SMO 0 0 ... 2 0 0 1 2 ... 1 ... ... 0 1 ...
(UNS S31254)
654 SMO 0 0 0 2 ... ... ... ... 0 0 0 ... 0 0 ...
(UNS S32654)
SAF 2304 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
(UNS S32304)
SAF 2205 0 1 2 2 0 1 2 2 2 2 2 2 2 2 2
(UNS S31803)
SAF 2507 0 0 1 2 ... 0 1 2 0 1 2 2 1 1 2
(UNS S32750)
Titanium 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
Concentration, % 60 60 60 70 70 70 80 80 80 85 85 85 85 90 90
Temperature, °C 20 40 70 20 40 70 20 40 60 20 30 40 50 20 30
Carbon steel 2 2 2 2 2 2 2 2 2 0 1 2 2 0 1
13% Cr steel 2 2 2 2 2 2 2 2 2 1 1 2 2 0 1
18-2 (UNS S44400) 2 2 2 2 2 2 2 2 2 1 1 1 2 0 1
3R12 (UNS S30400) 2 2 2 2 2 2 2 2 2 1 1 1 2 0 0
3R60 (UNS S31600) 2 2 2 2 2 2 1 2 2 1 1 1 2 0 0
18-13-3 2 2 2 2 2 2 1 2 2 1 1 1 2 0 1
17-14-4 2 2 2 2 2 2 1 2 2 1 1 1 2 0 1
2RK65 0 1 1 0 1 1 0 1 2 0 0 1 1 0 0
(UNS N08904)
Sanicro 28 0 0 1 0 0 1 ... 1 1 0 0 0 0 0 0
(UNS N08028)
254SMO 0 1 ... 0 1 ... 0 1 2 0 ... ... ... 1 ...
(UNS S31254)
654 SMO 0 1 ... 0 1 ... ... ... ... ... ... ... ... 1 ...
(UNS S32654)
SAF 2304 2 ... ... ... ... ... ... ... ... 1 1 ... ... ... 1
(UNS S32304)
SAF 2205 2 2 2 1 ... ... 2 2 2 1 ... ... ... 1 1
(UNS S31803)
SAF 2507 ... ... ... ... 2 2 ... 2 2 1 1 ... ... 0 0
(UNS S32750)
Titanium 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
Concentration, % 90 90 94 94 94 94 96 96 96 96 98 98 98 98
Temperature, °C 40 70 20 30 40 50 20 30 40 50 30 40 50 80
Carbon steel 2 2 0 2 2 2 0 1 2 2 1 1 2 2
13% Cr steel 2 2 0 1 2 2 0 1 2 2 1 1 2 2
18-2 (UNS S44400) 2 2 0 0 2 2 0 0 1 2 0 1 2 2
3R12 (UNS S30400) 2 2 0 0 1 1 0 0 0 1 0 0 2 2
3R60 (UNS S31600) 1 2 0 0 0 1 0 0 0 1 0 0 0 2
18-13-3 1 2 0 0 1 1 0 0 1 1 0 0 1 2
17-14-4 1 2 0 0 1 1 0 0 1 1 0 0 1 2
2RK65 1 2 0 0 1 1 0 0 1 1 0 1 1 2
(UNS N08904)
Sanicro 28 0 1 0 0 0 0 0 0 0 1 0 0 0 1
(UNS N08028)
(continued)
Table 1 (continued)
Concentration, % 90 90 94 94 94 94 96 96 96 96 98 98 98 98
Temperature, °C 40 70 20 30 40 50 20 30 40 50 30 40 50 80
254SMO 1 ... ... ... ... ... 1 ... ... ... ... ... 0 2
(UNS S31254)
654 SMO 2 2 ... ... ... 2 0 1 ... 2 ... ... 1 1
(UNS S32654)
SAF 2304 1 ... ... ... ... ... 1 ... ... 0 ... ... 0 1
(UNS S32304)
SAF 2205 1 ... 0 ... ... ... 0 0 1 ... 0 0 1 1
(UNS S31803)
SAF 2507 0 ... 0 0 0 1 0 0 0 1 0 0 0 1
(UNS S32750)
Titanium 2 2 2 2 2 2 2 2 2 2 2 2 2 2
Table 2 Corrosion table for fuming sulfuric When a crevice is permitted to exist, it mim-
acid (oleum), H2SO4 + SO3 ics the pH-altering action found within pits in
which transport restriction leads to a buildup of
Conc. H2SO4, % 100 100 100 100 100 100
Conc. SO3, % 7 11 11 60 60 60 metal and hydrogen ions and oxygen depletion.
Temperature, °C 60 60 100 20 70 80 All alloys undergo crevice corrosion under
Carbon steel 0 0 2 ... ... ... less-aggressive conditions than those required
13% Cr steel 0 0 2 ... ... 2
18-2 0 0 ... ... ... ... to induce pitting, so care must be taken to avoid
(UNS S44400) crevices.
3R12 0 0 1 0 0 0
(UNS S30400)
3R60 0 0 0 0 0 0 Intergranular Corrosion
(UNS S31600)
18-13-3 0 0 ... ... 0 ... Intergranular corrosion is a problem that can
17-14-4 0 0 ... 0 0 ...
2RK65 0 0 ... 0 0 ... be avoided entirely by correct alloy selection
(UNS N08904) and proper thermal processing. The principle
Sanicro 28 (UNS ... ... ... ... ... ... cause of grain boundary attack is alloy deple-
N08028)
254SMO ... ... ... ... ... ... tion at the grain boundaries. The most familiar
(UNS S31254) form of this problem occurs when austenitic
654 SMO ... ... ... ... ... ...
(UNS S32654)
alloys having carbon levels above 0.03% are
SAF 2304 ... ... ... ... ... ... welded. The region near the weld where tem-
(UNS S32304) peratures reach 600 to 900 °C (1100 to 1650 °F)
SAF 2205 ... ... ... ... ... ...
(UNS S31803) may have carbon migrate to and along grain
SAF 2507 ... ... ... ... ... ... boundaries, the fast diffusion paths, where it
(UNS S32750)
Titanium 2 2 2 2 2 2
combines with less-mobile chromium atoms
and precipitates as chromium carbide. This
Notes: 0, corrosion rate of less than 0.1 mm/yr. The material is corrosion proof. 1,
corrosion rate of 0.1–1.0 mm/yr. The material is not corrosion proof but useful in lowers the chromium level in solution, result-
certain cases. 2, corrosion rate of more than 1.0 mm/yr. Serious corrosion. The ing in poor corrosion resistance only at the
material is not usable. Conc., concentration. Source: Adapted from Ref 1
grain boundaries. This is easily prevented by se-
lecting alloys with low carbon levels. Duplex
the potential at which passive film breakdown alloys, curiously, undergo chromium carbide
occurs. This is especially important for welds, precipitation under the same conditions but do
which, if not annealed, can have maximum not undergo significant chromium depletion
deleterious segregation by both inclusions and because the neighboring ferrite grains, in which
solidification segregation. All austenitic and chromium diffuses more rapidly, contribute
duplex stainless alloys have best corrosion re- chromium, mitigating the depletion. Precipita-
sistance when quenched from the solution an- tion segregation of all types, not just by carbides,
nealing temperature. The precipitation harden- must be guarded against. Sigma phase, nitrides,
ing, martensitic, and ferritic alloys are more secondary austenite, and others can cause local
complicated but are less relevant to this topic. If breakdown of corrosion resistance if alloys are
information on them is needed, they are dis- heated to a dangerous temperature for sufficient
cussed in detail in their respective chapters. time. It is important to learn these potential
vulnerabilities by reviewing the metallurgy of austenitics. Designing within this limit is sensi-
any alloy selected for service. ble practice. And, if alloy selection uses a rule
of avoiding situations in which pitting can
occur, SCC will also be avoided even if stress
Stress Corrosion Cracking excursions occur since in general pitting is a
necessary precondition for SCC.
The theory of SCC is still under debate. The
reader will find the arguments confusing as the
debate generates more heat than light. We will Erosion
skip the theory; it can be found in the corrosion
chapters. SCC, like excessive general corrosion Flow velocities can reach levels at which ero-
or pitting, is avoided by referring to published sion becomes problematic, especially if hard
test data from the corrosion tables. If a material particles are suspended in a fluid. Assuming that
must be used where a risk of SCC occurs, then the material has sufficient corrosion resistance
stress levels must be managed to stay below the to survive well in the static environment, the
threshold stress for SCC. Figure 2 shows how best performance under erosive conditions is
various alloys resist SCC as a function of chlo- obtained by materials with higher surface hard-
ride concentration and temperature, the two ness. Accordingly, the duplex perform better
most important aggravating factors. Material than austenitic alloys of the same corrosion re-
comparisons are made difficult because tests are sistance level.
normally run at a given fraction of a material’s
yield strength. Thus, the data in Fig. 2 (Ref 1)
must be interpreted. Higher-strength duplex al-
loys, while having better SCC performance than Specific Environments
austenitics of equal corrosion resistance (e.g.,
316 vs. 2304), have much better SCC resis- The list of specific environments against
tance. Furthermore, the stress at which failure which stainless steels are sufficiently resistant
will occur is much higher since the yield to select for use in the chemical process indus-
strength at which the testing takes place is about tries is too long to provide here. Some of the
twice as high for duplex alloys. SCC also ex- most important specific corrosives, such as
hibits a threshold stress below which failure nitric, sulfuric, phosphoric, hydrochloric, and
does not occur. This is about 60% of tensile organic acids and others, are covered in the
strength for duplex and about 30% for chapter on corrosion. The main caution to the
designer is to make sure that the source mate-
rial from which design guidance is sought is
current. Many otherwise excellent handbooks
are somewhat obsolete in that they do not
include the very importance duplex stainless
steel family or only include the oldest alloys in
the group, such as 2205 (UNS S32205). Many
new alloys now exist that range in corrosion
performance from that of 316 to that of the
6Mo-plus-N austenitics. These alloys are us-
able in all gauges, have high strength and
toughness, resist SCC, and can achieve the
corrosion resistance levels of any ferritic or
austenitic alloy. They can also provide signifi-
cant savings in alloy cost at the same corrosion
level because they have lower nickel levels.
REFERENCES
Fig. 2 Stress corrosion cracking (SCC) resistance in neutral
chloride solutions containing 8 ppm oxygen. Testing
time, 1000 h. Applied stress equal to proof strength at testing 1. Sandvik Materials Technology, www.smt.
temperature sandvik.com, accessed June 2008
2. Outokumpu Corrosion Handbook for Stain- 4. D.B. Anderson and B.D. Craig, Handbook
less Steels, www.outokumpu.com, accessed of Corrosion Data, 2nd ed., ASM Interna-
June 2008 tional, 1995
3. P.A. Schweitzer, Corrosion Resistance
Tables, 5th ed., National Association of Cor-
rosion Engineers, NACE 37755, 2004 \aq2\
CHAPTER 24
Pulp-and-Paper Industry Applications
Summary exclusive domain of duplex stainless steels be-

cause of their lower cost per unit of corrosion
THE PULP-AND-PAPER INDUSTRY has resistance, high strength, and near immunity to
seen more benefits from developments in stain- stress corrosion cracking. Pricing changes
less steel than any other industry. The harsh mainly with alloying element costs, principally
chemicals used in this industry called for better those of nickel and molybdenum. At prices be-
materials than the normal austenitic stainless tween the highs and lows of the first decade of
steels without the expense of the 6Mo grades. the 2000s, duplex costs have been roughly one-
This need has been met through the use of the third less than that of an equivalent corrosion-
duplex alloys, which have become the new resisting austenitic stainless. This factored in
standard. with strength nearly double that of the equiva-
lent austenitic make them an overwhelmingly
superior choice for pulp-and-paper equivalent
Introduction except if very special corrosion requirements
differ from the norm, such as in bleaching.
The proximity of the Scandinavian paper in-
dustry to that region’s specialty steel industry
has been symbiotic. As a result of the strong Paper-Making Processes
interaction between engineers having well-
specified needs for improved materials and met- The kraft (German for “strong”) process was
allurgists capable of providing them, the ad- introduced in 1937, replacing the sulfite process.
vances in materials in the pulp-and-paper In the kraft process, the lignin-connecting wood
industry have been a model of rapid technology fibers are dissolved under conditions of elevated
transfer and innovation. Beginning in 1988, du- temperature and pressure in acidic conditions of
plex stainless steels first went into production in pH 2.0 to 4.0. This leaves a long fiber, which en-
kraft digesters, and there has been no turning ables paper of high strength, hence the name
back in the replacement of austenitic stainless kraft. Over the years the materials used for the
steels by duplex. So, a discussion of the materi- vessels, called digesters, in which this process is
als selection for the pulp-and-paper industry has carried out have been sequentially carbon steel,
changed from a fairly complicated analysis of stainless steel, and stainless steel clad onto car-
which austenitic steel to use while guarding bon steel. In the previous sulfite process, acid-
against stress corrosion cracking and pitting resistant brick vessels were used. Now, the di-
corrosion and when to use clad materials for gesters, essentially large vertical tanks, are
cost savings, to a fairly simple discussion of constructed of 2205 (UNS S32205) as a rule (see
which duplex stainless steel is most economical Fig. 1).
for a given piece of equipment. Since this revo- The digestion is typically carried out at 150
lution occurred in the 1990s before the latest and 180 °C (300 and 360 °C) and 10 to 12 bar.
surge in nickel prices, it is safe to say that future The pH of the sulfate is around 2.0 to 4.0. In
pulp-and-paper projects will be essentially the this environment, 316L can survive, but it
pulp suspension is injected at high velocity. The

environment is a mixture of alkaline liquid,
while the vapor phase can contain organic acids.
The hardness of the duplex helps mitigate ero-
sion, while the alloy level is beneficial against
corrosion. 2205 is the alloy of choice here, but
2003 (UNS S32003) would suffice.
The next step, washing and screening, has
seen increasingly severe environments as closed
systems required for pollution control have be-
come more common. This has rendered the pre-
vious choice of carbon steel untenable. This
stage also sees erosion potential from hard par-
ticles, such as sand, in the pulp. The optimum
solution is a lean duplex such as 2101, 2304, or
2003.
The delignification of the pulp comes next.
This oxygen process dates from the 1970s. At
first, highly alloyed austenitic alloys were
used. Subsequently, it was found again that du-
plex performed better in that they were suffi-
Fig. 1 The first kraft digester fabricated from alloy 2205. ciently corrosion resistant, but also offered
Courtesy of Outokumpu
freedom from stress corrosion cracking as well
as materials savings because of their higher
strength.
The bleaching of the pulp is important for
many types of paper, and this can be done by
requires maintenance and has a finite life. The chlorine bleaches or ozone/peroxide bleaches.
2205 is twice as resistant to corrosion, 0.005 The chlorine bleaching now must generally be
mm/yr versus 0.011 mm/yr (Ref 1). In the done in closed systems, which results in a
nonchloride environment, molybdenum is not buildup of chloride levels to a point at which
an essential alloying element, so the introduc- corrosion levels are unacceptable unless very
tion of the use of 2101 (UNS S32101) or 2304 highly alloyed materials are used. The 6Mo
(UNS S32304) is a logical cost-saving move grades have been successful, but now they can
without strength or corrosion compromises. The be replaced by duplex alloys such as 2507
reduction in wall thickness allowed by the (UNS S32750), which again save cost by virtue
higher-strength duplex depends on the engi- of their higher strength.
neering code required. The American Society of Bleaching can be accomplished without chlo-
Mechanical Engineers (ASME) code require- rine in the so-called TCF, totally chlorine free,
ment is based on tensile strength and permits process. This reduces the corrosivity of the en-
only a 24% reduction in wall thickness, while vironment as the ozone and hydrogen peroxide
the total kjeldahl nitrogen (TKN) code, based used in the process are relatively harmless to
on yield strength, would allow a 46% reduction. stainless steel. Alloys such as 316 are adequate
This large a difference in strength levels re- for this environment, but lean duplex, 2101 or
quired by codes is unfortunate and reflects an 2304, offer cost reductions through their greater
orientation to materials in which the yield/ strength.
tensile ratio is closer to unity, unlike either du- In plants that use recycled paper and mechan-
plex or austenitic stainless steel. In the more un- ical wood chip processing, the materials selec-
usual case of digesters using the sulfite process, tion criteria remain the same. Duplex stainless
the materials selected would be the same. has become the clear choice.
As one proceeds downstream in the process, Further downstream, containers and process
environments change greatly, but the optimal equipment benefit equally from duplex down to
materials remain duplex for various reasons. the handrails and walkways. This wholesale use
The subsequent stage is blow tanks in which the of duplex can make plants nearly maintenance
Chapter 24: Pulp-and-Paper Industry Applications / 267
free from a corrosion point of view, a dramatic REFERENCES

change in an industry in which the thousand-fold
greater corrosion rates of carbon steel presented 1. A. Tuomi et al., Duplex America 2000 Con-
operators with endless equipment downtime ference, Houston, KCI Publishing, 2000
problems. 2. H. Dykstra et al, Corrosion in the Pulp and
Additional detail about corrosion challenges Paper Industry, Corrosion: Environments
and the use of stainless steels in the pulp-and- and Industries, Vol 13C, ASM Handbook,
paper industry can be found in Ref 2. ASM International, 2006, p 762–802
APPENDIX 1
Compositions
Table A1.1 Composition of austenitic stainless steels
Composition, %
Name Designation(a) C N Cr Ni Mn Mo Si P S Other Other
201 S20100 0.15 0.25 16.0–18.0 3.5–5.5 5.5–7.5 ... 1.00 0.060 0.030 ... ...
201L S20103 0.03 0.25 16.0–18.0 3.5–5.5 5.5–7.5 ... 0.75 0.045 0.015 ... ...
201LN S20153 0.03 0.25 16.0–18.0 3.5–5.5 5.5–7.5 ... 0.75 0.045 0.015 ... ...
Gall-Tough S20161 0.15 0.08–0.20 15.0–18.0 4.0–6.0 4.0–6.0 ... 3.00–4.00 0.040 0.040 ... ...
202 S20200 0.15 0.25 17.0–19.0 4.0–6.0 7.5–10.0 ... 1.00 0.060 0.030 ... ...
230 EZ S20300 0.08 ... 16.0–18.0 5.0–6.5 5.0–6.5 0.5 1.00 0.040 0.18–0.35 Cu 1.75–2.25 ...
205 S20500 0.12–0.25 0.32–0.40 16.5–18.0 1.0–1.75 14.0–15.5 ... 1.00 0.060 0.030 ... ...
Nitronic 30 S20400 0.0–3 0.15–0.30 15.0–17.0 1.5–3.0 7.0–9.0 ... 1.0 0.040 0.030 ... ...
Nitronic 32 S24100 0.08 0.20–0.40 16.5–19.0 0.5–2.5 11.0–14.0 ... 1.00 0.060 0.030 ... ...
Nitronic 33 S24300 0.08 0.20–0.40 17.0–19.0 2.25–3.75 11.5–14.5 ... 1.00 0.060 0.030 ... ...
Nitronic 40 (219) S21900 0.08 0.15–0.40 19.0–21.5 5.5–7.5 8.0–10.0 ... 1.00 0.060 0.030 ... ...
21-6-9 LC S21904 0.04 0.15–0.40 19.0–21.5 5.5–7.5 8.0–10.0 ... 1.00 0.060 0.030 ... ...
Nitronic 50 S20910 0.06 0.20–0.40 20.5–23.5 11.5–13.5 4.0–6.0 1.5–3.0 1.00 0.040 0.030 Nb 0.1–0.3 V 0.1–0.3
Nitronic 60 S21800 0.10 0.08–0.18 16.0–18.0 8.0–9.0 7.0–9.0 ... 3.50–4.50 0.040 0.030 ... ...
Tenelon S21400 0.12 0.35 17.0–18.5 0.75 14.5–16.0 ... 0.30–1.00 0.060 0.030 ... ...
Cryogenic S21460 0.12 0.25–0.50 17.0–19.0 5.0–6.0 14.0–16.0 ... 1.00 0.060 0.030 ... ...
Tenelon
Esshete1250 S21500 0.15 ... 14.0–16.0 9.0–11.0 5.5–7.0 ... 1.20 0.040 0.030 Nb 0.75–1.25 V 0.15–0.40
216 S21600 0.08 0.25–0.50 17.5–22.0 5.0–7.0 7.5–9.0 2.0–3.0 1.00 0.045 0.030 ... ...
216L S21603 0.03 0.25–0.50 17.5–22.0 7.5–9.0 7.5–9.0 2.0–3.0 1.00 0.045 0.030 ... ...
301 S30100 0.15 16.0–18.0 6.0–8.0 2.0 ... 1.00 0.045 0.030 ... ...
301LN S30153 0.03 0.10–0.20 16.5–18.0 6.0–8.0 ... ... 0.045 0.030 ... ...
302 S30200 0.15 ... 17.0–19.0 8.0–10.0 2.0 ... 1.00 0.045 0.030 ... ...
302Cu S30430 0.08 ... 17.0–19.0 8.0–10.0 2.0 ... 1.00 0.045 0.030 Cu 3.0–4.0 ...
302B S30215 0.15 ... 17.0–19.0 8.0–10.0 2.0 ... 2.00–3.00 0.045 0.030 ... ...
303 S30300 0.15 ... 17.0–19.0 8.0–10.0 2.0 0.60 optional 1.00 0.045 0.15 min ... ...
303Se S30223 0.15 ... 17.0–19.0 8.0–10.0 2.0 0.60 optional 1.00 0.045 0.06 min Se 0.15 min ...
303 Plus X S30310 0.15 ... 17.0–19.0 7.0–10.0 2.5–4.5 0.60 optional 1.00 0.045 0.25 min ... ...
304 S30400 0.08 0.10 18.0–20.0 8.0–10.5 2.0 ... 1.00 0.045 0.030 ... ...
304L S30403 0.03 0.10 18.0–20.0 8.0–10.5 2.0 ... 1.00 0.045 0.030 ... ...
304H S30409 0.04–0.10 ... 18.0–20.0 8.0–10.5 2.0 ... 1.00 0.045 0.030 ... ...
304N S30451 0.08 0.10–0.16 18.0–20.0 8.0–10.5 2.0 ... 1.00 0.045 0.030 ... ...
304HN S30452 0.08 0.16–0.30 18.0–20.0 8.0–10.5 2.0 ... 1.00 0.045 0.030 ... ...
304LN S30453 0.03 0.10–0.16 18.0–20.0 8.0–10.5 2.0 ... 1.00 0.045 0.030 ... ...
304BI S30424 0.08 0.10 18.0–20.0 12.0–15.0 ... 2.0 0.75 ... ... B 1.00–1.20 ...
153MA S30415 0.04–0.06 0.12–0.18 18.0–19.0 9.0–10.0 0.8 ... 1.00–2.00 ... ... Ce 0.04 ...
305 S30500 0.12 ... 17.0–19.0 10.5–13.0 2.0 ... 1.00 ... ... ... ...
Cronifer 1815 S30600 0.018 ... 17.0–18.5 14.0–15.5 2.0 0.2 3.75–4.25 ... ... Cu 0.50 ...
RA 85 H S30615 0.16–0.24 ... 17.0–19.5 13.5–16.0 2.0 ... 3.2–4.0 0.045 0.030 Al 0.8–1.5 ...
308 S30800 0.08 ... 19.9–21.0 10.0–12.0 2.0 ... 1.00 0.045 0.030 ... ...
253MA S30815 0.05–0.10 0.14–0.20 20.0–22.0 10.0–12.0 0.8 ... 1.4–2.0 ... ... 1.0 Al 0.03–0.08 Ce
309 S30900 0.20 ... 22.0–24.0 12.0–15.0 2.0 ... 0.75 0.045 0.030 ... ...
309S S30908 0.08 ... 22.0–24.0 12.0–15.0 2.0 ... 0.75 0.045 0.030 ... ...
(continued)
Notes: All compositions include Fe as balance. Single values are maximum values unless otherwise noted. (a) Unified Number System, UNS numbers are S or N followed by 5 digits.
Table A1.1 (continued)
Composition, %
309H S30909 0.04–0.10 ... 22.0–24.0 12.0–15.0 2.0 ... 0.75 0.045 0.030 ... ...
309Cb S30940 0.08 ... 22.0–24.0 12.0–16.0 2.0 ... 1.00 0.040 0.030 Nb 10xC to 1.10 . . .
309HCb S30941 0.04–0.10 ... 22.0–24.0 12.0–16.0 2.0 ... 1.00 0.040 0.030 Nb 10xC to 1.10 . . .
309Si DIN 1.4828 0.04–0.10 0.11 19.0–21.0 11.0–13.0 2.0 ... 1.50–2.50 0.040 0.015 ... ...
310 S31000 0.25 ... 24.0–26.0 19.0–22.0 2.0 ... 1.00 0.045 0.030 ... ...
310S S31008 0.08 ... 24.0–26.0 19.0–22.0 2.0 ... 1.00 0.045 0.030 ... ...
310H S31009 0.04–0.10 ... 24.0–26.0 19.0–22.0 2.0 ... 1.00 0.045 0.030 ... ...
310Cb S31040 0.08 ... 24.0–26.0 19.0–22.0 2.0 ... 1.00 0.045 0.030 Nb 10xC to 1.10 . . .
310HCb S31041 0.04–0.10 ... 24.0–26.0 19.0–22.0 2.0 ... 1.00 0.045 0.030 Nb 10xC to 1.10. . .
310HCbN S31042 0.04–0.10 0.15–0.35 24.0–26.0 19.0–22.0 2.0 ... 1.00 0.045 0.030 Nb 10xC to 1.10 . . .
310MoLN S31050 0.02 0.09–0.15 24.0–26.6 20.5–23.5 2.0 1.6–2.6 0.5 0.020 0.010 ... ...
310Si DIN 1.4841 0.20 0.11 24.0–26.0 19.0–22.0 2.0 ... 1.50–2.50 0.045 0.015 ... ...
314 S31400 0.25 ... 23.0–26.0 19.0–22.0 2.0 ... 1.50–3.00 0.045 0.030 ... ...
316 S31600 0.08 0.10 16.0–18.0 10.0–14.0 2.0 2.0–3.0 1.00 0.045 0.030 ... ...
316H S31609 0.04–0.10 ... 16.0–18.0 10.0–14.0 2.0 2.0–3.0 1.00 0.045 0.030 ... ...
316F S31620 0.08 ... 16.0–18.0 10.0–14.0 2.0 1.75–2.5 1.00 0.20 0.10 min ... ...
316L S31603 0.03 0.10 16.0–18.0 10.0–14.0 2.0 2.0–3.0 1.00 0.045 0.030 ... ...
316LN S31653 0.03 0.10–0.16 16.0–18.0 10.0–14.0 2.0 2.0–3.0 1.00 0.045 0.030 ... ...
316N S31651 0.08 0.10–0.16 16.0–18.0 10.0–14.0 2.0 2.0–3.0 1.00 0.045 0.030 ... ...
316Ti S31635 0.08 0.10 16.0–18.0 10.0–14.0 2.0 2.0–3.0 1.00 0.045 0.030 5xC to 0.70 Ti . . .
317 S31700 0.08 0.10 18.0–20.0 11.0–15.0 3.0–4.0 2.0 1.00 0.045 0.030 ... ...
317L S31703 0.03 0.10 18.0–20.0 11.0–15.0 3.0–4.0 2.0 1.00 0.045 0.030 ... ...
317LN S31753 0.03 0.10–0.22 18.0–21.0 11.0–15.0 3.0–4.0 2.0 1.00 0.045 0.030 0.030 max P ...
317LM S31725 0.03 0.10 18.0–20.0 13.5–17.5 4.0–5.0 2.0 1.00 0.045 0.030 ... ...
317LMN S31726 0.03 0.10–0.20 17.0–20.0 13.5–17.5 4.0–5.0 2.0 0.75 0.045 0.030 ... ...
321 S32100 0.08 0.10 17.0–19.0 9.0–12.0 2.0 ... 1.00 0.045 0.030 Ti 5x(C+N) ...
to 0.70
321H S32109 0.40–0.10 0.10 17.0–19.0 9.0–12.0 2.0 ... 1.00 0.045 0.030 Ti 4x(C+N) ...
to 0.70
330 S33000 0.08 ... 17.0–20.0 34.0–37.0 2.0 ... 0.75–1.50 0.030 0.030 ... ...
332 N08800 0.03 ... 19.0–23.0 30.0–35.0 1.5 ... 1.00 ... ... Ti 0.15–0.60 Al 0.15–0.60
332Mo* S35125 0.10 0.10 20.0–23.0 31.0–35.0 1.0–1.5 2.0–3.0 0.75 ... ... Nb 0.25–0.60 . . .
334 S33400 0.08 ... 18.0–20.0 19.0–21.0 1.0 ... 1.00 ... ... Ti 0.15–0.60 Al 0.15–0.60
347 S34700 0.08 ... 17.0–19.0 9.0–13.0 2.0 ... 1.00 0.045 0.030 Nb+Ta 10xC ...
to 1.10
347H S34709 0.08 ... 17.0–19.0 9.0–13.0 2.0 ... 1.00 0.045 0.030 Nb+Ta 10xC ...
to 1.10
348 S34800 0.08 ... 17.0–19.0 9.0–13.0 2.0 ... 1.00 0.045 0.030 Nb+Ta 0xC Co 0.2
to 1.10
348H S34809 0.04–0.10 ... 17.0–19.0 9.0–13.0 2.0 ... 1.00 0.045 0.030 Nb+Ta 8xC ...
to 1.0
370 S37000 0.03–0.05 0.005 12.5–14.5 14.5–16.5 1.65–2.35 1.5–2.5 0.5–1.0 ... ... Ti 0.10–0.40 Co 0.05
384 S38400 0.08 ... 15.0–17.0 17.0–19.0 2.0 ... 1.00 0.045 0.030 Nb+Ta 10xC Co 0.2
to 1.10
353MA S35315 0.08 0.12–0.18 24.0–26.9 34.0–36.0 1.0 ... 0.6–1.0 ... ... Ce 0.03–0.10 . . .
(continued)
Appendix 1: Compositions / 271
Composition, %
AC66 S33228 0.04–0.08 ... 26.0–28.0 31.0–33.0 1.0 ... 0.03 ... ... 0.05–0.10 Ce Al 0.025
Incoloy 803 S25045 0.06–0.10 ... 25.0–29.0 32.0–37.0 1.5 ... 1.0 ... 0.015 Ti 0.15–0.60 0.15–0.60 Al
Incoloy 864 S35135 0.08 ... 20.0–25.0 30.0–38.0 1.0 4.0–4.8 0.6–1.0 ... 0.015 Ti 0.4–1.0 ...
21-4N S63008 0.48–0.58 0.28–0.50 20.0–22.0 3.25–4.50 8.0–10.0 ... 0.25 0.045 0.015 ... ...
21-2N S63012 0.50–0.60 0.20–0.40 19.25–21.50 1.50–2.75 7.0–9.50 ... 0.25 0.050 0.030 ... ...
21-12N S63017 0.15–0.25 0.15–0.25 20.0–22.0 10.5–12.5 1.0–1.5 ... 0.70–1.25 0.045 0.030 ... ...
23-8N S63018 0.28–0.38 0.28–0.38 22.0–24.0 7.9–9.0 1.5–3.5 ... 0.69–0.90 0.045 0.030 ... ...
19-9DL S63198 0.28–0.36 ... 18.0–21.0 8.0–11.0 0.75–1.50 1.0–1.75 0.03–0.80 ... ... 0.1–0.35 Ti 0.25–0.60 Nb
20Cb-3 N08020 0.07 ... 19.0–21.0 32.0–38.0 0.75–1.50 2.0–3.0 1.00 0.045 0.035 8xC to 1.00 Nb Cu 3.0–4.0
20Mo-4 N08024 0.03 ... 22.5–25.0 35.0–40.0 1.0 3.5–5.0 0.50 0.0035 0.035 Nb 0.15-0.35 Cu 0.5–1.5
20Mo-6 N08026 0.03 ... 22.0–26.0 33.0–37.0 1.0 5.0–6.7 0.50 0.030 0.030 8xC Nb 2.0–4.0 Cu
Sanicro 28 N08028 0.02 ... 26.0–28.0 29.9–32.5 2.5 3.0–4.0 1.00 0.030 0.030 ... 0.6–1.4 Cu
AL-6X N08366 0.035 ... 20.0–22.0 23.5–25.5 2.0 6.0–7.0 1.00 0.030 0.030 0.030 P ...
AL-6XN N08367 0.030 0.18–0.25 20.0–22.0 23.5–25.5 2.0 6.0–7.0 1.00 0.030 0.030 0.040 P ...
JS-700 N08700 0.04 ... 19.0–23.0 24.0–26.0 2.0 4.3–5.0 1.00 0.040 0.030 8xC to 0.5 Nb 0.5 Cu
904L N80904 0.02 ... 19.0–23.0 23.0–28.0 2.0 4.0–5.0 1.00 0.045 0.035 1.0–2.0 Cu ...
254SMO S31254 0.02 0.18–0.22 19.50–20.50 17.50–18.50 1.0 6.0–6.5 0.80 0.030 0.010 0.5–1.0 Cu ...
4565 S34565 0.03 0.40–0.60 23.0–25.0 16.0–18.0 5.0–7.0 4.0–5.0 1.0 ... ... 0.10 Nb ...
654SMO S32654 0.02 0.45–0.55 24.0–25.0 21.0–23.0 2.0–3.0 7.0–8.0 0.5 0.030 0.005 0.3–0.6 Cu ...
AL 20 N08020 0.07 ... 19.0–21.0 32.0–38.0 2.0 2.0–3.0 1.0 ... ... 3.0–4.0 Cu 8xC to 1.00
Nb+Ta
AL 22 N06022 0.015 20.0–22.5 balance ... 12.5–14.5 0.08 ... ... 2.5 Co,0.35 V W 2.5-3.5
Cronifer N08926 0.02 0.15–0.25 19.0–21.0 24.0–26.0 2.0 6.0–7.0 0.50 ... ... 0.5–1.5 Cu ...
1925 hMo
Cronifer 2328 ... 0.04 ... 22.0–24.0 26.0–28.0 0.75 2.5–3.0 0.75 0.045 0.035 2.5–3.5 Cu Ti 0.4-0.7
Nicrofer N08031 0.015 0.15–0.25 26.0–28.0 24.0–26.0 2.0 6.0–7.0 0.30 ... ... 1.0–1.4 Cu ...
3127HMo
URSB-8 N08932 0.02 0.15–0.25 24.0–26.0 24.0–26.0 2.0 4.5–6.5 0.40 ... ... 1.0–2.0 Cu ...
B66 S31266 0.03 0.35–0.60 23.0–26.0 21.0–24.0 2.0 5.0–7.0 1.0 ... ... 0.5–3.0 Cu W 1.0-3.0
NIC 25 S32200 0.03 ... 20.0–23.0 23.0–27.0 1.0 2.5–3.5 0.5 ... ... ...
CN-7M N08007 0.07 ... 19.0–22.0 27.5–30.5 1.5 2.0–3.0 1.5 ... ... 3.0–4.0 Cu ...
N08320 20Mod 0.05 ... 21.0–23.0 25.0–27.0 2.5 4.0–6.0 1.0 ... ... ... ...
Table A1.2 Composition of ferrite stainless steels
Composition, %
UNS
Name designation C N Cr Ni Mn Mo Si P S Ti Nb Other
405 S40500 0.08 ... 11.5–14.5 0.60 1.00 ... 1.00 0.040 0.030 ... ... Al 0.10–0.30
400 AK alloy 0.05 ... 12.0–13.0 ... 1.00 ... 1.00 ... ... ... ... Al 0.25 Al
409 S40900 0.08 ... 10.5–11.75 0.50 1.00 ... 1.00 0.045 0.045 6x(C+N) ... ...
to 0.75
409 S40910 0.03 0.030 10.5–11.7 0.50 1.00 ... 1.0 0.040 0.030 6x(C+N)to 0.5 0.17 ...
409 S40920 0.03 0.030 10.5–11.75 0.50 1.00 ... 1.00 0.040 0.030 8x(C+N) ... ...
to 0.15–0.50
409 ultraform AK alloy 0.02 0.020 10.5–11.7 0.50 0.75 ... 1.00 ... ... 8x(C+N) ... ...
466 S40930 0.02 0.020 10.5–11.75 0.05 1.00 ... 1.00 0.040 0.030 0.8+ 8x(C+N) ... ...
Ti+Nb
409Cb S40940 0.06 ... 10.5–11.7 0.50 1.00 ... 1.0 ... ... 10xC ... ...
to 0.75 Nb
409Ni S40975 0.03 0.03 10.5–11.7 0.5–1.0 1.00 ... 1.0 0.040 0.030 ... ... ...
11 Cr–Cb AK alloy 0.010 0.015 11.35 0.20 0.25 ... 1.30 ... ... ... 0.35 ...
typical
12 SR AK alloy 0.020 0.015 12.0 ... ... ... ... ... ... 0.30 0.60 AL 1.2
typical
Alfa I ATI alloy 0.025 ... 13.0 ... 0.035 ... 0.03 ... ... 0.40 ... AL 3.0
typical
Alfa II ATI alloy 0.025 ... 13.0 ... 0.035 ... 0.03 ... ... 0.40 ... AL 4.0
typical
4724 Outukumpu 0.08 ... 13.5 ... 0.70 ... 1.0 ... ... ... ... Al 1.0
typical
429 S42900 0.12 ... 14.0–16.0 0.75 1.00 ... 1.00 0.040 0.030 ... ... ...
430 S43000 0.12 ... 16.0–18.0 0.75 1.00 ... 1.00 0.040 0.030 ... ... ...
430F S43020 0.12 ... 16.0–18.0 ... 1.25 ... 1.00 0.060 0.150 ... ... ...
min
430Se S43023 0.12 ... 16.0–18.0 ... 1.25 ... 1.00 0.060 0.060 ... ... SE 0.15
430Ti S43036 0.10 0.04 16.0–19.5 1.00 1.00 ... 1.00 0.040 0.030 0.20+4x(C+N) ... AL 0.15
to 1.10
439 S43035 0.07 0.04 17.0–19.0 0.50 1.00 ... 1.00 0.040 0.030 0.20+4x(C+N) ... ...
to 1.10
439LT S43932 0.03 0.03 17.0–19.0 0.50 1.00 ... 1.0 0.040 0.030 0.20+4x(C+N) ... Al 0.15
to 0.75
Ti+Nb
439 HP ATI, AK 0.01 0.01 17.5 0.2 0.35 ... 0.45 ... ... 0.35 ... ...
439 ultraform alloys
468 S46800 0.030 ... 18.0–20.0 0.50 1.00 ... 1.00 0.040 0.030 Ti+Nb:0.20+ ... ...
4x(C+N)
to 1.10
18 Cr–Cb AK alloy 0.020 ... 18.0 ... 0.30 ... 0.45 ... ... 0.25 0.55 ...
typical
(continued)
Notes: All compositions include Fe as balance. Single values are maximum values unless otherwise noted
Table A1.2 Composition
(continued) of ferrite stainless steels
Composition, %
UNS
Name designation C N Cr Ni Mn Mo Si P S Ti Nb Other
18SR AK alloy 0.015 . . .. 17.30 0.25 0.30 . . .. . . .. . . .. ... 0.25 ... Al 1.7
typical
4742 Outukumpu 0.08 ... 18.0 ... 0.7 ... 1.3 ... ... ... ... Al 1.0
typical
434 S43400 0.12 ... 16.0–18.0 ... 1.0 0.75–1.25 1.0 0.040 0.030 ... ... ...
436 S43600 0.12 ... 16.0–18.8 ... 1.0 0.75–1.25 1.0 0.040 0.030 ... Nb+Ta ...
5xC:0.70
441, S44100 0.030 ... 17.5–18.5 ... 1.00 ... 1.0 0.040 0.030 0.1–0.6 9xC ...
4509, 0.3–1.0
430J1L
442 S44200 0.20 ... 18.0–23.0 0.6 1.0 ... 1.0 0.040 0.030 ... ... ...
436S ATI alloy 0.01 0.015 17.3 0.3 0.20 1.2 0.4 ... ... 8x(C+N) min ... ...
typical
444, YUS S44400 0.025 0.035 17.5–19.5 1.0 1.0 0.75–1.25 1.0 0.040 0.030 Ti+Nb:0.20+ ... ...
190-EM 4x(C+N)
to 0.80
433 ATI alloy 0.01 ... 20.0 0.25 0.30 ... 0.4 ... ... ... 10x(C+N) ...
typical
4762 Outukumpu 0.08 ... 24.0 ... 0.7 ... 1.4 ... ... ... ... Al 1.5
typical
453 ATI alloy 0.03 ... 22.0 0.3 0.3 ... 0.3 ... ... 0.02 ... 0.60 Al
typical 0.10 REM
E-Brite, 26-1 S44627 0.01 0.015 25.0–27.5 0.50 0.40 0.75–1.25 0.40 0.020 0.020 ... 0.5–0.20 0.2 Cu
0.5 Cu+Ni
Monit S44635 0.025 0.035 24.5–26.0 3.5–4.5 1.00 3.5–4.5 0.75 0.040 0.030 Ti+Nb:0.20+ ... ...
4x(C+N)
to 0.80
Sea–cure S44660 0.025 0.035 25.0–27.0 1.5–3.5 1.00 2.5–3.5 1.00 0.040 0.030 Ti+Nb:0.20+ ... ...
4x(C+N)
to 0.80
29-4C S44735 0.025 ... 28.0–30.0 0.5 1.00 3.5–4.5 0.75 0.040 0.030 Ti+Nb:0.20+ ... ...
4x(C+N)
to 0.80
446 S44600 0.20 0.25 23.0–27.0 0.6 1.50 ... 1.00 0.040 0.030 ... ... ...
Notes: All compositions include Fe as balance. Single values are maximum values unless otherwise noted.
Table A1.3 Composition of martensitic stainless steels
Composition, %
UNS
Name designation C N Cr Ni Mn Mo Si P S Other
403 S40300 0.15 max ... 11.5–13.5 ... 1.00 0.50 0.50 0.040 0.030 ...
410 S41000 0.15 max ... 11.5–13.5 ... 1.00 1.00 1.00 0.040 0.030 ...
410S S41003 0.03 ... 10.5–12.5 1.5 1.00 ... ... 0.040 0.030 ...
410 S41008 0.08 ... 11.5–13.5 ... 1.50 1.00 1.00 0.040 0.030 ...
410Cb S41040 0.18 max ... 11.5–13.5 ... 1.00 1.00 1.00 0.040 0.030 Nb 0.05–0.30
412 S41003 0.030 max ... 10.5–12.5 1.5 1.50 1.00 1.00 0.040 0.030 ...
414 S41400 0.15 max ... 11.5–13.5 1.25–2.50 1.00 1.00 1.00 0.040 0.030 ...
414 mod S41425 0.05 0.06 0.12 12.0–15.0 4.0–7.0 0.5–1.0 0.60 0.60 0.040 0.005 Cu 0.30
415 S41500 0.05 max ... 11.5–14.0 3.50–5.50 0.50–1.0 0.60 0.60 0.040 0.030 ...
416 S41600 0.15 max ... 12.0–14.0 ... 1.25 1.00 1.00 0.060 0.15 0.30 ...
416Se S41623 0.15 max ... 12.0–14.0 ... 1.25 1.00 1.00 0.060 0.060 Se 0.15 min
418 S41800 0.15–0.20 ... 12.0–14.0 1.80–2.20 0.50 0.50 0.50 0.040 0.030 W 2.50–3.50
420 S42000 0.15 min ... 12.0–14.0 ... 1.00 1.00 1.00 0.040 0.030 ...
4116 DIN 1.4116 0.50 ... 14.5 ... ... ... ... 0.040 ... ...
nominal
420F S42020 0.15 min ... 12.0–14.0 ... 1.25 1.00 1.00 0.040 0.15 ...
420FSe S42023 0.15 min ... 12.0–14.0 ... 1.25 1.00 1.00 0.040 0.06 Se 0.15 min
422 S42200 0.20–0.25 ... 11.0–13.5 0.50–1.00 1.00 0.75 0.75 0.040 0.030 0.75–1.25 W
424 S42400 0.06 max ... 12.0–14.0 3.50–4.50 0.50–1.0 0.30–0.60 0.30–0.6 0.040 0.030 ...
425 S42500 0.08–0.20 ... 14.0–16.0 1.00–2.00 1.00 1.00 1.00 0.040 0.010 ...
425 mod ... 0.50–0.55 ... 13.0–14.0 0.50 1.00 1.00 1.00 0.040 0.030 ...
Trinamet ... 0.30 max ... 12.0–14.0 ... 1.00 1.00 1.00 0.040 0.030 CU 2.0–3.0
HP13Cr-1 JFE nominal 0.025 ... 13.0 4.0 0.45 ... ... ... ... ...
HP13Cr-2 JFE nominal 0.025 ... 13.0 5.0 0.45 ... ... ... ... ...
NT-CRS Nippon nominal 0.03 0.040 12.7 4.5 1.45 ... ... ... ... 1.5 Cu
NT-CRSS Nippon nominal 0.02 0.015 12.3 5.8 2.0 ... ... ... ... 1.5 Cu
KL-12Cr JFE nominal 0.01 0.010 11.0 2.4 ... ... ... ... ... 0.5 Cu
KL-HP 12Cr JFE nominal 0.01 0.010 12.0 5.5 ... ... ... ... ... ...
431 S43100 0.20 max ... 15.0–17.0 1.25–2.50 1.00 1.00 1.00 0.040 0.030 ...
440A S44002 0.60–0.75 ... 16.0–18.0 ... 1.00 1.00 1.00 0.040 0.030 ...
440B S44003 0.75–0.95 ... 16.0–18.0 ... 1.00 1.00 1.00 0.040 0.030 ...
440C S44004 0.95–1.20 ... 16.0–18.0 ... 1.00 1.00 1.00 0.040 0.030 ...
440F S44020 0.95–1.20 ... 16.0–18.0 0.75 1.25 1.00 1.00 0.040 0.10–0.35 ...
440FSe S44023 0.95–1.20 ... 16.0–18.0 0.75 1.25 1.00 1.00 0.060 0.060 Se 0.15 min
BG-42 Nominal PM 1.15 ... 14.5 ... ... 0.3 0.3 ... ... 1.2 V
ATS-34 Nominal PM 1.05 ... 14.0 ... 0.4 0.35 0.35 ... ... ...
14-4 CrMo Nominal PM 1.05 ... 14.0 ... 0.5 0.3 0.3 ... ... ...
154 CM Nominal PM 1.05 ... 14.0 ... 0.45 0.3 0.3 ... ... ...
CPM S30V Nominal PM 1.45 ... 14.0 ... ... ... ... ... ... 4.0 V
CPM S60V Nominal PM 2.15 ... 17.0 ... 0.40 ... ... ... ... 5.5 V
CPM S90V Nominal PM 2.20 ... 13.0 ... ... ... ... ... ... 9.0 V
Table A1.4 Composition of selected precipitation–hardenable stainless steels
Composition, %
Name Designation C N Cr Ni Mn Mo Si P S Other Other
Stainless W S17600 0.08 ... 15.0–17.0 6.0–7.5 1.0 ... 1.00 0.040 0.030 Ti 0.4–1.2 Al 0.4
17-4 PH S17400 0.07 ... 115.5–17.5 3.0–5..0 1.0 ... 1.00 0.040 0.030 Cu 3.0–5.0 Nb .015–0.45
15-5 PH S15500 0.07 ... 14.0–15.5 3.5–5.5 1.0 ... 1.00 0.040 0.030 Cu 2.5–4.5 Nb 0.15–0.45
13-8 PH S13800 0.05 0.010 12.25–13.25 7.5–8.5 0.2 2.0–2.5 0.10 0.010 0.008 ... Al 0.90–1.35
Custom 450 S45000 0.05 ... 14.0–16.0 5.0–7.0 1.0 0.5–1.0 1.00 0.030 0.030 Cu 1.25–1.75 Nb 8XC
Custom 455 S45500 0.05 ... 11.0–12.50 7.5–9.5 0.50 0.5 0.50 0.040 0.030 Cu 1.5–2.5, Nb +Ta 0.1–0.5
Ti 0.08–1.4
Custom 465 S46500 0.02 ... 11.0–12.50 10.75–11.25 0.25 0.75–1.25 0.25 0.015 0.010 Ti 1.50–1.80 ...
Custom 475 S17600 0.01 ... 10.5–11.50 7.5–8.5 0.50 4.5–5.5 0.50 0.015 0.010 Co 8.0–9.0 Al 1.0–1.5
17-7 PH S17700 0.09 ... 16.0–18.0 6.5–7.75 1.00 ... 1.00 0.040 0.030 ... Al 0.75–1.5
15-7 PH S15700 0.09 ... 14.0–16.0 6.5–7.75 1.00 2.0–3.0 1.00 0.040 0.030 ... Al 0.75–1.5
AM-350 S35000 0.07–0.11 0.07–0.13 16.0–17.0 4.0–5.0 0.50–1.25 2.5–3.25 0.50 0.040 0.030 ... ...
AM-355 S35500 0.10–0.15 0.07–0.13 15.0–16.0 4.0–5.0 0.50–1.25 2.5–3.25 0.50 0.040 0.030 ... ...
A-286 S66286 0.08 ... 13.5–16.0 14.0–27.0 2.00 1.0–1.5 1.00 0.040 0.030 Ti 1.9–2.35 Al 0.35
V 0.10–.050 B 0.001–0.010
Ferrium S53(a) ... 0.21 ... 10.0 5.5 0.10 2.0 0.10 ... ... Co 14 W 1, V 0.3
Notes: All compositions include Fe as balance. Single values are maximum values unless otherwise noted. (a) Nominal value
Table A1.5 Composition of selected duplex stainless steels

Composition, %
Name Designation C N Cr Ni Mn Mo Si P S Cu W
329 S32900 0.08 ... 23.0–28.0 2.5–5.0 1.0 1.0–2.0 0.75 0.040 0.030 0.75 ...
44LN S31200 0.03 0.14–0.20 24.0–26.0 5.5–6.0 2.0 1.2–2.0 1.0 0.045 0.030 1.0 ...
DP3 S31260 0.03 0.10–0.30 24.0–26.0 5.5–7.5 1.0 2.5–3.5 0.75 0.030 0.020 0.75 0.1–0.5
3RE60 S31500 0.30 0.05–0.10 18.0–19.0 4.25–5.25 1.2–2.0 2.5–3.0 1.4–2.0 0.030 0.030 1.4–2.0 ...
2205 (old) S31830 0.03 0.08–0.20 21.0–23.0 2.5–3.5 2.0 2.5–3.5 1.0 0.030 0.020 1.0 ...
19 D S32001 0.03 0.05–0.17 19.5–21.5 1.0–3.0 4.0–6.0 ... 1.0 0.040 0.030 1.0 ...
2003 S32003 0.03 0.14–0.20 19.5–21.– 3.0–4.0 2.0 1.5–2.0 1.0 0.040 0.030 1.0 ...
2101 S32101 0.04 0.20–0.25 21.0–22.0 1.35–1.70 4.0–6.0 0.1–0.8 1.0 0.040 0.030 1.0 ...
2205 S32205 0.03 0.14–0.20 22.0–23.0 4.5–6.5 1.0 3.0–3.5 2.0 0.030 0.020 2.0 ...
2304 S32304 0.03 0.05–0.20 21.5–23.5 3.0–5.0 2.5 ... 1.0 0.040 0.040 1.0 ...
Uranus 52N+ S32520 0.03 0.20–0.35 24.0–26.0 5.5–8.0 1.5 3.0–5.0 0.8 0.035 0.020 0.8 ...
255 S32550 0.04 0.10–0.25 24.0–27.0 6.0–8.0 1.5 2.9–3.9 1.0 0.040 0.030 1.0 ...
2507 S32750 0.03 0.20–0.30 24.0–26.0 6.0–8.0 1.2 3.0–5.0 0.8 0.035 0.020 0.8 ...
Zeron 100 S32760 0.03 0.20–0.30 24.0–26.0 6.0–8.0 1.0 3.0–5.0 1.0 0.030 0.010 1.0 0.5–1.0
2906 S32906 0.03 0.30–0.40 28.0–30.0 5.8–7.5 0.8–1.5 1.5–2.6 0.5 0.030 0.030 0.5 ...
7-Mo Plus S32950 0.03 0.15–0.35 26.0–29.0 3.5–5.2 2.0 1.0–2.5 0.6 0.035 0.010 0.6 ...
DP3W S39274 0.03 0.24–0.32 24.0–26.0 6.0–8.0 1.0 2.5–3.5 0.8 0.030 0.020 0.8 1.5–2.5
AF 918 S39277 0.025 0.23–0.33 24.0–26.0 6.5–8.0 0.8 3.0–4.0 0.8 0.030 0.020 0.8 0.8–1.2
Table A1.6 Composition of Alloy Casting Institute (ACI) heat– and corrosion–resisting casting alloys
(continued)
Composition, %
Wrought UNS
Name equivalent(a) designation C N Cr Ni Mn Mo Si P S Other
Corrosion–resisting alloys
CA-15 410 J91150 0.15 ... 11.5–14.0 1.0 1.00 0.50(b) 1.50 0.04 0.04 ...
CA-15M ... J91151 0.15 ... 11.5–14.0 1.0 1.00 0.15–1.00 0.65 0.04 0.04 ...
CA-40 420 J91153 0.40 ... 11.5–14.0 1.0 1.00 0.50(b) 1.50 0.04 0.04 ...
CA-40F 420F J91154 0.2–0.4 ... 11.5–14.0 1.0 1.00 ... 1.50 0.04 0.04 0.20–0.40 S
CB-30 431,442 J91803 0.30 ... 18.0–22.0 2.0 1.00 ... 1.50 0.04 0.04 ...
CC-50 446 J92613 0.30 ... 26.0–30.0 4.0 1.00 ... 1.50 0.04 0.04 ...
CA-6N ... J91650 0.06 ... 10.5–12.5 6.0–8.0 0.50 ... 1.00 0.04 0.04 ...
CA-6NM S41500 J91540 0.06 ... 11.5–14.0 3.5–4.5 1.00 0.4–1.0 1.00 0.04 0.04 ...
CA-28MWV 422 J91422 0.20–0.28 ... 11.0–12.5 0.5–1.0 0.5–1.0 0.9–1.25 1.00 0.04 0.04 0.9–1.25 W, 0.2–0.3V
CB-7Cu-1 17–4PH J92180 0.07 0.05 15.5–17.7 3.6–4.6 0.70 ... 1.00 0.04 0.04 2.5–3.2 Cu, 0.2–0.35Nb
CB-7Cu-2 15–5PH J92110 0.07 0.05 14.0–15.5 4.5–5.5 0.70 ... 1.00 0.04 0.04 2.5–3.2 Cu, 0.2–0.35Nb
CD-3MN 2205 (S32205) J92205 0.03 0.10–0.30 21.0–23.5 4.5–6.5 1.50 2.5–3.5 1.00 0.04 0.04 Cu 1.0
CD-3MCuN 255 (S32550) J93373 0.03 0.22–0.33 24.0–26.7 5.6–6.7 1.20 2.9–3.8 1.10 0.04 0.04 Cu 1.4–1.9
CD-3MWCuN (S32760) J93380 0.03 0.20–0.30 24.0–26.0 6.5–8.5 1.00 3.0–4.0 1.00 0.04 0.04 Cu 0.5–1.0, W 0.5–1.0
CD-4MCu ... J93370 0.04 ... 24.5–26.5 4.75–6.0 1.00 1.75–2.25 1.00 0.04 0.04 Cu 2.75–3.25
CD-4MCuN ... J93372 0.04 0.10–0.25 24.5–26.5 4.7–6.0 1.00 1.75–2.25 1.00 0.04 0.04 Cu 2.75–3.25
CD-6MN ... J93371 0.06 0.15–0.25 24.0–27.0 4.0–6.0 1.00 1.75–2.25 1.00 0.04 0.04 Cu 1.75–2.5
CE-3MN 2507 (S32750) J93404 0.03 0.10–0.30 24.0–26.0 6.0–8.0 1.50 4.0–5.0 1.00 0.04 0.04 ...
CE-8MN ... J93345 0.08 0.10–0.30 22.5–25.5 8.0–11.0 1.00 3.0–4.5 1.50 0.04 0.04 ...
CE-30 312 J93423 0.30 ... 26.0–30.0 8.0–11.0 1.50 ... 2.00 0.04 0.04 ...
CF-3 304L J92500 0.03 ... 17.0–21.0 8.0–12.0 1.50 ... 2.00 0.04 0.04 ...
CF-3M 316L J92800 0.03 ... 17.0–21.0 8.0–12.0 1.50 2.0–3.0 2.00 0.04 0.04 ...
CF-3MN 316LN J92700 0.03 0.10–0.20 17.0–21.0 9.0–13.0 1.50 2.0–3.0 1.50 0.04 0.04 ...
CF-8 304 J92600 0.08 ... 18.0–21.0 8.0–11.0 1.50 ... 2.00 0.04 0.04 ...
CF-8C 347 J92710 0.08 ... 18.0–21.0 9.0–12.0 1.50 ... 2.00 0.04 0.04 Nb 8XC min
CF-8M 316 J92900 0.08 ... 18.0–21.0 9.0–12.0 1.50 2.0–3.0 2.00 0.04 0.04 ...
CF-10 304H J92590 0.04–0.10 ... 18.0–21.0 8.0–11.0 1.50 ... 2.00 0.04 0.04 ...
CF-10M 316H J92901 0.04–0.10 ... 18.0–21.0 9.0–12.0 1.50 2.0–3.0 1.50 0.04 0.04 ...
CF-10MC 316H J92971 0.10 ... 15.0–18.0 13.0–16.0 1.50 1.75–2.25 1.50 0.04 0.04 (10xC)–1.2 Nb
CF-10SMnN Nitronic™60 J92972 0.10 0.08–0.18 16.0–18.0 8.0–9.0 7.0–9.0 ... 3.5–4.5 0.04 0.04 ...
CF-12M 316 ... 0.12 ... 18.0–21.0 9.0–12.0 1.50 2.0–3.0 2.00 0.04 0.04 ...
CF-16F 303 J92701 0.16 ... 18.0–21.0 9.0–12.0 1.50 1.5 2.00 0.17 0.04 Se 0.2–0.35
CF-20 302 J92602 0.20 ... 18.0–21.0 8.0–11.0 1.50 ... 2.00 0.04 0.04 ...
CG-6MMN Nitronic™50 J93790 0.06 0.20–0.40 20.5–23.5 11.5–13.5 4.0–6.0 1.5–3.0 1.00 0.04 0.04 0.1–0.3 Nb, 0.1–0.3 V
CG-8M 317 J93000 0.08 ... 18.0–21.0 9.0–13.0 1.50 ... 1.50 0.04 0.04 ...
CG-12 308 J93001 0.12 ... 20.0–23.0 10.0–13.0 1.50 ... 2.00 0.04 0.04 ...
CH-8 309S J93400 0.08 ... 22.0–26.0 12.0–15.0 1.50 ... 1.50 0.04 0.04 ...
CH-10 309H J93401 0.04–0.10 ... 22.0–26.0 12.0–15.0 1.50 ... 2.00 0.04 0.04 ...
CH-20 309 J93402 0.20 ... 22.0–26.0 12.0–15.0 1.50 ... 2.00 0.04 0.04 ...
CK-3MCuN 254SMO™ J94653 0.025 0.18–0.24 19.5–20.5 17.5–19.5 1.20 6.0–7.0 1.00 0.04 0.04 Cu 0.5–1.0
CK-20 310 J94202 0.20 ... 23.0–27.0 19.0–22.0 2.00 ... 2.00 0.04 0.04 ...
CN-3M 904L J94652 0.03 ... 20.0–22.0 23.0–27.0 2.00 4.5–5.5 1.00 0.04 0.04 ...
CN-3MN AL–6XN® J94651 0.03 0.18–0.24 20.0–22.0 23.0–27.0 2.00 6.0–7.0 1.00 0.04 0.04 ...
(continued)
Notes: All compositions include Fe as balance. Single values are maximum values unless otherwise noted. (a) The wrought equivalent composition is not the same as the cast. (b) Mo is not an intentional addition.
Composition, %
Wrought UNS
Name equivalent(a) designation C N Cr Ni Mn Mo Si P S Other
CN-7M 320 N08007 0.07 ... 19.0–22.0 27.5–30.0 1.50 2.0–3.0 1.50 0.04 0.04 Cu 3.0–4.0
CN-7MS ... J94650 0.07 ... 18.0–20.0 22.0–25.0 1.50 2.5–3.0 3.50 0.04 0.04 Cu 1.5–2.0
CT-15C ... N08151 0.05–0.15 ... 19.0–21.0 31.0–34.0 0.15–1.5 ... 0.50–1.5 0.04 0.04 Nb 0.5–1.5
Heat resisting alloys
HA 504 J82090 0.20 ... 8–10 ... 1.0 0.9–1.2 0.35–0.65 0.04 0.04 ...
HC 446 J92605 0.50 ... 26–30 4 max 2.0 0.5(b) 1.0 0.04 0.04 ...
HD 327 J93005 0.50 ... 26–30 4–7 2.0 0.5(b) 1.5 0.04 0.04 ...
HE 312 J93403 0.20–0.50 ... 26–30 8–11 2.0 0.5(b) 2.0 0.04 0.04 ...
HF 302B J92603 0.20–0.40 ... 19–23 9–12 2.0 0.5(b) 2.0 0.04 0.04 ...
HH 309 J93505 0.20–0.50 0.2 24–28 11–14 2.0 0.5(b) 2.0 0.04 0.04 ...
HI ... J94003 0.20–0.50 ... 26–30 14–18 2.0 0.5(b) 2.0 0.04 0.04 ...
HK 310 J94224 0.20–0.60 ... 24–38 18–22 2.0 0.5(b) 2.0 0.04 0.04 ...
HK-30 ... J94203 0.25–0.35 ... 23.0–27.0 19.0–22.0 2.0 0.5(b) 2.0 0.04 0.04 ...
HK-40 ... J94204 0.35–0.45 ... 23.0–27.0 19.0–22.0 2.0 0.5(b) 2.0 0.04 0.04 ...
HL ... N08604 0.20–0.60 ... 28–32 18–22 2.0 0.5(b) 2.0 0.04 0.04 ...
HN ... J94213 0.20–0.50 ... 19–23 23–27 2.0 0.5(b) 2.0 0.04 0.04 ...
HP ... N08705 0.35–0.75 ... 24–28 33–37 2.0 0.5(b) 2.0 0.04 0.04 ...
HP-50WZ ... ... 0.45–0.55 ... 24–28 33–37 2.5 0.5(b) 2.0 0.035 0.035 W 4.0–6.0, Zr 0.1–1.0
HT 330 N08605 0.35–0.75 ... 13–17 33–37 2.5 0.5(b) 2.0 0.04 0.04 ...
HT-30 ... N08603 0.25–0.35 ... 13.0–17.0 33.0–37.0 2.5 0.5(b) 2.0 0.04 0.04 ...
HU ... N08005 0.35–0.75 ... 17–21 37–41 2.5 0.5(b) 2.0 0.04 0.04 ...
HW ... N08006 0.35–0.75 ... 10–14 58–62 2.5 0.5(b) 2.0 0.04 0.04 ...
HX ... N06050 0.35–0.75 ... 15–19 64–68 2.5 0.5(b) 2.0 0.04 0.04 ...
Notes: All compositions include Fe as balance. Single values are maximum values unless otherwise noted. (a) The wrought equivalent composition is not the same as the cast. (b) Mo is not an intentional addition.
APPENDIX 2
Physical and Mechanical Properties of

Select Alloys
Table A2.1 Physical properties of major stainless steel engineering alloys

Coefficient of Thermal Electrical
Modulus of thermal exp., conductivity, Specific heat, resistivity,
UNS Density, kg/dm3 elasticity, GPa 10–6 × K–1 W/M·°K J/kg·°K Ω·mm2/m
201 S20100 7.86 207 16.6 16.3 502 0.67
301 S30100 8.03 193 16.6 16.3 500 0.73
304 S30400 7.90 200 16.6 16.3 500 0.72
304L S30403 7.90 200 16.6 16.3 500 0.72
305 S30400 7.90 200 16.6 16.3 500 0.72
316L S31603 8.00 200 16.5 14.6 480 0.74
321 S32100 7.92 193 16.6 16.3 500 0.72
904L N08904 7.95 190 15.3 13.2 460 0.95
AL6-XN© N08367 8.06 200 15.3 11.8 474 0.89
409 S40920 7.76 200 10.5 25.0 477 0.60
430 S43000 7.70 200 10.3 23.9 460 0.60
439 S43035 7.70 200 10.2 24.2 460 0.63
468 S46800 7.76 200 10.5 25.0 477 0.60
410 S41000 7.65 200 10.5 24.9 460 0.56
2101 S32101 7.8 200 13.5 17.0 500 0.80
2003 S32003 7.72 210 13.5 17.0 510 0.80
2205 S32205 7.8 200 14.6 16.5 500 0.80
2507 S32750 7.8 200 12.5 13.5 500 0.80
Table A2.2 Typical minimum mechanical properties of representative stainless steel engineering alloys
Tensile strength,
Name Condition UNS Yield strength, MPa MPa Elongation, % Hardness
201 Annealed S20100 260 min 550 min 40 min 100 Rb max
201F 2B S20100 330 700 51 89 Rb
301 tensile 2D S30100 320 850 49 88 Rb
1 hard
301 /4 S30100 580 900 32 25 Rb
1
301 /2 hard S30100 815 1150 23 35 Rc
301 3
/4 hard S30100 1000 1270 17 40 Rc
301 Full hard S30100 1160 1380 12 42 Rc
301 sink 2D S30100 270 690 57 82 Rb
304 Hot rolled S30400 335 640 51 86 Rb
304 2D S30400 265 625 55 81 Rb
304 2B S30400 305 635 52 85 Rb
304 #4 polish S30400 325 650 51 85 Rb
304 2BA S30400 315 640 53 85 Rb
304 1
/4 hard S30400 705 890 23 29 Rc
304L Annealed S30403 170 min 485 min 40 min 92 Rb max
304L 2D S30403 255 590 53 80 Rb
304LT 2D S30403 255 600 51 81 Rb
304DD 2D S30400 270 610 55 82 Rb
304EDD 2D S30400 260 600 56 78 Rb
305 2D S30500 245 560 52 73 Rb
316L Annealed S31603 170 min 485 min 40 min 95 Rb max
316L 2B S31603 310 595 51 82 Rb
321 2B S32100 285 570 49 78 Rb
904L Annealed N08904 220 min 490 min 35 min 90 Rb max
904L 2B N08904 270 605 50 79 Rb
AL6-XN© Annealed N08367 310 min 690 min 30 min 100 Rb max
AL6-XN© 2B N08367 365 745 47 88 Rb
409 2D S40920 260 440 31 60 Rb
430 2B S43000 345 515 27 67 Rb
439 2D S43035 315 455 32 76 Rb
468 2D S46800 205 415 32 76 Rb
29-4C Annealed S44735 415 min 550 min 18 min 25 Rc max
29-4C 2D S44735 550 650 20 20 Rc
410 2B S41000 320 515 28 81 Rb
2101 Annealed S32101 530 min 700 min 30 min ...
(a) Finish conditions: 2D is cold rolled, annealed, and pickled; 2B is 2D with an added temper mill pass (approximately 0.5% reduction); 2BA is cold rolled, bright an-
nealed, and temper passed.
APPENDIX 3
Introduction to Thermo-Calc and

Instructions for Accessing Free
Demonstration Version
WITHIN THE MAIN BODY of this text- Thermo-Calc is used in conjunction with
book, a number of diagrams have been plotted thermodynamic databases containing polyno-
and attributed to a software package called mial functions that describe the Gibbs energies
Thermo-Calc. The purpose of this appendix is of the different phases according to certain
to give a brief introduction to Thermo-Calc, ex- models that take into consideration nonideal
plain what it is, and what are its uses. Also pro- chemical interactions in solution phases. These
vided are instructions for accessing a demon- databases are based on the critical evaluation of
stration version of the software. thermodynamic and phase equilibria data for bi-
nary, ternary, and some higher-order systems,
which are then assembled into self-consistent
What Is Thermo-Calc? databases. Different databases are available for
different broad classifications of materials, sys-
Thermo-Calc (Ref 1) is a powerful, flexible tems, or applications. For example, there are
software package available from Thermo-Calc databases for steels and iron-based alloys; iron-
Software AB for performing various kinds of based slags; nickel superalloys; aluminum,
thermodynamic and phase diagram calculations magnesium, titanium, and zirconium alloys; ce-
for multicomponent systems. mented carbides; nuclear materials; and more.
The software is based on the so-called CAL- Further information on the different databases
PHAD (CALculation of PHAse Diagrams) available can be found at the Thermo-Calc Web
method (Ref 2), which describes mathemati- site: www.thermocalc.com
cally the thermodynamics of a system through a The thermodynamic database for steels (Ref
representation of the Gibbs energies of the dif- 3), as developed by Thermo-Calc Software AB,
ferent crystalline phases relevant to that system was used in conjunction with Thermo-Calc for
and defined by the chemical composition of the all the calculations made during the preparation
system. Thermo-Calc minimizes the total Gibbs of this book. The version of the database used
energy of the system with respect to various for these calculations contains data for 20 ele-
constraints such as temperature, pressure, and ments and 85 phases.
chemical composition and thus predicts the Although the databases are based primarily
most stable energy state (or equilibrium state) on the critical assessments of binary, ternary,
that can form. By suspending certain phases and some quaternary systems, the CALPHAD
(i.e., manually removing certain selected phases methodology provides a theoretical framework
from the system and thus restricting the forma- on which extrapolations can be made to predict
tion of such phases), Thermo-Calc can also be the phase equilibria for higher-order, multi-
used to investigate meta-stable equilibria-type component systems (the higher the order of the
problems. system, the weaker the nonideal interaction
parameters become). Thermo-Calc can there- be used to perform calculations for most appli-
fore be used in conjunction with such databases cations involving phase equilibria, meta-stable
to make predictions for multicomponent sys- equilibria, phase transformations, phase dia-
tems and alloys of industrial relevance as illus- grams, and various thermodynamic properties,
trated by some of the examples given in the as well as critical assessments and data evalua-
main body of this book. These calculations can tions for multicomponent systems.
be validated against real alloy data if this infor- While many types of calculations can be
mation is available but is not based on (or ad- made using Thermo-Calc, the software typically
justed to fit) such higher-order alloy data. is used to predict:
Higher order in this sense means more than four
• Stable and meta-stable phase equilibria for
elements (i.e., larger than a quaternary system).
Four specific types of calculation can be per- binary, ternary, and higher-order systems
formed using Thermo-Calc, although the range (calculations for alloy compositions with 6,
of problems to which these can be applied is 10, 15 elements are not uncommon, as illus-
broader: trated by some of the examples in the main
body of this book).
1. Single-Point Equilibria: The temperature, • Amounts of phases (mass, volume and mole
pressure, composition/activity of a compo- fractions) formed (phase balance) as a func-
nent (or the amount of a phase) are fixed and tion of temperature, pressure, and composi-
the stable or meta-stable equilibrium for the tion and also the chemical compositions of
specified conditions is calculated. the phases formed
2. Step: The amount of one state variable pa- • Phase transformation temperatures such as
rameter (or condition) can be changed, while liquidus, solidus, and solvus temperatures.
the other conditions remain fixed. For exam- Phase transformation temperatures can be
ple, to see how the different phases and their predicted based on the actual chemistry (not
amounts and compositions would vary with nominal chemistry).
temperature for a given alloy, one would • Thermochemical data such as enthalpies,
“step in” temperature. Alternatively, one can heat capacity, and activities
vary the composition of one of the compo- • Driving forces for precipitation
nents/elements and calculate how the phase • Phase diagrams (isothermal and isoplethal
amounts change for a fixed temperature or sections for multicomponent, multiphase
predict how the solidus or liquidus would systems as illustrated in this book)
change with varying alloy composition. • Molar volume, density, and thermal expan-
3. Map: Two axis variables (such as tempera- sion
ture, pressure, composition, or activity of • Scheil-Gulliver (nonequilibrium) solidifica-
the components) are changed at the same tion simulations
time. Isoplethal sections are generated by
varying temperature and composition of one Thermo-Calc is not restricted just to model-
of the components. Isothermal sections are ing the alloy. Complex systems representing
the result of varying the amounts of two of processing, for example, can also be consid-
the components for a fixed temperature. Ex- ered. For example, another application is to cal-
amples of each of these kinds of diagrams culate the carbon potential of multicomponent
are given in the main body of the text. gas phase systems as a function of composition,
4. Scheil: Thermo-Calc includes a Scheil- temperature, and pressure and then predict what
Gulliver model for nonequilibrium solidifi- phases an alloy might form at a given tempera-
cation and a modified Scheil model that con- ture when exposed to such a carbon potential.
siders partial equilibrium for components Thermo-Calc can thus be applied to a number
that are selected by the user. of practical problems related to metallurgy, pro-
cessing, in-service performance, etc. as summa-
rized by:
Applications of Thermo-Calc • Alloy Design: Modification of alloy chemis-
tries to improve properties or reduce costs
Thermo-Calc is used around the world within using calculations to guide which composi-
academia, in government research laboratories, tions may be most suitable before preparing
and by commercial industry. The software can them for testing
Appendix 3: Introduction to Thermo-Calc / 283
• Heat Treatment: Prediction of formation and are supplied with only certain small data-
of problematic phases prior to thermal pro- bases that are for demonstration purposes and
cessing the evaluation of the software only.
• Casting: Calculation of liquidus and solidus A link to register and download the demon-
temperatures; calculation of thermodynamic stration version of the software can be ac-
properties of the alloy for input into casting cessed via a link on the Thermo-Calc web site
modeling codes at www.thermocalc.com. All fields in the regis-
• Welding and Joining: Prediction of the tration form should be completed before contin-
phases formed at the joining of two dissimilar uing to the download page, where further in-
materials or the interaction with filler material structions regarding installation of the software
• Quality Control: Investigation of properties will be provided.
and phase balance within designated compo- On installation of the software, additional
sitional tolerances documentation, including a Users Guide/Exam-
ples manual in the form of PDF files will also be
More examples are available in the literature installed. Technical support for the demo ver-
(search on key terms Thermo-Calc or CAL- sions of the software is limited, but problems
PHAD). A list of published articles citing related to installation or general inquiries can be
Thermo-Calc is available at www.thermocalc. addressed by visiting www.thermocalc.com and
com. linking to their support.
The demo version will run for approximately
1 month on a single computer, and installation
How to Obtain a Free Demonstration on a network system is not supported. If you
Version of Thermo-Calc wish to run the software after the demo license
has expired, it can be downloaded again (i.e.,
Thermo-Calc is available in two formats: obtaining a new demo license).
Thermo-Calc Classic, which has a command
line interface and can be run under a number of REFERENCES
different operating systems (including Mi-
crosoft Windows and Linux/Unix), and 1. J.O. Andersson, T. Helander, L. Höglund,
Thermo-Calc for Windows, which has an easy- P.F. Shi, and B. Sundman, Thermo-Calc and
to-learn graphical user interface but only oper- DICTRA, Computational Tools for Materials
ates in the Microsoft Windows environment. Science, Calphad, Vol 26, 2002, p 273–312,
Demo versions are available for both of these 2002
versions of the software. 2. N. Saunders and A.P. Miodownik, Pergamon
The demo versions are free to use, subject to Materials Series, CALPHAD (Calculation of
the terms outlined in the Thermo-Calc Software Phase Diagrams): A Comprehensive Guide,
End User License Agreement. It should be 1, Elsevier, 1998
noted that the demo versions are limited to 3. TCFE5—TCS Steel/Fe-Alloys Database,
using just three elements (whereas in the full Version 5.0, 2007, Thermo-Calc Software
product the current upper limit is 40 elements) AB, www.thermocalc.com
Copyright © 2008 ASM International®. All rights reserved.
Stainless Steels for Design Engineers (#05231G) www.asminternational.org
Index
A isocorrosion curves for, in sulfuric acid, 33(F)

isocorrosion curves for, sulfuric acid with chlorides, 34(F)
acetic acid alloy systems
corrosion rates for various alloys of, plus formic acid, alloying elements, 1
103(F) austenitic alloys, 69, 72
duplex alloys, 102 austenitic stainless family, 71(F)
isocorrosion curves in, 36(F) body-centered cubic (bcc) phase, 1
isocorrosion performances of various alloys, 102(F) families in perspective, 69–72
acids, corrosion in, and bases, 31–36 most widely used, 1
adsorption-induced brittleness, 51, 53 Schaeffler–Delong stainless steels constitution diagram,
adsorption-induced plasticity, 51, 52–53 5(F), 70(F)
aeration cells, differential, 38–39 alpha prime
aerobic bacteria, influencing corrosion, 55 formation kinetics for duplex alloys, 94, 96(F)
aesthetic finishes influence of, formation on hardness, 116(F)
Chrysler Building in New York City, 196, 197(F), 213(F) iron-chromium phase, 8
considerations, 217–219 martensite, 7–8, 73
flatness, 219 transition temperature change with, formation with
surface, 217–219 aging, 98(F)
aging treatments, precipitation hardening stainless alteration, surface, 199
steels, 168, 170 aluminum
aggressive chemical agents, refinery applications, 255 inclusions and pitting, 40, 41
Alloy Casting Institute (ACI) influence on thermodynamic activity of C, N, S and O,
composition of, heat- and corrosion-resisting casting 157(T)
alloys, 277(T), 278(T) metal migration into acetic solution from, 234(T)
naming system, 147 migration into acetic solution from stainless, aluminum
alloy design or carbon steel, 234(T)
avoiding unwanted phases, 94 oxidation resistance, 79, 226, 228
duplex alloys, 92–94 protective layer formation, 64–65
Thermo-Calc software, 93, 282 aluminum alloys, 1
alloying, sensitization, 47–48 aluminum oxide, 191
alloying elements aluminum-titanium-nitride (AlTiN), 191
alloys, 1 aluminum/titanium precipitates, possible, 138(F)
influence on alloy families, 147, 149 American Iron and Steel Institute (AISI), 240
influence on corrosion rate in contaminated sulfuric acid, American Petroleum Institute (API), 135
33 (F) American Society for Testing and Materials (ASTM),
influence on thermodynamic activity of C, N, S and O, 238–239
157(T) American Society of Mechanical Engineers (ASME), 266
influence on uniform corrosion, 29–30 anisotropy
martensitic precipitation-hardening stainless steels, deep drawing, 174
141(T) ferritic stainless steels, 120–121
martensitic stainless steels, 130, 131(T) Lankford ratio, 120, 175
alloy oxidation, behaviors, 66 stainless long products, 179
alloys annealing
influence of, content on corrosion rate in hydrochloric austenitic stainless steels, 162–164
acid, 35(F) bright, 198
286 / Index
annealing (continued) atmospheres, oxidation, 66–67

deep drawing, 177 atmospheric corrosion
duplex stainless steels, 170–171 uniform corrosion, 36–37
ferritic stainless steels, 165–166 atomic rearrangements, 2
long-term, of welds, 43 attraction, interatomic, 2
martensitic stainless steels, 166 austenite
precipitation hardening steels, 139 alloying elements, 5
anode carbide precipitation, 9
electrochemical reactions, 12 carbon and nitrogen, 6, 9
polarization, 20–21, 23 diffusion rates, 6
appliances face-centered cubic (fcc), 5
facades, 239–240 γ-austenite in precipitation hardening alloys, 138–139
kitchen, 237–240 interstitial elements, 6
laundry, 241 lattice expansions, 6(F)
stainless steels commonly used for, 239(T) lean alloy of martensite and, 73–74
architecture and construction mechanical properties, 7
aesthetic considerations, 217–219 metastable state, 6
average chloride concentration in rainwater in United phase in duplex alloy at room temperature, 91
States, 217(F) Schaeffler–Delong constitution diagram, 5(F), 70(F)
balancing corrosion resistance, processing and economy, Schaeffler diagram, 202(F)
214–215 secondary type, 7
balancing service environment, design and maintenance, semiaustenitic precipitation-hardenable stainless steel,
215, 217 143
cleaning methods for uncoated stainless steel, 220(T) sulfur and oxygen, 6
concrete reinforcing bar, 222 austenite conditioning, 169
corrosion resistance, 213–214 austenitic-ferritic “C” alloys, 151–152
design considerations, 216(F) austenitic “H” alloys
ecological considerations, 222 high temperature HE–HP, 152–154
environment, 216(F) precipitation hardening stainless steel, 170
fabrication and joining, 221 austenitic precipitation-hardenable stainless steels
fabrication considerations, 221 cold work and aging, 146(F)
fire resistance, 221–222 composition, 145(T)
flatness, 219 corrosion resistance, 145–146
grades recommended by expert system, 217(F) mechanical properties, 145
graphic depicting low release of metal ions from 304 austenitic stainless steels
and 316 stainless steels, 222(F) alloy families, 69–72
local weather pattern, 216(F) annealing, 162–164
maintenance, 220–221 automotive structural components, 229
maintenance schedule, 216(F) carburization, 82
roof, 219(F) composition of, 270(T), 271(T)
ranking common stainless steels by pitting resistance composition of high-temperature, 82(T)
equivalent number (PREN), 214(T) compositions of commonly used lean, 72(T)
repair, 221 compositions of corrosion-resistant, 86(T)
rolled-on stainless steel finishes, 218(F) corrosion resistance ratings, 87(T)
salt exposure, 216(F) corrosion-resistant alloys, 84–89
special finishes, 218(F) corrosive environments, 88–89
stainless steel selection expert system, 216(F) critical pitting temperature (CPT), 43, 44(F)
surface finish aesthetics, 217–219 drawability, 176
surface finish and corrosion resistance, 215 ductility, 180
argon oxygen decarburization (AOD) face-centered cubic (fcc), 174
adoption, 70 family, 71(F)
alloy adjustment, 157 forming limit diagram of carbon steel and, 176(F)
automotive industry, 225 halogens, 82
cleanliness, 184 heat exchangers, 246
control of nitrogen in refining by, 92 high-temperature alloys, 79–83
ferrite, 4 high-temperature mechanical properties, 82–83
first commercial use, 109 impact strength variation with temperature, 75(F)
foundry practice, 154 intermetallic phases, 82, 203
inclusions in steel, 40 isocorrosion curves for, in hydrochloric acid, 34(F)
production process, 155 kitchen appliances, 239
Index / 287
lean alloys, 72–78 tailpipe, 227(T), 228

machinability, 185 tensile properties of carbon and stainless steels for auto-
machining setup recommendations, 183(T) mobiles, 229(T)
martensite and austenite, 73–74 trucks, 231
mechanical properties, 74–76, 82–83
mechanical properties after cold work and annealing, 163
B
metal migration into acetic solution from, 234(T)
nitriding, 82 bacteria influencing corrosion, 55
oxidation resistance, 79–81 bacterial retention, food contact materials, 236(F)
petroleum industry applications, 249(T) bases, corrosion in acids and, 31–36
pitting resistance equivalent number (PREN), 43, 78, 85 basic oxygen furnace (BOF), 156
precipitation of carbides and nitrides, 76–78 Bauschinger effect, 164
recommended thermal processing temperatures, 162(T) biocorrosion, 55–56
resilience and toughness of carbon steel vs. for automo- biological neutrality, food contact, 235
tive components, 229(T) bleaching, pulp, 266
SCC (stress corrosion cracking), 49–50 body-centered cubic (bcc) phase
secondary phases in, 82(T) carbon and alloy steels, 1
sensitization, 46–47 change to face-centered cubic (fcc), 127(F)
soaking, 161–162 ferrite, 4, 110
stabilization, 78 ferritic material, 174
stainless steel in shipping, 245 metals, 2
SCC performance, 262 boron
stress-strain curve for single crystals of stable, 53(F) additions to ferritic stainless steels, 121
surface finish, 89 ferrite, 4
tensile properties of carbon steel vs. for automotive brazing, 211
components, 229(T) bright annealing, 198
thermal processing, 161–164 brightening stainless steels, 196
water vapor, 81–82 buffing, 197
weaknesses, 69 built-up edge (BUE)
welding characteristics, 201–204 austenitic stainless grades, 185
welding parameters, 207(T) carbon and, 182
austenitizing, martensitic stainless steels, 131, 132(F), coolants minimizing, 191
166–167 copper and, 183
automotive and transportation ferritic stainless steels, 185
alloy selection for exhaust systems, 226 grain sizes, 184
alloys for major elements of automotive exhaust sys- iron and tool, 182
tems, 227(T) nickel and, 183
automotive emission standards, 225 nitrogen and, 184
bus bodies, 231(F) precipitation hardening stainless steels, 185
car manufacturers, 230 bus
catalytic converter, 227(T), 228 life-cycle cost calculation for stainless vs. carbon steel,
center pipe, 227(T), 228 231(T)
decorative to highly engineered applications in automo- stainless steel body, 230, 231(F)
biles, 225 Butler–Volmer equation, 20–21
exhaust manifold and high-temperature, 227
exhaust systems, 225–228
ferritic stainless, 226
C
flexible pipe, 227(T), 228 calcium
front pipe, 227(T), 228 effect on machinability of 303, 188–189, 189(F)
fuel tanks, 231 inclusions and pitting, 41
life-cycle cost calculation for stainless vs. carbon steel calcium-fluoride based slag
for bus, 231(T) electroslag remelting (ESR), 158
microcar frame, 231, 232(F) “C” alloys
muffler, 227(T), 228 austenitic–ferritic alloys, 151–152
rail transport, 232 corrosion resisting, 147
resilience and toughness of carbon and stainless steels duplex alloys, 151
for automobiles, 229(T) mechanical properties of corrosion resisting cast, 150(T)
stress-strain curves for 301 variants vs. duplex steels and metallurgy of, 149, 151–152
transformation steel, 230(F) precipitation hardening, 151
structural components, 229–231 carbide, precipitation kinetics, 7(F)
288 / Index
carbides naming system, 147

duplex alloys, 94 precipitation hardening, 151
flatware, 240 room temperature mechanical properties of corrosion
precipitation, 76–78 resisting stainless, 150(T)
stainless steel, 9 welding, 154
tooling, 191 catalytic converter, 227(T), 228
carbon catastrophic oxidation, 65
alloying element, 1 cathode
austenite, 5 effect of, polarization, 24(F)
ferrite, 3–4 electrochemical reactions, 12
influence of alloying elements on thermodynamic activ- mass transfer limitations, 24
ity of, 157(T) polarization, 20–21, 23
influence on thermodynamic activity of C, N, S and O, caustic solutions, 50
157(T) center pipe of exhaust systems, 227(T), 228
influence on uniform corrosion, 29 cerium
interstitial atoms of, in austenite, 6 inclusions and pitting, 41
machinability of stainless steels, 182–183 oxidation resistance, 79–80
precipitation of carbides, 76–78 protective layer formation, 65
precipitation rates by, content, 76(F) Charpy V toughness
steel content, 156 high-temperature austenitic alloys, 83(F)
variation of martensite hardness with, 128(F) niobium-stabilized alloy, 119(F)
welding of austenitic stainless steels, 201 titanium-stabilized alloy, 119(F)
carbon diffusion, 115 chemical agents, 255
carbon dioxide, wet, 248 chemical and process industry. See also corrosion types
carbon solubility, austenitic stainless, 76(F) corrosion table for fuming sulfuric acid, 261(T)
carbon steel corrosion table for sulfuric acid, 259(T), 260(T), 261(T)
activities and activity coefficients of elements in, 40(T) corrosion types, 258–262
body-centered cubic (bcc) phase, 1 erosion, 262
corrosion rates of stainless vs., 135(F) forms of corrosion, 258
metal migration into acetic solution from, 234(T) intergranular corrosion, 261–262
resilience and toughness of, vs. stainless steel for auto- isocorrosion chart for sulfuric acid, 258(F)
motive components, 229(T) pitting and crevice corrosion, 258–259, 261
tensile properties of, vs. stainless steel for automotive single- and dual-environment systems, 257–258
components, 229(T) specific corrosives, 262
carburization, 82 stress corrosion cracking (SCC), 257, 262
carburizing, 199 chemical neutrality, food contact, 233–235
casting chemistry, machinability of stainless steels, 182–184
stainless steel processing, 158–159 chi
Thermo-Calc software, 283 intermetallic phase, 9
casting alloys precipitation kinetics, 7(F)
Alloy Casting Institute (ACI), 147 chip breaking, 187(F)
austenitic-ferritic alloys, 151–152 chloride concentration in rainwater, 217(F)
austenitic HE–HP alloys, 152–154 chloride-containing solutions, 50–51
chromium alloys, 148(T) chloride ion, aggressive against stainless steel, 85
chromium-nickel alloys, 148(T) chlorinated oils or waxes, 179(T)
composition of cast heat-resistant stainless and nickel chromium
base alloys, 149(T) alloying element, 1
compositions of cast stainless corrosion resisting alloys, austenitic alloys, 72
148(T) austenitic stainless steels, 85–86
duplex alloys, 151 chromium-oxygen system volatility vs. temperature and
ferritic HA, HC, HD alloys, 152 oxygen pressure, 64(F)
foundry practice, 154 corrosion resistance, 228
high-temperature mechanical properties of “H” alloys, depletion from austenite near grain boundaries, 77(F)
153(T) ferrite, 3
influence of alloying elements, 147, 149 ferritic stainless alloys with low, medium and high, 110
mechanical properties of heat-resistant stainless, at room inclusions and pitting, 42, 43
temperature, 152(T) influence on thermodynamic activity of C, N, S and O,
metallurgy of “C” alloys, 149, 151–152 157(T)
metallurgy of “H” alloys, 152–154 influence on uniform corrosion, 29, 30(F)
molten metal transfer, 153 ion release from stainless steel grades, 222(F)
Index / 289
machinability of stainless steels, 182 construction. See architecture and construction

migration into acetic solution from stainless, aluminum contamination, 235
or carbon steel, 234(T) continuous slab casting, 158
oxidation resistance, 71, 79, 80(F), 228 cookware, 237
paralinear oxidation from evaporation of chromium coolants, 191
superoxide, 64(F) copper
Pourbaix diagram, 17(F) acid resistance, 71
sulfide formation, 186 machinability of stainless steels, 183
thermodynamics of oxidation, 57–59 copper sulfate. See also sulfuric acid plus copper sulfate
volatile nature of Cr2O3, 63, 64(F) corrosion of stainless steel and titanium in, plus sulfuric
chromium alloys, cast stainless, 148(T) acid, 31(F), 32(F)
chromium-nickel alloys, 148(T) corrosion
Chrysler Building combating, in alloys for petroleum industry, 248, 250
architecture using stainless steel, 213 definition, 11
highly polished surface, 196, 197(F), 213(F) erosion, 262
cleaning intergranular, 261–262
passivation, 195–196 isocorrosion chart for sulfuric acid, 258(F)
recommended methods, 195(T) pitting and crevice, 258–259, 261
stainless steel, 194–196 pitting resistance equivalent number (PREN), 258
cleaning methods single- and dual-environment systems, 257–258
stainless steels, 235–236 stress corrosion cracking (SCC), 257, 262
uncoated stainless steel, 220(T) table for sulfuric acid, 259(T), 260(T), 261(T)
cleanliness tendency, 15–16
food contact materials, 235–236 types, 258–262
machinability of stainless steels, 184 corrosion cost, 252–253
coastal climates, 237–238 corrosion kinetics
coatings Butler–Volmer equation, 20–21
cookware, 237 introduction, 19–20
tooling, 191 mass transfer control, 21
cold heading, 179–180 migration and ionic diffusion, 21–22
cold work, 75 mixed potential theory and polarization diagrams,
coloring of stainless steels, 196 22–23
commercial use passivation, 23–25
applications, 237–242 Tafel regime: electrode-kinetics control, 21
cookware, 237 corrosion rate
flatware and cutlery, 240–241 alloys in simulated evaporator liquid, 37(F)
food contact, 233–237 influence of alloying element on, in contaminated
heating and water heating, 241–242 sulfuric acid, 33(F)
kitchen appliances, 237–240 stainless oil country tubular goods, 135(F)
laundry appliances, 241 stainless vs. carbon steel, 135(F)
stainless steel, 233 Tafel slope, 23(F)
composition vs. surface roughness, 215(F)
Alloy Casting Institute (ACI) heat- and corrosion- corrosion resistance
resisting casting alloys, 277(T), 278(T) architecture, 213–214
austenitic precipitation-hardenable (PH) stainless steel, austenitic precipitation-hardenable (PH) stainless steel,
145(T) 145–146
austenitic stainless steels, 270(T), 271(T) balancing, processing and economy, 214–215
duplex alloys commercially available, 97 (T) duplex alloys, 99–106, 204
duplex stainless steels, 276(T) ferritic stainless steels, 109, 110, 121–122
ferrite stainless steels, 111(T), 112(T), 273(T), 274(T) function of salinity and temperature, 244(F)
martensitic PH stainless steels, 140(T) martensitic PH stainless steels, 141–142
martensitic stainless steels, 124(T), 125(T), 275(T) material selection for desalination, 244–245
PH stainless steels, 276(T) pulp-and-paper industry, 265–267
semiaustenitic PH stainless steel, 143(T) rail transport applications, 232
tool and cutlery martensitic stainless steels, 134(T) ratings of austenitic stainless steels, 87(T)
concrete reinforcing bar, 222 semiaustenitic PH stainless steel, 144
constitution diagram stainless steel for refinery equipment, 254–255
Schaeffler–Delong stainless steels, 5(F), 70(F) stainless steel in shipping, 245
Schaeffler diagram, 202(F) sulfur hurting, 188
Welding Research Council’s 1992, 203(F) surface finish and, 215
290 / Index
corrosion resisting alloys creep rupture strength, 83, 84(F)

austenitic stainless steels, 84–89 creep strength, 83(F)
“C” alloys, 149, 151–152 crevice corrosion
composition of, austenitic stainless steels, 86(T) austenitic stainless steels, 85, 87
composition of Alloy Casting Institute (ACI), 277(T), corrosion type, 214, 258–259, 261
278(T) critical, temperature with alloy content, 45(F)
compositions of cast stainless, 148(T) critical crevice temperature (CCT) and critical pitting
duplex alloys, 91 temperature (CPT), 105(F)
mechanical properties of stainless, 150(T) dissimilar metals and differential aeration cells,
corrosion theory 38–39
corrosion tendency, 15–16 duplex alloys, 103–104
electrochemical reactions, 11–12 geometry, 45
Faraday’s law, 12 preventing, 45–46
galvanic vs. electrochemical cells, 14 critical crevice temperature (CCT), 105(F)
Nernst equation, 12–14 critical current density, 29, 30(F)
Pourbaix diagrams, 16–17 critical pitting temperature (CPT)
standard half-cell reduction potentials vs. normal hydro- austenitic steels, 43, 44(F)
gen electrode, 14(T) critical crevice temperature (CCT) and CPT, 105(F)
corrosion types duplex alloys, 103, 104(F), 105(F)
atmospheric, 36–37 stainless steels for unwelded and welded material,
biocorrosion and microbiologically induced, 55–56 44(F)
chromium influence, 29 vs. pitting resistance equivalent number (PREN), 85(F),
corrosion fatigue, 55 104(F)
corrosion in acids and bases, 31–36 cryogenic containers, 245
corrosion with fatigue or fraction, 48–55 current density, 23(F)
crack initiation, 48–49 cutlery
crack propagation, 49, 52(F) flatware and, 240–241
crevice, 38–39, 45–46 martensitic stainless steels, 133–134, 240
critical current density, 29, 30(F) stainless steels commonly used for, 241(T)
dissimilar metals and differential aeration cells, 38–39 cutting tools, 133–134
environmental variables, 50–51
environmental variables influencing uniform corrosion,
28–29 D
grain boundary, 46–48 deep drawing
hydrochloric acid, 33–34, 35(F) anisotropy, 174
hydrogen embrittlement, 54–55 forming stainless steel, 173, 174–179
influence of alloying elements, 29, 30(F) geometry, 174–175
localized, 37–38 hydroforming, 177–178
material variables, 29–31, 49–50 intermediate annealing, 177
molybdenum role, 29–30 materials composition, 175(T)
nickel, 30 schematic, 174(F)
nitric acid, 34, 35(F) strain rate, 177
nitrogen, 30 texture, 174
organic acids, 35, 36(F) tooling, 176, 178–179
phosphoric acid, 34–35, 36(F) defects, 160
pitting, 39–45 delignification of pulp, 266
pitting resistance, 43–45 demand for steel, 247
preventing crevice, 45–46 desalination
SCC (stress corrosion cracking), 48–54 materials selection for, 244–245
SCC mechanisms, 51–54 multi-stage flash (MSF), 243
sensitization, 46–48 reverse osmosis (RO), 243–244
sodium chloride/carbon dioxide environment, 30 technology, 243–244
strong bases, 35 design, balance, 215, 217
sulfuric acid, 31–33 designers
sulfuric acid plus copper sulfate, 31(F), 32(F) car manufacturers, 230
uniform, 27–37 pitting corrosion, 39
corrosive environments development
austenitic stainless steels, 88–89 precipitation–hardening stainless steels,
platforms, 254 137–138
refinery equipment, 254–255 welding, 211–212
Index / 291
differential aeration cells soaking, 170

active alloys, 38–39 sodium hydroxide, 101
microfouling, 56 stainless steel for line pipe, 253
schematic, 20(F) stainless steel in shipping, 245
differential aeration corrosion cell, 12 strength, 96
diffusion stress corrosion cracking (SCC) performance, 262
atomic rearrangements, 2 stress-strain curves for 301 variants vs., 230(F)
ionic transport, 21–22 structure and alloy design, 92–94
diffusion rates, austenite vs. ferrite, 6 sulfuric acid, 100
digesters thermal processing, 170–171
first kraft from alloy 2205, 266(F) umbilical tubing and risers, 253–254
pulp-and-paper industry, 265 variations of ferrite, austenite, and duplex with tempera-
disinfection, 236–237 ture, 98(F)
dissimilar metals, 38–39 welding characteristics, 204–205
dissociated ammonia, 198–199 welding parameters, 207(T)
dissolution wrought 2205 duplex microstructure, 91(F)
equation, 37
term, 27
dryers, laundry appliances, 241
E
dry film, forming stainless, 179(T) earing
dual-environment system, 257 deep drawing, 178
duplex alloys measuring tendency, 178
acetic acid, 102 ecological considerations, 222
annealing, 170–171 electrochemical cell
composition of selected, stainless steels, 276(T) closed circuit, 11–12
compositions, 97(T) potential, 38
concept, 91–92 electrochemical corrosion, 19
corrosion resistance, 99–106 electrochemical reactions, 11–12
corrosion-resistant “C” alloys, 151 electrode-kinetics control Tafel regime, 21
crevice corrosion, 103–104 electrodes, polarization, 20–21
deep drawing, 178 electrolysis cell, 14(F)
fastest-growing stainless steel family, 91–92 electrolyte resistance, 22
fatigue, 98 electrolytic cells, galvanic vs., 14
Fe-Cr-Ni phase diagrams, 92(F) electrolytic pickling, cold-rolled stainless, 194
formation kinetics, 96(F) electromotive force, 13
formic acid, 103(F) electropolishing, 196
forming and machining, 99 electroslag remelting (ESR), 157–158
heat exchangers, 246 embrittlement. See also hydrogen embrittlement (HE)
hot forming, 180 alpha prime, 8
hydrochloric acid, 100–101 high-temperature, 114
impact strength, 97–98 σ phase at higher temperatures, 151
impact strength variation with temperature, 75(F) engineering alloys
iron-nickel diagrams, 93(F) minimum mechanical properties of stainless steel,
machinability, 186 280(T)
machining setup recommendations, 183(T) physical properties of major stainless steel, 279(T)
mechanical properties, 94–98 environment, stainless steel selection expert system,
nitric acid, 101 216(F)
organic acids, 102, 103(F) environmental variables
partitioning of elements, 93–94 stress corrosion cracking, 50–51
petroleum industry applications, 248(T) uniform corrosion, 28–29
phosphoric acid, 101–102 epsilon martensite, 7–8, 73–74
photomicrographs, 95(F) equilibrium, argon oxygen decarburization (AOD),
pitting corrosion, 102–103 155–156
pitting resistance equivalent number (PREN), 43 equivalent weight (EW), 19
PREN influencing fatigue, 98, 99(F) erosion, corrosion, 258, 262
pulp-and-paper industry, 265, 266–267 expert system
recommended annealing and stress–relieving tempera- recommended stainless steel grades, 217(F)
tures, 170(T) stainless steel selection, 216(F)
SCC (stress corrosion cracking), 49, 104–106 exposure to salt, stainless steel selection expert system,
sensitization, 47 216(F)
292 / Index
F sensitization, 47
soaking, 165
fabrication, 221 stabilization, 109, 115, 118–120
facades of appliances, 239–240 stress corrosion cracking (SCC), 49
face-centered cubic (fcc) phase stress relieving, 166
aluminum alloys, 1 superferritics, 113
austenite, 5 texture and anisotropy, 120–121
austenitic materials, 174 time-temperature-transformation (TTT) curve for 430,
change to body-centered cubic (bcc), 127(F) 115(F)
metals, 2 titanium and niobium, 118
Faraday’s law, 12, 21 titanium for carbide and nitride formation, 115
fatigue toughness, 116(F), 117(F), 118–119
corrosion, 55 welding characteristics, 205–206
duplex alloys, 98, 99(F) welding parameters, 207(T)
fatty oils and blends, suitability in forming stainless ferromagnetism, 5
steel, 179(T) fingerprints, cleaning methods for uncoated stainless,
ferrite 220(T)
carbon and nitrogen, 3–4 fire resistance, stainless steel, 221–222
carbon diffusion rate, 115 flatness, surface aesthetic, 219
chromium, 3 flatware, 240–241
diffusion rates in austenite vs., 6 flexible pipe, 227(T), 228
δ-ferrite in precipitation hardening alloys, 138–139 flow lines, 252–253
hydrogen and boron, 4 flux cored wire (FCW) welding, 210
mechanical properties, 4 food contact
molybdenum, 4 bacterial retention by material and cleaning time, 236(F)
phase diagram of iron chromium, 3(F) biological neutrality, 235
phase in duplex alloy at room temperature, 91 chemical neutrality, 233–235
Schaeffler–Delong constitution diagram, 5(F), 70(F) cookware, 237
Schaeffler diagram, 202(F) flatware and cutlery, 240–241
stabilization with titanium, 4 heating and water heating, 241–242
thermal conductivity and thermal expansion, 4–5 kitchen appliances, 237–240
ferritic “H” alloys, 152 material cleanliness, 235–236
ferritic stainless steels metal migration into acetic solution, 234(T)
alpha prime formation, 116, 117(F) qualifications, 233–237
annealing, 165–166 stainless steels commonly used for appliances, 239(T)
automotive exhaust systems, 226 stainless steels commonly used for cutlery, 241(T)
body-centered cubic (bcc), 174 surface disinfection, 236–237
carbon diffusion rate in, 115 formability, ferritic stainless steel, 120
composition, 111(T), 112(T), 273 (T), 274 (T) formic acid
corrosion and oxidation resistance, 109, 121–122 austenitic stainless steels, 89
deep drawing, 178 corrosion in, 35
embrittling phenomenon, 116 corrosion rates for various alloys of acetic plus formic
forming limit diagrams, 176 (F) acid, 103(F)
groups of low, medium and high chromium, 110, 113 duplex alloys, 102
heat exchangers, 245–246 isocorrosion curves in, 36(F)
high-temperature properties, 121 forming limit diagram (FLD), 175–176
hot rolling, 159 forming technology
impact strength variation with temperature, 75(F) deep drawing, 173, 174–179
intermetallic phases, 116 deep drawing materials composition, 175(T)
iron-chromium phase diagrams, 113(F), 114(F) deep drawing schematic, 174(F)
kitchen appliances, 239 duplex alloys, 99
lowest cost and simplest stainless, 109–110 duplex stainless steel, 178
machinability, 185 ferritics, 120, 178
machining setup recommendations, 183 (T) flat, rolled stainless steel, 173–179
mechanical behavior, 116–117 forces for hot working, 180(F)
metallurgy, 113–116 forming limit diagram of carbon steel vs. austenitic
metal migration into acetic solution from, 234(T) stainless steel, 176(F)
petroleum industry applications, 247(T) forming limit diagrams for stainless steel categories,
pitting resistance equivalent number (PREN), 43 176(F)
recommended annealing temperatures, 165(T) hot, of stainless steel, 180
Index / 293
hydroforming, 177–178 grit sizes, 197(T)

hydrogen embrittlement, 177 Guinier–Preston (GP) zones, 138
limiting drawing ratio (LDR) vs. Lankford ratio, 175(F)
optimized 409 for forming vs. normal 409, 177(F)
orange peel, 178
H
stainless long products, 179–180 half-cell reactions
stainless steel, 173 reduction potentials, 13–14
stretch forming, 174 vs. normal hydrogen electrode, 14(T)
suitability of lubricants for use in, 179(T) “H” alloys
surface finish, 178 austenitic HE–HP alloys, 152–154
tooling, 176, 178–179 corrosion resisting, 147
foundary practice, casting alloys, 154 ferritic HA, HC, HD, 152
free energy, phases, 2 high-temperature mechanical properties of, 153(T)
friction stir welding, 212 mechanical properties of heat-resistant, 152(T)
front pipe, alloys in automotive exhaust systems, metallurgy, 152–154
227(T), 228 halogens, 82
fuel tanks, 231 hardening
fuming sulfuric acid, 261(T) austenitic stainless steels, 75
furnace, stainless steels in, 241–242 ferritic stainless steels, 116–117
heat-affected zone (HAZ)
austenitic stainless steel, 207
G chromium carbide formation in, 201
Gallionella, 55 duplex stainless steels, 204
galvanic cell ferritic stainless steels, 109, 205
electrochemical reaction, 38 laser welding, 210
schematic, 14(F) martensitic stainless steels, 206
vs. electrolytic cells, 14 secondary austenite, 7
gas metal arc welding (GMAW) heat exchangers, 245–246
joint design, 209(F) heat-resistant alloys
process, 210 composition of Alloy Casting Institute (ACI), 278(T)
gas tungsten arc welding (GTAW) compositions, 149(T)
joint design, 209(F) “H” alloys, 152–154
process, 208–210 mechanical properties of cast stainless, 152(T)
geometry heat tint, coloring stainless steels, 196
crevice corrosion, 45 heat tinting, cleaning method, 220(T)
deep drawing, 174–175 heat treatment, Thermo-Calc software, 283
pitting corrosion, 39 heat treatment and conditioning, 168–170
Gibbs free energy heavy-duty emulsions, 179(T)
electrochemical reactions, 12–13 heavy metals, elimination, 157
oxidation, 57, 58(F) high-frequency induction welding, 211
grade selection, corrosion resistance, processing and high-speed tool steels, 190–191
economy, 214–215 high-temperature alloys
graffiti, cleaning methods for uncoated stainless, austenitic stainless steel, 82(T)
220(T) intermetallic phases of austenitic stainless steel, 82
grain boundaries martensitic stainless steels, 133, 134(F)
austenite, 6 mechanical properties of austenitic, 82–83, 84(F)
austenite, of martensite, 8 oxidation resistance of austenitic, 79–81
boron additions to ferritics, 121 water vapor, 81–82
carbide precipitation, 9, 76–77, 77(F) high-temperature embrittlement, 114
corrosion, 46–48 high-temperature properties, ferritic stainless steels, 121
defects in stainless steel, 160 hopper cars, 232
depletion of chromium from austenite near, 77(F) hot ductility defects, 160
ferrite-austenite, 8 hot forming, 180
grain size hot mill defects, 160
austenitic stainless steel annealing, 163 hot rolling, 159–160
martensitic stainless steels and toughness, 131, hot Steckel mills, 159
132(F) hot strip tandem mills, 159
material structure, 184 hydrochloric acid
graphite, suitability in forming stainless steel, 179(T) austenitic stainless steels, 88, 89(F)
grinding, coarse polishing, 197 corrosion in, 33–34
294 / Index
hydrochloric acid (continued) phase diagrams with varying carbon, 114(F),

duplex alloys, 100–101 130(F)
influence of alloy content on corrosion rate in, 35(F) phase diagrams with varying chromium, 130(F)
isocorrosion curves for austenitic stainless steels in, 34(F) iron dissolution, 20(F)
isocorrosion curves for stainless steels in, 34(F) iron reduction, 15(F), 28(F)
isocorrosion performance of duplex, 101(F)
hydrofluoric acid, 193–194
hydroforming, 177–178
J
hydrogen/argon atmosphere, bright annealing, 198–199 JFI Steel, oil country tubular goods and line pipe
hydrogen embrittlement (HE) alloys, 135(T)
corrosion fatigue, 55 joining
crack growth, 49 stainless steel, 221
ferritic stainless steels, 121–122 Thermo-Calc software, 283
mechanisms, 54–55 joint design, 208, 209(F)
stress corrosion cracking, 51, 52–54
hydrogen ion reduction, 15(F), 28(F)
K
hypochlorite bleaches, 195
kinetics, alpha prime formation, 94, 96(F)
kitchen appliances
I austenitic stainless steel, 239
impact strength coastal conditions, 238
duplex alloys, 97–98, 204 exposure of stainless samples to North Carolina beach,
variation with temperature for stainless steels, 75(F) 238(F)
inclusion-related defects, 160 facades, 239–240
inclusions ferritic stainless grades, 239
chip breaking at sulfides, 187(F) food contact, 237–240
lead, selenium, tellurium, 186 interior or working parts, 239
oxides, 188–190 knife-line attack, 48, 202
pitting corrosion, 40–43 kraft process
role in machining stainless steels, 186–190 paper-making, 265–267
stainless steel, 10 pulp-and-paper industry, 36, 265
sulfur, 186–188
induction welding, high-frequency, 211
ingot method, 158, 159
L
initiation Lankford r, earing tendency, 178
pitting, 39–40, 43 Lankford ratio
stress corrosion cracking, 48–49 anisotropy measure, 120, 175
interatomic attraction, thermodynamics, 2 limiting drawing ratio as function of, 175
intermetallic phases lanthanum, protective layer, 65
austenitic stainless steel, 82, 203 laser welding, 210
ferritic stainless steels, 116 lattice expansions, 6(F)
stainless steel, 8–9 laundry appliances, 241
International Nickel Company (INCO) process, 196 laves, precipitation kinetics, 7(F)
interstitial elements, 6 laves phase, 9
ionic current, 11–12 leaching, elements from stainless to foods, 234
ionic diffusion, 21–22 lead, stainless steel machinability, 186
iron lean alloys
body- and face-centered cubic transformations, 2 austenitic, 72–78
electrochemical corrosion, 19 compositions of austenitic, 72(T)
ferrite, 3 martensite and austenite, 73–74
ion release from stainless steel grades, 222(F) lime content, electroslag remelting, 158
machinability of stainless steels, 182 limiting current
migration into acetic solution from stainless, aluminum electrode reaction kinetics, 21
or carbon steel, 234(T) increasing mass transfer, 23, 24(F)
penetration rates, 19(T) limiting drawing ratio, 175
Pourbaix diagram, 16(F) line pipe
pseudo-binary-phase diagram for, and sulfur, 41(F) martensitic stainless, 134–135
iron-chromium stainless steel application, 252–253
phase diagram, 3(F) liquefied natural gas (LNG) vessels, 254
phase diagram from Thermocalc, 113(F) localized corrosion, 37–38
Index / 295
lubricants machinability of stainless steels, 183

oxides, 189–190 sulfide formation, 186
suitability in forming stainless steel, 179(T) manganese sulfides
lubrication, 197–198 inclusions, 41
stress risers, 188
x-ray examination, 189(F)
M
Marangoni effect, 207
machinability, material’s, 181(F) marine systems
machining stainless steels corrosion resistance vs. salinity and temperature, 244(F)
austenitic, 185 desalination, 243–245
carbides, 191 heat exchangers, 245–246
carbon, 182–183 materials for desalination, 244–245
chromium, 182 shipping, 245
cleanliness, 184 typical analyses and properties of marine alloys, 245(T)
coatings, 191 martensite
coolants, 191 carbon and nitrogen, 9
copper, 183 composition range, 7
cross-section size, 185 formation, 126–127
duplex, 186 forms, 7–8, 73–74
duplex alloys, 99 lattice expansions, 6(F)
ferritic, 185 lean alloy of, and austenite, 73–74
high-speed tool steels, 190–191 platelets from surface, 126(F)
introduction, 181–182 reversion of, formed by cold work, 75(F)
iron, 182 Schaeffler–Delong constitution diagram, 5(F), 70(F)
lead inclusions, 186 Schaeffler diagram, 202(F)
machinability of stainless steel families, tempering, 7
185–186 varying hardness with carbon content, 128(F)
manganese, 183 martensitic alloys
martensitic, 185 “C” alloys, 149, 151
material’s machinability, 181(F) composition of, precipitation hardening (PH) alloys,
molybdenum, 183 140(T)
nickel, 183 corrosion resistance of, PH alloys, 141–142
niobium, 184 mechanical properties of, PH alloys, 139(T)
nitrogen, 184 microstructures of, PH alloys, 140(F)
oxide inclusions, 188–190 precipitation hardening stainless steels, 139–142
physical and mechanical properties, 182–185 martensitic stainless steels
precipitation hardening, 185–186 annealing, 166
process, 184 applications, 133–135
role of chemistry, 182–184 austenitizing, 166–167
role of inclusions, 186–190 composition, 124(T), 125(T)
selenium inclusions, 186 composition of, 275(T)
setup recommendations for turning wrought stainless compositions of tool and cutlery, 134(T)
steels, 183(T) corrosion rates of stainless oil country tubular goods
structure, 184 (OCTG) alloys, 135(F)
sulfur, 183 corrosion resistance, 123
sulfur inclusions, 186–188 distinction from other alloys, 123, 126
super stainless steels, 186 expanding austenite stability range with nickel, 131(F)
tellurium inclusions, 186 flatware and cutlery, 240–241
titanium, 184 hardness variation with carbon content, 127, 128(F)
tooling and coolants, 190–191 high-temperature use, 133, 134(F)
maintenance hot rolling, 159
balancing service, design and, 215, 217 influence of alloying elements, 130, 131(T)
stainless steel, 220–221 iron-chromium phase diagrams, 130(F)
stainless steel selection expert system, 216(F) machinability, 185
manganese machining setup recommendations, 183(T)
alloying element, 1 martensite formation, 126–127
austenite, 5 OCTG and line pipe, 134–135
inclusions and pitting, 41 passivation, 195
influence on thermodynamic activity of C, N, S and O, petroleum industry applications, 247(T)
157(T) phase structure, 127–128, 130–131
296 / Index
martensitic stainless steels (continued) metal oxides

photomicrographs, 129(F) parabolic rate constants for growth, 59(F)
recommended annealing, austenitizing, and tempering standard Gibbs free energy of, formation vs. tempera-
temperatures, 166(T) ture, 57, 58(F)
sensitization, 47 metals
smallest stainless steel category, 123 oxidation, 57
soaking, 166 with oxide scale, 61(F)
strain energy, 126(F), 127(F) metastable condition, 2
stress relieving, 167–168 metastable pitting, 40
tempering, 167 microbiologically induced corrosion, 55–56
tempering and toughness, 132, 133(F) microcar frame, 231, 232(F)
thermal processing, 131–133, 166–168 microorganisms, food contact, 235
tool and cutlery alloys, 133–134 microstructures, martensitic precipitation-hardening
toughness by austenite grain size and phosphorus, stainless steels, 140(F)
132(F) migration, ionic transport, 21–22
welding characteristics, 206 mineral resin, bacterial retention, 236(F)
welding parameters, 207(T) mischmetal, 41
mass transfer, 23, 24(F) mixed potential theory, 22–23
mass transfer control, 21 mold powder, continuous casting, 158
mass transport, 24–25 molybdenum
material selection, welding, 206–208 alloying element, 1
material structure, 184 austenitic alloys, 72
material variables carbide precipitation, 9
stress corrosion cracking (SCC), 49–50 corrosion resistance, 228
uniform corrosion, 29–31 disulfide, 179(T)
mechanical behavior, ferritic stainless steels, 116–117 ferrite, 4
mechanical properties influence on resistance to stress corrosion cracking,
austenite, 7 50(F)
austenitic precipitation-hardenable (PH) stainless steel, influence on thermodynamic activity of C, N, S and O,
145 157(T)
corrosion resisting cast stainless alloys, 150(T) influence on uniform corrosion, 29–30
deep-drawing stainless steels, 175(T) influencing critical pitting temperature in welded vs.
duplex alloys, 94–98 unwelded austenitic grade, 208(F)
ferrite, 4 machinability of stainless steels, 183
heat-resistant cast stainless alloys, 152(T) oxidation resistance, 226, 228
high-temperature, of austenitic stainless steels, 82–83, protective layer formation, 64–65
84(F) muffler, automotive exhaust systems, 227(T), 228
high-temperature, of “H” alloys, 153(T) multistage flash (MSF), 243
lean austenitic alloys, 74–76
machinability of stainless steels, 182–185
marine alloys, 245(T)
N
martensite, 8 naming system, Alloy Casting Institute (ACI), 147
martensitic PH stainless steels, 139(T) National Association of Corrosion Engineers (NACE)
minimum, of stainless steel engineering alloys, 280(T) alloy listing, 247
semiaustenitic PH stainless steel, 144(T) diagram showing alloy suitability, 250, 251(F)
stainless steels, 10 regulating high-strength alloys, 142
mechanisms restrictions in use recommendations for stainless steels,
pitting corrosion, 39–40 252(T)
precipitation-hardening, 138 natural gas, vessels for liquefied, 254
stress corrosion cracking (SCC), 51–54 Nernst equation
melting production process, 155–157 open circuit potential, 20
metal dusting, oxidation, 67 thermodynamics of electrochemical reactions, 12–14
metal flow directions, 208(F) neutrality
metal ions release, 222(F) biological, in food contact, 235
metallurgy chemical, in food contact, 233–235
“C” alloys, 149, 151–152 New York City’s Chrysler Building, 196, 197(F),
ferrite stainless steels, 113–116 213(F)
ferritic stainless, 226 nickel
“H” alloys, 152–154 alloying element, 1
introduction, 1–2 austenite, 5
Index / 297
austenitic alloys, 72 O
carbide precipitation, 9
corrosion rates for stainless steels and, base alloys, Occupational Safety and Health Administration
80(F) (OSHA), 211
influence on thermodynamic activity of C, N, S and O, Ohm’s law, 22
157(T) oil and grease marks, cleaning methods, 220(T)
influence on uniform corrosion, 30 “oil-canning,” 219
influencing oxidation of iron-chromium alloys, 79, oil country tubular goods (OCTG)
80(F) influence of chromium on corrosion rate of steel,
ion release from stainless steel grades, 222(F) 249(F)
machinability of stainless steels, 183 influence of copper and nickel on corrosion rate of
migration into acetic solution from stainless, aluminum martensitic alloys, 249(F)
or carbon steel, 234(T) martensitic stainless, 134–135
resistance to stress corrosion cracking, 50(F) stainless steels in petroleum industry, 250–252
nickel base alloys, 149(T) oleum, 33
niobium open circuit potential, 13
carbide former, 78 orange peel, 178
creep resistance, 71–72 organic acids
high-temperature martensitic stainless, 133, 134(F) austenitic stainless steels, 89
machinability of stainless steels, 184 corrosion in, 35
replacing titanium, 205 duplex alloys, 102
role in sensitization, 47 isocorrosion curves in, 36(F)
stabilization, 226 overpotentials, 20
stabilization of ferritic stainless steels, 118–119 oxidation
Nippon Steel, oil country tubular goods and line pipe effect of chromium, 57–59, 60(T)
alloys, 135(T) effect of rare earth additions, 65
nitric acid effect of silicon, aluminum, and molybdenum,
austenitic stainless steels, 88 64–65
corrosion behavior of high-silicon alloys in concen- electrochemical nature of, 60–61
trated, 35(F) influence of nickel on, of iron-chromium alloys, 79,
corrosion in, 34 80(F)
duplex alloys, 101 iron-chromium-oxygen phase diagram, 59(F)
isocorrosion curve for, 35(F) kinetics and, rates, 61–63
pickling oxide scale, 193–194 metal dusting, 67
nitrides metal with oxide scale, 61(F)
duplex alloys, 94 oxidation-resisting grades of stainless steel, 60(T)
stainless steel, 9–10 parabolic rate constants for growth of oxides, 59(F)
nitriding paralinear, from evaporation of chromium superoxide,
austenitic stainless steel and, 82 64(F)
surface alteration, 199 quasi-steady-state approximation of moving boundary
nitrogen problem of internal, 66(F)
austenite, 5 reaction at anode, 12
austenitic alloys, 72 schematic predicting thermal stresses, 65(F)
austenitic stainless steels, 71, 86 spalling and cracking of scale, 63–65
delay in carbide precipitation by, 78(F) standard Gibbs free energy of formation of metal oxides
ferrite, 3–4 vs. temperature, 58(F)
influence of alloying elements on thermodynamic activ- temperature dependence of metal dusting of iron,
ity of, 157(T) 67(F)
influence on thermodynamic activity of C, N, S and O, thermodynamics of, 57–60
157(T) transient, 60
influence on uniform corrosion, 29, 30 under less-oxidizing atmospheres, 66–67
interstitial atoms of, in austenite, 6 volatile nature of Cr2O3, 63, 64(F)
machinability of stainless steels, 184 Wagner’s theory, 61–63
solubility in austenite, 77–78 oxidation resistance
stainless steel for line pipe, 253 austenitic stainless steels, 79–81
nondestructive evaluation (NDE), 211 ferritic stainless steels, 109, 121–122
normal hydrogen electrode (NHE) isooxidation curves, 80(F)
half-cell reduction potential vs., 14(T) oxide film
standard hydrogen electrode, 14 coloring stainless steels, 196
North Carolina, exposure of stainless steel, 238(F) parameters for, coloring stainless steel, 196(T)
298 / Index
oxides ferritic stainless steels for, 247(T)

effect calcium on machinability of 303, 188–189, 189(F) line pipe and flow lines, 252–253
elongated, 190(F) liquefied natural gas (LNG) vessels, 254
inclusions, 10 martensitic stainless steels for, 247(T)
inclusions and pitting, 41 molybdenum influence on SCC susceptibility, 248,
machinability of stainless steel, 188–190 250(F)
metal with, scale, 61(F) National Association of Corrosion Engineers (NACE),
pickling to remove, layer, 25, 193–194 247, 248, 250, 252(T)
removal of oxide scale, 193–194 OCTG, 250–252
stabilization of austenitic alloy, 78 platforms, 254
Ugima, in 303 matrix, 189(F) precipitation-hardening stainless steels for, 248(T)
un-deformed, 190(F) presence of wet carbon dioxide, 248
x-ray examination showing Ugima, and manganese sul- refinery equipment, 254–255
fides, 189(F) restrictions in stainless steel use recommended by
oxyfuel gas welding (OFW), 210 NACE, 252(T)
oxygen stainless steels for refinery processes, 254(T)
austenite impurity, 6 stress corrosion cracking (SCC), 248
ferrite impurity, 3 umbilical tubing and risers, 253–254
impurity, 1 pH, corrosion tendency, 15–16
inclusions and pitting, 40, 41–42 phase diagrams
influence of alloying elements on thermodynamic activ- computer models, 2
ity of, 157(T) Fe-Cr-Ni, 92(F)
influence on thermodynamic activity of C, N, S and O, iron-chromium, 3(F), 73(F), 113(F), 114(F), 130(F)
157(T) iron-chromium-oxygen, 58, 59 (F)
influence on uniform corrosion, 29 iron-nickel, 93(F)
steel content, 156 pseudo-binary-, for iron and sulfur, 41 (F)
oxygen gas reduction, 15(F), 28(F) phases
oxygen pressure, 64(F) alloy systems, 1
ferrite, 3–5
free energy, 2
P intermetallic, of stainless steel, 8–9
paint, cleaning method for uncoated stainless, 220(T) structure of martensitic stainless steels, 127–128,
paper-making processes 130–131
bleaching pulp, 266 phosphoric acid
digestion, 265–266 austenitic stainless steels, 88
kraft process, 265 corrosion in, 34–35
process equipment, 266–267 duplex alloys, 101–102
washing and screening, 266 electropolishing solution, 196
partitioning elements, 93–94 isocorrosion curves in, 36(F)
passenger trains, 232 minimum temperatures for wet, with duplex alloys,
passivation 102(F)
effect on polarization diagrams, 23–25 phosphorus
removing surface contamination, 195–196 ferrite impurity, 3
stainless steel, 25 impurity, 1, 156–157
theory, 23 martensitic stainless steels and toughness, 131, 132(F)
transpassive regime, 24 photomicrographs
passive behavior, 27 duplex alloys, 94, 95(F)
passivity, 27 martensitic stainless steels, 127, 129(F)
penetration equation, 19 physical properties
penetration rates, 19(T) major stainless steel engineering alloys, 279(T)
petroleum industry stainless steels, 10
alloy suitability vs. H2S and CO2 partial pressure, 251(F) pickling
austenitic stainless steels for, 249(T) oxide layer removal, 25, 193–194
chromium influence on corrosion rate in environments uniform corrosion, 28
by oil country tubular goods (OCTG), 249(F) pigmented pastes, 179(T)
combating corrosion in applications, 248, 250 Pilling–Bedworth ratio (PBR), 63
copper and nickel influencing corrosion rate of marten- pitting
sitic stainless alloys for OCTG, 249(F) activities and activity coefficients in liquid steels, 40(T)
demand for steel, 247 austenitic stainless steels, 85, 88
duplex stainless steels for, 248(T) corrosion type, 39–40, 258–259, 261
Index / 299
critical, temperatures, 44(F) powder metallurgy, 159

geometry, 39 precipitated phases
inclusions, 40–43 Guinier–Preston zones, 138
influence of sulfur level on, resistance, 42(F) stainless steels, 8(T)
metastable, 40 precipitation
mischmetal, 41 carbides, 9
passive anode polarization curve, 40(F) carbides and nitrides, 76–78
pit initiation, 39–40 possible aluminum/titanium, 138(F)
pseudo-binary-phase diagram for iron and sulfur, 41(F) precipitation-hardening stainless steels
resistance, 43–45 advantage over martensitic, 137
“weakest link” phenomenon, 214 annealed condition, 139
pitting corrosion austenitic, 144–146
CPT (critical pitting temperature) vs. NaCl concentra- austenitic alloys, 170
tion, 103, 105(F) cast PH alloys, 151
CPT vs. pH, 103, 105(F) cold work influence on aging of A–286, 145, 146(F)
CPT vs. pitting resistance equivalent number (PREN), composition, 276(T)
103, 104(F) composition of austenitic, 145(T)
duplex alloys, 102–103 composition of martensitic, 140(T)
varying pitting potential with temperature, 103, 104(F) compositions of semiaustenitic, 143(T)
pitting resistance equivalent number (PREN) corrosion resistance of martensitic, 141–142
austenitic alloys, 43, 78, 85 corrosion resistance of semiaustenitic, 144
corrosion, 37, 258 development, 137–138
critical pitting temperature vs., 85(F) influence of alloying elements, 141(T)
duplex alloys, 43, 92, 102 machinability, 185–186
ferritic alloys, 43 machining setup recommendations, 183(T)
influence on duplex alloy fatigue strength, 98, 99(F) martensitic, 139–142
pitting corrosion, 43–45, 214 martensitic grades, 168–170
ranking stainless steels by PREN, 214(T) mechanical properties of martensitic PH alloys, 139(T)
umbilical tubing and risers, 253 mechanical properties of semiaustenitic, 144(T)
Pittsburgh Convention Center, 219(F) mechanism of PH, 138
platforms, stainless steel, 254 microstructures, 140(F)
polarization passivation, 195
anode, 23 petroleum industry applications, 248(T)
cathode, 23 phases in stainless steel, 10
influence on uniform corrosion, 29, 30(F) possible aluminum/titanium precipitates, 138(F)
overpotentials, 20 presence of δ-ferrite and γ-austenite, 138–139
passivating alloys, 39(F) processing routes for S15700, 142, 143(F)
passive anode, curve, 40(F) properties of A-286 vs. test temperature, 145(F)
stainless steel in chloride-containing solution, 40 recommended annealing and stress-relieving tempera-
polarization diagrams tures for martensitic grades of, 169(T)
effect of cathode polarization, 24(F) semiaustenitic, 142–144
effect of mass transport, 24–25, 25(F) specialized family, 137
mixed potential theory and, 22–23 stress corrosion cracking (SCC), 141, 142
passivation, 23–25 thermal processing, 168–170
schematic, 22(F) welding characteristics, 206
schematic of passive anode polarization curve, 23, 24(F) welding parameters, 207(T)
polishing precipitation kinetics, 7(F)
grit sizes for target surface roughness, 197(T) prevention, crevice corrosion, 45–46
surface finishing, 197–198 production processes
polycarbonate, bacterial retention, 236(F) basic oxygen furnace (BOF), 156
polythionate, 50 casting, 158–159
polythionic acid, 255 defects, 160
porosity, 52 electroslag remelting (ESR), 157–158
Porsche, auto components, 230 hot rolling, 159–160
postweld stress relief, 205 hot Steckel mills, 159
potassium hydroxide, 88–89 hot strip tandem mills, 159
Pourbaix diagrams impurities, 156–157
chromium, 17(F) influence of alloying elements on thermodynamics,
construction of, 16–17 157(T)
iron, 16(F) melting and refining, 155–157
300 / Index
production processes (continued) S

remelting, 157–158
semiaustenitic precipitation-hardenable stainless steel, Saab, auto components, 230
142, 143(F) safety, welding, 211
stainless steel, 155 salinity, corrosion resistance vs., 244(F)
thermodynamics, 156 salt exposure, stainless steel selection expert system,
vacuum arc remelting (VAR), 157–158 216(F)
vacuum induction melting (VIM), 157 sanitation, cleaning stainless steels, 195
vacuum oxygen decarburization (VOC), 156 scale
propagation metal with oxide, 61(F)
crack, rates of metals vs. current density, 52(F) spalling and cracking of, 63–65
stress corrosion cracking (SCC), 49 Schaeffler–Delong constitution diagram, 5(F), 70(F)
pulp-and-paper industry Schaeffler diagram, 202–203
duplex stainless steels, 265, 266–267 seawater. See also marine systems
kraft process, 36, 265 desalination, 243–245
paper-making processes, 265–267 secondary austenite, 7
pulsed arc transfer, 210 secondary phases, 82(T)
selenium, 186
semiaustenitic precipitation-hardenable stainless steels
Q austenite-stabilizing elements, 142
quality control, Thermo-Calc software, 283 compositions, 143(T)
corrosion resistance, 144
mechanical properties, 144(T)
R processing by T route, 142, 143(F)
rain, cleansing action of, 220 sensitization
rainwater, average chloride concentration, 217(F) austenitic, 46–47
rare earth metals duplex steels, 47
inclusions and pitting, 41 effect of alloying, 47–48
protective layer formation, 65 ferritic, 47
reduction, 12 ferritic stainless steels, 115
reduction potential heat treatment vs. time, 46(F)
iron and hydrogen ion reductions vs. pH, 15(F), intergranular corrosion, 9
28(F) knife-line attack, 48, 202
iron and oxygen gas reductions vs. pH, 15(F), martensitic steels, 47
28(F) schematic of, due to chromium-rich precipitates, 46(F)
reference electrode, 13–14 welding, 48
refinery equipment, 254–255 service, design, and maintenance, 215, 217
refining production process, 155–157 shielded metal arc welding (SMAW)
repair, 221 joint design, 209(F)
residential applications process, 210
cookware, 237 shielding gas
domestic goods, 233 welding austenitic stainless steel, 203–204
flatware and cutlery, 240–241 welding parameters for various stainless steels, 207(T)
heating, 241–242 shipping, 245
kitchen appliances, 237–240 short-circuiting transfer, 210
laundry appliances, 241 sigma
water heaters, 241, 242 intermetallic phase, 8–9
resistance precipitation kinetics, 7(F)
Ohm’s law, 22 silicon
pitting, 43–45 alloying element, 1
resistance welding, 210–211 content in cast alloys, 147
resistivity, 22(T) corrosion of high- austenitic steels in nitric acid, 34, 35(F)
reverse osmosis, 243–244 inclusions and pitting, 41
ridging, ferritics, 178 influence on thermodynamic activity of C, N, S and O,
rolled finishes 157(T)
applications, 198–199 oxidation resistance, 71, 79, 226, 228
benefits, 198 protective layer formation, 64–65
roping, ferritics, 178 single-environment system. See also chemical and
rust staining, cleaning method for uncoated stainless, process industry
220(T) aggressive chemical species, 258
Index / 301
slabs, 158 corrosion table for, in sulfuric acid plus copper sulfate,
slip dissolution, 51 31(F)
soaking deep drawing, 173, 174–179
austenitic stainless steels, 161–162 defects in, hot–rolled bands, 160
duplex stainless steels, 170 flat, rolled, 173–179
ferritic stainless steels, 165 hot rolling, 159–160
martensitic stainless steels, 166 inclusions, 10, 186–190
soap-fat pastes, 179(T) isocorrosion curves for, in sulfuric acid plus copper sul-
sodium chloride/carbon dioxide environment, 30 fate, 32(F)
sodium hydroxide lubricants for forming, 179 (T)
austenitic stainless steels, 88–89 machinability, 185–186
corrosion in, 35–36 machining setup recommendations, 183 (T)
corrosion rates of duplex alloys, 101(F) melting and refining, 155–157
corrosion rates of duplex alloys with contaminated envi- minimum mechanical properties of, engineering alloys,
ronment, 101(F) 280 (T)
duplex alloys, 101 nitrides, 9–10
sodium hypochlorite oxidation-resisting grades, 60(T)
cleaning stainless steels, 195 passivation, 25
disinfecting stainless, 236–237 penetration rates, 19(T)
software package, Thermo-Calc, 2, 281–283 physical properties of major, engineering alloys, 279 (T)
soldering, 211 precipitated phases, 8 (T)
solution treatment, precipitation-hardening stainless precipitation-hardening process, 10
steels, 168–170 precipitation kinetics in 316, 7 (F)
specialization, precipitation-hardening stainless steels, properties, 10
137 ranking by pitting resistance equivalent number (PREN),
Sphaerotilus, 55 214(T)
spots, cleaning method for uncoated stainless, 220(T) ranking common, by PREN, 214(T)
spray transfer, 210 refinery processes, 254(T)
stability remelting, 157–158
expanding austenite, with nickel, 131(F) resilience and toughness of carbon steel vs. for
lean alloy of martensite and austenite, 73–74 automotive components, 229(T)
stabilization Schaeffler–Delong constitution diagram, 5 (F)
ferrite, 4 selection expert system, 216(F)
ferritic stainless steel, 109, 115, 118–120 tensile properties of carbon steel vs. for automotive
ferritic steels for exhaust systems, 226 components, 229(T)
lean austenitic alloys, 78 thermodynamics, 2
stacking fault, 73–74 welding parameters, 207(T)
stainless long products stains, cleaning method for uncoated stainless, 220(T)
cold heading, 179–180 standard Gibbs free energy, 57, 58(F)
hot forming, 180 standard hydrogen electrode, 14
stainless steel alloys Steckel mills, hot, 159
austenite, 5–7 Steel Founder’s Society of America, 147
ferrite, 3–5 strain energy, martensitic stainless steels, 126(F), 127(F)
stainless steels. See also casting alloys strain rate, deep drawing, 177
activities and activity coefficients of elements in, 40(T) stress corrosion cracking (SCC)
bacterial retention by material and cleaning time, advantages of duplex alloys, 105–106
236(F) austenitic stainless steels, 87–88
carbides, 9 corrosion form, 258
casting, 158–159 crack initiation, 48–49
casting alloys, 147, 149 crack propagation, 49
categories for oxidation resistance, 59–60 crack propagation rates of metals vs. current density,
classifications by sulfur content, 187–188 52(F)
cleaning methods for uncoated, 220(T) debating mechanisms, 105
composition of austenitic, 270 (T), 271 (T) dilation of austenite due to hydrogen in solution, 53, 54(F)
composition of duplex, 276(T) duplex alloys, 91, 104–106
composition of ferrite, 273(T), 274(T) environmental variables, 50–51
composition of martensitic, 275(T) ferritic stainless steels, 121
composition of precipitation-hardenable (PH), 276(T) influence of molybdenum on resistance, 50(F)
concrete reinforcing bar, 222 martensitic precipitation-hardening (PH) stainless steels,
corrosion rates of, vs. carbon steel, 135(F) 141, 142
302 / Index
stress corrosion cracking (continued) sulfuric acid

material variables, 49–50 austenitic stainless steels, 88
mechanisms, 51–54 corrosion in, 31–33
petroleum industry, 248 corrosion table, 259(T), 260(T), 261(T)
resistance, 257, 262(F) corrosion table for fuming, 261(T)
stress-strain curve for single crystals of austenitic steel duplex alloys, 100
with and without hydrogen, 53(F) electropolishing solution, 196
susceptibility of martensitic stainless steels, 123 influence of alloying element on corrosion rate in con-
susceptibility to, with oxygen and chloride content for taminated, 33(F)
304 stainless, 51(F) isocorrosion, 88(F)
theory, 262 isocorrosion chart for, 258(F)
threshold stress for, for various alloys, 88(F) isocorrosion curves for various alloys in, 33(F)
varying resistance to, with nickel content, 50(F) isocorrosion curves for various alloys in, with chlorides,
water heaters, 242 34(F)
zones of susceptibility, 48(F) isocorrosion curves of duplex grades, 100(F)
stress relief annealing (SRA), 48 isocorrosion rates for various stainless steels, 32(F)
stress relieving oleum, 33
austenitic stainless steels, 164 pickling oxide scale, 193–194
ferritic stainless steel, 166 sulfuric acid plus copper sulfate
martensitic stainless steels, 167–168 corrosion table for stainless steels and titanium in, 31(F)
stress risers, 188 isocorrosion curves for stainless steel and titanium in,
stress sorption, stress corrosion cracking, 51, 53 32(F)
stretch forming, 174 sulfurized or sulfochlorinated oils, 179(T)
strip casters, 158–159 superaustenitic stainless steels, 164
strip tandem mills, 159 superferritics, 113
strong bases super stainless steels, 186
austenitic stainless steels, 88–89 surface finishing
corrosion in, 35–36 aesthetics, 217–219
structure aesthetic surface finishes, 196–199
duplex alloys, 92–94 austenitic stainless steels, 89
machinability of stainless steels, 184 bright annealing, 198
phase, of martensitic stainless steels, 127–128, 130–131 brightening, 196
submerged arc welding (SAW) cleaning, 194–196
joint design, 209(F) coloring, 196
process, 210 and corrosion resistance, 215
submerged entry nozzle, 158 deep drawing, 178–179
sulfate process, pulp-and-paper, 36 function of surface treatments, 193–196
sulfides introduction, 193
inclusions, 10 parameters for oxide film coloring of stainless, 196(T)
inclusions in stainless steel, 186–187 passivation, 195–196
size and shape and machinability, 187 pickling, 193–194
stabilization of austenitic alloy, 78 polished finishes, 197–198
sulfite process, 265 recommended cleaning methods, 195(T)
sulfur removal of oxide scale, 193–194
austenite impurity, 6 rolled finishes, 198–199
comparing machinability, 190(F) rolled-on finishes, 218 (F)
effect on stainless machinability, 187, 188(F) surface alteration, 199
ferrite impurity, 3 surface roughness, 215(F)
impurity, 1, 156 surface treatments
inclusions and pitting, 40, 41 brightening, 196
influence of alloying elements on thermodynamic activ- cleaning, 194–196
ity of, 157(T) coloring, 196
influence on thermodynamic activity of C, N, S and O, passivation, 195–196
157(T) removal of oxide scale, 193–194
machinability of stainless steels, 183
metal flow directions in weld pool with and without,
T
208(F)
pitting resistance of unannealed welds, 42(F) Tafel slope, 23(F)
pseudo-binary-phase diagram for iron and, 41(F) tailpipe, automotive exhaust systems, 227(T), 228
stainless steel machinability, 186–188 tellurium, 186
Index / 303
temperature thermodynamics
chromium-oxygen system volatility, 64(F) argon oxygen decarburization (AOD), 156
corrosion resistance vs., 244(F) influence of alloying elements on, activity of C, N, S,
critical crevice corrosion, with alloy content, 45(F) and O, 157(T)
critical pitting, (CPT), 43, 44(F) oxidation, 57–60
impact strength variation with, for stainless steels, stainless steel, 2
75(F) thiosulfate, 50
partitioning ratio varying with, 93(F) time-temperature-transformation (TTT) diagram
standard Gibbs free energy of metal oxide formation vs., high-alloy stainless steel, 94, 96(F)
58(F) unstabilized 430-type alloy, 115(F)
variation of pitting potential with, for duplex alloys, titanium
104(F) carbide and nitride formation, 115
tempering carbide former, 78
influencing martensitic stainless hardness, 132, corrosion table for, in sulfuric acid plus copper sulfate,
133(F) 31(F)
martensite, 7 deoxidizer in chromium–iron alloys, 156
martensitic stainless steels, 167 ferritic alloy stabilization, 4, 205
tensile properties inclusions and pitting, 41
austenitic precipitation-hardenable stainless steel, influence on thermodynamic activity of C, N, S and O,
145(F) 157(T)
austenitic stainless steels, 75 isocorrosion curves for, in sulfuric acid plus copper sul-
tensile strength equation, 74 fate, 32(F)
texture isocorrosion curves in phosphoric acid, 36(F)
deep drawing, 174 machinability of stainless steels, 184
ferritic stainless steels, 120–121 possible aluminum/titanium precipitates, 138(F)
thermal conductivity role in sensitization, 47
duplex alloys, 205 stabilization of ferritic stainless steels,
ferrite, 4–5 118–119
thermal cutting, 211 stabilization of ferritic steels, 226
thermal expansion titanium-aluminum-nitride (TiAlN), 191
austenitic stainless steels, 202 titanium carbonitride (TiCN), 191
duplex alloys, 205 titanium nitride (TiN), 191
ferrite, 4–5 tooling
thermal processing carbides, 191
annealing, 162–164, 165–166, 170–171 coatings, 191
austenitic stainless steels, 161–164 coolants, 191
austenitizing, 166–167 costs in deep drawing, 176
duplex stainless steels, 170–171 high-speed tool steels, 190–191
ferritic stainless steels, 165–166 lubricants, 189, 190
martensitic stainless steels, 131–133, 166–168 materials in deep drawing, 178–179
precipitation-hardening stainless steels, 168–170 tools, martensitic stainless steels, 133–134
soaking, 161–162, 165, 166, 170 toughness
stress relieving, 164, 166, 167–168 austenitic stainless steels, 75–76
tempering, 167 duplex alloys, 97–98, 204
thermal stresses, predicting, 64, 65(F) ferritic stainless steels, 117, 118(F), 118–119
Thermo-Calc high-temperature austenitic alloys, 83(F)
alloy design, 93, 282 martensitic stainless steels, 131, 132(F)
applications, 282–283 trains, 232
casting, 283 transient oxidation, 60
free demonstration version, 283 transpassive dissolution, 27
heat treatment, 283 transpassive regime, 24
iron-chromium phase diagram, 113(F) transportation. See automotive and transportation
map, 282 trucks, 231
phase determination program, 2 tubular goods, 134–135
quality control, 283 tungsten
Scheil–Gulliver model, 282 carbides for flatware, 240
single-point equilibria, 282 influence on thermodynamic activity of C, N, S and O,
software package, 281–282 157(T)
step, 282 tungsten inert gas (TIG), 208–210
welding and joining, 283 tuyeres, oxygen injection, 156
304 / Index
U ferritic stainless steels, 205–206

flux cored wire (FCW), 210
Ugima oxide gas metal arc welding (GMAW), 210
coating cutting tool and lubricant, 189–190 gas tungsten arc welding (GTAW), 208–210
comparing 304L chips with and without, 190(F) high-frequency induction, 211
machinability by sulfur levels with and without, 190(F) joint design, 208, 209(F)
x-ray showing, 189(F) laser, 210
umbilical tubing and risers, 253–254 martenistic stainless steels, 206
uniform corrosion. See also corrosion types material selection and performance, 206–208
environmental variables influencing, 28–29 metal flow directions in weld pool, 208(F)
material variables, 29–31 new developments, 212
stainless steel, 27–28 nondestructive evaluation (NDE), 211
United States, chloride concentration in rainwater, oxyfuel gas welding (OFW), 210
217(F) parameters for various stainless steels, 207(T)
unmixed zone, 208 practices, 211–212
precipitation-hardening (PH) stainless steels, 206
V processes, 208–211
recent developments, 211–212
vacuum arc remelting (VAR), 157–158 resistance, 210–211
vacuum induction melting (VIM), 157 safety, 211
vacuum oxygen decarburization (VOD) Schaeffler diagram, 202(F)
cleanliness, 184 sensitization, 48
refining process, 156 shielded metal arc welding (SMAW), 210
vanadium soldering and brazing, 211
carbides for flatware, 240 submerged arc welding (SAW), 210
high-temperature martensitic stainless, 133, 134(F) thermal cutting, 211
Volvo, auto components, 230 Thermo-Calc software, 283
tungsten inert gas (TIG), 208–210
Welding Research Council’s 1992 constitution diagram,
W
203(F)
Wagner’s theory, 61–63 weld shielding gas composition and crevice corrosion
washer tubs and drums, 241 resistance, 204(F)
water heaters, 241, 242 Welding Research Council, constitution diagram,
waterline corrosion, 38 203(F)
water marking, cleaning method for uncoated stainless, welds
220(T) influence of sulfur on pitting resistance of unannealed,
water vapor, 81–82 42(F)
wax-base pastes, 179(T) long-term annealing, 43
wax or soap plus borax, 179(T) wet carbon dioxide, 248
weather pattern, stainless steel selection expert system,
216(F)
weldability, 253
Y
welding yield strength
austenitic stainless steels, 201–204 austenitic precipitation-hardenable stainless steel, 145(F)
cast stainless alloys, 154 equation, 74
characteristics of stainless steels, 201–206 high-temperature austenitic alloys, 83, 84(F)
duplex stainless steels, 204–205 yttrium, 65



Fax 1-440-338-4634
Mail
www.ameritech.co.uk
Neutrino Inc.
Telephone: 81 (0) 424 84 5550
ASM International.

Stainless Steels For Design Engineers

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Stainless Steels For Design Engineers

Uploaded by

Copyright:

Available Formats

Copyright © 2008 ASM International®. All rights reserved.

Stainless Steels for Design Engineers (#05231G) www.asminternational.org

First printing, December 2008

CORROSION AND OXIDATION

Chapter 3 Corrosion Kinetics..........................................................................................19

Chapter 4 Corrosion Types..............................................................................................27

STAINLESS STEEL ALLOYS

Chapter 7 Duplex Stainless Steels...................................................................................91

Chapter 8 Ferritic Stainless Steels .................................................................................109

Chapter 9 Martensitic Stainless Steels ..........................................................................123

Chapter 10 Precipitation-Hardening Stainless Steels ......................................................137

Chapter 12 Melting, Casting, and Hot Processing...........................................................155

Chapter 13 Thermal Processing ......................................................................................161

Chapter 15 Machining ....................................................................................................181

Chapter 16 Surface Finishing ..........................................................................................193

Chapter 19 Automotive and Transportation Applications................................................225

Chapter 20 Commercial and Residential Applications ....................................................233

Chapter 21 Marine Systems Applications........................................................................243

Chapter 22 Petroleum Industry Applications ..................................................................247

Chapter 23 Chemical and Process Industry Applications ................................................257

Chapter 24 Pulp-and-Paper Industry Applications ..........................................................265

Appendix 2 Physical and Mechanical Properties of Select Alloys....................................279

Appendix 3 Introduction to Thermo-Calc and Instructions for

ASM International, Materials Park, Ohio, USA

To order products from ASM International:

Telephone 1-800-336-5152 (US) or 1-440-338-5151 (Outside US)

ASM International, Materials Park, Ohio, USA

To order products from ASM International:

Telephone 1-800-336-5152 (US) or 1-440-338-5151 (Outside US)

Summary phases coexisting simultaneously. Stainless

Phases and oxygen, have no influence on which phase

and is very difficult to identify microscopically. chromium or molybdenum, causing localized

Summary Electrochemical Reactions

THIS CHAPTER INTRODUCES THE funda- In electrochemical reactions, charge is trans-

often are described as half cells, for example,

If the cell in Fig. 1(a) was allowed to proceed

Here, i * t is the charge passed (in coulombs);

Fig. 1 Schematic illustration of (a) a differential aeration cell

the anode to the cathode.

ence is the reversible potential difference E;

Table 1 Standard half-cell reduction potentials

Fig. 3 Schematic of (a) galvanic cell and (b) electrolysis cell

Corrosion Tendency ing to Eq 13, which, assuming a Fe2+ activity of

cient for OH−: shaded in gray in Fig. 4(a). Hydrogen reduction

Figure 4(b) shows the condition in which corro- Fe(OH)2 = HFeO 2− +H + ,

rium constants): (Eq 20b)

The following electrochemical reactions will

Summary the penetration due to dissolution of element i

Consider the differential aeration cell dis-

a net corrosion current i will ﬂow, resulting in

EFe 2+ /Fe = EFe0 2+ /Fe − ln ⎜ ⎟

When a cell is not under open circuit (i.e., a

2 Fe ( s ) → 2 Fe 2 + + 4 e− (Eq 4) where the jo,i terms are the exchange current

and the respective equations describing the

The resistivities of some test solutions are j0 ,c

Viewing the electrochemical cell as an elec-

Table 2 Test solution resistivity

the potential drop across the electrolyte needs to Passivation

polarization level, the passive layer breaks j0 ,c

is the actual passivation; the iron removal is

Summary plex in their behavior because the influence of

Fig. 3 Influence of alloying element on corrosion rate as ex-