You are on page 1of 13

Seismic Design of Steel Railroad Truss Bridge

based on AREMA Design Criteria


Larry Wu, Bill Addlespurger & Dan Adams

ABSTRACT
Railroad bridges are functionally and behaviorally different from highway bridges.
Historically, railroad bridges have performed well in seismic events with little or no damage. The
American Railway Engineering and Maintenance-of-Way Association (AREMA) provides
seismic design guidelines for railroad bridges. This paper presents the seismic design of a twospan steel truss railroad bridge located in the Inland Empire of Southern California. First, a
detailed bridge type selection is carried out, and a two-span steel truss bridge with span length
200 ft (total 400 ft) and 20 degree skew is selected among other bridge types. Next, a three level
seismic hazard is evaluated based on the three performance criteria limit states. After getting
seismic input from the railroad, CSIBridge is used to perform 3-D dynamic analyses. Based on
the dynamic analysis results, the superstructure steel truss members and their connections are
designed by Allowable Stress Design Criteria, while the substructure with 10 ft x 8 ft drop cap, 6
ft diameter columns and 8 ft diameter (Type II) CIDH shafts are designed by Load Factor Design
Criteria. The Capacity Design Principle is applied to bent cap and CIDH shafts. The final design
was reviewed and approved by the railroad. Bridge construction is expected to begin in the
spring of 2013.

Larry Wu, Senior Bridge Engineer, Caltrans, 1081 30th Street, Sacramento, CA 95816
Bill Addlespurger, Bridge Engineer, Caltrans, 1081 30th Street, Sacramento, CA 95816
Dan Adams, Senior Bridge Engineer, Caltrans, 1081 30th Street, Sacramento, CA 95816
1

INTRODUCTION
California Department of Transportation (Caltrans), in cooperation with San Bernardino
Associated Governments (SANBAG) and Riverside County Transportation Commission
(RCTC), is proposing to widen the Inter-State Route 215 (I-215) freeway for approximately 7.5
miles to include high-occupancy vehicle (HOV) lanes in the median. The proposed
improvements would complete a regional HOV network to provide continuous HOV lanes from
San Bernardino to Los Angeles along the I-215/SR-91 corridor.
Under the scope of work, several highway bridges on the main line will be widened, and
a railroad bridge Highgrove Underpass, owned by Burlington Northern Santa Fe Railroad
(BNSF), needs to be replaced with longer spans due to the proposed roadway width under the
bridge. The existing structure is a four-span steel through girder bridge carrying two railroad
tracks. Currently, the roadway width under the bridge can only accommodate three traffic lanes
in each direction. In order to accommodate the HOV lanes near the median, bridge piers would
have to be relocated between traffic lanes. Also, the existing 2 column bent is inadequate for the
increased design loads associated with longer spans, thus it was determined the bridge has to be
removed and replaced by a longer span bridge. Since BNSF is the bridge owner, the design and
construction of the replacement bridge needs to follow the American Railway Engineering and
Maintenance-of-Way Association (AREMA) Manual of Railway Engineering guidelines.
The focus of this paper is on the seismic design of Highgrove Underpass (truss). The
seismic analyses for railroad bridges are different from highway bridges. The seismic responses
of railroad bridges are functionally and behaviorally different from highway bridges.
Historically, railroad bridges have performed well in seismic events with little or no damage due
to some unique characteristics. As example, railroad bridges are traversed by rail and tie
assemblies that help to restrain and dampen translational movement during earthquakes.
Compared with the movement towards displacement based design methods for seismic design of
highway bridges, elastic over-strength design philosophies and strategies are presented in
AREMA for the seismic design of railroad bridges.
The seismic design of Highgrove Underpass includes four parts. First Type Selection: a
detailed bridge type selection is carried out, and a two-span steel truss bridge with span lengths
of 200 ft (total 400 ft) and 20 degree skew is selected among other bridge types. Second
Seismic Hazard: a three event level seismic hazard is evaluated based on multiple performance
criteria specified in AREMA Manual. Third Force Demand: the bridge seismic response
demands are computed by 3-D dynamic analyses through program CSIBridge representing a
response spectrum analysis. Fourth Member Capacity: the capacities of the superstructure
(steel) truss members and their connections are estimated by Allowable Stress Design Criteria,
while the capacities of substructure (concrete) with 10 ft x 8 ft drop cap, 6 ft diameter columns
and 8 ft diameter CIDH shafts (Type II) are calculated by a Load Factor Design Criteria.
Capacity protection Design Principles are applied to the dropped bent cap and the deep
foundation CIDH shafts.
BRIDGE TYPE SELECTION
As shown in Figure 1, the existing bridge is a four-span through plate girder bridge
carrying two sets of railroad tracks. Three south-bound and three north-bound highway lanes
provide traveled way beneath Span 2 and Span 3, respectively. The fixed railroad alignment
2

traverses the I-215 alignment at a 55 o skew. The railroad bridge span lengths are: 47.75 ft
103.67 ft 122.00 ft 54.75 ft. The out-to-out width of the bridge steel flange plates is 36.0 ft.
Within the current project, only one HOV lane will be added in each direction. However,
the ultimate planned traffic layout will be six lanes, including the one HOV lane in each
direction. The task for bridge design is to accommodate the ultimate planned traffic layout.
3 lanes + shoulders

Abut 1

Bent 2

3 lanes + shoulders

Bent 3

2 RR track lines

Bent 4

Abut 5

Figure 1. Existing Four-Span Through Girder Bridge: Skew 55o


An initial choice of bridge type, as shown in Figure 2, is a four span reconstruction of the
existing bridge: four-span through girder bridge with 55 o skew and carrying two RR tracks. The
widened four south-bound traffic lanes would pass under Span 2, and the future two south-bound
lanes would pass under Span 1. The layout for six north-bound lanes would be similar under
Span 3 and Span 4. The maximum span length required would be about 130 ft. This type of
bridge is relatively straight-forward in design and familiar to construction. The drawback is that
Bent 2 and Bent 4 now must be located between the current four and future two traffic lanes.
From the travelers safety point of view, this proposal was rejected by Traffic Planning.
6 lanes + shoulders
2 RR track lines
6 lanes + shoulders

Abut 1

Bent 2

Bent 3

Bent 4

Abut 5

Figure 2. Four-Span Through Girder Bridge: Skew 55o


6 lanes + shoulders

6 lanes + shoulders

Abut 1

2 RR track lines

Bent 2

Abut 3

Figure 3. Two-Span Through Truss Bridge: Skew 55o


Naturally, a second choice is to remove Bent 2 and Bent 4, and build a two-span bridge
with a span length about 220 ft, so that total six south-bound lanes will travel beneath Span 1 and
six north-bound lanes will travel beneath Span 2. According to the AREMA Manual, the
recommended maximum span length for a steel through plate girder bridge is 150 ft. If the span
3

length is expected to exceed 150 ft, the preferred structure type would be a through truss. Since
the proposed two-span bridge has a span length about 220 ft, a steel through truss bridge
becomes the best choice, as shown in Figure 3. This alternative gets no objections from Traffic
Planning. However, when we sent this General Plan to many steel structure experts national-wide
to review and comment, their feedback indicated that a steel through truss with high skew (here
55o) and carrying two RR tracks is not a good structure type for resisting the live load fatigue
vibration and high seismic forces. This alternative was dropped during type selection.
In order to completely eliminate skew effects on this two span bridge, the third alternative
is to layout bridge supports perpendicular to railroad track alignment. Since the bridge width is
about 40 ft wide, after turning the single supporting bent in the median perpendicular to RR
alignment, the required median width along the roadway under the bridge would be over 50 ft.
Obviously, a significant roadway layout change has to be made in order to accommodate a
median over 50 ft in width. This alternative, which was rejected by Traffic Planning, didnt go
through the type selection.
The fourth alternative, as shown in Figure 4, is a single span bridge without skew at
abutments and without supports interfering with highway traffic under the bridge. The bridge
carries two railroad tracks and spans over the entire freeway in both directions with a span length
about 420 ft. During a preliminary design, it was found that the bridge height requirements
would reach about 45 ft due to the sway frame requirements. The overall dimensions of this
bridge are up to 420 ft x 40 ft x 45 ft. With a very limited space at the bridge site, assembling,
lifting and positioning this big truss span would be a challenging task during the construction.
Also, this big structure, doesnt match the overall environment aesthetically. This alternative,
again, failed to pass the type selection.
6 lanes + shoulders

6 lanes + shoulders

2 RR track lines

Abut 2

Abut 1
Figure 4. Single-Span Through Truss Bridge: Skew 0
6 lanes + shoulders

Abut 1

6 lanes + shoulders

1 track 1 bridge

Bent 2

Abut 3

Figure 5. Two-Span Through Truss Bridge (Two Bridges): Skew 20o


After summarizing various advantages and disadvantages from previous alternatives, the
final alternative, as shown in Figure 5, was proposed. This alternative includes two identical
4

bridges, each carrying one railroad track. The bridges consist of two twin spans with each span
length of 200 ft, for a total bridge length 406 ft. The bridge width is only 21 ft, and the bridge
height is 30 ft. Based on the narrower single track bridge width and a compromised median
width at Bent 2, the skew is reduced to 20o. The advantages of this final alternative include a
reasonable span length for easy construction, and a reasonable skew for reducing live load
fatigue and seismic effects.
SEISMIC HAZARD EVALUATION
The performance based seismic design criteria are given in the AREMA Manual. A threelevel ground motion and performance criteria are specified as in Table 1.
Table 1

Ground Motion Level and Performance Criteria

Ground
Motion Level

Performance Criteria (Limit States)

Serviceability

Ultimate

Survivability

1.
2.
3.
1.
2.
3.
1.
2.
3.

Critical structure members shall remain in elastic


Trains may continue at restricted speed over bridges
No structure damage is allowed
Strength and stability of critical members shall not be exceeded
Trains must stop until bridge inspections are made
Damage should be easily detected and accessible for repair
No structure collapse
Trains must stop until bridge inspections are made
Extensive damage to structure is allowed

The ground motion levels reflect the seismic hazard at the bridge site. The different
ground motion levels are associated with different earthquake return periods. Depending on the
importance of the bridge, the specified return periods are assigned to different ground levels. In
AREMA Manual, the importance of a bridge is determined by three measures: Immediate Safety,
Immediate Value and Replacement Value. Immediate Safety is a measure of the magnitude of
earthquake a structure should be able to survive without any interruption of service. Immediate
Value is a measure of the magnitude of earthquake a structure should be able to survive with an
interruption of service but with the ability to return to service after minor repairs. Replacement
Value is a measure of the magnitude of earthquake a structure should be able to survive. For this
bridge, based on the layout and the information provided by bridge Owner, the assigned various
factors are listed in Table 2. Then the values of three components are calculated according to the
combination formula specified in AREMA Manual. It can be seen from this Table that all three
factors reach to the maximum of 4.0, which means this bridge should be categorized as
Important Bridge.
Table 2
ICF Components
Immediate Safety
(ISF)
Immediate Value
(IVF)
Replacement Value
(RVF)

Components of Important Classification Factor (ICF)


Assigned Factors by Bridge Owner

1.
2.
3.
1.
2.
1.
2.
3.

Occupancy Factor
Hazard Material Factor
Community Life Line Factor
Railroad Utilization Factor
Detour Availability Factor
Span Length Factor
Bridge Length Factor
Bridge Height Factor

(OCF)
(HMF)
(CLF)
(RUF)
(DAF)
(SLF)
(BLF)
(BHF)
5

Value of ICF Components


4.0
2.0
4.0
4.0
1.0
3.0
1.5
1.0

ISF = (OCF + HMF + CLF) < 4.0


ISF = 4.0
IVF = (RVF x DAF)
IVF = 4.0
RVF = (SLF x BLF x BHF) < 4.0
RVF = 4.0

The earthquake return periods are determined based on the importance of the bridge. In
AREMA Manual, the Important Classification Factors are computed by the weighted
combination of three components. Table 3 lists the combination rules and the resulted Important
Classification Factors and the return periods corresponding to three ground motion levels. The
base acceleration coefficient maps with return periods 100 years, 475 years and 2400 years are
provided in AREMA Manual. For other return periods R (years), the formula to calculate the
base acceleration coefficient are

A475 ( R / 475) n

AR

en

n [ln( A100 / A475 )] /(1.558)

R 475

n ln( A475 ) [ln( A2400 ) ln( A475 )][0.606 ln( R) 3.73]

475 R 2400

(1)
Table 3

Ground Motion Level and Earthquake Return Period

Ground
Motion Level

Performance
Criteria

Important Classification Factor


(IC)

Return Period R
(Years)

Serviceability

ICF1 = 0.8ISF + 0.2IVF + 0.0RVF = 4.0

50 + ICF1(10050)/4 = 100

Ultimate

ICF2 = 0.1ISF + 0.8IVF + 0.1RVF = 4.0

200 + ICF2(500200)/4 = 500

Survivability

ICF3 = 0.0ISF + 0.2IVF + 0.8RVF = 4.0

1000 + ICF3(24001000)/4 = 2400

The soil conditions at the bridge site have influence on the bridge seismic responses.
Based on foundation soil characteristic, AREMA Manual specifies four Soil Types and
corresponding Site Coefficients (S). For this bridge site, according to a prepared Foundation
Report, the soil conditions satisfy Type 2 soil.
Another factor specified in AREMA Manual is called Damping Adjustment Factor. In
general, the input acceleration response spectrum (ARS) for bridge seismic analysis is given for
5% critical damping. Considering different structure types, abutment restraints, continuous
welded rail restraints, etc., the acceleration coefficient can be reduced by a factor expressed as:
D [1.5 /(0.4 1)] 0.5

(2)
For this bridge, after discussion with the Geotechnical Engineer, the critical damping of 8% is
used. Table 4 summarizes the coefficients and factors defining seismic hazard for three level
ground motions.
Table 4

Base Acceleration Coefficient, Site Coefficient and Damping Adjustment Factor

Ground
Motion Level

Performance
Criteria

Return Period
(R Years)

Base Acceleration
Coefficient (A)

Site
Coefficient (S)

Damping Adjustment
Factor (D)

Serviceability

100

0.330

1.20

0.86

Ultimate

500

0.591

1.20

0.86

Survivability

2400

0.905

1.20

0.86

Finally, in AREMA Manual, the Seismic Response Coefficient (or ARS) to be used as
input for bridge seismic analysis can be computed by
C m 1.2 ASD / T 2 / 3 2.5 AD

(3)
6

The three curves, representing three level seismic ground motions, are shown in Figure 6. Note
that short period structural response is assumed to attenuate to peak acceleration.

Figure 6. Seismic Response Coefficient (Acceleration Response Spectrum)


SEISMIC RESPONSE ANALYSIS
For Level 1 ground motion, the seismic design forces shall be determined by an elastic
response spectrum analysis. For Level 2 and Level 3 ground motions, the seismic design forces
may be computed by increasing the forces from Level 1 ground motion by a scalar ratio of the
Base Acceleration Coefficients. Therefore, only the Level 1 seismic analysis needs to be
performed.
A traditional Warren Truss is selected for the bridge superstructure, as shown in Figure 7.
The total 200 ft span is divided into 8 bays with 25 ft long bay spans. The first bay and the last
bay have different bay spans lengths to accommodate the required structure skew. The two
trusses are connected by bottom main floor beam with stringer system and top and bottom lateral
bracing systems, together, this assembly forms a through truss structure. Both Span 1 and Span 2
are simply supported on Abutments and Bent.

Figure 7. Truss Elevation View


For bridge substructure, Bent 2 is a two-column bent with column diameter of 6.0 ft,
supported by 8.0 ft diameter CIDH pile shafts. Both Abutment 1 and Abutment 3 are seat type
abutments supported by a group of 2.0 ft diameter CIDH pile shafts. Expansion Bearings are
used at the Abutments to connect the superstructure to the substructure, while Fixed Bearings
are used at the Bent.
7

The software program CSIBridge (SAP2000) is employed to calculate bridge member


forces, both statically as well as dynamically (through a response spectrum analysis as outlined
in AREMA). A 3-D analysis model is shown in Figure 8. The loads in static analyses include
Dead Load, locomotive and train Live Load (Cooper E-80), Wind Loads, Braking Force,
Traction Force, etc. As specified in the AREMA Manual, a multi-modal response spectrum
analysis is performed to compute member demand forces. During this dynamic analysis, an
iteration process was performed for abutment stiffnesses in order to limit the abutment force
responses not to exceed the resistances provided by soil behind the abutment backwall and by the
abutment shear key system. At the location of the column base terminus and the pile shaft cutoff
elevation of Bent 2, a 6 x 6 stiffness matrix is sub-structured within the model to represent the
stiffness of 8.0 ft diameter CIDH pile shaft vertically and laterally supported by the surrounding
soil layers profile. The software program LPile, by Ensoft is used to estimate linear CIDH pile
stiffnesses at the pile head.

Figure 8. CSIBridge 3D Bridge Model


Table 5 lists typical results, including dead load responses and seismic responses of truss
members, in which, P is axial load, MT is out-off plane bending moment and ML is in-plane
bending moment.
Table 5 Truss Member Dead Load and Seismic Responses
Dead Load Responses

Seismic Responses (+ / -)

Member
Location

Member
Name

P (kips)

MT (k-ft)

ML (k-ft)

P (kips)

MT (k-ft)

ML (k-ft)

Top
Chord

U1U3 & U5U7

-523

66

105

43

19

U3U4 & U4U5

-730

48

160

85

40

Bottom
Chord

L0L2 & L6L8

259

70

41

499

158

91

L2L4 & L4L6

499

48

50

658

78

45

L0U1 & U7L8

-547

111

29

153

909

131

U1L2 & L6U7

410

110

17

69

236

15

L2U3 & U5L6

-265

75

11

41

80

24

U3L4 & L4U5

106

65

34

130

18

U1L1 & U7L7

85

26

24

14

U2L2 & U6L6

10

26

Diagonal

Vertical

U3L3 & U5L5

82

15

U4L4

12

SEISMIC DESIGN OF TRUSS BRIDGE


Based on AREMA Manual requirements, only Level 1 seismic demands are calculated.
All members are classified as three types: Ductile member, Non-ductile member or
Capacity-Protected member. Level 1 seismic demands are used to design the ductile members.
Level 2 or Level 3 seismic demands are used to design non-ductile members or capacity
protected member, but the design forces should be limited to the maximum forces which can be
transmitted to the members. Design shall be used to ensure satisfactory performance for both the
ultimate (Level 2) and the survivability (Level 3) limit states.
Before getting into the detailed design of ductile members, the conceptual design is
performed. Table 6 summarizes the preferred and non-preferred features for this bridge. It can be
seen that most of preferred features are satisfied, except for the long and heavy spans seating on
the skewed bent. Considering these non-preferred features, a wider seat width than required per
AREMA is used in the design.
Table 6
Configuration
Alignment: Straight or Curved?
Straight (preferred)
Pier: Normal or Skewed?
Skewed (not preferred)
Pier Stiffness: Uniform or Varying?
Uniform (preferred)
Span Stiffness: Uniform or Varying?
Uniform (preferred)
Span Mass: Uniform or Varying?
Uniform (preferred)

Railroad Bridge Conceptual Design


Superstructure
Span: Simple or continuous?
Simple (preferred)
Span: Short or Long?
Long (not preferred)
Span: Light or Heavy?
Heavy (not preferred)
Hinge: with or without?
No hinge (preferred)

Substructure
Seat: Wide or Narrow?
Wide (preferred)
Bent Cap: Seated or Integrated?
Seated (preferred)
Column: Single or Multiple?
Multiple (preferred)

Figure 9. Typical Truss Member Section


Figure 9 shows a typical truss member section, which is a box section. Two web plates
are solid plates, but two flange plates are perforated plates. The load components include axial,
in-plan (L-L axial) bending and out-plan (T-T axial) bending. The design criterion is set as:
fa
f
f
bT b L 1.0
Fa FbT Fb L

(4)

in which fa, fb-T and fb-L are stress demands along axial, out-plan bending and in-plan bending,
respectively, and are calculated by

fa P / A;

f bT M T / ST ;

f b L M L / S L

(5)

Fa, Fb-T and Fb-L are allowable stress defined by Table 7 below.
Table 7

Allowable Stress Truss Members

Truss Member
Allowable Stress
Check

Allowable Stress
Axial

Allowable Stress
Out-Plan Bending

Allowable Stress
In-Plan Bending

Fa

Fb T

Fb L

Compression

0.55EQ Fy

0.55RT EQ Fy

0.55 R L EQ Fy

Tension
Gross Section Yield
Tension
Net Section Fracture

0.55EQ Fy

0.55EQ Fy

0.55EQ Fy

0.47EQ Fu

0.47EQ Fu

0.47EQ Fu

Here Fy and Fu are steel yielding stress and fracture stress; RT and RL are effective length factors.
It should be pointed out: AREMA Manual specifies that the allowable stress can be increased by
50% when designing steel members for (DL + EQ), which means EQ = 1.5 in Table 7. Some
typical member stress ratios are listed in Table 8. Surprisingly, (DL + EQ) doesnt control the
truss member design, mainly due to the strength increase factor EQ.
Table 8

Truss Member Stress Ratio

Member
Location

Member
Name

(DL+LL)
(= 1.00)

(DL+LL+WL+LF
)
(= 1.25)

(DL+EQ)
(= 1.50)

Controlling
Stress Ratio

Top
Chord

U1U3 & U5U7

0.811

0.718

0.313

0.811

U3U4 & U4U5

0.798

0.671

0.354

0.798

Bottom
Chord

L0L2 & L6L8

0.461

0.507

0.350

0.507

L2L4 & L4L6

0.558

0.536

0.348

0.558

L0U1 & U7L8

0.605

0.729

0.619

0.729

U1L2 & L6U7

0.633

0.597

0.322

0.633

L2U3 & U5L6

0.674

0.703

0.296

0.703

U3L4 & L4U5

0.293

0.307

0.144

0.307

U1L1 & U7L7

0.716

0.866

0.487

0.866

U2L2 & U6L6

0.336

0.320

0.217

0.336

U3L3 & U5L5

0.687

0.727

0.415

0.727

Diagonal

Vertical

10

U4L4

0.321

0.298

0.199

0.321

Gusset plate design under earthquake loading is a typical non-ductile element design.
The load demand is the lesser of truss member capacity and Level 3 seismic force. The force
ratio between Level 3 and Level 1 should be computed as the scaler: rF = A3/A1 = 0.905/0.330 =
2.74. The design requirements include: 1). Member end connection bolts; 2). Gusset plate gross
section (Whitmore Section) tension yielding; 3). Gusset plate net section tension fracture; 4).
Gusset plate compression yielding; 5). Gusset plate combined horizontal shear; 6). Gusset plate
combined vertical shear; 7). Member splice. For this bridge, a typical gusset plate connection
details is shown in Figure 10.

Figure 10. Typical Gusset Plate Connection Details


For designing railroad concrete substructures, Load Factor Design is specified in the
AREMA Manual. As in most other seismic design codes, the load factor for (DL + EQ)
combination is 1.0. The strength reduction factors are: for flexure = 0.9, for shear V = 0.85,
and for compression C = 0.75. However, for RC column design with a combined flexure and
axial load, the strength reduction factor is
M C C (M C )( Pb Pn ) / Pb

(6)

From CSIBridge 3-D static and dynamic analyses, the column moment demands by DL and EQ
are MDL and MEQ. From column section software to determine M-P capacity relations, the
nominal moment capacity under the nominal axial load Pn is Mn. Checking the column moment
capacity can be expressed as
( M DL M EQ ) M C M n

(7)

Since column shear failure is a non-ductile failure mode, the column seismic shear force
demand is taken as

11

VEQ min(VEQ 3 , VMp )

(8)
here, VEQ3 is the column shear force response under Level 3 earthquake, and VMp is the shear
force associated with column plastic moment Mp = 1.3Mn. The 1.3 value is traditional, and also
found within AREMAs reference to ACI guidance materials. Then the column shear capacity is
checked by
(VDL VEQ ) V Vn

(9)

The software program LPile is used to compute the maximum moment and shear force
demands for 8.0 ft diameter CIDH pile shafts as structural elements at Bent 2. The loads
(including moment and shear) applied at the top of pile are the lesser of Level 3 seismic forces or
column base over-strength forces. By pushing pile shaft laterally from top of pile, program LPile
will provide pile moment and shear distribution along the pile length. The design (maximum)
moment and shear can be shown easily with this analysis technique. Additionally, the AREMA
Manual specifies that a 25% capacity increase should be allowed for a drilled shaft. Therefore,
the pile shaft design is based on:
( M DL M EQ ) 1.25 M C M n

(10)
(VDL VEQ ) 1.25V Vn

(11)

Bent cap is a Capacity Protected member, so the demands are again limited to the
greater of column over-strength capacity or level 3 event forces. The design capacity
determination uses techniques typical for flexure and shears in flexural members, and as outlined
in AREMA, except no 1.25 strength capacity increase factor.
SUMMARY
Seismic responses of railroad bridges are different from highway bridges due to their
unique characteristics, such as continuously welded rail (CWR) restraints. The performancebased seismic design criteria are specified in AREMA Manual. This paper presents the seismic
design of a two-span steel truss railroad bridge located in the Inland Empire of Southern
California. First, a detail bridge type selection is carried out. The advantages and disadvantages
of five bridge types and layouts are discussed. A two-span steel truss bridge with span lengths of
200 ft (total 400 ft) and 20 degree skew is selected among other bridge types. Next, a three level
seismic hazard is evaluated based on the three performance criteria limit states. Then, CSIBridge
is used to perform 3-D dynamic analyses under the Level 1 seismic event. Based on the dynamic
analysis results, the superstructure steel truss members and their connections are designed by
Allowable Stress Design Criteria. It is found that, in general, the seismic load case will not
control the truss member design due to the stress increase factor of 1.5. The substructure with
10 ft x 8 ft drop cap, 6 ft diameter columns and 8 ft diameter CIDH shafts (Type II) are designed
by Load Factor Design Criteria. Level 1 seismic demands are used to design the ductile
columns. Level 2 or Level 3 seismic demands are checked to design bent cap and CIDH shafts,
but the design forces are limited to the maximum forces which can be transmitted to the
members. The final design was reviewed and approved by the bridge Owner. Bridge construction
will start soon.
12

REFERENCES
AREMA (2010). Manual for Railway Engineering, American Railway Engineering and Maintenance-of-Way
Association, Lanham, MD.

13

You might also like