You are on page 1of 171

CONTENTS

How to Use This Book IV


Acknowledgments I V
CHAPTER 1 The Geometry of Euclidean Space 1
CHAPTER 2 Differentiat ion 21
CHAPTER 3 Higher-Order Derivat ives; Maxima and Minima 43
CHAPTER 4 Vector-Valued Functions
63
CHAPTER 5 Double and Triple Integrals
77
CHAPTER 6 The Change of Variables Formula and
Appli cations of Integration
97
CHAPTER 7 Integrals over Paths and Surfaces 17
CHAPTER 8 The Integral Theorems of Vector Analysis 143
CHAPTER 9 Sample Exams 161
APPENDIX Answers to Chapter Te ts and Sample Exams 167
1
THE GEOMETRY OF EUCLIDEANSPACE
1.1: VECTORS IN TWO- AND THREE-DIMENSIONAL SPACE
GOALS
1. Be able to perform thefollowing operati ons on vectors: addition, subt raction,scalar multipli-
cation.
2. Given a vector and a point ,be able to write theequationofthe line passing through the point
in the direction ofthe vector.
3.Given t wo points, be able to wri te the equation of the line passing through them.
STUDYHINTS
1. Space notation. The symbol ]R or ]RI refers to all points on the real number line or a one-
di mensionalspace. ]R2refers toallordered pairs (X ,y)which liein theplane,atwo-dimensional
space. ]R3 refers to allordered t ri ples (x, y,z)which lie in three-dimensionalspace. In general,
the "exponent" in ]R71 tells you how manycomponents there are in each vector.
2. Vectors and scalars. A vector has both length (magnitude) and direction. Scalars are just
do not havedirection. Twovectors areequal
If and only Ifthey bot h have the same length and the same

direction. Pictorially, they do not need to originate from the
samestarting point. The vectors shown here are equal.
3. Vector notation. Vectors areoften denoted by boldfacelet ters, underlined letters, arrows over
letters, or by an n-tuple (Xl,X2 ,...,x
71
) . Each Xi ofthe n-tuple is called the lth component.
BEWARE t hat the n-tuple mayrepresent either a point ora vector. The vector (0,0,..., 0) is
denoted O. Your instructor or other textbooks mayuse other notationssuch as asquigglyline
underneath a letter. A circumflex over a letter is sometimesused torepresent a unitvector.
4. Vectoraddition. Vectors may be added componentwise, e.g., in ]R2
(XI,yr) +(X2,Y2) =(Xl +YI,X2+Y2).
Pictorially, two vectors may be thought of as the sides of a
parallelogram. St ar ting from the vertex formed by the two
vectors,we formanew vectorwhichendsattheoppositecorner
;;;?J
v
ofthe parallelogram. This new vector is the sumofthe other
t wo. Alternatively, one could simply translate v so that the
tailofv meets t he head ofu . The vector joiningthe tail ofu
U I
to the head of v is u + v.
?:j U + V I
________1
CHAPTER 1 2
5. Vector subtmction. Just as with addition, vectors may be subtracted componentwi se. Think
ofthis asaddinganegativevector . Pictorially, thevectorsa ,b
a b and a- b form a triangle. To determine thecorrect direction,
hb you should Ibe able to add a b and b to get a. - a- - Thus a b
go,"f<om the tipofb to the tipof8 .
6. Scalar multiplication. Here,each component ofa vector is multiplied by thesamescalar, e.g.,
in JR.
2
,
r(x,y) =(rx,ry) for any real numberr.
The effect of multiplication by a positive scalar is t.o change the length by a factor. If the
scalar is negative, the lengthening occurs in the opposite direction. Multiplication ofvectors
will be discussed in the next two sections.
7. Standard basis vectors. These are vectors whose components are all 0except for a single 1. In
JR.3, i, j and k denote the vectors which lie on the x, y and z axes. They are (1, 0, 0), (0,1,0)
and (0,0,1),respectively. Thestandard basis vectors inJR.
2
arei andj,which are vectors lying
on the x and y axes, and their respective components are (1,0) and (0,1). Sometimes, these
vectors are denoted by
8. Lines. (a) The line passing through a in the direction of v is l (t) = a+tv. This is called t he
point-direction form oft he line because the only necessary informationis the point a and the
direction ofv.
(b) Theline passing through a and b is l(t) =a+t(b-a) . Thisis called the poi nt-point form
ofthe line. To see ifthe direction is correct , plug in t = 0 and you should get the first point.
Plugin t = 1 and you should get thesecond point.
9. Spanning a space. If all pointsin a spacecan be written in theform AI VI + A2V2 + ...+ AnV
n
,
where Ai are scalars, then the vectors VI, ..., Vn span that given space. For example, the
vectors i andj span the xyplane.
10. Geometricproofs. Theuse ofvectors can often simplify a proof. Try to comparevector meth-
ods and non-vector methods by doing example 10 without vectors.
SOLUTIONS TO SELECTED EXERCISES
1. We mustsolve the following equations:
-21- x - 25
23 - 6 y.
We get x =4and y =17, so (-21,23)- (4,6)=(-25,17).
4. Convert -4i+3j to (- 4,3,0),so
(2,3,5)- 4i+3j =(2,3,5)+ (-4, 3,0)=(-2, 6, 5).
3 THE GEOMETRYOF EUCLI DEAN SPACE
7. To sketch v, tart at the origi n and move 2units along the x
axis , then move 3 uni ts parallel to the y axis, and then move
-6units paralleltothe z axis. Thevector w is sketched anal-
ogously. Thevector - v has thesamelengthas v,butitpoints
in the opposite direction. To sketch v + w, t ranslate the tail
of w to the head of v and draw the vector fromthe origin to
the head ofthe translated w. The vector v - w goes from the
y head of w to the head of v.
9. On t he y axis, points have the coordinat s (0 ,y, 0) , so we
must restrict x and z to be 0. On the z axis, points have
z
the coordinates (0,0,z) , so we must restrict x and y to be
(x,O,z)/ / y
0. In t he xz plane, points have the coordinates (:I: ,O, z).so
(
(o,y,o)
we mustrestrict y to be 0. In the yzplane, points have the
coordinates (0,y , z ), so we must restrict x to beO.
x
12. Every point on the planespanned by the given ve tors can be writ ten as aVI + bV2, where a
and b are real numbers;therefore, the plane is described by
a(3,-1,1)+b(O,3,4).
15. Given two points a and b , a line through them is l(t) =a + t(b - a). In this case, a =
(-1,- 1,-1)and b =(1, - 1, 2), so we get
l(t ) = (-1,-1,-1)+ t(2, 0,3)= (2t - 1, -1,3t - 1).
19. Substitute v =(x, y, z) =(2+ i, - 2+ t, -1+ i) into the equationfor x, yand z and get
2:1: - 3y +z - 2 2(2+ t) - 3(-2+ t) + (- 1+ t) - 2
4+ 2t + 6- 3t -- 1+ t - 2= 7.
Since 7f. 0,there are no points (:I: ,y, z)satisfying theequation andlyingon v.
23. J ust as the parallelogramofexample 17 was described by v =sa+ tbfor sand t in [0, 1], the
paralielpiped can be described by w =sa + t b + rc, for s, t and r in [0 , 1] .
Let a , band c be the sides ofthe triangle as shown, and let
Vij denote the vector from point i to poin j . Weassumethat
each median is divided into a ratio of 2 : 1 by the point of
intersection. Then we have
VI2 -a/ 2=- (c - b) / 2;
V23 = (1/3)(a/2+b);
V34 (-2/3)(a+b/2);
28.
1
V45 (a+b)/2.
The vector VI S should be thesumV1 2 + V23 + V34 + V45, or
c- b 1 (a ) 2 ( b ) a+ b c
V15 =--2- + 3 2+ b - 3 a+2' + - 2- =b - 2'
which is the median ofthe vector that ends on c. The other two median' are analyzed the
same way.
CHAPTER 1 4
30. (a) Using x, the numberofCatoms; y, the numberofH atoms; and z, the numberof0 atoms,
as coordinates, we get
p(3,4,3)+q(O, 0,2) =r(l,0,2)+s(O,2,1).
{b) To find the smallest integer solution for p, q, rand s, we balance the equation componen-
twise:
3p=r (equating x)
2s =4p i.e., s =2p (equating y)
2q +3p= 2r+s ~ 6p+2p i.e., q= (5/2)p. (equating z)
Let p =2, then the smallest integer solut ion is p =2, q= 5, r =6, s =4.
(c) Inthediagram,P is (6,8,6), Q is (0,0,10),R is (6,0,12)and
S is (0,8,4). Both sides ofthe equation add up to the vector
(6,8,16).
R
p y
x
1.2:THE INNER PRODUCT, LENGTH AND DISTANCE
GOALS
l. Be able to compute a dot product.
2. Be able to explain the geometric significance ofthe dot product.
3. Be able to normalize a vector.
4. Be abl e to compute the projection ofone vector onto another.
STUDY HINTS
l. Inne1' product. This is also commonly called the dot product, and it is denoted by a .b or
(a,b). Thedot product is thesum2.:::7=1 ajbj, where aj and bj are the ~ t h componentsofa and
b,respectively. Forexample, in lR2, a .b =alb
1
+a2b2. Note that thedot product is a scalar.
2. Length ofa vector. The length or the norm ofa vector x =(x,y, z) is Jx
2
+y2 +z2. Itis
denoted by Ilxll and is equal to VX-:-X.This is deri vable from the fact that x.y =X1YI +
X2Y2 +X3Y3 with x = y.
3. Unit vector. These vectors have length l. You can make any non-zero vector a unit vector by
normalizingit. To normalizea vector, divide the vector by its length, i.e., computea/liali .
4. Cauchy-Schwarz inequality. Knowingthat lahi Ilallllbllis most importantfor doingproofs
in the optionalsections ofthis text and in more advanced courses. .
5. Important geometri c properties. Know thatab=Ilallllbllcose, where Bis the angle between
the two vectors. As a consequence, a .b =0 implies that a and b are orthogonal. The zero
vector is orthogonal to all vectors.
6. Projections. Theorthogonal projection ofb onto a is the "shadow" of b falling onto a . The
projection ofbonto a is a vector oflength (ab)/llall ,in the direction ofa/liali. Thus, the
projection ofb onto a is
(a' b)) a (ab)
(
-W W ~ a .
.. .
THE GEOMETRY OF EUCLIDEAN SPACE 5
7. Problem solving. Since vectors have magnitudeand direction, they can be represented picto-
ri all y. It is often useful tosketch a diagramto helpyou visualize a vector word problem.
SOLUTIONSTO SELECTED EXERCISES
3. From t he definit ion oft he dot product, we get
uv (0 7 19) (- 2 - 1 0) -7
cos(j = = ' , " = ~ -01546
Il ullllvll V7
2
+ 19
2
V22 + 12 V410V5 . .
From a hand calculator, we find t hat (j ~ 99.
7. If w =ai+bj +ck, t hen Ilwll =va
2
+b
2
+ c
2
, and so
Il ull = vT+4 =..)5; Il vll = v1+l= h .
Using t heformula for the dot product,we get
uv =(-1)(1)+ (2)(- 1)=- 3.
10. Using t he same formulasas in exercise 7, we get
Il ull =VI+ 0+ 9=JiO; Il vl l =v O +16+ 0 =4.
Since u does not have aj component and v does not have any i or k component, t he vectors
are perpendicul ar; therefore n .v =O.
12. A vector w is normali zed by constructing the vector w/ll wll. For the vectors in exercise 7:
u 1 ( . 2') v 1 (' ' )
W = V5 -)+ J; IRI = V2 ]- J .
15. The projection ofv ontouis
nv _(-1)(2)+(1)( 1)+ (1)( - 3)(_, , k)-_i(_, , k)
Il ul1 2u - (V
1
+
1
+
1
)2 l + J + - 3 I+J + .
16. For ort hogonality, we want the dot product to be O.
(a) The dot product is (2i+ bj).(- 3i+ 2j + k)=-6+ 2b,so bmust be 3.
(b) Thedot product is (2i+ bj).k =0, so bcan be any real number.
21. (a) Looking at the x componen s, the pilot needs to get from 3 to 23. His velocity in the x
direction is the i component ,400km/hr. Thus,
Llt = d =23 - 3 =
v 400 20
The pilot flies over the airport (1/ 20) hour or 3 minutes later. The same answer could have
been obtained by analyzing the ycomponents.
(b) We look at the zcomponents and use theformula
~ d =vLlt , i.e., h - 5=(-1)( 1/20),
so h =99/20. Thus, the pilot is 4.95 kmabove the airport when he passes over.
24. (a) It is convenient to draw the diagram with A on t he .r axis.
CHAPTER 1 6
y
A x
(b) From the diagramin part (a), we get
A =150i and B =(llOcos60
0
)i+ (llOsin60
0
)j.
A + B = (150+ llOcos60
0
)i+ (llO si n60
0
)j = 205i+ 55V3j.
The angle that A +B makes with A is
() =tan-
1
( ; ) t an-
1
(0.4647) 250. =tan-
1
Alt.ernat.ively, the definition of the dot productgives us
() = cos- 1 ( A (A+ B) ) =cos- 1 (150)(205) + (0)(55V3)
IIAII IIA+ BII (150)(J51l 00)
25.
27. (a) Geometrically, we see that the i component of F is
IIFIIcos(). Similarly, thej component ofFis IIFIIsinB. There- y
F
fore,
F=IIFIIcos()i +IIFIIsi n()j,
where () is the angle from the x axis. Since t he angle from t he
y axis is 7r/4, () is also 7r/4, so F =3J2(i+j).
(b) We computeD= 4i+2j,soFD= (3)2)(4)+(3)2)(2) =
18"\/2. Also, II FII =6 and IIDII =J20. From the defi nition of
the dot product,
F D 18)2 3
cos() = II FII IIDII = 6J20= vT5 0.9487,
i.e., () 18, or equi valently,() 0.322 radians.
(c) Frompart (b), we hadcomputed FD =18)2. Knowing t hatcos() =3/.JfO, we calculate
IIFII liDII cos() =(6J20)(3/JIO)=18)2,also.
1.3: MATRICES, DETERMINANTS AND THE CROSS PRODUCT
GOALS
1. Be able to computea cross product.
2. Be able to explain the geometric significance ofthe cross product.
3. Be able to write the equation ofa planefrom given informationregarding pointson t he plane
or normals to the plane.
x
7 THE GEOMETRY OF EUCLIDEAN SPACE
STUDYHINTS
1. Matri ces and determi1wnts. A mat rix is just a rect angular array ofnumbers. The array is
written between a set of brackets. The determinant ofa matrix is a. number; a matrix has
no numerical value. The determinant is defi ned onl y forsquare matrices and it is denoted by
vert ical bars.
2. Computing determinants. Know that I ! I =ad- be. Also know that
abc
d e f
9 h
Note the minus sign in front ofthe second term on the right-hand side. The general method
for computing determinants is described next .
3. Computing n x n determinants. Use the checkerboard pattern shown here which begins wi t h
a plus sign in the upper left corner. Choose any column or
row - usually picking theone wit,h the most zeroes saves work.
+ +
Draw verti cal and horizontal lines through the first numberof
+
t he rowor column. Thenumbersremainingform an (n-1) x
+ +
(n-1) determinant, whi chshouldbemultipliedby thenumber
(wit h thesigndet ermined by t hecheckerboard) through which
bothlines are drawn. Repeatfor the remainingnumbersofthe
row or column . Finally,sum the results . This process, called
expansion by minors,works for any row or column. The best
way to remember the process is by practicing. Be sure to use
the correct signs.
4. Simplifying determinants . Deter minants are easiest to compute when zeroes are present.
Adding a non-zero mul tiple of one row or one column to another row or column does not
change the value ofthedeterminant, and this can oft en simplify the computations. See exa.m-
pIe 3.
5. Computing a cross product. If a = (aI,a 2,a3) and b = (b
1
,b
2
,b
3
), then
J k
a x b = a1 a2 a3
b
1
b
2
b
3
The order matt ers: a x b =-(b x a). The cross product is notcommutative. Also, note that
a x b is a vector , not a scal ar .
6. Properti es ofthe cross product. The vectors a, b and a X b form a right-handed system (see
figure 1. 3. 2of the text). The cross product a x b is orthogonal to both a and b. The length
ofa x b is II allllbill sinOI, where 0 is the angle between a and b. Note that the formula for
the cross product involves sin0, whereas t he dot product involves cosO.
7. More properties. Ifthe cross product is zero, then either: (i) t he length ofone ofthe vectors
must be zero, or (ii) sinO =0, i. e. , 0=0,so the vectors must be parallel.
8. Geometry. The absolute val ue of the determi nant I ! I is t he area ofthe parallelogram
spanned by the vectors (a, b) and (c,d)originat ing fromt hesamepoint. Theabsolutevalue of
a bc
thedeterminant d e f is thevolumeoftheparallelpipedspanned by thevectors (a, b, c),
g h i
(d,e,J) and (g,h,i) originat ingfromthesamepoint. Thelengthof the cross product 11a x hll
is the area oft he parallelogram spanned by the vectors a and b. The vector a x b gives a
vector normal to the plane spanned by a and b.
CHAPTER 1 8
9. Equation ofa plane. Recall t hat theequationofthe plane is ax+by+cz+d= O. Thevector
(a , b, c) is orthogona] to the plane. Knowing two vectors in the plane, we can deter mine an
orthogonal vector by using the cross product. Compare methods 1and 2 ofexample 11.
10. Distance from point to plane. You should understand the derivation ofthe equation in the
box preceding example 12. Ifnecessary, review the geometricproperties ofthe dot product in
section 1.2.
SOLUTIONS TO SELECTED EXERCISES
2. (b) We subtract 12 times the third row from the first rowand subtract 15 times the third row
from the second row. 'fhen we expand along the first column:
36 18 17 o -42 41
-42 41 1 . - . -
45 24 20 o -51 50
=3 -51 50 =3(-2100+2091) =-27.
1
3 5 -2 3 5 -2
5. The areaofthe parallelogram is /l a x hll. We compute
j k
a x b = 1 -2 1 =-3i+j +5k
2 1 1
andso the areaofthe parallelogramis
lin x hll =v'9+1+25 =55.
8. The volumeofthe parallelpiped is the absolute value ofthe 3x3 determinant madeup oft he
vectors' components. Expand along t he first row:
1 0 0
_1
3 -1
0 3 -1
- 2 'I =-l.
-1
4 2 -1
Thus, the volume is 1.
11. We want tofind the cross product and then normalizeit . We compute
So
J k
v = -5 9 -4 I =113i+17j - 103k.
789
There are twoorthogonal vectors in opposingdirections;they are given by
v/llvll=: (113i+17j - 103k)/ v'23667.
Since all vectors are orthogonal to 0, the inclusion of that vector into the problem does not
affect the answer.
15. (a) The equation of a plane with normal vector (A,B, C) and passing through the point
(xo,Yo, zo) is A(x- xo) +B(y- Yo) +C(z - zo) =O. In thiscase, t he equation is
l(x- 1) +l(y- 0) +l(z- 0) =0 or x+y+z =1.
(d) Here, the normal vector is parallel to the line, so it is (-1, - 2,3). Hence, t he equation of
the desired plane is
-1(x- 2) - 2(y- 4) +3(z+1) =0 or - x - 2y+3z+13 =O.
9 THE GEOMETRY OF EUCLI DEAN SPACE
16. (b) Two vectors in the desired plane are v = (0 - 1, 1 - 2, - 2 - 0) = (-1,-1, - 2) and
w = (4- 0,0- 1, 1+ 2) = (4,- 1,3). Thecross pr duct v x w is orthogonal toboth vectors,
and hence normalto thedesired plane. We compute
v xw =- 5i - 5j+ 5k,
so the desired equation is
-5(x - 1) - 5(y - 2)+ 5(z - 0) =0 or - x - y + z + 3=O.
22. (a) Let D be the matrix with rows u, v, w. Then
Ul U2 U3
detD = u(vxw)= VI V2 V3
WI Wz W3
Use the followingproperty ofdeterminants: v(w x u) corresponds to two rowexchanges of
the matrixD, 80 we hav
v .(w x u) = (-1)( -1)detD =detD and w (u x v) = (-1)( -1)detD =detD.
Toprove the other three, recall that u x v = -(v xu).
26. The line perpendicular to the plane is parallel to the normalof the plane,so t he equation of
the line is
l (t) =(1, - 2, - 3)+t (3, -1,-2).
29. Let u be the vector normal to the plane. Then u is perpendicular to 3i+ 2j + 4k since v
is on the plane. Alsu u is perpendicul ar t o 2i + j - 3k, because the vector Ai + Bj + Ckis
erpendicular to all vectors in t he plane Ax + By + Cz = D. To find u, we take t he cross
produ t of3i+ 2j+4kand 2i+j - 3k:
J k
u = 32 4 =-10i+ 17j - k.
2 1 - 3
When t = 0, we fi nd that a point on the plane is (- 1,1, 2),so theequation ofthe plane is
- 10(x+ l) +17(y -l)- (z-2}=0 or - 10x + 17y-z - 25 =0.
30. First, find thenormaloftheplane. Thenormaloftheplaneisperpendicular 0 theline passing
through (3,2,- 1) and (1,- 1,2).The equationofthelineis l(t) = (3,2,- 1)+t (2, 3,- 3). The
normaloft heplaneis also perpendicular to v =(1 ,- 1, 0)+t (3, 2, -2).Therefore, twovectors
on t hedesired planeare2i+ 3j- 3kand3i+ 2j- 2k,an the normalis (2i+ 3j- 3k) x (3i+
2j - 2k) = - 5j + 5k. Now, we need a point on the plane,say, (3,2,-1). Thus, the equat ion
oft he plane is
O( z - 3) - 5(y - 2) + 5(z + 1) =0 or - y + z =l.
34. The plane passi ng through the origin and perpendicular to i - 2j+ k is x - 2y + z = O. By
the dist anceformula with (A, B, C,D) = (1,- 2,1,0) and (Xl,Y1, zt) =(6,1,0),
d = IAxl + BYI + C Zl + DI = 6- 2 _ --- _2V6
VA2 + B2 + C2 VI+ 4+ 1 V6 3 '
37. Since all vectorsin thisexercise are unitvectors, li N xall = sin0
1
and II N xhll = sinO
2
. From
Snell ' s law, n l sin0
1
= n2 sinO
2
. Hence,
nlllN x all =n211N x hll.
To establish that N x a and N x h have the same direction, we assume t hat N, a and hall
liein the same plane, and a and h are on the sameside ofN. Hence N x aand N x h both
are perpendicular to this plane and parallel to ach other . Thus nl ll N x al l and n211N X hll
are equal.
10 CHAPTER 1
38. First, 4 times the first row is subtracted from the second row. Next, 7 times t he first row is
subtracted from the third row. The next step is expansion by minors along the first column.
Finally,the 2 x 2 determinant is computed.
1.4: CYLINDRICAL AND SPHERICAL COORDINATES
GOALS
l. Be able to convert back and forth between the cylindrical, spherical and cartesian coordinate
systems.
2. Be able to describe geometric objects with cylindrical and spherical coordinates.
3. Be able to describe the geometriceffects ofchanging a coordinate.
STUDYHINTS
l. Review. You should review polarcoordinates in your one-variablecalculus text.
. .
2. Cylindrical coordinates. Denoted (r,{}, z), this is just like polar coordinates except that a z
coordinate has been added. Know the formulas x =rcos{}, y =rsin{}, r =Jx
2
+y2 and
{} ::::; tan-1(y/x).
3. Spherical cooridinates. Denoted by (p, {}, , p is the distance from the origin, > is the angle
from the positive z axis and {} is the same as in cylindrical coordinates. Know the formulas
x=pcos{}sin>, y=psin{}sin> and z=pcos>.
Also know t.hat
p= Jx2+y2 +Z2, {} = tan-1(y/x) and > = cos-1(z/Jx
2
+y2 +z2)= cos-
1
(z/ p).
4. Graphs ofr, p = constant. Note that r = constant in cylindrical coordinates describes a
cylinder and that p = constant in spherical coordinates describes a sphere. You may have
suspected this from the nameofthe coordinatesystem.
5. Computing {} and >. Remember that, in this text, > takes values from 0 to rr and {} ranges
from 0 to 2iT. In someinstances, it is more convenient to define {} in the range -rrto iT. You
shouldbe verycareful aboutcomputing{}. If x = y = -1,t.hen tan-1(y/x) =iT/4, butplotting
the point (-1,-1)in the xy planeshows that, in reality, {} =5rr/4. This is why the authors
fuss with tan- 1(y/x) in the definition. Plotting the x and y coordinates is very helpful for
determining{}.
6. Negative r, p. Note that we have defined r and p to be non-negative. Ifthe distance is given
as a negative number, we need to reflect the given point across the origin.
7. Unit vector's in spherical and cylindrical coordinates. The unit vectors in cylindrical coordi-
nates are e
r
, eo and e
z
. The vector e
r
points along the direction of r, while eo goes in the
direction in which {} is measured, and e
z
=k. As one might expect., those three unit vectors
form an orthogonal basis, and e
r
x eo =ez . Those vectors, however, lire not fixed as is the
case with i, j and k, that is, ifyou change the point (7', {}, z), the set ofunit vectors rotates.
For spherical coordinates, there is also a set ofunit vectorse
p
, eo and e",. Those vectors, in
terms ofi, j, k and the cartesian coordinates ofthe point are worked out in exercise 7 (see
below) ofthis section.
11 THE GEOMETRY OF EUCLIDEAN SPACE
SOLUTIONS TO SELECTED EXERCISES
1. (a) To conver t to rectangular coordinates, use x =rcos& and y =1'Sin11:
x = 1cos45 = v2/ 2 and y = 1sin45 = V2/ 2.
Next , use p =J x
2
+ y2 + z 2 and rP =cos -
1
(z / p) toget thesphericalcoordinates:
Ir. - 1 ( 1l'
P :::;;
VI
2" + 2"
1
+ 1= V2 and = cos )2
1)
="4 '
Hence, for the cylindrical coordinates (1,45,1), the rect angul ar coordinates are given by
(/2/2,)2/2,1) and hesph r ical coordinates are (V2,rr/4,1l' /4).
(b) To convert to cylindricalcoordinates, we use l' =..jX2 + y2 and (J =tan-1(y/x):
r =.)4+1 =v'5 and B=tan-
1
(1/2).
Next , use t he sameformulasas in part (a) toget t h spherical coordi nates:
p=V4+ 4+ 1 = 3 and = cos-1(-2/ 3).
Hence, the rectangular coordinates (2,1,- 2) convert to the corresponding cylindrical coordi-
nat s (/5,t an-
1
(1/2),- 2) and t o thespherical coordinates (3,tan-
1
(1/2),cos-1(-2/3)).
2. (b) This mapping takes a point and rotatesit by 1l' radians about the z axis . This is followed
by a refl ection across the xy plane. The net effect is that the point is reflected through the
origin.
3. (b) Recall that the angle is measured from the "Nor th pole." If is 7i radians , t hen t he
locat ionis at the "Soutl pole." Theeffect. ofchanging to1[' - istakingapoint andreflecting
it across t he xy pl ane.
5. Let p ~ 0, t hen (p, 0,0) is the positi ve z axis. Now, let vary from 0 to 1l'. Then (p, 0,) is
the halfplane in the xz plane wi t h z ~ 0. By allowing & to vary from 0 to 21l', we rot ate t he
half plane descri bed above. Therefore, p ~ 0, (J < 21l' and 0 ~ ~ 1l' describes all points
in JR3.
If p < 0also,thecoordinates are not unique. Forexample, (x, y, z) = (1,1,0)hasspherical
coordi nates (V2, 1l'/4,1l'/2) and (- V2, 51l' / 4,1l' /2).
7. (a) First ,e
p
is the unit vector along t he vector (x, y, z) ; t herefore, t he formul a is
xi + yj+ zk
ep = .
J x
2
+ y2 + z2
z
y
k ~
r
y
x
x
(x, y,O)
Next, eo is parallelto the xy planeanddenotes t hedirection in which t he angle Bis measured .
It is perpendicular to r = (x , y, 0), so (ai + bj) .(xi+ yj ) = O. Since we want B measured
counterclockwise, a =-y (instead ofy) and b = x. Therefore,
- yi + zj
eo = J x 2 + y2'
12 CHAPTER 1
Tofi nde,p , we not e thate
p
, eo and e,p is a set oforthogonalvectors; they form a right-handed
coordinate system with e
p
x eo=- e,p. So .
(x,y,z )x(- y, x,0) xzi+ yzj- (x
2
+ y2)k
= .
rp J x
2
+ y2Jx2+ y2 + Z2
9. (a) Thelength of xi+ yj+ zk is J x
2
+ y2 + z2 , which is the definition ofp.
(b) Note that Ilvll =Jx
2
+ y2 + z2 = P and vk =z; therefore, <p = =cos-1(v.
k/llvll)
(c) Note that Ilull =J x
2
+ y2, which is the cylindrical coordinate rand u i =x; therefore,
() =cos-
1
(x/r) =cos-
1
(u .i /llull).
13. Note that <p will be between 7r /2 and 7r because the region lies in the lower hemisphere. From
the triangle, we see that cosa =(d/6) -;- (d/2)= 1/3; therefore, we have 7r - a ::; <p ::; 7r or
7r - cos-1(1/3) ::; ::; 7r . Now, p can be as large as d/ 2;however, as p gets smaller, its lower
limitdependson <p . Pickany,then +(3 =7r and accordingtothediagramcos(3 =(dj6)-;- p.
Rearrangement gives
d/(6cos(3) =p =d/(6cos(7r - )) =-d/(6cos).
Therefore, -d/(6cos) ::; p ::; d/2. Sofar ,we havedescribed thecross-section inonequadrant.
Theentire volume requires a revolution around t he z axis, so its description is
1
__d_ <p < 0<() <27r and 7r - cos-
1
(-3) <_ A, <_7r.
6cos - - 2 ' - - 'i'
1.5: n-DIMENSIONAL EUCLIDEAN SPACE
GOALS
1. Be able toextend the ideas ofthe previous sections t o jRn.
2. Be able to multiply matrices.
STUDYHINTS
1. The space ]Rn. Most ofthis textbook deals with the Euclidean spaces that we can visualize,
jR2 and jR3. Many ofthe same properties hold in jRn. Vector addition, scalar multiplication,
vector lengths, the dot product and the triangle inequality are defined similarly.
2. No cross product analog. The cross product in ]R3 does nothave an easy analogin jRn, n 4.
3. Standard basis vectors. The anal ogs ofi, j, k are defined ej . Thevector ej is (0,0,...,1,...0)
with 1 in the zth position. The vectors ej and ej are orthogonal if i =f j.
13 THE GEOMETRY OF EUCLIDEANSPACE
4. Matrices. A matrixis a rect angul ar array ofnumbers. Unlike a determinant, a matrixhas no
numerical value. You should remember that an n x m matrixhas nrows and mcolumns. The
(i,j)entry is the number located in row i, columnj.
5. Mat rix multiplication. Youshould practice until matrixmultiplicationbecomes second nature
toyou. Let thecomponents of A be ai j and let those of B be bkl, where A is an m x pmatrix
and B is a p x n matrix. Then the components ofAB are
p
(AB)mn =Lamjbjn.
j =l
We can onl y multiply an m x pmat rix with a p x nmatrix, i.e., [m x p][P x n]. Note that
the numberofcolumnsofA and the number ofrows ofB must be equal (p in t his case). The
result is an m x n mat rix ("cancelling" the p).
6. Non-commutativityofmatrixmultiplication. Ingeneral, AB#BA. Infact, ABmaybedefined
when BA is undefined. However ,matrixmultiplicationis associat ive; i.e., (AB)C= A(BC) if
the product ABCisdefined .
7. Matrices and mappings. An m x n matrix can represent a mapping from ]Rn to ]Rm. To see
this, let A be the mat ri xand let x be a vector in ]Rn, represented as an n x 1 matrix, and y
be a vector in ]Rm, an m x 1matrix. Then the matrixA takes a point in ]Rn to a point in ]Rm
by the equation Ax=y.
SOLUTIONS TO SELECTED EXERCI SES
2. (a) Use the properties ofleng hs and dot products:
(x+y).(x+y)+(x- y) . (x- y)
x.x+2x.y+y.y+x.x- 2x.y+y.y
2xx+2yy=211xl12 +211Y11 2.
The figure at the left depicts the equation geometrically. By
the law ofcosines, we have
12<tl
y and
We also note that a +j3 = 1r, so j3 = 1r - a . Therefore, cosj3 = cos(1r - a) = - cosa. Adding
the twoequations from the law ofcosines yields 211xl1
2
+211Yl12 ;:Ilx- Yl12 +I'lx+Y112.
4. To verify the Cauchy-Schwarz inequality, we compute
Ixyl 1(1)( 3)+(0)(8) +(2)(4) +(6)(1)1 =17.
Ilxll '1' 1+0+4+36= J41.
Il,yll J9+64+16 +1=V90.
Thus, we indeed have Ix.yl =v'I7v'I7<v4fJ90=II'xlillyll. For the triangleinequality, we
compute
x+y=(4,8,6,7) and II x+yll =J16+64+36 +49 =v'l65<13.
Indeed, we have Ilx+yll <13<15 = 6+9<AT+v'9O =Ilxllllyll .
14
CHAPTER 1
8. We compute
1
n
r:
-1
and

0
2
1 1 -1 1
Expanding by minorsacross the first row gives
detA 3:1
2
-1 I 11
II = 6- 2=4,
0
1 + 1 1
detE
11
0
1
1- 11 1=-3,
det(AE) 1-1 1 1_115 and
3 1
-1 1
1- 31
5
1
-1
1
1=-12
det(A+ B)
-
41 I =8.
11. (a) For n = 2,
I
'xall
det('xA) = 'xa21
For n = 3,
'xall Aal2 'xa13
det( AA) Aa21 'xa22 'xa23
Aa31 'xa32 'xa33
'xalldet('xAd - 'xa12 det('xA2 ) + 'xa13det(AA3)
,X .,X2(a11detAl - al2detA2+ a13detA
3
)
,X3 detA,
where AI, A2 and A3 are 2 x 2 matrices obtained by expandingacross the first row.
Assume that for n =k, det('xA) =Ak detA. Thefor n =k + 1, det(AA) can be found by
a process analogous to the 3x 3case:
det(AA) 'xalldet('xAd - 'xa12 det('xA
2
) + ... + (_l)k 'xal,k+ldet('xAk+d
,Xk+l detA,
where AI, A
2
, ..., Ak+l are k x k matrices obtained byexpanding across the first row.
By induction, det('xA) =An detA for an n x n matrixA.
14.Assume, as in the book, that det(AE) = (detA)(detE). Then det(AEC) =det[(AB)C] =
det(AE)detC = (detA)(detE)(detC).
17. Multiply the two matrices to get the identity matrix:
[
a b] 1 [d -b]
e d ad- be -e a
Similarly, we can show that
15 THEGEOMETRYOF EUCLIDEANSPACE
SOLUTIONS TO SELECTED REVIEW EXERCISES FORCHAPTER1
1. v+ w =(3,4, 5)+ (I ,-I ,1)=(4, 3, 6)=4i+ 3j+ 6k;v + w is thedi agonalofa parallelogram
whose sides are v and w.
3v =3(3,4,5)=(9,12,15) =9i + 12j+ 15k; 3vhas thesamedirection as v wi thlength 3
times the length ofv.
6v+ 8w =6{3,4,5)+ 8(1,-I,1)=(26, 16,38) =26i+ 16j+ 38k;6v+ 8wis thediagonal
ofa parallelogramwhose sides are 6v and 8w. 6v has the samedirection as v with length 6
times that ofv and simi larlyfor 8w.
- 2v =- 2(3, 4, 5) =(-6, - 8,-10) =- 6i - 8j - 10k; - 2v is a vector in the opposite
direct ion of v withlength twice thatof v.
y
y
x
v w =(3)(1) + (4)(-1)+ (5)(1) =4; v w is the number II v llll wll cos B, where B is the
angle between v and w.
v xw is perpendicular to both v and w. Its length is the area ofthe parallelogram spanned
by v and w.
4. (a) Using the point-direction form of the line, we get l (t )= (0, 1,0)+ t (3,0,1).
(b) Using t he point-pointfor m oftheline, we get
l(t)=a + t(b- a) =(0, 1,1) + t [( O,1,0 - 0, 1,1)]=(0, 1, 1)+ t (O,0,-1) .
(c) The normalto the planeis (a, b, c)=(-1,1,-I),so th equationofthe plane is
a(x - xo)+b( y- Yo) +c(z - zo)=0,
i.e.,
- 1( x- 1) +l(y- 1) - 1(z- 1) =or x- y+z=1.
5. (b) v w = (1)( 3)+(2)(1) +(- l ){O) =5.
i j k
1 1
6. (b) v x w = ; 1 =1 1j -I 1j + 1 1k =i - 3j- 5k.
7. (b) We compute IIvll = VI + 4+ 1= V6 and Il wll = V9+1+ 0=v'lO. Then the definit ion
ofthe dot product and the result of xercise 5(b) gives
v w 5 5 5
cosO = = =----
Jlvllll wil Y6v'lO V60 2.Jl5
16 (HAPTER 1
8. (b) The area ofthe parallelogramis the length ofv xw. Using the result ofexercise 6(b), we
get Ilv x wll =VI + 9+ 25 =v'35. ,
12. We compute the following dot products using the fact that u, v and w are orthonormal:
a .u (au + f3v + , 'w) .u =a.
av (au + (3v + ,w).v =,B.
aw (au + (3v + ,w).w =,.
v
Geometrically, a .u is the projection ofa on u; similarly for
the others.
15. From the definition ofthe dot product and the fact that U u = lIull
2
, we compute
v .a ~ lIall,b .a + IIblillall
2
_ ba + IIblili all
cos()
II vII IIall 11vlliiall IIvll
v . b lIallllbW+ IIblla.b lIalillbl ,1+ b .a
cos() =
IIvllllbll IIvllllbll Ilvll
Since t he angle between a and v is thesameas the angle between b and v, the vector v must
bisect the angle between a and b.
18. (a) Given that a b =a' .b, this implies that ab - a' .b =or (a- a') .b =0 for all b.
Choose b =a - a' toget lIa- a'1I2 =0, so a - a'=0, which means that a =a'.
(b) The answer is yes. Ifaxb =a' x b for all b, we can concl ude that (a- a') x b =0 for
all b. Choose b to be a unit vector orthogonal to a - a'. By definition ofthe cross product,
this implies lIa - a'il =0, so a = a'.
22. (e) Note thatthe x and y coordinates lie in t he thirdquadrant
ofthe xy plane. Thedefinitions from section 1.4are used: For
cylindrical coordinates, we compute
r JX2 + y2 =v'12+ 4 =4.
()
7r + tan-
1
(y/x) = 7r + tan-
1
(1/V3)
7r + 7r/6 = 77r/6.
Thus, the cylindricalcoordinates are (4,77r/6,3).
x
For spherical coordinates, we compute
p = Jx
2
+ y2 + z2 = V12+ 4+ 9= 5.
4> =cos-1(z/p) =cos-1(4/5).
Thus, the spherical coordinates are (5, 77r/6,cos-
1
(4/5)),
/
I
y
23, (b) Using the formulas from section 1.4, we calculate
z
x =rcos() =(3)(V3/2)
y
and
y = rsin()= (3)(1/2).
Thus, the corresponding cartesian coordinates are
(3V3/2,3/2, 4 .
For the spherical coordinates, we compute
THE GEOMETRYOF EUCLIDEAN SPACE 17
p= Jx2+ y2 +Z2 =/27
V4 4
cjJ = cos-
1
(z/p) = cos- 1( -4/ 5).
Hence, the corresponding spherical coordinates are (5,1r/6,cos-
1
(- 4/5)).
24. (b) Using th formulasfrom section 1.4, we compute:
z
x pcosBsincjJ = 2(0)( 1/ 2) = 0,
y psinBsin<fJ = 2(-1)(1/2)=- 1,
y
z Pcos<fJ =2(..;3/ 2) =v'3.
Thus, the corre ponding oordinates are (0,- 1,..;3).
x
For the cylindrical coordinates, we calculate:
r = J x
2
+y2 =JO+l=1.
We note that x = 0 and y = -1, so B = 311'/2. Therefore, the corresponding cylindrical
coordinates are (1,3rr/ 2,V3).
28. Using the methodsofsec ion 1. 5, we get
0 0 0
0 1 0
AB = [
=
n
n[!
n
0 0 0
0 0 0
BA= 1 0 0
and
[
;][ l
0
nr: n
0 0
Clearly, AB :j: BA.
33. (a) r is t he vector 7i+ 2j, so W =F r =70co B+ 20sinB.
(b) IfF has a magnitude of 6, then F = 6cosBi + 6sinBj. Since B = 11'/6, we have F =
6( J3/2)i+ 6(1/2)j ,andso W =F .r =(21V3+ 6) foot-lb.
36.Subtract t he fir t row from the second and third rows, t hen expand by minors along the first
column toget
2 2
1 X x 1 x x
2
I
0 y -
y2 x2 _
1
y2 =I y - X
y2 _ x
x
Y
z - x
z2 _ x
2
z2 _ x2
1 Z
Z2
0 z - x
(y - x)(z2 - x2) - (z - X)(y2- x
2
)=(y- x)(z- x)( z- y) :j: 0
ifX, y, zare al ldifferent. The last step used the fact that (z2 - x
2
)= (z - x)(z+ x).
39. (a) We recogni z a(b x c) as a t ripleproduct . Let a =(a1'a2 ,a3) and use a similarnot ati on
for bandc. Therefore,
1 a t a2 a3
V =6" b1 b2 b3
Cl C2 Cs
(b) Use t he formul a from part (a) andsubtract the first row fromth second row;
1 1 1 1
1 1 1 1 110 - 2 1 1
- 1 -1 1
6 = 6" 1 1 =3'
6 1 1 0
Thus, the volumeis 1/3.
18 CHAPTER 1
44. A vector which is normal to t.he first plane is v = 8i+j +k. A vector which is normai to the
second plane is w =i- j - k. The cross product v x w is orthogonal toboth normalvectors,
so itshould be parallel to both planes. We compute v x w = 2i - 7j- 9k,so the unit vector
in question is (2i- 7j - 9k)/ v'f34.
47. We want v=ai+.Bj +,kto be such that !lvll =1. From the definition ofthe dot product,
we know that
o v'3 v j
v'3, /3.' k
so v =2 1+ J +, .
cos30 =2 = Ilvllllill'
Since vmakes equal angles wi h j and k, we must have .B = , . Since Ilvll == 1, we determine
that .B =1 =1/20.Therefore,
TEST F OR CHAPTER1 (The answers are at the back ofthis book.)
1. True or false. Iffalse, explai n why.
(a) Given vectors u=i + j inIR2, v = i + j+kinIR3and w = 2i -kin IR
3
,we have uv= 2
and vw = 1.
(b) Gi ven two mat rices A and B, det(AB) is defined ifand only ifdet(BA ) is defined.
(c) For non-zero vectors a and b, the set a, b, and a x b, in that order, always forms a
right-handed syst em.
(d) Suppose a, b, and c are non-zero vectors in IR3 and ab =a c =0. Then either b is a
multipleofc or b and c are parallel vectors.
(e) Across product is 0 only ifone ofthe vectors is O.
2. Find an equation for the plane passing through the point (0,0,1) and parallel to the line
containing t he points (1,0, 2)and (-1,2,1).
3. Let A = (2,1,3), B = (-3,2,4) and C = (1 , 2, 2) be the vertices ofa triangl e. Let a be the
vector from A to B and let b be t he vector from B to C.
(a) Calculate a and b.
(b) Describe the triangle ~ A B C using vectors oftheform sa +tb. (Find restrictions on the
parameters sand t.)
(c) Calculate the areaof A B C
4. Let u =3i+ OJ + k and v = i- 2j+ k have a common t ail . Let () be the acute angle between
u and v. Calculatesin().
5. A parallelpiped spanned by the vectors (3, - 2,1), (0,2,-1)and (-1, 0,1) is filled with sand.
The sand will fill another parallel piped spanned by the vectors - i, 2j and zk. Compute all
possible values ofz.
6. (a) In]R4, fi nd anequation for thelinepassingt hrough A =(1,3,- 2,0) and B == (0,1,-1,1).
(b) Showthatthelinein part (a) isnotorthogonaltothelinel(r) = (1,3,-2,0)+r(2, 0,3,1).
7. Let
(a) Compute AB.
,. "
THE GEOMETRY OF EUCLIDEAN SPACE
(b) Computedet(AB) and interpret geometrically.
(c) If det(BA) exists, computeit and interpret it geometrically.
8. Convert the spherical coordinates (-2,-11"/4, 27r/3) to
(a) Cylindrical coordinates.
(b) Rectangular coordinates.
9. Let u =(2,-1,0,1) and v =(-1,- 2,1,2).
(a) Verify the Cauchy-Schwarz inequalityfor the given vectors.
(b) Calculate the projection of v ontou.
10. A unique house has an inclined fl oor where the owner
keeps his dog. The dog' s play area is a fenced en-
z
closure with vertices at A = (2,4,1), B = (1,2,0) ,
C = (0,0,-1), D = (1,-1,-1) and E =(3,0,0), as
shown in the diagram.
(a) Allofthe vertices lieon a plane. Find theequation
ofthe plane.
D
x
(b) Find the area ofthe dog's play area.
(c) An object is located at (1,1, 3). What is the min-
imum distance the dog must leap off the fl oor to
reach the object?
19
A
y
---------------- - - ------ --- --
21
2 DIFFERENTIATION
2.1: THE GEOMETRY OF REAL-VALUED FUNCTIONS
GOALS
1. Be able to define a graph, a level cur ve, a level set, and a section.
2. Be able to graph a function f : ]R
2
--+ ]R.
STUDY HINTS
1. Notation. f :A C ]Rn --+ ]Rm describes a functionf The domainis A, which is a subsetof]Rn.
Points in A are mapped to points in the range, which is a subset of]Rm. The only restriction
on nand m is that they have to be positive integers.
2. Real-valuedfunction. In thenot ation, f : A C ]Rn--+ ]Rm, we restrict m to be 1 in this section.
A real-valued function assigns a unique real number to each point in A.
3. Graphs. The graph of f : A C ]Rn --+ ]R is dr awn in the space ]Rn+l. If x is a pointofA, the
the graph consists ofall points in ]Rn+l wi th the form (x,f(x)), where f (x) is a real number.
4. Level set. This is the set ofx such that f(x) is a constant. In ]R2,such a set is called a level
curve and in ]R3, it is called a level surface. Level sets are importantfor graphing.
5. Sections. These are intersections ofgraphs with a vertical plane. Usually, the most helpful
sections for graphs in ]R3 are he intersections with the planes x =constant and y =consta.nt.
6. Graphing in JR3. Often, the best way to draw a graph in JR3 is to draw level curves for
z =constant. Then lift or drop the curves to the appropri at e "height" for z =constant.
Analyzing the sections helps complete the graph. It is a good idea to review how to sketch
the graph ofan ellipse, a hyperbola, a circle and a parabola fromyourcalculusorprecalculus
text.
7. Sketching planes. Many ofus are poor artists, and as a result , t hree-dimensional geometry
may be frustrating due to this problem rather than a lack ofmathematical understanding.
Planes are sometimes easily sketched by plotting t hree non-collinear points (usually on the
coordinate axes) and then passing a plane through them.
8. Spheres. Theequation (x-a)2+(y-b)2+(z-c)2= r2 represents asphere ofradiusreentered
at (a, b, c).
9. Cylinders. A surface in ]R3 is called a cylinder if x, yor z is missing from the equation. A
cylinder canbesketched by drawing the level curve in the plane where t he missing variable is
zero. Then move the curve along the axis ofthe missing variable.
10. Graphs in ]R4. Example5gives a function whose graph cannot be drawn on paper. Tosee the
"graph," one can make a movie which shows the concentric spheres expanding.
22 CHAPTER 2
SOLUTIONS TO SELECTED EXERCISES
1. (a) Todeterminethe levelcurves, we lookatt he equation c= x - y+2, where cis aconstant .
The equation is t he sameas y = x+ (2- c), which is the equation of a line with slope 1and
yintercept 2- c.
The graph offis a plane withintercepts at (-2, 0, 0), (0 ,2,0) and (0 , 0,2).
2
2. (b) We look at c =1- x - y 2 for c =constant . This rearranges to x
2
+ y 2 = 1- c, which
is the equation for a circle ofradius v'1='C, centered att he origin if c<1. Ifc=1, t hen the
levelset is a point at the origin. Th re is no level curve ifc> 1.
3. (3) Substitute x =rcos(} and y = rsinB to get
Since z = 1' 2 does not depend on () , theshape of t he graph does not depend on () .
x
2
c
2 2
5. For constant c, the equation c = y'100- - y2 is equivalent to = 100 - x - y2 or
2
x
2
+y2 =100- c . Th I vel curves are circles with radii VI00- c
2
, centered at the origin.
So, for c =0,2,4,6 8, 10 the radii are 10, v'96, j84, 8, 6 and 0, respectively. Dr awing the
level curves and raising them to t he appropriate zvalues, we obtai n t hefollowinggraph. The
graphoff(x ,y) is a hemisphere. The levelcurves and thegraphare shown here.
y ,
x
10. The level curves have the equation c= x/y or y = x/cfor onstant c. Th se level curves are
lines t hrough the origin with slope l/e. One restriction is that y f:. 0. Next we discuss the
graph z =x/yoNot i e t hat when x is held constant, the sect ion are the hyperbolas c =yz ,
and when y i held constant , we get t he lines c= x/z . Puttingall this information together,
we get t. he "twisted plane" as t he graph.
23 DIFFERENTIATIOI\l
y
12. The level surfaces have the equation c== 4:z;2 +y2 +9z
2
. There is no level surface ifc <O. If
c = 0, then t he level surface is the origin. If c > 0, then we should look at the level curves
for constant k, i.e., analyze c - 9k
2
=4x
2
+ y2 . We recognize that if 9k
2
< c, then the
level curves are ellipses which get smalleras Ikl approaches v'C/ 3. Similarly, we see that level
sections parallel to the yz plane have the equation c- 4k 2 = y2 +9z
2
, which are ellipses with
decreasing "radii" as Ikl approaches .;c/2. Also,the levelsect ions parallelto the xz planehave
k
2
the equation c- =4x
2
+ 9z
2
, which are again ellipses which get smaller as Ikl approaches
.;c. The level surfaces are elli psoidsif cis positive.
17. If c< 0, the level curve is empty. If c= 0, t he level curve is the x-axis. If c> 0, it is the pair
of parallel lines Iyl =c; that is, the lines y =c and y =-c. In the yz plane, we sketch the
graph ofz = Iyl. Since x does not appear in the equation,we get a "cy,)inder" and thissketch
is shifted along the x axis t o obtain the graph in 1R
3
.
y
x
22. The equation can be written as (x
2
- 2x + 1)+ y2 = 1or (x - 1)2 + y'2 = 1. In the xy plane,
this is a circle with r adius 1, centered at (1, 0). Since z is not in the equation, z may take on
any value, so the circle is shifted up and down the z axis.
24 CHAPTER 2
y
-.---:;,.-<-+-- (1,0,0)
x
,
/
25. First, wesketch the graph ofz ::: x
2
in the xz plane. Then shift the graph along the y axis to
get a parabolic ylinder.
x
I I
/ /
/ /
y
29. An equi alentequation is 4x
2
+ 2z2 ::: 3y2. When y ::: 0,t h levelcurve is theorigin. When
y :j:. 0, we have level sections which are ellipses centered around the y axis. The major axes
are parallel to the z axis and the minor axes are parallel to the x axis. The elli pses get larger
as IYIincreases. To complete the graph, we note that when x ::: 0, the section is the straight
lines z ::: y'372y. Thus our graph is two cones.
y
32. Subs itute x ::: rcosB and y ::: rsinB. Therefore, if x
2
+ y2 ::: r2 :j:. 0, then z ::: f (x, y) :::
2xy/ (x
2
+y2 )::: 2( rcos0)(l ' sinB) / r
2
::: r2(2sin0 cos8)j r2 ::: sin20. Thus, t he function reduces
to z ::: sin2B if r' :j:. O. If - 1 z 1, the level curve is one or two straight lines through the
or igin satisfying z ::: sinO (see sketch belowat left). The level curves are raised to a height
z ::: sin2B to obtainthe graph ofa "wrinkled plane." (See graph below atright. Thedot ted
li ne is a portion ofthe xyplane. ) If z > 1or z < - 1, t here is no levelcurve. Notice that the
plane becomes fl atteras I' gets larger.
25 DIF FERENTIATION
x
2.2: LIMITS AND CONTINUITY
GOALS
1. Be able to defi ne the following: open disk, open set , neighborhood, boundary point , limit ,
continuous.
2. Be able to determi ne where a function is cont inuous.
3. Given a function, be ableto compute a li mi t or show that a li mitdoes not exist .
STUDY HINTS
1. Theoretical section. Thissection is notessential for computational purposes. Your instructor
maychoose not to emphasize the materialin t his sect ion. Use your lectures to determine how
important the materialis for your course.
2. Definitions, (a) An open disk around Xo is the se of points x such that Il x - xoll <1' . Itis
denoted Dr(xo) . Note the strict inequality.
(b) An open set U is asetsuch thatevery xo has an opendisk entirely within U. You will need
to find an r when proving that a set is open.
(c) A neighborhood ofXo is an open set containing Xo.
(d) A boundary point of aset A has noneighborhood en irely inside or entirely outside ofA.
3. Review. You shouldreviewtheconceptsofli mitandcontinuityfromyourone-variablecalculus
textbook before continuing.
4. Limits. In thedefinition,be awarethatXo doesn'thaveto bein A;Xo maybe on t heboundary.
Also, f(xo) does not have to be defi ned. We are only interested in the points x n ar Xo . For
proofs, we need to choose U, which is dependent upon N.
5. Properties oflimits. Most of th se are what you would intuitively xpect . Note hat for
multiplication and division, the mappingis into}RI.
6. Conti nuity. Analogous to the one-variable definition, a functi on is continuous at Xoif
lim f(x)= f(xo),
~ o
I.e. , the li mitequals the funct ion value. The li mi t on the left-hand side is concerned about
points nearXo. The right-hand side, f(xo), is concerned about the point Xo itself.
26 CHAPTER 2
7. Non-existent limits. Showing that the limitoff(x,y)does not exist is sometimessi mple. To
show a limit does not exist , one can, for example, look at the limit off by first holding x
constant, then holdi ng yconstant. Ifthe two values differ, the limitdoes not exist.
SOLUTIONS TO SELECTED EXERCISES
2. Take<r ~ y, th n for allpoints (x , y), the open disk Dr(x,y)C B. Thus, B is open.
5. (c) Recall the definitionofthe derivative:
f
' ( )- I' f(xo+h) - f(xo)
Xo - 1m h .
h-tO
By letting f(x)= e"" and xo = 0, we get
h
e" - 1 - eO d I eO+
lim - h- =lim = _ eX =1.
h -t O "-to h dx x =o
6. (c) Fromone-variablecalculus recall that lirn.,-t 0(sin:z:/x)= 1. This fact an al 0 be verified
by using I'Hopital'srule. Using this fact and the properties oflimits,we compute
.2 (.)2 ( .)2
sm x _ I' SillX _ } ' smx _ 12 - 1
lim - 1m - 1m - - .
x
2
x-tO x -t O X x-t O X
g. (c) Itis obvious t hatthelimitofthenumeratoris 0, and thelimitofthedenominatoris 2# 0,
so t he limitof he quotient is 0/2= O.
10. (b) First, hold y = 0constant andlet x approach O. Then use I'Hopital'srule:
. cosx- 1 - (:z:2 / 2) . -sinx- x . - cosx - 1
hm = 11m =hm .
x-t O X4 ""-to 4x
3
x-t O 12:z:2
Thislastlimittends to -00since the numeratortends to - 2and the denominatortends to 0,
Thus, the limitdoes not exist.
11, (a) Let t =xyand use continuityofcompositions (theorem 5) to get
I
sinxy }' sint 1
1m --= 1m-- = ,
(x,y) -t(O, O) xy t -tO t
13. (c) Use the fact that the limitofa vector is the limitofeach component (theorem 3(v)). So
X
we getlim
x
-tl(x
2
,e ) =(l,e).
16. (a)We willuse theorem 5(continuityofcompositions). Notethat f(:z: )=(1- x)8+cos (1+x3 )
is thesumoftwo functions. Thefirst is u
8
with 'U = 1- x. Since tJ is continuous, theorem 5
says that (1- x)8 is continuous. Thesecond function is cosv wit h v =1+x
3
, Again, since v
3
is continuous, theorem5says thatcos (l +x ) is continuous. The sumofcontinuous functions
is continuous, so f(x) is continuous.
17. (a) We can make a function continuous by equating f(xo) and limx-t xof(x). As in exercise
11(a) , we can Ie t =x+y,so
sin(:z: +y) _ l' sint - 1
11m - Im-- - .
( x,y)-t(o,O) x +Y t-t O t
Thus, we let
sin(x+y)
- ~ - - - - - - =1 for x =y = 0,
x +y
"
27 DIFFERENTIATION
and we have a continuous function.
(b) First , we note that ifx =y, then
xy . x'.l 1
lim ') " =lim 2 ., =-2'
x- +y- x-
On the other hand, ifx =-y, then
x y . _x
2
1
lim =hm-- =--
x
2
+y2 2x
2
2
Since the Hmitingvalue depends on the direction ofapproach, the limitdoes not exist at the
origin,so itis not possible to make the function conti nuous at the origi n.
22. (b) We want to find a 6 for every N > 0 such that 0 < x < 6 implies that 1/1xl > N.
Let 0 <6 < l i N , then Ixl < liN implies that 1/1xl > N. This is not true if t he absolute
values are omitted,i.e., may be +ex> or -ex> dependingon which sideof0 we are
approaching from (see the figure below.)
y
x
x
y=l/lxl
y =l/x
27. (a) By the triangle inequality, 1a.
3
+3a
2
+al <la
2
1+31a
2
1+lal, since lal <1 (we assume itis
small, since6,is small), this isless than (or equal to) 514 Choose 6 <1/500,then for lal <6,
la
3
+3a
2
+al ::; 1/100 (note that this is a very rough estimate;a bigger6 would probably work
ifwe work harder to improve the inequality.)
2.3: DIFFERENTIATION
GOALS
1. Be able tostate the definit ion ofpartial derivat ives.
2. Be able tocompute a partialderivativeora mat rixofpartial derivatives.
3. Be abl e tocompute a gradient .
4. Be able to compute a tangent plane.
STUDYHINTS
1. Nota-tion. Class enmeans that the nth derivative is continuous.
2. Partial derivatives . Know thedefinition
of l ' f(Xl,"" Xi +h, ...,x n)- f(Xl' ...,Xi, ...,Xn)
- =In1
OXi h
TocomputeofI OXi inexamples,considerallvari ablesexcept Xi tobeconstantanddifferentiate
by one-variable methods. Differen iation is performed with respect to the variable Xi.
28 CHAPTER 2
3. Notation lor partial derivatives. In many texts, Ix is used for 8f/8x. Ifwe wish to evaluate
ata given point, we write
81
fxl(xo,yo) ' or ~ I if z = f (x,y).
a '
x 1(xo,Yo) (:Co,Yo)
4. Tangent plane. The tangent plane to the graph of a function f(x,y) at (xo,Yo, f(xo,Yo)) is
given by
8f l 81 1
z=f(xo,yo)+ ax (x -xo)+ {) (Y-Yo).
(Xo,Yo) y (Xo ,Yo)
This equation is also used to compute linear approxi mat ion . Compare this equation to the
equation ofa tangent line and the linear approximationin the one-variable case. See section
2.6for a gener lization.
5. Differ ntiability. Equat ion (2) in the text tells you that f :R
2
-+ is differentiable if t he
tangent plane approaches f(xo,yo)as (x,y)approaches (xo,yo). Now, if f :U C ]Rn -+]Rffl,
then
I
Il f(x) - I(xo)- T(x- xo)11 0
1m - (2)
X--+Xo Il x - xoll - ,
where T : ]Rn -+ ]Rm is t he derivative. You should be able t o get equat ion (2) from this
definition. This definit ion ofdifferentiab.ility is mostimportantfor theortical work.
6. Gradient. The gradient is a vector whose components are the partial derivatives of f , with
8f/8x; in the ,t h position. Here, fis a real-valuedfunction. Thi operation is denot ed by the
symbol 'V. Someti, es,itis denoted "grad." For a function I :]R3 -+JR. ,youshoul d remember
the formula
7. Derivative ofvector-valuedfu.nctions. Con ider a function I :Rn-+]Rm. The derivative is an
m x n mat rixofpartialderivativ s. Therange consists ofvectors with m components. Think
of t he componentsas real-valued vector functions;th n each rowof the derivat ive matrix is a
gradient. Thederivative matrixoff ,evaluatedatXo, is denoted Df(xo).
8. Important facts . Differentiability implies conti nui ty of the funct. ion, but continuity does not
implydifferentiability. Theexistence ofcontinuouspartialderivatives implies differentiability
but the converse is not true. Ifa funct ion is differentiable, then its partial derivatives exist ,
but the converse is, again ,not true.
SOLUTIONS TO SELECTED EXERCISES
1. (b) Holding y constant and differentiati ng with respect to x, we get 81/8x =ye
xy
. By
symmetry,8f/ 8y= xe
XY
Inthis problem, all we did to compute fJ!I fJywas to switch x and
y. This is whatwe mean by "symmetry" ;itonly works for functions thatare unchanged when
x and yareswapped.
2. (b) Hold y con taut and use the chain rule to differentiate wi th respect to x. We get
f}z 1 1 1 y x
. '1 1 8z
-- . - 'y - .
SImI ary, 8y
ox - J1+xy 2y'1+xy - 2{1+xy) ' 2(1 +xy) .
At (1,2),8z/ 8x =1/3and 8z/oy=1/6. At (0,0), az/ax =az/8y=o.
3. (b) Holdi ng yconstant and using the quotient rule, we calculate:
8w _ 2x(x
2
- y2)- 2x(x+2+y2)
8x (x
2
_ y2)2
29 DIFFERENTIATION
Holding x constant and differentiating with the quotient rule, we get:
ow _ 2y(x
2
- y2) - (-2y)(x+2+y2) _ 4x
2
y
ay (x
2
- y2)2 - (x
2
- y2)2 .
4. (b) We must show that the partials are continuous in the domain: of/ax =l/y- y/x
2
=
(x
2
- y2)/x
2
y, which iscontinuous for x f. 0and yf. OJaf/ay=_X/y2+l/x:;;:: (y2 - x
2
)/xy2,
which is continuous for x f. O. Thus, f(x) is Cl since its partial derivatives are continuous.
6. (b) Theequation of the tangent plane is given by
Z =Zo+[fx(xo,yo)](x - xo) +[fy(x o, yo)](y - Yo).
Using .the result ofexercise l(b) , we compute:
afl - l' afl =0', f( 0,1)= 1.
ax (0 , 1) - ,
oy (0,1)
Therefore, the tangent plane is z =1+l(x- 0) - O(y - 1) or z - 1+x.
7. (b) The first row contains the partial derivatives of xe
Y
+cosy. The second roW contains
those ofx,and the third row contains t hose ofx +e
Y
The fi rst columncontains the partial
derivatives wit h respect to x, and t he second column contains those wi th respect to y. Thus,
the matrixofpartial derivatives is
Y
eY xe - siny 1
1 0 .
[
e
Y
1
8. (b) The function f is a mapping from JR3to JR2, so the matrix ofthe partials is 2 x 3. Let
It(x, y, z) =x - y, the first component off. Similarly, let Iz(x,y, z) =y+z. Then
ay
Df(x,y ,z) =[ a;:
af l
%1 ] _[1 -1 ()]
olz - 0 1 1 .
aiz alz
ax ay az
xy
11. Using t.he result ofexercise 1(b), x(of/ax )=xye =y(af/ay).
12. (b) Use the linear approximation formula, which is the same as the equation ofthe tangent
plane: z =Zo +[fx(xo, yo)](x - xo) +[f y(xo,yo)](y - Yo). Let z = f(x,y) =x
3
+y3 - 6xy,
IO =1, I = 0.99, Yo =2and y =2. 01. We compute:
8flax=3x
2
- 6y, so fx(1,2)=-9.
af/ay=3y2_6x, so /y(1,2) =6.
Therefore, our linear approximation is z - 3+(-9)( -0.01)+6(0.01) =-2.85. The actual
value is -2.8485.
13. (c) Thegradient is defined as the vector (of/ox ,of/oy,af/az). Thus,
14. (c) The tangent plane is defi ned by \If(xo) .(x- xo) =O. Exercise 13(c) gives us
\l f(1, 0,1) =ei +2ek,
so the tangent plane is e(x - 1) +2e(z - 1) =0 or x+2z =3.
30
CHAPTER 2
17. We compute 'V/ (0,0, 1) =(2:1:, 2y,- 2z )1(0 ,0,1) =(0, 0,- 2) =- 2k.
20. We want to find T in equation (4). By linearity, /(x)- f(xo) =f{x - xo ). Denoting x- Xo
by h ,we want to fi nd T so that
I
II / (h) - Thll -
hl:To Il hll -.
If we choose T =f, t he numerator vanishes for all h, so this T sati ties the condition; that
is, t he deriv tive ofa linear map is t he map itself. For example, in one vari able, consider
f(x) = ax. Fromone-variablecalculus,T =f'(xo) = afor all Xo.
2.4: INTRODUCTION TO PATHS AND CURVES
GOALS
1. Given a path, be abl to compute t he velocity vector.
2. Be able to find a tangent line for a given path.
STUDY HINTS
1. Paths. A pathis a "formula" thatdescribes a curve in space. The pict ure of t he pat h, which
we can draw on paper, is called the imageofthe path or thecurve ofthe path.
2. Path images. Often, it is convenient to express a path in terms ofx and y when you want
to know the image of a path. This is done by eliminating the parameter. For example
t
4 2
(x, y) =(t 2,t4) means t =..;x,so y = =(-Jx)4=x . Caution: In this example, x =t
2
,so
x is always non-negative.
3. Circular functions. Ifa path is parametr ized by an expression involving cost nd sint , the
parametercanusuallybeeli minatedbysquaringandadding. Use theidentitycos
2
t +sin
2
t =1
and 0 her trigonometricidentities.
4. Velocity. The velocityvector 's componentsare fi rst deri vat ives of thecomponentsofthe path.
Thevelocity vector is t angent to the path.
5. Tangent lines. It is easy to find a tangent line if you remember that a line can be described
as x +tv. The vector x is chosen to be a point on the pat h at to and v is the velocity vector
c'(t o). Thus, the tangent line to a path i
l(t)=c(to)+ (t - to)c'(to).
SOLUTIONS TO SELECTED EXERCISES
1. Fromy= 4cost , we get y/4= cost. Use the fact that
andsubstitution toobtain
1
x
Since 0 t ::; 211" , t h curve is an ellipse wit h y intercepts at
4 and xintercepts at l.
-4
31 DIF FERENTIATION
3. As in example 1, c(t) has t he for m (xo,Yo ,zo) + tv,
where (xo,Yo,zo) =(-1,2,0) and v =(2,1,1). Thus, z
c(t) is a line in R3. Specificall y, it is the line t hrough
(- 1,2,0) withdirection (2,1,1).
2
(-1,2,0)
x
7. Apath'svelocity vector is found by differentiating the individual components. In this case,
d 2 d 3 d )
1" ( t)
(
dt(cos t), dt(3t - t ), dt(t)
(- 2costsint,3- 3t
2
, 1).
12. The tangent vector to a curve c(t) =(x(t), y(t is c'(t) =(x'(t),lI(t. In this case, the
tangent vector is (2t,0).
15. The equat ion for the tangent line is l (t) =c(to) + (t - to)c'(t
o
). Here, to =1 and c'(t) =
(3cos3t,- 3sin3t,5t
3
/
2
). Therefore, the tangent line is
l(t) = c(l)+ (t - l) c'(I) (sin3,co 3,2) + (t - 1)(3cos3,-3sin3,5)
(sin3- 3cos3,cos3+3sin3,-3)+ t (3cos3,- 3sin3,5).
17. First , we need to find the tangent lin at to. We compute c(2) = (4,0,0) and c'(t)
(2t ,3t2 - 4,0) ,so c'(2) =(4,8, 0). T hus, the tangent li ne is l (t) =(4,0,0)+ (t - 2)(4,8, 0).
The position ofthe particle at t l = 3 is 1(3)= (4,0,0)+ (3- 2)(4,8,0) = (8, 8,0).
2.5: PROPERTIES OF THE DERIVATIVE
GOALS
1. Be able tostate t he chain rule.
2. Be able tocomputea partialderivative by using the chain rule.
STUDY HINTS
1. Chain rule. Suppose f is a funct ion of Yl, Y2, ...,Yn and each Yi is a function ofx. Then
df of dYl of dY2 ' of dYn
-=-.-+-. -+... +- . -
dx aYI dx aY2 dx aYn dx '
Notice how each term appears to be dJldx wi t h ay,a.nd dYi "cane ling." However, bewar
that the "sum" on t he right-handside is df/dx, not n t imes df/dx. Also note the different d's:
"a" is for a function ofmany variables, while "d' is for a function ofone variable.
2. Multivariable chain rule. The mult ivar iablechain rule st ates that
D(f0 g)(xo)=::; Df(yo) Dg(xo),
where Yo =g(xo). This is he product oftwo deriv tive matri ces,so any desired partial m y
be obtained by multipli ca ion.
32 CHAPTER 2
3. Chain rule, gradient relationship. Know that iffis r ai-valued and h(t) == f( c(t)), then
dh
dt =\7f(c(t)) .c'(t).
SOLUTIONS TO SELECTED EXERCISES
2. (b) Thefunctionf is differentiable by thesumrule. Its derivative is
af af]
[
az'ay =[1 , 1] .
(f )Thefunction is differentiable by the chain rule. We know thatz2 and y2 are differentiable
by he product ruleand that 1- Z2- y2 is differentiable by thesumrule. Also, thesquareroot
functionisdifferentiable (where itsargumentispositive),so th entirefunctionisdifferentiable.
Itsderivativeis
3. (b) This is a spe ial case ofthe first special caseofthe chain rule:
dh af dx af du af dv
dx = axdz +au . dx +av . dx
af al du af dv
az +au . dx +av . dx
5. (b) First, we compute I(c(t)) =exp(3t
2
t
3
) =exp(3t
5
), so
f' (t) =15t
4
exp(3t
5
).
Next, by the chai n ruie,
dl a1 dx af dy '" '" 2
- =- .- +- . - =ye .6t+:te Y 3t Y .
dt ax dt ay dx
Subst itute x =3t
2
and y =t
3
toget
dl jdt = t
3
exp(3t
5
) 6t+3t
2
exp(3t
5
) 3t
2
= 15t
4
exp(3t6),
which is the sameas we gotfroma direct computation.
6. (b) Taketh derivativesofeach component toget c'(t )=(6t,3t
2
).
9. Substitute u= e"' - Yand v =x - y toget
By thechain rule,
D(f0 9)( X, y) =D/(u,v)D9(X,y).
First, we calculat
Df(u,v)= ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ] =[sec\uu- 1) =;:].
1
When (x, y) =(1 ,I ),we have g(l,1) =(e
1
- , 1- 1) =(1, 0) . Hence,
D/(I,O) =
[
1
2
-1

]
.
33 DIFFERENTIATION
Next , we calculate
1
-1 ]
so Dg(I,1) =
[
1
-1 .
Therefore
D(f0 g)(I,1) == =
Alternatively, we maycalcul ate D(f0 g)(x,y) directly from (f0 g)(x,V).
13. (a) By the chain rule,
dT =V'T(c(t)) .c'(t),
dt
where c(t) =(cost,sint). Differentiate:
Substituting x =cost, Y=sint gives
2
te
sin
V'T(c(t)) (2coste
sin
t - sin
3
t,cos t - 3costsin
2
t),
c'(t) (- sint,cost).
Thus,
dT . / 4 3 ' 2 ')
- =-2sintcoste
Sm
+sin t+cos te
smt
-3cos
dt
(b) Plug in x =cost, Y=sint into the expression for T and get
2
te
sin
T(t) =cos t - costsin
3
t.
Using techniques from one-variablecalculus, we have
dT . . t 3 it 4 2'?
- = -2S1l1 tcosteam + cos teS n + S1l1 t - 3cos t t
,
which is thesame as the answer obtained by the chain rule in part (a).
17. (a) Ify(x) and G are differentiable, then by the chain rule, and the fact that G(x,y(x)) is
constant,
dC =aC .dx +aC .dy =aC +oG .dy =0
dx ax dx ay dx ox oy dx .
Solve for dy/dx: If oC/ay0, then
dy oC/ox
dx - oC/oy'
(b) As in part (a), we differentiate C
I
and G
2
by the chain r ule:
dCI =aC
I
+ aC
I
.ciYI + oC
I
.dY2 =0 and dC
2
=aC
2
+ aC
2
.dYI + oC
2
.dY2 =O.
dx ox aYI dx 8Y2 dx dx ox OYI dx oY2 dx
Assuming YI (x), Y2(x) and Garedifferentiable and
aCt/aYI aC
I
/aY2 11=0 for all x,
oCdOYI 8C2/aY2
I
then we can solve for dyt/dx and dY2/dx. Rewrite the two equations as
oCI dYI oGI dY2 oC
I
-.-+_.-
(1)
oYI dx oY2 dx ox
oG2 dYI oG2 dY2 OC
2
-. - +_.-
(2)
OYI dx aY2 dx ax
34
Mul tiply (1) by OG2 / OYl and (2) by - 8GI/ OYl . Add t h two together to get :
8G
1
8G
2
8G
2
8G
l
--. -+_ . -
dY2 8x 8Yl 8x 8Yl
dx = - 8"G=- l ----; 8G=:- l"- "
8Y2 . 8Yl - 8Y2 . 8Yl
Similarly,
8G
l
8G
2
8G
2
8G
l
- _ ._+- . -
dYl 8x 8Y2 8x OY2
d;' - 8G
l
OG2 8G
2
8G
l
.
-._- -.-
8Yl 8Y2 8Yl 8Y2
18. Begin wit h (x, y, z) =0. L t x =I (y, z), y =g( x , z ) and z =h(x , y); this means'
F(f(y, z), y, z) =0,F(x, g(x, z), z) =andF(x y, h(x , y)) =0. DifferentiatingF(f(y, z) ,
wi th respect to yand z, we get
and
8x
8z +Fz =0.
Similarly,different iating F(x, g(x , z),z) and F(x,y, h(x , y)) withresp ct to x and z and y
z, respecti vely, we get
8y
Fx +Fy 8x =
0,
8y
Fy 8z + Fz
0,
and
8z
0,
Fx + Fz 8x =
8z
Fy +Fz 8y =
0.
Solving (2) for 8x/ 8z, (3)for By/ ox and (6) for oz/ 8y gives
ox
8z
assumingthatnone ofthe partials F
x
, Fy and Fz are O. Mul tiply andget -l.
20. (a) Use the definition ofthe partialderivative:
h(0) 2 _ 0
81 (0,0) =li m 1(0 + h, 0) - 1(0 , 0) =li m h
2
+0
2
=li m Q=O.
8x h-t O h h-tO h h-tO h
The laststep holdssince O/ h = for all h =F O. Similarl y,
O(h)2 _
81 (0,0)=lim 1(0 ,+ h) - 1(0 , 0) =lim 0
2
+ h
2
=lim Q=O.
8y h-tO h h-t O h h-tO h
Therefore, 01/ 8x and 01/ 8y exist at (0,0) and equal O.
(b) If g(t) = (at, btl, then
(at )(bt )2 ab
2
t
3
ab
2
(f 0 g)(t) = (at )2 + (bt )2 =(a2 +b2)t 2 = a2 + b2t,
35 DIFFERENTIATION
so
, ab
2
(J0 g) (t) =a
2
+b
2
.
On the other hand, from part (a) , we have "Vf(O, 0) =(of/ax,of/oy)(O, 0) = (0,0). Aliso, we
compute g'(t) =(a,b), so "Vf(O, 0) .g'(O) =(0,0) .(a, b) =O. This verifies, as required, that
the chain rule does not apply to t his function.
24. The t.erm ow/ ax on the left-hand side means the parti al deri vative ofw(x,y, g(x,y)) with
respect to x, hol ding yconstant. The term ow/ox on the right-hand side means the partial
derivative ofw(x,y ,z ) with respect t o x holding yand zconstant . These two terms are not
equal because different independent variables are held constant. Thus, the reasoni ng is indeed
flawed.
2.6: GRADIENTS AND DIRECTIONAL DERIVATIVES
GOALS
1. Be able to define a directionalderivat ive.
2. Be abl e to compute a directional derivative.
3. Be able toexplain thesignifi cance ofthe gradient.
4. Be able to underst and the relationships among the direct ional derivatives, the gradient , the
tangent plane and levelsets.
STUDYHINTS
1. Important example. Manyex mplesinthis bookusethefactthat"V1' =r/1', wherer =(x,y ,z )
and r= Ilrll = Jx
2
+y 2 +z2. This is derived in exampl e 1. Much time can be saved by
rememberi ng this result .
2. Definition. The directional derivative is defined to be
. f(x+hv) - f(x)
or I
1m h .
! f(x +tv)It=o
11-+0
The directional derivat ive gives a rate ofchange offin the direction ofv at x .
3. Geometri c interpretati on. Suppose v = ai+bj is a unit vector , (xo, YO) is a given point and
z := f(x,y) is a urface. The directional derivative tells us the "slope" ofa curve at (xo, Yo)
in the direction ofv . The curve is formed by the intersection ofthe surface with the pl ane
described by the set of points sv+tk. Ifv is not a unit vector , then the "slope" may be
determined by normalizingv to be a unit vector .
4. Relation to partial derivatives . The partials of/ox, of/oy and of/OZ are t he directional
derivatives offin the directions i ,j and k , respec ively.
5. Computing directional derivatives. To compute t he direct ional derivative offin the direction
ofv , theeasiest formulato use is"Vf(x)v . The direct ionalderivativeis a scalar, nota vector .
The vector v is often chosen to be a unit vector .
6. Gradient pmperties. Recall that "Vf =(af/ox)i +(of/oy) j +(af/oz)k. You should know
that "Vf pointsin thedirection inwhichfis increasingfastest and - \lf pointsin thedirection
offast est decrease. Thegradient offis always orthogonal to a level surface off
7. Tangent plane. In terms of t he gradient , the equati on ofthe tangent plane at Xo is "Vf(xo) .
(x - xo) =O. This generalizes the formul agiven is section 2.3.
36 CHAPTEP
SOLUTIONS TO SELECTED EXERCISES
1. Thedirectionalderivative of/(x) atXoin the direction v is "V/(xo) .v. In thi case,
V' / (1,1,2)(1/ V5,2/ V5, 0) = (z2,3y2, 2xz )I(1 ,1,2)' (l/V5,2/v'5,0)
(4, 3, 4)(1/vf:5,2/vI5,0)
= 1O/v15=2v'5.
2. (b) Given / (x,y) =\n(lx2 + y2 ), we compute
"VI (x,y) =
At (1 , 0) ,V'/ (1 , 0) =(1,0)and the directionalderivative i V' /(1 , 0) v =2/vI5.
3. (b) Given /(x, y, z ) =eX + yz, we calculate"V/(x, y, z) =eXi +zj+ yk and thus V'/(1 ,1,1) =
ei +j + k. Theunitvectorparallelto(1,-1,1)is (1,- 1,1)/..)3. Thus, the directionalderivati
is V' /( 1, 1,1) .(1,- 1,1)/..)3=e/..j3.
4. (c) For a function ofthree variables, I (x, y , z), the tangent plane to the surface is
V'/(xo, Yo , zo) .(x - xo, y - Yo, z- zo) =0.
To use this formula, we need to describe t he surface f(x, y, z) =constant. In this Cal:
f (x,y,z) =xyz =1, so V'/ (x,y,z) =(yz,xz , xy) and at (1,1 1), V'f =(1, ] ,1). Therefor
the desired tangent plane is (1,1, ] ).(x - 1,y - 1,z- 1) = 0, i.e., x +y +z =3.
5. (b) We have z =(cosx)(cosy), so Zx =- sinxeosyand, by symmetry, Zy =- sinycosx. .
(0,rr/ ,0),we compute zX = 0and zy = - 1. Therefore, the equation ofthe tangent planei_
z =Zo + [zx (xo, yo)]( x - xo) + [Zy(xo . yo)]( y - yo ) =0+O(x - 0)- l(y - rr/ 2),
i.e., z +y =7r /2.
6. (c) Given f (x, y, z) =1/(x
2
+ y2 + z2 ) =1/1,2, we have Ix =- 2x/ (x
2
+ y2 + z2)2 =- 2x/r
4
By symmetry, I y =- 2Y/ I and 10 =- 2z/ r4 . Then "VI =(- 2/ r
4
)( xi +yj +zk) =- 2r/r
where r =xi+ yj + zkand r =l x2 + y2 + z2.
7. ( ) The direction offastest increase is along the gradient vector. U ing the result of exerci:
6(c), we get thedirection offastest increase as 1(1,1,1 ) =(- 2/9)(i+j +k).
8. The gradient vector is normal to a surface. Here, we have I(x, y, z) =x
3
y'3 +y - z+2 =_
so Ix =3x2y'3, Iy =3x
3
y2 + 1and /z =-1. At thepoint (0 , 0,2),we compute Ix =0, Iy =
and f. = -1. Therefore, a normalvectoris V' / (O, 0,2)=j - k. Normalize this to get theun
normal: (l/.../2)(j- k).
13. (b) By definition, "VI =(81/ 8x,8// 8y,8f/8z), so "VI =(yzexyz,xzexY"xyexyO). Giv
g(t) = (6t,3t
2
,t
3
) , we differentiate each componen to get g'(t) = (6 ,6t,3t
2
) _ From Seetio
2_5, we know that
(log)' (I ) = V'/(g(I)) .g'( I) = exp(1 8t
6
)(3t
5
,6t
4
,18t
3
) . (6 ,6t ,3t
2
)lt=1
= e
18
(18+36+54) =108e
1S
_
17. Bydefinition,"VI = (Ix , Iy ). Sincef is independent ofy, I y = 0and given that I ( x, y) = g(z
we have Ix = 81/ 8x = g' (x ). Th refore, "VI(x, y) = (g'(x), 0).
37 DIFFERENTIATION
20. The direction in which the alt it ud is increasing most rapidly at the poi nt (x, y) is
\7z(x, y) =(-2ax , - 2by).
At the point (1,1),\7z(l , 1) =-2(a, b) , so the desired direction is - (ai +bj )/Ja
2
+b
2
Ifa
marble were released at (1,1) , it will roll in t he di rection at which the altit ude i decreasing
most rapidly,so the marble will roll down in the direction -\7z, i. ., (ai +bj )/Ja
2
+b
2
.
24. (a) We want to maximize I (c(t)) = (cost )(sin t ). Set thefirst derivative equal to 0: df/dt ::::
-(sint)(sin t ) +(cost )(cos t) = 0, so we get cos
2
t = si n
2
t or t :::: (7r/ 4 +n7r ), where
n =0, 1, 2,.. .. Since .::; t .::; 211' , we only want t =7r/ 4,37r/4,511' / 4,711'/ 4. Evaluat ing at
these points, we get l (c{7r/4)) =l (c(57r/4)) =1/2 and l (c(37r / 4)) :::: l(c(77r/ 4)) :::: - 1/2.
Therefore, the maximumvalueoflalongthecurve c(t ) is1/2and the minimumvalue is - 1/2.
SOLUTIONS TO SELECTED REVIEW EXERCISES FOR CHAPTER 2
1. (a) Since 3x
2
and y2 are non-negati ve, t here is no levelcurve ifz <O. Ifz :::: 0, then t he I vel
curve is the origin. Ifz >0, then t he levelcurve is an ellipse wit h the maj or axi parall el to
the y axis and the minor axisparall el to the x axis. Theellipses get larger as z increase . Put
all ofthese levelcurves together to get an elliptical paraboloid.
2. (c) First, consider the surface x yz :::: O. The surface consists ofthe planes x :::: 0, y :::: 0 and
z :::: O. Nowconsider x yz :::: 1. When z :::: k , a positive const ant, th I vel curve is xy :::: l/ k .
Thus we get a hyperbol a in thefi rst and thirdquadrants wit h asymptotesonthex and yaxes.
The hyperbol as get closer to the origin as z gets larger. Thus, t he surface in the first octant
has t heplane x :::: 0,y :::: 0and z :::: 0as asymptotes. Similarly,there isasurface int heoctant
where z >0, x <0 and y <O. Now, if z <0, then there is a si milar surface in either ofthe
octants where x <0 and y >0 or x >0 and y <0. The level surfaces for xyz :::: c, where c
is a positiveconstant ,consist offour similarsurfa.ces which move further from the origin as c
gets larger. Ifc <0, the level surface is positioned in the other four0 tants.
3. (b) Thefunction takes a pointfrom~ 1 to~ 2 so thesi ze ofthe derivative matrixis 2x1. The
derivative is
D / (x) :::: [ ~ ] .
5. Weneedtoshowthevectornormal tothet angentplaneofI (x , y) atthepoint(xo, Yo , I(xo, Yo) )
is parallel to the vector (xo, Yo , zo). The partialderivatives ofI are:
01 - 1 2 2 / ! -Xo
ox (xo, yo) 9 2x (1 - x - Y )-1 2 :::: 2 2 .
~ (xa,ya) vI- Xo - Yo
8I ( ) -1 2Y(I_X2_y2)-1/2! :::: -Yo
oy xo, Yo 2 ./ 1 2 2 '
(x a,Ya) V - Xo - Yo
So the normal to the tangent plane is
- Xo . -Yo .
. ,1 + J - k.
VI- X6 - Y6 VI- X6 - Y6
Multi ply t he above through by -(1- x5 - y6)1/2. We get (xo, Yo ,/(xo , Yo)) =(xo,Yo, zo).
Geomet ri cally, we are looking at t he sphere: x
2
+y 2 +z2 :::: 1. The v ctors normal to the
tangent planes are precisely the vectors r =(x, y, z). Those vectors that have the direction of
e
p
(see exercise 7(a) in section 1.4) are perpendicul ar to thesphere.
7. (c) The equation ofthe t angent pl ane is
z :::: I (xo, Yo) +[(01 / fJ x)(xo , Yo) ]( x ~ xo) +[(0I /oy)(xo, yo)](y - Yo).
In this case, we have 1(-1, -1) :::: 1; ol / ox :::: y, (of/8x )( - 1, -1) :::: -1; of/oy :::: x ,
(of/oy)( - I , -1) :::: - 1. Therefor , the equation oft he tangent plane is z :::: 1- l (x +1)-
l(y+ 1), i.e., x +y+ z + 1:::: O.
. .
38
CHAPTER.
8. (b) Iff(x , Y, z ) =constant, then theequation oft he tangent plane is "ilf(xo) .(x - xo) =
3
where x =(x ,y,z). In this case, f(x,y,z) =x - 2y3 +z3, so "ilf(x) =(3x
2
,-6y2,3;-
and "ilf(l,1,1) = (3,-5,3). Therefore, the tangent plane is (3,-6,3) .(x - 1,Y- 1,z - 1) =
3x- 5y+3z =0, i.e.,x - 2y +z =O.
11. (b) The strategy here is to find a few "paths," compute the limit along those paths. L-
x =2y; then the limi t as ygoes to 0is
lim FX+YI- r f
3Y
I- v'3
y';-=YI- Y1-;1- .
On theother hand, take the path x =4y. Then the limitas ygoes to 0 is
lim JI x +y1=lim f5
Y
I= f. =/: v'3.
(x, y )-+ (O ,O ) x - y Y13"Y1 V3
Since the limitingvain s depend upon the path taken, the limitdoes not exist .
12. (b) Hold y and z constant and use the hain rule to differentiate with respect t o x. Th
oflax= 10( x+y+z)9. By symmetry,oflay= oflaz= 10(x+y+z )9. The gradient is t
vector (of/ox,of/oy,of/oz) =10(x+y +z)9(i+j +k).
16. We compute
(1, - 2)=2x l =2; (1, - 2) =2yl =- 4.
x ( 1, - 2) Y ( 1,-2 )
Then, the tangent plane is z = f(l,- 2)+(8zlox )I(1, _2)(x - 1) +(azlay)I(1,_2) (y +2),
z =5+2(x-1)- 4(y+2)or2z -4y - z =5. Geometrically,thegradientoff(x,y) =x
2
+
at (1,- 2) is perpendicular to the level urve z = 5. The tangent plane of the graph off
t he plane that contains the line perpendicular to the gradient off at (1,- 2) and lyingin t
horizontal plane z = 5, and the tangent plane has slope .../22 +4
2
=2v5relative to the
plane.
18. (b) Thedirectional derivative in the direction ofvis "ilf(xo) , v/ll v il. We compute"ilf(x) =
(y +z ,x +z ,x +y) , so "ilf(l,1,2)=(3,3,2). Also, we have
Ilvll= ) 10
2
+(- 1)2+22 = .../105.
Thus, the directional derivative in the direction ofv is (3,3,2).(10, - 1,2)I v105=31/M
21. The bug should move in the direction of - "ilT(x,y), since this is the direct ion of fast
decrease. We compute - "ilT(x,y) =- 4xi+8yj and -"ilT(1,2) =4i+16j,so the bug shou
movein the direction 4i+16j .
24. Let u = :z: - y, then
a! du of 01
01 and 81
ou ay
au . dy - AU .
ox
Holding yconstant, we get
8z = (!:.) 81 = (!)01.
ax y ax y au
Holding xconstant and using the quotientrule, we calculate
az = [al . y _ 1(1)] = [_of.y - I].
ay y2 oy y2 au
Therefore,
39 DIFFERENTIATION
27. (ii)
(a) The sum rule tells us that x
2
+ y4 is differenti able. Also, x
2
y2 is di fferentiable by t he
product rule. Finally, x
2
y2 l(x
2
+ y4) is differentiable by the quotient rule since (x , y) 1- (0, 0),
and so x
2
+ y4 i- O. Holding y constant, we get
of _ 2xy2(X
2
+ y4) - 2X(x
2
y2) _ 2x y6
ax (X2+y4)2 (x
2
+y4)2
Holding x constant , we get
of 2x
2
y(x
2
+ y4) _ 4y.3 (X
2
y2) 2x4y - 2x
2
y5
oy (x
2
+ y4)2 (x
2
+ y4)2
At the origi n, we use the definition of the partial derivative:
o
o f (0 0) ::= lim f(O + h, 0) - f(O, 0) ::= lim h
2
- 0 = lim 0 =0,
ox' h-+ Qo h h-+O h h-+O
and similarly,
o
of (0, 0) = lim f(O , O + h) - f(O , 0) = lim = lim 0 =o.
By "-+0 h "-+0 h "-+0
(b) The partials exist at (0,0) and
lim f( x, y) =0,
(x , y)-+(O,o) II(x, y) - (0,0)11
so by the definition of differentiability (equation (2), section 2.3), f is differentiable at (0,0).
The function f is differentiable at all other points because the partials are continuous. Thus f
is differentiable. However, as (x,y) approaches the origi n, aflox and o f loy do not approach
o(take, for example, the path x = y); t hus t he partial derivatives are not continuous .
28. (b) The gradient vector is 'V f =(oflox, ofloy). If (x, y) i- (G, 0), then
2
of . ( 1) 2x y (1)
- =YSID - cos
ox x
2
+y2 (X
2
+y2)2 x
2
+y2'
and by symmetry
of . ( 1) 2xy2 (1)
- = X Sill - cos 1
oy X
2
+y2 (X
2
+y2)2 x
2
+y2
Now, if (x, y) = (0 , 0), then the definition of the partial derivative gives us
of _ . f(O + h, 0) - f(O , 0) _. (0 + h)(O) sin ((0 + + (0)2) I. 0
0
(0 , 0) - hm h - hm h = 1m - =o.
0X
-
"-+ 0 h-+O h-+O h
Similarly, (of l oy)(O, O) =O. Therefore, if (x,y) i- (0,0), then
2
[
. ( 1) y ( [ )] 2x
'Vf(x,y) = YSID - ') cos J
x2 + y2 (x2 + x2 + y2
[
. ( 1) (1)] . 2xy2
+ XSIl1 x 2+ y2 - (x 2+ y2)2 cos . x 2+ y2 J,
and thus 'V f(O, 0) Oi + OJ.
40
CHAPTER _
29. (b) The directional derivative is 'VJ (xo) . v. Here,
aJ =_ x sin(v x2 + y2 ),
ax V x2 +y2
and by symm try,
2
~ ~ vx
ZY
+ y2 sin( v x + y2 ) .
Therefore, 'V J(1, 1) = (( - 1/ V2) sin(V2), (- 1/.;2) sin(.;2)) and the directional derivative is
(( -1/ v2) sin(v2), (- 1/ v2) sin(h )) . (1/ Y2, 1/ 0 ) = - sin (v2).
33. (b) Direct ly, we first compute g(u) = J (h(u) ) = sin
2
31.1 +cos 8u . Then
~ =6(sin 3u)(cos3u) - 8sinSu.
When u = 0, dg/ du = O. By the chai n rule,
dg [ 3cos 3U ] . [ 3 cos 31.1 ]
-d = DJ(x,y) Dh(u) = [2x, 1] 8 ' S = [2sm3u, 1] S ' 8
u - sm u - sm u
= 6(sin 3u)(cos 3u ) - 8 sin Su.
Again, wh n u = 0, dg/ du = O.
35 . The normal to the surface J(x, y, z) = x
2
+ 2y2 + 3z
2
= 6 is
aJ aJ aJ )
(
ax' oy' oz = (2x, 4y, 6z ) = (x, 2y, 3z ).
At (1, 1, 1), the normal is (1,2,3), so he uni t normal is (1,2, 3)/ Jf4, the direc ion offiight . T
veloci y of the particle is the speed times the direction, or 10( 1,2, 3)/J14. The position of
particle at any time can be found by finding the equation of t he line through (1 ,1, 1) with
direction 10( 1,2, 3) / V14, and it is (1, 1, 1)+10t (l , 2, 3)/Jf4. This impli s tha.t x = 1+10t/ K
y = 1 + 20t/ v1A and z = 1 + 30t/ ..;u. At som t ime T, the particle is on he sph
x
2
+ y2 + z2 = 103, which means that (1+ 10T;./f4)
2
+(1 + 20T / ..;u)
2
+( 1+ 30T/J14)2 = 1
Simplifying, we get 3 + (120/ v'14)T + 100T2 - 103 = O. Solving for T using the quadra
formula., and taking t h positive T only, we get T = (-3 + V359)/5V14.
3S. (a) By substitution, Z = (x + y)(x - y) = x
2
- y2, so az/ ax = 2x and az/oy = - 2y.
(b) By the chain rule, we have
az oz aU. oz av ,
ax = au' ax + ov' ox = vl1) + u(l) =v+ u = 2x.
Also, we have
oz az ou az ov
ay = au . ay + av . By = ve l ) + u( -1) = v - u = - 2y.
41. Let u(t) = J(t )g(t) and h(u) = ell. By t he chain rule, we have
dh dh du [ d
J
d
9
] [d
J
d9] .
dt = du . dt =e
U
dtg (t) + J(t) dt = dtg(t) + J(t) dt exp[J(t)g(t)] .
46. The velocity is d fined as the derivative of displacement. Therefore, we want to compu
au/ot = - 6cos(x - 6t ) + 5cos(x + 5t ). When t = 1/3, x = 1 and the velocity is au./at =
-6 cos(-1) +6cos(3). Sine cos (- x) = cos x, the velocity is 6(cos (3) - cos( l )).
41 DIFFERENTIATION
49. (a) As given, P is a function of T and V. We can also write
so T is a function ofP and V. Finally, we can wri te
p = RTV2- a(V - {3)
V2(V - jJ) .
Upon rearranging, we get
3
pv - (PjJ - RT)V2+o:V - 0:{3 =O.
Since this is a cubic equation in V,it is theoretically possible to write Yin terms ofP and T.
Therefore, any two of V, P or T determines the third variable.
(b) From t heequati on for T, we hold Vconstant and get
(aT/aP)=(V- {3)/ R .
From the equationfor P, we hold T const ant and get
(fJP/fJV) =-RT/(V- jJ)2 +2Vo:/ V
4
=- RT/(V- jJ)2 + 20:/V
3
.
Now, hold Pconstant and differentiate the equation for Pby implici t differentiat.ion. We get
aV aV .
_ R(V- jJ) - 7jf(RT) 2V7jf0: _ ~ _ aV [ RT _ 20:]
o- (V - {3)2 + V4 - V- {3 aT (V - jJp V3'
or equivalently,
aV R
aT - (RT 20:) .
(V - jJ) (V _ {3)2 - V3
(c) Using the results ofpart (b), weget
(
aT) (BP) (BV) (V- {3) (-RT 20: ) [ R 1
aP BV fJT = ~ (V - jJ)2 + V3 (V _ jJ) v ~ ~ p _~ ~ ) =-1.
50. (a) Thequestion is askingfor thedirectional deriva ive in the direction ofa unitvector. Here,
our unit vector is (1,1)/v'2. Also V h(x , y) =(-0.00130x,-0.00048y) and Vh( -2,-4) ==
(0.00260,0.00196). Therefore, thedirectionalderi vati veat (-2,-4)inthedirect ionof(1,1)/0
is 'Vh( -2,-4).(1, 1)/v'2=0.00456//2. This means that the height increases (0.00456//2)
miles per horizontal mile traveled.
(b) The direction ofthe steepest upward path is V h( - 2,-4)=(0.00260,0.00196).
TEST FOR CHAPTER 2
1. True or false. If false, explain why.
(a) If a 3-variablefunction !(x ,y, z) has partial derivat ives l x,Iyand Iz at the origin, then
the function is differentiable at t he origin.
(b) A gradient vector for a function finIR2 is parallel tolevel curves off
(c) A function f is continuous at a point Xo if
lim I(x} =I(xo).
X---+Xo
42
CHAPTER_
(d) Ingeneral ,a ontinuous function is differentiable.
(e) Given a function !(w, x, y, z), th directional derivative in the direction of (0,0,1,0)
the sameas 8f18y.
2
2. Let u =x y +z and v =xyz. Also, let x =2a +h, y =ab
2
and z =a +2a
2
b.
let a = d
3
and b =c- d. Compute8(1., v)18(c , d) at(c, d) =(1,1) .
3.Use the li near approxi mation to estimate th distance between the origin and the poi..
(8.01,3.9 1. 04).
4. Let c(t ) be a mapping from ]R to ]R3 and let ! (x, y , z) be a mapping from ]R3 to R. 10::-
h(t ) =!(c(t )). When t =1, we have (z,y, z) =(-1,2,4) , c'( I ) =(3,4,-2),dyldt =1ar
dzl dt = 2. Is the rate ofdecrease fastest in the positive :c di r ction, the negative x directi
the positive 11 direct ion r the negative y direction?
5. Let z = (z+y) 2 - 5x
3
+2ye
X
be the equat ionofa surface in space. Find theequation ft
tangent plane at x =0, y =3.
6. Let
Y.., if x i= 0
g(x, y) = x .
{
o if x = 0
Compute8g18z and 8g 18y att he origin ifthey exist there.
7. Let u and v be functions ofx and V, and let x and y be functions ofs and t. Furthermore.
is known t hat
8(u, v)
8( z,y) = [2 1]
0 - 1]
8(s, t) 0 -1
and
8(z,y)
[ 3 2 .
Calculate8u. 18 sand8v 18t.
8. (a) Evaluatethe followinglimitfor !(w, z,y, z) =w - z2y3z:
lim !(5,2, -1,1)- !(5,2,- 1+h, 1).
h-+O h
(b) Is itpossi ble todefine g(O, 0) so thatg(x , y) =(x
2
+y2) /( x
2
+y) is cont inuous on all
1R2? Explain why or why not .
9. Let u (x, y) =2x +y2 +eX. At (1,1) show that u increases faster in the direction parallel
the x axis t haninth direction parallelto the 11 axis.
10. A recent survey showed that patient satisfaction, s, at a pharmacy depended mai nly up..
three factors - t he waitingtimein minutes (t) , the patient's perceiv d degree ofillness (i)
ascalefrom 0to 10,and t hedollarcost ofthe prescription (c) . Th patient satisfaction ind
is given by S(t, i,c) =(1000 - c) / it
2
.
(a) Inwhatdirectiondoespatientsatisfactionincrease mostrapidlyatthepoint(10,0.5,10C
(b) At the point (10,0.5,100) how fast and in what direction (positive or negative) d
patient satisfaction change for each extraminuteofwaiting time?
(c) The admi nistrators do not want s to decrease by more than 1per unit change oft, i
c. Is this goal met with a price decrease of3/ 5 dollars and an increase in waiti ngtime
4/ 5 minutes? Explain.
..
43
3 HIGHER-ORDER DERIVATIVES;
MAXIMA AND MINIMA
3.1: ITERATED PARTIAL DERIVATIVES
GOALS
1. Be able to compute iterated partial derivat ives.
2. Be abl e to explain when mixed parti als are equal.
STUDYHINTS
1. Iteratedpart ial deri vatives. These arehigh r-orderderivatives,such as second andthirdderiva-
tives. With several variables, higher-order deri vat ives may be taken wi th respect to different
variables. The notation :::y means (:x ) ~ ~ ) , whi ch is also denoted fyx.
2. EquaLity ofmixed parti als. If the second partial deri vatives are cont inuous, t hen an iterated
partial derivative may becomputedin any order.
3. Warning. Note t hat t he theorem on equal ity of mixed partials requires continuous partial
deri vatives. If t hisrequirement is viol a ed,differentordersofdi fferenti ation mayyield di fferent
results.
4. EvaLuating partials at a given point. Always rememberto di fferentiate completely before sub-
sti t ut ing gi ven values. With mixed partials,you may substitute for a variableonly after you
have completed differentiating in that variable.
5. Applications. The heat equation, the wave equation and the potenti al (Laplace) equation are
famous examples ofhow higher-order derivatives occur in nature. There is normally no need
to memorizethese equations in a vector calculus course.
SOLUTIONSTO SELECTED EXERCISES
2. Notethat when wedifferentiatewi t hre pect toeither xor y, e
Z
is "constant ," sothefi rstpartial
derivatives are of/ox =-1/x
2
+e-
Y
and of/oy =- xe-
Y
Finding t he part ial derivatives of
the fi rst part ial derivat ives gives us these second partial derivatives:
7.(b) Rewrite the func ion as
2X2 +7x 2y 2x 7x
z = =-+-
3xy 3y 3 '
provided x #-O and y #- O. W compute:
o z 2 7 fJz 2x fJ 2
z
0(Oz )
o x =3y+ 3; fJy =- 3
y
2; ox2 =ox ox =0;
--
44 (HAPTE
2 2
/)2z a ( oZ ) 2 a (az ) az a z 8 ( az ) 4x
8xay = 8x 8y = - 3y2 = ay ax = ay8x; ay'.! = ay ay = 3y3
Thefunction isnotcontinuous ifeitherx =0ory = O. Hence,thefunct ionis notdifferentia
whenever x =0or y =O.
3
. . a f 8 af oa f a ( Oh ) (0 (8
f
))
11. By defi rutlOn, oxayaz = ax ay az . Let h =az' so axayoz = Ox oy and
8 (8h) ah a ( a
f
) 0 ( a
f
t heorem 1,this is als equal to ay ax . Also,by theorem 1, ax = ox oz =8z ax
and t herefore
o3f a (8 (8
f
) ) oaf
8xayaz = ay 8z 8x =8yozax
15. (a) We are given f(x,y) = xarctan(x/ y), so we compute:
fx arctan + ::' 1 ( \ / = 2xy 2 + arctan .
y x+ y y
x - x _ x
2
fy
=
1+ (x
2
/ y2 ) .--:;J2 = x
2
+ y2
y(x
2
+y2)_2x
2
y 1 1 y3 _ x2y Y
fxx
=
(x
2
+ y2)2 + Y1 + (x
2
/y2 ) - (x
2
+ y2)2+ x
2
+ y2 - (x
2
+ y2) 2
-2x(x
2
+ y2)+ 2x
3
_2xy2
/:ey
fyx = (x2+ y2)2 = (x2 + y2)2
x
2
2x
2
y
fyy 2y= .
=
(x2+ y2)2 (x2 + y2)2
19. We have U
x
=3x
2
- 6xy,so U
xx
= 6x- 6x =O. Also, uy=3x
2
,soU yy =o. Substitutiongi
us 'U
xx
+'U yy =0+0=o. Thus, u(x,y) satisfies Laplace'sequationandso u(x,y) is harmon
20. (b) For u =;c2 + y2,we get U
x
=2x,so U
xx
=2. Also, u
y
=2y, so U
yy
=2. Substitution in
Laplace'sequation gives us'U
x
., +U
y y
=2+ 2:f. 0, so x
2
+ y2 is notharmonic.
(d) For U = 11 + 3x
2
y, we get u., = 6xy, so Uxx =6y. Also, U y =3y2 + 3x
2
, so U yy =
Substitution into Laplace's equation gives us Uxx + U
yy
=6y+ 6y "I 0, so y3 + 3x
2
y is r
harmonic.
23. Given V (x,y, z )= - GmM/Jx2 + y2 + z2 =- GmM/r, we compute:
GmMx
24. (a) If (x, y) "I (0,0), we can compute the first partialderivatives in the usual way:
af = (y(x
2
- y2)+ 2x
2
y)(x
2
+ y2)_ 2x
2
y(x
2
_ y2) _ x4y_ yS + 4x
2
y3
ax (x
2
+ y2)2 (x2 + y2)2
8f (x(x
2
- y2)_ 2xy2)(x
2
+ y2)_ 2xy2(x
2
_ y2) x
5
_ 4x
3
y2 _ xy4
ay (x
2
+ y2 )2 (x2+y2)2
45 HIGHER-ORDER DERIVATIVES; MAXIMA AND MINIMA
(b) To compute the part ial derivatives at (0,0), we need to use t he definition oft he partial
derivati es. First, hold yconstant atand differentiate withrespect tox at Xo =0. atethat
f(O,O) is defined to be 0. Then
h(0)( h
2
- 0
2
)
fj l (0 0) = li m 1(0 +h,0)- 1(0,0) =lim --..:hc::....
2
---' +h,--0,,- 2_ = lim Q= lim0=0.
fjx ' 11-+0 h h-+O h-+O h 11-+0
Similarly,we hold xconstant atand differentiate with respect to y at Yo=0. Then
0(h)(02 - h
2
)
{)I(OO) -I' I(O , O+ h)- I(O,O) _ l' 02 + h
2
, - 1m - 1m =lim Q=0.
fjy 11 -+0 h h-+O h " -+0 h
(c) By definition, :::y =:x~ : ) . First, we use ~ ~ from part (a) and then perform
di ffe rentiation as in part (b) . By definition,
h
5
{)f fj f - 4h3(0)2 - h(0)4
{PI . {)y (O+h,O)- {) y(O,O) (h2 + 02 ) 2 -0 . h5/ h4
-{)(0,0) = hm h =li m __'---_ ..,-!.... =hm-- =l. {) J____
x Y h-+O h -+O h 11 -+0 h
Similarly,
fjf of (0)4h - h
5
+ 4(0)2h
3
0
2
f - (0, + h) - -(0,0) (02 + h2)2 -
--(0,0)= li m {)% fjx =lim - - - - - - ~ - - - -
oyox h-+O h h -+O h
(d) Themixedparti alsarenotequal, which is consistent wi t h thefactthatthesecond partials
are notcontinuous at (0, 0) .
3.2: TAYLOR'S THEOREM
GOALS
1. Be able to write down t he first few terms ofTaylor'sformula for a given function.
STUDYHINTS
3
1. Notat ion. The summationsymbol L: means tosum all possible combinationsof (i, j) with
i ,j = l
i and j rangingfrom1 to 3, i.e., (i,j ) =(1,1),(1 ,2),(1 , 3),(2, 1),(2,2),(2, 3),(3, 1),(3, 2), and
(3,3) . In general , if m indices are summedfrom 1 to n, there will be nm terms; in our case
there are 3
2
= 9 terms.
2. Review. Before continui ng, you may wish toreview Taylor' s for mul a from your one-variable
calculus text. Recall t hat the Taylorseriescan beused toapproximatethe val ues offuncti ons.
3. Taylor 's formula. You should know t he pat tern for the general formula. As a reminder,
Taylor's for mul a is
The second term, which involves the second partial derivatives, will become important in
the coming sections. The term involving the third partials sums up 3
n
terms, so it may be
unreasonable to ask you tocompute all ofthese terms unless n =2.
46 CHAPTER
4. Computing Taylor 's f ormula. Remember that you will n ed t o compute all of the partl
deri vat ives ofthesameorder. Forexample,whencomputingsecond partials,onmustcompu
8
2
f 8
2
f 8
2
f .. 8
2
f 8
2
f (j2f
-82' -82'..., a2 aswellas alloft hemixedpartials8 8 '8 8 ' 8 a ,etc. Notetha
Xl Xn X l X2 X2 X l Xl X3 X
2
we do not need to compute !.l 8
2
! ,if. j, twice because mixed partials are equal (assumin
z
UXiUXj
continuity).
5. Taylor'sformula remainder. Recall thatinone-variablecalcul us, t he remainderis determina.
at somepoint between X o and X o + h. Now the r mainder isdet rmined atsomepointon th
line between t he vectors Xo andXo+ h, where h = (hl'h2,...,h
n
).
SOLUTIONS TO SELECTED EXERCISES
1. Recall thatTaylor'sformul ais a polynomialwhich approximatesafuncti on. Ifour functionis
itselfa polynomi al,then t his functionmust be itsown Taylorseries as wel l. Hence thesecond-
order Taylor formulafor fis f(h1 + h2 ) =(hi + h2)2= ~ + 2hlh2 + h ~ at X o =0, Yo =C
Alternatively, 8f/8x =af/ay =2{x + y) and 8
2
f/8x2 =8
2
f/8x8y =f}2f/8yax =
8
z
// 8y2=2. At the point (0,0), f(x,y) and t he first partials are all 0,and all ofthesecond
part ials are 2. Thus, the Taylor approximationis (1/ 2)(2)(hI+ 2h1hz + h ~ = (hi+ h2)2.
2
5. Here, 1(0,0) =1, f x =ycosxy- ysinxy,fy =xcosxy- xsinxy,In== -y
2
sinxy - y cosxy
2
fxy =fyx =cosxy - x ysinx y- sinxy- xycosx y and fyy=_ X2sinxy- x cosxy. At (0, 0)
we have Ix =fy =f xx =I yy =0and fxy =1. Thus, theTaylorseries is
1 2 2
f(h l ,hz) =1+ Ohl + Ohz + "2(Oh
1
+ 2hlh2 +Oh
z
)=1+ h1hz+ Rz(O,h).
7. (b) For x> 0and x < 0, f(x) = exp(- l/x) is infinitely differentiable. For x = 0,we mustuse
the defi nition ofthe derivat ive:
1'(0) = lim f(x)- f(O) = lim exp( - l / x) .
x--+o+ X x--+o+ X
Let u = l/x. We mustshowthatlim.. --> oo u exp(-u)= O. By I'H6pital 'srule,
u
lim ue- =lim (.!!:.. )=lim ( .-!.. ) =o.
e
U
e
U
u--+00 u--+oo tI--+00
In general,
(n- 1)(X) _ f( n-l)(o)
/
f(n)(o) = li m .
x--+o+ X
I"( x)= .-!.. exp ( -1), f(3)( x)= (-2 + .-!..) exp ( - 1) ,
x
2
X X
3
:1;4 X
so ifwecanshowt hat lim.. --+oo une-
u
=0for all n,then wecanconcl ude thatfisCoo atx :::: 0
as well. Again, use I'Hopit al's rule n times:
. _ . un . n!
lI m une " =hm - =. ..= lim - =O.
u--+oo u--+oo etl u--+oo e"
Thus, f is Coo ,wit h all derivatives equaling at x = 0. Now, 1(0 + h)=exp( -l/h) > 0, but
I f(k)(O) k
f(O)+ f (O)h+ ...+ - k!- h + .. .= o.
Hence, fis not analytic.
HI GHER-ORDER DERIVATIVES;MAXIMA AND MINIMA
EXTREMA OF REAL-VALUED FUNCTIONS
GOALS
1. Be able to find t h critical points ofa real-valued two-variablefun tion.
2. Be able to use the Hessian to classify the critical points ofa functi on as maxima,mini ma, or
saddles .
STUDYHINTS
1. Defini tions. (a) A local or relat ive extremumis a point Xo where ! (xo) is largest or smallest
in a small neighborhood of Xo.
(b) An absoluteextremum is a pointXowhere !(xo)islargestorsmalles on the entire domain
under consideration.
(c) Critical poi nts occu.r where allfirst partial derivatives are zero.
(d) A saddle point. is a critical poinLwhich is not a localextremum.
2. Criti cal point-extremum relatzonshlP. All extrema occur at cri tical points, but not all critical
points are extrema. Criti cal points may also be saddle points.
3. Real-valued f unctions. Note t hat we are comparing function val ues for re I-valued functions,
not vector-val ued functions.
4. Finding extrema. Ifa!/ ox =0and{}!/{}y =0have morethanonesolut ion,each combinat ion
of(x,y) which satisfies t hese conditions must be considered. You must he complete in your
analysis. See example 8.
5. Hessian. This is denoted by H! (xo)(h)anditis equal to thesecond term ofTayl or 'sformula,
. . 1 n [p!
which IS "2 L
X
t
;,j=1 xJ
6. Det er-mining definiteness. To determine definiteness in genet I, we need to know the
det erminants of the diagonal submatrices of the Hessian matrix.
Startingfromtheupperleft-handcorner ,i.e.,all,comput et hedeter-
minants. Iftheyare all positi ve, t hen the Hessian is positive defini te.
Ifa11 < 0 and t he signs alternate, then the Hessian is negative defi-
ni t e. ote that this t est includes t heorem 6.
7. Usefulness ofHessian. At a critical point, thefirst partialderivatives are all zero,so Taylor 's
formul a reduces to
f(xo +h) =! (xo) +Hessian+remainder,
where the remainder is small compared t o the Hessian. Thus , ifH!(xo) is positive definite,
then
!(Xo+h )=! (xo )+Hessian+remainder > ! (xo),
so ! (xo) is a rel ative mi nimum. Similarly, ifH! (xo) is negativedefinit e, then
! (xo +h) =!(xo)+Hessian+remainder<!(xo),
and f(xo) is a relati ve maximum.
8. Classifyi ng a critical point Xo. (a) If H!(xo) is posi ti ve definite, then Xo is a local minimum.
(b) IfH! (xo) is negative definite, t hen Xois a local maxi mum.
(c) IfHf (xo) does not satisfy (a) or (b) and not aU of thesubmatrices are zero, t.hen Xo is a
saddle point.
48 CHAPTER 3
9. Second derivative test. Theorem 6 is a special case ofhow the Hessian is used and is often
used for problem solving. To use this test , one must compute the discriminant
D =((J2f) (()2f) _(fl) 2
()X
2
()y2 ()X()y
If ()2f I ()X
2
> and D > at a critical point , then we have a local minimum. If ()2/1()x
2
<0
and D > 0 at a critical point, then we have a local maxi mum. And if()2/1()x
2
> and D <0
at a crit.ical point, then we have a saddle point.
10. Minimizing distance. Example 8 shows that distances may be mini mized by analyzing d
2
.
This is justified by the chain rul e. By differentiating d
2
, the chain rule gives 2d(adlox) and
2d(adlay), so we again aresolvingadlax=0 and adlay=O. Since d~ 0, the maximum(or
minimum) ofd
2
occurs at thesame points as t he maximum(or mini mum)ofd.
11. Guarant eeing absolute extrema. In lR!', if a domainis closed and bounded, andtiscontinuou5
on the domain, then there exists an absol ute minimumand an absolute maxi mum. All thre
conditionsare neces ary. Thinkabout whathappens if thedomai nis notclosed, ift hedomain
is unbounded or if/ is not continuous. Compare t his to the extreme value theorem ofone-
variable calculus.
12. Locating boundary extrema. In this section, one parametrizes the given boundary ofa region
andthendifferentiates todeterminethelocalminimumandmaximumpoints. Anothermethod
is introduced in the next section.
SOLUTIONS T O SELECTED EXERCISES
3. We compute the partials aflax =2x + 2y and a/lay= 2y + 2x . These partials vanish at
pointssuch t hat x =-v,so these are the critical points. Since f(x,y)=(x+ y)2 ~ 0 for all
(x, y), t.he ext remamust be minima.
5. By the chain rule a/ lox=2xexp(1 + x
2
- y2) and af/ ay =-2yexp(1 + x
2
- y2). Setting
these partials equal to 0, we find t hat (0,0)is theonly crit.ical poi nt . To classify thiscritical
point, recall that the exponential function is monotonic. Looking at the function along the %
2
axis, set y =t o get f(x,0) =exp(l+ x ). Thus, x =is obviously a minimum. Ifwe look
at the y axis , we set x = to get f(O,y) = exp(l- y2). The reader should be convinced that
y =is a maximum. Therefore, we concl ude t hat (0,0) is a saddle point .
9. At (0,Q 1 5 0 s ~ + y2) =l. The cosine funct ion cannot exceed 1, so (0,0) is a maxi mum.
At (v7r/2,V7r/2), cos(x
2
+ y2) = cos(7l') = - 1, so (v;J'i, v;J'i) is a minimum since the
cosine function is never less than - l. Now, consider the critical point (0,ft). We calcul ate
cos(x
2
+y2) =-1,so thiscriticalpointis alsoa minimum. Note: This exercise cannot bedone
by the max-mintest for functions of t wovariables (Theorem 6) because thesecond partialsall
vanish at the critical poi nts andso t. he t est is inconclusive.
10. The partial derivati ves are 0/1ox = inyand af/ ay = 1+ x cosy. Setting t hese equal to O.
we see t hat the critical points are (- 1,2n7r) and (1 ,(2 n+ 1)7r), where n is an integer. Since
0
2
f I ax
2
= 0, wecannot use theorem 6. Consider thecase when x =1. Ourfunction becomes
f(y) =y+siny. Thegraph off(y) shows t hatwhen y = (2n+1)7l', thereis aninflection poi nt
there. Therefore, we conclude that the point (1 ,(2n+ 1)7l') is a saddle point si nce, along t he
line x =1, the cri tical point is neither a maximum nor a minimum. Si mil arly,ifx =-I,th D
f(y) = Y- sinyhas infl ection points at y =2n7l'. Therefore, the cri tical points (-1,2n7r) are
also saddle points.
49 HIGHER-ORDER DERIVATIVES; MAXIMA AND MINIMA
fly) fly)
-11: 11: 2n
y y
fly) = y+siny .try)= y- siny
14. First, we compute of/ax=(ycosxy)/(2+sinxy) and of/oy=(xcosxy)/(2+sinxy). Since
thedenominatoris between 1and 3, we onlyneed tosolve ycosxy=andxcosxy== 0. From
ycosxy =0, we get y =or xy =(2n + 1)11"/2, where n is an integer. From xcosxy =0,
we get the additional solution x = 0. To classify the extrema, we look at f(x,y). Since
-1 $ sinxy :::; 1, we have In(l) :::; f(x,y) :::; In(3). So when xy ::: -311"/2,11"/2, ...,(4n+
1)11"/2, we have f(x,y) =In(3) and these points are local maxima. Similarly, when xy =
-11" /2,311"/3,711"/2,...,(4n+ 3)11"/2, we have f(x,y) =In(l) and we have local minima. Now,
look at (x,y) =(0,0). If x =y, then In(2 +sinx
2
) is a minimumwhen x =y =0. On the
other hand, when x =-y, we have In(2 +sin(-x
2
)) =In(2- sinx
2
), which is a maximum
when x =y =0. Therefore, the point (0,0) is a saddle point .
z
(31t12) 112
x
z=In(2+ sini )
(91tI2) 112 x
z= In(2- sini )
18. We shall find and classify the critical points in terms ofk. First, set the partial derivatives
equal to to find the critical points. We have
of =2x+ky=O and of=2y+kx==0.
ax oy
If k f:. 2,the only critical ponitis (0,0). We shall return to the case when k = 2later . The
second partialderivatives are
02f 02f o2f
ox2=2, oy2 = 2, and oxoy =k .
50 CHAPT ER .
The discriminant is D =4 - k
2
. If k
2
< 4, i.e., - 2 < k < 2, then D > 0 and sine
0
2
f /ox
2
> 0,the second derivati ve test tells us t hat (0,0) is a local mini mum. If k
2
> 4, i.E
k > 2 or k < - 2, t hen D < 0; therefore, (0,0) is a saddle point if k
2
> 4.
For t he case k =2, we have
Thegraph off is a paraboliccylinder t hat opens upward and whose vertex is the line x =-
on the xy Similarly,fo r t he case k =- 2, we get
In this case, the graph off is a paraboli ccylinder that opens upward and whose vertex is th
Line x =y on t he xypl ane. Whether k =2or k =- 2,f attains t he minimumvalue of0at i ;:
vertex. In summary,(0,0) is a local minimumif Ikl < 2 and (0,0) is a saddle poin ifIkl > :.
Minimaoccurat points where x =-yifk =2andthey occuratpoints where x =y ifk =-_
2 2
19. (a) We calculate the parti als: of/ox =-6x( y - x ) - 2x (y - 3x
2
) and o f/oy =(y - x ) -
(y- 3x
2
) . Setting of/oy equalto 0,wege 2y-4x2=0,whichimpliesy =2x
2
Substitutio
r
into of/ ox giv s - 6x( 2x
2
- x
2
) - 2x(2x2 - 3x
2
) = - 6x
3
+ 2x
3
= 0, which implies that x =.
and so y =O. Therefore, (0, 0) is a crit ical point.
(b) Let g(t) = (at,bt ). Then f(g(t)) = (bt - 3a
2
t
2
)(bt - a
2
t
2
) = b
2
t
2
- 4a
2
bt
3
+ 3a
4
t -l
Differen iati on gives f'(g(t)) =2t b
2
-12a
2
bt
2
+ 12a
4
t
3
=t (2b
2
- 12a
2
bt+12a
4
t2 ) whichimpIieo
that t =0 is one oft he soluti ons. Thesecond derivat ive is f" (t) =2b
2
- 24a
2
bt + 36a
4
t
2
and
f"(0)=2b
2
0, independent ofb(or a), except possibly at b=O. For b=0, f(g(t) =
which has a minimumat t =O. Thus, t = 0 is a relative mi nimumof f(t) along anyst raigh
line t hrough t he origi n.
(c) Look at the par abola y =2x
2
. Then f(x,y) =(_ X
2
)(:z; 2)=_x
4
< O. Thus, al ong t his
curve, f(O,O)is a maxi mum.
23. Given a vol ume V, suppose t he di mensions are x x y x z , then V =xyz and t he surface
area is S =2xy+ 2yz + 2x z. We want to mi nimi ze S. Solve for z, we get z =V/xv. Then
S =2x y+ 2V/x+ 2V/y. Taking partials, we get
oS
2y - 2V(1/x
2
) (I
ox
oS
2x- 2V(1/y2) (21
oy
Setting (1) equal to 0, we get y =V/x
2
. Substitute this int o (2) and set it equal to 0:
2x- 2V/(V/x2)2 =0, whi ch is equivalent to x - X4/V =0,or x
3
=V , or x =VI/3. Th n
3
y = V/x
2
= V/ V
2
/ =VI/ 3, and z =V/ x y = V/(V
I
/
3
V
I
/
3
) = VI/ 3. All t hree dimensionsare
equal; hence, the box is a cube.
28. We compute the parti al derivat ive: of/ox =anx"-
I
and af/oy =cnyn-l. Setting these
equal to0,we see that t hecritical point is (0, 0) . Thesecond partialderivatives are0
2
f / ox
2
=
n(n- 1)ax
n
-
2
, 0
2
f/oy2 =n(n- 1) cyn-2 and 0
2
f/8x8y =O. We cannot apply t he second
derivative test since all of the second parti alderivatives vanish at th origin. When nis even.
both xn and yn are positi ve if (x, y) :j:. (0,0); both x" and yn equal 0 only at t h origin. We
concl ude that t he origin is a maximumif(J and care both negative and nis even. The origin
is a minimumifboth aand care positiveand n is even. Forall ot her cases, we have a saddle
point at t he origin since theorigin is t heonly critical point .
:31. We want to find the extrema on the disk. First , check to see ifan xtremaoccurs inside t h
disk. Take the part ials off and set t hem equal to 0:
oj o f
- =2x + y =0 and
ay =2y+ x =o.
ox
51 HI GHER-ORDER DERIVATIVES; MAXIMA AND MINIMA
Thus (x, y) =(0,0) is &. relat ive extremaand f(O, 0) =0. On t he boundary, we let x =cosB,
Y =si nB. Then f(x ,y) =f(B) =1 + sinBcosB. Different iat ion gives f' (B) =C05
2
B-
sin
2
B. Setting f' (B) = 0, we get cos
2
B=sin
2
B, whi ch means that B =11'/ 4,311'/4,511' /4,
or 711'/ 4 on t he interval B 211'. For B =7r / 4, f(x,y) =!(1/ V2,1/V2) =1/ 2+
1/2+ 1/ 2 =3/ 2; for B =37r/4, !(x,Y) =!(- 1/V2, 1/V2) =1/2- 1/2+ 1/2 =1/2;
for B =511" /4, !('I: , y) =!(-1/ V2,- 1/V2) =1/ 2+ 1/2+ 1/ 2 =3/ 2; and for B =71r / 4,
!(x,y) =! (1/ V2,-1/V2) =1/ 2- 1/ 2+1/ 2 =1/2. Therefore, the absolute maxima oc-
cur at (1/ V2,1/V2) and (-1/V2,-1/ V2), with t he maximumvalue of3/ 2,and the absolute
mini mumoccurs at (0, 0), with the mini mumval ue of0.
34. First , it is always wise to check for extremainside the region. We set the part ial derivati v s
qual to 0: 8!/8x =Y =and8!/8y=x =0. Thus (0, 0)is a loc Iextremum,and at (0,0) ,
!(x,y)= 0. However, we have not checked the points on t he boundary. For he linesegment
y =1, we have -1 x 1. On this line segment !(x,y) =1 x = x ,so the minimumfor this
line segment occurs at (-1, 1) and t he maxi mum occurs at (1,1). Similarly,we can analyze
the ot her three line segments. Therefore, at two oft he corners , namely (1,1) and (-1, - 1),
we have !(x ,y) =1. And at the other two corn rs, !(x,y) =-1. So (0, 0) is not an absolute
extremum;t he maxi maareat(1, 1) and (-1, -1) andthemi nimaoccur at(1, -1) and (- 1,1).
39. Suppose \7
2
u= and uachievesits mini mumonD\8D. Let un(z,y) = u(x,y)-(l/n)e"',then
\7
2
u
n
= -( l/n)e
X
<0. Thus, Un is stri ctly superharmonic, and by Exercise 37, can have a
mini mumonlyon8D, ay, a.t Pn =(Xn,Yn ). Wehaveun (xn,Yn)=U(XY' Yn)- (l/ n) exp(xn)
U(X
n
,Yn)-e
1
/n,sinceall Xn 10n8D. If (xo, Yo) isapointinD,thenun( xO,Yo) > un (xn, Yn),
or
e e
u(xo,Yo) - - > u(xn ,Yn)- - .
n n
Estimating the left-handside upward (since el n>0),we have
e
u( xo, YO )> u(Xn,Yn)- - .
n
Since D is closed and bounded and u(x,y)is conti nuous, there must be a point q = (xoo,Yoo)
on 8Dsuch that the above inequal ity holds in an arbitrarilysmall neighborhood ofq. Hence,
u(xo, Yo) u(xoo,Yoo) and uhas a minimumon the boundary 8D.
42. First , notice that h = (y/2)tan0 and d = (y/2)secO. Thus, we want to maximize
A =xy+ ( /2)yh= xy+ (y2/4) tanB.
We aregiven P = 2x +y +2d=2x +y +ysecB. NotethatPis
constant. Solving forx,we ge x =(1/ 2)( P - y- YsecB) and
our area function becomes A(y, B)= (Py- y2 - y2secB)/2+
x
(y2 tanB) /4.Takingthe partialderivat ives, we get
8Aloy= P/2- y- yseeB+ (y/2)tan0
and
8A/8B = (_y2/2)sec0tan0+ (y2/4)sec
2
B=(y2/4)secB(- 2tanB+ secB).
Setting 8A/8B = 0, we get y =Oar 2tanB =seeB, i.e., sin O=1/2, i.e., B=11'/6. The
solution y=is impossible for thisgeometric problem. When 0=1r/6,we havesece=2/-va
and tanB=1/-va, so 8AI8y=PI2- y- 2y/-va+y/2-va. Setting 8A/fJ y=0, we get
P ( 2 1 P-va )-1
y=2 1+ -va - 2-va - 2-va- 5 .
Therefore, the maximumareais
52 CHAPTER
P y - y2 _ y2sece y2 tane
2 + 4
where
P..j3 7r
and B= 6'
y ;= 2V3- 5
3.4: CONSTRAINED EXTREMA AND LAGRANGE MULTIPLIERS
GOALS
1. Be able tofind theextremaifone or more constrai nts aregiven.
2. Be able to find mini mumand maxi mumdist ances in geometric problems.
3. Be able to an lyze thecritical poi nts ofa function with oneconstraint by using t he border _
Hessian.
4. For economicsproblems, be able to explai n thesignificance ofA.
STUDY HINTS
1. Notation. The not ation fl Smeans "restricted to. " For example, if g(x,y) =x+ y, th
gl(x =2) means we want to consider the function 9 =2+y.
2. Method for fi nding constrained extrema. Note that t here is an alternative method which
equations (3) rat her than (2). First , rewrite the constraint so that the right-hand side .
zero; for example, x+ y =2 becomes x+ y - 2 =O. Then consider the function h(x,>. )=
f( x) +>. .(constraint). Solve8h/8x; = 0 and 8h/8A =o. In example2,w analyze h(x,>.):::;
2
x - y 2 +>. (x
2
+y2 _ 1) and in example3,h(x,A) =x+Z +A(X
2
+y 2 +z2 - 1).
3. Solving equati ons. In general , we are not interested in the value of >.; only the values oftht:"
variables. Sometimes,solving for Ain terms of the variables, thus eliminating>., is the righ
thing to do.
4. Cautions. Remember ,all ofyour equations mustbe solved simultaneously. Solvingoneequa-
tion alone does not complete the jobof,finding an ext remum.
5. Generalization. If thereismorethanoneconstraint, then "Yf(xo)=>'1"Yg1(xo)+>'2"Yg2(XO) -
... + >'n"Ygn(XO)' The right-hand sides ofequations (2) will have theform t>'i g;i,in plact'
i = 1 J
of>.. : :'and there will be extra const rai nt equat ions . If you prefer to use equat ions (3), Ie
J
h(x, >. )= f(x)+I: Ajgj(X) and solve 8h/8xj = 0 and 8h/ 8>., = 0simultaneously.
6. Bordet'ed Hessian. This is a Hessian wit h a "border ," which are the additional top row and
the addi tional 1ft column. The entries from left to right or top to bottom are 0, - 8g/axl
-8g/8X2, ...,- 8g/8x
n
, where 9is t he const raint . Note that all entries except for the border
are second parti al derivatives.
7. Classifying critical poi nts Xo . (a) If the k X k submatrices ofthe bordered Hessian are aL
negativefor k 3,then Xo is a local minimum.
(b) If thesigns alternate: posi tive, negative, positive,.. .,starting with the 3x 3matrix, then
Xo is a localmaximum.
(c) If t he pattern does not sat isfy (a) or (b) , and the submatrices are not all zero, then Xo is
ata saddle point.
HI GHER-ORDERDERIVATIVES; MAXIMAAND MINIMA 53
8. Extrema on a regi on. The method of Lagrange multipliers is only good for locating extrema
on a boundary. Don'tforget to analyze t hecritical points inside the region.
9. Geometry. Asinexample8, section 3.3, we can analyze thesquare ofthedistance rather than
the di stance itself.
10. Economics. Isoquantsarecurves showing all possible combinationsofcapitaland labor which
produce the sam output.Int hese examples,). tellsyou how much morecanbe produced with
one extraunitof capital orlabor. Note that). has significanceonly at the opti malpoint.
SOLUTIONS TO SELECTED EXERCISES
1. Use the methodofLagrange multipliers. We have \If(x,y, z) = (1 ,- 1, 1),and the const raint
is g(x ,y,z) = x
2
+y2 +z2 - 2, so )'\lg(x,y, z) = ).(2x,2y, 2z). Thus, Vf = >'Vg gives
us 1 = )'2x, -1 = )'2y and 1 = )'2z. So we have x = z = - y = 1/2),. Substitute this
into the constr aint: {1/2>. )2 +(-1/2). )2 +(1/2). )2 = 3/4).2 = 2, or ). = (1/2) J3/2. For
). =+(1/2)J3/2, we have (x,y,z) =(J2/3, - J2/3,J2/3) ,and for)' =-(1/2)J3/2,we
have (x,y,z) = (- J2/3,.j2f3,- J2/3) .These are the twoextremepoints andt he maximum
is v'6,while the mi ni mumvalue is - v'6.
3. We want to fi nd t he extrema of f(x,y) = x subj ect to x
2
+2y2 = 3. Use the method of
Lagrange multipliers. From the constraint , let g(x,y) = x
2
+2y2- 3, so Vf = (1,0) and
Vg = (2x,4y). We wantt o si multaneously solve Vf= ).Vg and the constrai nt equation:
>.2x (1)
My (2)
3. (3)
From (2), we get y = O. From (1), x = 1/2), . Substituting for x and y in (3) gives us
(1/ 2). )2 =3,so 1/2>. = J3;therefore, x =v'3. At (v'3,0), f( x,y) = v'3 and at (-v'3,0),
f(x ,y) = - -/3. We concl ude that the maximum occurs at (v'3,0) and the minimum is at
(-y'3,0).
x
2
8. On S, yisrestricted tobecosx ,so f(x,y)= _ y2 =
x
2
- cos
2
x. Applying one-variable methods , we cal-
culate f'(x ) = 2x+2cosxsin:!.'. The derivative van-
ishes when x =- cosxsinx = - (1/2)sin2x. Thisis a
transcendent al equation and can be solved by graphi-
cal methods. The graphs ofy =x and y =- sin2x/2
only intersect at the origi n,so (0, 0) is an extremum.
Since x
2
0 and 0 cos
2
X 1, we conclude that
(0,0) is a minimum.
13. We want tominimize thesurf ce areaofth cylindersubject to t he constraint of the volume.
That is, we want to minimize S(r , h) = 2rrl' h +2rrr2 subject to rrr
2
h = 1000 cm
3
. Use
the met hod of Lagrange multipliers. From the constraint, we get g(r, h) = rrr
2
h - 1000.
We compute th following first partial derivatives: 8S/ 8r = 2rrh +4rrr , 8g/8r =2rrrh,
8S/8h = 2rr r)8g/8h = rrr2 . Now we wanttosolve the followingsystem ofequations:
2rrh +4rrr (1)
2rrr (2)
rr7
2
h 1000. (3)
54 CHAPTER 3
Factor out rrr from (2) t o get 2 =Ar or A =2/r. Substitution into (1) and factoring out 2rr gives
h + 2r =(2/ r)rh =2h , or h = 2r. Substitution into (3) gives rrr
2
h =rr7,2 . 2r =2rrr
3
= 1000,
so r = 10/(2rr)1/3 and h =2r =20/(2rr)1/3. To check that our result sat.isfies the const.raint
we calculate
2 100 20
rrr h = Tt (2rr)2/ 3 (2rr)l/3 =1000.
Therefore, the desired cylinder has height 20/(2rr)1/3 cm and base radius 10/(2rr)1/3 cm.
18. Use the method of Lagrange multipl iers. Let
f(x , y, A) =x + 2y sec B + A( XY + y2 tan B- A).
Then of/ax =1+ AY; of/oy =2 sec B+ A(X+2y tan B) ; and OflaA =xy+ y2 tan B- A. From
af/ox =0, we get A=-1/y; whereas from of/oy =0, we get
A = - 2sec B
x + 2y tan B
Hence, 2y = (x + 2y tan B)/(sec B) , so 2y(1 - sin B) = x cos B. Thus, x = 2y(sec B - tan B).
Substitute this into afI OA =0 to get 2y2(sec B-tan B)+y2 tan B =A. T hen y2 (2 sec B-tan B) =
A, so
2 A Acos ()
y =
2sec B - tan B 2 - sin B .
25. (a) Use the method of Lagrange multipliers on t he auxiliary function h(x, y, A) = x + y2 -
A(2x2 + y2 -1). We compute the following partial derivative :
hx 1 - 4XA == 0 (1)
hy 2y - 2YA = 0 (2)
h)., = _(2x2 + y2 - 1) = O. (3)
From (2) , we get either y = 0 or A = 1. For the case where A = 1, we get x = 1/4 from (1)
and then y =V778 from (3) . W hen y =0, we get x =1/V2 from (3) . Thus, the critical
point.s for the constrained function are located at (1/4, V778) and (1/V2, 0) .
(b) We let the const raint be g( x , y) =2x2 + y2 - 1, so h(x, y, A) = f(x , y) - Ag(X, V). Then
the bordered Hessian (Theorem 10) is
og og
0
ax oy
og
02h
a
2
h
/H/=
ax ox
2
oyox
og
02h 02h
ay oxoy oy2
At (x, y, A) = (1/4, ..ft78, I), we find
0 - 4x -2y
-4x - 4A 0
- 2y 0 2 - 2A
0 -1 -j7fi
-1 -4 0 = 14> 0, /H/=
-Jr72
0 0
SO (01: , y) = (1/4, V778) is a relative maxi mum point. At t he point (x, y, A) =(1/4, -V778,1).
we have
I 0 -1 Vf72
=14 > 0,
IHI =I ~ 2 ~ 4 ~
55
so (x, y) = (1/4, -ftl8) is also a relative maximumpoint. When (x, y) = (1/V2, 0) , equation
HIGHER-ORDER DERIVATIVES; MAXIMA AND MINIMA
(1) tells us that>. =-Ji78,so at the point (x, y, >.) =(1/V2, 0, - 078), we see t hat
-J8
=(2 + \1'2)( -8) < 0. IH,I:o:
I
-J8 V2


2+ V2

Thus, (1/V2,0) is a relative minimum. Similarly, when (x,y) =(-1/V2,0), >. =+Jl78, so
at the point (x,y,>.) =(-1/V2, 0, Ji78), we get
J8

=(2 - V2)(-8) < 0.


IHI = J8 -V2
2-

V2

Thus, (-1/V2,0) is also a relative minimum. Evaluating the function f(x,y) at the critical
points tells us that (1/4,ftl8) are absolute maxima, (-1/V2,0) is an absolute minimum,
and (1/V2, 0) is only a relative minimum.
26. With a hyperbola, there is only a minimumdistance from a point. With a parabola, there
is also only a minimum. There can be no maximumdistance from these geometric figures
because both figures extend to infinity.
31. Let the price oflabor be p and let the price ofcapital be q. We want tooptimize Q given the
constraint S::;;: pL + qK == B. We compute the partialsofQ and S: {)Q/{)K = AaK,,-l L1-",
{)S/ {)K =q, {)Q/ {)L =A(l - a)L-"K", {)S/ {)L =p. Use the methodofLagrange multipliers
to get the following systemofequations:
AaK,,-l L
1
-"
>.q (1)
A(l- a)K" L-" >. p (2)
pL+ qK B. (3)
From (1), we get q =(A/>')aK"-lL
1
-,,, and from (2), we get p =(A/>.)(l - a)K"L- ".
Substitutioninto (3) givesus (A/ >.)(l-a)K" L
1
-" +(A/ >.) aK" L1-" =B or (A/ >.)K" L
1
- " =
B or >. = (A/ B)K" L
1
-". Substitute for >. in (1) and (2):
AaK,,-l L
1
-" (A/ B)K" Li -"q
(4)
A(l - a)K" L-" = (A/ B)K" L
1
-"p. (5)
From (4), we get K =aB/q, and from (5), we get L = (1 - a)B/p. Thus, the point
(K, L) =(a: , (1 -pa)B)
optimizes the profit .
3.5: THE IMPLICIT FUNCTION THEOREM
GOALS
1. Be able todetermine ifan inverse function exists near a point x.
2. If an inverse exists, be able to find a derivative by implicitmethods.
STUDYHINTS
1. Advanced material. Thetheoremspresented inthissectionareusuallyprovedinmoreadvanced
courses. You should be mostconcerned with understanding the statements ofthe theorems .
-------- - - -
56 CHAPTER J
2. Notation. DxF is used in this section. Itis just another notation for " F with respect to x
although as in Chapter 2, DxF may be a matrixif F is vector-valued.
3. Local Theorems. The theoremsintroducedin thissection maynot applyifth rangeor domain
is too large.
4. Special implicitfunction theorem. If o f /oz=F 0,t hen zcan be wri tten in t ermsofxat agiven
point (xo,zo), and the derivat ive ofz = g( x) is
- DxF(x,z)
D ( )
-
_
. 9 x of
a.;- (x,z)
Note that we can differentiate z even though we don ' t have a formulafor z. Don'tforget the
minussign in the derivat ive.
5. Commonly usedformula. When z is a function ofx and y, we get t he formul a
dy oz/ox
dx - oz/oy'
Itlooks almost like division offr actions, except for the mi nus sign. Agai n, don' t forget the
minus sign.
6. General implicit function theorem. In general , z may be a vector. We form a matrix with
the top row being the parti al ofFl with respect to Zj, j =1, .. .,m (almost like gradient)
Similarly, the 0 her rows consist ofth part ials of Fk, k =I,... m. Ifthe determinant ofthi
m x m matrixis non-zero, t hen z is a funct ion ofx and a derivative exists.
7. Jacobian. This is thede erminantoft he mxm matrixdescri bed in item6above. Itisdenot d
8. Inverse functio n theorem. As stated in it em6 above, ifJf(xo) =F 0, then z can be written in
terms ofx. Itmay not be easy to express z in terms ofx , but it is possible in principle.
g. Example 3. This is a t ypical problem. St udy it carefull y. Note that when more t han one
function is given, you wi ll n ed to solve a system ofsimul taneousequations to find a parti al
derivative.
SOLUTIONS TO SELECTED EXERCISES
2. Let F( x,y ,z) =xy+z+3xz
5
- 4. Since we want to know if we can solve for zas a funct ion
of (x,y), we need to know t hat of /oz does not vanish near the desired point , so of/oz =
1+15xz4. ear (1,0,I), Fz =16 =F 0, so F =0 is solvable for z as a function of (x,y) .
Therefore,
oz - Fx y+3z
5
and Oz _ - Fy _ x
ax Fz 1+15xz
4
oy Fz 1+15xz
4
.
At (x,y) =(1 , 0), z =1,so oz/ox =- 3/16 and oz/oy=-1/16.
7. Let Fl = y+ x +uv and F2=uxy+v. Then we want
OFI/OU oFI/ ov I
= oFdou oF /ov =F 0 at (x,y ,u,v) =(0,0,0,0).
2
I
The entries ofthe determinant are aFI/ou =v, aFi/ ov=u, o F
2
/ o u =xy and oF
2
/8v =1.
We see that at (0,0,0, 0),
HIGHER-ORDER DERI VATI VES; MAXIMA AND MINIMA
so we may not b able to sol ve for tI, v in terms of x, y near (x, y, u , v ) = (0,0 , 0,0) . To check
directly, the first equation gives us ltv = - (y + x ), so v = -(y + x)/u. Combining t his with
2
the second equation, we get uxy = (x + y)/u, or 1.1 = (x + y)/ xy. For (x , y) near (0,0) , either
there is no solut ion for u small, or t here are 2 solutions for u.
10. (a) Using the definition of a(x, y) / o(r, 0) and computing the part ial derivatives, we get
a( x, y) _l ax/or ox/ ao I_I cosO -r sinS [_
8(r, O) - ay/or oy/oO - sinO rcos O - r .
At ( 1' 0 , (
0
) , r = 1'0 .
(b) By the inverse function t heorem, we can form a smooth inverse fu ncti on (r( x, y) , O(x,y))
as long as r I- 0. As a direct check, sol ve for r and 0 in terms of x and y: Jx
2
+ y2 = r
and 0 = arctan(y/x) . Since we have written r and () in terms of x and y, t he result above
is confirmed. Note t hat if x = 0, t hen 0 = 7r/ 2 or 37r / 2, depending on the sign of y. If, in
addition, y =0, then r =0, a.nd B can have any value, so we cannot fi nd an inverse, as we did
above.
12. Let Fl = xy2+XZU +yv
2
-3 and F2 = u
3
yz+2xv- u
2
V
2
- 2. T hen oFt/f)u = xz , oFt/ov = 2yv,
oF2/ol.l = 3u
2
yz - 2uv
2
, oF2/ov = 2x - 2u
2
v. At (x,y, z ) = (1,1,1) and (1.1, v) = (1, 1),
Since L\. I- 0, it is possible to solve for tJ, tJ in terms of x, y, z near the given point. To compute
ov/ ay, we use the chain rule:
and
oF2 201.1 3 OV AU 2 av 2
- =3u - yz + u z + 2x - - 2u- v - 2v - u =O.
oy oy oy oy oy
At (x , y, z) = (1,1, 1) and (u, v) = (1, I) , those equations become 2+ ou/ay+ 1+2(ov/oy) = 0
and 3(8u/oy) + 1 + 2(ov/oy) - 2(01l./oy) - 2(av/oy) = 0, or au/oy + 2(ov/oy) = - 3 and
ou/oy =- 1, so ov/8y =-1.
SOLUTIONS TO SELECTED REVIEW EXERCISES FOR CHAPTER 3
2. (c) Compute the partials of l and set t hem equal to 0:
(1)
2xy + 4y3 = 0. (2)
2x +1i =
From (I ), y2 = - 2x (so x 0). Substitute int o (2) to get y(_y2 ) + 411 = y3 = or y = 0.
T hus, x = also and (0,0) is t he critical point. The reader should verify that t he discri minant
Dis 0, so t his test is inconclusive. However, l(x, y) =x
2
+ 2xy2 + y4. - x y2 = (x + y2 )2 - xy2,
so 1(1: , y) > 0 for all x < 0, and for x positive, (z + y2 )2 < xy2 implies x
2
+y4 < _ z y2 but this
is impossible since Z2 + y4. is always positive and _ xy2 is always negative. We can conclude
that the mini mum value of l(x, y) is 0 at (0, 0).
4. The Taylor expansion is
58 CH APTEP
whereallofthepartialderivativesareevaluatedat (3': 0,Yo). Wecomputeth partialderivati
and evaluate at (xo , yo) =(0,0) as follow:
f (x, y) e"Ycosx; f (O , O) =1,
f ,, (x, y) ye"y cosx - e
XY
sinx ; fx (O,O) =0,
xy
fy(x , y) xe cosz; fy (0,0) = 0,
2 xy
XY
f" x(x, y) y e cosx - 2ye"Y sinz - e cosz; Ix;c(O, 0) =- 1,
f"y (x, y) e"Y cosx + zye
xy
cosx - xe
xy
sinz; f;c y(0, 0) = 1,
fy y(x , y) x
2
exy cosx; f yy (O , O) = 0.
Thus,the second-order Taylorexpansionis
Z2
f (x, y) =1-
2
+ x y.
7. Th criticalpoints are tho e where the first parti al derivatives vanish. Thus, we need toso
of
y1T cos(11' z)= 0,
ax
of
sin(11' x)= o.
oy
The equat ion si n(11' x) =0implies that x is an integer. Since the sine andcosine functions a
never 0 at the same angle, y11' cos(11'x ) = 0 implies that y = O. Thus, the critical points a
(n , 0),where n is an int eg r. Next, wecalculate thesecond parti 1derivatives. They are
02f
ox
2
8
2
f
11' cos (11'z ),
ozoy
f)2 f
oy2
= O.
At the critical points (n ,O), we have fxx (n, 0) =0, f;cy (n, O) = (- 1)"11' and fyy (n ,O) =
Using Theorem 6 from Section 3.3, we compute D = [I,,;c(n, O)][lyy{n , 0)]- (fxy(n , O) P =
o- ((_1) n11')2 =_11'2 . Since D < 0, we conclude that thepoints (n, 0) aresaddle points.
11. Use the method ofLagrange multipliers withg(x,y) =x
2
+ y2 =1. FromVf =>'Vg and t
constraintequation, weget
og
8x>' = 2x>.,
og
8y>' =2y>. ,
l.
If x and yare both nonzero, t hen the first two equations tell us that - sin(x
2
- y2 ) ar
sin(x
2
- y2) must bothequal>., so >. mustbe 0; therefore, x
2
- y2 must be 0also. Notice th
2
x - y2 cannot be a mul tipleof11' because x
2
+ y2 =1meansboth Ix I and Iyl must beless th
2
or equal to 1. From x - y2 =0, Z2 + y2 =1, Ixl 1and Iyl 1, we get four critical point
If x =0 and y # 0, then x
2
+ y2 =1tells us that y =1 andconsequently, >. = sin(- l
Thus, (0, l) arecriti alpoints. Simila.rly,when y =0, z =1and>'=sin(l). Therefor
-------------------------------------------
59
-
HIGHER-ORDER DERIVATIVES; MAXIMA AND MINIMA
(1,0) are also critical points. At the points (,ff, Vt), f has the value 1. At (1,0)
and (0, l) , f has the value cos(I). Thus, f has a maxi mum value of 1 at four points and a
minimum val ue of cost 1) at four other points.
13. By the method of Lagrange multipliers, we get he foll owing system of equations:
y >.
x >.
x+y 1.
Substituti ng into the last quation gives us 2>' = 1, or >. = 1/2. Thus, x = 1/2 and y = 1/ 2.
However, in this case, it would have been much easier to solve for x and substitute to get
z = x(1 - x ) = x - x
2
. By one-variable calculus methods, x = 1/2 is the critical point, and
the maximum value of z subject to x +y = 1 is 1/ 4.
16. (b) By the implicit function theorem,
dy of/ 8x
dx
- of/ 8y '
x
S
In this case, F( x, y) = - sin y + y4 - 4 =O. Thus,
18. (c) The rectangul ar parallelpiped (box) is symmetri ,so if x is a coordinate of a point (x, y, z)
on the corner of the box, then the di mension corresponding to that coordinate has to be 2x.
W want to maximize V (x , y, z) = (2x )( 2y)(2z ) =8xyz subject to x
2
/ a
2
+ y2 /b
2
+ z2/ e2 = 1.
Use the method of Lagrange multipliers:
oV/ 8x =8yz 2x>. / a
2
(1)
8V/oy =8xz 2y>"/ b
2
(2)
8V/ 8z =8xy 2z>. /e
2
(3)
x
2
/ a
2
+ y2/h
2
+ z2 / e
Z
1. (4)
Solve (1) for x: x =4a
2
yz/>.. Substitute into (2): 8(4a
Z
yz/ >.. )z =2y>' /b
2
, or z2 = >. 2/ 16a
2
b
z
.
Substitute for x in (3): 8(4a
2
yz / >.) y = 2z>.. / c
2
, or y2 = >..2 / 16a
2
e
2
. Then
Plug these resul ts into (4) to get
>.2 >.2 >. 2
16a2 b2c2 + 16a:lb2c2 + 16a
2
b
2
c
2
= 1.
i.e., xyz =abe/ 3v'3. Therefore, the maximum volume is
abe 8abe
V = 8xyz =8 . 3V3 = 3V3 '
60 CHAPTER
22. Firstcheckfor extremaon theinteriorofthecircleofradiusJ2. We havef( x ,y) =xy-y+x- _
so set its partial derivatives equal to O. We have
of of
- =y+ 1 and - =x-I ,
ax By
so the point (1,- 1)is an extremum. Since (1) 2+(_1)2 =2, thispoint is withintheconstrain
Tofind ot her extrema (ifany), we use the methodofLagrange multipliers:
y+ 1 2x"\
x-I = 2y..\
x
2
+ y2 = 2.
First consider the case when ..\ =O. Equations (1) and (2) gi ve us (x,y) =(1,-1), which
had considered earlier.
If ..\ :j; 0, then (1) and (2) added together yi elds x+y= 2"\ (x +y)or (2,,\ - 1)(x+y) =
Thus , either ..\ = 1/2 or x =-yo For x =- y, equation (3) gives us the points (1, -1) ~
(- 1,1). If >. = 1/2, then equation (1) reduces to x = Y + 1. Substit ution into (3) yiel
(y + 1)2 + y2 =2or 2y2+ 2y - 1 =0 or y =(-1 -/3) /2. Therefore, two other possib
extrema occur at ((1 + V3)/2,(-1+ V3)/2) and ((1 - -/3)/2,(-1- -/3)/2). Evaluatingat
ofthese possible extrema, we get
f e+2v'3, - 1;v'3)
- '
2'
f (
1- -/3 -1- v'3)
2 ' 2
1
2'
f(l,- 1) 0;
f(-I, I) -4.
Thus, the maxi mumpointsare ((1+v'3) /2,(-1+V3)/2)and ((1- -/3)/2, (-1- -/3)/2) wi
maximumvlaue of 1/2. The minimumpoint i (-1,1) with mi nimum value of-4.
25. We use t he implicitfunction theorem. We need to show that the determi nant
of of
au ov
oe oc
au av
is not 0 near (x,y, 1.1, v) =(2,-1, 2,1). The determinant is
- 31.12 2v I I - 12 2 I
-4u 121.13 = -8 12 = - 144+ 16:j; O.
I
Since the determinant is not 0, 1.1 and vexist as functions ofx and y. To compute ou/ox, "
implicitlydifferenti ate the given equations wit h respect to x. Keep in mind that 1.1 and v ~
functions ofx and y. We get
201.1 OV
2x- 31.1 - + 2v- 0
ox ax
aU 3 0V
2y- 4uax + 12v ax = O.
To make the calculationsi mpler,we can plug in (x,y,1.1, v) =(2, -1,2,1). Then
au av
4- 12ax+ 2ax 0
au ov
-2- 8- + 12- O.
ox ax
61 HIGHER-ORDER DERIVATIVES; MAXIMA AND MINIMA
Solve this simplesystem oftwo equations by your favorite method. You should get au/ox=
13/32.
30. (a) Using the given formula, we write
s;:f(m,b) = (1 - 1.m - b)2 + (3 - 2m- b)2 + (3 - 4m- b)2
= 19 - (46m+ 16b) + ((m+ b)2 + (2m+ b)2 + (4m+ b)2).
The problemis to find m and b which minimize f(m,b), so we take derivatives:
of
_ -46+2(m+b)+4(2m+b)+8(4m+b)
am
= -46+ 42m+ 14b,
and
of
-16+ 2(m+ b) + 2(2m+ b) + 2(4m+ b)
ob
- -16+14m+6b.
Next, we set themequal to 0:
-46+42m+14b 0
-16+14m+6b O.
Solving this system of2equations, we get b= 1/2and m =13/14. The reader is encouraged
to verify that this is indeed a minimumpoint. Therefore, the best-fitting straight line to the
points (I,I), (2,3), (4,3) is Y = 13x/14+ 1/2,as shown in thegraph below.
x
33. If y=mx+ bis the best-fittingstraight line, we must have,in particular,
Performing this differentiation, we get
"
-22)Yi- mXi - b) =0,
.=1
which impliesthat the summationhas to be 0, or the positive and negativedeviations cancel.
TEST FORCHAPTER3
1. True offalse. Iffalse, explain why.
(a) If yis adifferentiablefunction ofx, zis adifferentiablefunction ofxandY, andoz/oyi= 0,
then
dy oz/ox
dx - oz/oy'
62 CHAPTER3
(b) The function g(w, x,y, z) =w +xyz+y2has no critical points.
2
(c) For any f(x,y,z), we have o4floxoy oz=o4flozoyoxoy.
(d) A fourt h-order Taylorseries forf (x )y, z )w) =x
3
+y 4 - z2+w- Wz2is exactly thesame
as f it self.
(e) Suppose Dis a closed region on the xyplane. Iff(x y) is cont inuou and has a mini mllI!'
on D, t hen f(x,y) also has a maximum on D.
2. Use Taylor 's theorem tocalculate a fi rst-order and a second-order approxi mationfor xy2z3 a
the point (1. 1,2.03,0.98).
3. Consider the surface described by z =5x
3
y - 2x y2. Discuss the concavity ofthe
cross-section in the plane y =4.
4. Find all ofthe critical points ofh(u,v,w,r) =v
3
+w
3
+u
2
+7"2 - 3v- 12w+4r - 8.
x
5. Let f be a function ofx and yand of lax =3x
2
ye . Which of t h foll owing, if any, can Ix
aflay?
(a) x
3
-2y+cosy
(b) 3x
2
+y-8
(c) 3-e
Y
+x3
6. Find the minimumand maxi mumvalues of x+yon the curve where 2x
2
+y2 =l.
2
7. Find the minimum and maximum values ofx - x+y2 - Yon the followi ngsquares and thei!
interiors:
(a) The square with vertices at (0,0), (1,0), (1,1) and (0,1), but not including the x or
axes.
(b) T he square with ver tices at (0, 0), (2,0), (2,2) and (0,2), but onl y t he border on the
and yaxes is included in the region .
8. Let a particle move in a potentialfield in R2given by V( x,y) =x
2
+4xy+y2 - 2x+4y+E
Find the critical points of Vand classify t hem.
9. Consider the followingsystemofequations :
x
2
u+yv+uv o
2
u +x v+yu k,
where k is a constant. If.T and y may be wri tten in terms ofu and v, compute oxlov at t h
given point and for the given k.
(a) (u, v, x,y) = (1 ,-2,0, 1); k =2.
(b) (u, v, x,y) = (2,1,0,-2); k =-6.
10. FarmerJones owns a golden goose farm and he wants to maximizeproduction of gol deneggs
Heknows thatmoregeese will lay more eggs, but t oo manygeese will inhibitegg laying ast h
geese instinctively will not overpopulate. On t he other hand, too manycoyotes will prey upo
most ofthe gol den geese and too fe w will cause the geese tooverpopulate. FarmerJones h&
determined that t he golden egg production is propor tional t o
E(g , c) =10gexp(-O.lg)- 2cexp(-0.2c),
where 9 is the goose popul ation and cis t he coyote population.
(a) Find the critical poi nts of E.
(b) Classify the critical points ofE.
63
4 VECTOR-VALUED FUNCTIONS
ACCELERATION AND NEWTON'S SECOND LAW
GOALS
1. Given a path, be able tocompute the velocity vector , the acceleration vector and thespeed at
a gi ven point .
2. Be able to use Newton's second law and Kepler 's law.
STUDYHINTS
1. Velocity, acceleration and speed. Reca.ll t hat the velocity vector's components are t he first
derivatives of the components ofthe path with respect to time. Speed is the length ofthe
veloci ty vector. The second derivat ives make up the acceleration. Beware t hat t he derivative
ofspeed is not the accelerat ion . Note t hat speed is a scalar, while velocity and acceleration
are vectors.
2. Differentiation rules. Noticehow thedifferentiationrules for pathsarecomparabletotherules
you learned in your one-variablecalculus course.
3. second law. Thisstates t hatF =rna, where F is forceanda is the acceleration. You
should remember t his.
4. Regular path. If c'{t) 1= 0,t hen c{t) is a regular path and the imagecurve looks smooth.
5. Kepler's law. This law rel at es an orbiting body'speriod toits radius. Vector alculus is used
to derive the equation. It is probably not necessary to remember the equation, but ask your
instructor to besure.
SOLUTIONS TO SELECTED EXERCISES
2. Take the first derivative ofeach component to get the veloci ty vector; that is, c'{t) == v{t) =
(tcost + sint ,- tsint + cost,via). The accelerati on is composed ofthe second derivatives, so
c"{t) =a(t ) =(-tsin t + 2cost , - tcost - 2sint,0). Therefore, v(O) =(O,I,vIa), a(O) =
(2,0,0) and t he tangent line is 1(>..) = c(O) + >"v(O) = 0+ >"(0, 1,via) =>"(0, 1,via).
5. By the sum rule,we mayadd vector functions first and th n differentiate,or we may differen-
tiate first and then add t he derivatives. By addingfirst,
3
! [Cl (t)+ C2( t)] (e
t
+ e- t ,sint+ cost , t - 2t
3
)
(e
t
- e- t,cost - sint,- 3t
2
) .
On the other hand,
d d d
dt[Cl(t) + C2(t)]
dtcdt)+ dt
C2
(t)
(e
t
,cost,3t
2
) + (_e-
t
, - sint,- 6t
2
)
t
(e
t
- e- ,cost- sint,-3t
2
),
which is what we got by adding first and then different iating.
64 CHAPTER
10. We wan F(O) =ma(O), where m =1. The accelerat ion vector is given by a(t) =r l/(t ) -
(- cost ,- 4sin2t),so a (O) =(-1, 0) . Thus, F(O) =- i g-cm/ s
2
=- O.OOli newton.
13. Toshowthatthe speed is constant, we differentiate it withrespect to timeand show that
deriva.tiveis zero. Sincespeed isIIvll = .;v:v, it is more convenient to work with t he squ
of the speed. (Ifth square of t hespeed is constant, then the speed must be aconstant, t
We calculate
d d dv
dtII vll2 =dt(v v) =2v dt =2v . a,
but the acceleration a is perpendicular to the velocity Vi therefore v . a = O. Th
(d/dt)llv(t)W=0,so Il v(t)1I is constant.
17. Integrating each component ofc'(t) = (t,e
t
,t2)gives us c(t) = (t
2
/2+A,e
t
+ B,t
3
/ 3+
where A, B and C are constants. When t =0,we have e(O)=(A, 1+ B, C)= (0,- 5,1),
t
so A = 0,B =-6and C =1. Therefore, the pathis c(t)=(t
2
/2,e - 6 t
3
/ 3+ 1).
19. (b) Let X =2x,so X
2
+ y2 = 1. Recognizing this as a circle,
y
we let X =cost and y = sint, so x =X/ 2 =(cost)/ 2. OUf
path e(t ) is described by (x,y) =((cost )/2,sin t). From the
originalequation,we recognize 4x
2
+ y 2 =1 as an ellipse with
x-intercepts at 1/ 2and y-intercepts at 1.
1/2 x
20. By the cross product rule in the box preceding Example 1 ofthe text ,
d d ~ ~ ]
dt [mc(t) x v(t)] = mdt[c(t) xv(t )] = m dt x v(t )+e(t) x dt .
Butdc/dt =v(t), dv/dt = a(t ) and v(t) xv(t) =0, so the xpression reduces to
m[O+c(t)x a(t)].
If k is a constant,
k(u x w)=kux w=u x kw,
so
m[c(t)xa(t)] = c(t)x matt)= c(t) x F(c(t)).
IfF(e(t )) andc(t) are parallel,their ross product is 0, so t he angul armomentumi con
Thisis t he case ofplanetary motion; since
GmMc(t)
F(c(t)) =
Il c(t)1 13 '
wesee that F(c(t))is a mul tipleofc(t), so it is parallel to c(t).
4.2: ARC LENGTH
GOALS
1. Be abl to compute the arc length ofa givensegment ofa path.
STUDYHINTS
1. Notation. Often 5 is used to denote a path in space, rather than c.
2. Arc length. This isj ust t helengthofa curv . Lengthsmaybe added toget her, so we may
pute t he ar I ngthofcurves which are not differentiable atfinjtely manypoints by sum:
the lengt hs ofthe pieces.
65 VECTOR-VALUED FUNCTIONS
3. Arc length formula . You mayfinditeasier to rememberthatarc lengthis theintegralofspeed
(not veloci ty!) . This makes sense because arc lengthgives the distance traveled . In any case,
you should know that the for mulais
I
tl Itl
L(c) = Il e'(t) 11 dt = J (X'(t))2 + (y'(t))2+ (z'(t))2dt.
to to
4. Integration tricks.
(a) Due to t. he nat ure ofthe arc length formula,you may want to memorize the for mula
This formul a maybe derived by trigonometricsubstituti on ifyou do not wi h to memorize it .
Ask your instructor ifthis andsimilarformulas wi ll be providedon an exam.
(b) Lookforperfect squares. Radicalscan beeliminatedfrom the integrand ifa p rct square
occurs.
5. Positive lengths only. If youcomputea negati ve ora zeroarc lengt h, then you madea mistake.
Arc length is alwayspositive.
6. Riemann sum derivati on. If thederivationdoes not makesense now, you shouldreturn tothis
discussion after Riemann sums are explained more thoroughly in Chapter 5. Underst anding
the derivation of formulasgives much moreinsight into the t h ory.
7. Parametrization warning. If you need t o parametrize a curve, be sure the curve is traversed
once and only once. The orien'at ion is also importan ,and this will b come especially signif-
icant in chapter 7. The bad consequences of an incorrect parametrizat ion are illustrated in
example 1.
SOLUTIONS TO SELECTED EXERCISES
3. We compute e'(t) =(3cos3t,- 3sin3t,3Vt),so
lie' (t)II =[(3cos3t)2+(-3sin3t)2+(3vt) 2P/ 2=';9+9t =3v'l+t.
Thus, t he arc length is
1
1 2 11
3.Jf+t dt =3 - .(1+t)3/2 =2(23/2 - 1) =4h- 2.
o 3 0
6. Givene(t) =(t,t sint ,t cost) , we computee'(t) =(1,sint +tcost,cost-tsint ).Then li e' (t ) II =
[1 +(sint+tcost)2+(cost - t sint )2j1/2 =';2+t
2
. We want t.he arc length on the interval
[0,1t'], so we need to comput e:
fore J2+t2dt.
This may be integrated by he method oftrigonometricsubstitut ion: One would let tanB =
V2t,aq,d t hen performan integr at ioninvolvingsec
3
8. Thisis leftasanexercise. An alternati ve
is to use theformulagiven in the integration tables (See studyhint 4 above.):
i
re 1 Ire
o J2+t2dt = '2 [tJi2+2+2ln(t + Ji2+2)] 0
~ [1t' J1t'2 +2+2In(1i + ~ - 21nh].
66 (HAPTE
9. Given e(t) =(2t ,t
2
,logt), we compute c' (t) =(2, 2t, l/t). Then
Il c'(t)11 =(4 + 4t
2
+ l/t
2
)1/2 =[(4t
2
+ 4t4+ 1)/t2F/2 =(2t
2
+ 1)/t=2t + l/t.
Since c(l ) =(2,1,0) and c(2) =(4,4,log2), we want the arc length on [1,2]' so we n ed
compute
1 2(2t+ ) dt =(t
2
+ 10gt{=3+ log2.
12. (a) Since T(t) .T(t) = II T (t)112 =1, we can differenti ate both sides ofT( t ).T(t) =l. By
product rule for dot products (see the box preceding example 1 in section 4.1) , (d/dt)( T
T(t)) =(d/dt)( l), i.e., 2T(t) . T'(t) =0, which implies thatT(t) .T'(t)=O.
(b) Beginning with T(t) =c'(t)/lle'(t)ll, we differentiate with respect to t, using the qUOl
rule:
'()=!!.. ~ ) =c"(t)llc' (t)1I - c'(t)(d/dt)lle'(t)11
T t dt Ilc'(t)II lie'(t)112 .
Recall that Ilc'(t)112 = c'(t) .c'(t), so
d d 1 e'(t) .c"(t)
-lle'(t)11 =- y'c'(t) .c'(t) =- (e' (t ).e'(t))-1/2(2e' (t).e"(t)) = .
dt dt 2 Ile'(t)11
Substitution yields
, e"(t), c' (t) .e"(t) , e"(t)IIe'(t)112 - (c'(t) .e"(t))e'(t)
T (t) =lIe'(t )1I21Ie (i)11 - lIe'(t)11 3 e (t) = lI e'(tlIl 3
17. (a) From the definiti ons of k and N in exercises 13(b) and 14,
dT '() II' ( )II T' (s)
ds =T s = T s IIT'(s)11 =kN.
The vectors T, N and B have unit lengthand forma right-handed system ofmutuallyort
onal vectors, so we have
T x N =B, N x B =T and B x T =N.
Differentiate, using the product rul efor cross products , to get:
dN d dB dT dB dT
ds=ds(B x T) =ds x T + B x ds =dsx T - ds x B .
Using the fact that dB/ds = -TN (Exercise 15) along with the results derived earher i
exercise and then factoring out a constant from t he cross products, we get
-T(N xT) - k(N x B) =-T( - B) - kT=-kT+ TB.
Finally,
dB d dT dN
ds
= ds(T x N) =dsx N + T x ds=kN x N + T x (- kT + TB)
= kN x N - kT x T +TT xB =-TN
since N x N =TxT =o.
(b) Let w =WIT +W2N+W3B. We shall use the results from part (a) to find scalars
andW3. We have
dT/ds = w x T =wt{T x T)+w2(N x T)+w3(B x T)
= 0 - w2B +W3aN =kN.
67 VECTOR-VALUED FUNCTIONS
Here, we have used the facts that (1) B =T x Nand (2) T, Nand B form a right-handed
systemofmutuallyorthogonalvectors, andsothereexistsscalarsaand bsuchthatB xT =aN
and N x B =bT. Thus,
W2 =0 and aW3 =k.
Similarly,
dN/ds w xN =wdT xN)+w2(N x N) +w3(B x N)
w1B+ 0 - w3bT = - kT + rB.
Thisgives us
Wl =r, bW3:; k and a =b.
Finally,
dB/ds w x B = Wl(T x B) +w2(N x B) +w3(B x B)
wl(T x B)+0+ 0 =-TaN=-rN,
which gives a ::: 1, and this implies that b== 1andW3 =k. Therefore,
w == rT+kB.
4.3: VECTOR FIELDS
GOALS
1. Be able tosketch simple vector fields.
2. Understand the relationship between flow lines and a vector field.
STUDY HINTS
1. Vector field. This is a mapping from ]Rn to ]Rn. Note that the dimensions ofthe spaces are
equal. Each point x in the domainis assigned a vector. To depict a vector field, We draw the
assigned vector originatingfrom the point x.
2. Scalar field. A scalar field differs from a vector field in that each point x in the domain is
assigned a scalar, nota vector. An exampleis the annual rainfall ateach pointon the earth's
surface. The wind velocity at any instantoftimeis an exampleofa vector field.
3. Gradient vs. vector field. All gradient fields are vector fields, but not all vector fields are
gradient fields. For example, the vector field i + xj is not a gradient field because you cannot
find an fsuch that of/ox = 1 and of/oy = x. If a vector field Pi+ Qj is a gradient field
then oP/8y=8Q/ox. Thisstatement comesfrom the fact that {j2 f/oxoy=0
2
f/oyox for a
well-behaved function f.
4. Flow lines. This is the path a par ticle would t ake ifit was free to move along the vectors in
the field. Thinking ofthe vector field as velocity, the flow lines would show displacement . A
formula description ofa flow line can be obtained by integratingeach component ofa vector
field (or solving a system ofdifferential equations). Flow lines c(t) must satisfy the equation
c'(t) = F(c(t)).
68
CHAPTER
SOLUTION S TO SELECTED EXERCISES
1. At each point (x,y) , wesketch the vector (2,2)oriO"i-
natingfrom(x,y). Alternatively,we maydrawasmall
multiple of the vector fi eld. The orientation of the /' /' /' /' /'
vectors mustbe maintainedat45
0
from the positivex
/' /' /' /'
axis , and all oft he vectors must have the same ma.g-
/' /' /' /'
nitude.
/' /' /' /'
/' /' /' /' /'
5. At each point (x,y),we sketch the vector (2y,x)orig-
inating from (x,y). The magnitudeof the vectors in- y
creases as (x,y)moves away from the origin.
10. At each point (x,y), drawa li ttlearrowin thedirect ion (x,- y), thenconnect the littlearr
to get flow lines. Al ternativel y,one can solve the system ofdifferential equations dxldt =
dYI dt =-y and write y in terms of x, then plot th flow li nes y =CI x for various const a
C.Our computer-generated sket.ch is shown below.
x
13. Ifc(t) is aflow lineofF ,then c'(t) =F(c(t)) . Theleft-handsideis c'(t ) =(2e
2t
, l i t , -lil-
2t
t f::. O. On the right-hand side, we have F(c(t)) =(2e
2t
, lit, - 1It
2
) since x =e and z =
Wegot thesameresult for both sides of t he equation,so c(t ) is a fl ow line ofF .
69 VECTOR-VALUED FUNCTIONS
18. From the chain rule, we have
dV(c(t)) =\7V(c(t)) .c'(t).
dt
We are given that c(t) is a flow lineof F = -\7V, so c'(t) = F == -\7V and
d V ~ ~ t ) ) =-c'(t) .c'(t) = -llc
/
(t)112 :::; 0.
Thus, the derivativeofV :::; 0, which shows that V is a decreasing function oft.
20. First , we compute
At each point(x,y), we draw the vector -\7V. Notice thatfar from theorigin,the denomina-
torsof- \7V, (x
2
+y2) 2, will be muchlargerthant he numerators . Thus ,themagnitudeof- \7V
is verysmallatpointsfarfromtheorigin. Afewcomputationsrevealthat - \7V is very largein
magnitudenear theorigin. Forexample,- \7V(0.5, 0) == (4, -4)and-\7V(O.l,0.1) =(50,50).
Asketch of- \7V is shown below at the left.
To sketch the equipotential surface V( x, y) =(x + y)/(x
2
+ y2 ) =1, we rearrange and
complete thesquares toget (x
2
- x + i)+ (y2 - Y + i) =(x - t)2 + (y - t)2 =t. Therefore,
theequipotentialsurface V(x , y) =1is a circleofradius V"f, centered at ~ , t) . This isshown
below at the right .
y
,
, 2
/
/
-2 -1 0 2
x
x
-2 - ,
4.4: DIVERGENCE AND CURL
GOALS
1. Given any vector field, be able to compute its divergence.
2. Given any vector field in 1R
3
, be able tocompute its curl.
3. Be able to explai n t he physicalsignificance ofthe di vergence and the curl.
4. Be able to manipulate expressions involving the cross product , the dot product and the del
operator .
70
CHAPTER4
STUDY HINTS
1. The operation \1 . This operator , called "del ," tells you to assemble the vector of partial
deri vat ives: (%x, %y,% z).
2. Divergence. Note that the divergence is a scalar , not a vector. Know the t wo notations:
\1 .F =di v F. Youshould know t hat thedivergence is a rateofexpansion or, ifit is negative,
compression. The term incompressible means that div F =O.
3. Curl. Note that the curl is a vector , not a scalar. Know the two not ations: \1 x F ::: curl F.
You should know t hat the curl is associat ed wi th rotations and that the term irrotational
means that curl F =O.
4. Vali d space f orcurl. Note t hat curl is a property of]R3. We do not attempt to take the curl
of vectors in dimensions higher t han t hree. The two-dimensional vector xi+xyj is taken to
mean xi+xyj+Ok if a curl is desired.
5. Laplacian. The expression \1
2
f means \1 .(\1f) ,which is a scalar.
6. Theorems. The facts t hat \1 x (\1f ) =0 and V .(\1 x F) =0are useful ;it is nice to commit
t hese tomemory.Ifyou need to know one oft hese facts and you forget t hem, you can always
do t he computation.
7. Formulas in JR3. Note that t he cross product or curl occurs in some ofthe basic identities of
vector analysis in the t able preceding example 15; therefore, it is assumed that the formulas
are used in ]R3 and not in higher dimensions.
8. Basi c identities ofvectoranalysis. The for mulasin the table preceding Example 15 are useful
for developing the theory ofvectors; however, you should not memorize the t able. You can
refer back to the table as needed. Itwill become obvious which formulas are most important
as you refer back t o them frequentl y.
g. Exercise 30. These formul as are referred t o quite frequently in the examples. You should do
thisexercise even ifit is not assigned.
SOLUTIONS TO SELECTED EXERCISES
2. The divergence is \1 .V =tx(yz)+ :y(xz)+ : )xy) =O.
7. Recall t hat flow lines are defi ned by c'(t) =F(c(t)).
y t
In this case, we have c'(t) = (dx /dt,dy/ dt) =(y, O) .
ince dy/dt =0, we know y is some const ant, k
l
.
1

Then dx/dt =y =k
l
, so X =kI t +k
2
, where kl
.=r=;
and k2 are constants. Thus, the flow li nes have the
...
-
form (kIt+k
2
, kJ). Ifkl is positive, fluid flows from
left to right as tincreases , and ifkl is negative,fluid --..
x
flows from right to left, as shown in t he sketch. We
....
computethedivergence: \1.F =8(y)/8x+8(O)/8y=
O. The sketch shows that fluid appears t o be neither
....
-
expanding nor contracting, which is consistent with
....
...
our calculat ions.
11. The divergence is
\1.F =:x(sin (xy)) - :y(cos( x
2
y)) =ycos(xy) +x
2
sin(x
2
y).
71 VECTOR-VALUED FUNCTIONS
14. The curl is
i j k
() a 0
\7xF =i - - j -
ax oy oz oy oz ax oz
yz xz XY
+k(:x(xz) - :y(Yz))
(x - x)i- (y - y)j +(z- z)k =O.
17. Thescalarcurlis thecoefficient ofk when \7 x F iscomputed for a vector field F in lR2. Here,
we have
i j k
a a a
\7 x F = = [:x (cosx) - ;y(sinx) ] k = (-sinx)k.
ox oy oz
SillX cosx o
Therefore, the scalarcurl is - sinx.
22. First, we compute that \7J = (y +z)i+(x +z)j+(y +x)k. Then
i k
J
o o o
\7 x \7J
ax oy az
y+z x +z y +x
+x) - +Z)] i - +x) - +z)] j

oy oz ox . oz
+[:x(x +z) - ;y(y+z)] k
(1 - 1)i - (1 - l)j+(1- 1)k = o.
25. We knowthatthecurl ofagradientis thezero vector. Thus,it suffices toshow that\7 xF i= O.
We have
i j k
000
\7 x F = = (0 - O) i+(0 - O)j +(siny - cosx) k i= o.
ax oy oz
ycosx xsiny 0
28. We refer to the table ofbasic identities ofvector analysis, which precedes Example 15.
(a) From identity (5), div (F +G) = div F +di v G = 'V .F +\7 .G, which sums to zero,
according to the given hypot hesis. Therefore F +G does have zero divergence.
(b) Fromident ity (8), div (F x G) = Gcurl F - FcurlG, which,in general, does not equal
O. As a simple example, let F = xyi - zyk and G = i. We have \7 .F = \7 .G = 0 and
div (F x G) = z.
30. (a) We have \7(l/r) = \7(1/Jx
2
+y2+z2). Begin by findi ng (8/ox)(1/r) or
a ( -2x -x-x 1 )
ox Jx2+y2 +z2 = 2( x
2
+y2 +Z2)3/2 = (x
2
+y2 +z2)3/2 = -;:J'
By symmetry, (%y)(l/r) = _y/r
3
and ({)j oz)(l /r) = -z/r
3
Putting these together,
V'{l/r) = -(xi+yj+zk)/r
3
= - r / r
3
. In general,
\7(rn ) =\7((x
2
+y2+z2)n/2)
CHAPTER 4 72
and
(O/OX)( X2 +y2+Z2r/2) =(n/2)(x
2
+y2 +Z2)n/2-1)2x=nx1'
n
-
2
,
so bysymmetry,
\7(1'n) = nr
n
-
2
(xi+yj+zk)= n1'
n
-
2
r.
Finall y,
\7(logr) =\7(log(J x2 +y2 +z2)),
and
a 1 a x
-(log( Jx
2
+y2 +z2))= .-(J x2 +y2+z2) = 2'
ox J x2+y2+z2 ox l'
so bysymmet ry,\7(logr) =(xi+yj +zk)/1'
2
=r/1'2.
(b)Using the results ofpart (a),
\7
2
(1/1') = \7.\7(l / r )= \7.(-r /1'
3
)
= - [: x C JX2+:2+Z2 )3 ) +:y
+ :z CJ x
2
+ +z2)3) ].
Thefirst partialderivative is
x (x
2
+y2 +z2)3/2 _ +y2 +z2 .2x
:x Cx2 +y2 + z2).3/2) (x
2
+y2 +z2)3
x
2
+y2 +z2 _ 3x
2
(x2 +y2 +z2)5/2 .
By symmet ry,
a( y ) x
2
+y2 +z2_ 3y2
oy (x
2
+y2 +z2)3/2 = (x2 +y2+z2)5/ 2
and
a( z ) x
2
+y2 +z2- 3z
2
OZ (x
2
+y2 +z2)3/2 = (x2 +y2 +z2)5/2 .
Then
\7. (-r)=3(x
2
+y2 +z2) - 3(x
2
+y2+z2) =O.
1' 3 (x2 +y2 +z2)5/2
Simil arly, part (a) tells us that in the general case,
\7
2
(1'n) = \7 .\7(1'n) =\7 .(nr
n
-
2
r) .
We compute
:x x
2
+y2 +z2)n/2-1 .x )=n [( x
2
+y2 +z
2
t/
2
-
1
+2x2 _ 1)(x
2
+y2 +z2)n/2-2] .
Again,by symmetry,
\7 .(n1'
n
-
2
r) n[3(x2+y2 +z
2
t/
2
-
1
+(x
2
+y2 +z2)(n_ 2)(x
2
+y2 +z2r/2-2j
n1'n- z(3 +n- 2)=n(n+1)1'n-2.
(c) The identi ty \7 .(r/r
3
) = 0follows immediatelyfrom part (b) since n = -1 in that case.
For thegeneralcase,use part(b)again tocompute \7.(1'''r). Note thatweonly need todivide
73
the general result by n andchange n - 2to k (Why?) . Then \7.(rkr ) = (k +3)rk = (n +3)rn.
VECTOR-VALUED FUNCTIONS
(d) By a dired computation, we get
i k
J
0 0 8
'\7 x r = = 0.
8x 8y az
x y z
Using the fact that the urI of a gradient is the zero vector, the calculation j ust completed,
and t he generalcasein part (a), we get
\l x (rnr ) = rn ('\7 xr )+\l(rn ) x r = 0+m
n
-
2
(rx r )= o.
32. (a)
1 j k
o 8 o
curl F = \l x F
ox oy oz
3x
2
y x3 +y3 o
k(3x
2
- 3x
2
) =o.
(b) Note that if F = \l I, then 3x
2
y = (%x) /(x, y) and x
3
+11 = (% y)/(x, y). Integrate
each equation:
f( x, y) =J3x
2
y dx = x
3
y+g(y),
where 9 is a function ofy only, and
where h is a function of x only. In both cases, I (x , y) must be the same, so compare both
sides. Ifwe let g(y) =y4 / 4 and h(x) be an arbitrary constant , then f(x,y) =x
3
y + y4 / 4 + C
satisfies \l f = F.
iy
33. (c) Let F( x, y) =e'"cosyi - eX sinyj , where thej component is thereal partofe
x
- and th j
component is the imaginarypart . We calculate div F =(0/ ox)( eX cosy) - (%y)( e
X
siny) =
eX cosy - eX cosy =0and
i j k
o o a
curl F
ox oy 8z
eX cosy -e'"siny 0
k(- ex siny +e" siny) =o.
Thus, F is both incompressi bl and irrotational.
SOLUTIONSTO SELECTED REVIEW EXERCISES FORCHAPTER4
2. The velocity vector is vet) =e'(t) =2ti +(- 2t sin(t
2
))j +4t
3
k. The acceleration vector is
aCt) = e" (t ) = 2i +[-2sin(t
2
) - 4t
2
cos(t
2
)lj +12t
2
k. When t = ."fi, we have v( y'ir) =
2y'1ri+47r."fik and a(y'ir) = 2i+47rj +127rk. Thespeed is the length ofthe velocity vector.
When t =."fi, thespeed is [47r +1611"3]1/
2
=2-/11" + ~ The equationofthe tangent lin IS
2
given by l(t) =e(y'ir) +tv(y'ir) =(r. - 1, - 1,r. ) +t(2."fi, 0,4r.y'ir).
7. We use t he equation F = rna = me". Here, e" = (2,- sint, - cost) . At t = 0, a(O)
(2,0,-1),so F(O) =rn(2, 0,- 1).
74
_ ....a. _____ .
CHAPTER4
2 3 2 5
9_ From x
2
= .;= Zll , we get y = X / and z = X / . Let x = t, so the path of the curve is
c(t) =(t, t
2
/
3
, t
2
/
5
). Here, we have 1 t 4and c' (t ) =(1 ,it-1/3, i t-3/S). The arc length is
4I1C'(t)11 dt = t VI+i t -2/3 +i.t -6/ 5 dt.
1 11 9 25 1
13. We set the given equation equal to t and solve for x, y and z: x- I = 2y +1= 3z +2 = t.
From x -I = t, we get r = t +1. Similarly, 2y +1= t yields y = (t - 1)/2 and 3z +2= t
yields z =(t - 2)/3. Therefore, we get (.7:, y, z) =(t +1,(t - 1)/2,(t- 2)/3) .
15. We want toshow that c'(t) = F(c(t) ). Here, x = 1/(1- t) , Y = 0and z = e
t
/(1- t). We
dx/ dt = 1/(1- t) 2, dy/ dt = 0 and dz/ dt = (2e
t
- t e
t
)/(l- t)2. Thus, dx / dt = .:z:2 , dy/ dt = 0
and dz/dt = z(l +x) = let /(1- t )][l +1/(1- t)l = (2e
t
- tet) / (l - t )2.
18.We computedi v F = \l .F = fx( .:z:2) +/y (y2 ) +f. (z2 ) = 2x +2y +2z . Also,
i j k
888
curlF = \l x F = I - - -
f}.:z: 8u 8z
.:z:2 y2 z2
[
a(2 a 2)] ' [8 2 8 2]. [ {} 2 8 2]
= - Z ) - -(y 1 - - (z ) - - (x ) J+ - (y ) - -(r ) k = o.
ay 8z 8x f}z 8x ay
21. Thedivergence is \l - F =f}y/ f}x +az/ 8y +8x / 8z =O. The curl is
i k
J
a 8 8
= ({}X_{}Z ) i + ( f}y _ {}x ) j + (8Z _ 8
Y
) k = -i-j-k.
\l x F =I {}x
8y {}z {} Z ax (}x 8y ay 8z
y z x
25. We compute
\lJ = i+ j + k =[2.:z: exp(z2)+y2 sin(zy2)]i+[2xysin(xy2)li+Ok.
Next, we compute
I k
J
{}
8 8
\l x \lJ =
8x ay 8z
2x exp(x
2
) +y2 sin(xy2) 2xysin(zy2) 0
[:Y (0) - :z (2xYSi n( X
y2
))] i+[:z(2x exp(z2)+y2 sin(xy2 )) - :z(0)] j
+ [:x(2x
y
sin(x
y2
)) - :y (2xexp(x
2
) + y2Sin(Xy2 ))] k
= (0 - O)i +(0 - O)j+[2y sin(xy2 ) +2xy3 cos(xy2 ) - 2ysin(xy2) - 2xy3 cos(xy2)] k
= o.
Thus, for J(x, y) =exp(x
2
) - Cos(xy2 ),we have\l x \l J =O.
28. (b) Ifsuch J exists, then we know that F =\l1 =({}f/{}x )i +(al/ ay )j +(8f/8z)k. Thue,
e
Z
we have {}1/8x = 2xye
Z
, 01/ oy = x
2
and 81/ oz = +z2. Integrating aI/ox = 2xye
Z
withrespect tox givesus I(x, y, z ) = +A(y, z ), where A is somefunction involvingyand
Z only. (You can check this result by computing (}f/ox. ) Similarly, integrating (}f/ ay with
respect to y gives us l (x, !I , z) = +B (x, z ), where B is some function involving x and z
only. Next, integrating oj/ 8z with respect to zgives us I( x , y, z ) = z2ye
z
+z3/ 3 +D(x , y),
where D is some function involving x and y only. Comparing the three resulte for I(z,y, z ),
we see that A(y, z ) = z3 / 3 +C, B(x , z ) = z3/ 3 +C and D(x , y) = C, where Cis an arbitrary
constant. Therefore, f (x, !I , z ) = +z3 / 3 +C.
75 VECTOR-VALUED FUNCTIONS
32. (a) At the intersection, we know that y =1, so t he intersecting curve has the equation
x
2
+( 1)2+z2 =;3, orx
2
+z
2
=2. Equivalently,we havex
2
/2+z
2
/2=1, or(x/V2)2+(z/V2)2 =
1. Since we know thatcos
2
t+sin
2
t =1, we get x/V2 =cost and z/V2 =sint,or x =V2cost
and z =V2sint. To get one revolution of the intersecting curve, we let t vary from 0 to
2r.. Thus, a parametrization ofthe int ersecting curve is x = V2 cost, Y =1, z =V2sint,
o<t <2r. .
(b) FromSect ion 2.4, recall that an equation for a tangent line is l(t ) =c(to) + (t - to)c'(to).
Here, c(to) = (1, 1,1), where c(t) is the parametrization given in part (a). We want
(V2costo, 1, V2sinto) = (1, 1,1), so t = r. /4. Differentiating each component of the
parametrizat ion, we get cl(t) =(-V2 sin t,O,V2cost) . Thus, the tangent line is
l(t) =(1,1,1)+ (t - r./4)( - 1,0,1).
(c) In par t (b), we computed c'(t), so Illc'(t)11 =V2. Thus, the arc length is
f 2rr
2rr
1
i o Ilc'(t)11 dt = 0 h dt =2hr..
As expected, this is the circumference ofa circle of radius V2.
36. (a) The direction ofthe rotation vector w is the sameas the axis ofrotation, in this case, the
positive z axis, or k direction. vVe're given the magnitude as 4, so w =4k.
,(b) When r = 5 v ' 2 ~ i - j), the velocity is
j k
v =w x r = 0 0 4 =20hi- 20hj.
5V2 -5V2 0
(c) The vector r is the vector from the axis to the point , so r is the vector from (0,0,5) to
(0,5V3,5), or r = .5V3j. For the point (0,.5V3,.5), we get
J k
v =w x r = 0 0 4 =-20V3i.
o .5V3 0
TEST FORCHAPTER 4
1. Trueor false. Iffalse, explain why.
(a) The curl ofany vector field in ]R3 is another vector field.
(b) One revolution ofthe path c(t) = (cos2t,sin2t) has arc lengt h f;rr Ilc'(t)11 dt.
(c) Ifbandc are vector-valued funct ions oft, then d(b x c)/dt =db/dt x c+ dc/ dt x b.
(d) Two particles which have t he same acceleration must travel at the samespeed.
(e) Suppose G(x, y) is the velocity field ofa gas. IfG(x, y) = xi, then the gas is expanding
and ifG(x, y) = yi, then the gas is not expanding.
2. Let c(t) =(2t, t - 3,t
2
+ 1) be a possible fl ow line for a veloci ty vector field. Whi ch ofthe
following, ifany, could be such a velocity vector field?
(a) F(x, y, z) =(2,x - 2y +7,x).
(b) F(x,y,z)=(yiz -l-y+5,1, 2y -6).
3. Suppose the velocity vector fields represent the flow ofa fl uid. At what poi nts in ]R3, ifany,
does the flui d lack rotation?
--
76
-'---- -
CHAPTER4
(a) F( x, y, z) =4xyi +2x
2
j +k.
(b) F(x , y, z ) = xyi +zj+xk.
4. A particle travels counterclockwise along t he ellipse 9x
2
+4y2 =36.
(a) Express the distance traveled going from (2,0) to (0,-3) as an integral of the form
Jva +bcos
2
t dt for constants a and b.
(b) Suppose t he part icle had fl own off on a tangent line at (-l ,3V3/2) . Where will the
particle hi t the x-axis?
5. For a gi ven funct ion,g(t), let a part icle's posi tion be described by 2g(t)i +[g(t)j2j- k.
(a) What force is act ing on t heparticle ifits mass is 3 units?
(b) What conditions must be imposed on g(t) for the force to be O?
6. Let F(x , y, z) =xi+yj +(z2 +y) k be the velocity vector field for a gas. Where,in ~ 3 is the
gas expanding?
7. InR
4
, apathis descri bed by y =2x, z =4x
2
- 6 and w =e
Z
Find the arc length ofthe path
as an integral in y for 0 ~ x ~ 2. Do not evaluate.
8. Which ofthe following, if any, is the curl ofsomevector field G such that F =curl G?
(a) F(x, y, z) =2i+3j- 2k.
(b) F(x, y,z ) =xi + 2j - zk.
(c) F(x , y,z )= yi + yj + x
2
k.
9. (a) Let v(t) = (sin
2
t)i +(t
4
- t
3
)j +3k. What is t hederi vative of v( x
3
- x
2
+3x - 4) with
respect t o x?
(b) Let c(t) be a pat h in ]R3 and let p(t) be a scalar function. Find a formula for the
acceleration ofp(t )c(t) .
10. Att heannualcamelraces,camelA 's positionisgiven by (cos5t, 2sin5t) andcamelB's position
is given by (cos4t , 2sin4t).
(a) Show that t he camels are running on the samepath.
(b) Suppose camel A 's massis 500. What force is being generated by camel A att =7r/5?
(c) Camel B saw an oasis at t = 7r /5. Itran offon a tangent line. What is the equation of
the tangent line?
(d) Suppose camel B had continued to run on the track. How far behind (in distance) was
camel B when camel A had completed one lap? Leave your answer in the form of an
integral.
77
5 DOUBLE AND TRIPLE INTEGRALS
5.1: INTRODUCTION
GOALS

1. Be able to calculate double integrals over rectangles.


2. Be able to use Cavalieri' principle.
STUDYHINTS
1. Some notation. (a) The cartesian product of two intervals in lR2 is a rectangle. Ifa x b
and c y d, then it is denoted [a,b]x [c,d].
(b) Sometimesdxdy or dydx is abbreviated dA, the "differentialofarea."
2. Review. Before continuing, you should re iew integration techniques for one variable. Itis
essential that you remember how to integrate by parts and by subst it ution.
3. Geometric interpretation. Iff(x,y) 0,then thedouble int grallID f(x,y)dA is thevolume
under thesurfacedefined by the graph ofz =f(x, y). Recall that with one variable, If(x)dx
is the area under the curve y = f(x).
4. Computing a double integral. Just as with partial d ri vatives, all but one variable is held
constant in each step. We integrate from the inside and work our way to the outside. For
example,
i d ,b
i
a
b
c f(x,y)dydx = Ja[F(x,d)-F(x,c)]dx,
where Fisanantiderivati veoffwhen xisheld constant. Now,we computethesecond x-integral
using one-variable methods.
5. Cavalieri's principle. This principle is used in most one-variable calculus courses to derive
volumeformulas. These formul as are oft en referred to by names: disk or slice method.
6. Riemann sums. Be aware that in the summation 2:7:=0 f(Ci)(Xi +! - Xi), ci can be chosen
anywherein [Xi ,xi+d . In a Riemannsum, we take the limi t as n -+ 00, so in mostcases, !(c;)
is almostindependent ofCi because Xi+! and Xi eventually getclose together.
SOLUTIONS TO SELECTED EXERCISES
1. (b) First hold x constant and integr ate wi th respect to y, then integrate with respect to x:
r/
2
t
io Jo(ycosx+2)dydx
r/
2
(1 )
1 ) 17f/2 1
io 2' cosx+2 dx =
(
2' sinx+2x 0 ='2 +'fr.
CHAPTER5 78
2. (b) We can change t he order of integration. Thus , we hold y constant and integrate with
respect to xfirst:
i 2
r r/
t 1"'/2]
Jo Jo (y cosx+2) dxdy Jo [(ysinx+2x )x=O dy
1\y+7r) dy = (y;+7rY) I:= +7r.
As expected, we get the same answer regardless ofthe order ofintegration.
3. By Cavalieri ' s principle, the volumeofa solid is
1
b
V = a A(x)dx,

where A(x) is the cross-sectional area cut out by a plane. Note that at the same height, a
cross-sect ion ofthe left-hand side is a circle ofradius 7' and a cross-section ofthe right-hand
side is also a circle of radius r . Since A(x)is equal in both cases, the volumes must be equal.
5. With the setup in figure 5. 1. 12, we slice W vertically by planes to produce triangles Rx of
area A(x) in the fi gure. The base boft he triangle is b=.Jr2 - x
2
and its height is given by
h =btan(j =.Jr2- x
2
tan(j, Thus A(x) = = - x
2
)t an(j. Hence, the volume is
r
3
j
r jT 1
2
1 (
2
x ) I
_rA(x)dx = _ r
2
(r
2
-x)tan(jdx= 2(tan(j) r x -'3 -r
3 3
1 ( 2r ) 2r
2tan(j 2r
3
- 3 =3 tan(j.
8. Note that when y is in the interval [-1,0], then Iyl =-y,and so
2
Jl (Iylcos dydx =
rjO (-YCOS 7r4X) dydx
Jo -1
r
2
[(_y2 7rX)I
O
1
Jo -2- cos 4" y=-l dx
1
2
2 (1
- cos-
7rx)
I dx = -
2.
sm -
7rX 12
o 2 4 7r 4 0 7r
10. Since f(x,y) 0for al l points in R, ffRf( x,y) dydx is the desired volume. Itis
2
11 \1+2x+3
Y
)dYdx = 12 [(Y+2XY + dx
11
12 dx= (x2+ 5
2
X) I:
2
5. 2: THE DOUBLE INTEGRAL OVERA RECTANGLE
GOALS
1. Be able to computea double integral over a rectangular region.
2. Understand Fubi ni 's theorem.
79 DOUBLE AND TRIPLE INTEGRALS
STUDYHINTS
1. Definition. The definition ofthe integral is more important for theoretical rat her than com-
putational work. It is defined to be a limi t of Riemannsums:
n-1
J1!(x,y)dA = L f ( Cjk)
R j,k=O
Although itis not required, it is usuallyconvenient to use a regular partition, i.e., a partit ion
with ualspacings.
2. Properties. Manyoftheproperties thathold forsingle integrals alsohold for doubleintegrals.
Someoftheseincludetheintegrabili tyofanycontinuousoreven piecewise conti nuousfunction.
3. Warning. As in one-variable integrat ion, [fJRfdA] [fJR g dA] =1= JJR!gdA,in general. In
other words, t he product oftwo integrals does not usually equal the integral ofthe product.
4. New terminology. Th nam s oft he proper ties are known as lineari ty, homogeneity, mono-
tonici ty andadditivity. Youshould know thestatements even ifyou don' t know t he names.
5. Fubini's theorem. This tells you t ha.t, for most reasonable funct ions, you can int grate one
variable at a t imeand the or der ofintegrationdoes not matter.
6. Double integrals and volumes. Recall that if!(x,y) 2: 0, t he double integral is simply the
volume of the region between the graph of!(x,y) and the xy plane. If f(x,y) < 0, then we
subtra t the corresponding volumebetween the graph of!(x,y) and t he xy plane.
SOLUTIONSTO SELECTED EXERCISES
1. (b) We compute t he double integral as an iterated integral:
Note that we chose to integrate in x fi rst be ause that integral is simple (integrat ing with
respect to yfirst would require integrationby parts).
2. (b) Thisis the integrated integral
fa 1[ +(by+ C) x)[=al dy
fa1 +by +c) dy = + +c) y JI
a b
"2 +"2 +c.
5. We will use t h fact t hatJcf(x) d:c = cJ f( x )d:t for anyconstant c. If we integrate in xfirst,
then g(y) is hel d constant in the fir t st p, andso
Jk[J(IX )g(y)]dxdy = ldib f(x)g(y) dxdy =ld[i
b
f(x)g(y)d:c1dy
ld[g( y)lb !(x)dx 1dy,
80
y
CHAPTER 5
where J:I( z) dz is a constant in y. Factoringout theconstant integral gives us
[ibI(X)dXl [i
d
g{Y)dyl .
7. The function x
2
+ y is positive over n,so t he double integralrepresents the desired volume.
111\x2 +y)dydx = 11[(YX
2
+y;) [=J =11(x2 dx dx
(
x3 + 3X) 11 = + =
3 2 0 3 2 6
Thus, the volumeis 16
1
.
11. Begin by substituting y danO, so dy xsec
2
OdO and
J x
2
+ y2 = Xsece. We get
J
x2 - Jx Z2tan
2
() y2
2
-
2
" 2)"d
y
= ,> 411 (xsec 0dO)
( + y . .. x'sec u
2 2 2
J
x3(1 - tan B) dB =Jcos 0- sin (J dB
X4sec
2
B z x
J
cos2BdO=sin2B + C.
x 2x
From the triangle, we have sine= y/';'2 and CO! 0= z/Jz2 + y2. Then z-::-
sin2B + C = sinecose+ C= . __y
x +C= Y
2x x x Vx2+ y2 J x2 + y2 X
2
+ y2
+C
'
Evaluating, we get
[I x
2
_ y2 y 11 1
10 (x
2
+ y2 )2 dy = Xl + yl y=0 =x
2
+ l '
Then
1
111x2 -y2 11 1 1 11"
(2 2 ) 2 dydx= =tan-
1
xl
o
=-4 '
o 0 x +y 0 x +1
x
l
For thesecond integral, we substitutez = ytan</>, SO dx =ysee
2
d</> &nd J + y 2 =ysee</>.
This gi ves
J
z 2 - y2
2 2
J
y2 tan </> - y2 ( sec2</>#) =J11(tan - 1) d</>
dx
(x2 + y2 )2 y4 sec
4
Y y4 sec
2
</>
l
-J11(1 - tan </ d</>.
y4sec
2

Using the method above, t his becomes - x /{x


2
+ y2)+ C. Evaluating,we get
1
1 X2 - y2 - x 11 - 1

o (xl+ y2)2 - x
2
+ y2 .: =0 - 1+ y2 .
Finally,
111x2 11 - y2 -1 1 11"
( 2 2)2 dxdy= - -2dy =- tan-1 yI0 = - -4.
o 0 x +y 0 1+y 1
Fubini'stheorem does not apply in this case because (x
2
- y2)/(Z2 + y2)2 is not bounded on
the entiredomain,namely,at (0, 0) .
81 DOUBLE AND TRIPLE INTEGRALS
5.3: THE DOUBLE INTEGRAL OVER MORE GENERAL R EGI ONS
GOA LS
l. Be able to compute a double integral over an elementary region on the xyplane.
STUDY HINTS
l. Region types. A y-simple regi on is bounded by the lines x =a and x =band two curves which
are functions of x, for a x b. Such a region is called y-simple because the curves that
form the boundary can be described by yas a "simple' function of x. Similarly, an x-si mple
region is bounded by the lines y = c and y := d and two curves which are functions of y, for
c y d. Such a region is called x-simple because the curves that form the boundary can be
described by x as a "simple" function of y. A simple region may be classified as both y-simple
and x-simple regions. See figure 5.3.3 of the text .
2. Classifying regions. For purposes of integrating, it is more important to be able to recognize a
region type rather than being able to name it. Your primary concern should be learning how
to perform double integration.
3. Simplifying complicated regions. Most plane regions may be broken up into regions each of
which is y-simple or x-simple. For example, the region at the
left is divided into six peices, each of which is either y-si mple
or x-simple. In fact, many of the subregions are both y-simple
andx-simple.
4. How to integrate. Itis best to use the iterated integral J: f(x,y)dydx for y-si mple
regions. For x-simple regions, it is best to use the iterated integral f(x,y) dxdy.
5. Choosing integration limits. When you perform multiple integration, be sure that the limits
of integration do not include any previously integrated vari ables. In particular, no variable
should appear in the li mits of the outermost integral. For example, the integral of (x +y)2
2
over the region Ddefined by 9 :s y x , 0 x 1, should be written as
x' 1
not i10 (x+y)2dxdy.
Sketching the region often heJps in choosing your limits. Also, it helps to read the limits of
integration in this example as follows : "while yranges from 0 to x
2
; xranges bet ween 0 and I"
or "for each x between 0 and 1, y ranges between 0 and x
2
." Draw a picture and think about
it.
6. Definition of the integral. As in t he last section, this is important primarily for theoretical
purposes. Up to this section, we only know how to integrate over rectangles, so we cover a
region D with a large rectangle and let f(x,y) = 0 outside of D.
7. Double integrals and area. Iff( x,y) = 1 on a region D, then JIDf(x,y)dA is the area of D.

82
CHAPTER 5
SOL U TIONS TO SELECTED EXERCISES
l. (b) Theiterated integralis
1
3x 1
2 [1 +
dy
]
dx =
12(
y
13x+1)
dx
1 2x 1 y=2x
12 (3x +1- 2x)dx
1 \ x +1) dx = +x) I:=
At each Xo in [1,2], the region extends from 2xo t o 3xo +1,
so we get the region shown in the sketch. This is a y-si mple
region because itcan be described by
1 x 2 and 2x y 3x+1.
2. (b) First, we will sketch t he region. At each Xo in [-1, 1], the region extends from -2'lxol to
Ixol, sowe get theregion in thesketch. Theintegrationis most
easilydone by di\ridingthe region into two parts, so
1
1
1X 1
1
x
l
e
X
+
Y
dydx=
i -X
e
X
+
Y
dydx+ e
X
+
Y
dydx .
' 10 1
- 1 - 2/x l - 1 2x 0 -2x
1x
Note that when xis in [-1 , 0]' IxI=: -x. The first integral of
t hesum is
1
X
0
0 ( ex+yl- ) dx = 1(1- e
3x
)dx
- 1 y=2x -1
2 1
(x- [1
"3 - 3e3 '
Thesecond integral ofthe sumis
2
1
1 (
e
)
dx=
11
- e- ) -,- +e-
) 11
=- +-
1
- -.
X
+
Y IX
(e
2x X
dx ::::
(e
2X
x
e 3
o y=-2x 0 2 0 2 e 2
Therefore, the entire integral is e
2
/ 2+l/e- 1/3e
3
- 5/6.
(e) The iterated integralis
Y
11 [l>x
n
+ym)dX]dy t [( _xn+1 +xym)I 1dy
Jo n+1 X=y2
t (_y'n_+_1 +ym+1 _ _y2_n+_2 _ ym+2) dy
Jo n+1 n+1
yn+2 ym+2 y2n+3 ym+3 ) 11
(
(n+l)(n+2) + m+ 2- (n+ 1)(2n+3)- m+3 10
1 1 1 1
(n+ 1):-:- (n-+----::-:-2) + -m-+- 2- (n + 1)(2n+3) - -m-+-3'
Tosketch the region, note t hat for every Yoin [0 ,IJ ,x extends fromx = to x =Yo.
y = 3x+1
x
-1
83 DOUBLEAND TRIPLE INTEGRALS
y
x
4. Theequationofan ellipsewithserniaxes aand bis x
2
/ a
2
+
y2/b
2
=1or x
2
/b
2
+y2/a
2
=l. Ineither case, the areas
are equal. We will use the first equat ion, find t he area of
the region in the fi rst quadrant , and t hen multiply by 4.
The region is descri bed as
Theareaofaregion Dcan becomputedby A = ffDdx dy.
In this case, we have
b
2 o
o
1 1
a
1..j1- x /a 1"( l
b
..j1-x /a 1
a
g2
-A= dydx = y dx =b 1-
2
dx.
4 a a a y=O a a
Let u = x/a to get bf01 a..jl - u
2
du =abf0
1
.Jl=U2duo The integral f01 ~ du is the
areaofa quarter ofa circle ofradius I, which is If/4. So A/4=ablf/4. Therefore, the area of
the ellipse is A = ablf.
7. The region D,shown in the sketch,can be descri bed as
- V3/2 y ~ V3/2 and 0 ~ x ~ _4y2+3,
3
x
so
4yO 3
.rx
3
ydxdy = j,;3/2 r- + x
3
ydxdy
JJD -,;3/2JO
3
,;3/2 (x41-
4y2
+ )' j,;3/2 (_4y2+3)4y
= -y dy = dy.
j
- ,;3/2 4 x=O - ,;3/2 4
Let u =_4y2+3and wegetan integralwhoseli mitsofintegration areand0, so theintegral
is 0.
12. From section 2.1, we know that an equation of the
form z2 = -a(x
2
+y2) is a cone. To get t he t ip at
(0,0,h) as shown in the figure, the equat ion becomes
z
z2 =h-a(x
2
+y2). When z =0,we get h =a(x
2
+y2).
Ifa =h/r2, we get r 2 =x
2
+y2, a circle ofradius
r. Thus, the equation of the cone is z =f(x,y) =
Jh- (h/r
2
)( x
2
+y2). The region of integration is
the cone's base, which is descri bed as
--- - - - - -- ~ . ~ .
y
x
__a_ _ _
CHAPTER 5
84
Therefore, the volumeof the cone is

/
2
-r - -..!r'-x' Vh- :2(X +y2) dy dx .
Fortunately, we areonly asked to set up the integral. We can "simplify" the integral by using
t rigonomet ricsubstitut ion or by fi nding the followingequat ion from an integral table:
2
1
j
Ja2-x2dx= 0 sin- for a>O.
2 2 a
By using the equation above, the volumeof the cone becomes
J:r [(rJh-V; x2) sin-
1
x2)1dx.
Changingthevolumeintegral topolarcoordinates, a technique introduced inchapter 6, makes
t he calculation easy.
14. Let Dbe a y-si mple region ;t hen D is described by
and
and the integral IID f(x)g(y) dA becomes
l
bl2(X)
f(x)g(y) dy dx.
a ,(x)
Now, we integrate in y. Let Gbe an antiderivative ofg and not e that f(x ) is constant in y.
We get
b
[ i 2(X) 1 l b
f(x) g(y)dy dx = f(X)[G(4>2(X)) - G(4)l(X))]dx.
la ,(x) a
Gdepends on x, so we can'tconsider it to be const ant and we cannot factor it outside ofthe
integral sign . Thus ,in general, IID f(x)g(y) dydx is not t he product of two integrals.
5.4: CHANGING THE ORDER OF IN TEGRATION
GOALS
1. Be able to evaluate a double integral by changing the order of integrat ion.
2.Be able tost ateand understand t he mean value t heoremfordouble integrals.
STUDY HINTS
1. Rati onale for changing order. Recall t hat Fubini's t heorem allows you t o change the order of
integrat ion. Somet imes ,changing the order of integrat ion si mplifies a problem. Try doingex-
ample1 wi thoutchangingt heorder ofintegration and compare theefficiency ofbothmethods.
(You will probably need to use trigonometricsubstitution.) At other t imes, a double integral
can only be comput ed ifthe order is changed.
2.Beginning the change ofmYier. It is useful tosketch t he region ofintegration from the given
li mitsbefore choosing new limit s.
3. Mean value theorem for int egrals. If two conditions hold: (i) f is continuous amd (ii) D is an
elementary region, then the conclusion is
JIn f( x ,y)dA =f(xo, yo) .area(D)
for some point (xo,Yo) in D.
85 DOUBLE AND TRIPLE INTEGRALS
4. Mean value inequality. If m is the mini mumvalue of!(x ,y) on D and M is the maximum,
then
marea(D) Jl !(x,y) dA M . area(D).
This allows you to estimatethe value of a dou Ie integral .
SOLUTIONS TO SELECTED EXERCISES
1. (b) Fir t, we recall thatcos
2
B= (1 +cos2())/2and compute the integral as written:
2 9
r/ ros
io io cos()drdB =
iT
4
Fromthe graph, we see thatifwe choose an TO, then B ext ends from 0 to cos- 1(ro) . Thus, the
region can also be described as
and
r
r
a
Therefore, changing the orderof integration gives us
11COS-l(r) 11 ( ICOs-
1
(r)) 11
1
cosBdOdr = sin() dr= si n(cos- 1(r))dr.
a 0 0 9=0 0
From the triangle, we see thatsin(cos-
1
(7))) We recogni ze t he integral s he area
of a quarter ofa circle of radius 1,so
1
1 iT

o 4
2. (c) The region of integration is sketch here. There is no obvious
function whose derivativeis exp(x
2
), so we try changing the order of Y
integration. Theregion can beexpressed as a y-simpleregion:
o x 2 and 0 y 2x.
Thus , the integral becomes
2X
(2 rx exp(x
2
) dydx = r [ yexp(x
2
)1 ] dx.
ioio io y=o
2 x
2
Now let u =x and we get
2 4
12 2xexp(x ) dx =14 e
ti
du = e
ti
I: =e - 1.
4
86 CHAPTER5
3. Thisformulaisanapplicationofthemeanvalueinequa.li ty. The regionDis[-11",11"]X [-11" , 11"], eo
A(D)is t he areaofD, which is411" 2. Thefunct ion I(x,y)=esin(x+y) is largest whensin(x+y)
is largest, i.e., when x+y =rr/22n1l" , where n is an integer. Similarly, I(x,y) is small est
when x+y =- 11" / 22n1l". Over the regi on D, we have -1 sin(x +y) 1,so m =li e and
M =e
1
. Therefore, substitutingeverything into t he mean val ue inequality gives us
4!2 J L l(x ,y) dA e.
6. The areaofthe triangleis tThefuncti on I(x,y)=11 (y- x+3) issmallest when y- x+3
is largeston D.Notice that y x in D, so y - x +3is largest when y =x ,and so m =1/3.
Similarly, we see t hat Moccurs at (1,0) where 1(1,0) =t.Now, the mean val ue inequality
gives us
or
10. We need to di vide the t riangle into two parts: one part for x in
[0,1]and the other part for xin [1,2]. When xisin [0 , 1], we have Y
x Y 3x, so the integral over this region is
(2.2)
[1[3'"
J J., eX- Ydydx
o
11 (-ee - yI::Jdx
11 (_e-
2e
+l)dx
x
2
e- ", ) 11 1 1
( - 2-+
x
0 = 2e2 +2
For the second region, we have x y 4- x, so the integralover t hat region is
2[4- " 4 X
X 2X 4
1 J., eX- Ydydx 2 ( - e_
Y
I - ) dz= 1\-e- +l)dx
1 11 11 = " 1
2X 4
_ e - ) 12 1 1
( 2 +X 1 = 2e2 +2
Adding the two integrals together, we get 1+ 1/ e
2
.
13. First, we sketch the region. For a.ll Yoin [0,1],xgoes
from Yo to -/2- To describe the region as an
x-simple region, we have to subdivide the region at
y
x =1. When xisin [0,1]'we have 0 y x . When x
is in [1,)2], we have 0 y )2- x
2
. Thus,we have
2112
x
[1 1 t[r ] 1v'2 [ 1
Jo Y I( x ,y)dx dy =Jo J I (x,y)dy dx + 1 Jo I( x ,y)dy dx .
o
In manycases, the left-handside will be easier t o work with.
87 DOUBLE AND TRIPLE INTEGRALS
15. This proofrequires using the chain rule. Let
G( x ,u) =l
x
jd f(u, y, z) dzdy.
We want to find dG/dx (the "totalderivative," ifyou will ) when u =x. By the chain rule,
dG _ aGI +aGI
dx - ax u=x au u=x'
is simple; itis equal to :x J: [t f(u, y,z)dZ] dy. From thefundamentaltheoremof(one-
variable) calculus, we simply take the "integrand" (the dz integral here) and replace y by x
(because we are integrating with respect to y). At u =x,
aGI jd
= f( x ,x,z) dz.
uX u=x c
is a little tricki er. First,
aG
al
X
{d
au aJc f(u,y,z)dzdy
au
l
X
a
j
d
a c auf(u,y,z )dzdy,
assuming thatf is a nice functionsuch that the orderofintegration and differentiation can be
interchanged. Now we want toevaluate everything at u =x, so we simply replace uby x, and
get
l
X

aGI
=
d ()
-;;-f(x,y,z)dzdy.
j
uU u=x a c uX
Add the two results and you'vegot it.
5: THE TRIPLE INTEGRAL
GOALS
l. Be able tocompute a triple integral over geJleral regions in space.
2. Be able tochange the order ofintegration for computing triple integrals .
STUDY HINTS
l. Notation. We use dVfor the differential dxdydz since it represents a vol ume.
2. Properties. Many ofthe properties ofthe tri ple integral are the same as those ofthe double
integral. The tripleintegralmaybeconsidered asa t hreefold iterated integralandwe integrate
from inside to outside. Fubini's theorem still holds and we can still integrate all piecewise
continuous functions.
3. Balls. Example 3demonstrates one way todescribe the unit ball. Another way todescribe it
IS
Yet another description is
Thesolut ionofexample3usesa y-si mpledescri ptionofD.Usinganx-simpledescription would
generate three moredescriptionsofthe unit ball.
88 CHAPTER 5
4. How to integrate. As with double integrals , knowing how to set up the limits is important.
Drawing a picture of the region is helpful for fi nding the limits ofintegration, although in
many instances itis easier said thandone.
5. Triple integrals and volumes. IfI(x, y, z) =1, then fffwIdV is the volume of W.
6. Integration trick. Atime-savingdevice is shown in example 4. Ifwe realize thatan integral is
the areaofall or part ofa circle, then we don' t have to go through the process offi...nding an
antiderivati ve.
7. Factoring out integrals. Some special t riple integrals can be easier to compute ifyou use the
followingfact: Ifthe limits ofintegration ofthe innermost integral do not involve a certain
variableandthelimitsof integrationforthatparticularvariableareconstant, thenthatvariable
may be integrated separately and mul t iplied by the remainingintegral. For example,
2 3 3 1
1
1
1:&1I(x)g(y)h(z )dzdydx= (1h(Z)dZ) (11'2XI( x)g(y)dydX)
because the variable z does not appear in any of the limits ofintegration and its limits of
integration arethe constants 2 and 3. Wecannot integrate X' separately because it appears in
the limitsofintegration for y.
SOLUTIONS TO SELECTED EXERCISES
2. Itiseasier to integrate in x before integratingin y:
JJl e-xYydV = 111111e-X'Yydxdydz =1111[(e=:Y) yl:=J dydz
1
111 11[ 1] 111
(l-e-
Y
)dydz= (y+e-y)1 dz=- dz;;:::-.
1
o 0 0 y=o e a e
ote that you would have had to use integration by parts ifyou had int egrated in y before
integratingin x.
6. Itis useful to makeasketchoftheregion. The elliptic
cylinder runs parallel to the y axis and its int ercepts
on the coordinate axes occur at x =1/.j2and z =
l. The ball is a sphere ofradius 2 centered at the
origin. To describe t he region inside t hecylinder and
the ball as an elementary region, let x vary between
-1/ J2jnd1/J2. Next, for anygiven xo, we let zgo
from - 1- 2x5to J I- 2X6. Finally, for any given
(xo, zo), we let yvary between twosurfacesoftheball ,
namely, y =-J4 - x5- Z6 and y =J4-x5-z6.
Therefore, t he desired region is descri bed by
-1/V2 x 1/ V2, - Jl- 2x2 Z J l- 2X2, - \1"4 - X2- z2 Y J4_X2- Z2,
Another possiblesolution is
x
2
x
2
-1 Z 1, - J(l- z2)/2 z J(1- z2)/2, - J4- - z2 Y J4- - z2.
x
2
9, Thesurface z =x
2
+y2 is a pa.raboloid opening upward. The surface Z =10 - - 2y2 is a
paraboloid opening downward. When we equate the two surfaces, we see that they intersect
where x
2
+y2 =10- x
2
- 2y2 or 2X2 +3y2 =10, which is an ellipse. The ellipse can be
described by
- V5 x V5 and - J (10- 2x
2
)/3 y J(10- 2;2)/3.
DOUBLE AND TRIPLE INTEGRALS 89
As a tripleintegral ,thedesired volumeis
Vsjv'(lO-2XJ) /3 1 10-X2 -y2


dzdy dx = (10 - 2x2 - 3y2) dydx
j
-V5 x' +y'
- Vs - y"-(l-O---2x---::'-:- )/:-3
l
Vs
[ (lOy - 2x2y_ dx
-Vs -y(lO- 2x
2
)/ 3
_4_ j Vs _ 2v'2x
2
/5_ X2] dx.
3V3 -Vs
From the integral table, we have J d:c =(x/2)J5- x
2
+ sin-l(x/V5) +C and
Jx
2
J5=X2'dx =(x/8)(2:c
2
- 5)J5- x
2
+28
5
sin-
1
(x/V5) +C. Therefore, we get
v =
_4[10..;2 _x
2
+ sin-
l

3V3 2 2 V5
V5
(x r;;---.; 25. 1 ( x))] 1 251l'V2 501l'
- 2v2 _( 2x2 - 5)V5- x
2
+- sm- - =--=-.
8 8 v5 -V5 V3 y6
12. We will compute the volume ofthe region in the first 0 tant where x, y and z are positive.
Due to t hesymmetryoftheregion,this result multipliedby8gives the correct answer. In the
first octant, we let x go from 0 to a. Since the region lies within the cylinder x
2
+y2 ::; a
2
,
one boundary is y =v'a
2
- x
2
Since we areonly looking at the first octant, we have 0::; y ::;
2 2 2
v'a - x . Finally, the zvalues oftheregi on gofromthe xyplane to thecylinder x
2
+z2 =a. .
Thus, 0::; z ::; v'a
2
- x
2
. The volume ofthe region in the fi rst octant is the triple integral


dzdydx =
a 0 a
Therefore, the volumeoftb entire region is 8(2a
3
/3) =16a
3
/3.
17. The plane x+y =1can bewritten as x =1- y. Therefore, Wcan be described by
0::; y::; 1, O::; :c::; l-y and 0::; z ::; 1l'.
Therefore, the desired integralis
1
r r1l1-Y
7r l 1 (31 -
Y
10 10 0 x
2
coszdxdydz
(cos z) x3
)
dydz
1
o 0 x=o
Alternatively, one can integrate in zfi rst:
1
Y
l
1f 1 1
Y
2
1
1
1- x cos ;;dzdxdy= l1 - (x
2
sinz[=J d:cdy = 1111-YOd:cdY=0.
90 CHAPTER 5
21. Evaluateas an iterated integral:
1 { 2X
Y
112X1X+Y
dzdydx =
t { 2X [ I
X
+ J
dy dx = 0 J (x+y-x2_y2)dydx
1
o 0
J
o
Jo z z=x
2
+y' 1
x'+y
2
o
r
1
2 ( 2 14
11 [((x- x - [:al Jo 4x +
3)
dx )y+y; ~ dx = -
4x
3_7X4)11 =
(
3 6
o
6
22. (a) St arting with the interval 0 ::; x ::; 1, sketch the area D
z&
whichextends from y =0to Y=Xo for each xoin [0, 1].Then
over each point (xo,Yo) in D,extend t he region Wfrom z =0
to z =Yo.
y
x
(1, 1,0)
(b) We will now fi nd an equivalent integral of the form
fffw !(x,y, z) dxdydz .
First, note that z goes from 0 to 1. Now, for each Zo, Y goes from Zo to 1. This gives us a
triangular areain the yzplane. Finally, foreach (Yo ,zo) in the triangulararea, xgoes from Yo
to 1. Thus, we describe Was follows:
o::; z ::; 1 and z::;y ::; 1 and y::;x ::; 1.
Thus,
Y 1 1
1
X
1
1
1
1
1 1 !(x, y,z )dz dy dx =1 !(X,y,Z)dxdY dz.
As aquick check, we cansee t hat both regions have thesamevolumeby letting!(x ,y,z) =1.
In this case, the volumeis i.
26. Wenote that x
2
+y2 +z2 ::; 1descri bes the uni t ball
centered atthe origin. Ifz ~ 0, then the region Wis
the upper hemisphere as depicted in the sketch. We
z
are given t hat Ixl::; 1 and Iyl::; 1, but this does not
provide any extra informat ion. To find t he limits of
integration, note that x varies between - 1 and 1, so
-1::; x ::; 1. For each x, t he variable y extends from
-Vf=X2to+Vf=X2. Finally, foreach (x,y) in the
unitdisk,thevariablezgoesfrom0to+VI - x
2
- y2. "
,
,
\
. . - - - - - - - - - - ~ ~
Thus, W can bedescri bed by
y
-1::; x ::; 1, - ~ ::; y ::; ~
and x
o:S z ::; VI - x
2
- y 2 .
Therefore,
1
1 1 ~ lVI-X,-y2
!dV = !(x ,y,z) dz dy dx.
jii
W - 1 - ~ 0
91 DOUBLE AND TRIPLEINTEGRALS
27. The volume, as a triple integral,is
11
[I
f(X' Y) ]
dz dxdy
D 0
and integrating with respect to zyields
JL!(x ,y) dxdy,
which is the double integraloff over D.
SOLUTIONS TO SELECTED REVIEW EXERCISES FORCHAPTER5
2. Evaluatethis as an iterated integral:
6. The region is sketched here. As an x'-simple region, it
can be described as 0::; y ::; 1and 0::; x ::; y'l , 0 we
get Y
1
(1,1) 1 [(X+y)3IY2 ]dy
a 3 x=o
9. Evaluate as an iterated integral:
X
l
1
1[(yz+x;2)[=Jdydx
X
1
1
1(y2 + dydx
= 11 [ (1 + y;[=Jdx
11 + dx = + I: =
x
92
CHAPTER 5
13. From one-variable calculus, we expect t.o get
y
l
b
(f( X)- g(x)]dz
if1 9for z E [a,b]. The region Dcan be described
as a z band g(x) y I(x ) because it is a
y-simpleregion. Therefore, the areaof Dis
a b x
f
l
b
l
.
(X)
dy dx =
lb[
yl
f(X)]
dx =
Ib
[/(x)- g( x) ]dz .
a g(",) a !I =g( 3; ) a
15. (b) As a y-simpleregion, the unitcircle can be described by
and
Thus, we have
11 [2 211
2
1
1v'1=X'2 x
2
y2dydx = 1dx == - x
2
(1 - X ) 3
/
2 dx.
1
-1 -1 3 3 -1
Use the trigonometricsubsti tution: =cose,x =sineand dz =cosedo, 80
2
Jx
2
(1 - x ) 3
/
2dx = Jsin
2
Bcos
3
B(cosedO)= J(sin2e- 2sin
4
B+sin
6
e) de,
which was obtained by using the identity sin
2
B+cos
2
B=1 and expanding. Now, use the
identities: sin
2
e= (1- cos2e)/ 2and cos
2
e= (1+cos2e)/2. We get
1- cos2fJ _ 1- 2cos2e_ cos
2
2B 1- 3cos2B +3cos
2
2fJ - cos
3
2e] dB
J
[
2 2 2 + 8
- cos4B+cos20sin 20) dO = .i._sin4B _ sin 2e+C.
16 16 8 16 64 48
Now,we use t hedoubleangleidenti tiesandthefollowingwhich
we getfrom the triangle
J
2 3

e== sin- 1:1:, sine== z and cose=

Substitut ing and eval uatingat 1, we get
I 2
sin- z _ Z2-x ) _ 8z
3
(1-
[
16 64 48 16
-I
Therefore, the originalintegral becomes (2/ 3)(11"/16) == 7r/ 24.
93 DOUBLE AND TRIPLE INTEGRALS
20. As the integral is written, the regi on is described as 0 x 2 and
- 3v'4- x
2
/ 2 y 3v'4- :z:2/2,which is a y-simpleregion. When we
rearrange y =3v'4- x
2
/2,we get y2/9+x
2
/ 4=1, which we recognize
as an ellipse, so oursketch is asshown. We evaluateas follows:
r
2
j3"f4-X
2
/2 ( 5 + y3) dydx
io v'2+X
l' [(vA--.+ d.
12 = - 10(2 - x)3/21: = 20V2.
24. As written, the region de cribed by 0 y 1 and 0 :z:
3y is an x-simple region. The region is sketched as shown.
We evaluate t he double integral as follows:
t [3
y
X
io i o e+
Y
dxdy = 11 (e
X
+
Y
1::0) dy
4l1
t (e
4Y
- e
Y
)dy = ( e _e
ll
)1
1

Jo 4 0 4 4
27. The region is sketched h reo Note that is consists ofthe following two regions:
o :z: 1 and 0 y _ x
2
+X
1 :z: 2 and - x
2
+ X Y o.
We evaluateas a sumofintegrals:
y
tl-X2+X [2jO
io 0 !(x ,y)dydx+ il _X2+x I(x ,y) dyd:z: .
x
The first integral is:
X2
X
1
1
1- + (:z:2 + 2xy2+ 2) dydx
l' [(x'Y+ 2xt + 2Y)C+<] dx
The second integral is:
0
y3
[ 2j O (x
2
+ 2xy2 + 2) dydx = [2 [ (x
2
y+ 2x + 2V) ] dx
i1 - x' +x il \
3
1
y=-x'+:r
(-1) 12 + 2x6- 2:z:5 - + X
3
- 2X
2
+2X) dx
8 6 5 3
_x 2x7 x x X4 2x ( 584 27) 2)11
(-1) (12+ 7 - 3 - 15 + 4 - 3 + x 0 =(-1) - 105 - 70 .
Add t he two integrals together toget Ii .
94
.....
CHAPTER 5
31. Thefunction f(x, y) =x
2
+ y2+ 1has the value r2+1where r is thedistance from the origin.
Thus, on D, 1::; f(x,y) ::; 5. The area ofDis411", so by the mean value inequal ity, we have
thedesiIed result.
35. Dividing thIough by x2, the integrand becomes (l/x)/[1+ (z /x )2]. Let u =z/ x, so 3; du =dz
and the integral becomes
[1 [Y t/../3 ( 1 ) t r( 1 11/../3) [1 [11 11"
10 J 10 1+ u2 dudxdy
J J tan- u u=o dxdy ="6 J 10dxdy
o o o o
-
11" 11(
x
I
Y
)
dy =-
11" 11
ydy=-
11" ( y
-
211)
=-
11"
.
6 0 x=O 6 0 6 2 0 12
TESTFOR CHAPTER5
1. Trueor false. IT false, explain why.
(a) The areabetween the parabolay = 4- :r
2
and the x- axis can beexpressed as the double
integral
4 _ X2 2
[ j dxdy.
Jo -2
(b) If g(x,y,z) is integrableover a region Q in1R3, then
JJkg(x,y,z )dxdydz =JJkg(x,y,Z)dydZdX.
(c)
y 3
tj4jS_y3(z+ y)3/2dzdydx =(j41
5
- (z + y)3/2 dzd
Y
) .4.
Jo - 2 y2 - 2 y2
(d) For any integrablefunction f(x,V), thedoubleintegral lID f (x,y) dx dy isthe vol umeof
the region bounded by t he sUIfaces z = f( x,y)and z= 0over D.
(e)
1
2jl(X2+ 4x+ .JX)(y3 _ y2 +4y) dydx
o - 1
1 1
[1
2
11( 3;2 + 4x+ .JX)dydX] [1
2
11(y3- y2+ 4y)dydX].
y
2. Changetheorderofintegrationandrewritethe triple integral10
1
1:2Io
1
- dzdydx in fi ve ways.
Computeits value.
3. Findthevolumeofthe region bounded by z =x
2
+1,z =0,y =0and y =x
3
for-1::; x ::; 2.
4. Evaluate the double integral I/
1
- COS(I ) Ic
1
os-
I
( I _ 2) ydxdy.
y
x
2
5. An amateUI magician'shatisdescribed by z::; 9- - y2 and z ~ O.
(a) Use Cavalieri'sprinciple tocalculate how muchmilk he can pour into hishatifhewants
to bake a magiccake.
(b) Express the volumeofthe hat as a double integralin t heform IIf(x,y) dxdy.
(c) Use parts (a) and (b) to quicklycalculate I ~ 3 9 - t
2
)3/2 dt.
DOUBLEAND TRIPLE INTEGRALS 95
6. Evaluate the double integral JJT Ix- yldy dx, where T is the triangle wit h vertices at (2,0),
(0,0) and (0,2) .
7. (a) Suppose f :W -t is continuous and W is an element ary regi on in IR . Let V( W) be
t he volumeof W, m the minimum val ue of fon Wand M the maximum value of f on W.
Make a statement relating m, M, V and the triple integral off over W.
(b) Show t hat
v'2 l
1
1
Y
l"/ 4cos(xy) 7r
:S +z dx dz dy :S -8'
16
e
2
0 a 0 e
Y
8. Eval uate J04JolJ3z exp(x3) dx dydz.
9. Eval uate JJD x dx dy where D is the region shown.
The bOWldary on the left is y =4 - x
2
. All of the
other boundaries are straight lines.
x
(4.- 3)
10. A cake-eating monster saw a pyramid-shaped cake which he easily devoured. Th cake had
a square cross-section and at a height h, the side of the square had lengt h 8- h. The cake
monster'sson could only eat the top ofa. similarlyshaped cake.
(a) Use Cavalieri's principle to calculate t he cake monster's minimumstomach capacity.
(b) How t all is t he remainder ofthe cake that the son started eating?
.. .
97
6 THECHANGE OFVARIABLES FORMULA
AND APPLICATIONS OF INTEGRATION
6.1: THE GEOMETRY OF MAPS FROM]R2 TO JR2
GOALS
1. Be able todetermine whet her a mapis one-to-one.
2. Be able to graph a region D which is the imageofa region D* under a mapping.
STUDYHINTS
1. Notation. Theequality D =T (D* ) expresses thatfact that T t akes a pointin D* and mapsit
to a point in D, and t hat allsuch points ofDareobtainedin this way. T is called a mapping
or a transformation. Thinkingof T as a function, this is similarto y =f(x) for f :JR -tJR.
2. One-to-one. If two distinct points always get mapped to distinct points, i.e., distinct points
never get mapped to a single point, then the funct ion is said to be one-to-one. Adomainmay
need to bechosen carefully for thisproperty tohold. (See example3.) For integration, this is
a nice property;otherwise, the double integralover D maynot equal the double integral over
DO.
3. Onto. If every point ofa rangeDcan beobtainedfromsomepointin thedomainD",then the
mappingis onto. Onto does not implyone-to-one because t wo points ofD* mayget mapped
to one point of D.By t he samet oken, one-to-one does not implyonto.
4. Finding D f rom D* . In many cases, the range Dcan be determined by simply mapping the
boundary of D* . Then decide whet her the map takes D* to points inside or outside the
boundary ofD. For example,let D be the unit disk. If
then T maps the uni circle to the unit circle, and T maps a point such as ~ , ~ ) to (2,2), a
point outside t he circle. So T maps t he inside ofthe circle to the outside [and is not defined
at(O, O)J .
5. Import ant exercise. The result of exercise 8 is used in the examples ofsection 6.2. It states
thatifT(x) =Ax,then T is one-to-oneifandonlyifdetA =f:. O. Theresult is truefor all n xn
determinants.
6. Lineartransformations. If T( x)=Ax+b and A isa 2x2mat rix, t hen T takesparallelograms
to parallelograms. If D is a parallelogram, one can often obtain it from the unit square D*
withsuch a T.
98
-
CHAPTER 6
SOLUTIONS TO SELECTED EXERCISES
2. The transformation is given by the followingequation:
* [1/-/2 - 1/-/2] [ X: ] .
T( x, y )= Ax = 1/-/2 1/V2 y
We computedetA =1i- 0, so by theorem 1, the linear mapping T maps vert ices to vertices.
The origin ofD* gets mapped to T(O, 0) =(0,0) in D. Similarly,the points (1,0), (1, 1) and
(0,1) ofD get mapped to (1/V2, 1/V2), (0,-/2) and (-1/V2,1/)2) ,respectively, in D. Itis
easy to show that the length ofeach side ofthe transfor mation is 1and t he dot product can
be used to show that the angles are 7r/2. Thus , T rotates the unitsquare by 45 degrees.
y!
y

T
1
x
* x
4. The four vertices (0,0), (1,0), (1,1) and (0,1) ofD should be mappedfrom the four vertices
2 2
(0,0) , (15 , \6) and (t, 15 ) ofD. We want to find a mat rix A so that
T( x*, yO) =Ax = [:: ].
When 1) gets mapped to (1,0),we get 8a/5 +4b/5 =1and 8e/5+4d/ 5 =0, so d = -2e.
When (i,Y5
2
) gets mapped to (0,1),we get 4a/5+ 12b/5 =and 4c/5+ 12d/5 =-4c =1, or
a = -3b and c= -i, d = t. Substitutinga =- 3b back into8a/5+4b/5 = 1gives us - 4b = 1
or b =-t,and so a = Therefore,
( *) (3 * 1. 1 *
+'2
1y*)
T x = 4'x - 4'y ,- 4'x .
7. Since sines and cosines appear, we will try to use t he identity sin
2
t +cos
2
t = 1 to eli mi nate
the parametersand obtain a recognizable form: x
2
+y 2 = p 2 sin
2
(cos
2
B+sin
2
B) =p2 si n:!
and x
2
+y2 +z2 = p2(si n
2
+cos
2
) =p2 . We recognize that x
2
+y2 +z2 =p2 is a sphere
ofradius p; since :s p :s 1, D is the unit ball.
The map T is not one-to-one. For example, (0, 7r, 7r /2) and (0 ,7r/6,7r/ 5) both mapto the
origin. Also, (I,O,7r/2) and (1, 27r , 7r/2) both map to (1, 0, 0).
Since p =alwaysgets mapped to theorigin,we want toelimina.tethatpoint . Also, si nce
B =and () =27r give the same mapping, we want to eli minate either one of those points.
Hence, T can be made one-to-one by using the followingintervals: E [0, 7rJ, B E (O,27r] and
pE(O,l].
10. Using the hint and applying T to the parallelogramdescribed by q =p + .xv + /-,W, we have
T( q) =Aq =Ap+>.Av +/l Aw.
Since A is a 2x2matrixand p,v, w are all vectors in JR2, then Ap, Avand Aw are all vectors
in JR
2
. Since A is nonsingular (detA i- 0), Av cannot be a scalar multipleofAw ifv is not a
scalarmultipleofw. Thisis because ifAv =aAw , ascalar, then Av -aAw =A(v - aw) =0,
which impliest hateither detA =ofv =aw. Thus,theimageofT is a parallelogramin JR2.
99 THECHANGE OF VARIABLESFORMULA AND APPLICATIONS OF INTEGRATION
6.2: THE CHANGE OF VARIABLES THEOREM
GOALS
1. Given a transformation T, be able to computeits Jacobian.
2. Be able to use the J acobian to change variables in double and triple integrals.
3. Be able to state and use t he change of variables formulafor pol ar, cyl indrical and spherical
coordinates .
STUDYHINTS
1. Review. You should review the definitions ofpolar,cylindrical and spherical coordinates from
section 1.4.
2. Jacobian. The Jacobian is a determinant, not a matrix. In this section, we introduce the
notation
a(x,y) o(x,y,z )
a(u,v)
and
o(u,v,w),
which are the determinants
ax/au ox/ov ax/ow
ax/au ax/av
ay/au oy/av By/ow
ay/au oy/ov
and
l
az/au 8z/ Bv Bz/ 8w
respect ively.
3. Changing variables . As stated,the changeofvariables theorem requires that T be one-to-one.
However, thet heorem still holds if Tis notone-to-oneon the boundaryofD,asit uationthat
occurs in a number ofexamples . Asimilar t heorem holds for triple integrals.
4. Useful change ofvari ables. Ifx
2
+y 2 occurs in the integrand, try usi ng polar or cylindrical
coordinates , which have the Jacobian r. Ifx
2
+y2 +z2 occurs in the integrand, try using
spherical coordinates, which has t he Jacobian p2 sin1jJ . These two Jacobians should be memo-
rized.
5. Integrating over ellipsoi ds and ellipses. Ifyou are integratingover an elli psoid wit h equation
(x/a)2+(y/b)2 +(z/e)2=1, the subsitutions u =x/ a,v =y/band w =z/e may be made
to transform the ellipsoid into a sphere. Spherical coordinates may now be used to integrate
2
over u
2
+v
2
+w =1. Similarmethods can transforman ell ipse into a circle and then pol ar
coordinates may be used.
6. Limits of int egrati on. When you set up t he limits of integration, remember that ; is mea-
sured from the "north pole," not from the "equator ." Also, recall that () is measured in a
coun erclockwise direction.
SOLUTIONSTO SELECTED EXERCISES
2. We know t hat x =u+v and y=u- v. Adding the two equations gi ves us (x+y)/2 =u and
subt raction gives (x - y)/2 =v. By mapping the boundaries ofthe triangle in the x1Tpl ane,
we get another tri angle in t he uv-plane as shown.
y v
~ . ~ )
u
100 CHAPTER 6
TheJacobian is the absolute value of
l
a(X,y)I_I ax /au ax/av 1_11 1 1- 2
a(u,v) - ay/au ay/av - 1 - 1 - - ,
so the Jacobian is 2. In this case, we integrate x + y =2'11 as follows:
J
o [ t/2[I - V {1/2 {I-V
JD(x+y)dxd
y
= Jo i" (2u)(2)dudv=4
Jo
J" u dudv
1
1/2 (u
2
11-V) 11/2 1/2 1
4 - dv=2 (1 - 2v) dv= 2(v-v
2
)1 = 2'
o 2 ti = v 0 0
Don' t forget t hat the Jacobian is the absolute value ofthe det erminant. A direct calculation
givesus
y2
1
111 11 [(x2 ) 11 1 11 (1 3 ) ( y y2 y3)11 1
(x+y)dxdy = - + xy dy = - + y- - dy = - + - - - =-0
o y 0 2 x = y 0 2 2 2 2 2 0 2
3. (b) We compute that x - y =2'11 - 3v and the Jacobian is
I
a(x,y)I= 14 a I=12.
a(u,v) 2 3
Therefore, the integral becomes
12111\2U- 3v)dvdu = 1211 [(2UV- 3 ~ 2 ) [=J du = 1211(2U - ~ ) du
12 (u
2
_ 9; )1: = 1 2 1 - ~ ) = - 42.
Notice that we did not even need t o know what D looks like to perform the integration. By
theorem 1ofsection 6.1 ,this linear transfor mationmaps the rectangle D to a parallelogram
D. We mapthe vertices and then connect them to get D.
v y
(4,8)
2 (4,5)
u x
7. Wecomputet hatx
2
+y2 = (u
2
_v
2
)2+4u
2
v
2
=u
4
+2u
2
v
2
+v
4
=
v
(u
2
+v
2
)2, andso';x2 + y2 = u
2
+v
2
. Inthiscase, theJacobi an
1
2u
IS
l
a(X'Y)1= - 2v I=4u
2
+4v
2
.
a(u,v) 2v 2u
Therefore,
4(u2 + v )
1 u
Jj
2 2 dudv =4 dudv.
D. u +v D. J1
2
Thisis just4 times the areaofD*,which is a quarter of a unit
circle,so the answer is 7r .
THE CHANGE OF VARIABLES FORMULA AND APPLICATIONS OF INTEGRATION 101
11. The transformation takes (x, y) to (rcosO,rsinO),so
the Jacobian is
y
2 8
(x,y) I= I -rsinO 1=r.
8(u, v} Sillf) rcosf)
1
Recall that t he area is t he double integral fID dxdy.
In polar coordinates, this becomes
1
2lrl1+sinB
rdrdO
J10. rdrdO =
o 0
f21r (r211+Sin B) dO
Jo 2 r = O
x
1271: (1 + sin0) 2dO.
Use the halfangleformulato get
21r
11271: ( 1 2 0 I- COS O) dO- 1 (30 2 sin20)1 _ 311" II
- + sm + - - - - cosrJ - -- - --.
20 2 22 40 2
14. WhenD- = [0,1] X [0 ,211"] gets mapped. toa unitcircle, we knowwe areusingpolarcoordinates.
Subst itute u =1+ r2 to get
17. Here, we want to take the uni t square and find a transformation which maps it 0 the giv n
region R, which is a parallelogram. Note that we can map (0, 0) 0 (1,0) , (1,0) to
(0,1) to (0,1) and (1,1) to The transformaion should be linear, so we should get
(x,y) = (au+ bv + c,du + ev + f). When (u,v) = (0,0) gets mapped to (x,y) =(1 , 0) , we
see that c =1. The mapping of (1,0) to gives us a = he mapping of (0,1)
to (0 , 1) gives us b = -1. Similarly, we can determine d, e and l The transformation is
(3u/2- v+ 1,3u/2+ v), so the Jacobian is
8
(x, Y) I=II -11 I =3.
1
8(u,v) 2
y
v
(Yz-X)

1 u 1 x
Also, we compute x + y =3u+ 1and x - y =1- 2v, so the change ofvariablesgives us
102 CHAPTER 6
Since the integrand can be factored into separate factors which contain only one variable and
the limits of integration are all constants, we get
1 1
1
3 (1(3u + 1)2 dU) (1e -
2v
dV)
3 ( (3U; 1)31: ) ( I:)
1
3(64; 1) e ) = 221 (e _ .
20. The Gaussian integral tells us that
1:exp(-x
2
) dx =,fo.
We substit ute u =2x and du =2dx to get
a J2a
1:exp( _ 4x ) dx lim exp(-4x
2
) dx == lim exp(-u
2
)(du/2)
2
J -2a a-+oo -a a-+oo
2 1 Ja
-
2 a-+oo
li m
_ exp( _u ) du = ,fo
a 2 .
Since the function exp( -4x
2
) is symmetric about t be y axis, we deduce that
00
1exp(-4x2) dx = 1:exp(-4x
2
) dx = V;.
23. We will use spherical coordinates. The region S can be described by
a p b, 0 B 211" and 0 11".
The Jacobian for spherical coordinates is p2 sin , so the integral becomes
l
b
1
lr
1
2lr
!......--3
p2 sin
...:....

dB d dp.
a 0 0 P
Since all of the limits of integrati on are constant, we can "factor" t he integral as follows:
b
(l (1'" sin d) (1
2
". dB) = log . 2 . 211" = 411" log ( .
26. The Jacobian for cylindrical coordi nates is r.
(a) After sketching B, we see that B lies over the unit disk (see figure at left, below) and z
extends from 0 to y'x
2
+ y2 =r. Therefore,
2 r
ffl zdV = 1"'1
1
1rz dz drdB= 1
2
"'1
1
(r;2[=J drdB
"'1
1
r3 1
2
". (r411) 11
2
". 11"
1
2
- d7' dB = . - dB = - dB = - .
o 0 2 0 8 r=O 8 0 4
103 THECHANGEOFVARIABLES FORMULA AND APPLICATIONS OFINTEGRATION
z
x
y y
x
(b) The boundary x
2
+y2 +Z2 ::; 1 is the same as r2 +z2 ::; 1 (see the sketch of W above).
Solving for 7', we get r ::; ~ Al so, the integrand is l/vx2 +y2 +z2 =1/v'r
2
+z2.
Therefore, the change ofvariables gives us
r
2
7< t r ~ I' drdzd().
10 11/210 vr2+z2
Let u =1'2 +z2 in the inner r integral to get
1
2"11 11 1 27< 11( 1 ) 127< fl
--dudzd() = v'ul dzd() = (1- z) dzd()
o 1/2~ 2VU 1o 1/2 u = z ~ a J 1/2
{ 27r [ ( z2 ) ll 1 1 {27< 1T
10 z - 2 z =I/2 d() ="8 10 d() =' 4'
27. We want to fi nd a mappingof the unit square B* to the rectangle B. We want to map (0,0)
to (1 , 0), (1,0) to (4, 3), (1, 1) to (3, 4) and (0, 1) to (0,1). We want a linear transformation
which maps a parallelogramto a parallelogram. O UI desired mapping is (x,y) =(au +bv +
c,du + ev + f) . Mapping (u,v) =(0,0) to (x,y) =(1,0) gives us c =1. Mapping (1,0) to
(4,3) gives us a =3. And mapping (0,1) to (0,1) gives us b =-1. Similarly,we get d, eand
f We get (x, y) =(3u - v +1,3u +v). The J acobian is
IO(X'Y)I=1 3
- 1
8(u , v) 3 1
1=6.
Y
v
1
(3,4)
(4,3)
1 u 1 oX
Therefore, the integral gets changed into
..
104
30. In the first octant ofJR3,the spherical coordinate Bgoes from0 to rr /2. The desired region of
integration is sketched below. Recall that t he J acobian for spherical coordinates is p2 sin,so
the desired integral is
7f/21tan-1(2) 1.../6 (1)
- /sindpddB
1o 7f/4 a P
I
7f
/
2
jtan- (2) 1.../6
psindpd dB
1a 7f/4 a
7f/21tan-I(2) (2 1.../6 )
sin d dB
1a 7f/4 p=O
7f /21tan-I(2) 1 7f/2 ( I
tan
-I(2))
3 sinddB =3 - cos dB.
a 7f/4 a 1>=rr/4
Fromthe triangle below showing =tan-
1
(2), we get the following:
1 1 ) r/2 3rr ( 1 1 )
1
3 (- vis + Vi.1 dO =2 Vi - v'5 .
o
z
2
y
1
6.3: APPLICATIONS
GOALS
1. Be able to use double and triple integrals to compute averages, centers ofmass . moments of
inerti a and gravitational potentials.
STUDY HINTS
1. Averages. In general, an average is defined to be the integral offdivided by the integral of1.
Hence, the two average formulas given in t his section are
IffII' fdV ffD f dA
and
fffll'dV ffD dA '
2. Cente1' ofmass. This is a weighted average. In general, the center ofmass ofa region in JRn
is (Xl,X2,X3,...,X
n
), where Xj is
fRxj c5 dV
fR c5 dV
and dV =dXl dX2 ... dX
n
and R is the region. Since c5 is the density, fRc5 dV is the mass. If
the region is in JR2 use double integrals, and ifthe region is in IRs use t riple integrals.
CHAPTER 6
105 THE CHANGE OF VARIABLES FORMULA AND APPLICATIONS OF INTEGRATION
3. Moment ofintertta. A good way to remember Ix = IIIw(y2 +z2) 8dV is tonotice that the
integrand lacks an x term. Similarly, Iy and l z lack a y and a z term,respectively.
4. Gmvitational potential. Its formula is V = - OMm/R. Although you do not ne d to under-
stand the physic 1t heory behind t he discussion the mathematicsshould be clear to you. You
probably wi ll not need to reproduce the discussion for an exam.
5. Geometry. Recall thatift he integrandis 1,then the triple integral IIIwdVgives the volume
of W. Simil arly, thedouble integral lIDdA gives the areaof D.
SOLUTIONS TO SELECTED EXERCISES
3. Theformulasused tocomputet hecenterofmass(x,y)
for a region in the plane are
_ IIDxtS(;z;,y)dxdy _ lIDycS (x,y)dx dy
x = and y -
lIDtS (x,y) dxdy
- lID8(x ,y) dxdy .
In thiscase, we sketchtheregi on Dandsee that itcan
be descri bed by
o~ x 1 and x
2
~ y ~ x.
The numeratorofxis
x
[11"x(x+y)dydx
J
o
rl
The denominatoris
(l"(X +y)dydx
11 [ (Xy +y; )[=".]dx
Jo r'
X3_ X4 _ X
11 (3;2 _ _ ~ 4 dx =
(
5
)1
1
18
x 3
2 4 10 0 120
65
Therefore, x= i! .We leave it to you to calculate y=126'
6. Th mass ofthe plateis lID8dA, which is
[21< r 2 [27< (y3 2 )1" {27< ( 3 2 )
Jo Jo (y2sin 4x+2)dydx=Jo '3sin 4x+2y y =O =Jo ~ sin 4x+21r
dx.
Recalling t hat in
2
t = (1 - cos2t )/2, we have
1r
3
1r
3
2
{ 2" 1- cos4x {21r (x sin4X) 1 " 2 1r4 2
'3J 2 dx +21r J dx = '3 '2 - -8- a +41r = '3+41r .
o o
The area ofthe plateis 21r x 2r. =4,,2 Thus the averagedensity is mass/ area, or r.
2
/12+1.
7.(a) Let 8 be t he density of t he box, then t he massi IIfwIfdV,which is
t/2 t [2 (1)
tS Jo Jo Jo dz dy dx =If '2 (1)(2) = 8.
--
106 CHAPTER 6
(b) Again, the massis JJJw6dV , which is
[1/2t [2
2
10 1010 (x +3
y
2+
Z
+1)dzdydx
1
1/ 2 1
2
10 [ ((x + 3y2 + l)z+ z:)[=J dydx
tl2 [1 t/2[
2
1 ]
2
1
10 102(x + 3y2 + 2)dydx = 210 ((x + 2)y+ y3) Iy=o dx
1/2 (X3 )11/2 ( 1 3) 37
2 (x
2
+ 3)dx = 2 - + 3x 2 - + - = - .
o 3 0 24 2 12
11. Theaverage value off over W is JJJwf(x,y, z )dV / JJJw dV. Thenumerator is
6
1
2
1
4
1
2 2
sin 7r Zcos 7rX dzdydx.
Since f(x,y, z )can befactored so that each fact or containsonly one variableand thelimitsof
integrat ion are constants, we get
2 2
(1
2
cos 7rX dX) (1
4
d
y
) (1
6
sin 7r ZdZ ) .
Using t he half-angleformulas, we get
[1
2
C+ 27r X) dX] [1
4
dy] [1
6
C-c;s2rrz) dZ]
6
1 [( sin27r X) 12] [ 14] 1 [( sin27r Z) 1] 1 1
= 2 x + 0 Yo 2 z - 0 =2(2)(4)2(6) =12.
The denominatorisjust t he vol umeof W, which is 2(4)(6) = 48,so the average is = i.
14. The momentof inertiaof Waround the yaxisis given by
Iy= fffw 6( x2 + Z2) dxdydz.
Since W is the ball ofradius R, we will use spherical coordina.tes. Since x
2
+y2 + z2 = p2, we
can rewrite the integrand as 6(x
2
+z2)=6(p2 - y2). Thus, the momentofinertiais
[21f r [R
2
6
10
1010(p2 - p
2
sin sin
2
B)p
2
sin dp d dB
[ 21f r [R
610 10 10 (p4sin - p4sin
3
sin
2
B) dpd dB
(j 1 21f 1" [ (sin - sin
3
sin
2
B) ] ddB
5
<iR 3 r " r
-5- 1010 (sin - sin sin
2
B) ddB.
To finish the integral ,use the identities si n
2
t +cos
2
t = 1 and the half-angle formulasin
2
t =
(1 - cos2t)/2. The above integral becomes
<i R
5
[2" r
2
510 10 [sin- sin(1 - cos )sin
2
B] ddB
107 THECHANGEOF VARIABLES FORMULA AND APPLICATIONS OF INTEGRATION
5
2
6R r21r [( ( COS
3
fjJ)) I"1
-5- Jo - cosif> +sin B cosif> - - 3- =o dO
6R5121r ( 4. 2 ) 6R5121f [ 2 ]
-5- 0 2 - 3'SlD B d8=-5- 0 2 - 3' (1 - cos2B) dB
6R5 [2B_~ +sin2B] 121r 86R5
rr
5 3 3 0 15
19. From t he discussion in the text, we know that thegravitational potentialacti ngon a mass m
10-
11
lying outside ofC.M.W.is - GmM/R, where G =6.67 X Nm
2
/kg
2
, m is t he mass of
the small er object in kg, M is t he mass of the planet C.M.W. and R is t he distance between
the object and the planet's center. When t he object lies outside of a spherical planet , the
gravitational potent ial acts as if the mass M is concentrated at the planet's enter. Given the
pl anet'sdensity as J, we use spherical coordinates to calculate
21r X108
JJ
'r r rr
M = Jw J(x,y,z)dxdydz=J JoJ 6(p,fjJ, e)/si n<jJ dpdde.
o o
Since 6 is independent of and e, we have
Note that 10
12
and 10
8
are insignificant when added toa numberwith the magnitudeof10
17
.
Therefore, the potential is
v= -(6.67 x 10-
11
)(4. 71 x 1019) m/ R = -(3.04x 109)m/ R (meter/ second)2
.4: IMP ROPERINTEGRALS
GOALS
1. Be able to defi ne and computean improper integral.
STUDYHINTS
l. CompaTis on. One-variable integrals are call ed improper if the fun tion becom s infinite at
some point or if the limits ofintegr ation are infi ni e. The difference here is that a function
may become infiniteon a curve or surface.
2. Computation. To find an improperintegral,computealimit ingvaluerather thansubstit uting
the limitsdirectly. Fubini's theoremis valid for improper integrals of positivefunctions.
3. f becomes infi nite. Weremindyout hat f maybecomeinfinite betwe n t heli mitsofintegration.
In this case, we need todivide the region atthe points wherefis infinite. For exampl ,
0
1
1]1x 1
1
1 1111
- dydx = ~ dydx+ ~ dydx.
o - 1 Y 0 - 1 Y 0 0 Y
(1,1)
(1,1/2)
.. .&-
108
(HAPTER 6
See what happens ifyou evaluate the antiderivative at y = 1 and do notconsider y = 0; you
would get the wrong answer.
SOLUTIONS TO SELECTED EXERCISES
3. This integralis improperat x =O. Wesketch the region andsee
y
that it is a y-simpleregion described by
x 1 and x/2 y x.
Thus, theintegral is
[1 1X 1 x
11 ( [=x/ J dx
Jo x/2 X
1(x2 X2/4) _113x _3x211 3
- - -- dx - - dx - -
1o 2x 2x a 8 16 0 16
6. (a) AssumingD isay-simpleregion, itwouldbereasonable tothinkofb = 00 as an "endpoint."
Then we can define f fD fdA by
l
01 >2(X)
f(x, y) dy dx.
a >,(X)
(b) The region D is like the one described in part (a) . Since the limitsofintegration are con-
stant, we can integrate each variable separately and then multiply the results. The integrand
is factorable as follows:
xy exp(-x
2
_ y2) =[xexp(-x2)][yexp(-y2)],
so
1
JI f(x,Y)dA [1yex
p
(-y2)dY] [/00xexp(-x
2
)dX]
b
2
exp(-y2)11 )lim (ex
p
(-x ) I ) =(. (!_ .
(
- 2 a b-t oo -2 1 4 e e
2
7. Fubini's theorem tells us t hat multiple integration can be performed in any order as long as
the integrand is positive. In this case, e-
xy
>0, so we have
00
/21 e-XY dx dy =1
00
/2 e-
xy
dydx .
For the left-hand side, we compute
r[00 e-xy dx dy
2
lim
-xy Ib )
dy
(

J1 Jo 11 y x=O b-too
by
2 ( _e- 1) 12dy 12
/,
lim --+- dy= -=logy
1 b-too y y 1 Y 1
On the ri ght-hand side,we compute
1
2 00
1
00
XY X 2x
100 _e- 12 1 - e- e-
e-
XY
dy dx = - - dx = dx .
a 1 o :Z: y=1 0
Since Fubini's theorem tells us t he two integrals are equal, we have the desired result.
109 THE CHANGE OF VARI ABLES FORMULA AND APPLI CATIONS OF INTEGRAT IOI\l
9. Theintegrandsimplifi esto 1/(x+y). Theintegralis improperwhenever y== -x,wh'ich occurs
in our region at t he point (0,0). We compute
1
1
111_1_dydx =11 [ (InIx + yl)1 _]dx =11 [In(1 + x) -In(x)]dx,
o 0 x + y 0 y-O 0
which we interpret as
li m 11 [In(1 + x) -In(x)]dx.

Integrating by parts, we get
{ [((x + 1) In(x+ 1) - x) - (x Inx- x)]I: }
Using I'Hopital 's r ule, we know hat
I
I I' Inf 0
1m f nf = 1m -/= ,
E--+O+ [--+0+ 1 f
and so the integral is 21n 2.
11. Since x
2
+ y2 + z2 appearsin theintegrand, try usingsphericalcoordi nates. In thefirst octant,
we have
o () 7r/2 and 0 cp 7r/2.
Since x
2
+ y2+ z 2 a
2
, we also have 0 p a. Recall that the Jacobian for spherical
l
coordinates is p2 sine/; . Subs itute p2 =x
2
+ y2+ z2 and pcoscp =z to get
1f/21 ff /2 1 Q pl/2. l 1f/2 1 1f/21 Q p2 sincp
----r====p2smcp dpde/; d() = dpdcp d().
o 0 0 Jpcoscp + p4 0 0 0 Jcoscp + p3
Substitute u =coscp + p3:
3
1f/2 1 1f/2 j COS +a d n'+'
uSI '+' dcp d()
lo 0 3y'U cos
1f 2 3
r/
2
1/ (y'Ul
cos
+a ) dcpd()
Jo 0 3 u=cos
1f/2 r/2 2
o Jo "3 sinCPJcoscp + a
3
dcp d()
1
2 2
2 r/ r/
- Jo Jo 3" sinCP Jcoscp dcp d()
2
r/
2
[_i(coscp+a
3
)3/zl1f/ + i(cosCP)3/2' 11f/2] d()
Jo 9 =o 9 =o
27r [(1+ a
3
)3/2 _ a
9
/
2
_ 1]
9 .
14. When 0 g(:I:, y) f(x,y), integrationover D should give us
.lin 0dA .ILg(x,y) dA .ILf(x,y) dA.
However , since t he integral of 9 is smaller than the integral off and lIDg(x,y) dA does not
exist , we mustconclude t hat fi D f(x,y)dA also does notexist.
110 CHAPTER 6
16. Since x
2
+ y2 + z2 appears in the integral ,we will try touse spherical coordinates. The region
D can be described by
I:=;p and 0:=; O:=;27r and 0:=;:=; 7r .
In spheri cal coordinates, the integrand is 1/p4 andrecall that the J acobian is p2 sinr/!. There-
fore, we get
27f 1 7f Joo p2 sin
1
'----;-4-'- dpddO.
o 0 1 P
Since all ofthe limits of integration are constant , we may integrate each variable separately
and multiply the results:
{27f f"Joo sinriJ ( { 27f ) (f" . ) ( roo dP)
Jo Jo 1 ~ dpd<f; dO = Jo dO Jo smd Jl p2
27r ( - cosI:) b ~ ~ ( - ~ I:)=27r(2)(I) =47r.
SOLUTIONS TO SELECTED REVIEW EXERCISES FOR CHAPTER6
2. (a) This is a linear transformation ofa paralellogram, so by t heorem 1 ofsection 6.1 , all we
need todois to mapthevertices . WecomputeT(O, 0) =(0,0) ,T( 1,0) =(2,1),T( 1,1) =(2, 4)
and T(O, 1) =(0,3) . The transformat ion is depicted below:
y,
v,
(2.4)
(1,1)
(0.3)
T
(2,1)
r ~
x u
(b) Using the notation shown in t he sketch in part (a), our formulais
J10 f(u,v) dudv =J10.f(u(x,V),v(x,V))1~ ~ ~ : ~ ~ 1 dxdy.
In this case, u(x,y) =2x and v(x,y) =x +3y, so
a(u,v) 1=1 au/ox au/ay 1=1 2 6 1=
1
a(x,y) av/ax av/ay 1 3 .
Thus, our formulais
J10 f(u,v) dudv =6J10.f(2 x,x+ 3y) dxdy.
4. (d) Recall that t he Jacobian for spherical coordinates
is p2 sinqS and sin2qS =2sinqS cosqS. We factor out
the Jacobian from the integrand and we are left with
2pcosqS, which converts in rect angul ar coordinates to
2z. From our sketch ofthe region , we see that the
region ofintegration is over the disk ofradius ~ cen-
tered atthe origin, so we have
-I/V2< x < I/V2 and - J ~ ~ x-
2
< Y < J ~ -x
2
.
- - - - 2
y
x
111 THE CHANGE OFVARIABLES FORMULA AND APPLICATIONS OF INTEGRATION
Nowzextends from thecone to the sphere ofradius 1. Theequationof thecone is z =x
2
+y'l
and the upper hemisphere has theequat ion z =/1 - x
2
- y2 so
Therefore, the integral is converted to
Y2
1/-/2 lv' I-X
2
-
2zdzdydx.
]
-1/-/2 -Jl/2-x' E' +y'
5. The desired region is shown here. The surfaces inter-
sect where z =x
2
+y2 =2- z2 or z2 +z - 2 =0
or z = 1. We disregard z = - 2 as a sol ution
z
since we want z O. Looking at our drawing, we
can describe it in cylindrical coordinates as follows.
Notice t hat the entire region lies over the unit disk.
Also, the region lies between z = x
2
+y2 = r2 and
z =/2- x
2
- y2 =-J2=T2, so the r gion can be
described by
y
o r 1, 0 e 27l'
and
Since the Jacobian for cylindricalcoordinates is r, we get
8. For convenience, we will rotate C
2
so that its axis lies on the z axis. The volume remains
unchanged. Using the resul t of exercsie 12 in section 5.5, we substitute a =1 since the
diameter is 2 and the radius is 1. Thus, the answer is \6 .
11. When the cut is made by the pl ane x +y +z =a, the
volumeofthe (shaded) solid below t hat plane is
a
-
x
f
O
-
y
far -
x
fafO-x
ioio io dzdydx= io io (a- x -y)dydx
r [( Ill _Xl r( r
= io (a - x)y-
2 \
) y=o dx =io a x - dx.
Substitute u =a- x toget
x
1
y
Thus, the volume for the entire solid, when a =1 is i.Ifthe solid is to be cut into n equal
volumes, then t he volume under ;I: +y +z =a should be k/6n , where k is an int eger such
that 1 k n - 1. Therefore, a
3
/6 ::::: k/6n implies that cuts should be made in the planes
x +y+z =(k/ n)1/3.
112 CHAPTER 6
15. Let u=x+y, v =y- x or x =(u - v)/ 2, y =(u+v)/2. The reader should veri fy that the
Jacobian 18(u,v)/8(x,y)\ is Fromthe figures, 0 u :S 1and -u v u.
v y
(1,1)
u
(1,-1)
>
x
Then
u
Jlexp : Jdxdy
tjUexp(v/u)dvdu= t u ( ev/ul ) du
10 - u 2 210 v =-u
(e - 11U du = (e - .
16. The tot al mass is the integral of the density. Since we are integrating over a sphere, use
spherical coordinates. Thesphere ofradius R is described by
o P:S R, 0 r/J 1r and 0 0 21r.
The di st ance dfrom t he origin is the same as p and the Jacobian is p2si n. Therefore , the
totalmassis
2" 1"lRp2sinr/J
'----=- 3 dpdr/J dO.
1o 0 0 l+p
Since all oft he limits of integration are constants, we can integr ateeach variable separately
and multiply t he results. We get
(1
2
"dO) (1"sinr/J dr/J) (1
R
1:2p3 dP)= 21r [ - cosI:J =
19. (a) Let T( x,y, z) be t he temperature at (x,y,z). Then the average t emperature is
fffeTdV/ fffe dV. By definition, d= J x
2
+y2 +z 2, so d
2
= x
2
+y2 +z 2 and T(x,y, z)=
32(x
2
+y2 +z2). The numeratoris
32 [11 [11 [1
1
(x
2
+y2 +z2)dxdydz 32[:[ 11 [ +(y2 +z2)x) [=-J dydz
64 [1
1
[1
1
y2 +Z2 ) dydz
64 [ 1 +z2 ) y+ Y:J[=-J dz
1
j
1(2 ) (2Z Z3 )11
128 -1 3+z 2 dz =128 "3+3 -1 =256.
Thedenominatorisjustthe volumeofthe cube C, which is 8, so the average temperature on
Cis =32.
(b) The average temperat ure is attained wherever 32d
2
=32,i.e., d =l. Thus the average
temperatureis att ained at all pointsofthe cubeon asphere ofr adius1,centered attheorigin.
113 THE CHANGE OF VARIABLES FORMULA AND APPLICATIONS OF INTEGRATION
20. The inequality y2 + z2 tdescribes the inside ofa circul arcylinder ofradius t centered on
the xaxis . Theinequality (x- 1)2 +y2 +z2 1describes a ballofradius 1centered at(1,0,0).
We also want x 1,so we get he region shown. By symmetry, we easilysee thatfi = z= O.
For conveni nee, we will shift the region in question so thatitsaxisofsymmetryis the z'
axis and we will place the ball'scenter at the origin. For thecenter ofmassoft he new region
D+, it is the integral ofz over D" divided by he volumeofD- . Th region D* is described
in cylindrical coordinatesby
o< 7' < ! 0 <_ B_< 211" and 0< z<
- - 2' V! -
z
y
-1/2
.
z
112
The Jacobian is r,so the integral ofz over D is
211"11/21Yl-rl
r 1l" [1/2 drdB
rz" dzdrde =
1o 0 0 Jo Jo 2 z.=O
27r11/ 2 (r- r
3
)
- 2- drde
1o 0
11
2
". [(r2 r4 )11/ 2] _ 7 1211" _711"
- - - - dB - - de - -.
2 0 2 4 r=O 128 0 64
The olume f DO is


211"11/2 (
r dzdrdB rz drdB
o 0 0 1o 0 z =O
211" 11/2 . 1" [-(1-r2)3
/
2Il /2]
rvh - r2 drdB = de
1
2
o 0 0 3 r=O
(
! _ v'3) [2'" dB =211" (! _ .;3)
3 8 Jo 3 8 .
D* was shifted down t uni t, so forthe original region, xis
24. (a) Since we are integrating a spherical regi on, we will use spherical coordin tes and the
Jacobian is p2sin. We get
ffl (:1:
2
+ y2+ z2)xyzdV = 1
2
"1"'l
R
p2(pcos{Jsin4J)(psinesin)(pcos)p2sindpdde
114 CHAPTER 6
( r. r[R
10 10 10lcosOsinOcosJsin3JdpdJdO.
Since all of the limits ofintegration are constants, we can integrate each variable separately
and multiplythe results. Thus, the integral is
27r 7r
(l
R
p7 dP) (1 COS 0sin0dO) (1cosJ sin
3
J dJ)
2
(p;C)( sin 0C) (sin:J [) O.
2
28. Since x
2
+y2+z2 appearsin theintegrand,we will use spherical coordinates, and the Jacobian
is p2 sinJ. To integrate over all of we will integrateover asphere ofinfinite radius. Thus,
we describe by
o p<00, 0 J 1T, 0 0 21T.
Therefore,
(Xi r [21f p2 sinJ
J J !a/(x,y,z)dXdYdZ
10 1010 {,. dO difJ dp
00
1f
21f
(1 (1 + d
P
) (1sinJ dJ) (1 dO)
(1
00
(1 ( - cosJI:) ( oe)
00 p';idp
41T .
o (1 +p3)3/2
1
Substitute u =1+p3 to get
00 du 41T. 1b3/" 41T .
41T
1
-- =- hm u- du =- hm
(- Ju)I:= 81T/3.
1 3u
3
/
2
3 b-+oo 1 3 b-+oo
31. (a) Thisimproper integralisevaluated as follows:
Y
1
001
11
xexp(_y3)dxdy
100 (
2
x2
1 )
exp(_y3)dy
o 0 0 x=O
11
00 b
1
- y2 exp( _y3)dy =lim
(exp(_y3)I )
2 0 b-+oo -6 0
- 6'
33. Notice that x
2
+y2 occurs in the integrand,so we will use polar coordinates, which we recall
has Jacobian r. Consider to be a disk with infinite radius, so is described by
o rand 0 0 21T.
Thus,
[21f [00
J !(x,y) dA 10 10 r drdO
2
r.
121f [ dO =10
r
dO =21T.
115 THE CHANGE OF VARIABLES FORMULA AND APPLI CATIONS OF INTEGRATION
TEST .FOR CHAPTER 6
L Trueorfalse. If false, explain why.
(a) In polarcoordinates, thevolumebounded by the paraboloid z =a- (x
2
+ y2) with a >
and the plane z = is given by t he integr al
2
""l
Fa
-. r(a- r2)drdB =
rra2
8
100
(b) The xistenceoftheimproperintegralJ1
2
Jo
oo
co:x dxdyisguaranteedbyFubini 'stheorem.
(c) When changing from Cartesian coordi nates to pol ar coordinates, we have
1
1rv'4-y2 xdxdy=J,,"/6 [2r2cosBdrdB.
-l lo -,,"/6lo
(d) Thevolumeof the unit sphere centered at theorigin with t he cone z :=:; Jx
2
+ y2 deleted
IS
[211" [1rp2 sinif; dif; dpdB.
lo l ol 1f /4
(e) Theparallelogramwith vertices at ~ 2 2),(4,3), (0,5) and(2,6) canbe transformedtothe
square [0,1] x [0,1]by T(x,y) -+ ( x+2l -10, -x+;y-2) and itscorresponding Jacobian is
1
s
2
2. Compute t he integral of 1/(1+ x
2
+ y2 + x y2) over all ofJR
2
(Hint : Factor thedenominator.)
3. Aregioninspaceis described by:=:; x :=:; 3, - 2:=:; y :=:; 2and:=:; z :=:; ~ . Use cylindrical
coordinates to ompute the average ofzover the region.
4. Find the volume of the region which lies between the surfaces z = 10 - x
2
- y2 and
z =_(x
2
+ y 2 ) 2 and which lies within t hecylinder x
2
+ y2=4.
5. LetS be the t ransformation defined by S :(x,y,z) -+ (u
2
, uw,V + w). Suppose this transfor-
mationtakes Dinto t he region [0,1] x [0, 2]x [-1,1]. ComputeJJJD(y +z)dxdydz.
6. As wit hparallelograms,alinear ran formationwillmapverticesofa parallelepipedto vertices
ofanother parall elepiped. Let T be the t ransformation T : (x , y, z) -+ (3u +2v+ w,2u -
w,v+2w) and let Pbe t heparallelepiped withvertices at (0,0,0)' (3,2,0), (2,0,1), (1,-1,2) ,
(5, 2, 1), (3,- 1,3), (4,1,2) and (6,1,3).
(a) ConvertJJJp(x - y- z)dxdydz to tJvw coordinates. [Hint : Map the points (x,y, z) ==
(0,0,0),(3,2,0), (2,0,1) and (1,-1,2).]
(b) Evaluate the triple integral.
7. Integrate x+ 2y over t he interior ofthe ellipse (x - 2)2/4+ (y - 1)2/9 = 1 by making an
appropriatesubsti tut ion and then using polar coordinates.
8. The shap of a swimming pool can be described by :=:; x :=:; 2, ~ 3 J l - x
2
/4 :=:; y :=:;
3J1- x
2
/4and - J1- x
2
/4 - y2/9 :=:; z :=:; O. Compute the volume by first substituting
u =x/ 2and v=y/ 3and t hen changing tospherical coordi nates.
9. Explain whatis wrong with this calculat ion:
[lJY"!. dxdy=t In Ixll
Y
_ dy =[ 1 (In lyl- lnIY!) dy= tody =0.
l o -y x lo l o lo X_-Y
10.A near-sighted mosquitosaw a t attoo on a man's arm which appeared to have the combined
form ofa rect angle, [-2, 2] x [0,1], and a semiellipse described by x
2
/ 4+ y2/9 :=:; 1, y:=:; 0.
Where should the mosquito aim its probiscus to consume blood out ofthe tattoo's center of
mass? You mayassume t hat t he mosquitosees a planarsurface and t he tattoohas a constant
density.
117
7 INTEGRALS OVERPATHS AND SURFACES
1.1:THE PATH INTEGRAL
GOALS
1. Be able to compute a path integral.
STUDYHI NTS
1. Not ation. The pat h integral is denoted by Ief ds, where ds =Ile'(t)lldt and e(t) is a path.
2. Definition. Thepathintegralistheintegralofreal-valuedfunctionsoveracurve. Theintegrand
f only needs to bedefined on c andf must be piecewise continuous on e. In addition, e needs
to be piecewise continuously differentiable. The integration of vector-valuedfunctions will be
discussed in section 7.2.
3. Computation. If e(t) is defined for a t bandf is defined on e, then the path integral is
lbf( c(t)) lie' (t)11 dt.
4. Relation to arc length. If f = 1, then the pathintegralis thearc length formulaofsection 4.2.
SOLUTIONS TO SELECTED EXERCISES
2. (b)The path in egral is defined by
lbf(e(t)) Ile'(t)111dt.
In this case, f(e(t)) =cost ,e' (t) =(cost,- sint, 1) and lie' (t)11 =[cos
2
t+ sin
2
t+ 1P/2== V2.
Therefore, t he pathintegral is
2.. 1271'
1cost.v'2dt =v'2sint 0 == O.
3. (b) Let x =t, Y=3t andz =2t. Here, f =yz =(3t)(2t) and ds =Ile'(t)11 =[12+3
2
+22P / Z=
V14. Then the pat h integral is
3
3 3 2 1
1f ds =J(3t)(2t)V14dt "'" 6v14j t dt ::::; 2v14t
3
1 =52Ji4.
6. (a) Recall that in polar coordinates,
x =r(O) cos0 and y = r(O) sinO.
Heie we treat r as a function of0, andso
dx =(r'(B) cosB- r(B) sinB) dB and dy =(r'(B) sin0 + r(B) cos0) dB.
118 CHAPTER 7
Also, we know that
ds = Vdx
2
+ dy2 dO
= ([7/(0) cos0- r(O)sin(W + [1'1(0) sin0+ 1'(0)cos0]2}1/2dO
= [(1" (0))2(cos
2
0+sin
2
0)+(i(0))2(cos
2
0+sin
2
0)j1/2dO
= Vi2 + (drjdO)2dO.
Hence, t he pat h integraloff(x ,y) =f(1' cos0, rsin0) is
11(x,y) ds = f(rcose, l' sinO)V i
2
+ rde.
7. (a) From t =- 1 to t =0,t he pat h is a straightlinegoingfrom (1,1,1) to (0,0,0);from t =
to t =1, t he pat h is a straight line goingfrom (0, 0,0) to (1,1, 1), so act ually the same path
was covered t wice. Hence, the path integral is
lb f(x (t ),y(t))y'(x'(t))2 + (y'(t))2dt = 211(2t4 - t
4
)V {4t
3
)2 + (4t
3
)2 dt
= 2114t 4Y'2t6dt =8h11 t
7
dt =h.
Theanswer could bethought ofas the areathatneeds to be covered ifone wants topaintthe
two sides ofa fence erected along the path x =t4,Y =t 4, ofheight f( x( t),y(t)) .
(b) The arc length function is the pat h integral when f =1, so it is
lb f( c(r))ll c'(r)lldr ,
where a is the st arting point. To evaluate s(t), we need to break up the interval into [-1,0]
and [0 , 1] . For - 1 r 0, Il c' (r)1I =[( 41"3 )2 + (4r3)2j1/ 2=4V2I1"13 =-4y1r3. Then
s(t) = - [tl 4V2r3d1" = - V2(t
4
- 1) .
Thus, t
4
=- s jy1+ 1and4t
3
dt =- ds/y1. When t =- 1, s =0;and when t =0, s =y1.
For :::; r 1, Il c'(r)11 =[(41"3)2 + (41'3)2]1 / 2=4y1r3. We also need to add the length from
- 1 to 0,so we get, for :::; t :::; 1,
s(t) =V2 +1t 4V21"z d1' =-/2(t
4
+ 1).
So t
4
=s/V2- 1and 4t
3
dt =ds/y1. Whent =0, s =V2; and when t =1, s =0. Putting
all this toget her, thepath integral becomes
1V2(- ds+ ds = +
10. (a) When density is a constant k, we have
Mass =k x lengt h ofwi re.
Thelength ofthe wire is
lb Il c'(O)11 dO,
119 INTEGRALS OVER PATHS ANDSURFACES
so the mas is
(b) Recall that a coordinateofthecenter ofmass, Xj, is
Here, we have a thin wire in the yz plane. By symmetry, x =z =O. In this case, y =a sin0,
so the center ofmassfor y, with k =2, is
f kyds fo" asin(J 2adO _ a 1". 0_ 2a
-f-- = - - smOd --.
kds 2a7r 71' 0 71'
Thus, the center ofmass is (0,2a/7r, 0) . Note that the center ofmass is noton t he wire.
13. The t rick of t his xercise is to use the correct coordinate system. Th intersection of the
sphere and th plane is a unitcircle, butinan unfamiliarorientation. The unit normalvector
ofthe pl ane x+y+z =0 is (i+j +k)/J3,so we need tofind two orthogonal vectors in the
plane x +y+z =O. For example, (1,0,- 1) and (1,- 2,1) are a pair. Normalizationof h e
two vectors yields u = (1,0,- 1)/0 and v = (1,- 2,1) / V6. The circle is now parametrized
as c(B) =(x, y z) =cosBu +sinOv, 0 0 < 271' . Take the first component of c to get
x = (cosB)//2+(sinB) /)6.Then the mass is
2
2
1x ds = 1" Il c'(B)l ldB
2
1" cosB+ sinB)2 . 1dB
2.. (1 20 2sinBcos0 1. 2 B)
o 2cos + v'i2 +6 ill dB.
1
You should verify that Ilc' (B)11 =1. Now use the half-angleformul as to get
7.2: LINE INTEGRALS
GOALS
1. Be able to compute a line integral.
2. Be able toexplai n thedifference between a line integral and a path integral.
STUDYHINTS
l. Definition. The line int gral deals with vector fields whereas the path integral dealt with
scalar funct ions. Itis required that c be Cl and F be continuous on c (or at least piecewise
continuous for both c and F) .
2. Physical interpretation. Thelineintegralis mostcommonlyinterprete as the work performed
by F along a path.
120 CHAPTER 7
3. Computation. IfF is defi ned on c and c is defined fora t b, then theline integral is
lb F(c(t )) .c' (t ) dt.
Note that the integrand is a dot product, so the integral is a scalar . The formula involving
T (t ), the unit tangent, maybe moredifficult touse since you need tocomputeT (t) as well as
Il c'(t)ll
4. Sign interpretation. Posi tive work means that t heforce fi eld did a net amountofwork on t he
object; such is the case when the motionof the object isin thedirection oftheforce. Ifwork
is negative, then this amountofwork is done by theobject on the force field.
5. Reparametrization and orientation. Ifc is reparametrized and the orientation is preserved,
the value of a line integral does not change. Ift he orient ation is reversed, only the sign
ofthe valueof the line integral changes. Youcan substitute for t he endpoints to be sure the
direction iscorrect. Orientationisnotimportantwiththepathintegraloftheprevioussection.
Reparametrization allows us to break up a curve and integrate over each segment separately
and with convenient parametrizations (see example 11). Be sure the path is traversed the
correct number of times.
6. Gradients. The line integral ofa gradient field depends only on the endpoints. We will see
howto make use ofthis fact in section 8.3. Given this fact, we can see that t he line integral
ofa gradient along a closed path is O.
SOLUTIONS TO SELECTED EXERCISES
1. (a) The line integral ofF alongcis
11 F(c(t)) .c'(t ) dt .
We have c(t} =(t, t, t), so c'(t) =(I ,I,1) and F( c(t )) =(t, t, t ). Therefore, the desired li ne
integralis
1
1 11 11 3 3
(ti +tj+tk ) .(i+j +k)dt = (t +t +t)dt =_t
2
= - .
o 0 2 0 2
2. (c) We need tobreakuptheintegralintotwoparts. Forthesegmentj oining(1, 0,0) to (0, 1,0),
the easiest way to parametrize the line is t ofi nd c(t) =u(t) +t v (t ) such t hat c(O) =(1,0,0)
and c(l) =(0,1,0) . Letting u(t ) =(1,0,0) and v (t )=(0,1, 0) - (1,0,0)=(-1, 1, 0) gives us
C1(t ) =(1,0,0) +t (- 1,1, 0) or (x, y, z ) =(1- t, t, 0) for 0 t 1, so dx = -dt, dy =dt and
dz = O. Substitute these values to get
1
(yz dx +xz dy +xydz) =11[t 0(- dt ) +(1 - t) 0dt +(1 - t) .t 0] =O.
C l 0
For the line segment joining (0, 1,0) to (0, 0,1), use the same method. We find C2 (t)
(0,1,0) +t(O , - 1,1)or (x, y, z) =(0,1- t, t ) for 0 t 1,so dx =0, dy =- dt and dz =dt.
Substitute these values toget
1
(yz dx +xz dy + xy dz ) = { I[( 1- t) . t 0+0.t .(- dt ) + 0.(-t) .dt] =O.
h
Addi ng the two li ne integrals together, we get 0+0= O.
4. (a) By definition, we have
1Fds =1F( c(t) ) .c'(t) dt .
121 INTEGRALS OVER PATHS AND SURFACES
We are given thatF ande'areperpendicular, so Fe'(t) =O. Therefore, t helineintegralis O.
(b) Again, we use the definition oftheline integral. In addition, we use the definition ofthe
dot product to get
1F ds = 1F(e(t )) .c'(t )dt =lIIF(e(t ))llll e'(t)II'cos(O) dt = l IIF"ds.
This result because the cosine between parallel ve tors is 0and by defi nition ds = Ilc'(t )11 dt.
7. We aregiven that c(t) =(x,y,z)=(t,tn, O), 0$ t $ 1,where n =1,2, 3,.. .. Differentiationof
eachcomponentgivesus c' (t) = (1,nt
n
-
1
, 0) ,sodx = dt,dy =nt
n
-
1
dt anddz =O. Therefore,
1[ydx+(3
y3
- X)dy + z dz
1
= fa\ tn dt+(3t3n _t)ntn-1 dt+0)
fa1 (t n+ 3nt
4n
-
1
- ntn)dt

3 n-l
1+n 4 0 4'-I+n'
11. Given thate(t) is a path and T is a unit tangent , we have

Ile'(t)11
by definition. Since v v =Il v11 2, we get the li ne integral
r re'(t) , rlIe'(t)112 l '
Jc T ds = Jc Il e' (t )11 .e (t) dt = Jc Ile'(t)11 dt = c li e (t)/I dt .
The lastintegralis simply the arc length ofc(t).
14. If a vector field is a gradient, then the integral only depends on t he endpoints. For a closed
curve, thestartingpointis thesameast h ending point,so theorem3tells us thatt heintegral
is O.
17. From the fact that V(I/7' ) = - r/ r
3
, we see that F =Vf , where f = l/ r. According to
theorem 3, the line integral is
1 1
f(c(b)) - f(e(a)) =-R - -R .
:! 1
7.3: PARAMETRIZED SURFACES
GOALS
1. Be able to parametrize a given surface.
2. Be able todetermine ifa surface is regular and/ordifferentiable.
3. Be able to compute a tangent plane for parametrized surfaces.
STUDY HINTS
1. Parametrized surfaces. Recall thatcurves in the plane can be parametri zed by wo functions
ofone variable, namely, x =f(t ) and y =g(t). A curve in space is given by t hree functions
ofon variable. Now, we extend t his idea to surfaces. We do his by letting J,' , y and z be
fune ions of two variables, i.e., now we hav t hree functions of t wo variables, x =f(u,v),
y =g(u, v) and z = h(u, v).
122 CHAPTER 7
2. Differentiable surfaces. Such a surface is parametrizedby (f( u,v) , g( u,v),h(u,v)) and j, gand
h are all differentiable functions.
3. Review. Recall t hata vector n which is normaltoa planegives thecoefficientsoftheequation
ofthatplane. See Section 1.3.
4. Tangent vectors. Like a partial derivative, we can hold uconstant to get T I) (fl), g",hI)).
Likewise, holding vconstantgives us Tu =(fu,9u,hu).
5. Tangent plane. By holdingeither parameter, u or v, constant, we get a curve on the surface
and ifthe cur ve is "nice," i t will have a t angent line. Simil arly,we can get another t angent
lineby holding the other parameterconstant. Those two tangentlines determine the tangent
plane, and t he normalvector to t he plane is TuxT".
6. Regular. A surface is said to be regular ifTuxT" f:. O. When the cross product is not
0, a tangent plane exists and consequently, t he surface has no pointed regions (li ke a cone).
Differenti ability does not implyregular (see exercise 1 i), but regularness requires differenti a-
bility,for otherwise Tu and T" would notbedefined.
7. Important parametrizations. Know the followingparametrizations:
(a) Circle of radius r: x = rcosO, y= rsine,0 f) 21T.
2
(b) Thehyperbola x - y2 =1: x =cosht, y =sinht.
(c) Sphere ofradius p: x = pcosOsinifJ, y = psinBsinifJ, z = pcosifJ, 0 0 <21T, 0 ifJ <1T.
Thisis thesameas the definition ofspherical coordinates.
(d) Asurface z = g(x,y): x = u,y = v, z = g(u,v).
8. Normal to a graph. If z = ! ( x,y), the normal to t he graph is (- - %&' 1).This is useful
to know for the comingsections. Note thatifwe letg= z- !(x,y), then the (unnormalized)
normal to the surface is n ='il g.
SOLUTIONS TO SELECTED EXERCISES
3. Thenormaltot hedesiredtangentplaneisTuxTI) ' WecomputeTu= (/Jx/ /Ju,/Jy/ /Ju ,/Jz/ /Ju) =
(2u,sine", cose" ). Simil arly, T" = (0,e"ucose
V
, - e" ). We compute
i j k
Tu xTv 2u sine" 1 cose"
3
o ue"cose" (-u/3)e"sine"
2
(( - u/3)e"sin
2
e" - (u/3)e
V
cos eV) i+ ((2u
2
/3)e"sineV )j + (2u
2
e" cose")k
(u/3)e"(- i+ 2usine"j+ 6ucose"k).
To find the tangent plane, we only need to evaluate - i + 2usine"j + 6ucose"kat t he given
point (uo,vo). However, we are only given (ZO,yo, zo), so we must be clever . We know tha
usine" =- 2 and (u/3)cose
U
=1 at the given point , so 2u si n e" =- 4 and 6ucose" =18.
Hence, the tangent plane is
-(x - 13)- 4(y+ 2)+ 18(z - 1) =0, or 18z - 4y- x =13.
4. (1)Asurfaceis regularifTuxT" f:. O. We calculateTu = 2i + 2uj +Ok and T" = Oi+j+2vk,
so
j k
Tu X T" =I 2 2u 0 I= 4uvi- 4vj+2k,
o 1 2v
123 INTEGRALS OVERPATHS AND SURFACES
which is never O. Therefore, the surface is regular.
(2) Again, we calculate T... =2ui+j +2ukand Tv =- 2vi+j +4k, so
i J k
T... x Tv = 2u 1 2u = (4 - 2u)i- (8u +4uv)j-+ (2u +2v)k.
- 2v 1 4
Ifu =-v=2, then T... x Tv =O. Therefore, the surface is notregul ar at (0,0,-4).
7. From t he partials ofx, y and z with repsect to tI , we compute T... = j,and similarly, Tv
(cosv) i - (sinv)k. Then
j k
T uxT v = 01 0 =-(sinv)i-(cosv)k.
cosv 0 -smv
By our luck, the magni t ude ofthis vector is 1 already, so a unit normal is n = -(sinv)i -
(cosv)k. [Notethatanother unitnormalis (sinv)i+(cosv)k,dependi ngon whichcrossproduct
(T... x T v or 1\ x T... you take.]
2
To identify the surface, note that x
2
+z2 = sin
2
v+cos V = 1. As vvaries between 0and
21l" , this becom s a circle of radius 1centered at the origin on the xz plane. Since y = u, we
have -1 Y 3. Therefore, t he surface is a cylinder ofradius 1 centered around the y axis
and it extends bet ween y = -1 and y =3.
9. (a) Parametrically, the surface is described by x =h(y,z), y =y , z =z . In this case,
Ty = (hy,1,0) and Tz = (hz,0,1),so
1 j k
Ty x Tz = hy 1 0 =(1, -hy,-h
z
).
hz 0 1
The vector T y x Tz is normal to t he tangent plane, and so the tangent planeis
(x - xo) - hy(Y- yo) - hz{z- zo) =0,
where hy and hz are evaluated at (xo, Yo, zo).
11 . The surface is parametrized by (x, y, 3x
2
+8xy), so Tx = (1,0,6x+8y) and Ty = (0,1,8x).
At t he point (1 , 0, 3),T x =(1,0,6) and Ty = (0,1,8). The normalvector is given by
j k
n = Tx x Ty= 1 0 6 = -6i- 8j+k.
018
Thus, the desired equationis (x - Xo, Y- Yo ,z- =0) .n =0; that is, 0 =(x - 1,y- 0, z- 3) .
(-6,-8,1) =-6x- 8y+z +3, i. e., z =6x+8y- 3.
Thinki ng ofthe surface as the graph oft he equation f( x,y) = 3x
2
+8xy, we calculate
Oflox = 6x+8y,ofloy=8x,which become6and8,respectively, atthepoint (x,y) =(1,0).
Then the tangent line is z - 3 =6(x- 1) +8(y- 0); that is, z =6x+8y- 3.
13. (a) When 0= 0, 4>(1',0) is t he line segment from (0, 0,0) to (1 , 0, 0) . As 0 increases, the line
segment rotates around the zaxis and the segment also moves to t he height z = O. Thus , we
get the sketch ofa helicoid. Thesketch in figure 7.4.2shows a helicoidfor 0rangingfrom 0 to
21l". The hel icoid for 0 0 41l" is similarexcept the graph extends to z =41l" and makes an
extrarevolution around the z axis.
(b) Differentiating the components of <I (r, 0) gives us Tr (cosO,sinO,O) and
To =(- rsin0,rcos0,1). A normal vector to 4> is
1 j k
cos0 sin0 o =(sinO,-cosO, r). T r x T o=
- rsinO rcosO 1
124 CHAPTER 7
Bence, t he unit normal vector is (T
r
x T9) /II T
r
x T911 = (sinB,- cosB,
(c) Multiply the vector Tr x T ocalculated in part (b) by r to get another normal vector:
(rsinB,- rcosB,r
2
)=(y,-x,x2+y2).
Then the equation ofthe tangent plane at (xo,Yo, zo) is
Yo( x - xo) - xo(y - Yo)+ +Y6)(z- zo) =0.
17. (a) Itis obvious thatthe is the xyplane (or the uvplane, ifyou prefer). Since u
3
and v
3
can have any real value, and the third coordinate is 0, the image of also is the xy
plane.
(b) A surface is not regular is li T" x Tvll = 0. For we have T" = i and Tv = j, so
li T" x Tvll = Ilkll = 1. Therefore, descri bes a regular surface. For T" = 3u
2
j and
Tv = 3v
2
j. Ifu = or v = 0, then T" x Tv = o. Bence, is not regular. (This problem
illustrates t hat "regularness" depends on parametrization,not necessarily the image.)
(d) The answer is no. A "regular" parametrizat ion cannot "round out" the "corners" ofa
graph. Indeed, by (c), a regular parametrization gives a smoothsurface.
7.4: AREA OF A SURFACE
GOALS
1. Be able tofind thesurface areaofa givensurface.
STUDYHINTS
1. Riemann sum argument. Ifyou underst and the Riemann sum argument ,you should be able
toderi ve thesurface areaformula.
2. Formulas. Youshould know at least onesurface areafor mula. One is
A(S) = J10 liT" x Tvll dudv.
Equation (3) follows immediately from this formula, and (4) is a special case of (3). Ifyou
choose to remember equation (3), note that the integrand has the Jacobian matrix of each
possible pairchosen from (x,y,z),i.e., (x,y),(y, z) and (x,z).
3. One-lo-one. J ust as wit h curves, the chosen parametrization must be one-to-one; otherwise,
thesurface may becovered more thanonce. See exercise 2.
4. Sur/ace area 0/a graph. Ifz = !(x,y), use the parametrization x = u, y = v and z = !(tt,v).
Then thesurface area formulabecomes
A(S) =J10 /11 + +1dA.
This is important to know oryou should be able to derive itfrom ffDli T" x Tvll dudv.
SOLUTIONS TO SELECTED EXERCISES
2. Usingformula(3), we compute the followingdeterminants:
8( x,y)
8x/8B Bx/B I_I -sinBsin cosBcos</> / _ . ; </>
8(B, )
By/8B By/B</> - cos8sin sinBcos - -sm cos,
/
8(y, z)
cos8sin sinBcos / _ . 2 A, (j
_sin - - sm If'cos ,
8(B, )
/
125 INTEGRALS OVERPATHS AND SURFACES
and
8(x,z) _ 1- sinBsinJ cosOcosJ 1_ . 2 ", . l1
8(e,J ) - 0 - sinJ - sm 'I'sm(l.
Squaring the abovedeterminants, adding themand taking the square root gives us
liTe x T1>11 = j sin
2
I,b cos
2
J + si0
4
tP(cos
2
B+ sin
2
B) = j Sin
2
( cos
2
tP + sin
2
I,b )= 1 sintP l
Ifwe allowJ tovary from - 7r/2 to 7r/2, then we get
2
2'1fl1f/ 2
Isin4> 1ddO = 27r [1
0
(- sinJ)dJ + r/ SintPdl,b]
1
o - 11" /2 - 'If/2 io
27r [ cOSJIO + (.- cos4 1'If / 2] = 47r .
- 'If/ 2 0
The answer is 47r because when 4> varies from -7["/ 2 to 7r/ 2, t he parametrization covers the
upper hemisphere twice, so thesurface area is the sameasone compl tesphere.
Ifwe vary I/; from 0to 27r we should get 87r ,since the sphere is paramet rized twice:
2
11[1211" IsinIdJ dB = 27r [111" sintP d + 12'1f (- sinJ) d]
27r[ (- cosJ)I : + costP C ] = 27r (2 +2)= 87r .
Ifwe hadforgo ten theabsolute values and simplyintegrated sin1/;, we would have computed
a surface area of0in bothcases, but tills is not possible. Surface areas have to be positi ve.
5. The ar a of of/ (D) is lID II Ttl x TvII du dv. Given of/(u, v) = (u - v,u + v,u'v) ,we get
8(x,y) = I 1 1 1= 2; 8(y, z) = 1-1 1 1=- u_ v; 8(x,z) = 11 1 1=u- v.
8(u,v) -1 1 8(u,v) v u 8(u,v) v u
Squaringt hedeterminants, addingthemandthentakingthesquar rootgives ustheintegrand
II Ttl x Tvll = V2
2
+ (- u- V)2+ (u- v)2=V4+ 21.12 + 2v
2
=V2Vu
2
+ v
2
+ 2.
Since we are integrating over the unit disk, use polar coordinates. Let 1.1 =rcosB and v =
rsin() ,0 ~ r ~ I ,0 ~ B 27r. Then
JLh Vu
2
+ v
2
+ 2dA = 111211" h Vr2+ 2rdBdr= 2V27r11 v,z+ 2rdl'
2
2V27r' i (r
2
+ 2)
3/2
1: = 2V;7r(3
3
/ - 2 3/2) = i (6V6- 8).
8.We choose the parametrizationx = ~ , y =t, Z =z, 0:S Z ~ 1,- 1~ t :S 1. Tocompute
t he sur face areaintegrand, we first compute
8 x , y ~ 1=0,
8
(y,Z) I- l 18(x,Z)I - t
1
8(t, z } 8(t, z) - , 8(t, z) - Jt2+T'
1
t hen,
J
2
[ffi t t2 +1
II Tt x TzII= 0
2
+ 12 + --Z-l dtdz = - 2--dtdz.
t + t +1
Thus , thesurface area is
+1 11 [ffi
1
111 [ffit 2 _ t2 +1
- 2- - dtdz - - 2- - dt.
a -1 t +1 -1 t +1
Wedo not attempttosimplifyfurther. An al ternativeparametrizationis touse thehyperbolic
functionssinht and cosht; theintegral onlygets nastier.
126
(H APTER 7
10. Let x=ucosv, Y = f(u), z=usinv,a:::; u:::; b, 0:::; v:::; 2r. . The reader should veri fy that
8
(X,Z) 1= u.
8
(X,y) 1= - U!, (U) sinv,
8(y,z) 1=u!,(u ) cosv, and 8(u,v)
8(u,v) 8(u,v)
1 1 I
Thus, the surface area is
A(S) =JIn.ju
2
+ u
2
(f'(u)) 2 du dv.
Since the integrand does not depend on v, the v integral can be performed, and we get the
desired formula:
A(S)= 211" lb lul.jl+ (f'(u))2 duo
We arerotatinga. curve about the y axis,so consider thedistance from the y-axis to thecurve
as t he "height ," which is Ix I. Thus, a cross-sectional circumference of the surface at a fixed
Yo is 211" 1xl Next, describe the cur ve y = f (x ), a < x < b as a path c(t) = (t, f (t )) . Then an
infinit esimalarc length can be expressed as J1 + (f'(t ))2 dt or simply ds . The surface area
is obtained by integrating the cross-sectional circumferences along the path c and the above
formul areduce to A(S ) =Ie 2r. lxlds.
13. We are interested in the areaofthesurface z(x, y) =f(x, y) =1- x - y, inside x
2
+ 2y2 :::; l.
First,compute
VI+ r; + fJdx dy =v'3dx dy.
To compute thesurface area, we need to parametrizethedisc z =0, x
2
+ 2y2 :::; 1using polar
coordinates: x =rcosB , y =(r/V2)sinB, 0 :::; r:::; 1, 0 :::; e:::; h, and the Jacobian is r/V2.
Our integral then becomes
1
2" fl
o Jo v'3dr dB = r. V6
2 .
17. Completing squares, the equation x
2
+ y2 = x becomes
(x
2
- x+ t )+ y2 =t, i.e. , (x - t )2+ y2 = This
equation represents a cylinder whose base circle is centered
at 0) wit h radius as shown. To find the surface area
of5] , we need to consider where the cylinder "sticks out"
of the sphere. Consider the positive octant. The surface
area is IID ) 1 + r;+ f3 dx dy , where D is halfofthe base
2
circle (shaded), and z = f (x, y) = J1 - x - y2 is the
sphere. Since wewi ll beintegratingoveracircularregion, we
can use polar coordinates: x
2
+ y2 =x is the same as r2 =rcosBor r =cosB. From the
figure, one can see that D is described by :::; r :::; cosBand 0 :::; 8 :::; r. / 2. Also, we com-
2 2
pute f x = - x/J1 - x - y2 and by symmetry, fy = -y/Jl- x - y2. So ) 1 + n+ =
2
1/ J1 - x - y2 , whichbecomes in polar coordi nates. RememberingthattheJ aco-
bian is r and that SJconsists offour equal surfaces, we get
I
cos e
) 1f /21cos e r 1 7r / 2(
A(St}
4 v'1=T2 drdB =4 - dB
1o 0 1- r2 0 r=O
f7r/2 7r/2
4Jo (l-sinB)dB= 4(8+cosB)lo =21T' - 4.
By high school geometry, we know thatA(S:d =411" - (211" - 4) =211" + 4,so A(5
2
)/A(SI) =
(r.+ 2) /(r. - 2).
".. ,-
x
127 INTEGRALS OVER PATHS AND SURFACES
20. First, we aregoingtofigure thevolumeofthematerial removed tomakethe hole in thesphere.
Use cylindricalcoordinates todescribe the hole:
0::; r ::; 1, 0 ::; B ::; 21!' - ::; z ::;
The limits for z were fou nd from the equations of the upper and lower hemispheres: z =
J4- x
2
- y2 and z =- /4- x
2
- y2 and substit uting 1' 2 =x
2
+y2. Remembering that the
Jacobian is r, we get
Vhole = ('" l ' dzdrdB=t " t2r/4- r
2
drdB
Jo Jo Jo Jo
2 2
1" ( -2{4 1'2) 3/2[=JdB= - 3\1'3)1" dB = 4;(8 - 3\1'3).
Since thevolumeofthesphereis3211" / 3, th volumeofthecoupler is3211" /3 - (3211" /3 -4rrvS) =
4rrV3.
Forthesurfacearea, itsuffices tocalcula ethesurface areaofone "cap" ofthehole. Inrect-
x
2
- angularcoordinates, the surface areaofI (x,y) =J 4- y2 over D,the ci rcle ofradi us 1,
2 2
is ffD VI +Ii+ dxdy. We calculate Ix =- x/J4- x - y2and Iy =-y/ / 4- x - y2,
x
2
so VI +Ii +n:= 4/J4 - - y2, Changing to polar coordinates, the surface reaofone
capis
2 2
1
2"11 h drdB =1" [ - ] dB =(8 - 4V3)1". dB:= rr(16 - 8V3).
w
o 0 4- r 0 r=O 0
Si.nce the surface area ofthe sph re is 411"1'2, or 16rr, and the surface area ofthe two caps is
2rr(16 - 8V3), the outer surfa.ce areaof t he coupler is 161!'(V3-1).
22. (b) We computel:r: = y+ l/( y+ 1) and Iy =x - x/(y+ 1)2. Thus
1+ (y2 + y + 1)2+ (x(y+ 1)2 - x)2 dA
VI+Ii+fJ dA =
(y + l )2 (y+l)4
(
1
)2 J(y+ 1)4 + (0+ 2y2 +2y +1)2 + (x(y + 1)2 - xFdA,
y+l
and the surface areais
1
412 1
( \2 / (y + 1)4 + (0+ 2y2+ 2y+1)2 + (x(y+ 1)2- x)2dxdy.
1 1 Y+I,
7.5: INTEGRALS OF SCALARFUNCTIONS OVERSURFACES
GOALS
1. Be able tocompute theintegral ofa given scalarfunction over a given surface.
2. Understand why the integral is defined the way it is and tointerpret it physically.
STUDYHINTS
l. Notation. In t his section, we introduce dS, which stands for li T" x T vII dudv, which was
discussed in section 7.4.
2. Importance. Thissection and the next will be used extensively in chapter 8. In this section,
we integratescalarfunction as we didinsection 7.1, andin the nextsection, we will integrate
vector-valued functions .
128 CHAPTER 7
3. Computation. The formulafor the scalarsurface integral of a scalarfunctionf is
fisfdS=flf liTu XTtlll dudV.
Here, fis usually given as a func tion of(z,y, z)and we rewrite it in t ermsoft he parameters
tI and vby substituting x, y, z as functions oftI, v.
4. Physical interpretation.
(a) Iff =1,then we get thesurface area. This may helpyou t o remember t heformula.
(b) Iffis the massdensity per unit area ateach pointoft he surface, we get the massofthe
surface.
5. Scalar integral over a graph. Ifz =g(x,y),then the formulabecomes
fl f(z,y,g(x'Y))V
1
+ (:!r + ~ : r dxdy.
You should remember this or be able t o derive itfrom lID fli T... xTvlldudv.
O. Integrating over a plane. IfS is a plane, we can simplify the integrati onformulain equat ion
(5) ofthe text :
fisf dS =fL~ () dxdy ,
where cos() =D .k and D is the unitvector normalt o theplane. (Review t he geometryofthe
dot product, section 1.2. ) Theregion Dis the projection ofSonto the xy plane.
SOLUTIONS TO SELECTED EXERCISES
3. Since we're integating over a hemisphere, it is wise to use spherical coordinates. For th
hemispherical surface, we have p = a,so x = acos() sin</>, y = asin() sin4> and z = acos</> for
o~ () ~ 271" and 0 ~ </> ~ 71"/2 . Thus
liTexT>Il II(- asi n() sin</>i +acos() sin</>j - Ok)
x(acos()cos4>i +asin() cos</>j - asio</>k)11
a
2
sin</> .
Then
'{ {2'1f r
l 2
2
f1
zdS
JD acos</> II
T
e x T4> 11 d4> d8 = J J acos</> .a sind4> dB
J
r
o o
/2
( . "/2 3 3 ry )1
271"a
3
Jo si n4> cosd = 21ra sm; 0 = 7I" a
5. (a) The equation ofthe sphere is x
2
+y 2 +z2 =2Rz, R >O.
Upon completing the squares, we see t hat it is equivalent to
z
the equati on z2 +y2 +(z - R)2 =R2, wbich is a spbere of
radius R centered at (0,0,R). The tip of thecone z2 =x
2
+y2
intersects (by design) thesphere at theorigin. Tofind theother
intersections,notethatx
2
+y2+(z - R) 2 =R2 and x
2
+y2 =z2
implies R2 - (z - R)2 =z2or z =R. Parametrize t he cone as
follows: x =p cos B, y =psin8, z =p wi th 0 ~ p ~ Rand
o~ B 271" . You should verify that II Tp x Tell =V2p. Thus ,
theareaoft he portion ofthe cone that is inside the sphere is
x
y
1
2"lR p2 1R
V2pdp dB =271"V2- =7I" V2R2.
o 0 2 0
129 INTEGRALS OVER PATHS AND SURFACES
(b) By "areaof the portion of the sphere inside the cone," the authors presumably mean the
areaofthepiece ofthesphere that isthe "icecream" partofthisconfiguration. This is si mply
the areaofthe hemisphere ofradius R, or 21l'R2 .
7. Parametrize S usin polar coordinates: x = rcosB,y = 1'sinB, 0 ~ r ~ 1, 0 ~ B 21l' . Since
the surface isdescribed by z = x
2
+ y2,we substitute in x and y, and get z = 1' 2. Then,
T r x T 8= (cos B,sin B,21' ) x (-1' sin B,rcosB,O) = (-2r
2
cosB,-2r
2
sinB, 1' ),
and so
dS= II T r x Telld1'dB= V41'
4
+r
2
d1'dB=1'V4r
2
+1d1'dB.
Calcul at,jng IIs ;; dS:
This integralcan be done using integration by parts (or the tables) : let
u = r2, dv =rV4r
2
+1dr; du=21' d1' , v= 1
1
2 (41' 2+1)3/2.
Then t he integral becomes
8. First, integrateover the portion ofthe cube in the plane z = 1,
which we will call 51 . We have -1 ~ x ~ 1 and -1 ~ y ~ 1, so
Tx x Ty = i x j = k and II Tx x TyII = l. Then
J1z211Tx x Tyll dxdy
y
1
1 Jl11dx dy =4.
-I -1
The integral over t he portionofthe cube in the plane z = -1, which we wIll call 52 , is done
in thesameway,so
.{z2 d5 =4.
2
Jis,
ow, for S3, which is in the pl ane x = l. Let D be the square -1 ~ y ~ 1 and -1 ~ z ~ 1,
and no e thatTy x Tz = j x k = j and IITy x Tzil = l. Then
Similarly,
J
.{ z2 dS
4
= J' { z2 d5
5
= J'f z2 dS
6
= ,
J ~ J ~ hs 3
where S4is the par t ofthe cube that is in the plane x = -1,S5is in the plane y = 1 and 56
is in the plane y = -1. Therefore, I Isz2 d5 is the sumofthe integrals over the six surfaces,
which is 4+ 4+ + + + = ~ o
130 CHAPTER 7
11. (a) Theequation ofthe spbere is x
2
+y2 +z2 = R2. One can substit ute xfor y, y for z, and
z for x, or any other permutation, and still get the sameequation. Thus, t he three integrals
ought to be equal. (This is what "by symmetry" usuaUy means.) Geometrically, a sphere
"looks thesame" no matterhow you look atit.
(b) Using part (a), we have
2 2 2
Jis(x +y2 + z 2) dS=Jisx dS+ Jisy2 dS+Jisz2dS=3Jisx dS.
Substitute x
2
+ y2 + z2 = R2 and recall that ffsdSis the surface area of5 to get
x 1 fl -1 flR
2
dS =- R2 -R
4
.
11
2
dS =- (x
2
+ y2 + z2) dS = 411" R
2
= 411"
s 3 s 3 s 3 3
(c) Due to symmetry ofthe sphere, ifwe integrate x
2
+ y2 over theentire sphere, we should
get twice the massdesired in exercise 10. So thedesired mass is
2 2
~ Jis(x ~ 2 Jisx 1 1 R4. + y2) dS= dS =
14. By exercise 12, the average zcoordinateis
AtS) JiszdS.
In thiscase,A(S), thesurfaceareaofahemisphereofradius r,is2 11"1' 2. By exercise 3, ffszdS
is 11" 1'3, so Z= 1I"r3/211"1'2 =r/2. Since there is as much ofthesphere on one sideofthe z axis
as there is on the other (by symmetry),x= y= O.
20. If z =g(x,y),then we can use t heformulastated in equation (4), namely,
Jis!(x,y,z)dS= JInf( x,y,g(x,y) )J1+ (8g/ ox)2 + (og/ oy)2 dxdy.
Since 5is defined implicitlyby F(x,y,z)=0,we can use implicitdifferentiation to get
8z 8g -(oF/ox) oz og -(oF/oy)
and oy
8x ox of/oz oy of/oz .
Given that !(x,y,z) =loF/8zl,weget
f
r10F IdS = Jr 10F I r::[- (OF/8x)]2+ [-(8F/8
y
)]2dxdy
Js oZ JD 8z y1 i 8F/8z 8F/8z
f
"f 18F I J (8F/8z)2+ (8F/oX)2 + (8F/8yF dxdy
JD OZ J(8F/8z)2
JInJ(8F/8zP+ (oF/8x)2 + (8F/8y)2 dxdy.
7.6: SURFACEINTEGRALS OF VECTORFIELDS
GOALS
1. Be able to computea surface integral ofa vector funct ion.
2. Understand its derivation and physicalinterpretation.
131 INTEGRALS OVER PATHS AND SURFACES
STUDYHINTS
1. Notat ion. Thesymboln( (uo,vo)) is used to describe the unit vector which is normal to (a
param trized surface) at (uo,va); note that va ) is the base point ofn.
2. Orientation. As wit h line integrals, the orientation is important. The sign of t he integral
changes with the opposite orientation.
3. ParametrizatlOn. As with line integrals , we can reparametrize a surface. As long as the
orientation is retained, the value ofthe integral is unchanged.
4. Definition. Thesurfaceintegralofa vec orfield F over a parametrized surface wit h domain
Dis
flF.(TuXTtJ)dudv.
The integral ofa vector field is a scalar.
5. Generalizations. _'otice thatscalar integrals (sections 7.1 and 7.5) do not depend on orienta-
tion. However the sign will change in the integration ofa vector field (sections 7.2 and 7.6)
iftheorientation has been reversed. Finally, note thatscalar integrals involve t he lengthofa
vector and the integral ofvector fields involve the dotproduct.
6. Reduction to calor integrals. If we know he unit vector to t he surface t hen the surface
integral reduces to IIs(F .n)dS. Lettingf = F . n, we get the scalar integral of section 7. 5.
If he orientationswitches, n changes sign andsodoes thisf.
7. Surface integral ouer a graph. If z =g(x,y) and F =(FI 'F2, F3), then a normal vector 1S
(- ag/ox, - og/oy,l) and thesurface integral becomes
fin[Fl (- + F2 (- + F3] dxdy.
8. Physical interpretation. If F represents the velocity field ofa fluid, the surface integral ofF
over gives the rateofflow ofthe fluid across the surface This is known as flux.
9. Unit normalto unit sphere. It is useful toknow that for the un1tsphere, n = r = (x,y,z).
10. Good example. Example6shows three ways ofcomputi ng thesame integral.
11. Summary. Theformulas followingexample6give a nice summaryofsections 7.5 and 7.6. An
understanding ofhow to use these formulas is important. If any oft he formul as do not make
sense to you.itis time to review.
SOLUTI ONS TO SELECTED EXERCISES
1. As in example4, the heat flux across thesurface S is IIs- 'i1TdS. Parametri ze the surface
x
2
+z2 =2. Let x =V2cos8, z =v'2sin8 with 0 0 211". Thesurface can be parametri zed
by S(8,y) =(V2cosO,y,V2sin8), 0 8 211" and 0 y 2. Then t he outward pointing
normal to S is To x Ty =V2cosBi + v'2sinBk. Given T(x,y,z) =3x
2
+ 3z
2
, we compute
- 'il T =- 6(.1:,0, z) =- 6(v'2cos8,0,v'2sinB). Thus
- 'ilT.dS = - 'ilT(Tox Ty )dydB fl JL
1
2""12 [ 2 2]
-6 0 0 ( J2cos8) + ( v'2sinB)
dydB
-1212""12dydO = -4811".
If the Ty x To cross product (hence the opposite orientation) is used, the answer would be
+4811" .
132 (HAPTER 7
6. First, we compute
j k
%x %y 0/ 8z = - 2zj+ (3y - l) k .
x
2
+ y - 4 3xy 2xz+ z2
Usespherical coordinates to parametrize S: x = 4cosBsin</J, y = 4si nBsin</J, z = 4cos</J wit h
\7xF=
o </J 7r/ 2and 0 8 27r . Then To x T =16(- sin
2
</J cosBi - sin
2
</J sinOJ - sin</J cos</Jk).
Thus
J1(\7 x F).dS J1(0,- 2z ,3y- 1) .(Te x T)ded</J
- 16J1 [(0, - 8cos</J, 12sinBsin</J - 1)
.(sin
2
</J cosB, sin
2
</J si nB, sin</J cos8)] dB d</J
r
2
r/
2
1'(
- 16io io (4sinBsin
2
</Jcos</J -sin </Jcos</J) dd8
2
r [(4 1 )11'(/2 ] 1'(
3sin3</Jsin8- 2"sin2</J <1> =0 dB -16
io
21r
r (4 1)
3
sinB
- 16io - 2" d8
2
[
-4 8J 11'(
-16 Tcos
8
- 2" 0 =167r .
8. (a) The wall lies under the circle z = 4R
2
, x
2
+ (y - R)2 = R2 and above the mountain
x
2
+ y2 + z = 4R2. From the top view, we mayparametrize the circle by
x = Rcose, y- R= Rsin8,
Then the mountainbecomes
z = 4R
2
- (x
2
+ y2) =4R2 - [(Rcos0)2 + (R+ Rsin0) 2]
4R2 - [2R2 + 2R2sinBJ=2R2 - 2R
z
sinO.
To find the surface areaofthe "cylindrical" wall oftherestaurant ,we parametrizet he wall by
x = RcosO, y = R +Rsin8, z = z
o B 27r, 2R2 - 2R2sin8 z 4R2,
and the surface areabecomes
1
271'14Rl
li Te x T211dzd8.
o sine
Thereader should verify that li Te x Tz il =R. Thus, the integral becomes
2"14Rl 12"-
RdzdB = R(4R2- 2R2+ 2R2sin8)dO =47rR3.
1o 2Rl- 2R'lsin e 0
(b) Parametrizethe restaurant interior by
x = rcos0, y = l' + 1" sinB! z = z,
where 0 1" R, 0 8 27r, 4R2 - (21"2 + 21"2 sin B) z 4R2. TheJacobian is
osB - rsinB 0
8(x,y,z)
1 +sinB rcos0 o =r+ rsin8.
8(1", 8,z)
o o 1
133 INTEG RALS OVER PATHS AND SURFACES
Thus, the volume of the restaurant is
2
f2" fR
1
4R (1' + l' sin B) dz d1' dB
Jo Jo 0)
2
1 " 1 R (1' + l' sin (;1)( 4R2 - 4R2 + 2r2 + 2r2 sin B) d" d(;l
211"
1 Jo
fR
(2r
3
+ 4r2 sin 8 + 27,3 in
2
8) d1' d(;l
211" (R4 R4)
a 2 + R4 sin (;I + 2 sin
2
8 d8.
1
Since f02" sinBdB =0 and fo
27f
sin
2
8d8 = rr , the volume is Thus, as long as R > 0, the
volume rr R
4
is always greater than rrR4 j 2.
(c) The heat flux is JIsV . dS, where V = -k(6x, 2y - 2R, 32z). The roof of the restaurant
can be parametrize by x = l' cos f) J y = "sin (;I + R, Z = 4R2 for 0 l' Rand 0 (;I 2rr .
We have T r = (co 8)i + (sin 8)j and T o = (- 1' sin (;I) i + (r cos 8)j , so Tr x To = 1' k. Therefore,
he heat flux through the roof is
-k Jis (6:z:, 2y - 2R, 32z ) (0 , 0, 1' ) drdB = - k 1 211"lR 32zr d1'd8
Z
-k 1
2
"l
R
128R
2
r drdB =- k f 1I" (64R2r21R ) dB =-k(64R4 )(2rr )
a a Jo r=O
- 128rr R
4
k.
The side which makes contact with the mountain can be paramet rized by x = l' cos B, y =
rsinB + R, Z = 4R2 - (r2 + 2RrsinB + R2) for 0 r R and 0 8 2rr. We have
Tr = {cosB)i+ (sin 8)j + (- 21' - 2Rsin B)k and T o = (-1'sin (;I) i + (reos 8)j + (- 2Rr cos 8) k, so
Tr x T 8 = (21'2 co O)i + (2Rr + 21' 2si n 8)j + rk. Therefore, the heat flux through the mo untain
side '
- k Jis (6x, 2y - 2R, 32z) . (2r2 cosB , 2Rr + 2r2 sin 8, r ) d1' dB
{R127T
- k Jo a (12 r3 cos
2
8 + 4Rr
z
sin e+ 41'3 sin
2
e+ 128R
2
" - 32,,3
- 64R
2
7'Sin e- 32R
2
1' ) dB d1'
-k 1R(12rr,,3 + 4rr1'
3
+ 256rr R 2r - 64rrr
3
- 64rr R
2
,, ) dr
R
-k 1(-48rrr
3
+ 192rrR
2
r ) dr =- 841rR4k .
Finally, t h curved gJass wall can be parametrized as in part (a) . We have T o x T z
(Rcos B)i + (Rsin B)j. The heat flux through t he wall is
- k Jis (6x,2
Y
- 2R,32z) (RcosB ,Rsin8,0) dzd8
- k 1Z1I'14R2 (6 RZ cos
2
8 + 2R2 sin
2
8) dz d8
o 2R2 - 2R' sin8
-k 1211' (2R2 + 4R
Z
cos
2
B)(2R2 + 2R
z
si n e) d8
211"
-k 0 (4R4 + 8R
4
eos
2
B+ 4R4 sin 8 + 8R
4
cos
2
Bsin B) d8
1
- k(8rrR4 + 8rrR 4) = -16rr R
4
k.
134
CHAPTER7
Adding these results, we find that the totalflux is -2287TR
4
k.
11. The surface S is the unitsphere, so S can be parametrized by x =cos0sin, y =sin0sin
and z =cos for 0 0 27T and 0 11". Differentiate each component with respect to0
to get T o= (- sinOsin,cosBsin,O). Similarly, = (cosOcos,sinOcos,- sin). The
normalvector for S is T ox T =(sin
2
cos0, sin
2
sin0, sincos). Fact or out sin to get
(sin)(x , y , z) = n = (sin)r . Suppose F = (F
r
, Fe, F), then F .n = F .(sinr)= Fr sin.
Here, we make use oft he fact that (er , ell,e) form an orthonormal basis (see section 1.4).
Hence
J1F .dS =1 21r L'" Fr sind dB.
Thecorresponding formulafor real-valued functions is
Jl!dS = 1 21r l" !sinddB.
14. A parametrization ofthesurface v) = (u, v, 0) ,so liT" x Tvll = Ii i x jll = Il kll = 1. The
scalar integral is
J1! (x, y , z ) dS =JIv ! (tI,v, 0)li T" x T vll dudv.
Bysubstitut ing (x, y) for (u,v), we get ffD! (x , y, 0) dx dy. For a vector fi eld F =(F"" Fy, F
z
)
(NOTE: thesubscripts here do not denote partials!),
Jis F .dS =JIv F .kdz dy =JIv F. dx dy.
Thatis, only the zcomponent ofF enters intothe formula.
16. (a) Parametrize the cone as follows: x = reose, y = rsinO, z = (x
2
+y2)1/2 = r, 0 r I ,
o B 211'. Then T ox Tr = (- rsinO, rcosB,O) x (cosO,sinB,l ) = (rcosO,rsinB,-r) . The
flux is the surface integral of F over S, whichis
21r 1
r
21r
t r r
Jo J (0, 0,- 1)(r cosO, rsinB,-r) drdB = Jo Jo rdr dO=71'.
o
For the hot shots, note that the same amountofrain falling through the cone must also go
through thedisk x
2
+y2 1, z = O. Theproblem then could be done in your head.
(b) Using thesame parametrization as in part (a), t he total flux through the cone is
2tr 1
r
21r
r
1
(-Vi -Vi) r r
Jo Jo - 2- ,0' - 2- (rcosB,rsinO, - r) dr dB =
.J2(- r cosB+r) dr dB
J
o
Jo 2
Vi {21r _ cosO) dO =
2 Jo 2
7.7: APPLICATIONS TO DIFFERENTIAL GEOMETRY, PHYSICS AND FORMS OF
LIFE
GOALS
1. Be able to compute the Gauss curvat ure fora given surface.
2. Be able tocompute the meancurvature for a given surface.
3. Be abletoexplain when theGauss-Bonnet theorem applies for calculat ing the totalcurvat ure
ofa surface.
135 INTEGRALS OVER PATHSAND SURFACES
STUDY HINTS
1. Old calc ulations. In the process ofcomputing the Gauss and mean curvatures, you already
learned how to do the calculat ions for many ofthe ingredients earlier in the course. In fact ,
you should consider the calculations a review.
2. New funct ions. Ifyour instruct or expects you to memorize how to compute E, F, G, i, m, n,
it may be helpful to know that E, F and G involve manipulations ofthe first partials ofC).
Also, note that N is dottedwit h the second partialsofC) to computei, m and n.
3. Calculating Gauss cur vature. Note that E, F and G are not needed to compute the Gauss
curvature since W =IIC) u X If1 v112.
4. Curvature ofspecial surfa ces. You should know the results ofexamples 1 and 2. Example 1
shows that a plane has no Gauss or mean curvature. Example 2shows t hat the curvature is
the sameat any pointon a sphere.
5. Gauss-Bonnet theor m. For a "sphere-like" surface which is closed and bounded, we have
(1/271') lIsK dA =2. Furthermore, asurfaceof genus g resultsin heintegral(1/271') Ifsf{ dA ==
2- 2g.
SOLUTIONS TO SELECTED EXERCISES
1. We begin by computing allfi rstand second partialderivativesof C) (u,v) =(ucosv,usinv,bv),
that is,
C)u =Tu =(cosv,sinv,O); C)U=Tu =(-usinv,ucosv,b);
C)uu =(0,0, 0) ; lf1 "v =(-sinv,cosv,O); and C)uu =(-ucosv,-usinv,O).
Using these results, we can compute
2
T u x Tv =(bsinv,-bcosv,u); E = IIC)"W =cos v+sin
2
v+ =1;
F =C)U .C)v =-usinvcosv + ucosvsi nv+ =0;
2 2
G=II C)v112 =u
2
sin
2
v+u cos V +b
2
=u
2
+b
2
; and
0
2 2
+b
2
W =BG - F =(l)(u
2
+b
2
) - =u
Then, we calculate N = (T " x Tv) /v'W=(b si nv, -bcosv,'1.I.)/Ju
2
+b
2
. Now, we calculate
i, m and n:
f N .C)"U = N 0 = 0;
m N 'C)uv =(-bsin2v-bcos2v)/)u2+b2=-b/)u'J+b2; and
n N .C)v v =(-ubsinvcosv+ubsinvcosv)/Vw=O.
Finally, t he Gauss curvature is
K =in- m
2
(0)(0) - m
2
-[b
2
/(U
2
+b
2
)]
W W u
2
+b
2
and the meancurvature is
H =: Cf +En - 2Fm =
2W
since i =n =F =0.
5. Let x =u and y =v, so z = u
2
/a
2
- 1I2/b
2
. Thus, a parametrization of the hyperbolic
paraboloid is lf1( u,v) =(u,v,u
2
/a
2
- v
2
/b
2
). We now compute the first and second partial
derivatives of1f1:
C)u =T " =(1,0,2u/a
2
); C)V =Tv =(0, 1,-2v/b
2
);
lf1u" =(0, 0, 2/a
2
); C) "V = (0,0,0); and c)"v =(0,0,-2/b
2
).
.. .
136
CHAPTER 7
Using t hese results, we can compute
4
Ttl X T v =(- 2u/a
2
, 2v/ b
2
,1) and W =liT" X Tvl1
2
=1+4u
2
/a
4
+4v2/b .
Now, we calculate e, m and n:
e = N ()uu =[(Tu xT
lI
)/VW](0,0,2/a
2
) =2/a
2
/W.
m = N ()uv =N .(0 , 0,0) =0; and
n = N 4'1111 = [(T" x T,,)/ Vw] .(0, 0,_2/b
2
) ;::: -2/b
2
J'{V.
Finally, t he Gauss curvatUIe is
){ =
en - m
2
[( 2)(-2) 2] .
W = b2JW - '7" a2J'{V W
- 4 - 4a
6
b
6
a
2
b
2
W2 - (a
4
b
4
+4b
4
u
2
+4a
4
v
2
)2 .
Notice that we did not have tocompute E, F and Gt o fi nd t he Gauss curvature.
8. The torus T is defined by
~ , ) ) =(x ,y,z) =(( R + cos ) cos B, (R +cos)sin(),sincP)'
so
~ =T =(- sincosf), - si nrf; si n() ,cos) ;
4' e= Te=(- (R + coscP) sinIf, (R + cos.p) cos() ,0).
Throughout this exercise, we use the identity C05
2
A+ si n
2
A = 1, so we get the cross product
Ttl> x T o - (R + cos)coscosei- (R + cos)cossinIfj
-(R +cos)sink.
Next , we calculate
W = liT</> X Tol12
(R+ COS ) 2 cos
2
1; cos
2
e+ (R+coscP )2C05
2
sin
2
e+ (R+ cos)2sin
2

(R + cos)2cos
2
+ (R + cos)2 sin
2
if;
(R+cosif; f .
Then, we calculate
N = (T x T8)/J'{V= - cos1> cos(lj - cossinIj - sin.pk.
Now, we compute e,m and n:
=N () N .(- coscosIf, - cossinIf, - sinrP)
cos
2
cos
2
B+ cos
2
si n
2
() + sin
2
.p
cos
2
if; + sin
2
=1;
m =N ~ o N .(sinsin() ,- sincos0, 0)
- sincossin() cos(I + si nif; cos1> sin() cose+ 0 =0;
n = N 4' OB N .(- (R + cos)cos() ,- (R + cos)sin(), 0)
(R + cosif; )cosif; cos
2
B+ (R + cosif; )cossin
2
If + 0
(R + cos)cos.
Finally, the Gauss curvatUIe is
m
2
0
2
Cn - (1)(R+ cos)cos - cosif;
K = HI
(R +cos)2 R +cos
137 INTEGRALS OVER PATHS AND SURFACES
For the torus, which has genus 1, the Gauss-Bonnet theorem tells us that we should get
(1/ 271") ffs K dA =0. In this case, we have
= {27r {27r osl II Tt,6 x T ell dBdl
271" io 10 R +cosl
1 127r127r cosl
= -2 R (R+cos)dedl
71" 0 0 +cos
= {2r. {27r cosded = 271" {2rc cosl dl
io 10 271" 10
= sinC =o.
9. We begin bycomputingallofthe first andsecond partialderi vatives of(I(u,v):
(I"
=
Tu =(l,h'(u)cosv,h'(u)sinv);

=
=(0,- h(u)sinv, h(u)cosv);
(I"u =
(0,h"(u)cosv, hI!('U ) sinv) j

=
(0,- h'(u)sinv, h'(u)cosv); and
(I"" =
(0, - h(u)cosv, - h(u)sinv) .
Using these result we cancompute
T uX T il =(h'(u)h(u) ,-h(u)cos v,- h(u)sin v).
Furthermore, we calculate
2
w = II TuX =(h'(u) 2(h(u2+(h(U))2cos V +(h(U)2sin
2
v
= (h(u2[ (h'(u)2+1].
Next, we calculate t . mand n:
e=N c)uu = [(Tu x TtI)/Vw](0,h"(u)cosv, h"(u)si nv)
2
= (- h(u)h// (u)cos V - h(u)h"(u)sin
2
v)/Vw =[-h(u)hl!(u)]/Vw;
m = N .(lUti = [(Tu x Tu)/Vwj(0,- h'(u)sinv, h'(u) cosv)
= (h(u)h'(u)sinvco v- h(u)h'(u)sinvcosv)/Vw =0;
n = N(ln = [(TuxTu) / Vw](O, - h(u) cosv,-h(u)sin v)
= [(h(U2cos
2
v +(h(U))2sin
2
v]/Vw =(h(u))2/Vw.
Hence, the Gau curvature is
K
= en - m
2
W
=en - =
W
(-h(h
ll
) )
VW

VW

'W
=- h
3
(h")
W
2
= - h3(h" ) _ - hI!
[( 1+(h')2)h
2
J2 - [(1+(h
'
)2)2h]'
SOLUTIONS TO SELECTED REVIEW EXERCISES FORCHAPTER 7
l. (b) First, we calculate c'(t ) = (- sint,cost,1), so lI e'(tlll =.;2.Then the path integralis
{{2'1r {27r
1c
fds
= i o xyzll e' (t) ll dt = io V2(cost )(sint)t dt .
138
CHAPTER7
Use integration by parts with 1.1 =t, dv =costsintdt, du = dt and v = sin
2
t. The integral
becomes
127r In
(
2 0 0 2 2 i o 2 2 .
V2 -sin
t
2t _
127f

)
(
2
7rl- cos2fdt =-7r V2
2. (b) As inexercise 1(b), we have ds= IIc'(t)II dt = V2dt. Then, the path integral is
{27f
2
{27f ( 1+ cos2t)
c fds i o (sint+ cos t)V2dt = V2i o sint+ 2 dt
1
t'2t)127r
V2
(
- cost+ "2 + SI: 0= V27r .
3. (a) Compute t he line integral over each segment of C and add them t ogether. Be careful
with the orientation. Recall from chapt er 1 that the equation ofthe line I(t) = (x,y ,z) =
...
(1 , 0,0)+ t( - 1,1,0), 0:5 t :5 1 satisfies the condition1(0) = (1 , 0,0) and1(1) = (0,1,0). On
this segment , x = 1- t, Y =t and z = O. Also dx = - dt, dy = dt and dz = O. Substitute
these values into thegiven line integral toget
1
1 11 11 1 1 2
[sin7r (1 - t)dt - (cos7rt)(O) ]= sin7r( 1- t)dt = - cos7r (1- t) = - (1- (- 1)) = - .
o 0 7r 0 7r 7r
Thenextsegment can be parametrized by I(t) = (x,y,z) = (0,1,0)+ t(O, -1,1) = (0,1- t,t )
1
for 0:5 t :5 1. Also, dx = 0, dy=- dt, and dz = dt,so we get
1 1 11
[sin7r (0)( -dt )- cos7r( 1- t)dt ]= - sin7r (1 - t) =O.
o 7r 0
Finally,the lastsegment can beparametrized by theequation
I(t)= (x,y,z) =(0,0,1)+ t(l,0,-1)=(t,0,1- t)
for 0:5 t :5 1. Also, dx=dt, dy=0 and dz = - dt ,so we get
11 [sin7rt(O) - cos7r(0)(- dt)] = 11 1dt = l.
Therefore, the line integral around the triangleis 2/7r +0+ 1=2/7r+1.
5. We start at (0,0) and compute t he line integral over each side of the square. For t he line
segmentfrom (0,0) to (a,0), we have y=0, dy = 0and:5 x :5 a. Thus
3
J
r r 2 2 a
F .ds =io [( x - 0
2
)dx+2x(0) .(0)] =i o x dx =a-
For the linesegment from (a, 0) to (a,a), we have x =a, dx = 0and 0:5 y:5 a. Thus,
a a
2 3
JFds =l [( a - y2)(0)+2(a)ydy]=l 2aydy = a
y2
1: =a .
For t helinesegment from (a,a)to (0,a), we have y= a, dy= and xgoes from a to O. Thus,
a
3 3
2 2 3
J
F ds =ia
r
(x
2
- a ) dx =-
l
0 (x
2
- a ) dx =-
(a
a- - a
)
=3
2a
'
Finally,for t he line segmentfrom (0,a)to (0,0) ,x = 0,dx= 0and ygoes from a to O. Thus
JF ds =O.
By addition, we find t hat the lineintegralaround thesquare is a
3
/3+a
3
+2a
3
/3+0= 2a
3
.
I
INTEGRALS OVER PATHS AND SURFACES 139
7. (b) First complete the square: (2x2 - 8x + 8)+ y2 + z2 =1+ 8or 2(x - 2)2 + y2 + Z2 =3
2
.
This has the form X
2
+ y2 + z2 =p2, so use spherical coordinates: For the x parametrizat ion,
let X =v'2(x - 2) =3cos Bsin<,h , so x = 2+ (3cos11 sin<,h )/ v'2. Thus, the parametrizat ion is
x 2+ (3coshin<,h)/ v2
y 3sin11 sin<P
z 3cos<p,
with 0 B 271' and 0 <P 71'. Thisis a general strategy in at tacking manyparametrization
problems: changing cartesian coordinates into either cylindrical or spherical coordinates by
completing the squares.
10. Thesurfaceareaofthegraphlyingover D islIDII TxxTy II dx dy andtheareaofD is lID dx dy.
The'parametrization" oft hegraph is x =x, y = y, z =/ (x, V) . Thus,
Tr= i + (8// 8x)k and Ty= j + (8!18y) k.
Therefore,
Tx x Ty = (8/18x)i + (8//8y)j + k and II Trx Ty ll = [( 8/18x)2 + (8//8y)2 + l f/2.
Since (8/18z)2 + (8//8y)2 = c, II Tx x Tyll = JI+C.Returning to t he origi nal formula, we
have lID IITr xTylldxdy = lID JI+Cdxdy. Since cis a constant, we factor the constant
from the integral to get VI+ClID dx dy = v'I+C.(areaofD) .
12. (b) Use cylindrical coordinates. Let x = rcosB, y = rsinl1. In addition,Z = x = "cos l1 , and
the intervals are 0 l' 1, 0 11 27l' since we want to be inside the cylinder x
2
+ y2 =1.
We calcuJate
(cosBi + sinOJ + cos11k) x (-rsinOi+ l'cosOJ - l'sinOk)
(-ri+ rk)=r(-i+ k) so IITr xTell=v2r .
Therefore,
15. We want to compute lIs x dS, where S is the trian-
gle with vertices (I,I,1), (2,1,1) and (2,0,3). First ,
we need to find the normal to the triangle: Two vec-
tors on the triangleare (1 , 0,0) and (0,I,- 2) (found by
subtracting the coordinates ofthe vertices). Take t heir
cross product and normalize it. We get the unit normal
n =(0,2,l)/v'5, so cosB =n k = 1/v'5. Next, the
projection ofS onto the X!t"plane can be described by
- y + 2 x 2, 0 y I ,as shown. Thus,
2
Jis xdS =
v'sJ'r x dx dy = v's1
1
1 X dx dy
JD 0 -y+2
-
v'51
1
4- (2 - y)2 dy
2 0
v'5 [4 (2 - y)311] =5V5.
2 + 3 6
o
(1,1,1)
y
,(1,1,0)
x
(2,1,0)
140
CHAPTER 7
19. Here, we have F =e'i+ tj+ t
2
k and ds =etj+ j + 2tk. Thus
1 1
(e
t t
+ t.1 + t
2
(e
2t
1F . ds = 10 . e . 2t)dt =10 + t+ 2t
3
) dt
e
2t
1 2 t2 t4)11
( 2 + "2 +"2 0 = 2(e + 1) .
2
24. Oursurface is z2 =1 - x - y2. We want the part above the xyplane ,so we want z ~ 0, or
2
- f(x,y) =z =(1 - x y2)1/ 2. For thesurface areaofa graphover D, the formula is
A=fLV
1
+
f
; + f;dxdy.
In this case,
_ - 1 2 2 - 1/ 2 _ -x _ -y ( )
fx - - 2 . 2x (1 - x - y ) - ( 2 2)1 / 2 and fy - ( 2 2)1/2 by symmetry .
I - x - y I - x - y
Hence, the integrand is
x2 + y2
1+ 1_ x 2 _ y2 - (1 _ x2 _ y2)1/ 2'
The areawe wantis
f
afa 1
_ 2 _ 2) 1/ 2 dydx.
- a -a
(
1 X Y
2
We can eval uateone ofthe integrals. Let u =(1 - x
2
)1/2. Since tI is independent ofy, the
integral in ybecomes
f
a 1 f a 1 1
(
u2 _ 2) 1/2 dy=~ I." '" dy.
- a y -a
Substitutingv =y/u with dy =tI dv, we get
f
a ti
/ dV . -1 a/ ti . -1 a . - 1 ( a )
v'f=tj2= sm (v)1 =2sm (-)= 2sm ~
- a/ ti 1- V - a/ti U 1- ;z;
Plug this back into the original integral and get t he formul a for t hesurface area:
2I:sin-
1
(b) dx.
27. (c) Use cylindrical coordinates: x =2cosO, y =2sinO, z =z, where 0 :5 (J :5 21r and
o:5 z:5 x+ 3= 2cos(J + 3. (Note the difference in parametrizations, as compared t o exercise
12). We calcul a te IIT8x Tzil = IIC- 2sinOi+ 2cosOJ) x kll =112cosBi + 2sinBj il =2. T hen ,
thesurface integral becomes
211' 12cos8+3 f211' 2
1
2z2dzd()
lo a (2cosO+ 3)3dB
o 0
2 f211'
3 2
a lo (8cos () +36cos () + 54cosB+ 27)dO
2f21r [. (1+ cos2B) J
al 8( 1 -S10
2
(J)cosO+36 2 +54cosB+27 dO
o
2 [ ( . Sin
3
(J ) ( Sin20) . J/
2
11'
a 8 smO - - 3 - +18 ()+ - 2- + 54smO+ 270 0
2
a [18.21r + 27 .21r]= 601r.
141 INTEGRALS OVER PATHS AND SURFACES
TEST FORCHAPTER 7
1. True or false. If false, explain why.
(a) Let c(t ) =(e
t
+ 1,sin2t , - 2t + 4), t 1 and let b (u) =(e
u
/
2
+ 1,sinu, -u + 4),
u 2. Then the line integrals IeF .ds and IbF .ds areequal.
(b) On a givendomai n,jfg( x , z) andf is a positive scalar,t hen the integraloff over the
surface y =g( x, z) is negative.
(c) Over a trianglewith vertices at (-1, 1),(2, 3)and (5 ,- 3),thesurface areaofthegraphof
I( x, y) =2xy+y+5 isthesameas t hesurfaceareaofthegraphofg(x , y) =x
2
+ y2 + x +3.
(d) Consider t he cone z = J x2 + y2. Cl t( u, v) = (u, v, ";U
2
+v2) yields a one-to-one
parametrization ifu 0 and v 0,but tl2 (r , 0) =(rcos(I , rsin(I , Jr) for 0 (I < 211' ,
r 0is not one-to-one.
(e) If the li ne integral of F along one closed curve is 0,then F is a gradi ent.
2. An ellipsoid centered at the origi n wi th intercepts at x = 1, Y =2, z = 3 can be
paramet rized by tI (B,</J ) =(cos(1 sin</J,2sin(1sin</J,3cos</J), (I < 211' , 11'. If this
ellipsoidhassurface areaA, howmuch largeris thesurface areaofan ellipsoidcentered at the
origin wi th intercepts a.t x =2, Y = 4,z =6?
3. Find the workdone by a force F =xi+ x yj alongthe path beginningat (0, 0), going to (1,1),
to (2,0) and t hen back to (0,0) all alongstraight line segments.
4. LetF(x, y, z) =xi+yj + (x +z) k. Computet hesurfaceintegralofF over theellipticalcylinder
x
2
/4 + y2 =1, 0 z l.
5. At a point (x , y, z) on aspherical shell ofradius 2,the density is Iz I.
(a) Find a paramet rization for t hesphere using thespherical coordinates (I and .
(b) Write the mass ofthe sh II as asurface int egral.
(c) Compute the mass.
e
Z
6. (a) Compute the gradientofI (x ,y,z) = + 2xz sin y.
(b) Let p be the pathfrom the origint ot hepoint.(1r / 2, 1,0) alongthecurve y =sinx. From
there, p continues along a st raight line to (5,e, - 3). Then p cont.inues on a straight
line from (5, e, - 3) and ends at (2,37r/2, 1). Compute the line integral of F(x , y,z)
2zsi nyi + 2xz cosyj+ (e
Z
+ 2;1; siny)k over p.
7. Let C) (u v) =(u
2
, V - 2u, 3uv
2
) bea parametrized surface.
(a) Find the equat ion ofthe tangent plane when (u v) =(1,1) and (x, y , z) =(1,-1,3) .
(b) Wri te an expression for thesurfaceareaoftI over theregion (u, v) E [0,2]x [-1,1]. (Do
not evaluate the integral. )
8. Find the flux of F(x , y, z) = ;ri- 3yj + 2zk across the surface z = x
3
y + y2 on [- 1,1]x [0,2].
t
9. Let r (t) =(t
2
, e , t + 3,2t ), 0 t 1b . a path in 1R4and let F(w, x , y, z) =(z , 3,w + y, 2x).
Compute the li neintegral of F over the path r.
10. Anewamusementparkisbeing builtin t own. Itsperimeterhasanelli pt icalshape andcan be
descri bed by x
2
/9+ y2 /4=1. A roller coaster ride is to be builton t opofthefence enclosing
the park. Its height is IxYI. Use the path integral t o compute the surface area ofthe park's
fencing.
143
8 THEINTEGRAL THEOREMS OFVECTORANALYSIS
8. 1:GREEN'S THEOREM
GOALS
1. Be able tostate Green's t heorem.
2.Be able t o use Green's theorem to compute a lineintegral or a double integral.
3. Be able to use Green's theorem to fi nd an area.
STUDYHINTS
1. Notati on. Recall thata, t hesamesymbol used for part ialderivatives, meansboundary. Thus,
aD is the boundaryofD.
2. Green's the07'em. Under c rtai n conditions, a line integral is converted intoa double integral:
L (PdX+QdY)= iL ~ ~ -:) dxdy.
Ofen,one ide ofthe equation will be much easier toevaluat than t he other side.
3. Required conditions.
(a) Cmust be a closedcurve wit h an outside counterclockwise orientation. Interior "holes"
have a clockwise orientation. In bothcases, the region is on yourleft ifyou walk around
Cwith thecorrect orientation.
(b) P and Q must have continuou first derivatives.
IfGreen's theorem does not applydi rec ly to a region, the region can usually be subdivided
so that the theorem can be applied. See figure 8.1.5 ofthe text .
4. Green's theorem and area. The are ofD may be computed by the formula
A = ihD(xdy-ydX).
Again, aD is oriented counterclockwise. This for ulais most useful if t he boundary has a
simple parametrization;otherwise,double integration is usually si mpler.
5. Vector form. IfF =Pi+Qj ,then \7 x F =(aQ/ ax - ap/ ay)k, so another formulati onfor
Green's theorem becomes
hDF .ds = (\7 x F) .kdA, iL
where ds =dxi+dyj.
6. Divergence theorem in the plane. Ifn is an outward uni t normal to aD,t hen
hDF nds=iLdiv FdA.
Compare this with Gauss' di vergence theorem in space studied later in Section 8.4.
144 CHAPTER 8
SOLUTIONS TO SELECTED EXERCISES
2. Accordi ng to Green's theorem, the areaofa region is
A=
Since D is a diskcentered at (0,0) wi t h radius R, the boundary aD can be parametrized by
x =R cose and y=R sinB, 0 e 27r.
Then, dx = - RsinBde,dy = Rcosede ,and so the areais
1 f2Tr 1 f2Tr
A =2" lo [(RcosB) (Rcose) - (RsinB)(- Rsine)] dB = 2" l o R2 dB =7r R2,
which is indeed the area that is calculated by using elementary geomet ry.
3. (b) We use thesameparametri zati onfor aD as inexercise 2. Theleft-handsideoftheidentity
in Green's t heorem is
f (P dx +Qdy) f (x+ y) dx+ f ydy
JaD JaD JaD
f2Tr f2Tr
Jo (Rcose+ Rsine)(-RsinB)dB+ Jo (R sinB)(R cose)de
f2Tr
Jo (-R2cosBsinB- R2sin
2
e+ R2sinecosB) dB
2Tr 2Tr
_ R21 sin
2
B dB= _ R21 1- cos2B dB =_R27r.
2
0 0
By Green's theorem, thesameintegralshould be equal to
'f (OQ- OP)dxdy.
JD oX oy J
Again,we usepolarcoordinates. Dcanbedescribed by0 r Rand0 B 27r. Calculating
oQ / Ox = 0 and fJ P/fJy =1 and rememberingthat theJ acobian for polarcoordinates is r, the
right-hand sideofGreen's t heorem becomes
f2Tr f R
2
" fR 1
Jo Jo (0-1)rdrdB= 0 Jo -rdrdB=-7rR
2
.
Therefore, Green' s theorem is verified for this case,
5. Green's theorem t ells us that the area is
A = f (x dy- ydx).
2laD
From the given parametrization ofthe cycloid, we compute dx = a(1 - cosB) dB and dy =
asinBdB. Remember that t he x axis forms partofthe boundary. This part ofthe boundary
can be described by x =27ra(1 - u) and y =0 for 0 u 1. Notice that our cycloid is
t raversed clockwise, t hat is, from left to right, starting at the origin along the cycloid, and
t hen back along the 3.; axis from right to left. Since y =dy =0, the desired areais
1 f2Tr 1 t
"2 J [a(B - sin(1) .asinBdB - a(1 - cosB) .a(1- cose) dB] +2" J (x0- 0.dx)
o
o
sinB- a
2
si n
2
B- a
2
+ 2a
2
cosB- a
2
cos
2
B) dB
112Tr
- (a
2
f) sin e- 2a
2
+ 2a
2
cose) dB,
2 0
145 THE INTEGRALTHEOREMS OF VECTORANALYSIS
Using integration by parts, we get
To get t he correct answer, recall that t he di rection should be counterclockwise. By simply
changing the sign,we get the areaspanned by one are of the cyeliod, 3a
2
'1T".
8. Let D be t he union of simple regions, Di, i = 1,2,3,...,n,where each boundary, aD;, is
oriented counterclockwise. Suppose Pand Qare continuouslydifferent iableon D.We want to
show that the boundary of Dis the same as the boundary of all ofthe Di. Each intersecting
boundarylieson theboundaryoftworegions. Whenorientationsareconsidered, contributions
to the integralcancel out as in figure 8.1.5. Hence
{ (P dx +Qdy)
t {(Pdx+Qdy)=til . - dxdy
J&D
.=1 J&D, . = 1 D, Y
Jl - dxdy.
11. (a) Usingpolarcoordinates ,we let x =r cos() andy =rsin(). Then Dis described by0 r 1
and 0 () 2'1T" . Wecomputethat di v F =1+1=2. (Ifnecessary, youshould review how to
compute a divergence in section 4.4.) Since the J acobian for polar coordinates is 1', we get
Jldi v FdA=127<11(2)1' drd() =211" .
Ontheotherhand ,weknowt hat t heoutwardunitnormalto t heunitcircle is n::o: (cos(), sin()),
so
27< 127<
F .n ds = (cos() , sin()) .(cos() ,sin()) ds = ds =2'1T".
1 1 a &D a
(b) We want to compute
{ F nds,
J&D
where F =2xyi- y2j. By the divergence theorem, this integral is equal to
Jldi v FdA =Jl(2y- 2y)dxdy=O.
15. Since cos
2
() + sin
2
() =I , let x/a =cos() and y/ b =sin(), 0 () 211" . Then x =acos(),
dx =- asin() dO,y =bsin() and dy=bcos()d(). By Green's t heorem,t he areaof the ellipse is
A =
1127<
"2 a [( acos())( bcos()) - (bsin())(-asin())1d()
! (21r abd() =ab'lT".
2Jo
16. In polar coordinates, x = l'cosBand y =rsin(). Since r is a function of (), we get dx =
(1" cos() - l'sin()) d() and dy =(1" sin() + rcos()) d() (here 1" denotes the derivative ofl'with
CHAPTER 8
146
respect t o e). Substituting into Green's t heorem, we get
A [ (xdy-ydx)
2Jc
f b (1' cose)(r'sine+1'COS e) de- (rsinB)(r'cose- rsine)de
2 Ja
1 fb ') . ') 1 f b
2J e+ e) de == 2Ja 1'2 de.
a
19. Toform one loop oftherose, we go frome=0toe== 7r/2. Using theresult ofexercise 16,the
area IS
1
2 2 2
r/
(9 sin
2
2B)dB (1+cOS 4e )
2Jo (3si n2e)2 dB
r/ = r/ de
2Jo 2Jo 2
(B+ Sin4e)11f/2 =97r
4 4 0 8
22. Use the given definition of Bu/an and the divergence theorem in the pl ane for the regIOn
B = Bp:
f oou ds =jf \7u.n ds =J'f \7 .(\71/,)dA =J'{\7
2
udA.
J8B n 8B JB JB
26. (a) Suppose u(p) is a maximumpoi nton D.From exercise 25,u(p) is t he average valueofuon
a disk ofradius Rcentered at p. This is possible only ifu(p )= u( q)for all q in D.Indeed, if
u(q) <u(p)forsomeq in D,there must be a u(r ) >u(p)t o maintaint he average. Therefore,
tI must be constant on some disk centered at p.
8. 2: STOKES' THEOREM
GOALS
1. Be able to state and use St okes ' t heorem.
2. Be abletouseStokes ' theorem t ocalculatealineintegralon aclosed curveorasurface integral
over a surface wi th a closed curve as its boundary.
STUDY HINTS
1. Review. Surface integrals are used in t his section. You should review section 7.6 ifyou have
forgott en how to calculate a surface integral.
2,Relation to Green's theorem. Like Green's t heorem, Stokes ' t heorem converts an integral in
onedimension to an integralin two dimensions. St okes' theoremis a generalizat ionofGreen's
theorem.
3.Stokes ' theorem for graphs. If z =f(x,y),then
Jiscurl F dS =lasF .ds.
This is a formul a you should memorize. As with Green's theorem, oS is oriented so that the
surface is on your left as you walk upright around as,which must be a closed curve.
4. Genemlized surfaces. Ifthesurface S is not the graph ofa function, t hen Stokes' t heorem is
still t rue ifScan be descri bed using a one-to-one parametrization. The boundary ofD, oD,
gets mapped t o as, so oD should have t he correct orientation. For an example when the
orient ation becomes important ,see the solution ofexercise 5 ofsection 8.1.
147 THEINTEGRALTHEOREMSOF VECTOR ANALYSIS
5, Application, Accordi ng to St okes' theorem, to evaluate ffscurl F ' dS, we can change the
surface 5 to any other surface with thesame boundary 85, In mostcases, we will change to
a planar surface or another simplesurface, Imagine a loop ofwire with an elasticsheet S on
it , Thesurface integral of the curl of a vector fi eld over any surface formed by a deformation
oftheelasticsheet will be equal tot he line integral over the wi re (assuming the wire itselfis
not deformed),
6, Circulation, You should know t hat curl V,n is the circulation of V per unit area ofsurface
perpendicular ton ,
SOLUTIONS TO SELECTED EXERCISES
1. By Stokes' theorem, we only need to evaluate
r F .ds,
Jas
where 8S is thecircle x
2
+y2 =1, Z = 0, the boundary ofOUI surface. Parametrize thecircle
usi ng polar coordinates, i.e., x =cosB, y =sinB, Z =0, 0 ~ B ~ 27r. Then the integral
becomes
1
2 ~ 2 ~
F ds = [sin8(- sin8)- cos8(cosB)] d8=- d8=-27r.
&S 0 0
4. We want to show that thederivativeof he magneticflux with respect to timeis O. We begin
wi t h
~ J'{H .dS =J' {8H .dS =- J'r-8H .dS.
8t Js Js at Js at
Th first step is justified because Sis not afunction oftime. By Faraday'slaw,
-J is ('v xE).dS.
Then by Stokes' theorem and thefact t.hat Eds = 0,since E isperpendiculartot he boundary
of5, we get
- { Eds = O.
Jas
5. The boundary of t h surface is a closed curve,so we may take advan ageofStokes' theorem:
Jis(V' x F).dS:;:: lasF .ds.
The boundary is the circle x
2
+y2 = 1, Z = O. The right-hand side becomes
{ F ds = { (x,y) .(dx,dy).
Jas Jas
Now use polarcoordinates: Let x = cos8, dx = - si n8d8;y = sin8,dy = cos8d8; 0~ 8~ 27r .
Substitution into the last integral gi ves us
Jo
r
21r
(- cos8sin8+sinBcos8)dB = O.
9. Use Stokes' theorem. Theboundary of S is thecircle y2 +z2 = I ,x = O. On the boundary,F
becomes - i'j . Use polar coordinates: y = si nO, Z =cosB, 0 ~ f) ~ 27r. By Stokes ' theorem,
we get
148
(HAPTER 8
12. Theflowrateis JJs curl I)dSand byStokes'theorem,thisisJasI).ds.We have theboundary
on the xyplane where z O. Parametrize the boundary by
x = (R/4)cosO, y = (R/4)sinO.
Then the flux is
1
f2Tr R2 11"R2
2 2
asl)ds=io 16(sin B+cos B) dO=-8-
14. By Stokes' theorem,
J1s(V' x F).dS lasF .ds.
SinceF isperpendiculartothetangenttotheboundaryS,Fds 0. Hence, JJs('V'xF)dS O.
If F is an electric field, this means the rate ofchange ofmagnetic flux is zero by Faraday's
law. See example5 in the text .
18. Thesolutionrelies on the basic identities ofvector analysis (see table in section 4.4) . Since C
is a closed curve which is the boundary of a surface S,we have C = asin Stokes' theorem.
Thus,Stokes' theoremtells us JJs(V' x F).dS = JeF .ds .
(a) Here, we useStokes'theorem where F = IV' g. We need toshow that\lxfV'9 V'IxV' g,
orcurl(JV'g)= \llx\lg. Bybasicidentity10,wehavecurl(lV'g)=/curl(V'g)+\llxV' g, but
by basicidentity 11, curl( V' g) = 0,and80 we havecurl(IV' g)= 1(0)+V'IxV'9 V'IxV'g,as
required. Al ternati vely, we could have computed the curl ofIV'g = I(ag/ox)i + I(og/oy)j+
I( og /oz) k and t hen used theequality ofmixed partialssince Iand 9 are C
2
functions.
(b) Here, we useStokes' theorem withF =IV' g+gV'I. By basicidentity6, curl(fV' g+gV'f) =
curl(fV'g) + curl (g\l I). Applying basic identity 10,we have
curl(IV'9+ 9V' f) f cUrl(V'g)+V' I x \lg+ 9 curl (V' f)+ V'g x V'I
I curl( V'g)+9 curl (\lf)
since V' I x V' 9 =-(V'9 x V' f) by properties ofthecross product. According to basic identity
11, curl(\ll)=curl (V'g) =0, andso curl(l\lg+ gV'f) = O. By Stokes' theorem,
[(IV'9+ gV' f). ds isV' x (IV'9+ 9V'/) .dS J
JIso dS O.
2l. First ,we wili calcuiateJJs (V' xF) .dS. (1,1, 1). Alsol)r =(cosO, sinO,O)
and 1) 0=(- rsin0,"cos0, 1), so I)r x I) e=(sine,- cos0,r). Therefore, we get
1
Jis (1,1,1) .{Sin 8, - cOs8, r )d7>dO =1 1"/2 (sino-coso+ r) dBdr =11Cir)
On the other hand, the boundary, aS, is composed of four parts. First, when r =1, we
have 1)(1,B)= (cose,sinfJ, B), so F = (e,cosB, sine) and ds = dl)(1,8) = (- sin e,cosB, 1)dB.
Therefore,
1 l
Tr /2
Fds = (0,cos8,8inB) . (- sin0,cosB, 1)dO.
BS, 0
Using integrat ion by parts tointegrate -esinBand the half-angleformul ato integratecos
2
8,
we get
11"
(
0 .20) ]1"'/2
[
(BcosB- sine) + "2 + SI: - cosO 0
4
When e= 11"/2, orientation is maintainedby letting r go from 1to O. Thus, we have
f F . ds =
iaS2 il
THE INTEGRAL THEOREMS OF VECTOR ANALYSIS 149
When r = 0, B goes from 7r/2 to 0, so we get
[ F ds = [0 (B , 0,0) .(0, 0,1)dB = o.
las. l-rr/2
Si mi larly, when B= 0, we get
[ F ds= [l(O,,, O).(l , O,O) dr =O.
las. l o
Adding all t he parts together, we get IasF .ds = 7r/4 + 0+ 0+0= 7r/4, so theorem 6 is
verified.
25. For a direct computation,paramet rize the sUl'face as follows: Le x = T' cos B and y = rsinB,
so z = t(x2+y2 )= r
2
/2. Also,we want z ~ 2, so 0~ ,.
2
/2 ~ 2, or r ~ 2. In addi tion,
we have 0 ~ B ~ 210 . W calculate T o= (- 1'si nB,rcosB,O) and Tr = (cosB, inB, r),so the
out ward normal is T8 X Tr = (1' 2cose, r2sinB, -1'). Also, we calculate
i j k
2 (-1 -1 )
-(-z + x 0 -z - 3) - - r
4
+ 1'cosB 0-1' 2 -3 \7 x F = a/ax a/ay a/az
- , , - 4 ' , 2 .
3y - xz _
yz
2
Finally, I Is(\7 x F).dS becomes
1
2 r2rr
o lo (\7 x F) .(T o x T r ) dBd1'
12(27r 1
3
+ 67rr )dr= ( 7r;4 + 37r 7'2)I: = 207r .
On the ot her hand, by Stokes' theorem, I1s(\7 x F).dS = IasF .ds. The boundary is as,
which is the circleofradi us 2in the plane z = 2. Itcanbe parametrized by (2cost,- 2sint ,2)
for t ~ 27r. We use t his orientation b cause the surface lies below t he boundary, so we
should traverse itin a clockwise orientation. We computeds = (-2sint ,- 2cost ,0) dt,so
211'
[ F ds = 1(- 6sint, - 4cost,8sint )(- 2sint, -2 cost ,O) dt
las
r
2
-rr
l o (12sin
2
t+ 8cos
2
t )dt
[12 ~ - Si : 2t )+ 8 ~ + Si:2t)] C= 207r.
Ifone choo es the other orientation, t he answer should be - 201T.
8.3: CONSERVATIVE FIELDS
GOALS
1. Understand that the line integral ofa gradient field is path independent.
2. Be able to determi ne whether a field is conservative.
150 CHAPTER 8
3. Gi ven a conservative vector field, be able to fi nd a scalar function whose gradient is equal to
that vector field.
STUDYHINTS
1. Theorem 7. This is a very important theorem. To summarize, itstates thatifF is agradient,
then \l x F =0, the line integral depends only on theendpoints, and all lineintegrals around
closed curves are O. Also, ifone these conditions hold, then F isa gradient. If the conditions
oft he t heorem hold,then wesay t hat t he lineintegrals are "independent ofpath." Note that
ifa single lineintegral is 0, then F may not necessarily be a gradi ent.
2. Example 1. In method 2, part (a), notice how after integrating in x, we add a "constant"
h
1
(y,z). It is a "constant" because it only involves the other variables. Study example 1
carefully. You shoul d knowhowto use at least oneof the two methods.
3. Gradi ents in1R2. Thecoroll ary precedi ng example3tells you that ifthe integrand in Green's
theorem, oQ/Bx - BP/By, is 0, then F =Pi+ Qj
is a gradient . Be careful!! F must be C
1
on all of
]R. 2, unlike theorem 7, which allows some exceptional
points in1R3. Exercise 12 el aborates this point . If the
x
origin is an exceptional pointin IR2, then the integral
of F over t.he pat h on the left (a closed path) is 0
because the path does not include the origin, while
theintegral ofF over the path on the right is not.
4. Is F a curl? F is a curl ofsome vector field ifdi v F =O. Exercise 16 explains t he procedure
for finding a G such t hatF =curl G.
SOLUTIONS TO SELECTED EXERCISES
2. (a) Si nce y = 2X2, we have dy= 4xdx, and so
[ [1 t 1 8 19
2 3 6
lcFds =10(x2X2)dx+ (2X ) 2.4xdx=10(2x + 16x )dx=2" + 3' =6'
(b) The answer is yes, since
j k
\7 x F =I B/ ox a/ay o/Bz 1= (0, 0,-x)f- O.
xy y2 0
Alternatively, one can pick a different path and show t hat theline integral is not equal to li.
This would mean t hat t he li ne integral ofF is path dependent.
3. If F =\l f , then t he x componentofF must be of /ax, i.e.,
of/ox=2xyz+ sinx. (1)
Similarly, the ycomponent ofF must be oj/ oyand the zcomponent ofF must be oj/ oz,
%.e.,
of/ay x
2
z (2)
Bj/oz x
2
y. (3)
Integrating (1) with respect to x, we get f =f(2xyz + sinx) dx = z 2yz - cosx + h(y,z),
where h is a function of y and z only. When we integrate with respect to x we should restore
all the terms with x. Theterms without x are treated as a constant when differentiated, and
151 THE INTEGRAL THEOREMS OF VECTOR ANALYSIS
integration with respect to xcannotrestore them. Similarly, integrate (2) with respect to yto
get
2
1=Jx
2
zdy= x zy+ g(X,z),
where g(x,z) is a function of x and z only. Integrating (3) wit h respect to z gives us I =
f x
2
ydz =x
2
yz + k(x,V), where k(x,y) is a function of x and y only. Compare t he three
results:
2 2 2
I(x,y, z) =x yz - cosx+ h(y,z)=x yz+ g(x,z) =x yz+ k(x,V) .
We conclude (by insp ction) t hat g(x,z) =k( x,y)=- cosx+C and h(y,z) =C, where Cis
a constant. Thus,
I(x, y,z) =x
2
yz - cosX +C.
6. (a) Thisfact shoul d already be familiar (see Example 6 in Section 2.6). Here are t he details.
We use the chain rul e to compute (alox)(1/Jx2+ y2 + z2 ) =-xl(x
2
+ y2 + z2)3/2. Then,
by symmetry, we get
"V ~ ) = '\7 ( 1 ) = - (xi + yj + zk) = =-r
l ' J x2 + y2+Z2 (x
2
+ y2 + z2)3/2 IIrl13 r3 .
(b) Since F is the gradient of a function j, fcF .ds = I(c(a)) - I(c(b)) for all paths c(t)
beginning at a and endi ng at b, unless the pat h passes t hrough the origin, where 1/r is not
continuous. Thus , t he work done by F in moving a particlefromro "to 00" is
1 _ lim ~ ) _ 1
J x
2
+ y2 + z2 r-+oo r -..;x
2
+ y2 + z2.
Notice that t his tells you it is moredifficultto move a particlefroma position neartheorigin.
9. The reader should verify that '\7 x F = O. Hence, F is a gradient ofsome function I(x,y,z).
Integrate the i component with respect to x, the j component with respect to y, and the k
component wi h respect to z. Comparingthe results, we see t hat I(x,y, z) =eX siny+z
3
/3.
Also,we compute c(O) =(0,0,1) and c(l )=(1 ,1,e). Since F is a gradient, we have
1
e3 1
cFds =l(c(l)) - l(c(O)) =esin1+ 3 - 3'
13. (b) F is not the gradient ofa scalar functionf. If such j were to exist, then aI I ax =xy and
ofI 8y =xv By t h equalityofmixed partials,we wouldexpect that8
2
I I 8x8y=8
2
I I 8y8x.
But in this case, (818y)(8f!8x)=x and (818x)(8118y) =y.
15. (b) To show that F is conservative, we can show t hat it is a gradient . (We can also show
that anyone ofthe four conditions in theorem7 is met) . We will use the same method as in
exercise 13. The partialderivat ives are:
8 ( 2X ) (2x)(- 2y) -4xy
8y y2+1 = (y2 +1)2 = (y2+1)2;
8 (-2y(x
2
+ 1)) - 4xy
8x \ (y2 + 1)2 =(y2 + 1)2 '
Sincethese part ials areequal, F is conservative. FindingI so thatF = "V Imakestheevaluation
ofthe path integral easy. You should verify t hat if I(x,y) =(x
2
+1) /(y2 +1) , then F = '\7f.
3 6
[(t
3
Then substitute x=t - 1 and y =t - t to get I(t) = - 1)2 +1]/[(t
6
- t)2 + 1], and so
1
[ [( t
3
_1)2+
1
]1
lcFds = (t6_ t)2 + 1 o::::;;-l.
152
CHAPTER 8
18. First, we compute "V .F =(8/8x)xz+ (8/oy)(-yz)+ (%z)y=z- z+ 0 =O. Thus, there
exists a G such that F = "V x G. To find G,use the result ofexercise 16:
l
z
lY _ y2 lz -
2
yz2 xz
G
1
= -ytdt - tdt =-- - -, G
2
=- xtdt =--,and G3 =O.
o 0 2 2 0 2
Hence, G = + y2)i + xz
2
jJ. It is a good idea to compute curl G to verify your
answer:
1 I J k
= (- 2xzi+ 2yzj+ (z2 - z2 - 2y) k)
yz2 + y2 xz2 0
xzi- yzj+ yk = F.
"Vx G
2 a/ax a/oy 8/8z
Not e that G is not unique;for example,arbitraryconstants may be added toeach component
ofG, or any gradient may be added toG, and "V x G would still beequal to F.
23. (a) Recall from chapter 4 that F is not irrotational when curl F # O. Thus, we compute
curl F :
J k
"V x F =I 0/ox 0/8y 8/OZ I =2k #O.
- y x 0
(b) Let c(t) =(x(t),y(t),z(t)) be the trajectory ofthe cork. By the definition offlow lines,
c/(t)=F(c(t)), and so c/(t) =(-y(t),x(t),O). Equi valently, we have the following system of
differential equations:
Xl (t) -y(t) (1)
yl(t) x(t) (2)
Zl (t) o. (3)
Equati on (3) canbesolved easily. Itssoluti onis z(t) =constant. Taki ngone morederi vativeof
(1) withrespect t o tyields X"(t) =_yl(t),but (2)impli esthatx"(t) =- x (t), or x"(t)+x(t )=
O. This equation represents a harmonic oscillator. The famous solution to the differential
equation x"(t) + x(t) =0 is
x(t) =Asint+Bcost,
where A and B are cons t ants. If t he reader is unfamiliar with the techniques ofsolving
ordinarydifferentialequations ,thereadermayverifythisisindeed thesolutionbysubstitution.
Similarly,
y(t) =Csint+ Dcost ,
where C and D are constants. Since x"(t) =-y'(t), we differentiate x and y and compare
t erms to find that C =B and D =-A. Therefore, y(t) =Bsint- Acost. Squaring xand y
and addi ng themgives us
x2 + y2
A2sin
2
t+ 2ABsintcost+ B2cos
2
t+ B2sin
2
t
-2ABsintcost + A2cos
2
t = A2+ B2.
Since A2+ B2 is a constant, we recognize t he equation x
2
+ y2=A2+ B2as the equation of
a circle ofradius VA2 + B2 centered at (0,0) . Thus, the cork has a ci rcular trajectory about
the zaxis in a plane parallel to the pl ane.
(c) As y increases, x decreases, since xl(t) =-yo We also know t hat the cork is going in a
circle. Thus t he cork is revolving counterclockwise.
24. (c) The property ofbei ng rotational is a local property, t.hat is, the field is rot ational at a
point. In exercise 23, t he cork whi rls whil e going around in a circle, but in exercise 24 , the
cork does not. Single trajectories have little to do wit h the r otationalness ofthe fluid.
153 THE INTEGRAL THEOREMS OFVECTOR ANALYSIS
8.4: GAUSS'THEOREM
GOALS
1. Be able tos ate and use Gauss' theorem.
2. Be able t o use Gauss ' t heorem to computea double integral over a closed surface or a triple
int egral over a volumeenclosed by a surface.
STUDYHINTS
1. Definition. A closed surf ace is a surface which, roughlyspeaking, must bepunctured in order
to get into the region it encloses. The enclosed region is denoted Wand the closed surface is
denoted oW.
2. Gauss} dwergence theorem. IfoW is a closed surface, then
fffw (diV F)dV =fhw(F.n)dS.
Thus, a triple integral is reduced to a double integral , or vice versa. Compare this to the
divergence theorem in the plane (section 8.1) .
3. Physlcal i nterpretation. div F(P) is t he net outward flow at the poi nt P per unit vol ume. If
div F(P) >0 materialflows out from P,and ifdiv F (P) <0,material fl ows in towards P.If
div F(P) =0,t he vector field is divergence free, t hat is, what goes in mustcome out .
SOLUTIONS TO SELECTED EXERCISES
4. (a) For the faces parallel to the yz plane, the out ward pointing unit normal vectors are j and
- i,respectively. For those two faces , we have
fiaw
F
'
dS
= 1
1
1 \ i +
j
+k) . idYdZ +1
1
1 \i +
j
+k).( - i)d
y
dz
111\i +j +k).(i - i)dydz=o.
Theoutward pointing unit normalvectors n for any twoparallelfaces are the exact opposite,
so the integrals over any t wo parallel faces ofthe cube cancel. Therefore, the int gral is O.
Next , we calculate thatdiv F =0, so by the di vergence theorem, t he desired integral is
fffwdiVFdV=O.
6. (b) First , we see that div F =1+1+1=3,so by thedi vergence theorem,
flawF .dS = 1ffw div F dV = 3fffw dV,
which is t hree ti mes the vol umeof W. The problemnow is to fi nd th volumeofthe region of
interest. Use "cylindrical" coordinates. We will let f) range from -1r/2 t o 7r /2since x ~ O. In
addition, we have :S r :S 1and x
2
+y2 =1'2 :S Z :S 1. Therefore,
7r
,"/21111 11 (11)
dV =J rdzdrde=7r r (l - 1'2) dr =7r - - -
fliW -1f'/ 2 0 r 2 0 2 4
4
Thus,
flaw FdS =3 ~ ) =3:.
154 CHAPTER8
10. Use the divergence theorem. We have div F =(x
2
+y2)2, so
fhs F .dS = ffis(x
2
+y2)2dV.
Using cylindrical coordinates, thecylinder S can be described by 0::;: T ::;: 1,0::;: B ::;: 21!', and
o::;: z ::;: 1. Since the Jacobian is r, we get
5
fhs F .dS =121< 1110
1
T' r
4
drdzdB =21!' 11 r dr=i'
16. From section 8. 3, we know that curl F = 0 implies t hat F = \7f, so t aking the divergence
of both sides gives us div F =\7 .(\7J) =\7
2
f. But we are given that div F = O. Hence,
\7
2
f = O.
18. By the divergence theorem, fffw div FdV =ffow F .dS. Given t hat F is t angent to the
surface, we knowt hat F is perpendicular t o t he unitnormalofthesurface S, andso F dS= O.
Hence, IIIwdi v F dV = O.
8. 5: SOME DIFFERENTIAL EQUATIONS OF MATHEMATICAL P HYSI CS
GOALS
1. Be abl e to use vect or analysis techniques in applicat ions such as conservation laws, heat
t ransfer and electromagnetism.
STUDYHINTS
1. Notation. There'sa lotofnew notation,soyoumayfind itusefultomake a listofnewsymbols
and t heir defini t ions.
2. Oldconcepts. No newvector calculusmaterial is present ed in this section. Only the t erminol-
ogy has changed. You should be abl e t o just ify each step with the knowl edge t hat you have
acquired in t he previous chapters.
SOLUTI ONS TO SELECTED EXERCI SES
3. Using t he t r ansport equation for p, we have
:t fJfw p(x, y,z,t )dx dydz = fJfw (: +pdiv F)dx dydz
Jf/w +\7P.F +pdiv F) dxdydz
Jffw +\7 .(PF)) dxdydz
o
for any HI in R3. Thus, theintegrand must be identically 0, i.e.,
ap +\7 .(pF) =0,
at
which is t he lawofconservat ion of mass.
8. Given \7
2
rjJ =0 and V =\7rjJ,we want toshow t hat p(aV l Ot +V\7V)=- \7pifav lot =O.
Using exercise 7,V .'VV=!'V(II VI1 2)+('V x V ) x V. Since V is a gradi ent ,'V x V =O. It
suffi ces t o show t hat =- 'Vp. Choose p = -(pI2) IIV W.
THE INTEGRAL THEOREMS OF VECTOR ANALYSIS l 55
9. Start with Ampere's law: \7 x H - fJE / fJt =J. Take the divergence ofboth sides: di v J =
\7 .(\7 x H) - \7 .(8j&t )E =0- (fJ/ fJt )(\7 .E) (the change oforder oft he two derivatives is
j ustified since \7 takes thederivativesofspatialvariablesonly). Now,byGauss'law,\7.E =p.
Combiningthetworesults, we getdiv J =- op/ at , ordiv J+ fJp/ fJt =0,which is theequation
ofcontinuity.
10. (a) By definition,the integral ofthe Poynti ng vector field is
By Gauss' theorem, we mayrewrite the right-hand side as
J J 1\7 .(Ex H)dV.
Equation 8 from the table of basic identities ofvector analysis in section 4.4 tells us that
div(Fx G)= G .curl F - F .curl G. Thus, the integrand above is
V(E x H) div(E x H) =H .curl E - Ecurl H
H .(\7 xE)- E .(\7 x H)
H O- EJ = -EJ
since x E = 0 and V x H = J. Therefore,
Jis p .dS=J J 1\7.(E x H)dV= - J J 1E .J dV.
8.6: DIFFERENTIAL FORMS
GOALS
1. Be able toadd and wedge multiplyk-forms.
2. Be able tointegrate and differentiate k-forms.
3. Be able to tate Greens, Stokes' and Gauss' t heorems in termsof k-forms.
STUDYHI TS
1. Notation. In this textbook, w is used for I-forms, 7} is used for 2-forms and IJ is used for
3-forms. (Thisis generally the case, but not always .)
2. Definition .
(a) O-forms arescalarfuncti ons.
(b) l-formscontain dx, dy, or dz .
(c) 2-formscontain dxdy,dydz, ordzdx in theorder specified.
(d) 3-formscontain t he expression dxdydz .
These definitions will differ slightly inhi gher dimensions.
156
CHAPTER 8
3. Int egration.
(a) I-formsare integrated like lineintegrals.
(b) 2-formsare integrated likesurface integrals. Notice thateach t ermofthedouble integral
is equa.tion (2) contains a Jacobian. The variables in the Jacobian's "numerator" follow
the cyclic natureof the differentials.
(c) 3-formsare integrated li ke ordinary triple integral s.
4. Wedge products. Different formsmay be mul tiplied. In our discussion ,you can only multiply
a k-form and an I-form if 0 k +I :::; 3. Otherwise, products are zero. Be aware that
w/\ J1 = (-1)101 (I-' /\ w). This property is cal led anticommutativity. Notice that dx /\ dx =
dy/\dy = dz/\dz = O. Theother properties are similartot hose ofrealnumbermultiplication.
5. Differentiation. Iffis a O-form, df is likethe ordi nary derivative. The "sum" rule (linearity)
remainst hesameas inone-variablecalculus. The"product" rulediffersslightlyandd(dw) = 0;
tbislast ident ity captures the t woidentities 'V x 'Vf = 0 and 'V .('V x F )= 0in one formula.
SOLUTIONS TO SELECTED EXERCISES
l. (d) Since this is a 2-form wedged with a I -form, i.e., a 3-form,the product is a (2 +I)-form.
By thedistribut ive property, we get
w/\ry = (xy dy dz + x
2
dxdy)/\(dx + dz )
xydydz/\ dx +x
2
dxdy/\dx+x ydydz/\ dz +x
2
dxdy/\dz
xy(dydz/\dx)+x
2
(dxdy/\dx)+xy(dydz/\dz) +x
2
(dxdy/\dz)
= xy dxdy dz+x
2
dxdydz= (xy+x
2
)dxdydz.
We used the anticommutativity fact that dydz /\ dx = (-I)2( dx /\ dydz). Also, we used the
associativity property along wi t h the identities dx/\ dx = dz/\ dz = O.
3. (b) When we differentiate this I-form, we get a 2-form:
dw = [d(y2cosx)/\ dy+(_I)0(y2cosx )/\ d( dy)]+ [d(xy)/\ dx +(-l)o(xy) /\d(dx)]+d(dz)
= [(_y2sinxdx+2ycosxdy) /\dy+0] +[(ydx+xdy)/\dx+OJ +0
= _y2sinxdxdy- xdxdy= (_y2sinx - x )dxdy
since dx/\dx =dy/\ dy=0 and dy/\ dx =- dxdy.
(e) When we di fferentiate this 2-form,we get a 3-form:
dw =
2
= [d(x /\dy + (_1)0 (x
2
+y2)/\ d( dy)]/\dz +0
= [( 2xdx+2ydy)/\dy]/\dz = [2xdx/\dy+2ydy /\dy]/\dz = 2xdxdydz
since d(dz) =d(dy)=dy/\dy =O.
4. Notice t hat dx/\ dxdy=dy/\ dxdy =dy/\dydz =dz /\ dydz =dz/\dzdx =dx /\dzdx = O.
Thederi vative of Fdxdy is
oF of OF ]
[
-8 dx+-;:;- dy+ -0 dz /\ dxdy+(-1) F/\d(dxdy)
x ay z
of of of of
ox dx/\ dxdy+oy dy /\ dxdy+ OZ dz /\dxdy== oz dz /\ dxdy
of of
oz dz /\ (d:c /\ dy)= az (dz/\dx )/\dy
of of
- 7h(d:c /\dz)/\ dy =- 7hdx /\ (dz /\dy)
of of
8zdx /\ (dy/\dz)= fudxdydz.
157 THE INTEGRAL TH EOREMS OF VECTOR ANALYSIS
Si milarly the derivative of Gdy dz is (oG j ax) dx dy dz and the derivative of H dz dx IS
(oHj oy) dx dydz . Therefore,
OG 8H OF) .
dl1= ox + ay + Tz dx dydz= (d1v V)dx dy dz .
(
8. By a directcalculation,w notethatasis theunitcircle inthe xy plane. Theparametrization
is (x,y, z) =(cost, sint , 0), 0S; t S; 27r . Therefore,
isw = 12Tr [(cost + sint)O + (sint + 0)( - sint dt ) + (cost + O)(cost dt)]
2Tr 2Tr
r
2
r ' 2tl2Tr 2
= in (- sin t + cos t) dt =io cos2t dt =S1; 0 =O.
By Stokes theorem, the above calculationis equal to ffs dw. We compute dw:
d = [d(x + y) !\ dz + (x +y) !\ d(dz)] + [d(y + z) !\ dx + (y + z) !\ d(dx)]
[d(x + z ) !\ dy + (x + z ) !\ d( dy)]
= [(dx + dy ) !\ dz) + [(dy + dz) !\ dx) + [(dx + dz) !\ dy)
because d(dx) =d(dy) =d( dz ) =O. Thissimplifies to
dx A dz + dy !\ dz + dy !\ dx + dz !\ dx + dx !\ dy + dz !\ dy
= -dz dx + dy dz - dx dy + dz dx + dx dy - dy dz = O.
Therefore, we have ffs dw =0, aLso.
l!. By Stokes theorem, we have fJaR T} =f f f R d1J. We will look atthe right-handside:
d(x dy dz + y dz dx + z dx dy)
= d[(x dy Adz) + (y dz !\ dx ) + (z dx Ady))
= [d(x dy)!\ dz + (-l)l(Xdy A d{dz ))] + [d(ydz ) !\ dx + (-1/ (y dz!\ d(dx)))
+[d(zdx)!\ dy + (-1)1( z dx !\ d(dy))]
= [dx A dy + x!\ d(dy ))!\ dz + [dy !\ dz + y!\ d(dz )) !\ dx + [dz A dx + z!\ d(dx))!\ dy.
Since d(dx) =d(dy) =d(dz ) =0, we get
(dx Ady) !\ dz + (dy !\ dz ) !\ dx + (dz !\ dx)!\ dy
= dx dy dz + dx !\ (dy !\ dz ) + ( -dx !\ dz ) !\ dy
= dx dydz + dxdydz - dx !\ (dz Ady)
= 3dx dydz.
Thus, the right-hand side is ffJR3dxdydz. We recognize f ffRdx dy dz as the volume ofR.
Dividing by 3gives us the desi red result.
SOLUTIONS TO SELECTED REVIEW EXERCISES FORCHAPTER 8
2. By Gauss' t heorem, we have ffaw HdS =fffw C'v H)dV =fffw( div H)dV, so we need
to how thatdiv[F x (\7 x G)]= (\7 x F).(\7 x G)- F (\7 x \7 x G). By formula 8in the
table ofbasic ident ities ofvector analysis in section 4.4, we have
div(A x B)= B .curl A - A .curl B = B(\7 x A)- A (\7 x B).
Now let A =F and B =\7 x G. This gives the desired result since a .b =ba.
158
CHAPTER 8
4. According to Green's theorem, Ie Pdx+ Qdy =IID(8Q/ox- {)P/fJy) dxdy, where C is the
2
boundary ofDand Chas acounterclockwise orient ation. With P(x,y) =x yandQ(x, y) =y,
the right-band side is easi ly computed as
2 2
fL(0 - x ) dxdy = 1
1
1:(_ x ) dydx = 11 (
- X2yl:=x3) dx
1
1\-x
3
+ x
5
) dx= (_;4
12
3
t
3
For the left-hand side, first parametrize the curve C
1
: Y =x as x =t, y = , 0 t l.
Then
1
1 11 2 4t 11
1
x y dx+y dy= t
2
(t
3
)(dt)+(t
3
)(3t
2
dt ) = 4t
5
dt =-
6
2
C,OD 6 0
3
For the curve C
2
: Y =x, a counterclockwise parametrization is x =1- t, Y=1- t ,0 t l.
Then
2
[ x y dx +ydy =([(1 - t)2(1 - t)(-dt)+(1- t)(-dt)l= t[(1 - t)3 +(1 - t)l(-dt).
le. l o
Substitut ing u = 1- t yields
(u
4
0
2
[ 0 u) 1
3
3
I I (u +u) du= 4+'2 1 -4'
Thus, the left-hand side is - =- 1
1
2' and Green's theorem is verified for this case.
7. (a) To show that F is conservative, the easiest. thing to do is toshow that Y' x F =0:
k
J
Y'x F a/ax %y %z
6x y(cos z) 3x
2
(cosz) - 3x
2
y(sinz)
(- 3x
2
sinz+3x
2
sinz,5xysinz- 6xysinz , 5xycosz - 6xycosz)
O.
(b) IfF is the gradientofsome f(x, y, z), then /(x, y, z) must sat isfy
6xycosz o//ox (1)
3x
2
cosz of/oy (2)
-3x
2
ysinz of/oz. (3)
Integration of (3) with respect to z gives f( x,y, z) =3x
2
ycosZ +g(x,V); integration of (2)
with respect to ygi ves f(x,y, z) =3x
2
y cosz +h(x,z); integrati on of (1) with respect to x
gives f( x ,y,z)= 3x
2
ycosz+ k(y,z). Comparing these three f's, we can see that f( x , y,z)=
3x
2
ycosz+ C .
(c) Since F is a gradient, we only need to evaluat e f at the endpoints t o calcula.te the line
integral . We get
1
F ds =f(c(1r/2)) - f(c(O)) = 3(cos
3
tI)2(sin
3
tI)
n/2
=(J.
c
I
0
9. We want to compute IIIwV' .F dV where W is the unitcube. By Gauss' theorem,
fffw Y' .F dV =
t t ( (5z +x
2
+y) dxdydz
lolo
1111 +5z+y) dydz
1 (1 1) (1 1 ) 23
-+6z+- dz = - +-+3 =-.
o 3 2 3 2 6 1
159 THE INTEGRAL THEOREMS OFVECTOR ANALYSIS
13. (a) We compute
'\71= (al/ax)i+(al/ay)j +(af/az)k=3yexp(z2)i+3xexp(z2)j+6xyzexp(z2)k.
(b) '\71 i a gradient, so the integral is l(c(11")) - l(c(O)). Since I(x ,y,z) = 3xyexp(z2),
we have I(c(t)) =3(3cos
3
t)( in
2
t ) exp(e
2t
) ; thus l(c(11")) =0 and l(c(O)) =O. Th refore,
Ie'\71 . ds=O.
(c) Let A be the region whose boundary is c(t). Then, by Stokes' theorem,
1'\71 ds =1['\7 x ('\7f)] .dS.
Thisis 0becau_ the uriofa gradientis alwaysOJ theoth r partofStokes' theorem has been
illustrated in part (b).
14. By Green' heorem, Ie x
3
dy - dx =IID(3x
2
+3y2)dxdy, where Dis the unit disc. Now
use polar coordinates: x = rcos8, y =rsin8. The unit circle is described by 0 r 1 and
o 8 2- . ince the Jacobian is r, the integral b omes
1
2"1
1
3r
2
. r drd8= 211"
11
3r
3
dr = 211" .-
3
=-11"
3
.
o 0 0 4 2
17. By Green' eo em.
Area = Jldzdy = l(xdy- ydx)
= 1(asin8cos8)(2asin8cos8) d8- 1'fr(asm
2
8)[a (cos
2
8- sin
2
8)] de
_
1a2 fo ( 2 2 2 2 4
= Jo 2sin 8cos 8- sm ()cos () +sin 8)de
2
= f in
2
O(cos
2
() +sin
2
() ) d8= a rsin
2
8d8
2 Jo 2 J
o
1C- 2() )d()
= Q ? I:=11":2.
2
20. (c) IfF ' a gradient, then curl F =O. We compute
, .
I J
curl F = a/ax a/ay
2 3
x z
Excep a theorigin,curl F is not 0,so F is not a gradient.
TEST FORCHAPTER8
1. Trueor false. Iffalse, explain why.
(a) For anyregion to which Gre n's theorem applies and ifCis t raversed counterclo kwise,
the line integral Ie(4x - sinx+eX +2y)dx +(4x +cosy+2
Y
) dy has positive value.
(b) Thevector field
F(z.y z) = (y
2
ecosz)i+(2xye
Z
cosz+2ycos(y2))j+(xy
2
e
z
cosz- xy
2
e
z
sinz)k
is a conservative vector field.
160
CHAPTER 8
(c) If F is a gradient, then curl F =O.
(d) The line integral IePdx+Qdy is independent ofpath ifaP/oy =oQ/ax on all oflR2.
(e) If y = l(x )is a posi tive function on [0, 1] ,then the areaunder thecurve is
~ 11[t (J' (t) dt)- I(t)dt].
2. Let C be the perimeter of a trapezoid with vertices at (0 , 0), (4,0) , (4, 1) and (2,1). IT C is
traversed in t he counterclockwise direction, starting and ending at the origin, calculate
3
fc (2x+x - Y +sinx)dx + (e
Y
- x - tany+ Vii) dy.
3. (a) Compute 'il l for I(x , y, z )= 2X
2
Z
2
sinycos(e
X
) .
(b) Let F ='il l as in part (a). Compute IcF. ds where C is t he portion of t he pl ane
2x+y + Z =3 thatlies inside the sphere x
2
+y 2 + Z2 =16.
4. Let F(x,y,z)= (3x+ysinz,x
2
+ y+ ze
x
, xe:ll y+ z). Let P bet he parallelepiped spanned by
the vectors (1,2, 1), (3,0,1) and (- 1,3,2) and having one vertex at t he origin. Compute the
i.ntegral ofF over oP.
5. (a) Computecurl F for F( x,y,z) = - 2yi+ 2xj+xyz
3
k.
(b) For F as in part (a), compute IIscurl FdS where S is the semiell ipsoid
x2 2)1/ 2
z= l - - - !L
( 4 9 .
6. Acurve is described paramet rically by x =sint ,y =si n2t,0 ~ t ~ 1r. Find the areaenclosed
by t he curve. (Hint: sin(mx)cos(nx) = [ s i n m - n)x +sin(m+n)x].)
7. A solid W is described by -1 x ~ 1, 0~ y ~ 4- x ,0~ z ~ 10 - x
2
- y . Computethe fl ux
of F( x ,y,z)=(y+ z2)i+ (xy- sinz)j+ (5xe!l )k across oW.
x
2
8. Suppose t hat Iiscurl F dS =5 for a given F and S is the cone y =4- - Z2 , 0~ Y~ 4.
(a) Compute f ITcurl FdS,where T isthesemiellipsoid y= 5(1- x
2
/4- z2 /4)1/2,0 ~ Y ~ 5.
(b) Compute Ilucurl F .dS,where Uis t he paraboloid (y +2)2 =x
2
+ z2,- 2 ~ y ~ O.
9. For a closed curve Ct hat is traversed in a counterclockwise di rection, it has been determi ned
that Ieydx- xdy is -6. Calculate Ie(x +2y)dx+ (5x - ye!l) dy.
10. (a) Show thatF(x,y, z)={2xyz3 +z)i+(X
2
Z
3
+l)j+(3x
2
yz2 +x)k is aconservative vector
fi eld.
(b) Find a fUDctionI such t hat 'il l = F .
(c) Suppose the same force fi eld F as in (a) causes a baby to make a random crawl from
(0,0,0) to (5,e,0) where she knocks over a priceless obj ect . How much work is done by
the force field on the baby?
(d) Theforcefield F then causes t he baby tocrawlupsomecabinetdrawers to (I,1, 1). How
much work was done by the force field in taking the baby from (0,0, 0) to (1 , 1,I) ?
161
9 SAMPLE EXAMS
9.1: COMP REHENSIVE TEST FOR CHAPTERS 1 - 4
x
2
1. (a) Find the relativeextremaof f(x, y) = - 2x + y2 - 1.
(b) For the samet, find the relativeextrema on the curve x
2
+ y2 = 1.
2. (a) Fido is on a strange mountai n which has a shape described by M (x,y) = e-Y xsiny. If
Fido is at (1,1':/ 2, e-
rr
/
2
) , whatis thedirection Fidomusttake toget down the mountain
as fast as possible?
(b) Fido has to foll owa certain trail oftree trunks (which he has previously marked) along
e(i) =(2t , eexp(-t
2
), 2tsin(t
2
)) . How fast is Fido'salt it udechanging at t =I?
3. (a) Given the vector field F(x, y) =(y, -x) , find anexpression for a flow line c(t) for F by
taking a good guess.
(b)Do thesamefor F(x, y) = (2x , - y).
4. (a) Find the arc length ofe(t) =(2t,t2), O:'S t:'S l.
(b) Find the arc I ngt h ofe(t) =(2sint,sin
2
t ), 0 :'S t:'S 1': / 2.
(c) Explain your answers.
5. A wooden box is to be made with $120.00 worth ofwood. The lid is to be made from wood
that costs $2.00 per square inch, and the rest ofthe box is to be madefrom wood that costs
$2.50 per square inch. Whatis t he biggest box that could be made?
6. (a) Findanequat ionofthelineofintersectionoftheplanes 3x +2y -z =7andx -4y+2z = O.
(b) Find the equation ofthe plane which contains the points (1 ,3,-2) and (0,- 2, 1) and is
perpendicular tothe plane3x - y - 2z = 5.
7. Let t: B C ~ 2 -t 3 be defined by (u,v) -t (uv
2
,v
3
- u,vsinu), and g: B C IR3 -t IR2 be
defined by (x , y, z) -t (yz e"', y3 cosxz).
(a) Calculate D(J 0 g)(O , 1,0).
(b) Calculate D (g 0 f )(O, 1) .
8. An aquarium manufacturer bas advert ised that its tanks hold exactly 0.49 m
3
of water . Due
to production error , t he tanks have di mension 1. 01 m x 0.72 m x 0.67 m rather than the
specified 1. 00 m x 0.70m x 0. 70m. Use the linear approxi mationtoestimatet he error in the
advertised capacity oft he t ank.
9. A budding bioengineering st udent who is an amateur ornit hologist wanted to fl y south for
Christmasvacation, but couldn't afford an airline ticket. He knew that his pet swallows also
wan ed tofly south for t he winter, so his plan was to tie lis feathered friends to a hang glider
and have them fl y hi m home. His well-trained swallows had a veloci ty vector of3i+ 4j + k
(km/ hr) unt il they r ached their cruising altitude at (0, 0, . At that point, they continued
along the samepath, but with no change in altitude, (Assume the Earth is fl at. )
(a) What was the originalstarting point at (x, y, O)?
(b) How long did it take to reach cruising altitude?
162
CHAPTER 9
(c) When they reached (0, 0, , a strong wind with veloci ty 2i +j affected the flight. What
was theswallow's veloci ty vector to maintaina total velocity of 3i+4j?
(d) After 3hours, the bioengineer calculates thatthey will passover hisworst enemy'shorne,
where some of his pets will give a 21-bird dropping salute. What is the position ofthe
enemy'shouse?
(e) Assuming that the swallows were unable to compensate for the wind, how far from the
t arget poi nt would they be after 3 hours?
10. Tightwad Terry, the miser , has performed a computer analysis ofhis earning power. He has
ti
det ermi ned thathisearning function is $(t, u, v, w) =t
2
e +v sinw.
(a) Compute $t, $u , $t1 and $w
(b) Assuming Tightwad Terry wants to increase his earnings as fast as possible and he was
at (1,0,2,0) on his $ graph, how should he change t, u, v and w?
(c) Ifhe has exactly one unit ofeach resource (represented by t, u, v and w), what. is the
largest increase in earning that he can expect? Note that he can use fractional amounts
ofresources, such as toft and ~ ofw.
9.2: COMPREHENSIVE TEST FOR CHAPTERS 5 - 8
1. (a) Eval uate
Jin 3(x +y)e(x-v) dx dy
over the region bounded by the lines x +y = 1, x +Y =- 1, x - y = 1 and x - y = ~ l .
(b) Evaluate
1111siny2dydx .
2. A block has a slanted top described by x +y +z = 2. Its edges are perpendicular to the xy
plane, and the bottom of t he block is formed by the triangle with vertices (1,0,0), (0, -1,0)
and (0 , 1,0). What is the volumeofthis block?
3. Find t he mass of a wall described by a$ y $ _x
2
- 2x +3, - 3 $ x $ 3, having cJiensity
o(x, y) =21yl +3.
4. Let F (x , y) =(eX cos3y, - 3e
x
sin3y).
(a) Find an f( x,y) such that Vf= F for aJi (x,y).
(b) Evaluate IeF .ds for thepath c(t ) =(cost, sint), $ t $ 211' .
(c) ComputeV x F.
(You should be able todo parts (b) and (c) in your head.)
5. (a) Suppose informationfl ows through an idiot 'sskull (asurfacedescri bed by x
2
+y2 +2z2 =
1) atthe veloci ty 2xi +3yj +zk. After brai nsurgery, his mW1bed skull, which remained
the sameshape, receives infor mationatthe velocity 5xi+yj +3zk. Thinkrngofthis as
a fl ux ofinformation through asurface, di d the surgery help? Explain.
(b) A particle moves in a path c(t ) =(2t, 3t , t ) in a force field F =(2x, 2y, 4z). What is the
work done in the ti meinterval 0 $ t $ 5?
6. (a) Acont actlenscanbe described as acapofasphereofradius R cutout by acone ofangle
1T/ 4. Find thesurface areaofthe lens. (Hint: Set up the lens in acorrect and convenient
coordinate system. )
163 SAMPLE EXAMS
(b) Slice a sphere anywhere with 2 parallel plane which are separated by
a fixed di tance d. The bands obtained always have the same surface
area. Prove this. What is the area? (Hint : Use spherical coordinates.)
7. (a) Find the volumeenclosed by the surface
a;2 +9y2 +z2- 2x+54y- 10z+106 =O.
(b) Find an expression for thesurface areaofthe surface above.
8. Which of thefollowi ngvector fields are conservative?
(a) F(x,y, z )= (x
2
+l/x,In(y+1), z)
- 3x -3Z) -3y
(b) F(x,y, z) = ( (x2 +y2 +z2)3/2' (x2 +y2 +z2)3/2' (x2+y2 +z2)3/2
(c) F( x,y, z) = (3yz ,2xz ,5xy)
9. Complete t he followingst atement: A vector field F is conservative if __.
(i) There is a vector field G such that F =curl G.
(ii) There is a scalar function fsuch that \7f = F.
(iii) div F = O.
(a) (i) only (b) (i) and (iii) (c) (ii) only (d) (ii) and (iii).
10. Consider the followingargument : given a vectorfield F,
by Stokes' theorem but
\7.(\7 xF).dS = { (\7 xF).dS
JJ
.f
}w }s=&W
by Gauss' t heorem, so everyt hing is 0because the divergence ofthe curl ofany vector field is
O. What's wrong?
9.3: COMPREHENSIVE TEST A FORCHAPTERS 1 - 8
l. Iftrue, state true. Iffalse, expl ain why.
(a) The path integral Ie27l'ds equals t hesurface areaofa cylinder ofradius 1and height 27l'
ifc(t ) =(cost,sint,t ), 0 t 27l'.
(b) A continuous fun tion is a different iable function.
(c) The line integral ofa mass density function along a curve is the massofthe curve.
(d) curl F =\7 .F .
(e) 8
2
f /8x8z =8
2
f/8zax for all conti nuous functions f(x,y,z).
164 CHAPTER9
2. Mult iple choice. Choose the correct answer.
(1) Which pairofvectors have the smallest angle between them?
(a) i- j ,j +2k
(b) 3i+2k,- 2j
(c) 2i- j +k, i- j +k
(d) There is insufficient informat ion.
(2) Green's theorem requires
(i) continuous fi rst parti al deri vat ives
(ii) conti nuous funct ions
(iii) any closed curve which is the boundary ofa region
(a) (i) only (b) (ii) onl (c) (ii) and (iii) (d) (i) and (iii) .
(3) Let !(x , y, z) = z2xex cos(yz) . A particle t ravels along a path c from (0,- 2/5,1T/4) to
(3,5, -1T/2). Iftheforce actingon t he part icle is 'VI, then t he work done on the particle
IS
(a) negat ive
(b) zero
(c) positive
(d) unknown si nce c is unknown
(4) For any vectors u and v,U (u x v) =
(i) v(u x v)
(ii)
(iii) (v x u).u
(a) (i) only (b) (ii) only (c) (i) and (ii) (d) (i) ,(ii) and (iii ).
3. Let F = xi+yk. ComputeIIs curl F .dS for the followi ngsurfaces S:
(a) x
2
+y2 +(z - 3)2 = 1, z 3
(b) 9x
2
+9y2 +z2 = 9,z
4. Consider the system
2
F1(u, v, x, y) u - v
2
+2x +x y =
F2(U, v, x, y) 2u +3v - 5x
2
= 0
(a) Show that one cannot solve for t l and v in terms ofx and yat (u, v) = (0,0).
(b) Show that au/ax exists at (u , v, x, y) = (3, 2,-1,0). Computeit .
5. Suppose g(u,v) = (uv
2
,u+2v,v) . At (x, y, z ) = (0,0,0), (u ,v) = (1,2) and
[
1-1 0]
D! = 2 3 1 .
Let h = 9 0 f. Whatis Dh(O, 0,o)?
6. A surface is parametri zed by
2 2
x = u , Y = v , Z = uv , u 2, v 2.
165 SAMPLE EXAMS
(a) Find the tangent planeat (u, v) =(1, 1) as a function ofxand y.
(b) When u = 1, find the arc length ofthe curve in space for 0 ~ v ~ 2.
7. Let ~ be thesurface parametrized by
x:;:;;: e
U
cosv, y =e
U
sinv, z =v, 0 ~ u ~ 1, 0 ~ v ~ 1r/2.
(a) Find the areaofthe surface ~
(b) Compute the average ofzover ~
8. Letf(x,y)=x
2
-3x+y-y2+5.
(a) Atallcritical points (xo,Yo), express (xo,Yo, f(Xo, Yo in spherical coordinates.
(b) Find the extremaoff on thecircle x
2
+ y2 = ~
9. A surface is described by the equation z = 2y+ cos1rX - ft. Consider the point (4, -1,1).
(a) Find the directional derivative in the direction ofi + 2j.
(b) Find the equation ofthe tangent plane.
10. A region inspace lies between the graphs ofz =16+ x
2
+ y2 and z =Jx
2
+ y2. The region
also lies inside the cylinder x
2
+ y2 =4.
(a) Express the volumeofthe region as a triple integral withcartesian coordinates.
(b) Rewrite your answer in (a) with cylindrical coordinates.
(c) Fi nd t he volumeofthe region.
9.4: COMPREHENSIVE TEST B FORCHAPTERS 1 - 8
1. Physical and geometric interpretations.
(a) What do you know about u and v ifu .v =O?
(b) What doyou know about u and v ifu x v =O?
(c) Whatphysical interpretation is associated with a negative divergence?
(d) Givea phy ical interpretation ofcurl V.
2
2. Consider the point (1, 1,1 ,1)ofthe function f(x,y, z,t) = x y+ zt- z.
(a) Compute the directional derivative in the direction of(2,0,6,4).
(b) Isfincreasingfaster in thedirectionof(2, 0,6,4) orin thedirectionof(1, 1,2,I)? Explain
youranswer.
3. (a) Thecyli nder x
2
+ y2 = 4 is cut by the plane x + y+ z =1. Show that the arc length of
the intersecting curve is
J8127rVI- cos0sin0 dO.
(b) For the intersecting curve, find an equation for the tangent line at (1,)3,-V3).
4. Find t he minimum and maximum of x + yz subject to the constraints x + z = 2y and
{( x,y) I (x,y) E [3, 5]x [0,2]}.
5. Let S be the boundary of a box B =[-2,2] x [- 1,1] x [-3,3], F 2xi+ 3zj+ 2yk and
G =x
3
i + 3zj+ 2yk.
(a) Compute the integral of\l . F over B.
(b) Compute ffsG .dS.
166 CHAPTER 9
(c) Supposet.he originatthecenterofBisshiftedto(8,-15,20) andthenrotated30
0
around
the yaxis. Compute IIsF .dS .
6. Aholeofradius is drilled through the axisofsymmetryofthe hemisphere x
2
+ y2 +z2= 1,
z O.
(a) Writethe vol umeofthe remainingpiece in Cartesian coordinates.
(b) Write the volumeofthe remainingpiece in cylindrical coordinates.
(c) Computethe volume.
7. A painter is scared of heights, so he charges z2 dollars per square
metertopaintobjectslocated atheight z. His latestjobis topaint
thesilo5shownhere. Theheightofthecylindricalpartis 3meters
and the radius ofthe hemispherical partis 2 meters.
(a) ComputeIIsdSand interpret geometrically.
(b) Computehow much t he paintercharges.
3
8. (a) ComputeI;Ix1cos(y2+3)dydx.
roo roo roo ('+'+ ')'/' d
(b) ComputeJo Jo Jo e- x Y Z dxdy z.
9. (a) Verify Stokes' theoremfor F =z3j + (x
3
- y3)j+ y3k over the hemisphere x
2
+ y2 + z2,
z O.
(b) For the same F as in (a), evaluate IIs('V' x F) .dS for the surface x
2
+ y2 +5z
2
, z O.
10. (a) Find a vector-valued function /(x,y,z) such that
- x zsin(xy)eCDs(x
y
) eeDS(XY) ]
D/(x,y, z) = [ - yzsin(xy)e
CD8
(x
y
)
y2sinz 2xysinz xy2cosz .
(b) For the region D shown at the right, let Vbe
the volumeofthe solid lyingbetween /(x,y) =
x
3
sinyandthexyplaneandlyingover D. Write
Vin the form IIIg(x,y,z)dzdydx.
(c) Rewrite your answer to part (b) in the form IIIg(x,y,z)dzdxdy.
y
1 x
167
Appen dix
ANSWERS TO CHAPTER TESTS
AND SAMPLE EXAMS
ANSWERS TO T EST FOR CH APTER 1
1. (a) False; dot products need to be defined for vectors in the same space, so u v is undefined.
(b) True.
(c) True.
(d) False; let t he vectors be i, j and k.
(e) False; v x a v =0 for a being any real number .
2. - 2x - 2y +z = 1
3. (a) a =-5i+ j + k;h = 4i - 2k
(b) s = [0,1 and t =[0, s)
(c) v'I4
4. } 11/15
5. 6V5
6. (a) l(t) = (1,3, - 2, 0) + t( - 1, --2, 1,1)
(b) (-1, - 2,1, 1) (2, 0, 3, 1)= 2 #0
7. (a) [!3 ~
(b) 1; area of th parallelogr am spanned by (1,0) and (3, -1).
(c) O' volume of th paralJelpiped spa.nned by (-2, 0, - 3), (8, 1, 5) and (2, 0,3) is 0; all three
vectors lie in a single plane.
8. (a) (-/3, 311"/4, 1)
(b) (- .../6/2, .../6/2, 1)
9. (a) lu ,vi =2 ~ 2V15 =II ull II vll
(b) (2, -1, 0, 1)/3
10. (a) x +y - 3z =3
(b) 5IIT
(c) 10/ IIT
ANSWERS TO T EST FOR CH APTER 2
1. (a) False; let f (x, y, z) = 1 if x, yor z = and f (x , y, z) = 0 otherwise.
(b) False; should be perpendicul ar.
168 APPENDIX
(c) True.
(d) False; consider z =Ix!-
(e) True.
3
2. [ ; 16 ]
3. 8.969
4. Positive xdirection.
5. z =12x+8y- 9
6. 8g/ 8xdoes not exist ;8g/8y== 0.
7. 8u/8s= 0; 8v/8t= 1
8. (a) 12
(b) lim(x,y)-+(O,O)!(x,y) does not exist because limx-+o!(x,O) =1 and iimy-+o!(O,y) =0;
t herefore, !(x, y) cannot be madecontinuous.
9. ~ ~ (1, 1) = 2+e> ~ ~ (1, 1)= 2
10. (a) (-3. 6,-36,- 0.02)
(b) J)ecreases by 3.6
(c) No;the directional derivative is - ~ ; ~ -1.
ANSWERS TO TEST FORCHAPTER3
1. (a) False; a mi nussign is missing.
(b) True.
(c) False; the fourt h-order derivati ves must be continuous to guaranteeequality.
(d) True.
(e) False; let !(x , y) = x
2
+y2 on the whole plane. Ithas a minimumbut no maximum.
2. 4. 28; 4.2665
3. Since 8
2
z/8x
2
= 30xy, concave up if x > 0;concave down if x <0.
4. (0,1,2,-2),(0, -1,2,-2) , (0, 1,- 2, - 2), (0,-1,-2,-2)
5. (a) and (c) since 8
2
f/ 8y8x = fj2 f/8x8y.
6. Minimum= - v'6/2; maximum= v'6/2.
7. (a) Minimum= ~ ; maximum=0.
(b) Minimum= ~ ; no maximum.
8. Theonly cri ticalpoint is at (- ~ , ~ . Also, D =(8
2
!/8x
2
)(8
2
!/8y2)- (8
2
f/8x8y)
2
=- 12 <
0, so (- ~ , ~ is a saddle point .
9. (a) 8x/8v=2
(b) Can'tbe done.
10. (a) (g , c) = (10,5)
(b) (10,5) is a maximumpoint .
ANSWERS TO CHAPTER TESTS ANDSAMPLE EXAMS 169
ANSWERS T O TEST FOR CHAPTER 4
1. (a) True.
(b) False; li mits ofintegration should be and 11'.
(c) Fa.lse; the last termshould be b x dc/dt.
(d) False; v(t) = (2,0,0) and w(t) = (0,1,0) both have 0 acceleration.
(e) True.
2. Both satisfy the desired conditions.
3. (a) All of 3
(b) Nowhere.
2
4. (a) ft
r
/ J4+5cos
2
tdt
(b) (-4,0)
5. (a) 6gl/(t)i+6[(g'(t))2 +g(t)gl/(t)Jj
(b) g(t) must be a constant.
6. z >-1
7. fa
4
J ~ +4y2 +4y2exp(2y2 - 12)dy
8. (a) and (b) since div F =0.
9. (a) (3:c
2
- 2:c +3)[( 2sin(x3- x
2
+3x- 4)cos(x
3
- x
2
+3x- 4))i+(4(x
3
- x
2
+3x - 4)3_
3(x
3
- x
2
+3x- 4))j]
(b) pl/(t)c(t) +2p' (t)c'(t) +p(t)cl/(t)
10. (a) They are bot h running counterclockwise on the ellipse 4x
2
+y2 = 4.
(b) 12500i
(c) l (t) =(cos( 4rr/5),2sin(4rr/5))+(t - rr/5)(-4sin(4rr/5) , 8cos(4rr/5))
2
(d) 4f211:11:/
5
4 VI+cos
2
4tdt
ANSWERS TO TEST FOR CHAPTER5
2 4 x ~
1. (a) False; shouJd be 1-2fa - dydx.
(b) True.
(c) True.
(d) False; I( x,y) should be nonnegative at all pointsofD.
(e) False;t he right-hand side should be
[1
2
(x
2
+4x+v'x)dX] [[11(y3 - y2 +4y)d
Y
] .
2.
Y 1
i11VYl
1
=
l
1
1
Z
lVY
- dzdxdy
-
dxdydz
a a a a a a
1
1
1
dxdzdy
=
l -
Y
iVY
a a a
1 z
=
1 1 i ~ l -
dydxdz
a a x'
4
=
i111-X'11-Y
dydzdx :
a a x'
15
170
APPENDIX
181
3
. 12
4. (1 - sin(1))/2 by changing the order ofintegration.
S. (a) 1811"
( ) f
3 f'; 9-,,2( 2 2
b -3 _';9-,, 2 9- x - y )dydx
(c) 2711"/2
6. i
7. (a) mv(W) :S fffw f( x,y,z) dxdydz:S MV(W).
(b) The minimumvalue offon Wis V2/2e
2
at (11" /4,1,1) and the maximumvalue is 1 at
(0,0, 0). The vol umeof Wis 11"/8. Apply these values to the statement in part (a).
8. (e
8
- 1)/3 by changing the order ofintegrationofx and z.
9.
3
10. (a)
(b) 8- if-
ANSWERS TO TEST FOR CHAPTER 6
1. (a) True.
(b) False; fooo cosxdxdoesn't converge,so Fubini'stheorem doesn't apply.
(c) False; the integral in polar coordinates is only being integrated over the region between
x =2yand x =J4- y2.
(d) True.
(e) True.
2. 11"2
3. 8/311"
4. 16011"/3
5.
6. (a) f01
(b) 2
7. 4811"
8. 211"
9. Theintegralisimpropernearz = 0, sotheintegralshouldbecalculatedas f01 dxdy+
f o1 (l/x)dxdy,which diverges.
10. (0,-10/(4+311"))
ANSWERS TO TEST FOR CHAPTER 1
1. (a) True.
(b) False; liTti x TIiII is positive andfis positive, so thescalarsurface integralis positive.
(c) True.
5\ ERS TO CHAPTERTESTS AND SAMPLE EXAMS
171
(d) True.
(e) False' Ie F(x , y) == xi +xyj and choose any closed curve in the plane x == o.
2. 4A
3. ~
4. 41T
5. (a) C)(O,r/ == (2cosOsinr/>,2sinOsinr/>,2cosr/, 0 S0 S 21T, 0 Sr/> S 1T.
(b) f fs Iz l dS == fa" f02" 12cosr/>I (4sinr/ dO dr/>
(c) 321T
6. (a) V/== (2zsiny,2xzcosy,e
Z
+2xsiny)
(b) e - 5
7. (a) -15x-12y+2z==3
(b) f02 f ~ l (144u
2
v
2
+72uv
3
+9v
4
+144u
4
v
2
+4U
2
)1/2dv du
8. 1 ~ 8
9. 7e+3 1
6
10. 8(3V3- 2V2)/ fJ
ANSWERS TO TEST FORCHAPTER 8
1. (a) True.
(b) True.
(c) True.
(d) True.
(e) False; the formula A == ~ f aD x dy - y dx must also be applied to the lines x == 0, x == 1
and y == o.
2. -6
3. (a) (4xz
2
siny cos(eX) - 2e
x
x
2
z2 siny sineeX))i +2x2z2 cosY cos(eX)j +4x
2
z siny cos(eX)k
(b) 0
4. 70
5. (a) xz
3
i - yz3j +4k
(b) 241T
6. ~
7. _ 15
8
8. (a) 5
(b) - 5
9. 9
10. (a) V x F == 0
(b) I (x y , z) == x
2
yz3 +y +x z
(e) e
172 APPENDIX
(d) 3
ANSWERS TO COMPREHENSIVE TE ST FOR CHAPTERS 1 ~ 4
1. (a) (1 , 0) is a minimum
(b) (1 , 0), (-1,0)
2. (a) (0, _e-
7r
/
2
, -1)
(b) (3e-
1
cos 1 - 2e-
i
sin 1)/(e-
2
sin
2
1 -
3. (a) c(t) =(cos t ,sint)
(b) c(t) = (exp(2t) , exp(-t))
4. (a) V2 + In(1 + V2)
(b) V2 + In(1 + V2)
8e-
2
sin 1cos 1 + 1 + 4e-
2
)1/2
(c) The two parametrizations actually trace out the same path.
5. 32V5/3 cubic inches
6. (a) (2 , 1,1)+t(0,:,2)
(b) l:3x + 7y+ 16z =2
0 OIl
7. (a) 0 9 -1
[
o 0 1
(b) [1 0]
- 3 9
3
8. 0.0021 m
9. (a) - ~ , - 2 )
(b) ~ hour
(c) i + 3j
(d) (1
2
5,10, 0)
(e) 5V5/ 2 km
t
2 u
10. (a) $t = 2te", $u = e , $v = sinw, $w = vcosw
(b) Go in the direction of (2, 1, 0, 2)
(c) (2 e
1
/
3
/3,4e
1
/
3
/9,0,0)
ANSWERS TO COMPREHENSIVE T EST FOR CHAP TERS 5 - 8
1. (a) 0
(0) (l-cosl)/2
2. ~
3. 1848
5
4. (a) J(x , y)=excos3y+C
(b) 0
(c) 0
ANSWERSTO CHAPTER TESTS AND SAMPLE EXAMS 173
5. (a) ar e. helps; theflux is 6V before and itis 9V after , where V = 41211"/3 is t he volume
0" e-ull
(b) 3,5
6. (a) '! - R- 1- -:-/2)
(b) ea 2- Rd.
7.
9. Ie)
10. In he urf ce should no beclosed (so it would have a real boundary), wher-
m. the surface has to be closed (toenclose a volume).
ANSWERS TO CO. HE. SIVE TEST A FORCHAP TERS 1 - 8
1. areaofsuch a cylinderis 211", while the valueofthe integral is 21211" .
pie.in::-3 let z = Ixl ; it 's not differentiable at x = o.
(e)
(d) a product , not a dot product .
(e) eed continuously differentiablesecond parti als.
2. (1) (c (3) (b) (4) (d)
3. (a)
(b) 0
(b)
Fl / 0
4.
F./8
- 2 3
0
1-
-
0
1
-1
5.
[
6. a z -
(b)
-
- ..IU)l
7. (a) - 1)
(b) ~
8.
9. a)
'4
(b
10. (a)
- T
-
- z'!- y2 - Jx
2
+y2dydx
f
16 r2 r4 ~
(b)
2. 2 Jo rdrdz+211" JoJo r dzdr
(e) 4 - /3
174 APPENDI'X
ANSWERS TO COMPREHENSIVE TEST B FOR CHAPTERS 1 - 8
1. (a) The two vectors are orthogonal, orone of themis O.
(b) The two vectors are parallel, oroneoft hem is O.
(c) There' s more material going in tha.n out; i.e., we have a sink.
(d) curl V .n tells you about the circulation ofVonthe surface with normal n.
2. (a) 8
(b) In t he direction of(1,1,2, 1) because the directional derivative using normalized vectors
is larger in the second case.
3. (a) The intersecting curves can be parametrized by
(x, y, z) =(2cosB, 2sinB, 1- 2cosB-2 sinB).
(b) (x , y,z ) =(1,V3,-v'3)+t(-V3,1,V3-1)
4. Maximum=6, minimum=3
5. (a) 96
(b) 192
(c) 96
f
l Iv'1'=? n/
1
-
x 2
-y'
6. (a
)
4 1/2 Vl/4-x2J O dz dydx
(b) f01/2 fo"l-r
2
rdzdrdB
(c) %(1 _
7. (a) 161!', whi ch is the surface areaofthe silo.
(b) (5001!'/3) dollars
8. (a) [sin(4) - sin(3)J/2
(b) 1!'
9. (a) 31!'/4
(b) -31!'/4
10. (a) f (x ,y,z) =(ze
cos
(xY),xy
2
sin z)
(b) tlfO-X'+l fox'sin
y
dzdy dx
(c) f01 dzd:r: dy

You might also like