Chapter 5

Double integrals

5.1 Introduction
Before we look at the multiple integrals, let us recall a few things about integrals for functions
of one variable. Suppose function y = f (x) is non-negative and continuous on the interval
[a, b]. The figure bounded by the graph of the function y = f (x), the x-axis and the vertical
lines x = a and x = b as shown in Fig 5.1 is called a curvilinear trapezoid. The area of the
curvilinear trapezoid is Z b
A= f (x)dx. (5.1)
a

Figure 5.1: A curvilinear trapezoid.

Example: Find the area of the figure bounded by the ellipse

x2 y 2
+ 2 = 1.
a2 b

Solution. The area of the ellipse is 4 times larger than the shaded area shown in Fig.r5.2. The
x2
shaded area can be calculated using the formula (5.1) by taking a = 0, b = a, f (x) = b 1 .
a2
Therefore,
Z ar
x2
A = 4b 1 dx.
0 a2

65
Calculus of Several Variables (G5210) 2018-19 Lecture notes

x2 y2
Figure 5.2: An ellipse a2
+ b2
= 1.

In order to calculate the integral, we can use the substitution x = a sin t, 0  t  ⇡/2, which
gives
Z ar
x2
A = 4b 1 dx.
0 a2
The integral can be calculated using the substitution
x
x = a sin t ) t = arcsin , ) dx = a cos tdt.
a
Thus,
Z r Z r
a ⇡/2
x2 a2 sin2 t
4b 1 dx = 4b 1 a cos tdt
0 a2 0 a2

Z ⇡/2
= 4ba cos2 tdt use {cos2 t = 1/2 + (cos 2t)/2}
0

Z ⇡/2  
1 cos 2t 1 ⇡/2 sin 2t ⇡/2
= 4ab + dt = 4ac t +
0 2 2 2 0 4 0

h⇡ i
= 4ab + 0 = ab⇡.
4
⌅

Consider the figure shown in Fig. 5.3. It is bounded by the segments OA and OB, and a
continuous curve given (in polar coordinates) by the formula

r = r('), '1  '2.

This curve is called a curvilinear section. Its area can be found as

Z
1 '2
A= rr(')d'.
2 '1

Dr Yuliya Kyrychko 66 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

Figure 5.3: A curvilinear section.

5.2 Double integrals

Suppose G is a bounded subset of points in Rn , in particular, on a plane, and let M1 and M2
be two arbitrary points from G. A number

d = sup{d(M1 , M2 ) : M1 , M2 2 G}

is called the diameter of the subset G. For example, the diameter of a rectangle is equal its
diagonal; the diameter of an ellipse is equal its major axis.

Suppose G is a subset (open or closed) in the plane (x, y), and let a bounded function z =
S
n
f (x, y) = f (K) (K = (x, y)) be defined on G. Let us divide G into n parts, i.e. G = Gi
i=1
(shown in Fig. 5.4) in such as way that any two subparts Gi and Gj do not have any common
inner points. In each subpart Gi we can arbitrarily choose a point Ki (⇠i , ⌘i ) and write down

Figure 5.4: Division of the subset G into n parts.

the sum n
X
I(Gi , Ki ) = f (Ki )A(Gi ),
i=1

where A(Gi ) is the area of Gi .

The number I(Gi , Ki ) is called an integral sum of the function f (x, y), which corresponds to
this division of G and this particular choice of the points Ki . Let di be the diameter of Gi , so
that d = max1in di .

Dr Yuliya Kyrychko 67 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

Definition 22 If there exists a limit limd!0 I(Gi , Ki ) = I, then the number I is called a
double integral of the function f (x, y) in the region G, and is denoted as
ZZ
I= f (x, y)dxdy.
G

If a function f (x, y), (x, y) 2 G is a continuous non-negative function, then

ZZ
f (x, y)
G

is the volume of the solid shown in Fig. 5.5.

RR
Figure 5.5: A solid with the volume G
f (x, y).

5.2.1 Properties of double integrals

(1) If f (x, y) = Af1 (x, y) + Bf2 (x, y), then
ZZ ZZ ZZ
f (x, y)dxdy = A f1 (x, y)dxdy + B f2 (x, y)dxdy.
G G G

S
(2) Let region G be subdivided into two regions G1 and G2 , i.e. G = G1 G2 , then
ZZ ZZ ZZ
f (x, y)dxdy = f (x, y)dxdy + f (x, y)dxdy.
G G1 G2

(3) If f (x, y)  g(x, y), then the following inequality holds

ZZ ZZ
f (x, y)dxdy  g(x, y)dxdy.
G G

Dr Yuliya Kyrychko 68 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

5.3 Repeated integration

There are several ways for evaluating double integrals. In this section we consider the so-called
repeated di↵erentiation.

Case 1. Suppose function f (x, y) is defined in the rectangle Q = {(x, y) : a  x  b, c  y 

d}. Then, the following theorem holds.

Theorem 21 Suppose

(i) a double integral ZZ

f (x, y)dxdy
Q

exists;

(ii) 8x 2 [a, b], the definite integral

Z d
I(x) = f (x, y)dy
c

exists.
Z b
Then, there exists a definite integral I(x)dx (called repeated), which is written as
a
Z b Z d
dx f (x, y)dy,
a c

and the following formula holds

ZZ Z b Z d
f (x, y)dxdy = dx f (x, y)dy.
Q a c

The last expression means that the double integral is equal to a repeated integral in this case.

Corollary. If we interchange x and y in Theorem 21, the following formula holds

ZZ Z d Z b
f (x, y) = dy f (x, y)dx.
Q c a

Example: Let Q = {(x, y) : 0  x  1, 0  y  1}. Evaluate

ZZ
xexy dxdy.
Q

Dr Yuliya Kyrychko 69 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

Solution. We have
ZZ Z 1 Z 1 Z 1 y=1
xy xy
xe dxdy = dx xe dy = dx · exy
Q 0 0 0 y=0
Z 1 1
= (ex 1)dx = (ex x)
0 0

=e 1 1=e 2.
⌅
Case 2. Suppose function f (x, y) is defined in the region G = {(x, y) : a  x  b, y1 (x)  y 
y2 (x)}, where y1 (x) and y2 (x) are continuous functions as shown in Fig. 5.6.

Figure 5.6: A region G = {(x, y) : a  x  b, y1 (x)  y  y2 (x)}.

Then, the following theorem holds.

Theorem 22 Suppose
(i) a double integral ZZ
f (x, y)dxdy
Q
exists;

(ii) 8x 2 [a, b], the definite integral

Z d
I(x) = f (x, y)dy
c

exists.
Then, there exists a repeated integral
Z b Z b Z y2 (x)
I(x)dx = dx f (x, y)dy,
a a y1 (x)

and it is equal to the double integral

ZZ Z b Z y2 (x)
f (x, y)dxdy = dx f (x, y)dy.
G a y1 (x)

Dr Yuliya Kyrychko 70 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

Example: Let a region G be bounded by a line y = x and a parabola y = x2 . Evaluate

ZZ
xy 2 dxdy.
G

Solution. The region G is sketched in Fig. 5.7.

Figure 5.7: A region G be bounded by a line y = x and a parabola y = x2 .

We have
ZZ Z 1 Z xZ
2 1 1 2
xy dxdy = xy dy = x(x3 x6 )dx
G 0 x2 3 0
✓ ◆ 1
1 x5 x8 1
= = .
3 5 8 40
0

5.3.1 Area and volume: useful facts

Suppose that f (x, y) is continuous on a region G in the plane z = 0. Then

(i) The area A of the region G is ZZ

A= dxdy.
G

(ii) The volume V of the solid that lies below the surface z = f (x, y) and above the region G
is (assuming that this integral exists)
ZZ
V = f (x, y)dxdy.
G

Dr Yuliya Kyrychko 71 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

(iii) The volume of the solid that lies below the surface z = ztop = z2 (x, y) and above the
surface z = zbottom = z1 (x, y) is (assuming that this integral exists)
ZZ ZZ
V = [ztop zbottom ]dxdy = [z2 (x, y) z1 (x, y)]dxdy.
G G

Example: Find the volume of the solid region under the graph f (x, y) = 3xy and above the
triangle G bounded by x- and y-axes and a line y = 2 x.

Solution. The volume is can be found as the double integral

ZZ ZZ
3xydxdy ⌘ 3xydydx.
G G

The region is sketched in Fig. 5.8. Let us first integrate in x, i.e.

Figure 5.8: A region G bounded by x- and y-axes and a line y = 2 x.

ZZ Z 2 Z 2 y Z 2 
3 2 2 y
3xydxdy = 3xydxdy = xy dy
G 0 0 0 2 0
Z 2 Z 2
3 3
= (2 y)2 ydy = (4y 4y 2 + y 3 )dy
0 2 2 0
 
3 4y 2 4y 3 y 4 2 3 32
= + = 8 +4
2 2 3 4 0 2 3
= 34.

We will obtain the same answer if we integrate in y first by evaluating

ZZ Z 2Z 2 x Z 2
3
3xydxdy = 3xydydx = (2 x)2 xdx = 34.
G 0 0 0 2
⌅
Example: For an unknown function f , find
ZZ
f (x, y)dxdy,
G

Dr Yuliya Kyrychko 72 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

where G is the region bounded by the left semi-circle, a line y = x 1 and a line y = 1.

Solution. The region is shown in Fig. 5.9. We have two choices: integrate with respect to x
first, and then in y, or integrating with respect to y first, and then in x. Since the left and

Figure 5.9: A region G bounded by the left semi-circle, a line y = x 1 and a line y = 1.

right boundaries are represented by single equations, i.e. semi-circle and the line y = x 1,
integrating with respect to x requires one repeated integral. Looking at the Fig. 5.9, we can
see that y goes from 1 pto 1. Moreover, since the equation for the circle is x2 + y 2 = 1, the left
boundary in x is x = 1 y 2 , and the right boundary in x is the line y + 1. Therefore, we
have ZZ Z 1 Z y+1
f (x, y)dxdy = p f (x, y)dxdy.
G 1 1 y2

Now, itpcan be seen from Fig. 5.9 that the bottom boundary consists of two parts (a semicircle
y = 1 x2pand a line y = x 1), and the bottom boundary consists of two parts (a
semicircle y = 1 x2 and a line y = 1). Thus, in order to integrate in y first, we need to split
the region into parts, say, the part to the left of y-axis and the part to the right of the y-axis,
i.e. ZZ ZZ ZZ
f (x, y)dydx = f (x, y)dydx + f (x, y)dydx.
G left part of G right part of G
p p
Over the left part, x is between 1 and 0, and y is between 1 x2 and 1 x2 . Over the
right part of G, x is between 0 and 1, and y is between x 1 and 1, which gives
ZZ Z 0 Z p 1 x2 Z 1Z 1
f (x, y)dydx = p
f (x, y)dydx + f (x, y)dydx.
G 1 1 x2 0 x 1

5.4 Coordinate transformation in double integrals

ZZ
Suppose we need to evaluate a double integral f (x, y)dxdy, and we want to change from
G
variables (x, y) into variables (u, v), where
x = '(u, v), y = (u, v), (u, v) 2 g. (5.2)

Dr Yuliya Kyrychko 73 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

Using the change of variables above, the double integral can be evaluated as
ZZ ZZ
D(x, y)
f (x, y)dxdy = f ('(u, v), (u, v)) dudv, (5.3)
G g D(u, v)

where
D(x, y) 'u 'v
=
D(u, v) u v

is the Jacobian of the functions ' and in variables u and v. Geometrically, modulus of the
Jacobian represents the area stretching coefficient using the transformation (5.5) as illustrated
in Fig. 5.10.

Figure 5.10: A graphic representation of the Jacobian.

Example: Evaluate ZZ
(x2 y 2 )dxdy,
G

where G is a diamond-shaped region with vertices (1, 0), (0, 1), ( 1, 0) and (0, 1).

Solution. The region G is plotted in Fig. 5.11, where the boundaries are formed of four lines:
y = x 1, y = 1 x, y = x 1 and y = 1 x. If we were to compute the integral directly, we
would need to split the region into parts. However, we can use the coordinate transformation
1 1
u = x + y, v=x y, =) x = (u + v), y = (u v),
2 2
and the region G now becomes a square g = {(u, v) : u = ±1, v = ±1}. In the new coordinates,
we have
x2 y 2 = (x y)(x + y) = uv.
We need to calculate the Jacobian
D(x, y) xu xv 1/2 1/2 1 1
= = = = .
D(u, v) yu yv 1/2 1/2 2 2

Dr Yuliya Kyrychko 74 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

Figure 5.11: A region G is a diamond-shaped region with vertices (1, 0), (0, 1), ( 1, 0) and
(0, 1).

Finally, we have
ZZ ZZ
2 2 D(x, y)
(x y )dxdy = vu dudv
G g D(u, v)
Z Z Z 1
1 1 1 1 2 1
= vu = vu dv
2 1 1 1 4 1

= 0.

5.4.1 Polar coordinates

The formulas for transforming cartesian coordinates (x, y) into polar coordinates (r, ') are

x = r cos ', y = r sin ', r 0, 0  '  2⇡. (5.4)

Consider a point M0 drawn in Fig. 5.12. On one hand, a pair (r0 , '0 ) give polar coordinates
of the point M0 . On the other hand, formulas (5.4) allow to transform a shaded strip in r '
plane (drawn in Fig. 5.13) into the x y plane.
The Jacobian in polar coordinates is

D(x, y) cos ' r sin '

= = r.
D(r, ') sin ' r cos '

Example: Evaluate a double integral

ZZ
I= (x2 + 2y 2 )dxdy,
G

where G = {(x, y) : a2  x2 + y 2  b2 } is a ring plotted in Fig. 5.14.

Dr Yuliya Kyrychko 75 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

Figure 5.12: Cartesian coordinates of the point M0 .

Figure 5.13: A polar coordinates transformation.

Figure 5.14: A ring G = {(x, y) : a2  x2 + y 2  b2 }.

Solution. In polar coordinates, we have x = r cos ', y = r sin ', (r, ' 2 g = {(r, ') : a  r 
b, 0  '  2⇡}) - a rectangle shown in Fig. 5.15. Using the polar coordinates, the integral can

Dr Yuliya Kyrychko 76 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

Figure 5.15: A rectangle x = r cos ', y = r sin ', (r, ' 2 g = {(r, ') : a  r  b, 0  '  2⇡}).

now be evaluated as
ZZ
I= (r2 cos2 ' + 2r2 sin2 ')rdrd'
g
Z b Z 2⇡ ✓ ◆
1 + cos 2' 1 cos 2'
= r3 dr +2 d'
a 0 2 2
Z b Z 2⇡ 
3 cos 2' r4 b 1 1 2⇡
= r3 dr d' = · 3' sin 2'
a 0 2 4 a 2 2 0
3⇡ 2
= [b a2 ].
4
⌅
Example: Determine the value of
Z 1
x2
I= e dx.
1
2
The function f (x) = e x is called a Gaussian function, and knowing the area under the graph
is important in applications of probability and statistics. It is plotted in Fig. 5.16.

Figure 5.16: A Gaussian function.

Solution. In this example, we can use polar coordinates to calculate I 2 , i.e.

✓Z 1 ◆ ✓Z 1 ◆ Z 1Z 1
2 x2 y2 2 2
I = e dx · e dy = e (x +y ) dxdy,
1 1 1 1

Dr Yuliya Kyrychko 77 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

where we used the fact that since the first integral is a constant with respect to y, we can bring
it inside the second integral to get the last expression. Now we can calculate I 2 using the polar
coordinates
Z 1Z 1 Z 2⇡ Z 1
(x2 +y 2 ) 2
e dxdy = re r drd'
1 1 0 0
Z 2⇡ Z 2⇡
1 2 1 1
= e r d' = d'
2 0 0 0 2
= ⇡.
p
Since I 2 = ⇡, we find I = ⇡, i.e.
Z 1
2 p
e x dx = ⇡.
1

5.5 Triple integrals

Triple integrals are defined in a similar way as double integrals. Let function u = f (x, y, z) =
f (M ) be a function defined in the region T 2 R3 . Now, we split the region T into n parts, in
Sn
such a way that any two parts do not contain any common inner points, T Ti . In each part
i=1
Ti we choose a point Mi (⇠i , ⌘i , ⇣i ) and construct an integral sum
n
X
I(Ti , Mi ) = f (⇠i , ⌘i , ⇣i ) · V (Ti ),
i=1

where V (Ti ) is the volume of T1 . Let di denote the diameter of Ti , d = max1in di .

The limit of the integral sum when d ! 0 is called a triple integral of the function f (x, y, z)
over the region T and is written as
ZZZ Z
f (x, y, z)dxdydz or f (M )dV.
T T
RRR
If f (x, y, z) = 1, then T
dxdydz = VRRR(T ) is the volume of the region T . If ⇢(x, y, z) is the
density of T at the point (x, y, z), then T
⇢(x, y, z)dxdydz = m is the mass of T .

5.5.1 Repeated integration of triple integrals

Consider a region T given by

T = {(x, y, z) : (x, y) 2 G, z1 (x, y)  z  z2 (x, y)},

where G is the region in the x y plane, z1 (x, y) and z2 (x, y) are continuous functions in G as
shown in Fig. 5.17.
Let a bounded function u = f (x, y, z) is defined in the region T . Then, the following Theorem
holds.

Dr Yuliya Kyrychko 78 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

Figure 5.17: A region G the x y plane.

Theorem 23 Assume the following conditions are satisfied

RRR
(1) there exists a triple integral T
f (x, y, z)dxdydz;

(2) 8(x, y) 2 G, there exists a definite integral

Z z2 (x,y)
I(x, y) = f (x, y, z)dz.
z1 (x,y)

RR
Then, there exists a repeated integral G
I(x, y)dxdy, and it is equal to the triple integral, i.e.
ZZZ ZZ Z z2 (x,y)
f (x, y, z)dxdydz = dxdy f (x, y, z)dz.
T G z1 (x,y)

RR
Corollary. If the conditions of Theorem 21 for a double integral G I(x, y)dxdy are satisfied,
then it can be written in the form of the repeated integrals, i.e.
ZZ Z b Z y2 (x) Z b Z y2 (x) Z z2 (x,y)
I(x, y)dxdy = I(x, y)dy = dx dy f (x, y, z)dz.
G a y1 (x) a y1 (x) z1 (x,y)

Example: Let region T be a tetrahedronRRR bounded by the planes x = 0, y = 1, z = 0, and

x + 2y + 3z = 8. Write the triple integral T
f (x, y, z)dV as a repeated integral.

Solution. The tetrahedron is shown in Fig. 5.18. There are 6 possible orders to find the solu-
tion. Let us do this by writing the outer integral first, i.e. we can ”slice” the surface parallel
to the xy plane, parallel to the xz-plane, and parallel to the xy-plane.

(1) Slice parallel to the yz-plane.

Dr Yuliya Kyrychko 79 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

Figure 5.18: A tetrahedron bounded by the planes x = 0, y = 1, z = 0, and x + 2y + 3z = 8.

In
R 6 this case, we are slicing the interval [0, 6] on the x-axis, so the outer integral will be
0
unknown dx. In order to write the inner integrals, we think of x as being constant
(as within as slice x is constant). Each slice is a triangle, with one edge on the plane y = 1,
one edge on the plane z = 0, and one edge on the plane x + 2y + 3z = 8. The slice is shown
in Fig. 5.19, and we can write a double repeated integral, which describes this figure. There
are two ways to do this.

Figure 5.19: A slice parallel to the yz-plane


8 x
(i) If we go vertically, we are slicing the interval 1, on the y-axis, so the outer in-
2
Z 8 x
2
tegral will be unknown dy. Along the z-axis, the slice goes from 0 to the line
1
Z 8 x 2y
8 x 2y 3
2y +3z = 8 x, or z = , and the inner integral will be f (x, y, z)dz.
3 1

Dr Yuliya Kyrychko 80 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

Therefore, the triple integral can be written as

ZZZ Z 6Z 8 x Z 8 x 2y
2 3
f (x, y, z)dV = f (x, y, z)dzdydx.
T 0 1 1

(ii) If we go horizontally, on the z-axis we go from 0 to 6 3 x , so the outer integral will be

Z 6 x
3
unknown dz. On the y-axis, we go from y = 1 to the line 2y + 3z = 8 x, or
1
Z 8 x 3z
8 x 3z 2
z= , so the inner integral will be f (x, y, z)dz. Finally, the triple
2 1
integral can be written as
ZZZ Z 6 Z 6 x Z 8 x 3z
3 2
f (x, y, z)dV = f (x, y, z)dydzdx.
T 0 1 1

(2) Slice parallel to xz-plane.

In
R 4 this case, we are slicing the interval [1, 4] on the y-axis, so the outer integral will be
1
unknown dy. In order to write the inner integrals, we think of y as being constant
(as within as slice y is constant). Each slice is a triangle, with one edge on the plane x = 0,
one edge on the plane z = 0, and one edge on the plane x+2y +3z = 8. The slice is shown in
Fig. 5.20, and we can write a double repeated integral, which describes this figure. Again,
there are two ways to do this.

Figure 5.20: A slice parallel to xz-plane.

(i) If we go vertically,
Z we are slicing the interval [1, 8 2y] on the x-axis, so the outer inte-
8 2y
gral will be unknown dx. Along the z-axis, the slice goes from 0 to the line
0
Z 8 x 2y
8 x 2y 3
2y +3z = 8 x, or z = , and the inner integral will be f (x, y, z)dz.
3 0
Therefore, the triple integral can be written as
ZZZ Z 4 Z 8 2y Z 8 x 2y
3
f (x, y, z)dV = f (x, y, z)dzdxdy.
T 1 0 0

Dr Yuliya Kyrychko 81 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

(ii) If we go horizontally, on the z-axis we go from 0 to 8 32y , so the outer integral will be
Z 8 2y
3
unknown dz. On the x-axis, we go from x = 0 to the line x = 8 2y 3z,
0
so the triple integral can be written as
ZZZ Z 4Z 8 2y Z 8 2y 3z
3 2
f (x, y, z)dV = f (x, y, z)dxdzdy.
T 1 0 1

(3) Slice parallel to xy-plane.

In
R 2 this case, we are slicing the interval [0, 2] on the z-axis, so the outer integral will be
0
unknown dy. In order to write the inner integrals, we think of z as being constant
(as within as slice z is constant). Each slice is a triangle, with one edge on the plane x = 0,
one edge on the plane y = 1, and one edge on the plane x+2y +3z = 8. The slice is shown in
Fig. 5.21, and we can write a double repeated integral, which describes this figure. Again,
there are two ways to do this.

Figure 5.21: A slice parallel to xy-plane.

(i) If we go vertically,
Z 6 we are slicing the interval [0, 6 3z] on the x-axis, so the outer
3z
integral will be unknown dx. Along the y-axis, the slice goes from 1 to the
0
line x + 2y = 8 3z, and the triple integral can be written as
ZZZ Z 2 Z 6 3z Z 8 3z x
2
f (x, y, z)dV = f (x, y, z)dydxdz.
T 0 0 0

(ii) If we go horizontally, on the y-axis we go from 1 to 8 23z , so the outer integral will be
Z 8 3z
2
unknown dy. On the x-axis, we go from x = 0 to the line x = 8 2y 3z,
1
so the triple integral can be written as
ZZZ Z 2Z 8 3z Z 8 2y 3z
2 2
f (x, y, z)dV = f (x, y, z)dxdydz.
T 0 1 1

Dr Yuliya Kyrychko 82 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

5.6 Coordinate transformation in triple integrals

ZZZ
Suppose we need to evaluate a triple integral f (x, y, z)dxdydz, and we want to change
T
from variables (x, y, z) into variables (u, v, w), where
x = '(u, v, w), y = (u, v, w), z = ⇠(u, v, w), (u, v, w) 2 g. (5.5)
We assume that functions ', , ⇠ have continuous first order partial derivatives in g and
D(x, y, z)
6= 0, 8(u, v, w) 2 g.
D(u, v, w)
Let g and T be bounded regions, and let the function f (x, y, z) be bounded and continuous in
the region T . Then, using the change of variables above, the triple integral can be evaluated as
ZZZ ZZZ
D(x, y, z)
f (x, y, z)dxdydz = f ('(u, v, w), (u, v, w), ⇠(u, v, w)) dudvdw, (5.6)
T g D(u, v, w)
where
'u 'v 'w
D(x, y, z)
= u v w
D(u, v, w)
⇠u ⇠v ⇠w
is the Jacobian of the functions ', and ⇠ in variables u, v, w. Geometrically, modulus of the
Jacobian represents the volume stretching coefficient.

5.6.1 Cylindrical coordinates

A tuple (r, ', z) is called cylindrical coordinates of the point M in Fig. 5.22.

Figure 5.22: A cylindrical coordinates transformation.

In cylindrical coordinates we have

x = r cos ', y = r sin ', z = z, (r 0, 0  '  2⇡, 1 < z < 1).

Dr Yuliya Kyrychko 83 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

The Jacobian factor is

cos ' r sin ' 0
D(x, y, z)
= sin ' r cos ' 0 =r
D(r, ', z)
0 0 1
and dV = rdrd'dz.

Example: Evaluate
p the integral of f (x, y, z) = xz over the cone-shaped region T bounded by
the cone z = x2 + y 2 and the sphere x2 + y 2 + z 2 = 2z.

Solution. First, we re-write the equation of the sphere as x2 + y 2 + (z 1)2 = 1. The

intersection of the surfaces is plotted in Fig. 5.23. We will evaluate this integral using the

p
Figure 5.23: An intersection of a cone-shaped region T bounded by the cone z = x2 + y 2 and
the sphere x2 + y 2 + z 2 = 2z.

cylindrical coordinates in the order dzdrd'. The projection of this region onto the xy-plane
is the disk with boundary circle x2 + y 2 = 1, which is the circle lying directly below the
intersection of the cone and the sphere. This intersection is at z = 1, and it is obtained by
using the cone and sphere equations to solve for z. Thus, in order to describe this projected
disk, we note that ' makes one full revolution, and r goes from 0 to 1. The pgiven region z
starts atpthe bottom cone, and moves to the top sphere, so z goes from z = x2 + y 2 = r to
p
z = 1 + 1 x2 y 2 = 1 + 1 r2 . Therefore, we have
ZZZ Z Z Z p
2⇡ 1 1+ 1 r 2
xzdV = zr2 cos 'dzdrd'.
T 0 0 r

Dr Yuliya Kyrychko 84 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

Evaluating this integral directly might be tricky, but noticing that there are no ' in the inner
and middle integrals means that the result of the inner and middle calculations will be some
constant C, so the integral becomes
ZZZ Z Z Z p Z
2⇡ 1 1+ 1 r 2 2⇡ 2⇡
2
xzdV = zr cos 'dzdrd' = C cos 'd' = C sin ' = 0.
T 0 0 r 0 0

Is this surprising? Not, really! Since f (x, y, z) = xz is odd with respect to x, and the region T
is symmetric in the yz-plane, the integral should indeed be zero. ⌅

5.6.2 Spherical coordinates

A tuple (r, ✓, ') is called spherical coordinates of the point M shown in Fig. 5.24. The coordinate
plane r = constant is a sphere.

Figure 5.24: A spherical coordinates transformation.

In spherical coordinates we have

x = r sin ✓ cos ', y = r sin ✓ sin ', z = r cos ✓, r 0, 0  ✓  ⇡, 0  '  2⇡.

The Jacobian factor is

cos ' sin ✓ r cos ' cos ✓ r sin ' sin ✓
D(x, y, z)
= sin ' sin ✓ r sin ' cos ✓ r cos ' sin ✓ = r2 sin ✓,
D(r, ✓, ')
cos ✓ r sin ✓ 0

and dV = r2 sin ✓drd✓d'.

Example: Calculate volume of the region T bounded by the surface

✓ 2 ◆2
x y2 z2 x2 y 2
+ + = + . (5.7)
4 9 4 4 9

Dr Yuliya Kyrychko 85 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

✓ ◆2
x2 y 2 z 2 x2 y 2
Figure 5.25: A surface + + = + .
4 9 4 4 9

Solution. The region is plotted in Fig. 5.25.

Intersection between this surface and z = 0 is an ellipse
x2 y 2
+ = 1,
4 9
and the point (0, 0, 0). Intersection between this surface and x = 0 is a curve
✓ 2 ◆
y z2 y2
+ = .
9 4 9
In order to calculate the volume, we have to evaluate the following triple integral;
ZZZ
V = dxdydz.
T

In this case, it easier if we use generalised spherical coordinates, i.e.

x = 2r sin ✓ cos ', y = 3r sin ✓ sin ', z = 2r cos ✓, 0  ✓  ⇡, 0  '  2⇡.

In new coordinates, the equation of the surface is

✓ 2 2 ◆2
4r sin ✓ cos2 ' 9r2 sin2 ✓ sin2 ' 4r2 cos2 ✓ 4r2 sin2 ✓ cos2 ' 9r2 sin2 ✓ sin2 '
+ + = + ,
4 9 4 4 9

Dr Yuliya Kyrychko 86 y.kyrychko@sussex.ac.uk

Calculus of Several Variables (G5210) 2018-19 Lecture notes

or
r4 = r2 sin2 ✓. =) r = 0, r = sin ✓.
The Jacobian factor for the generalised spherical coordinates is

2 cos ' sin ✓ 2r cos ' cos ✓ 2r sin ' sin ✓

D(x, y, z)
= 3 sin ' sin ✓ 3r sin ' cos ✓ 3r cos ' sin ✓ = 12r2 sin ✓.
D(r, ✓, ')
2 cos ✓ 2r sin ✓ 0

Finally.
ZZZ Z Z ⇡ Z sin ✓
D(x, y, z)
V = drd✓d' 2⇡d' d✓ 12r2 sin ✓dr
g D(r, ✓, ') 0 0 0

Z ⇡ Z ⇡
1 sin ✓ 24
= 24⇡ sin ✓ · r3 = ⇡ sin4 ✓d✓
0 3 0 3 0

In order to calculate the last integral, we use trigonometric formula sin2 x = (1 cos 2x)/2 and
cos2 x = (1 + cos 2x)/2 to obtain
✓ ◆2
1
cos 2✓ 1 cos 2✓ cos2 2✓
sin4 ✓ = (sin2 ✓)2 = = +
2 4 2 4
1 cos 2✓ 1 1 3 3
= + + cos 2✓ = cos 2✓.
4 2 8 8 8 8
The integral becomes
Z ⇡ Z ⇡✓ ◆
24 4 24 3 3
V = ⇡ sin ✓d✓ = ⇡ cos 2✓ d✓
3 3 8 8
✓0 ◆ 0
24 3 3 ⇡
= ⇡ ✓ sin 2✓ = 3⇡ 2 .
3 8 16 0

Dr Yuliya Kyrychko 87 y.kyrychko@sussex.ac.uk

Chapter 6

Line integrals

6.1 Length of a curve

Suppose we have a plane in x y coordinates. Consider a set of points {M (x, y)}, which have
coordinates given by the following equations

x = '(t), y = (t), ↵t , (6.1)

where '(t) and (t) are continuous functions on the interval [↵, ].

If some point M (x, y) form this set corresponds to several values of t 2 [↵, ], then this point
is called multiple.

Suppose di↵erent values of t 2 [↵, ] correspond di↵erent values of the point M (x, y), so that
the set M (x, y) does not have multiple points. Then the set {M ('(t), (t))}, where ↵  t 
is called simple plane non-closed/open curve. The variable t is called the parameter, and the
equations (6.1) defined the curve parametrically.

The points A('(a), (a)) and B('(b), (b)) are called end or boundary points of the curve. If
the points A and B coincide, and all other points are not multiple points, then the curve is
called simple closed curve.

Example: Let x = cos t, y = sin t.

(a) If 0  t  ⇡, then the curve is a simple open curve (a half-circle).

(b) If 0  t  2⇡, the the curve is a simple closed curve (a circe).

(c) If 0  t  4⇡, then the curve is not simple, since all its points repeat twice, and the point
(1, 0) repeats three times.

Consider an open/closed curve, which is given by the equations (6.1), and divide an interval
[↵, ] into n parts so that
↵ = t0 < t1 < ... < tn = .

88