Anda di halaman 1dari 120

Table of contents

Differential calculus for functions of one real variable

1.1

The derivative of a function. Basic differential formulas .

1.2

Higher order derivatives . . . . . . . . . . . . . . . . . .

1.3

Taylors formula for functions of one variable . . . . . . .

1.4

Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . .

Differential calculus for functions of several real variables 13


2.1

Partial derivatives . . . . . . . . . . . . . . . . . . . . . .

13

2.2

Higher order partial derivatives . . . . . . . . . . . . . .

14

2.3

Derivatives of composite functions . . . . . . . . . . . . .

15

2.4

Differential operators . . . . . . . . . . . . . . . . . . . .

17

2.5

Differential of a function . . . . . . . . . . . . . . . . . .

19

2.6

Taylors formula for functions of several variables . . . .

20

2.7

Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . .

21

3 Extrema of functions

33

3.1

Local extrema of a function of one real variable . . . . .

33

3.2

Local extrema of a function of several real variables . . .

34

3.3

Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . .

36

4 Implicit functions

41

4.1

Implicit functions of one real variable . . . . . . . . . . .

41

4.2

Implicit functions of two real variables . . . . . . . . . .

42

4.3

Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . .

43

2
5 Change of variables
5.1 Change of the independent variable for functions of one
real variable . . . . . . . . . . . . . . . . . . . . . . . . .
5.2 Change of the independent variables for functions of several real variables . . . . . . . . . . . . . . . . . . . . .
5.3 Interchange of variables . . . . . . . . . . . . . . . . . . .
5.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . .

47

6 Integral calculus for functions of one real variable


6.1 Antiderivatives . . . . . . . . . . . . . . . . . . . . . . .
6.2 Riemann integral on a real interval . . . . . . . . . . . .
6.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . .

59
59
61
64

7 Line integrals
7.1 Line integral with respect to arc length . . . . . . . . . .
7.2 Applications of the line integral with respect to arc length
7.3 Line integral with respect to the coordinates . . . . . . .
7.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . .

67
67
68
70
72

8 Double integral
8.1 Calculus by iteration . . . . . . . . . . . . .
8.2 Changes of variables for the double integral
8.3 Applications of the double integral . . . . .
8.4 Exercises . . . . . . . . . . . . . . . . . . . .

.
.
.
.

79
79
82
83
84

.
.
.
.

95
95
97
99
99

9 Triple integral
9.1 Calculus by iteration . . . . .
9.2 Changes of variables . . . . .
9.3 Applications of triple integral
9.4 Exercises . . . . . . . . . . . .

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

47
48
49
50

10 Surface integrals
107
10.1 Surface integrals of scalar fields . . . . . . . . . . . . . 107
10.2 Applications of the surface integral of a scalar field . . . 109

3
10.3 Surface integrals of vector fields . . . . . . . . . . . . . .
10.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . .
Bibliografie

109
111
120

CHAPTER 1

Differential calculus for functions of


one real variable
1.1

The derivative of a function. Basic differential


formulas

Let A be a subset of R, f : A R a function and x0 A a limit point


of the set A. If the limit
lim

xx0

f (x) f (x0 )
x x0

exists, it is called the derivative of the function f at the point x0


df
and is denoted by f 0 (x0 ) or
(x0 ).
dx
If, for each x0 A, the function f has a finite derivative, then it is called
differentiable on A. The function x 7 f 0 (x) is denoted f 0 and is called
the derivative of f .
The process of finding the derivative is called differentiation.
We present below some of the basic differentiation rules and formulas:
1. (f + g)0 = f 0 + g 0 , (f g)0 = f 0 g 0 ,
5

6
2. (kf )0 = kf 0 , for k R,
3. (f g)0 = f 0 g + f g 0 ,
n
X
0
4. (f1 f2 . . . fn ) =
f1 f2 . . . fk1 fk0 fk+1 . . . fn ,
k=1
0
f 0g f g0
f
5.
=
,
g
g2
6. (f g)0 = f 0 g g 0 , or in another writing, (g(f ))0 = g 0 (f )f 0 ,
1
1
7. (f 1 )0 = 0 1 = 0
,
f (f )
f f 1
8. (f g )0 = gf g1 f 0 + f g g 0 ln f ,
9. c0 = 0, where c R is a constant,
10. x0 = 1, for x R,
11. (xn )0 = nxn1 , for x R , n Z,
12. (x )0 = x1 , for x (0, ), R,

1
13. ( x)0 = , for x (0, ),
0 2 x
1
1
14.
= 2 , for x R \ {0},
x
x
1
0
15. (ln x) = , for x (0, ),
x
16. (ex )0 = ex , for x R,
17. (ax )0 = ax ln a, for x R, a (0, ), a 6= 1,
18. (sin x)0 = cos x, for x R,
19. (cos x)0 = sin x, for x R,
1
20. (tan x)0 =
= 1 + tan2 x, for cos x 6= 0,
cos2 x
1
21. (cot x)0 = 2 , for sin x 6= 0,
sin x
1
0
22. (arcsin x) =
, for x (1, 1),
1 x2
1
23. (arccos x)0 =
, for x (1, 1),
1 x2
1
24. (arctan x)0 =
, for x R,
1 + x2
1
25. (arccotx)0 =
, for x R.
1 + x2

7
1.2

Higher order derivatives


Let f : A R be a differentiable function and f 0 : A R its derivative.
The function f is called twice differentiable at a point x0 A if f 0 is
differentiable at x0 . The second derivative at x0 is denoted f 00 (x0 ).
The higher order derivatives of the function f are defined by induction:
f 00 = (f 0 )0 , f 000 = (f 00 )0 , . . . , f (n+1) = (f (n) )0 .
f (n) is called the derivative of order n of the function f , it is also
d nf
denoted
.
dxn
A function f that admits continuous derivatives up to order k is called
a function of class C k :
C k [a, b] = {f | f : [a, b] R, f (k) , continuous}.
Theorem 1.1 (Leibnizs formula) If the functions f and g have derivatives of order n, then their product possesses a derivative of order n, given
by:
n
X
n (nk) (k)
(n)
f
g .
(1.2.1)
(f g) =
k
k=0
n
(by k are denoted the binomial coefficients Cnk ).
1.3

Taylors formula for functions of one variable


Let f : [a, b] R be a function of class C n that possesses also the
derivative of order n + 1 on (a, b). The value of the function f , in a
neighborhood of the point a, can be approximated using the given values
f (a), f 0 (a), . . . , f (n) (a).
More precisely, for all x (a, b], there exists (a, x) such that
f (x) = Tn [f ; a](x) + Rn [f ; a](x),

(1.3.1)

8
where
Tn [f ; a](x) = f (a) +

xa 0
(x a)n (n)
f (a) + +
f (a)
1!
n!

is called the Taylor polynomial of degree n, for the function f , at the


point a, and
(x a)n+1 (n+1)
Rn [f ; a](x) =
f
()
(n + 1)!
is the reminder in Lagranges form. Formula (1.3.1) is called Taylors
formula.
If a = 0, one gets MacLaurins formula:
f (x) = f (0) +

x 0
xn
xn+1 (n+1)
f (0) + + f (n) (0) +
f
().
1!
n!
(n + 1)!

1.4

Exercises
1.1 Using the definition of the derivative at a point, study if the following functions are differentiable at the indicated points:
i). f : R R, f (x) = x3 , at x0 = 2,

ii). f : [1, ) R, f (x) = x 1, at x0 = 5,


iii). f : R R, f (x) = |x + 1|, at x0 = 1,

iv). f : [0, ) R, f (x) = x, at x0 = 0.


1.2 Determine the derivatives of the functions f : D R below, mentioning the set D and also the differentiability set D1 :
i). f (x) = (x3 1)7 ,
2 +x+1

ii). f (x) = ex

iv). f (x) = ln

x2 + 1
,
3x

iii). f (x) = xe cos x,

v). f (x) =

1
,
ln x

9
r
vi). f (x) =

x1
,
x+1

vii). f (x) = arctan

x,

viii). f (x) = sin2 (x3 + 1),


ix). f (x) = esin x cos x,
x). f (x) = 3

1.3 Determine the derivatives of order n, n N, for the following functions:


i). f (x) = e2x ,
1
ii). f (x) =
, a R,
xa
iii). f (x) = ln(x a), a R,
iv). f (x) = (x a) , a R,
R \ N,
1
,
v). f (x) = 2
x 3x + 2

2x
,
x+1

vii). f (x) = x,
vi). f (x) =

viii). f (x) = sin x,


ix). f (x) = cos x,
x). f (x) = x3 ex .

1.4 Approximate the following functions using MacLaurins polynomial


of order n:
i). f (x) = ex ,
ii). f (x) = sin x,
iii). f (x) = cos x,

1
,
1x
1
v). f (x) =
,
1+x
vi). f (x) = ln(x + 1).
iv). f (x) =

Solutions and answers


f (x) f (2)
x3 8
= lim
= lim (x2 + 2x + 4) = 12,
x2
x2 x 2
x2
x2
0
so f is differentiable
at
2
and
f
(2)
=
12.

x1 4
x14
0

ii) f (5) = lim


= lim
=
x5
x5 (x 5)( x 1 + 2)
x5
1.1 i) We have lim

10
= lim
x5

1
1
= .
4
x1+2

iii) The function is f (x) =

x 1, if x 1,
so we have
x + 1, if x > 1,

f (x) f (1)
x 1
= lim
= 1,
x%1
x%1 x + 1
x+1
lim

f (x) f (1)
x+1
= lim
= 1.
x&1
x&1 x + 1
x+1
Since the two lateral limits are not equal, it means that f is not differentiable at 1.

1
x
f (x) f (0)
iv) lim
= lim
= lim = +, so f is not differenx&0
x&0 x
x&0
x
x
tiable at 0.
lim

1.2 i) D = D1 = R, f 0 (x) = 7(x3 1)6 3x2 .


2
ii) D = D1 = R, f 0 (x) = (2x + 1)ex +x+1 .
iii) D = D1 = R, f 0 (x) = ex cos x + xex cos x xex sin x.
3x 2x 3x (x2 + 1) 3
x2 1
iv) D = D1 = (0, ), f 0 (x) = 2

=
.
x +1
(3x)2
x(x2 + 1)
1
1
v) D = D1 = (0, 1) (1, ), f 0 (x) =
.
2
(ln x) x
vi) D = (, 1) [1, ), D1 = (, 1) (1, ),
1
x + 1 (x 1)
1

f 0 (x) = q
.

=
2
(x + 1)
(x + 1) x2 1
2 x1
x+1

1
1
.
1+x 2 x
0
3
viii) D = D1 = R, f (x) = 2sin(x +1) cos(x3 +1)3x2 = 3x2 sin 2(x3 +1).
ix) D = D1 = R, f 0 (x) = esin x (cos2 x sin x).

1
x) D = [0, ), D1 = (0, ), f 0 (x) = 3 x ln 3.
2 x
vii) D = R, D1 = R \ {0}, f 0 (x) =

1.3 i) f 0 (x) = 2e2x , f 00 (x) = 4e2x , by induction f (n) (x) = 2n e2x .


ii) Considering f (x) = (x a)1 , we get f 0 (x) = (x a)2 ,
f 00 (x) = 2(x a)3 , f 000 (x) = 2 3(x a)4 , and so on,
1
f (n) (x) = (1)n n!(x a)n1 = (1)n n!
.
(x a)n+1
1
iii) Since f 0 (x) =
, we have, using the previous exercise,
xa

11
1
.
(x a)n
iv) f (n) (x) = ( 1) . . . ( n + 1)(x a)n .
1
1
v) f (x) =

, so using the result from ii), we have


x 2 x 1

1
1
(n)
n
f (x) = (1) n!
.

(x 2)n+1 (x 1)n+1
2
2
vi) f (x) = 2
and f (n) = (1)n n!
.
1+x
(1 + x)n+1

f (n) (x) = (1)n1 (n 1)!

vii) f (x) = x1/2 , so we can use the result from iv), with =

1
. It
2

(2n 3)!! 1 n
x2 .
follows that f (n) (x) = (1)n1
2n

viii) f 0 (x) = cos x = sin(x + ), f 00 (x) = sin x = sin(x + ),


2
n
3
000
). Generally, f (n) (x) = sin(x +
).
f (x) = cos x = sin(x +
2
2
n
ix) f (n) (x) = cos(x +
).
2
x) Using Leibnizs formula (1.2.1), we get f (n) (x) = ex (x3 + 3nx2 +
3xn(n 1) + n(n 1)(n 2)).
1.4 i)f (n) (x) = ex , for all n N, f (n) (0) = e0 = 1, so we have
ex 1 +

x2
xn
x
+
+ + .
1!
2!
n!

x
x3 x5
x2k+1

+
+ (1)k
.
1!
3!
5!
(2k + 1)!
iii) Using the result from 1.3, ix) we get
ii) sin x

cos x 1

x2 x4
x2k
+
+ (1)k
2!
4!
(2k)!

1
1 + x + x2 + x3 + + xn , for x (1, 1).
1x
1
1 x + x2 x3 + + (1)n xn , for x (1, 1).
v)
1+x
(n 1)! (n)
, f (0) = (1)n1 (n 1)! so
vi) f (n) (x) = (1)n1
(x + 1)n
iv)

ln(x + 1) x

x2 x3
xn
+
+ (1)n1 ,
2
3
n

for x (1, 1).

CHAPTER 2

Differential calculus for functions of


several real variables
We will study in this chapter only functions of two or three real variables,
but all the definitions and properties can be extended to functions of an
arbitrary number of real variables.
2.1

Partial derivatives
Let E be a subset of R2 , f : E R a function and (x0 , y0 ) E. If the
limit
f (x, y0 ) f (x0 , y0 )
lim
xx0
x x0
exists, it is called the partial derivative of the function f , with
respect to x, at the point (x0 , y0 ). It is denoted
f
(x0 , y0 ) or fx0 (x0 , y0 ).
x
In the same way is defined the the partial derivative of the function
f , with respect to y, at the point (x0 , y0 ):
f (x0 , y) f (x0 , y0 )
f
(x0 , y0 ) = fy0 (x0 , y0 ) = lim
.
yy
y
y y0
0
13

14
For a function f : E R3 R, of three real variables, are defined:
f
f (x, y0 , z0 ) f (x0 , y0 , z0 )
(x0 , y0 , z0 ) = fx0 (x0 , y0 , z0 ) = lim
,
xx0
x
x x0
f
f (x0 , y, z0 ) f (x0 , y0 , z0 )
(x0 , y0 , z0 ) = fy0 (x0 , y0 , z0 ) = lim
,
yy0
y
y y0
f
f (x0 , y0 , z) f (x0 , y0 , z0 )
.
(x0 , y0 , z0 ) = fz0 (x0 , y0 , z0 ) = lim
zz0
z
z z0
The partial derivative fx0 (x0 , y0 , z0 ) is in fact the ordinary derivative
of the function f (, y0 , z0 ) (which is a function of one variable), at x0 ,
when y0 and z0 are fixed. In the same way, for fy0 , the variable is y, x and
z are considered fixed. For fz0 , only z is considered variable, x and y are
fixed. As a consequence, all the differentiation rules known for functions
of one real variable remain true also for functions of several variables.
2.2

Higher order partial derivatives


If f is a function defined on E R2 , that admits partial derivatives at
each point of E, then the partial derivatives of order two are defined
as follows:

2f
f
00
(x, y) = fx2 (x, y) =
(x, y),
x2
x x

f
2f
00
(x, y) = fy2 (x, y) =
(x, y),
y 2
y y
called pure derivatives and
2f

00
(x, y) = fxy
(x, y) =
xy
x

2f
00
(x, y) = fyx
(x, y) =
yx
y

f
y
f
x

(x, y),

(x, y).

called mixed derivatives.


For a function of three variables, the second order derivatives ( 9 in
number) are defined in a similar way.

15
00
00
In many cases, the mixed partial derivatives fxy
and fyx
are equal, the
order of successive partial differentiation is not important. The following
result is well-known:

Theorem 2.1 (Schwarz) If the function f : E R has continuous


2f
2f
and
in a neighborhood of a point
mixed partial derivatives
xy
yx
(x0 , y0 ) E, then the two derivatives are equal at the point (x0 , y0 ):
2f
2f
(x0 , y0 ) =
(x0 , y0 ).
xy
yx
The higher order partial derivatives are defined by induction. For
example,

3
3f
2f
4f

f
=
,
=
, etc.
2
2
2
2
xy
x y
x y
x xy 2
2.3

Derivatives of composite functions


We will present the rule for calculating the derivative of a composite
function (called the chain rule) in four different situations.
I. The first case is well known from high-school, each of the two
functions to be composed are functions of one real variable.
Let A R be an open interval, f : A R a continuous function. Let
u : (a, b) R be such that u(x) A, and u possesses a derivative at
each point x (a, b).
Theorem 2.2 If the derivative f 0 : A R is continuous then the function F : (a, b) R defined by F (x) = f (u(x)) (F = f u) has a derivative
that can be expressed as
F 0 (x) = f 0 (u(x)) u0 (x).

(2.3.1)

II. The second case deals with the composition between a function of
two variables and a vector-valued function of one variable.

16
Let D R2 be a domain and f : D R a continuous function. Let
u, v : (a, b) R two functions such that (u(x), v(x)) D and that they
possess partial derivatives, for each x (a, b).
Theorem 2.3 If at least one of the partial derivatives fu0 , fv0 is continuous, then the function F : (a, b) R, defined by F (x) = f (u(x), v(x))
has a derivative and this is equal to
F 0 (x) = fu0 (u(x), v(x)) u0 (x) + fv0 (u(x), v(x)) v 0 (x).

(2.3.2)

Formula (2.5) can also be written, with other notation:


F 0 (x) =

dF
f 0 f 0
=
u +
v.
dx
u
v

III. Another situation is when a function of two variables is composed


with a vector-valued function, also of two variables.
Let D R2 be a domain and f : D R a continuous function. Let
E R2 , u, v : E R two functions such that (u(x, y), v(x, y)) D and
they possess partial derivatives at each point (x, y) E.
Theorem 2.4 If at least one of the partial derivatives fu0 , fv0 is continuous, then the function F : E R, defined by F (x, y) = f (u(x, y), v(x, y))
has partial derivatives given by:
F
(x, y) =
x
F
(x, y) =
y

f
(u, v)
u
f
(u, v)
u

u
(x, y) +
x
u
(x, y) +
y

f
(u, v)
v
f
(u, v)
v

v
(x, y),
x
v
(x, y).
y

(2.3.3)

IV. In the last case that we present here, one of the functions to be
composed has one variable and the other has two variables.
Let A R be an open interval, f : A R a continuous function,
E R2 and u : E R, such that u(x, y) A and u possesses partial
derivatives at each point (x, y) E.

17
Theorem 2.5 If the derivative f 0 is continuous, then the function F :
E R defined by F (x, y) = f (u(x, y)) has partial derivatives given by:
F
(x, y) = f 0 (u(x, y))
x
F
(x, y) = f 0 (u(x, y))
y

u
(x, y),
x
u
(x, y).
y

(2.3.4)

2.4

Differential operators
An operator associates to a function another function, by a given rule.
We present here the most frequently used differential operators.
Let E Rn be an open set, f : E R, differentiable, with continuous partial derivatives. (For n = 2 or n = 3 we get a function of two or
three variables). The function f : E Rn , defined by

f =

f f
f
,
,...,
x1 x2
xn

is called the gradient of f .


The operator that associates to the function f its gradient f is called
the gradient operator. Notice that f is a real valued function of n real
variables, while the function f is a vector-valued function of n variables.
The sign is read "nabla". The gradient of f can also be denoted by
gradf .
Consider now a vector-valued function of n real variables,
F = (f1 , f2 , . . . , fn ) : E Rn , possessing continuous partial derivatives.
The function divF : Rn R, defined by
divF =

f2
fn
f1
+
+ +
x1 x2
xn

18
is called the divergence of f .
The operator that associates to the function F its divergence divF is
called the divergence operator. Notice that divF is a real valued function
of n real variables. Another used notation is divF = F .
Let f : E R, two times differentiable, with continuous partial
derivatives up to order II. The function f : E R, defined by
f =

2f
2f
2f
+
+

+
x21 x22
x2n

is called the laplacean of f .


The operator that associates to the function f its laplacean is called
the Laplace operator. Notice that f is a real valued function of n real
variables.
In fact, the Laplace operator is a combination between the divergence
and the gradient operators. Indeed, it is easy to check that:
f = div(f ) = (f ).
Let F = (f1 , f2 , f3 ) : E R3 R3 and define the function

curlF =

f3
f2 f1
f3 f2
f1

x2 x3 x3 x1 x1 x2

The function curlF is a vector-valued function of 3 variables.


The operator that associates to the function F curlF is called the curl
operator Another form of writing this operator (easier to remember) is:

~i ~j ~k

not

not
curlF = x 1 x 2 x 3 = F = rotF.

f1 f2 f3
curlF shows the rate of rotation of the vector field F . A vector field for
which curlF = 0 everywhere is called irrotational.

19
2.5

Differential of a function
A function f : Rn R is said to be (Frchet) differentiable at a
point a if there exists a linear function T : Rn R such that
lim

h0

|f (a + h) f (a) T (h)|
= 0.
khk

p
(If n = 2, khk = h21 + h22 , for h = (h1 , h2 ) R2 . )
If it exists, the function T is unique and it is called the differential
of f at the point a, and is denoted df (a), df (a) : Rn R.
Theorem 2.6 Let D Rn be an open set, f : D R a continuous
function, a D. If f has continuous partial derivatives at a, then it is
differentiable at a and
df (a)(h) = f (a) h,

for every h Rn .

For functions of two or three variables, we have


f
f
df (x0 , y0 ) =
(x0 , y0 ) dx +
(x0 , y0 ) dy,
x
y
f
f
f
df (x0 , y0 , z0 ) =
(x0 , y0 , z0 ) dx +
(x0 , y0 , z0 ) dy +
(x0 , y0 , z0 ) dz.
x
y
z
Higher order differentials are defined by induction:

d
d k f (a)(h) = (d k1 f (a + th)(h))t=0 .
dt
For a function of two real variables, the differential of the second order
is expressed
2f
2f
2f
2
(a)
dx
+
2
(a)
dx
dy
+
(a) dy 2 .
2
2
x
xy
y
For a function of three real variables:
2f
2f
2f
2
2
d 2 f (a) =
(a)
dx
+
(a)
dy
+
(a) dz 2 +
x2
y 2
z 2
2f
2f
2f
+2
(a) dx dy + 2
(a) dx dz + 2
(a) dy dz.
xy
xz
yz
d 2 f (a) =

20
2.6

Taylors formula for functions of several variables


As for functions of a real variable, Taylors formula gives the possibility to approximate a function by a polynomial, using the values of the
differentials at a point.
Let f : D R, where D is a domain in Rm , be a function that
possesses continuous partial derivatives up to order n + 1, in a convex
neighborhood of a point a D. Then, for an arbitrary point in this
neighborhood, Taylors formula holds:
f () = Tn [f ; a]() + Rn [f ; a](),
where
df (a)( a)
d n f (a)( a)
+ +
1!
n!
is called the Taylor polynomial of degree n, for the function f , at the
point a, and
d n+1 f ()( a)
Rn [f ; a]() =
(n + 1)!
is the reminder, with a point in the neighborhood of a.
If a = 0, one gets MacLaurins formula.
For instance, for a function of two variables, Taylors polynomial of
order 3, at a point (a, b), can be written

1
f
f
f (x, y) f (a, b) +
(x a) (a, b) + (y b) (a, b) +
1!
x
y

2
1
f
f
+
(x a)2 2 (a, b) + 2(x a)(y b)
(a, b)+
2!
x
xy

2
2 f
+(y b)
(a, b) +
y 2

3f
1
3f
+
(x a)3 3 (a, b) + 3(x a)2 (y b) 2 (a, b)+
3!
x
x y

3
3
2 f
3 f
+3(x a)(y b)
+ (y b)
(a, b) .
xy 2
y 3
Tn [f ; a]() = f (a) +

21
2.7

Exercises
2.1 Determine the partial derivatives of first order for the given functions, at an arbitrary point:
i). f : R2 R, f (x, y) = xy 3 + e2xy .
ii). f : R2 \ {(0, 0)} R, f (x, y) =
iii). f : E R, f (x, y) =

x2

xy
.
+ y2

x
.
y

iv). f : E R, f (x, y, z) = ln(1 + xy + z 2 ).


v). f : E R, f (x, y, z) = y x

2 3z

vi). f : E R, f (x, y) = xy + y x .

x
vii). f : E R, f (x, y) = x y sin .
y
p
viii). f : R2 R, f (x, y) = arctan x2 + y 2 .
x
y
ix). f : E R, f (x, y, z) = ln sin .
y
z
x). f : E R, f (x, y, z) = arcsin xyz.
2.2 Determine the partial derivatives of the second order for the given
functions, at an arbitrary point:
p
i). f : E R, f (x, y) = ln x2 y 2
ii). f : E R, f (x, y) = ex sin y
iii). f : E R, f (x, y) = arctan
iv). f : E R, f (x, y) =

v). f : E R, f (x, y, z) =

x
y

x2 + y
x+yz
xy+z
2

vi). f : E R, f (x, y, z) = exy z .

22
2.3 Check if the functions below satisfy the given relations:
i). f (x, y) = ln
ii). f (x, y) =

x2 + y 2 1, y

(2y + x)2
f
f
xy, x
+y
= f xy.
2x
x
y

iii). f (x, y, z) = y 2 + z 2

2xy 2xz
f
f
f

, yz
+y
+z
= 0.
z
y
x
y
z

iv). f (x, y) = ex cos y + 4xy,


v). f (x, y) =

f
f
x
= 0,
x
y

2f
2f
+
= 0.
x2
y 2

x
y 2f
2f
ln(x2 + y 2 ) y arctan ,
+
= 0.
2
x x2
y 2

2.4 Determine the derivatives of order one and two for the function F ,
F (x) = f (u(x), v(x)) where:
i). f : R2 R is a function that admits continuous partial derivatives
up to order II, u, v : (0, +) R, u(x) = x2 + 1, v(x) = ln x,
ii). f : R2 R is a function that admits continuous partial derivatives

up to order II, u, v : (0, ) R, u(x) = sin x, v(x) = x,


iii). f : R2 R is a function that admits continuous partial derivatives
up to order II, u, v : R R, u(x) = e2x , v(x) = ln(x2 + 1),
iv). f : R2 R, f (u, v) = u v 2 , u, v : (0, +) R, u(x) = x2 + 1,
v(x) = ln x,

v). f : R2 R, f (u, v) = u2 + v 2 , u, v : R R functions that admit


derivatives up to order II,

vi). f : R2 R, f (u, v) = u2 + v 2 , u(x) = x2 , v(x) = e x .


2.5 Determine the partial derivatives of order one and two for the function F , F (x, y) = f (u(x, y), v(x, y)) where:
i). f : R2 R is a continuous function that admits continuous partial
y
derivatives up to order II, u(x, y) = xy, v(x, y) = ,
x

23
ii). f : R2 R is a continuous function that admits continuous partial
derivatives up to order II, u(x, y) = x cos y, v(x, y) = y 2 ,
iii). f : R2 R is a continuous function that admits continuous parp
tial derivatives up to order II, u(x, y) = ln x2 + y 2 , v(x, y) =
arctan xy,
iv). f : R2 R, f (u, v) = u2 v 2 , u(x, y) = x2 y 2 , v(x, y) = exy ,

v). f : R2 R, f (u, v) = u + v, u(x, y) = tan xy, v(x, y) = 2x y.


2.6 Determine the partial derivatives of first and second order of the
function F , F (x, y) = f (u(x, y)) where:
i). f : R R is a function that admits continuous derivatives of order
one and two, u(x, y) = arctan xy,
ii). f : R R is a function that admits continuous derivatives of order
xy
one and two, u(x, y) =
,
x+y
iii). f : R R is a function that admits continuous derivatives of order
2
2
one and two, u(x, y) = ex y ,
2.7 Determine the partial derivatives of first and second order of the
function f , where g admits continuous derivatives of order one and two:
x
i). f (x, y) = xyg( ),
y
ii). f (x, y) = sin2 xy + yg(x2 + y 2 ).
2.8 Check if the following functions satisfy the corresponding partial
differential equations:
f
f
f
3y
z
= 0 ( is an
x
y
z
arbitrary function possessing continuous partial derivatives),

i). f (x, y, z) = (x3 y, xz) satisfies x

f
f
2y
) satisfies (x2 + y 2 )
+ 2xy
= 0 ( is an
2
y
x
y
arbitrary function possessing a continuous derivative),

ii). f (x, y) = (

x2

24
iii). f (x, y) = (x ay) + (x + ay), a > 0 satisfies fy002 = a2 fx002 ( and
are arbitrary functions possessing continuous derivatives up to
order II),
iv). f (x, y) = xy(x2 y 2 ), satisfies xy 2

f
f
+ x2 y
= z(x2 + y 2 ).
x
y

2.9 Determine the gradient f for the following functions, at the indicated points:
i). f (x, y) = xy ln(x + y), at (1, 1),
p
ii). f (x, y) = x3 y 2 , at (2, 2),
iii). f (x, y) = sin2 xy, at (0, ),
2

iv). f (x, y, z) = ex yz , at (1, 1, 1),


x+y+z
v). f (x, y, z) =
, at (0, 1, 2).
2x y + z
2.10 Prove that, for the following functions, the laplacean is identically
equal to zero. (Such functions are called harmonic.)
i). f (x, y) =

x2

y
,
+ y2

ii). f (x, y) = ex cos x + 4xy,


x
iii). f (x, y) = arctan ,
y
1
iv). f (x, y, z) = 2
x + y2 + z2
2.11 For the following vector-valued functions, find the divergence and
the curl.
i). F (x, y, z) = (xy + xz, xy + yz, xz + yz),
ii). F (x, y, z) = (y 2 + z 2 , x2 + z 2 , x2 + y 2 ).
2

2.12 For the function f (x, y) = ex y , determine the differentials: df (x, y),
df (1, 1), df (x, y)(3, 0), df (1, 1)(3, 0).

25
2.13 For the function f (x, y, z) = xy + yz + zx, determine the differentials: df (x, y, z), df (0, 1, 2), df (x, y, z)(1, 0, 1), df (0, 1, 2)(1, 0, 1).
2.14 Determine df and d 2 f for the following functions:
i). f (x, y) = x5 y + 2xy 2 ,
p
ii). f (x, y) = x2 y 2 ,
iii). f (x, y, z) = x + 2y 2 + z 3 .
2.15 Approximate the function f (x, y) = xy , in the neighborhood of
the point (1, 1), using Taylors polynomial of order 2.
2.16 Approximate the function f (x, y) = ln(x2 + y 2 ), in the neighborhood of the point (1, 1), using Taylors polynomial of order 2.
2.17 Approximate the function f (x, y) = exy , in the neighborhood of
the point (0, 1), using Taylors polynomial of order 3.
2.18 Write MacLaurins formula of order 2 for the function f (x, y) =
cos(2x + y).

Solutions and answers


f
f
(x, y) = y 3 + 2e2xy ,
(x, y) = 3xy 2 e2xy .
x
y
f
y(x2 + y 2 ) xy 2x
y 3 x2 y
ii)
(x, y) =
=
,
x
(x2 + y 2 )2
(x2 + y 2 )2
f
x(x2 + y 2 ) xy 2y
x3 xy 2
(x, y) =
=
.
y
(x2 + y 2 )2
(x2 + y 2 )2
2.1 i)

iii) To calculate the partial derivatives, we write f (x, y) = x


use the differentiation rule for products:
f
1
(x, y) = ,
x
y

f
x
(x, y) = 2
y
y

1
and we
y

26
iv)
y
f
1
(x, y, z) =
(1 + xy + z 2 )0x =
,
2
x
1 + xy + z
1 + xy + z 2
f
x
(x, y, z) =
,
y
1 + xy + z 2
v)

2z
f
(x, y, z) =
z
1 + xy + z 2

f
f
2
2
(x, y, z) = 2x y x 3z ln y,
(x, y, z) = (x2 3z)y x 3z1 ,
x
y

f
2
(x, y, z) = 3 y x 3z ln y.
z
f
1
y
f
1
y 1 x
y
ix)
(x, y, z) = sin ,
(x, y, z) = sin + ln cos ,
x
x
z
y
y
z z y
z
x
f
y
y
(x, y, z) = 2 ln cos .
z
z
y
z

x2 y 2
x2 y 2
2xy
00
00
,
f
=
, fxy
= 2
.
2
y
2
2
2
2
2
2
(x y )
(x y )
(x y 2 )2
00
ii) fx002 = ex sin y, fy002 = ex sin y, fxy
= ex cos y.
y
1
iv) fx002 = p
, fy002 = p
,
x2 + y(x2 + y)
4 x2 + y(x2 + y)
x
00
fxy
= p
.
2 x2 + y(x2 + y)
2
2
2
vi) fx002 = y 4 z 2 exy z , fy002 = 2xzexy z (2xy 2 z + 1), fz002 = x2 y 4 exy z ,
2
2
00
00
fxy
= 2yzexy z (xy 2 z + 1), fxz
= y 2 exy z (xy 2 z + 1),
2
00
fyz
= 2xyexy z (xy 2 z + 1).
2.2 i)fx002 = (fx0 )0x =

f
x
f
y
= 2
and
=
.
x
x + y2 1
x
x2 + y 2 1
Then we substitute the results in the differential expression and obtain
2.3 i) We calculate first

iii) We have

x2

x
y
x 2
= 0.
2
+y 1
x + y2 1

2y 2z f
2x 2xz f
2xy 2x
f
= ,
= 2y + 2 ,
= 2z + 2 .
x
z
y y
z
y
z
z
y

1
2.4 i) F 0 (x) = fu0 (u(x), v(x)) 2x + fv0 (u(x), v(x)) . For the second
x
derivative, we differentiate again, taking account that fu0 (u(x), v(x)) and

27
fv0 (u(x), v(x)) are also composite functions.

1
00
00
00
F (x) = fu2 2x + fuv
2x + fu0 2+
x

1
1
1
00
00
0
+ fuv 2x + fv2
+ fv 2 =
x
x
x
1
1
00
= 4x2 fu002 + 2 fv002 + 4fuv
+ 2fu0 2 fv0 .
x
x
1
ii) F 0 (x) = fu0 cos x + fv0 ,
2 x

1
00
00
00
F (x) = fu2 cos x + fuv
cos x fu0 sin x+
2 x

1
1
1
00
00
fv0 .
+ fuv cos x + fv2
2 x 2 x
4x x
iii) F 0 (x) = fu0 2e2x + fv0
F 00 (x) = 4e4x fu002 +

2x
,
+1

x2

2x 2x2 0
4x2
8xe2x 00
00
2x 0
f
f
+
4e
f
+
f .
+
2
u
(x2 + 1)2 v
x2 + 1 uv
(x2 + 1)2 v

1
2 ln x
= 2x
,
x
x

1
1
2
2 ln x
F 00 (x) = 2 2v 0 (x) 2v(x) 2 = 2 2 +
.
x
x
x
x2
u
v
v) F 0 (x) =
u0 (x) +
v 0 (x),
2
2
2
u +v
u + v2
1

F 00 (x) =
(v 2 u0 (x) uv v 0 (x)) u0 (x)+
2
2
2
2
(u + v ) u + v
u
+
u00 (x)+
2
u + v2
1

(uv u0 (x) + u2 v 0 (x)) v 0 (x)+


+
2
2
2
2
(u + v ) u + v
v
+
v 00 (x).
u2 + v 2
1
vi) We substitute in the previous exercise: u0 (x) = 2x, v 0 (x) = e x ,
2 x

x
e
1
u00 (x) = 2, v 00 (x) =
1 .
4x
x
iv) F 0 (x) = 2x 2v(x)

28
y
1
2.5 i) Fx0 (x, y) = fu0 y + fv0 2 , Fy0 (x, y) = fu0 x + fv0 .
x
x

y
00
Fx002 (x, y) = fu002 y + fuv
( 2 ) y+
x

y
y
2y
00
00
+ fuv y + fv2 ( 2 ) ( 2 ) + fv0 3 ,
x
x
x

1
1
1
00
00
00
00
00
x + fuv x + fv2
,
Fy2 (x, y) = fu2 x + fuv
x
x
x

y
00
00
( 2 ) x + fu0 +
Fxy
(x, y) = fu002 y + fuv
x

y 1
1
00
+ fuv
y + fv002 ( 2 ) + fv0 ( 2 ).
x
x
x
0
0
0
0
0
ii) Fx (x, y) = fu cos y, Fy (x, y) = fu (x sin y) + fv 2y,
Fx002 (x, y) = fu002 cos2 y,
00
Fy002 (x, y) = x2 sin2 yfu002 + 4y 2 fv002 4xy sin yfuv
x cos yfu0 + 2fv0 .
x
y
iii) Fx0 (x, y) = fu0 2
+ fv0
,
2
x +y
1 + x2 y 2
x
y
+ fv0
.
Fy0 (x, y) = fu0 2
2
x +y
1 + x2 y 2
iv) Fx0 (x, y) = 2u(x, y) 2x 2v(x, y)yexy = 4x(x2 y 2 ) 2ye2xy ,
Fy0 (x, y) = 4y(x2 y 2 ) 2xexy ,

Fx002 (x, y) = 2 2x 2x + 2u(x, y) 2 2 yexy yexy 2v(x, y) y 2 exy =


= 12x2 4y 2 4y 2 e2xy ,
Fy002 (x, y) = 4x2 + 12y 2 4x2 e2xy ,
00
Fxy
(x, y) = 2 (2y) 2x 2 xexy yexy 2v(x, y)(exy + yxexy =

= 8xy 4xye2xy 2e2xy .

y
x
1
1
0
0
v) Fx (x, y)
+ 2 , Fy (x, y) =
1
2 u + v cos2 xy
2 u + v cos2 xy
y
2.6 i) Fx0 (x, y) = f 0 (u(x, y)) u0x (x, y) = f 0
,
1 + x2 y 2
x
,
Fy0 (x, y) = f 0
1 + x2 y 2
Fx002 (x, y) = f 00

y2
2xy 3
0
+
f

,
(1 + x2 y 2 )2
(1 + x2 y 2 )2

29
Fy002 (x, y) = f 00

x2
2x3 y
0
+
f

,
(1 + x2 y 2 )2
(1 + x2 y 2 )2

xy
1 x2 y 2
0
+
f

.
(1 + x2 y 2 )2
(1 + x2 y 2 )2
2y
2x
ii) Fx0 (x, y) = f 0 (u(x, y))
, Fy0 (x, y) = f 0 (u(x, y))
.
2
(x + y)
(x + y)2
2
2
2
2
iii) Fx0 (x, y) = f 0 2xex y , Fy0 (x, y) = f 0 2yex y ,
00
Fxy
(x, y) = f 00

2 2y 2

+ f 0 (2ex

2 2y 2

f 0 (2ex

Fx002 (x, y) = f 00 4x2 e2x


Fy002 (x, y) = f 00 4y 2 e2x

00
Fxy
(x, y) = f 00 4xye2x

2 2y 2

2 y 2

+ 4x2 ex

2 y 2

4y 2 ex

f 0 4xyex

2 y 2

),

2 y 2

),

2 y 2

2.7 i) We apply the differentiation rules for products and for composite
functions.




x
x 1
x
x
0
0
0
fx (x, y) = yg
+ xy g
= yg
+ xg
y
y y
y
y


x
x2 0 x
0
fy (x, y) = xg
g
,
y
y
y



x
x
x
1
1
00
0
0
00
fx2 (x, y) = yg
+g
+xg
=
y
y
y
y
y


x
x
x
= 2g 0
+ g 00
,
y
y
y

x
x
x2 0 x
x2 00 x
x
00
0
fy2 (x, y) = xg
2 + 2g

g
2 =
y
y
y
y
y
y
y

3
x
x
= 3 g 00
,
y
y

x
x
x
x
x
00
0
00
fxy (x, y) = g
+ yg
2 +xg
2 =
y
y
y
y
y



2
x
x
x
x
x
=g
g0
2 g 00
.
y
y
y
y
y
ii) fx0 (x, y) = y sin 2xy + 2xyg 0 (x2 + y 2 ),
fy0 (x, y) = x sin 2xy + g(x2 + y 2 ) + 2y 2 g 0 (x2 + y 2 ).

30
f
f
f
(x, y, z) = 0u 3x2 y+0v z,
(x, y, z) = 0u x3 ,
(x, y, z) =
x
y
z
0v x. Substituting in the given expression, we get
2.8 i)

x(0u 3x2 y + 0v z) 3y 0u x3 z 0v x = 0
which is obviously true.
2
2
f
f
4xy
0 2x + 2y
ii)
,
.
(x, y) = 0 2
(x,
y)
=

x
(x y 2 )2 y
(x2 y 2 )2
iii) fx002 (x, y) = 00 (x ay) + 00 (x + ay), fy002 (x, y) = a2 00 (x ay) +
a2 00 (x + ay), clearly check the equation fy002 = a2 fx002 . This equation is
called the wave equation and it describes the propagation of different
waves, such as light waves, sound waves or water waves.
f
iv)
(x, y) = y(x2 y 2 ) + xy 2x 0 (x2 y 2 ),
x
f
(x, y) = x(x2 y 2 ) xy 2y 0 (x2 y 2 ).
y

xy
xy
2.9 i) f (x, y) = y ln(x + y) +
, x ln(x + y) +
,
x
+
y
x
+
y

1
1
f (1, 1) = ln 2 + , ln 2 +
.
2
2
!
3x2
y
p
ii) f (x, y) =
,p
, f (2, 2) = (3, 1).
2 x3 y 2
x3 y 2
iii) f (x, y) = (2y sin xy cos xy, 2x sin xy cos xy) = (y sin 2xy, x sin 2xy),
f (0, ) = (0, 0).
2
2
2
2
2
2
iv) f (x, y, z) = (2xyz 2 ex yz , x2 y 2 ex yz , 2x2 yzex yz ),
f (1, 1, 1) = (2e,e, 2e).

3y z
3x + 2z
x 2y
v) f (x, y, z) =
,
,
,
(2x y + z)2 (2x y + z)2 (2x y + z)2
f (0, 1, 2) = (5, 4, 2).
2.11 i) The divergence is a scalar: divF = 2(x + y + z). On the other
hand, the curl is a vector:

i
j
k

= i(z y) + j(x z) + k(y x),


x
y
z

xy + xz xy + yz xz + yz

31
or written in another way, curlF = (z y, x z, y x).
ii) divF = 0, curlF = (2y 2z, 2z 2x, 2x 2y).
2

2.12 df (x, y) = 2xyex y dx + x2 ex y dy, df (1, 1) = 2e dx + e dy,


2
df (x, y)(3, 0) = 6xyex y , df (1, 1)(3, 0) = 6e.
2.13 df (x, y, z) = (y + z) dx + (x + z) dy + (x + y) dz, df (0, 1, 2) =
3 dx + 2 dy + dz, df (x, y, z)(1, 0, 1) = (y + z) + (x + y) = x z,
df (0, 1, 2)(1, 0, 1) = 2.
2.14 i) df (x, y) = (5x4 y + 2y 2 ) dx + (x5 + 4xy) dy,
d 2 f (x, y) = 20x3 y dx2 + 2(5x4 + 4y) dx dy + 4x dy 2 .
y
x
ii) df (x, y) = p
dx p
dy,
2
2
2
x y
x y2
1
p
d 2 f (x, y) =
(y 2 dx2 2xy dx dy + x2 dy 2 ).
2
2
2
2
(x y ) x y
iii) df (x, y, z) = dx + 4y dy + 3z 2 dz, d 2 f (x, y, z) = 4 dy 2 + 6z dz 2 .
2.15

f
f
= yxy1 ,
= xy ln x,
x
y

2f
2f
2f
y2
y1
y1
=
y(y

1)x
,
=
x
+
yx
ln
x,
= xy (ln x)2 ,
x2
xy
y 2
f
f
2f
2f
2f
(1, 1) = 1,
(1, 1) = 0, 2 (1, 1) = 0,
(1, 1) = 1,
(1, 1) = 0.
x
y
x
xy
y 2
Finally we get
xy 1 + x 1 + (x 1)(y 1) = xy y + 1.
2.16

f
2x
f
2y
= 2
,
= 2
,
2
x
x + y y
x + y2
2f
2x2 + 2y 2 2 f
4xy
2f
2x2 2y 2
=
,
=
,
=
,
x2
(x2 + y 2 )2 xy
(x2 + y 2 )2 y 2
(x2 + y 2 )2

Finally we get
ln(x2 + y 2 ) ln 2 + (x 1) 1 + (y + 1) (1)+
1
+ [(x 1)2 0 + 2(x 1)(y + 1) 1 + (y + 1)2 0] =
2
= ln 2 + xy + 2x 2y 3.

32
3
3f
3f
3f
3 xy f
3 xy
xy
2 xy
=
y
e
,
=
x
e
,
=
=
2ye
+
xy
e
,
x3
y 3
x2 y
xy 2
2xexy + x2 yexy . The approximation is

2.17

1
1
exy 1 + x + [x2 + 2x(y 1)] + [x3 + 3x2 (y 1) 2] =
2
6
2
x3
x
=
+ x2 y
+ xy + 1.
6
2
1
sin(2 + )
2.18 cos(2x + y) = 1 2x2 2xy y 2 +
(8x3 + 12x2 y +
2
6
6xy 2 + y 3 ).

CHAPTER 3

Extrema of functions
3.1

Local extrema of a function of one real variable


Let A R be an open interval, f : A R a function and x0 A. The
point x0 is a point of local minimum (maximum) of f if there exists
a neighborhood U of x0 such that
f (x0 ) f (x), for each x U,
(f (x0 ) f (x), for each x U ).
The points of local minimum or maximum are called points of local
extremum.
Theorem 3.1 (Fermat) If the function f : A R is differentiable on
A and x0 A is a point of local extremum for f , then f 0 (x0 ) = 0.
A point x0 A for which f 0 (x0 ) = 0 is called a critical or stationary
point. The converse of Fermats Therorem is not always true. There
exist critical points that are not points of extremum.

33

34
Theorem 3.2 Let f : A R be a function that is two times differentiable on the interval A and x0 A a critical point of f . If f 00 (x0 ) < 0,
then x0 is a point of minimum. If f 00 (x0 ) > 0, then x0 is a point of
maximum.
3.2

Local extrema of a function of several real variables


For simplicity, we will present the theory for a function of two real variables.
Let D R2 be an open set and f : D R a function.
The point (x0 , y0 ) D is called a point of local minimum (maximum) of f if there exists a neighborhood U of (x0 , y0 ) such that
f (x0 , y0 ) f (x, y), for each (x, y) D,
(f (x0 , y0 ) f (x, y), for each (x, y) A).
The value f (x0 , y0 ) is called a local extremum of f .
Assume that the function f has partial derivatives at (x0 , y0 ). If
f
f
(x0 , y0 ) =
(x0 , y0 ) = 0, the point (x0 , y0 ) is called a critical or
x
y
stationary point of f .
Theorem 3.3 If (x0 , y0 ) D is a point of local extremum for f , then is
also a stationary point for f .
As for the functions of one variable, the converse is not true: not
all the stationary points are points of extremum. To determine whether
they are or not, the second order differentials are used.
The second order differential d2 (v) is called:
positive definite at v if d2 f (v)(h) > 0, for each h R2 , h 6= 0; negative
definite at v if d2 f (v)(h) < 0, for each h R2 , h 6= 0; indefinite at v if
there exist h1 , h2 R2 such that d2 f (v)(h1 ) > 0 and d2 f (v)(h2 ) < 0.

35
Theorem
If d2 f (v0 )
If d2 f (v0 )
If d2 f (v0 )

3.4 Let v0 = (x0 , y0 ) D be a critical point of f.


is positive definite, then (x0 , y0 ) is a point of local minimum.
is negative definite, then (x0 , y0 ) is a point of local maximum.
is indefinite, then (x0 , y0 ) is not a point of local extremum.

Theorem 3.5 (Sylvesters criterion) Let v = (x, y) and denote

f 00 (v) f 00 (v)

2
xy
D2 = x00
D1 = fx002 (v),
.
fxy (v) fy002 (v)
d2 f (v) is positive definite if and only if D1 > 0 and D2 > 0.
d2 f (v) is negative definite if and only if D1 < 0 and D2 > 0.
According to the results above, we can use the following ALGORITHM to determine the local extrema of a function:
Find the stationary points (x0 , y0 ) by solving the system of equations
(
fx0 (x, y) = 0
fy0 (x, y) = 0
For each stationary point (x0 , y0 ), calculate D1 and D2 .
If D1 > 0, D2 > 0 then (x0 , y0 ) is a point of minimum.
If D1 < 0, D2 > 0 then (x0 , y0 ) is a point of maximum.
If D2 < 0 then (x0 , y0 ) is not a point of extremum. (it is called a
saddle point).
If D1 or D2 are zero, we can not tell, using this method, if (x0 , y0 ) is
a point of extremum.
Remark 3.6 For a function of three real variables, the algorithm is similar:
Find the stationary points (x0 , y0 , z0 ) for which the partial derivatives
are zero,
For each stationary point (x0 , y0 , z0 ), calculate D1 , D2 and

f 002 (v0 ) f 00 (v0 ) f 00 (v0 )


xz
xy

00
00
00
D3 = fxy
(v ) f (v ) f (v0 )

00 0 y002 0 yz
fxz (v0 ) fyz (v0 ) f 002 (v0 )
z

36
If D1 > 0, D2 > 0, D3 > 0 then (x0 , y0 ) is a point of minimum.
If D1 < 0, D2 > 0, D3 < 0 then (x0 , y0 ) is a point of maximum.
If one of D1 , D2 or D3 is zero-undecided.
In other situations, (x0 , y0 , z0 ) is not a point of extremum.
3.3

Exercises
3.1 Determine the points of local extrema for the functions of one real
variable:
i). f : R R, f (x) = x arctan 2x,

ii). f : (0, ) R, f (x) = sin x2 ,

iii). f : R R, f (x) = x2 2x + 4
iv). f : R R, f (x) = x3 3x2 9x + 1,
v). f (, 1) R, f (x) =

x+2
,
x1
1

vi). f : (0, ) R, f (x) = xe x ,


vii). f : R R, f (x) = (x 1)3 .
3.2 Determine the points of local extrema for the functions of several
real variables:
i). f (x, y) = x3 y 3x2 4y 2 ,
ii). f (x, y) = x2 + y 2

2
,
xy

iii). f (x, y) = x3 + y 4 ,
iv). f (x, y) = 3x2 y + y 3 3x2 3y 2 + 2,
v). f (x, y) = ln(16 x y) + 2 ln x + 5 ln y,
vi). f (x, y) = (1 + ey ) cos x yey .

37
2 +y 2 )

vii). f (x, y) = (x + y)e(x

viii). f (x, y, z) = 2x2 y + xz + y 2 + z 2 + 9z

Solutions and answers


2
4x2 1
=
.
1 + 4x2
1 + 4x2
1
From f 0 (x) = 0, we get 4x2 1 = 0, so the stationary points are x1 =
2
1
and x2 = . We determine the second derivative
2
3.1 i) The derivative of the function is f 0 (x) = 1

f 00 (x) =

8x(1 + 4x2 ) 8x(4x2 1)


16x
=
,
2
2
(1 + 4x )
(1 + 4x2 )2

1
1
1
and obtain f 00 ( ) > 0 and f 00 ( ) < 0. This means that is a point
2
2
2
1
2
1
of local minimum, with the minimum value f ( ) =
and is a
2
4
2
2
1
.
point of local maximum, with the maximum value f ( ) =
2
4
x cos x2
ii) We solve the equation f 0 (x) = 0, that is
= 0. The only solusin x2
r

. The second order derivative


tion from the interval (0, ) is x0 =
2
is
cos x2 sin x2 2x2 sin2 x2 x2 cos2 x2

f 00 (x) =
,
sin x2 sin x2
r
r

00
and f (
) = < 0. So
is a point of maximum.
2
2
x1
iii) f 0 (x) =
, so the stationary point is x0 = 1. The second
x2 2x + 4
3
> 0, so 1 is a minimum point, with
derivative is f 00 (x) = 2
3/2
(x 2x + 4)
the value f (1) = 3.
iv) f 0 (x) = 3x2 6x 9, solving the equation x2 2x 3 = 0 we get two
stationary points x1 = 1, x2 = 3. f 00 (1) = 12 < 0, f 00 (3) = 12 > 0,
so 1 is a point of local maximum, f (1) = 6 while 3 is a point of local

38
minimum, f (3) = 26.
3
v) Since f 0 (x) =
6= 0 for any x (, 1), the function does
(x 1)2
not have extremum points on the open interval (, 1). The derivative
is negative on this interval, so the function decreases from lim f (x) = 1
x

to lim f (x) = .
x%1

vi) Using the differentiation rule for products, we get f (x) = e

1
x

1
1
x

1 1
and f 00 (x) = 3 e x . The equation f 0 (x) = 0 has the solution x0 = 1, and
x
f 00 (1) = e > 0 so x0 = 1 is a point of minimum.
vii) f 0 (x) = 3(x 1)2 , so x0 = 1 is the stationary point. f 00 (1) = 0, also,
so we can not decide if this is en extremum point by this method. But,
since f 0 (x) 0, the function is increasing and it follows that x0 = 1 is
not a point of extremum.
3.2 i) We find first the stationary points, by solving the system
(
fx0 (x, y) = 3x2 y 6x = 0
fy0 (x, y) = x3 8y = 0
We obtain three solutions for the system:
(
(
(
x1 = 0
x2 = 2
x3 = 2
y1 = 0,
y2 = 1,
y3 = 1,
We determine the partial derivatives of order II:
fx002 (x, y) = 6xy 6,

fy002 = 8,

We study now each stationary point.

00
fxy
= 3x2 .

6 0

For (x1 , y1 ) = (0, 0) we have D1 = 6, D2 =


= 48. We have
0 8
D1 < 0 and D2 > 0, so (0, 0) is a point of local maximum, the maximum
value is f (0, 0) = 0.

6 12

For (x2 , y2 ) = (2, 1) we have D1 = 6, D2 =


= 192. Since
12 8
D2 < 0, (2, 1) is not a point of extremum.

39
For (x3 , y3 ) = (2, 1) the values are the same D1 = 6, D2 = 192, so
this not an extremum point.
ii) We find first the stationary points, by solving the system

fx0 (x, y) = 2x + 2 = 0
xy
2

0
fy (x, y) = 2y +
=0
xy 2
1
= x3 and substituting in the
3
x
second one, x8 1 = 0, which has only two real solutions: 1 and 1. So
the stationary points are (x1 , y1 ) = (1, 1) and (x2 , y2 ) = (1, 1). The
4
4
2
00
= 2 2 . For
second derivatives are fx002 = 2 3 , fy002 = 2 3 , fxy
xy
xy
xy
both stationary points D1 = 6 > 0 and D2 = 32, so they are points of
local minimum, with the local minima f (1, 1) = f (1, 1) = 4.
iii) The system
(
f 0 (x, y) = 3x2 = 0
,
f 0 (x, y) = 4y 3 = 0
From the first equation we have y =

has the unique solution (0, 0). For this stationary point, we get D1 =
D2 = 0, so we can not decide if (0, 0) is an extremum point or not, using
this method. We notice that f (0, 0) = 0 and f (x, 0) = x3 , for any real
number x. For positive values of x, very close to zero, we have that
f (x, 0) > 0 and for negative values f (x, 0) < 0. This implies that (0, 0)
is not a point of extremum.
iv) (0, 0) is a point of maximum, (0, 2) a point of minimum and (1, 1),
(1, 1) are saddle points.
v) We write the system

1
2

+ =0
16 x y x
,
1
5

+ =0
16 x y y
5
2
= and finally the stationary point (x0 , y0 ) = (4, 10). The
x
y
1
2
second order partial derivatives are fx002 =

, f 002 =
(16 x y)2
x2 y
It follows

40
1
5
1
00
2 and fxy
=
. For the stationary point,
2
(16 x y)
y
(16 x y)2
3
4
we get D1 = < 0, D2 =
> 0 so the point is a maximum one.
8
16
2

2 y 2

vii) We have fx0 (x, y) = ex y (12x2 2xy) and f 0 (x, y) = ex


2
2
2y 2 2xy). Since ex y 6= 0, we solve the system
(
2x2 2xy + 1 = 0
,
2y 2 2xy + 1 = 0

(1

1 1
1 1
obtaining the stationary points
and , . The second
,
2 2
2 2
00
x2 y 2
3
(4x + 4x2 y 6x 2y), fy002 =
order partial derivatives are fx2 = e
2
2
2
2
00
ex y (4y 3 + 4xy 2 2x 6y), fxy
= ex y (4x2 y + 4xy 2 2x 2y).

1 1
For
,
we have D1 = 3e1/2 < 0, D2 = 8e1/2 > 0, so this
2 2

1 1
is a point of maximum. For ,
we have D1 = 3e1/2 > 0,
2 2
D2 = 8e1/2 > 0, so this is a point of minimum.
viii) The only stationary point is (1, 1, 4), but since D2 < 0, it is
not a point of extremum.

CHAPTER 4

Implicit functions
4.1

Implicit functions of one real variable


Consider the equation
F (x, y) = 0,

(4.1.1)

where F is a function of two variables, F : A B R, A and B are


subsets of R.
If, for each x A fixed, the equation F (x, y) = 0, with the unknown y has a unique solution, then a function f : A B that satisfies
F (x, f (x)) = 0, for all x A is called an implicit function defined by
equation (4.1.1).
The following result gives conditions in which such an implicit function exists and admits a derivative.
Theorem 4.1 Let x0 , y0 R, U a neighborhood of x0 and V a neighborhood of y0 . Let F : U V R be a continuous function that satisfies
the following conditions:
(a) F (x0 , y0 ) = 0,
(b) there exist the partial derivatives Fx0 and Fy0 , continuous on U V ,
(c) Fy0 (x0 , y0 ) 6= 0.
41

42
Then there exist some neighborhoods U1 , V1 of x0 , respectively y0 and
a function f : U1 V1 such that:
(i) F (x, f (x)) = 0, for each x U1 ,
(ii) f (x0 ) = y0 ,
(iii) f has a continuous derivative on U1 and this is equal to
Fx0 (x, f (x))
f (x) = 0
.
Fy (x, f (x))
0

(4.1.2)

If F has partial derivatives of the second order, then the second derivaF 0 (x, y)
tive of f can be determined from f 0 (x) = x0
by differentiating
Fy (x, y)
with respect to x and taking account that y is a function of x.
4.2

Implicit functions of two real variables


Implicit functions of two (or several ) real variables are defined in a similar
way as above. We have:
Theorem 4.2 Let x0 , y0 , z0 R, U , V , W neighborhoods of x0 , y0 and
z0 respectively. Let F : U V W R be a continuous function that
satisfies the following conditions:
(a) F (x0 , y0 , z0 ) = 0,
(b) there exist the partial derivatives Fx0 , Fy0 and Fz0 continuous on
U V W,
(c) Fz0 (x0 , y0 , z0 ) 6= 0.
Then there exist some neighborhoods U1 , V1 , W1 of x0 , y0 , z0 respectively, and a function f : U1 V1 W1 such that:
(i) F (x, y, f (x, y)) = 0, for each x U1 , y V1 ,
(ii) f (x0 , y0 ) = z0 ,
(iii) f has continuous partial derivatives on U1 V1 and they are given
by:

43

fx0 (x, y) =

Fx0 (x, y, f (x, y))


,
Fz0 (x, y, f (x, y))

fy0 (x, y) =

Fy0 (x, y, f (x, y))


.
Fz0 (x, y, f (x, y))

(4.2.1)

4.3

Exercises
4.1 Show that the equation x2 3xy + y 2 x + y 2 = 0 defines an
implicit function y(x) in a neighborhood of the point (0, 2). Determine
y 0 (x), y 00 (x), y 0 (0) and y 00 (0).
4.2 Show that the equation x2 2xy + 5y 2 2x + 4y + 1 = 0 defines an
implicit function y(x) in a neighborhood of the point (1, 0). Determine
y 0 (x), y 00 (x), y 0 (0) and y 00 (0).
4.3 Find the derivative y 0 for the function y(x) defined implicitly by the
equation (at an arbitrary point where the conditions of Theorem 4.1 are
satisfied) :
i). x3 + y 3 xy + 5 = 0,
ii). x2 y 3xy + y 2 2x + y = 0,
iii). exy x2 y 2 + x + y = 0,
y p
iv). arctan + x2 + y 2 = 0,
x

v). ln(x + y x y) = 0.
4.4 Determine the equation of the tangent line to the curve
x(x2 + y 2 ) + 5(y 2 x2 ) = 0
4
at the point (4, ).
3
4.5 Determine the first and second order partial derivatives at the point
(2, 0, 1) for the implicit function z(x, y) defined by the equation
xy ln z y 2 + z 2 + xz 3 = 0.

44
4.6 Find the partial derivatives zx0 and zy0 for the function z(x, y) defined
implicitly by the equation (at an arbitrary point where the conditions of
Theorem 4.2 are satisfied):
i). x2 + 3y 2 + z 3 + 5z 8 = 0,
ii). xy + 2z +

z2
= 0,
xy

iii). xyez x y + z = 0,
iv). x y 2 + z + tan z = 0.
4.7 The implicit function z(x, y) is defined by the equation
(y z 2 , 2z 3 + 3x2 3yz) = 0, where = (u, v) is a function that
admits continuous partial derivatives of order one. Prove that z satisfies
the relation
y z
z
+ 2xz
= x.
(4.3.1)
2 x
y
4.8 The implicit function z(x, y) is defined by the equation
x
( , y 4 x2 z 2 ) = 0, where = (u, v) is a function that admits continz
uous partial derivatives of order one. Prove that z satisfies the relation
xy 3

z
z
+ x2 z 2
= y 3 z.
x
y

Solutions and answers


4.1 Denoting F (x, y) = x2 3xy + y 2 x + y 2 = 0, we check the
conditions of Theorem 4.1.
F (0, 2) = 4 2 2 = 0,
Fx0 (x, y) = 2x3y1,

Fy0 (x, y) = 3x+2y+1, and Fy0 (0, 2) = 3 6= 0.

Using formula (4.1.2) we get


y 0 (x) =

2x 3y 1
5
and y 0 (0) = .
3x 2y 1
3

45
For the second derivative we differentiate the relation above with respect
to x, taking account that y is a function of x:
(2 3y 0 )(3x 2y 1) (2x 3y 1)(3 2y 0 )
=
(3x 2y 1)2
5y + 1 + (1 5x) 2x3y1
5y + 1 + y 0 (1 5x)
3x2y1
=
=
=
(3x 2y 1)2
(3x 2y 1)2
30xy 10x2 10y 2 + 10x 10y 2
=
,
(3x 2y 1)2

y 00 (x) =

y 00 (0) =

22
.
27

xy1 0
, y (1) = 0, y 00 (1) = 1 (so x = 1 is a
x 5y 2
point of maximum for the implicit function y).
4.2 We have y 0 (x) =

4.3 i) Denoting F (x, y) = x3 + y 3 xy + 5, we have Fx0 (x, y) = 3x2 y


and Fy0 (x, y) = 3y 2 x. According to formula (4.1.2) we have
y 3x2
dy
= 2
.
y 0 (x) =
dx
3y x
3y 2xy + 2
yexy 2x + 1
ii) y 0 = 2
. iii) y 0 = xy
.
xp
3x + 2y + 1
xe 2y + 1

y x2 + y 2 x
12 xy
0
0

. v) y =
iv) y = p
.
1+2 xy
x x2 + y 2 + y
4.4 We denote F (x, y) = x(x2 + y 2 ) + 5(y 2 x2 ) and check that
4
F (4, ) = 0, which means that the given point really belongs to the
3
curve. The derivative of the implicit function y(x) is
y 0 (x) =

Fx0 (x, y)
3x2 + y 2 10x
=

,
Fy0 (x, y)
2xy + 10y

and, at the given point y 0 (4) =


y

11
. So the equation of the tangent line is
27
4
11
= (x 4).
3
27

46
4.5 F (2, 0, 1) = 0, Fz0 (2, 0, 1) = 4 6= 0. The first derivatives are
y ln z + z
,
zx0 (x, y) = xy
+ 2z + x
z

x ln z + 2y
zy0 (x, y) = xy
,
+ 2z + x
z

(4.3.2)

1 0
and at the specified point zx0 (2, 0) =
, zy (2, 0) = 0. For the sec4
ond order partial derivatives we differentiate relations (4.3.2) taking ac3
count that z is a function of x and y. Finally we obtain zx002 (2, 0) = ,
32
1 00
1
00
zxy (2, 0) = , zy2 (2, 0) = .
8
2
2x
6y
4.6 i) zx0 (x, y) = 2
, zy0 (x, y) = 2
.
3z + 5
3z + 5
z 2 x2 y 2
z 2 x2 y 2
ii) zx0 (x, y) = 2
, zy0 (x, y) =
.
2x y + 2xz
2xy 2 + 2yz
1 xez
1 yez
0
(x,
y)
=
,
z
.
iii) zx0 (x, y) =
y
xyez + 1
xyez + 1
1
2y
iv) zx0 (x, y) =
, zy0 (x, y) =
.
2
2 + tan z
2 + tan2 z
4.7 Denote u = y z 2 , v = 2z 3 + 3x2 3yz and F (x, y, z) = (y
z 2 , 2z 3 + 3x2 3yz). Using the chain rule for the differentiation of composite functions, we get:
Fx0 (x, y, z) = 0u u0x + 0v vx0 = 0v 6x,
Fy0 (x, y, z) = 0u u0y + 0v vy0 = 0u + 0v (3z),
Fz0 (x, y, z) = 0u u0z + 0v vz0 = 0u (2z) + 0v (6z 2 3y),
and

z
6x0v
=
,
x
2z0u + (6z 2 3y)0v
z
3z0v 0u
=
.
y
2z0u + (6z 2 3y)0v
Plugging these into the equation (4.3.1), it follows that it is satisfied.

4.8

z
=
x

1 0
2xz 2 0v
z u
,
x 0
2 z0
+
2x

2
v
u
z

z
=
y

4y 3 0v
.
x 0
2 z0
+
2x

2
v
u
z

CHAPTER 5

Change of variables
Mathematical expressions that contain derivatives can be sometimes simplified by making a change of variables.
5.1

Change of the independent variable for functions of


one real variable
Let D, D1 be subsets of R and consider the function y : D R, y = y(x).
Consider a bijective, continuous differentiable function : D1 D,
possessing a continuous differentiable inverse (diffeomorphism) and with
0 (t) 6= 0, for each t D1 .
We make the change of variable x = (t), composing the functions y and
, and we get
y(t) = y((t)) = y(x(t)).
The new independent variable is t, so it is necessary to express the derivatives of y, using this new variable. Using the differentiation rules for
composite functions, we get:
dy dx
dy
=

dt
dx dt
47

48
so
y 0 (x) =

dy
1 dy
1
dy
= dx
= 0 .
dx
dt
x (t) dt
dt

d
1
d
() = 0
(). For the second
dx
x (t) dt
dy
derivative of y, this operator is applied to the function y 0 (x) =
(x):
dx

1
1
d dy
d dy
d
1
dy
00
= 0
= 0
=
y (x) =

dx dx
x (t) dt dx
x (t) dt x0 (t) dt

1
x00 (t) dy
1
d 2y
= 0
0

.
+

x (t)
(x (t))2 dt x0 (t) dt2

We have obtained the operator

In the same way, the operator can be applied to express the higher order
derivatives y 000 (x), y IV (x), etc. So an expression
E = f (x, y(x), y 0 (x), y 00 (x), . . . )
becomes
E = g(t, y(t), y 0 (t), y 00 (t), . . . ).
5.2

Change of the independent variables for functions of


several real variables
Let D, D1 be subsets of R2 and consider the function z : D R, z =
z(x, y). Consider the homeomorphisms : D1 R, : D1 R such
that ((u, v),
(u, v))
D, for each (u, v) D1 and with the Jacobian

0
0
D(, ) u v
= 0 0 different from zero.
u v
D(u, v)
We make the change of variables x = (u, v), y = (u, v); u and v will
be the new independent variables:
z(u, v) = z((u, v), (u, v)) = z(x(u, v), y(u, v)).
The partial derivatives of the function z have to be expressed in terms
of the new variables (u, v). It is necessary to find relations between the

49
differential operators
operators


,
with respect to the old variables and the
x y


,
with respect to the new variables. We have:
u v

z
z x z y

u
x u y u
z
z x z y

.
v
x v y v

Considering this a system with the unknowns


Cramers rule, we get

x =
z

y =

1
D(x,y)
D(u,v)

D(x,y)
D(u,v)

z
z
and
and applying
x
y

z y z y

u v v u

z x z x
.

v u u v

The same operators are used also for higher order partial derivatives.
Finally, an expression
E = f (x, y, z,

z z 2 z
, ,
,...)
x y x2

E = g(u, v, z,

z z 2 z
, ,
, . . . ).
u v u2

becomes

5.3

Interchange of variables
Let D be a subset of R and y : D R a function, y = y(x). Consider a
differential expression in which y is the function and x the independent
variable. In some situations, it is useful to change the roles of the variables: x will become the function and y the independent variable, that is
we will have the function x = x(y). To express the derivatives, we have:
y 0 (x) =

dy
1
1
=
= 0 ,
dx
dx
x (y)
dy

50

d
1
1
d
1
d dy
y =
=
= 0

=
dx dx
dx x0 (y)
x (y) dy x0 (y)
1
x00 (y)
x00 (y)
= 0
0
=

.
x (y) (x (y))2
(x0 (y))3
00

5.4

Exercises
5.1 Transform the given equations, where y = y(x), by the indicated
change of the independent variable:
i). x2 y 00 xy 0 + 10y = x, x = et ,
ii). (x + 1)2 y 00 + 3(x + 1)y 0 5y = ln(x + 1), x = et 1,
iii). x3 y 000 2x2 y 00 + xy 0 2y = 0, x = et ,
iv). (2x 3)2 y 00 5(2x 3)y 0 + 10y = 0 , 2x 3 = t,
v). x2 y 00 3xy 0 +

4
1
y
=
0,
x
=
x3
t

vi). (x2 1)y 00 + xy 0 + 1 = 0, x = sin t,


vii). (1 + x2 )y 00 + xy 0 y = 0, x = sinh t,
viii). (1 + x2 )2 y 00 x(1 + x2 )y 0 + y = 0, x = tan t.
5.2 Transform the following equations by interchange of variables:
i). y 0 cos y(sin y x) = 1,
ii). (2xy + y y 3 )y 0 = 1,
iii). y 00 x(y 0 )2 + (y 0 )2 sin y = 0,
iv). y(y 0 )3 + y 00 = 0.
5.3 Transform the given expressions, where z = z(x, y), by changing
the independent variables x, y into polar coordinates , :

51
i). E = x

z
z
y
x
y

ii). z =

2z 2z
+
x2 y 2

5.4 Transform the equations with the unknown function z = z(x, y), by
changing the independent variables x, y into u, v according to the given
relations:
i). 2

2z
2z
2z
5

7
+
5
=
0;
u
=
x + y, v = x + y,
x2
xy
y 2
2

ii). 2

2z
2z
2z
z
z
+

+2

= 0; u = x y, v = x + 2y,
2
2
x
xy y
x y

iii).

2z
2z
2z

4
= 0; u = 2x + y, v = y,
+
4
x2
xy
y 2

2
2z
2z
z
2 z

2xy
+
y
+ 2y
= 0, u = xy, v = y,
2
2
x
xy
y
y
z
z
v). x
+y
= 0; u = x2 + y 2 , v = x,
x
y
z
z
y
vi). x
2y
= 0; u = x, v = ,
x
y
x

iv). x2

2
z
z
x
2 z
vii). 4x
y
+ 2x
= 0; u = 2 , v = xy 2 ,
2
2
x
y
x
y
2

5.5 Transform the equations with the unknown function y = y(x), by


changing the function with z(x), according to the given relations
i). xy 0 + 2y = 4y 2 , z = y 1 ,

ii). y 0 + 2y = 2x y, z = y,
iii). y 0 tan y =

ex
, z = sin y,
cos y

iv). x2 y 00 + 4xy 0 + (2 x2 )y = 4x, z = x2 y.


z
2z

cz = 0, a, c R, with the
2
x
y
unknown function z = z(x, y), by changing the unknown function into
w(x, y) = ecy z(x, y).
5.6 Transform the equation a2

52
5.7 Find

dy
for the function defined in parametric way
dx
(
x = arctan t,
y = ln(1 + t2 ), t R.

Solutions and answers


dy
1
dy
1 dy
dy
= 0

, that is y 0 = t
= et
. For
dx
x (t) dt
e dt
dt
d
1
d
the second derivative y 00 we apply the operator
() = 0
() to
dx
x (t) dt
dy
dy
= et
and we get:
dx
dt

1
d dy
1
d
00
t dy
y = 0
= 0
e
=
x (t) dt dx
x (t) dt
dt

2
dy d 2 y
1
t dy
t d y
2t
+e
=e
+ 2 .
= t e
e
dt
dt2
dt
dt
5.1 i) y 0 =

Plugging these into the given equation, we obtain

dy
dy d 2 y
2t 2t
+ 10y = et , or
e e
+ 2 et et
dt
dt
dt
d 2y
dy

2
+ 10y = et ,
dt2
dt
which is the new differential equation, with the unknown function y and
the independent variable t.
ii) In the same way as in the previous exercise, we have

dy d 2 y
0
t dy
00
2t
y =e , y =e
+ 2 .
dt
dt
dt
d 2y
dy
Substituting in the equation it follows 2 + 2 5y = t.
dt
dt
iii) In addition to exercise i), we also have

d 2y d 3y
dy
000
3t
y =e
2 3 2 + 3 ,
dt
dt
dt

53
and the equation becomes
d 3y
d 2y
dy

5
+ 5 2y = 0.
3
2
dt
dt
dt

dy 00
1
d
dy
1 0
1
d
dy
0
iv) x (t) = , y = 2 , y = 0

=2
2
=
2
dt
x (t) dt x0 (t) dt
dt
dt
d 2y
4 2 . Substituting in the equation we get
dt
d 2y
dy
10t + 10y = 0.
2
dt
dt

dy
d 2y
0
2 dy
00
2 d
2 dy
t
= 2t3 + t4 2 , and the new
v) y = (t ) , y = (t )
dt
dt
dt
dt
dt
equation is
d 2y
dy
t2 2 + 5t + 4t3 y = 0.
dt
dt
1 dy 00
sin t dy
1
d 2y
vi) y 0 =
,y =

. We also have x2 1 =
cos t dt
cos3 t dt cos2 t dt2
d 2y
2
cos t and finally we get the equation 2 1 = 0.
dt
et et
et + et
vii) We remind that sinh t =
and cosh t =
. Similar
2
2
to the previous exercise, taking account that x0 = cosh t and x2 + 1 =
d 2y
sinh2 t + 1 = cosh2 t, we obtain the equation 2 y = 0.
dt
0
2 dy
viii) y = cos t ,
dt

dy
d 2y
d
2 dy
00
2
y = cos t
cos t
= 2 cos3 t sin t + cos4 t 2 .
dt
dt
dt
dt
4t2

Also 1 + x2 = 1 + tan2 t =
d 2y
dy
3 tan t + y = 0.
2
dt
dt
0

1
and the equation becomes
cos2 t

5.2 i) Substituting y (x) =


x0 = cos y(sin y x).
ii) x0 2xy = y y 3 .

dx
dy

1
=

1
x0 (y)

in the equation we get

54

1
d
1
1
d
1
x00 (y)
00
iii) y (x) = 0 , y (x) =
=

x (y)
dx x0 (y)
x0 (y) dy x0 (y)
(x0 (y))3
00
x (y)
1
1
and the new equation is 0
x 0
+ 0
sin y = 0, or
3
2
(x (y))
(x (y))
(x (y))2
x00 (y) + x0 (y)(sin y x) = 0.
iv) The transformed equation is x00 y = 0.
0

5.3 i) The relations between the cartesian coordinates (x, y) and the
polar coordinates (, ) are:
(
x = cos
y = sin .
Using the differentiation rule for composite functions for z(x(, ), y(, ))
we have:

z x z y
z

x y
z
z x z y

x y
or

z
z
z

=
cos +
sin

x
y
z
z
z

=
sin +
cos .

x
y
Solving the system, it follows:

z
z sin z

= cos

z
z cos z

= sin
+
.
y

Plugging these into the given expression, we get

z sin z
z cos z
E = cos cos

+ sin sin
+

z
z
z
= cos2
+ sin2
= .

ii) As in the previous exercise, we have the operators:

sin

() = cos ()
()
x

cos

() = sin () +
().
y

55
For the second order partial derivatives:



2z
z
z
sin z
=
= cos

=
x2
x x
x
x

z sin z
= cos
cos

sin
z sin z

cos
=

2 z sin z sin 2 z

= cos cos 2 + 2

sin
z
2z
cos z sin 2 z

sin
+ cos


2
2 z sin2 2 z 2 sin cos 2 z
= cos2 2 +

2 2

2 sin cos z sin2 z


+
+
2


and

2z

z cos z
= sin
sin
+

+
y 2

cos
z cos z
+
sin
+

2 z cos2 2 z 2 sin cos 2 z


= sin2 2 +
+

2 2

2 sin cos z cos2 z

+
.
2

So the laplacean expressed in polar coordinates is


z =

2z
1 2 z 1 z
+

+ .
2 2 2

5.4 i) Since z(x, y) = z(u(x, y), z(x, y)), we have:


z u z v
5 z z
z
=

=
+
,
x
u x v x
2 u v
z
z u z v
z z
=

=
+
,
y
u y v y
u v

56
and the second order partial derivatives

2z
2 z v
5 2 z u
2 z u 2 z v
=

=
x2
2 u2 x uv x
uv x v 2 x

5 2z 5
2z
5 2z
2z

+
+
+
=
=
2 u2 2 uv
2 uv v 2
25 2 z
2z
2z
=
2 +5
+ 2,
4 u
uv v
2z
2z
2z
2z
=
+
2
,
+
y 2
u2
uv v 2

2z
5 2 z u
2 z v
2 z u 2 z v
+
=

=
xy
2 u2 y uv y
uv y v 2 y
5 2z 7 2z
2z
= 2+
+ 2.
2 u
2 uv v
2z
Finally, substituting in the given equation we get
= 0.
uv
ii)
z
z z z
z
z
=
+
=
+2
x
u v y
u
v
2z
2z
2z
2z
=
+
2
+
,
x2
u2
uv v 2
2z
2z
2z
2z
=

4
+
4
,
y 2
u2
uv
v 2
2z
2z
2z
2z
= 2 +
+ 2 2,
xy
u
uv
v
The equation becomes 3
2z
iii)
= 0.
v 2
iv)

2z
z
+
= 0.
uv u

z
z
=y ,
x
u

z
z z
=x
+
,
y
u v

2
2z
2z
2 z
=
y

y
=
y

,
x2
u2
u2
2

2
z
2z
2z
2z
2z
2z
2z
2 z
=
x

x
+

x
+
=
x
+
2x
+
,
+
y 2
u2
uv
uv
v 2
u2
uv v 2

57
2z
z
=
+y
xy
u

2z
2z

x
+
u2
uv

= xy

2z
2z
z
+
y
+
.
u2
uv u

2z
z
The new equation is v 2 2 + 2v
= 0.
v
u
z
z z z
z
v)
= 2x
+ ,
= 2y
. Substituting into the equation we
x
u v x
u
z
z
z
z
get 2(x2 + y 2 )
+x
= 0, that is 2u
+v
= 0.
u
v
u
v
z
z
vi) u
3v
= 0.
u
v
2z
1 z
vii)

= 0.
uv 4v u
1
z0
5.5 i) y =
so y 0 = 2 . Substituting in the equation we get
z
z
z0
1
4
2
4
0
x( 2 ) + 2 = 2 , or z z = .
z
z
z
x
x
ii) y 0 = 2zz 0 , the equation becomes z 0 + z = x.
z0
iii) We have z 0 = cos y y 0 , so y 0 =
, and finally the equation is
cos y
z 0 z = ex .
iv) We have y = zx2 , so the derivatives are
y 0 = z 0 x2 2zx3 ,
y 00 = z 00 x2 2z 0 x3 2z 0 x3 + 6zx4 = z 00 x2 4z 0 x3 + 6zx4 .
Substituting these into the given equation it follows z 00 z = 4x.
z
w z
5.6 Since z(x, y) = ecy w(x, y), we have
= ecy
,
= cecy w +
x
x y
2
2
2
w

w w

z
ecy
and
= ecy 2 . The equation will be a2 2
= 0.
2
y
x
x
x
y
5.7 Considering the function y as a composite function, y = y(t(x)),
we get
1
dy dt
dy
dx
dy
=

=
dx
dt dx
dt
dt

1
2t
1
= 2
2
= 2t.
t +1
t +1

CHAPTER 6

Integral calculus for functions of one


real variable
6.1

Antiderivatives
Let [a, b] be a real interval and f : [a, b] R a function defined on it.
The function F : [a, b] R is called an antiderivative (or primitive)
of the function f if F is differentiable on [a, b] and F 0 (x) = f (x), for each
x [a, b].
The antiderivative of a function is not unique. If F : [a, b] R is
an antiderivative of f , then F + c is also an antiderivative of f , for any
constant c R.
If G is another antiderivative of f , then there exists a real constant c
such that G = F + c.
The set of all antiderivatives of a function f is called the indefinite
integral of f and is denoted
Z
f (x) dx.
Z
In fact, f (x) dx = {F (x) + c | c R}, where F is a fixed antiderivative
59

60
(primitive) of f .
Theorem 6.1 If f, g : [a, b] R are functions that admit antiderivatives
and , are real numbers, then:
Z
Z
Z
(f (x) + g(x)) dx = f (x) dx + g(x) dx.
the antiderivatives for some usual functions:
ZWe present below
n+1
x
1.
xn dx =
+ C, x R, n N,
n+1
Z
x+1
2.
x dx =
+ C, x (0, ), R \ {1},
+1
Z
1
3.
dx = ln |x| + C, x I R ,
x
Z
4.
ex dx = ex + C, x R,
Z
ax
x
5.
a dx =
+ C, x R, a > 0, a 6= 1,
ln a
Z
dx
1
x
6.
=
arctg
+ C, x R, a 6= 0,
2 + a2
x
a
a

Z
x a
dx
1
+ C, a 6= 0, x R, x 6= a, x 6= a,
7.
=
ln

2 a2
x
2a
x
+
a
Z

dx

= ln(x + x2 + a2 ) + C, x R, a 6= 0,
8.
2
2
x
+
a
Z

dx

9.
= ln |x + x2 a2 | + C, a > 0, x R \ [a, a],
2
2
Z x a
dx
x

10.
= arcsin + C, a > 0, x (a, a),
a
a 2 x2
Z
1
11.
sin ax dx = cos ax + C, x R, a R ,
a
Z
1
12.
cos ax dx = sin ax + C, x R, a R ,
a
Z
13.
tan x dx = ln | cos x| + C, x I, (2k+1)

/ I, k Z,
2
Z
14.
cot x dx = ln | sin x| + C, x I, k
/ I, k Z,
Z
dx
15.
= tg x + C, x I, (2k+1)

/ I, k Z,
2
2
Z cos x
dx
16.
= ctg x + C, x I, k
/ I, k Z.
sin2 x

61
6.2

Riemann integral on a real interval


Let [a, b] be a closed, bounded interval from R and f : [a, b] R a
function.
If the function f is continuous and has positive values, we would like
to determine the area of the surface between the curve y = f (x), the axis
0x and the straight lines x = a, y = b. To approximate the area of this
surface, it is divided in thin pieces like this: Consider a division of the
interval [a, b],
= {x0 , . . . , xn } with a = x0 < x1 < < xn = b,
and a system of intermediate points i [xi , xi+1 ]. The union of rectangles having the dimensions xi+1 xi and f (i ) approximates the initial
surface, so the area of the surface is approximated by the sum
f (, ) =

n1
X

f (i )(xi+1 xi ),

i=0

called the Riemann sum attached to the function f , division and


system .

f(x)

f (x i )

xi xi xi+1

The number || = max{xi+1 xi | i = 0, . . . , n1} is called the norm


of the division . The smaller || is, the better the approximation of
the area becomes. This is the basic idea in the definition of the integral.

62
The function f : [a, b] R is called (Riemann) integrable on [a, b] if
there exists the real number
I = lim f (, ),
||0

for each choice of the system of intermediate points {i }. The number I


(which is unique) is called the (Riemann) integral of the function f
on the interval [a, b] and is denoted
Z

f (x) dx
a

Theorem 6.2 If f, g : [a, b] R are integrable functions on [a, b] and


, are
Z breal numbers, then
Z b
Z b
f (x) dx +
g(x) dx,
(i) (f (x) + g(x)) dx =
aZ
a
a
Z a
b
(ii)
f (x) dx =
f (x) dx,
a

(iii) If f (x) g(x) for each x [a, b], then


Z

f (x) dx
a

g(x) dx
a

(iv) If f is continuous on [a, b], then there exists c [a, b] such that
Z

f (x) dx = f (c)(b a).


a

Theorem 6.3 Let c [a, b] and f : [a, b] R integrable on the intervals


[a, c] and [c, b]. Then f is integrable on [a, b] and
Z

f (x) dx =
a

f (x) dx +
a

f (x) dx.
c

The connection between Riemann integrals and antiderivatives is given


by the following theorem, called the fundamental formula in integral calculus.

63
Theorem 6.4 (Leibniz-Newton) If f : [a, b] R is an integrable function on [a, b], and it admits antiderivatives, then the following equality
holds:
Z b
b
not
f (x) dx = F (b) F (a) = F (x) ,
a

where F is an antiderivative of f .
We present in what follows the methods used to calculate Riemann
integrals: integration by parts and changes of variables.
Theorem 6.5 (integration by parts formula) If f, g : [a, b] R are
differentiable functions, with continuous derivatives, then the equality
holds:
Z b
Z b
b
0

f (x)g (x) dx = f (x)g(x)


f 0 (x)g(x) dx.
a

Theorem 6.6 Consider J an interval of R and the functions u : [a, b]


J, g : J R having the properties:
(i) g is continuous,
(ii) u is differentiable, with continuous derivative.
Let f = (g u) u0 . Then
Z

f (x) dx =
a

u(b)

g(u(x)) u (x) dx =
a

g(t) dt.

(6.2.1)

u(a)

In practice we denote t = u(x) and by differentiation we have u0 (x) dx =


dt.
Theorem 6.7 Consider I an interval of R and the functions f : [a, b]
R, u : I [a, b] having the properties:
(i) f is continuous,
(ii) u is bijective, u and u1 are of class C 1 and u0 (t) 6= 0 on I.
Then
Z
Z 1
b

(b)

f (x) dx =
a

f (u(t)) u0 (t) dt.

u1 (a)

In practice, we denote x = u(t) and we have dx = u0 (t) dt.

(6.2.2)

64
6.3

Exercises
6.1 Calculate the following integrals:
Z

i).

x ln x dx,
Z

2x

iii).

arctan x dx,
Z

e cos x dx,
e

v).
Z

1
3

vi).

/2
1
0

x).

/2

cos5 x dx,

iv).

ix).

0
0

ex
dx,
1 + e2x

viii).

xe dx,

ii).

vii).

ln x
dx,
x

xi).
Z

2xex dx,

1
0

dx

,
4 x2

x
dx.
1+ x
x2

xii).

dx
,
2 + sin x

x2
dx,
1 + x6

Solutions and answers


6.1 i) We integrate by parts:
Z

e
1

Z e

0
e Z e x2 1
x2
x2

x ln x dx =
ln x dx =
ln x
dx
2
2
x
1
1
1 2
Z e
2
2
2 e
2
2
e
x
e
x
e
1
e2 + 1
e
=

dx =
=
+ =
.
2
2
4 1
2
4
4
4
1 2

ii) Choosing f = x, g 0 = e2x , we have f 0 = 1, g =


Z

1
0

xe2x 1
xe dx =

2 0

2x

e2x
and
2

e2x
e2 e2x 1 e2 + 1
dx =

=
.
2
2
4 0
4

65
Z
iii)
Z

arctan x dx =
0
1

1
(x)0 arctan x dx = x arctan x0

1
1
x 2
dx = ln 2.
x +1
4 2
0
iv) We integrate by parts, two times:
Z 0
Z 0
x
I=
e cos x dx =
(ex )0 cos x dx =
/2

0
= ex cos x/2
Z

0
/2

/2

=1+e

/2
0

ex ( sin x) dx =

/2

0
(e ) sin x dx = 1 + e sin x/2
x 0

=1+

ex cos x dx =

/2

I.

1 + e/2
Finally I =
.
2
1
v) We denote t = ln x and we have dt = dx. The limits of the integral
x
Z 1
1
change x = 1 t = 0, x = e t = 1 and we get I =
t2 dt = .
3
0
Z 9
vi) t = x2 , dt = 2x dx, I =
et dt = e9 e4 .
4

vii) t = ex , I = arctan e .
4
viii) We write the integral in the form
Z
Z
4
I=
cos x sin x dx =
(1 sin2 x)2 cos x dx
/2

/2

and we denote t = sin x, with dt = cos x dx, so I =


1

(1t2 )2 dt =

ix) t = x3 and dt = 3x2 dx; x = 0 t = 0, x = 1 t = 1. Then


Z
Z
1

1 1 3x2
1 1 dt
1

I=
dx =
dt = ln(t + 1 + t )
6
2
3 0
3 0
3
0
1+x
1+t

1
1
= [ln(1 + 2) ln 1] = ln(1 + 2).
3
3
x) Using trigonometric formulas, we write the integral in the form:
Z
Z
2
2
1 + tg2 x2
dx
dx.
=
x
2 x
2 tg x2
0 2+
0 2 tg 2 + 2 tg 2 + 2
2 x
1+tg

7
.
15

66
x
1
x
We make the substitution t = tg and get dt = (1 + tg2 ) dx;
2
2
2

x = 0 t = 0, x =
t = 1.
2
Z 1
Z 1
Z 1
dt
dt
dt
=
dt
=
I=
=

2
1
3
1
2
2
0 t + 22t + 4 + 4
0
0 t +t+1
t + 21 + 34

2
1 1
2
1
2
= arctan (t + ) = [arctan 3 arctan ] =
2 0
3
3
3
3
2

= ( )= .
6
3 3
3 3
1
xi) Denote x = 2 sin t, with dx = 2 cos t dt; x = 1 sin t = and
2

t = ; x = 2 t = . We have:
6
4
Z
Z
Z
4
4
2 cos t dt
2 cos t
1 4 1
p
I4 =
=
dt =
dt

4 6 sin2 t
4 sin2 t 2 cos t
4 sin2 t 4 4 sin2 t
6
6

31
1
1
4
.
= cot t = [cot cot ] =
4
4
4
6
4
6
Z

xii) x = t , dx = 2t dt, I =
0

t
2t dt = 2 ln 2 1.
1+t

CHAPTER 7

Line integrals
7.1

Line integral with respect to arc length


Let : [a, b] R2 be a (piecewise) smooth, plane curve, with A = (a),
B = (b).
Let f : D R be a continuous function, where D R2 is such that
([a, b]) D.
_
: A = A0 , A1 , . . . , An = B
Consider a division of the arc AB,
and a system of intermediate points Mi (i , i ) with Mi A\
i Ai+1 for each
\
i = 0, . . . , n 1. Denote by l(A\
i Ai+1 ) the length of the curve arc Ai Ai+1 .
Consider the integral sum
{Mi }) =
f (,

n1
X

f (i , i )l(A\
i Ai+1 ).

i=0

The function f admits a line integral with respect to arc length


(or of the first type) along if the limit
{Mi }),
I = lim f (,

||0

exists in R, for each choice of the system of intermediate points {Mi }.


67

68
The integral is denoted
Z
Z
I = f (x, y) ds (or
f ds)

In the same way can be defined the line integral with respect to arc length
of a function f : D R, with D R3 , along a curve in space and is
denoted
Z
Z
I = f (x, y, z) ds (or
f ds)

A line integral with respect to arc length can be calculated by reducing


it to a Riemann integral on a real interval.
Theorem 7.1 Let f : D R, D R2 be a continuous function and
: [a, b] R2 a (piecewise) smooth curve, with ([a, b]) D, defined
parametrically by (t) = (x(t), y(t)), t [a, b]. Then
Z b
Z
p
f (x(t), y(t)) x02 (t) + y 02 (t) dt.
f (x, y) ds =
(7.1.1)
a

A similar formula holds for line integrals along curves in space:


Z

f (x, y, z) ds =

p
f (x(t), y(t), z(t)) x02 (t) + y 02 (t) + z 02 (t) dt.

(7.1.2)
In the case when the curve is given explicitly, the following is true:
Theorem 7.2 If f : D R, D R2 is a continuous function and
: [a, b] R2 is a (piecewise) smooth curve, represented by the equation
y = y(x), x [a, b], then:
Z
Z b
p
f (x, y) ds =
f (x, y(x)) 1 + y 02 (x) dx.
(7.1.3)

7.2

Applications of the line integral with respect to arc


length

I. Length of a curve

69
If is a smooth curve, plane or in space, then it has the length
Z
l() = ds.

II. The mass and the mass center of a thin wire


Consider a thin wire, represented by a smooth curve : [a, b] R3
(or R2 ). The wire is supposed to be nonhomogeneous, with the density
given by the function, : R3 R (that is, at a point P (x, y, z) of the
wire, the density is (x, y, z)). The mass is given by the formula:
Z
M () = (x, y, z) ds.

The coordinates xG , yG , zG of the mass center are given by the formulas:


Z
Z
1
1
x(x, y, z) ds, yG =
y(x, y, z) ds,
xG =
M
M
1
zG =
M

Z
z(x, y, z) ds,

III. The inertia moments of a thin wire


The inertia moments with respect to the coordinate planes of a wire
with the density are:
Z
Z
Z
2
2
Ix0y = z (x, y, z) ds, Iy0z = x (x, y, z) ds, Ix0z = y 2 (x, y, z) ds.

The inertia moments with respect to the coordinate axes are


I0x = Ix0y + Ix0z , I0y = Ix0y + Iy0z , I0z = Ix0z + Iy0z .
The inertia moment with respect to the origin is I0 = Ix0y + Ix0z + Iy0z .
The inertia moment with respect to the point P (x0 , y0 , z0 ) can be
determined by the formula
Z
IP0 = [(x x0 )2 + (y y0 )2 + (z z0 )2 ](x, y, z) ds.

70
7.3

Line integral with respect to the coordinates


Let : [a, b] R2 be a plane, (piecewise) smooth curve, with A = (a),
B = (b).
Let F~ : D R2 be a vector valued function, having the components
P, Q : D R, where D R2 is such that ([a, b]) D:
F~ (x, y) = P (x, y)~i + Q(x, y)~j

(F~ = (P, Q)).

: A = A0 , A1 , . . . , An = B,
Consider a division of the arc AB,
Ai (xi , yi ); a system of intermediate points Mi (i , i ) A\
i Ai+1 for each
i = 0, . . . , n 1 and the sum
{Mi }) =
Sf (,
=

n1
X
i=0
n1
X

P (i , i )(xi+1 xi ) + Q(i , i )(yi+1 yi )


F~ (i , i ) (~ri+1 r~i ),

i=0

where r~i = (i , i ) is the position vector of the point Ai and denotes


the inner product between the two vectors.
The function F~ admits a line integral with respect to the coordinates (of the second type) along the curve if there exists a real
{Mi }), for any choice of the intermediate
number I, I = lim Sf (,

||0

points system {Mi }.


The integral is denoted
Z
I = P (x, y) dx + Q(x, y) dy

Z
F~ d~r)

(or

In the same way is defined the line integral with respect to coordinates
of a function F~ : D R3 , D R3 , F~ = P~i + Q~j + R~k along a curve
in space:
Z
P (x, y, z) dx + Q(x, y, z) dy + R(x, y, z) dz

71
The line integral with respect to coordinates can be calculated by reducing it to a Riemann integral on a real interval.
Theorem 7.3 Let F~ = P~i + Q~j : D R2 be a function with continuous
components, with D R2 and : [a, b] R2 a (piecewise) smooth curve,
with ([a, b]) D, defined by (t) = (x(t), y(t)), t [a, b]. The following
formula holds
Z

P (x, y) dx + Q(x, y) dy =

[P (x(t), y(t))x0 (t) + Q(x(t), y(t))y 0 (t)] dt.

(7.3.1)

Likewise, for an integral along a curve in space, (t) = (x(t), y(t), z(t)),
t [a, b], it holds:
Z
P (x, y, z) dx + Q(x, y, z) dy + R(x, y, z) dz =

[P (x(t), y(t), z(t))x0 (t) + Q(x(t), y(t), z(t))y 0 (t)

(7.3.2)

+ R(x(t), y(t), z(t))z 0 (t)] dt.


In the case when is a plane curve, represented explicitly, it holds:
Theorem 7.4 If F~ : D R2 , D R2 is a continuous function and
is a (piecewise) smooth curve, represented by the equation y = y(x),
x [a, b] then the following holds:
Z

P (x, y) dx + Q(x, y) dy =

[P (x, y(x)) + Q(x, y(x)) y 0 (x)] dx.

The work of a force


Consider a point of mass m, moving along a smooth curve : [a, b]
3
R under the action of a force F~ = P~i + Q~j + R~k : D R3 , D R3 .
The work made by the force F~ in the action is
Z
Z
~
L = F d~r = P (x, y, z) dx + Q(x, y, z) dy + R(x, y, z) dz.

72
7.4

Exercises
7.1 Calculate the following line integrals with respect to arc length,
along the given curves:
Z
i).

(y + z) ds, where is the line segment [AB], A(1, 3, 1) and

B(1, 2, 2),
Z
ii). (3 x)ey1 ds, where is the line segment [AB], with A(3, 1)

iii).
iv).

v).

vi).

and B(1, 3),


Z
xz
ds, with (t) = (ln t, t2 , 2t), t [1, e],
2y
+
1

Z
y
ds, where is formed by the parabola y = x2 and the straight
x

line y = x, situated in the first quadrant,


Z
x2
xy ds, is the part of the ellipse
+ y 2 = 1 from the first
4

quadrant,
Z
y ds, where is the astroid arc x2/3 + y 2/3 = a2/3 from the third

quadrant,
Z p
vii).
z 1 x2 y 2 ds, where is the intersection circle between the

sphere x2 + y 2 + z 2 = 1 and the plane x = y,


Z
viii). (x + y + z) ds, is the intersection circle between the paraboloid

z = x2 + y 2 and the sphere x2 + y 2 + z 2 = 2.


7.2 Determine the length of the arc given by the parametric equations
x = et cos t, y = et sin t, z = et , t [0, 1].

73
7.3 Determine the mass
and the coordinates of the mass center for the

x = 3 cos t,
helix arc represented by y = 3 sin t, , with t [0, 2] having the density

z = 4t
(x, y, z) = 8 z.
7.4 Determine the mass and the coordinates of the mass center for a
thin wire represented by the equation y = 2x + 1, x [0, 1] and having
the density (x, y) = x + y.
7.5 Calculate the following line integrals with respect to coordinates,
along the given curves:
Z
i).
zy dx + 2x dy y dz, where (t) = (t3 , 3t, t2 + 2), t [1, 1],
Z

y dx + x2 dy, where is the ellipse arc 9x2 + y 2 1 = 0, y 0,

ii).

in direct trigonometric orientation,

2
2
2

(y 2) = x + z

iii). (x+y) dx+ xz dy z dz, along the curve y = 1


,

x 0, z 0
in the sense of z increasing,
Z
z dx + xy dz, where is the intersection between the plane y = z
iv).
Z

and the cylinder x2 + y 2 = 1, with x 0, y 0 and z 0,


Z
dx
v).
+ x dy, where is the parabola arc y 2 = x, x [0, 1],
4
1
+
y

Z
x
vi).
F~ d~r, where F~ = ~i + ~j, and A(1, 2), B(2, 1),
y
[AB]
Z
dx + dy
vii).
, where is the curve y = x2 x + 1, x [1, 1].
x
+
2

Solutions and answers

74
7.1 i) We write first the parametric equations of the segment [AB].
From
y3
z1
x1
=
=
= t, t [0, 1]
1 1
23
21
we have

x(t)
=
2t
+
1

x (t) = 2
and
y(t) = t + 3
y 0 (t) = 1

0
z(t) = t + 1
z (t) = 1
and

Z 1
I=
(t 3 + t + 1) 6 dt = 6
(2t 2) dt
0
0
1
2

= 6(t 2t) = 6(1 2) = 6.

ii) We have x = 2t + 3, y = 2t + 1, with t [0, 1] and I = 2(e2 + 1).


2t2 + 1
iii) ds =
dt, I = 2.
t
iv) We will have the sum of two integrals, one along the curve 1 : y = x2 ,
x [0, 1] and the other along 2 : y = x, x [0, 1]. For 1 the arc element

is ds = 1 + 4x2 dx and for 2 , ds = 2 dx.


Z
Z 1 2
Z 1
x
y
x 4x2 + 1 dx.
ds =
4x2 + 1 dx =
0 x
0
1 x
Performing the change of variable u = 4x2 + 1, du = 8x dx, we have
Z
Z
Z
3
y
1 5
1 u 2 5
1 5 1/2
1
ds =
u du = 3 = ( 125 1).
u du =
8 1
8 1
8 2 1 12
1 x
For the second integral,
Z
Z 1
1
y
x
ds =
2 dx = 2x = 2.
0
2 x
0 x

1
Finally, I = 2 + ( 125 1).
12
v) The parametric equations of the curve are x = 2 cos t, y = sin t,

t [0, ]. Then ds = 3 cos2 t + 1 dt and


2
Z 1
Z

4 14
2
1
1 2
2 cos t sin t 3 cos2 t + 1 dt =

u du = u3/2 1 =
I=
3
3 3
9
0
4

75
(in the last integral we made the change of variable u = 3 cos2 +1, with
du = 6 cos t sin t dt).
vi) The parametric representation of the astroid is:
(
x(t) = a cos3 t,
3
, t [, ]
3
2
y(t) = a sin t,
We have x0 (t) = 3a cos2 t sin t, y 0 (t) = 3a sin2 t cos t and so
p
ds = 9a2 cos4 t sin2 t + 9a2 sin4 t cos2 t dt =
3
= 3a cos t sin t dt, for t [, ]
2
Then
Z

3/2

I=

Z
3

3/2

a sin t 3a cos t sin t dt = 3a

sin4 t cos t dt

We make the change of variable u = sin t, du = cos t dt and obtain


Z 0
Z 1
u5 0
3a2
4
2
u4 du = 3a2 =
u du = 3a
I = 3a
.
5 1
5
1
0
vii) Substituting y by x it follows 2x2 + z 2
this:

x = cos t

2
1
y = cos t with t [0, 2]

z = sin t

= 1. We can parametrize like

x0 = sin t

2
1
0
y = sin t

z 0 = cos t

O
x

76
r

1 2
1
sin t + sin2 t + cos2 t dt = dt and
2
2
r
Z 2
Z 2

1
1
2
2
I=
sin t 1 cos t cos t dt =
sin t 1 cos2 dt
2
2
Z0 2
Z
Z 20
=
sin t | sin t| dt =
sin2 t dt
sin2 t dt

Z0
Z 20
1 cos 2t
1 cos 2t
=
dt
dt
2
2
0

1
sin 2t
sin 2t 2 1
1
1
1
t
t
=

= 2 + = 0.
2
4
2
4
2
2
2
0

Then ds =

viii) Intersecting the surfaces we get z = 1 and the parametric form of


the curve x = cos t, y = sin t, z = 1, with t [0, 2]. We find
Z

I=

(cos t + sin t + 1) dt = 2 2.

7.2 x0 (t) = et cos t et sin t, y 0 (t) = et sin t + et cos t, z 0 (t) = et ,


Z
Z 1

l = ds =
et 3 dt = 3(e 1).

7.3 We have, for the curve ,


p
ds = (3 sin t)2 + (3 cos2 t) + 42 dt = 5 dt, so
Z

M=

(8 z) ds =

2
t2 2

(8 4t)5 dt = 40t 20 = 40 2 .
2 0
0

For the coordinates of the mass center, we calculate:


Z
Z 2
1
1
xG =
x(8 z) ds =
3 cos t(8 4t)5 dt
M
M 0
Z 2
Z 2
1
[120
cos t dt 60
t cos t dt] =
=
M
0
0
Z 2
2
1

=
t(sin t)0 dt] =
[120 sin t 60
M
0
Z0 2

2
1

=
sin t dt] = 0.
[60t sin t + 60
M
0
0

77
The other coordinates are obtained similarly: yG =

3
,

zG =

8
.
3

5 5
3
11
7.4 M =
, xG = , y G = .
2
5
5
Z

7.5 i) I =

Z
2

[(t +2)3t3t +2t 33t2t] dt =


1

(9t5 +24t3 6t2 ) dt =

4,
ii) We write first the parametric equations of the curve

x = 1 cos t
, t [0, ].
3
y = sin t
1
We have x0 = sin t, y 0 = cos t, so
3
Z

1
1
sin t( sin t) + cos2 t cos t dt
3
9
0
Z
Z
1
1
=
sin2 t dt +
(1 sin2 t) cos t dt
3 0
9 0
Z
Z
Z
1 2
1 1 cos 2t
1
cos t dt
sin t cos t dt
=
dt +
3 0
2
9 0
9 0

1 sin3 t
1 1
sin 2t 1

=
t
+ sin t
= .
3 2
4
9
9 3 0
6
0
0

I=

iii) We write the parametric equations of the curve (the intersection


between a plane and
a cone). Replacing y = 1 in the first equation we

x = cos t
2
2
get x + z = 1, so y = 1
, with t [0, 2 ] (the interval [0, 2 ] is given

z = sin t
by the conditions x 0, z 0 and the fact that z has to increase.)

78
z

We have x0 = sin t, y 0 = 0, z 0 = cos t and


Z

(x + y) dx + xz dy z dz =

2 cos t sin t dt
sin t dt =
0

cos 2t 2
2
=
cos t = 2.
2 0
0
0

sin 2t dt

sin t dt

iv) The curve is written with parametric equations x = cos t, y = sin t,

1
z = sin t, t [0. ]. We obtain I = .
3 4

Z 1 2
1 1 2
1 1
1
1
v) I =
dx = arctan x0 + x3/2 0 = + .
+x
2
x +1
2 3
4 3
2 x
0
vi) 15
2 10.
Z 1
Z 1
Z ln
1
2x + 1
3
dx + 2x dx
=
dx =
(2
) dx = 4 3 ln 3.
vii)
x+2
x+2
1 x + 2
1
1

CHAPTER 8

Double integral
8.1

Calculus by iteration
Let D R2 be a bounded, measurable set. Let = {D1 , . . . , Dn }
be a division of the set D and = {1 , . . . , n }, Di a system of
intermediate points of .
Let f : D R be a function and consider the Riemann sum
n
X
f (, ) =
f (i )area(Di ) associated to f , and .
i=1

It is said that the function f is Riemann integrable on D if there


exists the real number I,
I = lim f (, ),
||0

for any choice of the system of points .


The real number I defined like this is unique and it is called the
(Riemann) integral of f on the set D. It is denoted
ZZ
f (x, y) dx dy
D

79

80
Theorem 8.1 If D is a bounded, measurable, closed set and the function
f : D R is continuous on D, then f is Riemann integrable on D.
A double integral can be calculated by decomposing it into two simple
integrals. We present three cases when the iteration is possible.
Case I: rectangular domain
Consider the set D of the form D = {(x, y) | a x b; c y d} =
[a, b] [c, d].
y

d
D
c

Theorem 8.2 Let D = [a, b] [c, d] and f : D R, integrable on D.


Z d
If for each x [a, b] there exists h(x) =
f (x, y) dy, then the function
c

h : [a, b] R is integrable on [a, b] and we have

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

(8.1.1)

This theoremZisZ a particular case Zof Fubinis


Theorem.
Likewise, the
d Z b
following holds:
f (x, y) dx dy =
f (x, y) dx dy.
D

Case II: simple domain with respect to 0y


If D is of the form
D = {(x, y) R2 | a x b; 1 (x) y 2 (x)}
where 1 , 2 : [a, b] R are continuous functions, D is called simple
with respect to 0y.

81

y=j2(x)

DD

y=j1(x)

Theorem 8.3 If f is integrable on D, simple with respect to 0y, and for


Z 2 (x)
each x [a, b] there exists h(x) =
f (x, y) dy, then the function
1 (x)

h : [a, b] R is integrable on [a, b] and


ZZ
Z "Z
b

f (x, y) dx dy =
D

f (x, y) dy dx.
a

f (x, y) dx dy =
D

(8.1.2)

1 (x)

ZZ
(It is also written

2 (x)

2 (x)

dx
a

f (x, y) dy.)
1 (x)

Case III: simple domain with respect to 0x


If D is of the form
D = {(x, y) R2 | c y d, 1 (y) x 2 (y)}
where 1 , 2 : [c, d] R are continuous functions, D is called simple
with respect to 0x.

y
.

d
x=y1(y)

x=y2(y)
D

c
O

82
Theorem 8.4 If f is integrable on D and for each y [c, d] there exists
Z 2 (y)
h(y) =
f (x, y) dx, then the function h : [c, d] R is integrable on
1 (y)

[c, d] and we have


ZZ

Z
f (x, y) dx dy =
D

2 (y)

dy
c

f (x, y) dx.

(8.1.3)

1 (y)

8.2

Changes of variables for the double integral


Let D and A be two bounded, measurable, closed sets from R2 .
Let : A D be a bijective function. If = (1 , 2 ), to each point
(x, y) from D corresponds a unique point (u, v) from A (and conversely,
too),
(
x = 1 (u, v)
.
y = 2 (u, v)
The determinant

D(1 , 2 )
u (u, v)
(u, v) = J (u, v) =
2 (u, v)
D(u, v)

1
(u, v)
v

2
(u, v)
v

is called the Jacobian or functional determinant of at the point (u, v).


Theorem 8.5 Let D, A R2 be bounded, measurable, closed sets. Let
: A D be a continuous bijection, with continuous partial derivatives
D(1 , 2 )
of order I and II, with
6= 0 on A. If f : D R is integrable
D(u, v)
on D, then

ZZ
ZZ
D(1 , 2 )
du dv.
f (x, y) dx dy =
f (1 (u, v), 2 (u, v))
D(u, v)
D
A

Polar coordinates

83
In some situations, it is useful to change the variables from cartesian
coordinates into polar coordinates.
The relations between the cartesian coordinates (x, y) and the polar
coordinates (, ) are
(
x = cos
with [0, +), [0, 2].
y = sin ,
( is the polar radius and the polar angle).
y

q
O

The Jacobian for this change of variables is

D(x, y) cos sin


=
= cos2 + sin2 = .
D(, ) sin cos
So the formula becomes:
ZZ
ZZ
f (x, y) dx dy =
f ( cos , sin ) d d.
D

(8.2.1)

Another frequently used change of variables uses generalized polar


coordinates:
(
x = a cos
D(x, y)
, with
= ab.
D(, )
y = b sin
8.3

Applications of the double integral


I. The area of a bounded, plane domain
ZZ
areaD =
dx dy.
D

84
II. The mass and the mass center of a plane domain
Consider a plane domain, having the shape of the bounded, measurable set D R2 and the density at each point (x, y) given by the function
: D R. Denoting with M the mass of the plane domain and with
xG , yG the coordinates of the mass center, we have:
ZZ
M=
(x, y) dx dy,
D

1
xG =
M

ZZ
x(x, y) dx dy,
D

1
yG =
M

ZZ
y(x, y) dx dy.
D

III. The inertia moments of a plane domain with respect to the


axes:
ZZ
ZZ
2
I0x =
y (x, y) dx dy, I0y =
x2 (x, y) dx dy.
D

With respect to the origin: I0 = I0x + I0y .


IV. The volume of a cylindric body
Let f : D R+ , D R2 a continuous function on D.
The body V = {(x, y, z) R3 | 0 z f (x, y), (x, y) D} is
a cylindric body having the generators parallel to the axis 0z, bounded
above by the surface z = f (x, y) and below by the plane x0y. Its volume
is given by
ZZ
vol(V ) =

f (x, y) dx dy.
D

8.4

Exercises
8.1 Calculate the following double integrals, using the iteration method.
ZZ
(x2 + y) dx dy, D = [1, 1] [0, 2],

i).
ZZ

ii).
D

x
3
dx dy, D = [1, 0] [ , ],
2
cos y
4

85
ZZ

iii).
ZZ

1
dx dy, D = [0, 1] [1, 2],
x+y

y
dx dy, D is the domain bounded by the parabola y 2 = 2x
x
+
1
D
and the straight line y = x,
ZZ
dx dy
v).
, where D is the domain bounded by the coordi2
D (3 + x y)
nate axis Ox and the straight lines x = 1, x = 2, y = x + 1,
ZZ
y
vi).
dx dy, where D is the domain bounded by the para2
D x +x+1
bolas y = x2 and x = y 2 ,
ZZ
vii).
sin(x + y) dx dy, D = {(x, y) R2 | x 0, y , y 2x},
iv).

ZZ

x dx dy, D = {(x, y) R2 | x2 + y 2 2, y x2 },

viii).
ZZ

ex+y dx dy, where D is the interior of the triangle AOB, with

ix).
D

A(1, 1) and B(1, 1),


ZZ
x).
xy dx dy, D = {(x, y) R2 | x2 + y 2 2x, x2 + y 2 1, y 0}.
ZZ

xi).

xy dx dy, where D is the domain bounded by the parabolas


D

y 2 = x, y = 2 x2 and the axis 0x.


ZZ
1
xii).
dx dy, where D is given by D = {(x, y) | x + y 2; x
D x+y
0; y 1}.
8.2 Using a change of variables, calculate the following double integrals:
ZZ
dx dy
p
i).
, where D = {(x, y) R2 | 1 x2 + y 2 9, x
2
2
x +y
D
0, y 0},
ZZ
ii).
y dx dy, where D = {(x, y) R2 | 6x x2 + y 2 8x, y 0},
ZZ

ln(x2 + y 2 ) dx dy, where D is given by e x2 + y 2 e2 ,

iii).
D

86
ZZ
iv).
ZZ

dx dy
p
, D : x2 + y 2 a2 , y x.
2
2
1+x +y
x2 y 2 dx dy, D : x2 + y 2 1, x 0, y 0

v).
ZZ

vi).
ZZ

vii).

x2

y
dx dy, D : x2 + y 2 2y, x 0,
+ y2

x dx dy, where D is the disc with the center at the point


D

A(1, 2), of radius 2,


ZZ
viii).
(x2 y + xy 2 ) dx dy, where D is the domain bounded by the hyD

perbolas xy = 1, xy = 2 and the straight lines xy = 1, xy = 1.


8.3 Determine the area of the plane domain D = {(x, y) R2 | x2 +y 2
2y, x2 + y 2 2x}.
8.4 Determine the area of the plane domain bounded by the ellipse
x2 y 2
+ 2 = 1.
a2
b
8.5 Determine the mass and the mass center for the domain bounded
by the curve y = sin x, x [0, ] and the axis 0x, knowing that it has
the density (x, y) = y.
8.6 Determine the coordinates of the mass center for a plane domain
with the density and having the shape of a quarter of a circle of radius
R.

Solutions and answers


Z

8.1 i) We have, according to Theorem 8.2, I =


Z

(x + y) dy dx.
1

y 2 2
We calculate first
(x + y) dy = x y + = 2x2 + 2. Further on,
2 0
0
Z 1
1
3
2x
16

I=
(2x2 + 2) dx =
+ 2x = .
3
3
1
1
2

87
The same integral can be calculated in reverse order:
Z

I=

(x + y) dx dy =
Z

=
0

2
2
2

( + 2y) dy = y + y 2
3
3
0

1
x3

( + yx) dy
3
1
0
16
= .
3
2

1
.
2

4
iii) (3 3 2 2).
3
iv) The intersection points between the parabola and the straight line
are O(0, 0) and A(2, 2). D is simple with respect to 0y, that is 0 x 2

and x y 2x.
ii)

2x=y2
y=x

Z
We have I =

dx
0

Z
x

2x

2x

1
y dy. We calculate first
x+1

1
1 y 2 2x
1
y dy =
=
(2x x2 ).

x+1
x+1 2 x
2x + 2

From here it follows


Z
Z 2
1 2
3
1 x2 + 2x

dx =
(x + 3
) dx
I=
x+1
2 0
x+1
0 2

1
x
2
=
+ 3x 3 ln(x + 1) = 2 ln 3 3.
2
2
0

88
Z

Z 2
x+1
dy
1

v)
dx
=
dx
=

2
(3 + x y)
3+xy 0
0
1
Z 11
1
1
1
4
) dx = + ln .
=
(
3+x
2
5
0 2
Z 1 Z x
Z 1
y
1
x x4
vi) I =
dx
dy
=

dx =
2
x2 + x + 1
2
x2
0 x +x+1
Z 1 0
1
1
=
(x x2 ) dx = .
2 0
12
2
vii) .
3Z
Z 2x2
Z 1
1
viii)
dx
x dy =
x( 2 x2 x2 ) dx = 0.
2

x+1

x2

ix) The equations of the three sides of the triangle are AO : y = x,


BO : y = x, AB : y = 1.
y

-1

We decompose the integral in simple integrals:


Z

e e dy +
dx
ex ey dy =
1
x
0
x
Z 0
Z 1
1
1

=
dx ex ey +
dx ex ey =
x
x
1
0
Z 0
Z 1
Z 0
x
x
x
x
=
e (e e ) dx +
e (e e ) dx =
(ex+1 1) dx+
1
0
1
Z 1
0
e2x 1 e2 3

) =
.
+
(ex+1 e2x ) dx = (ex+1 x) + (ex+1
2 0
2
1
0

I=

x)

5
.
48

dx

x y

89
xi) The domain D is simple with respect to 0x. We determine the intersection points between the two parabolas by substituting x = y 2 in
y = 2 x2 and get the equation y 4 + y 2 = 0. The only convenient
solution of this is y = 1, so the intersection point is A(1, 1).

y
2

1
2

y =x

y=2-x

We decompose the integral using formula (8.1.3):


Z

x2 2y
I=
dy
xy dx =
dy y
2 y2
y2
0
0

Z 1
y
1
y 3 y 6 1 1
4
2
=
(2 y y ) dy =
y

= .
2
3
6
4
0
0 2
1

2y

xii) We have

y
2
y=2-x
1

90
Z

2y Z 2

dy
I=
dy ln(x + y)
=
(ln 2 ln y) dy
0
1
0
1
1
Z 2
2 Z 2
2

= y ln 2
ln y dy = ln 2
y 0 ln y dy = ln 2 y ln y
1
1
1
1
Z 2

1
2
+
y dy = ln 2 2 ln 2 + y = 1 ln 2.
y
1
1
2

2y

1
dx =
x+y

8.2 i) The domain is a part from a circular annulus, situated in the


first quadrant.
y

3
D
1
O

We pass to polar coordinates: x = cos , y = sin .


From the condition 1 x2 + y 2 9, we get 1 2 9, that is [1, 3].

From x 0 and y 0 we have cos 0 and sin 0, That is [0, ].


2

So the new domain is A = [1, 3] [0, ]. It follows


2
Z 3 Z
ZZ
2
1
p

d
d
=
d
d
I=
2 cos2 + 2 sin2
A
1
0
Z 3

3
=
d = = .
2
1
1 2
ii) x2 + y 2 = 6x can be written in the form (x 3)2 + y 2 = 9 and
is the equation of the circle centered at A(3, 0), of radius 3. Likewise,
x2 + y 2 = 8x represents the circle with the center at B(4, 0), of radius 4.
The set D is situated outside the first circle and inside the second one,
under the axis 0x.

91
y

. .
3 4

x
D

We pass to polar coordinates.Using the condition 6x x2 + y 2 8x we


have 6 cos 2 8 cos and dividing by > 0,
6 cos 8 cos .
This implies also cos 0, which together with y 0 takes to
3
[ , 2]. It follows
2
Z 2
Z 2 Z 8 cos
3 8 cos
d
sin d =
d sin
I=
=
3
3
3 6 cos
6 cos
2
2

Z 2
296
296
cos4 2
74
3
sin
cos d =

=
3 = .
3
3
3
4
3
2
2
iii) e2 .

iv) ( 1 + a2 1).
v) In polar coordinates, we get
Z 1 Z
Z
2
2
2
2
d
cos sin d =

6 1

6 0

d
0

Z
5

sin2 2
d =
4

1 cos 4

d = .
8
96

vi) [0, ], [0, 2 sin ] and I = .


2
2
vii) The equation of the circle is (x 1)2 + (y 2)2 = 4. We pass to
generalized
Z polar
Z coordinates: x = cos + 1, y = sin + 2 and obtain
2

that I =

d
0

( cos + 1) d = 4.
0

92
(
viii) We make the change of variables

u = xy
. It is clear that
v =xy

u [1, 2] and v [1, 1].


y

x-y=-1
x-y=1

xy=2
xy=1

D(x, y)
To determine the Jacobian, we take account that
=
D(u, v)
where

u u

D(u, v) x y y x
=
=
= y x
D(x, y) v v 1 1

x y

D(u, v)
D(x, y)

1
,

D(x, y)
1
=
. We have:
D(u, v)
x+y
ZZ
ZZ
1
xy(x + y) dx dy =
u (x(u, v) + y(u, v))
du dv
x(u, v) + y(u, v)
D
A
ZZ
Z 2 Z 1
Z 2
Z 2
1

=
u du dv =
du
u dv =
du uv =
2u du = 3.

So

8.3 Passing to polar coordinates, 2 cos and 2 sin . From


here sin 0 and cos sin , so [ , ]. For [ , ] we have
4
4 2

[2 cos , 2 sin ] and for [ , ], [0, 2 sin ]. So,


2
Z
Z 2 sin
Z Z 2 sin
2

area(D) =
d
d +
d
d = + 1.

2
2 cos
0
4
2

93
ZZ
8.4 The area is given by the formula aria(D) =

dx dy. We pass
D

to generalized polar coordinates x = a cos , y = b sin .


x2 y 2
From 2 + 2 1 we get 2 1, that is [0, 1]. For the polar angle,
a
b
[0, 2].
Z

area(D) =

d
0

ab d =
0

1
2

ab 2 d = ab = ab.
0

8.5
ZZ

y 2 sin x
M=
y dx dy =
dx
y dy =
dx
2 0
D
0
0
0

Z
Z
2
sin x
1 cos 2x
x sin 2x
=
dx =
dx =

= .
2
4
4
8
4
0
0
0

sin x

y
1

p
2

For the coordinates of the mass center we have


ZZ
Z Z sin x
Z
1
1
sin2 x
1
xy dy =
xG =
dx
dx x
xy dx dy =
M
M 0
M 0
2
0
Z

Z D
Z
1
1 cos 2x
1
x
1
x
dx =
dx
x cos 2x dx
=
2M 0
2
2M 0 2
2 0
2
The first integral has the value
; for the second one we integrate by
4
parts:
0
Z
Z
Z
sin 2x
x sin 2x
sin 2x
x cos 2x dx =
x
dx =
dx

2
2
2
0
0
0
0
cos 2x
=
= 0.
4 0

94
1 2

So xG =

= (which was expected, due to the symmetry of the


2M 4
2
domain).
yG =
=
=
=

ZZ
Z
Z sin x
Z
1
1
1
sin3 x
2
2
y dx dy =
dx
y dy =
dx
M
M 0
M 0
3
D
0
Z
Z
1
1
2
sin x sin x dx =
(1 cos2 x) sin x dx
3M 0
3M 0
Z
Z
1
1
sin x dx +
cos2 x(cos x)0 dx
3M 0
3M 0

1 cos3 x
16
1

( cos x) +
.
=
3M
3M 3 0
9
0

8.6 We choose the coordinate axes such that the center of the circle is
at the origin and the domain is situated in the first quadrant. The mass
R2
4R
and xG = yG =
.
is M =
4
3

CHAPTER 9

Triple integral
9.1

Calculus by iteration
The triple integral is defined in a similar way like the double one.
Let V R3 be a bounded, measurable set.
n
X
Let f : V R be a function and f (, ) =
f (i )vol(Vi ) the
i=1

Riemann sum associated to the function f , division and intermediate


point system .
The function f is called Riemann integrable on V of there exists the
real number
I = lim f (, ),
||0

for each choice of the system .


The number I defined like this is unique, is called the Riemann
integral on V and is denoted
ZZZ
f (x, y, z) dx dy dz
V

A triple integral can be calculated by decomposing it into a double


integral and a simple one.
95

96
Case I: rectangular domain
Theorem 9.1 Let V = [a1 , a2 ] [b1 , b2 ] [c1 , c2 ] and f : V R, integrable
on V . If for each (x, y) [a1 , a2 ] [b1 , b2 ] there exists h(x, y) =
Z
c2

f (x, y, z) dz then the function h : [a1 , a2 ] [b1 , b2 ] R is integrable


c1

and
ZZZ

ZZ

f (x, y, z) dx dy dz =
V

c2

dx dy
Z

[a1 ,a2 ][b1 ,b2 ]


Z b2
a2

dx
a1

f (x, y, z) dz
c1

c2

dy

(9.1.1)

f (x, y, z) dz.

b1

c1

Case II: simple domain with respect to 0z


If V is of the form
V = {(x, y, z) R3 | (x, y) D; 1 (x, y) z 2 (x, y)}
where D R2 is a measurable set and 1 , 2 : D R are continuous
functions, is said that V is simple with respect to 0z.
z
z= j1(x,y)

z= j2(x,y)
O

y
D

D is the orthogonal projection of V on the coordinate plane x0y. Any


parallel to 0z, passing through a point (x, y) of D, intersects each of the
surfaces z = 1 (x, y) and z = 2 (x, y) at a point.
Theorem 9.2 If the function f : V R is an integrable one V and
Z 2 (x,y)
for each (x, y) D there exists h(x, y) =
f (x, y, z) dz, then the
1 (x,y)

97
function h : D R is integrable on D and we have
ZZZ

ZZ

f (x, y, z) dx dy dz =

2 (x,y)

dx dy

f (x, y, z) dz.

(9.1.2)

1 (x,y)

9.2

Changes of variables
The general formula for changing the variables in the triple integral is
similar to the one for double integral (see Theorem 8.5). We present here
two particular cases of variable changes.
Spherical coordinates
Consider a point M (x, y, z). The spherical coordinates are:
the radial distance (the distance OM from the point to the origin),
the zenith angle formed by OM with the positive axis 0z,
the polar (azimuth) angle formed by OM 0 with the positive axis 0x,
where M 0 is the orthogonal projection of M on the plane x0y. The
connection between spherical coordinates and cartesian coordinates is
given by

x = cos sin
y = sin sin , with [0, +), [0, 2], [0, ].

z = cos
z
M

O
q
x

y
M

98
We determine the Jacobian

cos sin cos cos sin sin

D(x, y, z)

= sin sin sin cos cos sin

D(, , )
cos

sin
0
= 2 sin3 sin2 + 2 cos2 cos2 sin + 2 sin2 cos2 sin
+ 2 sin3 cos2 = 2 sin ,
and obtain the formula for change of variables into spherical coordinates:
ZZZ
f (x, y, z) dx dy dz =
V
ZZZ
=
f ( cos sin , sin sin , cos )2 sin d d d.
A

(9.2.1)
Cylindrical coordinates
Another frequently used system of coordinates is the cylindrical one.
For a point M (x, y, z) the cylindrical coordinates are: the polar radius
= OM 0 (where M 0 is the orthogonal projection of the point M on the
plane x0y), the angle formed by OM 0 with the positive axis 0x and
the height z. The relations between the cylindrical and the cartesian
coordinates are:

x = cos
y = sin , with [0, +), [0, 2], z R.

z=z
z
M

O
q
x

99
The formula for the change of variables is:
ZZZ
ZZZ
f (x, y, z) dx dy dz =
f ( cos , sin , z) d d dz. (9.2.2)
V

9.3

Applications of triple integral


I. The volume of a body
ZZZ
vol(V ) =

dx dy dz.
V

II. The mass and the mass center of a body


Consider a nonhomogeneous body, having the shape of the measurable, bounded set V R3 and the density at each point (x, y, z) given
by the function : V R. Then
ZZZ
ZZZ
1
M=
(x, y, z) dx dy dz, xG =
x(x, y, z) dx dy dz,
M
V
V
ZZZ
ZZZ
1
1
y(x, y, z) dx dy dz, zG =
z(x, y, z) dx dy dz.
yG =
M
M
V
V
III. Inertia moments of a body
For the body having the shape of V and the density , the inertia
moments with respect to the three coordinate planes can be calculated
by
ZZZ
ZZZ
2
Ix0y =
z (x, y, z) dx dy dz, Ix0z =
y 2 (x, y, z) dx dy dz,
V

ZZZ

x2 (x, y, z) dx dy dz.

Iy0z =
V

9.4

Exercises
9.1 Calculate the following integrals, by the iteration method:

100
ZZZ
i).

cos(x + z) dx dy dz, with V = [2, 3] [1, 2] [0, 1],


ZZZ

ii).
ZZZ

dx dy dz
, V = [1, e 1] [0, 1] [0, 1],
(x + y + z)3
(1 + x + y + z)2 dx dy dz, V is the domain described by

iii).
V

x + y + z 1, x 0, y 0 and z 0,
ZZZ p 2
x + y2
iv).
dx dy dz, where V is the domain bounded by the
z2
V
surfaces z = 1 and z 2 = x2 + y 2 ,
ZZZ
dx dy dz
v).
, where V : x + y + z 1, x, y, z 0,
4
V (1 + x + y + z)
ZZZ
vi).
(x2 + y 2 )2 dx dy dz, V is the body bounded by the paraboloid
V

z = x2 + y 2 and the plane z = 4,


ZZZ
p
xyz dx dy dz, with V given by z x2 + y 2 , z 2,
vii).
ZZZ

ey dx dy dz, V : x 0, 0 y 5, z 0, x + z 1,

viii).
ZZZ

exy dx dy dz, V : 0 z x 1, 1 y 2.

ix).
V

9.2 Calculate the following integrals, by an appropriate change of variables:


ZZZ
z
p
i).
dx dy dz, where V is given by
x2 + y 2 + z 2
V
1 x2 + y 2 + z 2 9 and z 0,
ZZZ
ii).
dx dy dz, V : x2 + y 2 + z 2 2z,
V

ZZZ r

x2 y 2 z 2
+ 2 + 2 dx dy dz, V being the interior of the ellipsoid
a2
b
c
V
2
2
2
y
z
x
+ 2 + 2 = 1,
2
a
b
c
ZZZ
iv).
z 2 dx dy dz, where V is given by x2 +y 2 2y and 1 z 1,

iii).

101
ZZZ
v).
ZZZ

z2
dx dy dz, V : x2 + y 2 2y, y 1, 1 z 1.
y
ln(x2 + y 2 + z 2 ) dx dy dz, V : 1 x2 + y 2 + z 2 e2 ,

vi).
V

9.3 Determine the volume of the body

2
2

z x + y
(z 2)2 x2 + y 2

z 2.
9.4 Determine the mass of the part of the sphere V = {(x, y, z)
R3 | x2 + y 2 + z 2 1, x 0, y 0, z 0}, of density (x, y, z) = xz.

Solutions and answers


9.1 i) We have, according to Theorem 9.1:
Z 3 Z 2 Z 1
Z 3 Z 2
1

I=
dx
dy
cos(x + z) dz =
dx
dy sin(x + z)
0
0
2
Z 3 1
Z2 3 Z1 2
2

dx [sin(x + 1) sin x] y
[sin(x + 1) sin x] dy =
dx
=
1
2
1
Z2 3
=
[sin(x + 1) sin x] dx = 2 cos 3 cos 4 cos 2.
2

1 4(e2 1)
ii) ln
1.
2
3
iii) We decompose the triple integral in a double integral and a simple
one.
z

B 1

D
x

A
1

1
C

102

ZZ

I=

dx dy
ZZ

1xy

(1 + x + y + z)2 dz

ZZ
(1 + x + y + z)3 1xy 1
dx dy
[8 (1 + x + y)3 ] dx dy.
=

3
3
0
D
D

D is the interior of the triangle AOC. The equation of the straight line
AC being x + y = 1, we have
Z
Z 1x
1 1
I=
dx
[8 (1 + x + y)3 ] dy =
3 0

Z 1 0
1
(1 + x + y)4 1x
8y
dx =
=

3 0
4
0

Z
1 1
(1 + x)4
1
(1 + x)5 1 31
2
=
4 8x +
dx =
4x 4x +
= .
3 0
4
3
20
60
0
iv) We have
ZZ
Z 1
I=
dx dy

1
dz =
z2
D
x2 +y 2

ZZ
p
1 1
=
dx dy x2 + y 2
2 2 =
z
x +y
D

!
ZZ
ZZ p
p
1
2
2
x + y 1 + p
dx dy =
( x2 + y 2 + 1)
=
x2 + y 2
D
D
x2 + y 2

D
1

D is the projection of V on the plane x0y, in this case, the disc centered
at the origin, of radius 1. Passing to polar coordinates, x = cos ,

103
y = sin , we have [0, 1] and [0, 2], so
Z

d
( + 1) d =
(2 + )2 d =
0
0
03

1 1

2 1

= 2 +
= .
= 2 +
3
2
3 2
3
0

I=

1
.
48 Z Z
Z
vi) I =
dx dy
v)

ZZ

2 2

4(x2 +y 2 )2 (x2 +y 2 )3 dx dy,

(x +y ) dz =
x2 +y 2
2

where
disc x + y 2 4. Passing to polar coordinates,
Z 2D isZthe
2

I=
d
(44 6 ) d = 26 .
3
0
0
vii) 0.

ZZ

viii) I =

dx dy
D

1x

ZZ
y

ey (1 x) dx dy =

e dz =
0

e5 1
ey (1 x) dy =
.
2 ZZ
0
Z0Z
Z x
Z 1 Z 2
xy
xy
ix) I =
dx dy
e dz =
xe dx dy =
dx
xexy dy =
D
D
0
1
Z 1
2 Z 10
2
1
e

=
dx exy =
e+ .
(e2x ex ) dx =
2
2
1
0
0
=

dx

9.2 i) We use spherical coordinates, that is x = cos sin , y =


sin sin , z = cos and get 1 2 9 and cos 0, which leads

to [1, 3], [0, 2], [ , ]. We have


2

104

I=

d
Z

Z 2

d
Z

d
Z

cos 2
p
sin d =
2
2 sin cos d =

Z
Z 2
1 3
sin2
=
d
d
d
2 d =
=
2
2 1
1
0
0
2
Z 3
3 3
26

1
2 2 d = =
.
=
2 1
3 1
3
3

4
ii)
.
3
iii) We pass to generalized spherical coordinates and obtain abc.
iv) V is a part of the interior of the cylinder, bounded by two parallel
planes).
z
1

-1

We pass to cylindrical coordinates x = cos , y = sin and z = z.


From x2 + y 2 2y we get 2 2 sin and further 2 sin , so
[0, 2 sin ]. Also, sin 0, that is [0, ]. Obviously, z [1, 1].
We have
Z Z 2 sin
Z Z 2 sin Z 1
z 3 1
2
d
d
d
d
z dz =
I=
3 1
1
0
0
0
0
Z Z 2 sin
Z
Z

1
2
2 sin 1
d
d =
d 2
=
4 sin2 d
=
3 0
3 0
3 0
0

Z 0

1
sin 2
2
2
= 2

.
(1 cos 2) d =
=
3
3
2
3
0
0
1
2 sin . Then, from
sin

1
2
sin 0 and
2 sin , which leads to sin
, we obtain
sin
2
v) Using cylindrical coordinates we get

105
3
2
[ , ]. Finally, I = ( 2).
4 4
3
vi) Using spherical coordinates, we have
Z e Z 2 Z
I=
d
d
ln 2 2 sin d =
0
Z1 e Z0 2
Z e
8 3
2
2
(2e + 1).
=
d
2 ln d = 8
2 ln d =
9
1
0
1
9.3 We have
ZZZ

ZZ

vol(V ) =

dx dy dz =
V

ZZ
=

(2

x2 +y 2

dx dy

dz
x2 +y 2

x2 + y 2 x2 y 2 ) dx dy.

D
z
2

D is the projection of V on the plane x0y. To find the radius of the disc
D, consider the system
(
z = x2 + y 2
(z 2)2 = x2 + y 2
that gives (z 2)2 = z. Among the two roots of the equation z 2 5z +4 =
0, z = 1 is convenient. For it we get x2 + y 2 = 1, so the radius of the disc
D is 1. Using polar coordinates, it follows
Z 1 Z 2
Z 1
2
vol(V ) =
d
(2 ) d =
(2 2 3 ) 2 d
0
0
0 3

4
1

.
= 2 2
=
3
4
6
0

106
9.4 The mass is:
ZZZ
M=
xz dx dy dz =
V
Z 1 Z
Z
2
2
=
d
d
cos sin cos 2 sin d =
0
0
0
Z 1
Z
Z
2
2
4
=
d
cos d
cos sin2 d =
0

sin3
1
5 1
2
2
= sin
= .
5 0
3 0
15
0

CHAPTER 10

Surface integrals
10.1

Surface integrals of scalar fields


Let S : D R3 be a smooth surface, given by the parametric equations

x = x(u, v)
y = y(u, v) ,

z = z(u, v)
where D R2 is a measurable domain.
To each point of the surface M we associate its position vector
~r(u, v) = x(u, v)~i + y(u, v)~j + z(u, v)~k
The vectors ~ru0 (u, v) = x0u (u, v)~i + yu0 (u, v)~j + zu0 (u, v)~k and ~rv0 (u, v) =
x0v (u, v)~i + yv0 (u, v)~j + zv0 (u, v)~k belong to the tangent plane to the surface
S, at the point M (u, v).
For a surface given by parametric equations, the following notation
will be used
E = (x0u )2 + (yu0 )2 + (zu0 )2 ,
107

F = x0u x0v + yu0 yv0 + zu0 zv0 ,

108
G = (x0v )2 + (yv0 )2 + (zv0 )2 ,
and

y0 y0
z0 z0
x 0 x0

A = u0 v0 B = u0 v0 C = 0u 0v .
zu zv
xu xv
yu yv

Let f : V R be a continuous function, with V a subset of R3 ,


containing the surface (S).
The real number
ZZ
ZZ
def
f (x, y, z) d =
f (x(u, v), y(u, v), z(u, v))k~ru0 ~rv0 k du dv.
S

(10.1.1)
is called the integral of f on the surface S (with respect to the surface
element).
Theorem 10.1 If the function f : V R is continuous, then
ZZ

ZZ

f (x(u, v), y(u, v), z(u, v)) A2 + B 2 + C 2 du dv

f (x, y, z) d =
S

Z ZD
=

f (x(u, v), y(u, v), z(u, v)) EG F 2 du dv,

(10.1.2)
with the notation mentioned above.
In the case when the surface has an explicit representation, the following result holds:
Theorem 10.2 If z = z(x, y), (x, y) D is the explicit equation of the
surface S, with D R2 , then
ZZ

ZZ

p
f (x, y, z(x, y)) 1 + p2 + q 2 dx dy,

f (x, y, z) d =
S

where p = zx0 and q = zy0 .

(10.1.3)

109
10.2

Applications of the surface integral of a scalar field


I. Area of a surface
ZZ
area(S) =

d.

(10.2.1)

II. The mass and the mass center of a surface


Considering (S) as made of a certain material, denote by : (S) R,
(x, y, z) the density at each point (x, y, z). The mass is given by
ZZ
M=
(x, y, z) d.
S

The coordinates of the mass center are:


ZZ
ZZ
1
1
x(x, y, z) d, yG =
y(x, y, z) d,
xG =
M
M
S
S
ZZ
1
zG =
z(x, y, z) d.
M
S
III. The inertia moments of a surface
The inertia moments of a surface S, of density , with respect to the
coordinate planes are:
ZZ
ZZ
2
Ix0y =
z (x, y, z) d, Ix0z =
y 2 (x, y, z) d,
S

ZZ

x2 (x, y, z) d.

Iy0z =
S

10.3

Surface integrals of vector fields


Let S be a smooth surface, having the parametric equations

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

z = z(u, v)

110
where D R2 is a measurable domain.
Consider a point M on the surface, having the position vector
~r(u, v) = x(u, v)~i + y(u, v)~j + z(u, v)~k.
The unit normal vector to the surface S, at the point M is
~n =

~ru0 ~rv0
k~ru0 ~rv0 k

(10.3.1)

At each point of the surface, two unit normal vectors can be considered,
having the same direction and opposite sense.
If ~n is a continuous function on (S), (S, ~n) and (S, ~n) are called
oriented surfaces. In this way, a surface has two faces.
For a surface given by parametric equations, the unit normal vector
has the formula

B
C
A
~
~
~
i+
j+
k .
~n =
A2 + B 2 + C 2
A2 + B 2 + C 2
A2 + B 2 + C 2
For a surface represented by the explicit equation z = z(x, y), (x, y)
D, D R2 , we have

!
p
q
1
~k ,
~i + p
~j + p
~n = p
2
2
2
2
2
2
1+p +q
1+p +q
1+p +q
where p = zx0 , q = zy0 .
For a surface represented by the implicit equation F (x, y, z) = 0,
(x, y, z) R3 , the unit normal vector is:
~n =

gradF (x, y, z)
.
kgradF (x, y, z)k

Let (S, ~n) be a smooth, oriented, simple surface.


Let F~ : V R3 be a continuous, vector-valued function (vector field),
F~ = P~i + Q~j + R~k.
The surface integral (with respect to the coordinates) of F~ on the
surface S, is the real number
ZZ
I=
F~ ~n d.
(10.3.2)
S

111
(where by F~ ~n is denoted the inner product of the two vectors). The
integral is denoted
ZZ
P (x, y, z) dy dz + Q(x, y, z) dz dx + R(x, y, z) dx dy
S

and is also called the flux of F through S.


10.4

Exercises
10.1 Calculate the following surface integrals for scalar fields:
ZZ
i).
z 2 d, where S is the surface represented by the parametric
S

x = u cos v
equations: y = u sin v , u [0, 1], v [0, 2],

z = 2u
ZZ
ii).
xz d, S = {(x, y, z) R3 | x = a cos u sin v, y = a sin u sin v,
S

z = a cos v, u [0, 4 ], v [0, 2 ], a > 0},


ZZ
iii). I =
(2xy+z) d, where S is the part of the plane xy+z = 1
S

cut by the coordinate planes,


ZZ
iv).
(x y + z) d, where S is the part of the plane x + y + z = 3
S

cut by the coordinate planes,


ZZ
p
d
v).
, where S is the part of the cone z = x2 + y 2 cut by
S z +1
the cylinder x2 + y 2 = 1,
10.2 Determine the area for the part of the paraboloid z = x2 + y 2
situated in the interior of the cylinder x2 + y 2 = 9.
10.3 Determine the area for the part of the conic surface z =
situated in the interior of the cylinder x2 + y 2 4y = 0.

p
x2 + y 2

112
10.4 Determine the mass of the nonhomogeneous surface
S = {(x, y, z) R3 | x2 + y 2 + z 2 = R2 , x 0, y 0, z 0}, of density
(x, y, z) = z.
10.5 Calculate the following surface integrals for vector fields:
ZZ
i).
2z dy dz 3y dz dx 3 dx dy, on the upper face of the plane
S

surface S = {(x, y, z) | x + 3y + 2z = 6, x 0, y 0, z 0},


ii). Determine the flux of the field F~ = ~i + z~k through the exterior face
of the half sphere S = {(x, y, z) | x2 + y 2 + z 2 = r2 ; z 0},
ZZ
iii).
x2 dy dz + y 2 dz dx + z 2 dx dy, S is the face of the surface x +
S

y + z = 0, x2 + y 2 1 for which the normal makes a pointed angle


with the positive semi-axis 0z,
ZZ

iv).
y dy dz + x dz dx + z + 2 dx dy, where S is the exterior face
S

of the parabolic surface z = x2 + y 2 , z 2.

Solutions and answers


10.1 i) We have:

xu = cos v
yu0 = sin v

0
zu = 2,

xv = u sin v
and
yv0 = u cos v

0
zv = 0,

E = sin2 v + cos2 v + 4 = 5, G = u2 sin2 v + u2 cos2 v = u2


F = u sin v cos v + u cos v sin v + 0 = 0,

so d = 5u2 du dv = u 5 du dv. The integral is transformed into a


double integral
ZZ
Z 1 Z 2

2
I=
4u u 5 du dv =
du
4 5 u3 dv
D
0
Z0 1
2

Z 1

du u3 v = 4 5 2
u3 du = 2 5.
=4 5

113

4
2
a
.
ii) d = a2 sin v du dv, I =
6
iii) The surface being given by an explicit equation, z = 1 x + y, we
use the formula (10.1.3).

-1

1
x

We have p = 1, q = 1 and d =
ZZ

3 dx dy. It follows

Z
(2x y + 1 x + y) 3 dx dy = 3 (x + 1) dx dy,

I=
D

where D is the projection of the surface S on the plane x0y, that is


the interior of the triangle AOB, with A(1, 0, 0) and B(1, 0, 0). We
calculate the double integral by iteration:
Z
I= 3

0
Z 1

dx
(x + 1) dy = 3
dx (x + 1)y
x1
0
x1
0

x3 1 2 3
(1 x2 ) dx = 3 x
.
= 3
=
3
3
0
0
1

9 3
iv)
.
2

v) d = 2 dx dy, I = 2 2(1 ln 2).


10.2 We have p = 2x, q = 2y, d =

1 + 4x2 + 4y 2 dx dy, so

114
z
9

3
x

area(S) =

ZZ p

1 + 4x2 + y 2 dx dy.

To calculate the double integral, we pass to polar coordinates:


(

x = cos
with [0, 3], [0, 2],
y = sin
Z

so area(S) =

d
0

1+

42 d

= 2

p
1 + 42 d. We make

the change of variable 42 + 1 = t, with 8 d = dt and obtain


Z

37

area(S) =
1

1 1/2
t3/2 37 3/2
t dt =
= (37 1).
4
4 3/2 1
6

10.3 4 2.
p
R
10.4 We have z = R2 x2 y 2 and d = p
.
2 x2 y 2
r
ZZ
ZZ
R2
R3
So M =
d =
R dx dy = R
=
. (we used the fact
4
4
S
D
ZZ
that
dx dy) is the area of D, which is a quarter of a disc of radius
R.

115
10.5 i) The normal to S is the normal to the plane x + 3y + 2z = 6,
~ = (1, 3, 2), with the unit vector
that is N
~n =

1
1
(~i + 3~j + 2~k) = (~i + 3~j + 2~k).
1 + 32 + 2 2
14

Since the integral is on the upper face of the surface (S), we choose the
1
unit vector ~n = (~i + 3~j + 2~k).
14
It follows that
ZZ
ZZ
1
1
I=
(2z 3y 3 3 2) d =
(2z 9y 6) d.
14 S
14 S
We transform the surface integral into a double integral. We have
6 x 3y
1
3
z=
and p = zx0 = , q = zy0 = , so
2
2
2

p
14
d = 1 + p2 + q 2 dx dy =
dx dy.
2
ZZ

14
1
6 x 3y
I=
9y 6 dx dy =
2
2
2
14
D
ZZ
1
(x 12y) dx dy,
=
2 D
where D is the projection of S on the plane x0y, that is the interior of
the triangle AOB, with A(6, 0, 0) and B(0, 2, 0). The equation of the line
x
AB is y = 2 , so we have:
3
Z 6 Z 2 x
Z
2 x
3
1
1 6
3
2
dx
dx (xy + 6y )
I=
(x + 12y) dy =
2 0
2 0
0
Z 6 0

2
x
1
x
x 2
=
+6 2
dx
2 0
3
3

Z
1 6
x2
=
24 6x +
dx = 30.
2 0
3
ii) The parametric equations of the surface are:

x = r cos sin

y = r sin sin , [0, 2], [ , ].

z = r cos

116
z

a
n

We have

and

x = r sin sin
y0 = r cos sin ,

0
z = 0

x = r cos cos
y0 = r sin cos

0
z = r sin

z0 z0
y0 y0


A = 0 0 = r2 cos sin2 , B = 0 0 = r2 sin sin2 ,
x x
z z

x0 x0

C = 0 0 = r2 sin cos , A2 + B 2 + C 2 = r4 sin2 .
y y

The vectors of the normal are ~n = ( cos sin ~isin sin ~j cos ~k).
To choose correctly the sign, notice that for = , that is the lowest
point of the sphere, we have ~n = ~k. Since the face is the exterior one,
~n = ~k is needed. At an arbitrary point, we take
~n = cos sin ~i + sin sin ~j + cos ~k.
It follows:
ZZ

ZZ

I=

dy dz + z dx dy =
ZZ

(cos sin + r cos cos ) d


S

(cos sin + r cos2 )r2 sin d d


ZD
Z 2
Z
Z
2
2
3
=r
d
sin cos d + r
d
=

Z
= r2

Z
2

3
2
sin sin d + r

= 0 + r3 2

cos
2r3
.
=

3
3
2
3

cos2 sin d

cos2 sin 2 d

117
1
iii) ~n = (~i + ~j + ~k), I = .
3
1
2
iv) ~n = p
(2x~i 2y~j + ~k), I =
.
3
4x2 + 4y 2 + 1

References
[1] Alexandra Ciupa, Calcul integral, Editura Mediamira, Cluj-Napoca,
2006.
[2] S. Cobza, Analiz matematic, Presa Universitar Clujean, 1997.
[3] Francine Delmer, Mathmatiques: rappels du cours et exercices resolues, Ed. Dunod, Paris, 1996.
[4] Eugenia Duca, Calcul diferenial, Editura U.T.Pres, 2006.
[5] G.M. Fihtenhol, Curs de calcul diferenial i integral, Editura
Tehnic, Bucureti, 1964.
[6] I. Gavrea, Probleme de analiz matematic, Editura Mediamira,
2006.
[7] A. Guzman, Derivatives and Integrals of Multivariable Functions,
Birkhuser, Boston, 2003.
[8] D. M. Ivan, Calculus, Editura Mediamira, Cluj-Napoca, 2002.
[9] D. M. Ivan, Elemente de calcul integral, Editura Mediamira, ClujNapoca, 2003.
[10] Viorica Murean, Analiza matematic , Editura Mega, Cluj-Napoca,
2006.
[11] S.M. Nikolsky, A Course of Mathematical Analysis, Mir publishers,
Moskov, 1981.
119

120
[12] D. Popa, Calcul integral, Editura Mediamira, Cluj-Napoca, 2005.
[13] D. Popa, Calculus, Editura Mediamira, Cluj-Napoca, 2006.
[14] Gh. Sirechi, Calcul diferenial i integral, Editura tiinific i Enciclopedic, Bucureti, 1985.
[15] O. Stnil, Analiz matematic, Editura didactic i pedagogic,
Bucureti, 1981.
[16] T. Trif, Probleme de calcul diferenial i integral n Rn , Casa Crii
de tiin, Cluj-Napoca, 2003.
[17] www.wikipedia.org (Wikipedia Free Encyclopedia)

Anda mungkin juga menyukai