Bai Giang Math4 PDF

H ANOI U NIVERSITY OF S CIENCE AND T ECHNOLOGY
S CHOOL OF A PPLIED M ATHEMATICS AND I NFORMATICS
N GUYEN T HI T HU H UONG AND T RAN M INH T OAN
Lecture on
M ATH 4
M ULTIPLE I NTEGRAL , I NTEGRAL THAT DEPENDS ON A PARAMETER ,

L INE I NTEGRAL , S URFACE I NTEGRAL , F IELD T HEORY AND S ERIES
Summary, Examples, Exercises and Solutions
Ha Noi - 2008
C ONTENTS
Contents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Chapter 1 . Multiple Integral . . . . . . . . . . . . . . . . . . . . . . . . 5

1 Double Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1 Calculation of a double integral in Cartesian coordinate system . . . 5
1.2 Change of variables in double integrals, polar coordinate . . . . . . . 8
1.3 Applications of double integrals . . . . . . . . . . . . . . . . . . . . . . 12
1.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.5 Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2 Triple Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.1 Calculation of a triple integral in Cartesian coordinate system . . . . 20
2.2 Change of variables in triple integrals . . . . . . . . . . . . . . . . . . 22
2.3 Calculate the triple integrals in cylindrical coordinate . . . . . . . . 22
2.4 Calculate the triple integrals in spherical coordinate . . . . . . . . . 24
2.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.6 Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Chapter 2 . Integrals that depend on a parameter . . . . . . . . . . . . . . 29
1 The definite integrals that depend on a parameter . . . . . . . . . . . . . . . 29
1.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
1.2 Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2 The generalized integarls that depend on a parameter . . . . . . . . . . . . . 32
2.1 The uniformly convergent integrals . . . . . . . . . . . . . . . . . . . . 32
2.2 Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.3 Euler’s integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4 Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Chapter 3 . Line integral . . . . . . . . . . . . . . . . . . . . . . . . . . 39
1 Line integral of the first kind . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
1
2 CONTENTS
1.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
1.2 Calculation formulae . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2 Line integral of the second kind . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.3 Theorem of four equivalent propositions . . . . . . . . . . . . . . . . . 45
2.4 Area of a plane domain . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4 Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Chapter 4 . Surface integral . . . . . . . . . . . . . . . . . . . . . . . . . 53
1 Surface integral of the first kind . . . . . . . . . . . . . . . . . . . . . . . . . 53
1.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2 Surface integral of the second kind . . . . . . . . . . . . . . . . . . . . . . . . 54
2.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.3 Stokes’ formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4 Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Chapter 5 . Field theory . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
1 Scalar field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2 Vector field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
4 Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Chapter 6 . Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
1 Number series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
1.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
1.2 Convergent criterion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
1.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
1.4 Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
2 Function series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
2.1 Function sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
2.2 Function series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
2.3 Power series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
2.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
2.5 Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
3 Fourier series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
CONTENTS 3
3.1 Decomposition theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

3.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3.3 Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
4 CONTENTS
CHAPTER 1
M ULTIPLE I NTEGRAL
§1. D OUBLE I NTEGRAL
1.1 Calculation of a double integral in Cartesian

coordinate system
Consider the integral ZZ
I= f ( x, y)dxdy. (1.1)
D
1. (The Corollary of Fubini’s theorem)

Suppose that D = [ a, b] × [c, d] and f : D → R is a continuous function on D. Then
Zb Zd Zd Zb
I= dx f ( x, y)dy = dy f ( x, y)dx
a c c a

a ≤ x ≤ b
2. If D is described as follows: D = ,
 ϕ( x ) ≤ y ≤ ψ( x )
where y = ϕ( x ), y = ψ( x) are continuous and have continuous derivatives on [ a, b]
Zb ψZ( x )
then I = f ( x, y)dy dx or
 

a ϕ( x )
Zb ψZ( x )
I= dx f ( x, y)dy. (1.2)
a ϕ( x )
5
6 Chapter 1. Multiple Integral

c ≤ y ≤ d
3. If D is described as follows: D = ,
 ϕ(y) ≤ x ≤ ψ(y)
where x = ϕ(y), x = ψ(y) are continuous and have continuous derivatives on [c, d]
then
Zd ψZ(y)
I= dy f ( x, y)dx. (1.3)
c ϕ(y)
Example 1.1. Calculate the double integral

ZZ
I= x2 ydxdy,
D
where D = [0, 1] × [0, 2].
Solution. We have
Z1 Z2 Z1 2 2
2y
ZZ
2 2
I= x ydxdy = dx x ydy = x dx

2 0
D 0 0 0
Z1
24 x3 1 2
= x dx = 2. = .
2 3 0 3
0
ZZ
x3 + xy dxdy where D is bounbed

Example 1.2. Calculate the double integral I =
√ D
by the curves y = x2 and y = x.
√
Solution. We have the region D = 0 ≤ x ≤ 1, x2 ≤ y ≤ x (Figure 1.1).

y
y = x2
√
y= x
1
O 1 x
Figure 1.1
1. Double Integral 7
Therefore
√
Z1 Z x √
y2
Z
x

3 3
I= dx x + xy dy = x y+x 2 dx
2 x
0 x2
Z1 √

3 1 2 1 5
5 5
= x x − x + x − x dx = .
2 2 36
0
Example 1.3. Interchange the order of the following integrals:

Z 2 Z 2x Z e Z ln y
i) I = dx f ( x, y)dy; ii) I = dy f ( x, y)dx.
0 x 1 0

 x = 0, x = 2



Solution. i) We have D = y=x (Figure 1.2)


y = 2x

O 2 4 x
Figure 1.2
From above figure, we have
Z2 Zy Z4 Z2
I= dy f ( x, y)dx + dy f ( x, y)dx.
0 y/2 2 y/2




 y = 1, y = e
ii) We have D = x=0 (Figure 1.3).


x = ln y


y
O 1 x
Figure 1.3
Hence
Z1 Ze
I= dx f ( x, y)dy.
0 ex
1.2 Change of variables in double integrals, polar

coordinate
1. In general
ZZ case
Put I = f ( x, y)dxdy.
D
To calculate I, we can perform the tranformation

 x = x (u, v)
(1.4)
 y = y(u, v).
The two equations in (1.4) define a mapping which carries a point ( x, y) ∈ D ⊂ Oxy
′ (or inversion).
to (u, v) ∈ D ⊂ Ouv
We shall consider mapping for which the functions x = x (u, v), y = y(u, v) are con-
tinuous and have continuous partial derivatives on D. Then
ZZ ZZ
I= f ( x, y)dxdy = f ( x (u, v), y(u, v)) | J | dudv,
D D

∂x ∂x
∂v = D ( x, y) 6= 0.

where J = ∂u

∂y ∂y D (u, v)
∂u ∂v

2. Polar coordinate
In this case we write r and ϕ instead of u and v and discrible the mapping by the
two equations 
 x = r cos ϕ
; | J | = r ≥ 0. (1.5)
y = r sin ϕ
Then ZZ ZZ
I= f ( x, y)dxdy = f (r cos ϕ, r sin ϕ) rdrdϕ.
DOxy D Orϕ
y
ZZ
Example 1.4. Calculate I = dxdy, where the region D is bounded by
x
D
y = x, y = 2x, xy = 1, xy = 3 ( x > 0).

y
Solution. Because x > 0, put = u, xy = v (u > 0, v > 0). Therefore we perform the
x
tranformation
√ √

v
v 1 1
x = √
 − . √ √ √
D ( x, y) u√ 2 u 2 √v u 1
u =⇒ J = = =− .
√ √ D (u, v) v u u
2√ u √
y = v. u

2 v

1 ≤ u ≤ 2
The region D → D = .
1 ≤ v ≤ 3
Hence
Z3 Z2
1 3 2
I= dv u − du = v u = 2.
u 1 1
1 1
y
y = 2x
y=x
1 xy = 3
xy = 1
O 1 x
Figure 1.4
Example 1.5. Tranform each of the given integrals to one or more interated integrals in
polar coordinate
√
Z1 Z1 Z1 Z1− x2
1. I = dx f ( x, y)dy; 2. I = dx f ( x, y)dy.
0 0 0 1− x
Solution. 1. We have D = [0, 1] × [0, 1] (Figure 1.5).
O 1 x
Figure 1.5
To transform to polar coordinate, we put x = r cos ϕ; y = r sin ϕ.

We devide the region D into two subregions by the line y = x : D = D1 ∪ D2 , where
π 1
D1 : 0 ≤ ϕ ≤; 0≤r≤ ;
4 cos ϕ
π π 1
D2 : ≤ϕ≤ ; 0≤r≤
4 2 sin ϕ
Therefore
1/cos ϕ 1/sin ϕ
Zπ/4 Z Zπ/2 Z
I= dϕ f (r cos ϕ, r sin ϕ) rdr + dϕ f (r cos ϕ, r sin ϕ) rdr
0 0 π/4 0




 x = 1, x = 2
2. Rewrite the region D = y = 1−x (Figure 1.5)

 √
y = 1 − x2


O 1 x
y = 1−x
Figure 1.6
π
 
 x = r cos ϕ; 0 ≤ ϕ ≤

Put , we have 2 .
1
y = r sin ϕ  ≤r≤1
sin ϕ + cos ϕ

Hence
Zπ/2 Z1
I= dϕ f (r cos ϕ, r sin ϕ) rdr.
0 1/(sin ϕ+cos ϕ)
Note: Some regions used frequently in polar coordinate
i) If DOxy = x2 + y2 ≤ R2 then

D Orϕ = {0 ≤ ϕ ≤ 2π; 0 ≤ r ≤ R} .
n o
ii) If D = x2 + y2 ≤ 2ax; a > 0 = ( x − a)2 + y2 ≤ a2 then

− π ≤ ϕ ≤ π

D= 2 2
0 ≤ r ≤ 2a cos ϕ.
n o
iii) If D = x2 + y2 ≤ 2ay; a > 0 = x2 + (y − a)2 ≤ a2 then


0 ≤ ϕ ≤ π
D=
0 ≤ r ≤ 2a sin ϕ.
iv) If
n o
D= x2 + y2 ≤ 2ax + 2ay, a > 0, b > 0
n o
= ( x − a )2 + ( y − b )2 ≤ a2 + b2
− arctg a ≤ ϕ ≤ − arctg a + π

then D = b b
0 ≤ r ≤ 2 ( a cos ϕ + b sin ϕ) .
v) If D = x2 + y2 = −2ax, a > 0 = ( x + a)2 + y2 = a2 then

 π ≤ ϕ ≤ 3π

D= 2 2
0 ≤ r ≤ −2a cos ϕ
x 2 y2
vi) If D is ellipse + 2 = 1 then we perform the transformation
a2 b

 x = ar cos ϕ
, | J | = abr.
 y = br sin ϕ

0 ≤ ϕ ≤ 2π
Then D → D = and
0 ≤ r ≤ 1
Z2π Z1
I= dϕ f ( ar cos ϕ, br sin ϕ) abrdr. (1.6)
0 0
1.3 Applications of double integrals

1. To compute the area of a plane domain D
The area of the domain D in the plane Oxy is computed by the formula:
ZZ
S= dxdy (1.7)
D
Example 1.6. Compute the area of the domain D bounded by the curves
5
xy = a2 , x + y = a.
2
5a
Solution. The curve xy = a2 cuts the line x + y = at two points that have abscis-
2
a
sas x = and x = 2a, respectively (Figure 1.7).
2
Therefore the area of D is
5a/2− x
Z2a
15
Z
S= dx dy = − 2 ln 2 a2
8
a/2 a2/x
1. Double Integral y 13
O 1 4 x
Figure 1.7
2. To compute the area of a curved surface

Suppose that S is a curved surface whose equation is z = f ( x, y) (Figure 1.6) and D
is the projection of S on the plane Oxy. Then the area of S is
ZZ q
′ ′
S= 1 + z x2 + zy2 dxdy (1.8)
D
z = f ( x, y)
z
O y
Figure 1.8
3. To compute the volume of a object

Suppose that the object Ω is bounded by the smooth curves
z = f ( x, y), z = g( x, y), f ≥ g ∀( x, y) ∈ D
and the surrounding cylinder has the directrix that is the border of the region D :
ϕ( x, y) = 0 and has the element that is parallel with Oz.
Then the volume of Ω is
ZZ
V= [ f ( x, y) − g( x, y)] dxdy. (1.9)
D
z = f ( x, y)
z
Ω
z = g( x, y)
O y
x
Figure 1.9
Example 1.7. Use the double integral to compute the volume of the object bounded by
z = 1 + x + y, z = 0, x + y = 1, x = 0, y = 0.
Solution. We have
Z1 1Z− x Z1 1 5
1h 2
i 1 1 3
I= dx (1 + x + y) dy = 4 − (1 + x ) dx = 2 − . (1 + x ) = .
2 2 3 0 6
0 0 0
1.4 Exercises
Exercise 1.1. Interchange the order of the following integrals
Z1 1Z− x2
1. dx f ( x, y)dy;
√
−1 − 1− x 2
√ 2
1+
Z1 Z 1− y
2. dy f ( x, y)dx;
0 2− y
√
Z2 Z 2x
3. dx f ( x, y)dy;
√
0 2x − x2
√ √ 2
Z 5 Z y Z2 Z4−y
4. dy f ( x, y)dx + dy f ( x, y)dx.
0 √
0 2 0
Exercise 1.2. Calculate the following integrals

π πo
ZZ n
1. x sin ( x + y) dxdy, D = ( x, y) ∈ R2 : 0 ≤ x ≤ ; 0 ≤ y ≤ ;
2 2
D
ZZ
2. x2 (y − x ) dxdy, where D is bounded by curves y = x2 and x = y2 ;
D
ZZ
| x + y| dxdy, D = ( x, y) ∈ R2 : | x | ≤ 1, |y| ≤ 1 ;

3.
D
ZZ p
4. |y − x2 |dxdx, D = {| x | ≤ 1, 0 ≤ y ≤ 1};
D
ZZ
5. (| x | + |y|) dxdy.
| x |+|y|≤1
Exercise 1.3. Transform the integral to polar coordinate and compute its value
√
ZR Z2 − x2
R
ln 1 + x2 + y2 dy, ( R > 0);

1. dx
0 0
√
ZR Z − x2
Rx
p
2. dx Rx − x2 − y2 dy, ( R > 0);
√
0 − Rx − x2
ZZ n o
2
3. xydxdy, where D = ( x, y) ∈ R2 : ( x − 2) + y2 ≤ 1, y ≥ 0 .
D
ZZ
4. xy2 dxdy, where D is bounded by the curves x2 + (y − 1)2 = 1 and x2 + y2 − 4y = 0.
D
Exercise 1.4. Calculate these following integrals:


ZZ
dxdy 4y ≤ x2 + y2 ≤ 8y
1. 2
, where D : √ ;
D
( x 2 + y2 )  x ≤ y ≤ 3x



 x2 + y2 ≤ 12

ZZ
xy  x2 + y2 ≥ 2x

2. dxdy, where D : √ ;
x 2 + y2 
 x 2 + y2 ≥ 2 3y
D 


x ≥ 0, y ≥ 0

s
1 − x 2 − y2
ZZ
3. dxdy, where D : x2 + y2 ≤ 1;
1 + x 2 + y2
D
2 2
9x2 − 4y2 dxdy, where D : x + y ≤ 1;
ZZ
4.
4 9
D

ZZ 1 ≤ xy ≤ 4
4x2 − 2y2 dxdy, where D :

5.
 x ≤ y ≤ 4x
D

x
y = 2



Exercise 1.5. Compute the area of the domain D bounded by y = 2− x


y = 4.


y = 0; y2 = 4ax
Exercise 1.6. Compute the area of the domain D bounded by
 x + y = 3a; y ≤ 0, ( a > 0).

 x2 + y2 = 2x; x2 + y2 = 4x
Exercise 1.7. Compute the area of the domain D bounded by
 x = y, y = 0
Exercise 1.8. Compute the volume of the object bounded by the surfaces

3x + y ≥ 1



3x + 2y ≤ 2


y ≥ 0, 0 ≤ z ≤ 1 − x − y.

Exercise 1.9. Compute the volume of the object bounded by the surfaces

0 ≤ z ≤ 1 − x 2 − y2
√
y ≥ x, y ≤ x 3.
1.5 Solutions
√ 2
√
Z0 Z1−y Z1 Z1−y
Solution 1.1. 1. I = dy f ( x, y)dx + dy f ( x, y)dx;
−1
√ 0
√
− 1− y2 − 1− y
√
Z2 Z − x2
2x
2. I = dx f ( x, y)dy;
1 2− x
√ 2
1−
Z1 Z 1− y Z1 Z2 Z2 Z2
3. dy f ( x, y)dx + dy f ( x, y)dx + dy f ( x, y)dx;
0 0
√
y2/2 1+ 1− y2 1 y2/2
√ √
Z2 Z4− x2
4. I = dx f ( x, y)dy.
0 x
Zπ/2 Zπ/2
π
Solution 1.2. 1. I = dx x sin( x + y)dy = ;
2
0 0
√
Z1 Zx
1
x2 y − x3 dy = −

2. I = dx ;
504
0 x2
3. Divide D into two regions D = D1 ∪ D2 , where
D1 = {−1 ≤ x ≤ 1, − x ≤ y ≤ 1}
D2 = {−1 ≤ x ≤ 1, −1 ≤ y ≤ − x } .
Then
Z1 Z1 Z1 Z− x
8
I= dx ( x + y)dy − dx ( x + y)dy = .
3
−1 −x −1 −1

n o
D1 = −1 ≤ x ≤ 1, x2 ≤ y ≤ 1
n o
D2 = −1 ≤ x ≤ 1, 0 ≤ y ≤ x2 .
Hence
2
Z1 Z1 q Z1 Zx q
3π + 4
I= dx y − x2 dy + dx x2 − ydy = .
12
−1 x2 −1 0
5. Note that D is axisymetricZZand the function f ( x, y) = | x | + |y| is even with respect

to x and y. Therefore I = 4 (| x | + |y|) dxdy, where
D1
D1 = {( x, y) : 0 ≤ x ≤ 1; 0 ≤ y ≤ 1 − x } .
Hence
Z1 1Z− x
4
I=4 dx ( x + y)dy = .
3
0 0
n √ o
Solution 1.3. 1. We have D = 0 ≤ x ≤ R; 0 ≤ y ≤ R2 − x2 .

 x = r cos ϕ π
Put =⇒ 0 ≤ ϕ ≤ , 0 ≤ r ≤ R.
y = r sin ϕ 2
Hence
Zπ/2 ZR
2
π h 2
2 2
i
I= dϕ ln 1 + r rdr = R + 1 ln( R + 1) − R .
4
0 0
n √ √ o
2. We have D = 0 ≤ x ≤ R, − Rx − ≤ y ≤ RX − x2. x2
 x = R + r cos ϕ
 
0 ≤ ϕ ≤ 2π
Put 2 =⇒ | J | = r, .
y = r sin ϕ 0 ≤ r ≤ R
2
Hence s
Z2π ZR/2
R2 πR3
I = dϕ − r2 rdr = .
4 12
0 0
 
 x = 2 + r cos ϕ 0 ≤ r ≤ 1
3. Put =⇒ . Then
y = r sin ϕ 0 ≤ ϕ ≤ 2π
Z2π Z1
I= dϕ (2 + r cos ϕ) r sin ϕrdr = 0
0 0
Note: The domain D is Ox-axisymetric and the function f ( x, y) = xy is odd with

respect to y. Therefore we have I = 0.
 
 x = r cos ϕ 0 ≤ ϕ ≤ π
4. Put =⇒ . Hence
y = r sin ϕ 2 sin ϕ ≤ r ≤ 4 sin ϕ
Zπ sin ϕ
4Z
I= dϕ r cos ϕ (r sin ϕ)2 rdr = 0.

0 2 sin ϕ
Note: The domain D is symetric to the axis Oy and the function f ( x, y) = xy2 is odd
with respect to x. Therefore we have I = 0.
π ≤ ϕ ≤ π
 
 x = r cos ϕ
Solution 1.4. 1. Put =⇒ 4 3 . Hence
y = r sin ϕ 4 sin ϕ ≤ r ≤ 8 sin ϕ
Zπ/3 sin ϕ
8Z
1 3 1
I= dϕ 4
rdr = 1− √ .
r 128 3
π/4 4 sin ϕ

n π √ o
D1 = 0 ≤ ϕ ≤ , 2 cos ϕ ≤ r ≤ 2 3
6
nπ π √ √ o
D2 = ≤ ϕ ≤ , 2 3 sin ϕ ≤ r ≤ 2 3 .
6 2
Then
√ √
Zπ/6 2Z 3 Zπ/2 2Z 3
r2 cos ϕ sin ϕ r2 cos ϕ sin ϕ 11
I= dϕ rdr + dϕ rdr = .
r2 √ r 2 8
0 2 cos ϕ π/6 2 3 sin ϕ
 
 x = r cos ϕ 0 ≤ ϕ ≤ 2π
3. Put =⇒ . Then
y = r sin ϕ 0 ≤ r ≤ 1
Z2π Z1
s
1 − r2 π2
I= dϕ rdr = .
1 + r2 2
0 0
 
 x = 2r cos ϕ 0 ≤ ϕ ≤ 2π
4. Put =⇒ | J | = 6r, . Then
y = 3r sin ϕ 0 ≤ r ≤ 1
Z2π Z1
I = 6 × 36 |cos 2ϕ| dϕ r3 dr = 216.
0 0
 
u = xy 1 ≤ u ≤ 4 r
u √
5. Put =⇒ and x = , y = uv. Therefore
v = y 1 ≤ v ≤ 4 v
x
Z4 Z4 h
u i 1 45
I= du 4 − 2uv . dv = − .
v 2v 4
1 1
Solution 1.5. Divide the domain D into two subdomain D = D1 ∪ D2 , where
D1 = −2 ≤ x ≤ 0, 2− x ≤ y ≤ 4 ; D2 = {0 ≤ x ≤ 2, 2x ≤ y ≤ 4} .

Then
Z0 Z4 Z2 Z4
3
S= dx dy + dx dy = 2 8 − .
ln 2
−2 2− x 0 2x
Solution 1.6. We have

Z0 Z−y
3a Z0
y2

S= dy dx = 3a − y − dy = 18a2 .
4a
−6a y2/4a −6a
0 ≤ ϕ ≤ π ,
 
 x = r cos ϕ
Solution 1.7. Change to polar coordinate =⇒ 4 .
y = r sin ϕ 2 cos ϕ ≤ r ≤ 4 cos ϕ
Hence
Zπ/4 cos ϕ
4Z
3π 3
S= dϕ rdr = + .
4 2
0 2 cos ϕ
Solution 1.8. We have

(2−2y)/3
Z1 Z1
1 1
Z
V= dy (1 − x − y) dx = 1 − 2y + y2 dy = .
6 18
0 (1−y)/3 0
Solution 1.9. We have V = 2V1 , where

Zπ/3 Z1 π
V1 = dϕ 1 − r2 rdr = .
48
π/4 0
π
Hence V = .
24
§2. T RIPLE I NTEGRAL
2.1 Calculation of a triple integral in Cartesian

coordinate system
Consider the integral ZZZ
I= f ( x, y, z)dxdydz, (1.10)
V
2. Triple Integral 21
where f ( x, y, z) is three-variables function that is continuous on V.

If
n o
V = ( x, y, z) ∈ R3 |( x, y) ∈ D; z1 ( x, y) ≤ z ≤ z2 ( x, y) ,
where D is the projection of V on the plane Oxy and z1 , z2 are continuous on D then
 
ZZ z2Z( x,y)
I= dxdy  f ( x, y, z)dz . (1.11)

 
D z1 ( x,y)
dxdydz
ZZZ
Example 2.1. Calculate the integral I = , where V is bounded by the
( x + y + z )3
V
planes x = 0, y = 0, z = 0 and x + y + z = 1.
Solution. V is tetrahedron bounded by two planes z = 0 and z = 1 − x − y, ( x, y) ∈ D,

where D is the triangle OAB in the plane Oxy (Figure 2.1). Hence we have
1−
Zx−y
dz (1 + x + y + z)−2 z=1−x−y
ZZ ZZ
I= dxdy = dxdy
( x + y + z )3 −2

z =0
D 0 D
Z1 1Z− x ! Z1
1 1 1 1 1−x 1 1
=− dx − dy = − + − dx
2 4 (1 + x + y )2 2 4 2 1+x
0 0 0
Z1
1 3 x 1 1 5
=− − − dx = ln 2 − .
2 4 4 1+x 2 16
0
O B y
x
Figure 2.1
2.2 Change of variables in triple integrals

Consider the tranformation:

 x = x (u, v, w)



y = y(u, v, w)


z = z(u, v, w).

Suppose that the following conditions are satisfied:
i) (u, v, w) ∈ V ′ in O′ uvw-plane and x (u, v, w), y(u, v, w), z(u, v, w) are continuous and
have continuous partial derivatives on V ′ .
ii) The vecto-valued mapping Φ : V ′ → V is one-to-one.
iii) The Jacobian determinant

x′ x′ x′
D ( x, y, z) u v w
J= = y′u y′v y′w 6= 0 in V ′ .

D (u, v, w) ′
zu z′v z′w

Then
ZZZ ZZZ
I= f ( x, y, z)dxdydz = f ( x (u, v, w), y(u, v, w), z(u, v, w)) | J | dudvdw. (1.12)
V V′
2.3 Calculate the triple integrals in cylindrical

coordinate
Here we write r, ϕ, z for u, v, w and define the mapping by the equations:
x = r cos ϕ, y = r sin ϕ, z = z. (1.13)
In other words, we replace x and y by their polar coordinate in the plane Oxy and retain
z.
Again, to get a one-to-one mapping we must keep r > 0 and restrict ϕ to be in an interval
of the form: ϕo ≤ ϕ < ϕo + 2π.
The Jacobian determinant of the mapping in (1.13) is

cos ϕ −r sin ϕ 0

J = sin ϕ r cos ϕ 0 = r cos2 ϕ + sin2 ϕ = r > 0

0 0 1
and therefore we have the tranformation formula

ZZZ ZZZ
f ( x, y, z)dxdydz = f (r cos ϕ, r sin ϕ, z) rdrdϕdz. (1.14)
V V′
Note: In some cases we use the generalized cyclindrical coordinate
x = ar cos ϕ, y = br sin ϕ, z = z
and J = abr.
ZZZ
x2 + y2 dxdydz to cylindrical coordinate

Example 2.2. Transform the integral I =
V
and compute its value, where V is the region bounded by the surfaces x2 + y2 = 2z and
z = 2.
Solution. Transform to cylindrical coordinate : x = r cos ϕ, y = r sin ϕ, z = z, 0 ≤ ϕ ≤ 2π.
−2 O 2 y
Figure 2.2
We note that the paraboloid x2 + y2 = 2z cuts the plane x2 + y2 = 4 by the circle x2 + y2 =

4, therefore 0 ≤ r ≤ 2.
r2
On the other hand on the paraboloid we have r2 cos2 ϕ + r2 sin2 ϕ = 2z =⇒ z = .
2
r2
So that in B we have ≤ z ≤ 2.
2
Therefore we have
Z2π Z2 Z2 Z2
r2

3 3 16π
I= dϕ dr r dz = 2π r 2− dr = .
2 3
0 0 r2/2 0
2.4 Calculate the triple integrals in spherical

coordinate
In this case the symbols r, θ, ϕ are used instead of u, v, w and the mapping is defined
by the equations 
 x = r sin θ cos ϕ



y = r sin θ sin ϕ (1.15)


z = r cos θ

To get a one-to-one mapping we keep r > 0, 0 ≤ ϕ < 2π and 0 ≤ θ < π.

The Jacobian determinant of the mapping is

sin θ cos ϕ r cos θ cos ϕ −r sin θ sin ϕ

J = sin θ sin ϕ r cos θ sin ϕ r sin θ cos ϕ = −r2 sin θ

cos θ −r sin θ 0
Therefore we have the tranformation formula

ZZZ ZZZ
f ( x, y, z)dxdydz = f (r sin θ cos ϕ, r sin θ sin ϕ, r cos θ ) r2 sin θdrdθdϕ (1.16)
V V′
ZZZ p
Example 2.3. Transform the integral I = x2 + y2 + z2 dxdydz to spherical coordi-
V
nate and compute its value, where V is the sphere x2 + y2 + z2 ≤ z.
Solution. We have
1 2 1

2 2 2 2 2
x +y +z −z = x +y + z− − .
2 4
( 2 )
1 1
So that V = ( x, y, z) : x2 + y2 + z − ≤ , i.e V is sphere whose center is the
2 4

1 1
point 0, 0, and radius R = .
2 2
Transform to spherical coordinate.
2.5 Exercises
Exercise 1.10. Calculate the following triple integrals:

0 ≤ x ≤ 4


ZZZ 
1. zdxdydz, where the region V is defined by x ≤ y ≤ 2x .

V
 p
0 ≤ z ≤ 1 − x 2 − y2


ZZZ  x 2 + y2 + z2 ≤ 1
x2 + y 2

2. dxdydz, where V : .
 x 2 + y2 − z2 ≤ 0
V

ZZZ  x 2 + y2 ≤ 1
x2 + y2 zdxdydz, where V :

3. .
1 ≤ z ≤ 2
V
ZZZ
x2 + y2 zdxdydz, where

4.
V
(a) V is the region bounded by the cylinder x2 + y2 = 2x and the planes z = 0, z =

a ( a > 0).
(b) V is a half of the sphere x2 + y2 + z2 ≤ a2 , z ≥ 0 ( a > 0).

x 2 + y2 z2
(c) V is a half of the ellipsoid + 2 ≤ 1, z ≥ 0, ( a, b > 0).
a2 b
ZZZ √
5. ydxdydz, where V is bounded by the cone y = x2 + z2 and the plane y = h, ( h >
V
0).
ZZZ 2
y2 z2 x 2 y2 z2

x
6. + + 2 dxdydz, where V is bounded by + 2 + 2 = 1 ( a, b, c > 0).
a2 b2 c a2 b c
V
2.6 Solutions
√
1/4 1− x 2 − y2
Z2x
43
Z Z
Solution 1.10. 1. I = dx dy zdz = .
3072
1 x 0

 x = r sin θ cos ϕ



2. Transform to spherical coordinate y = r sin θ sin ϕ =⇒ | J | = r2 sin θ.


z = r cos θ

π
We have 0 ≤ ϕ ≤ 2π, 0 ≤ r ≤ 1 and 0 ≤ θ ≤ . Hence
4
√
Z2π Zπ/4 Z1 8 − 5 2
2π
I = dϕ dθ r2 sin2 θ.r2 sin θdr = . .
5 12
0 0 0
 
 x = r cos ϕ 0 ≤ ϕ ≤ 2π

 

 
3. Transform to cylindrical coordinate y = r sin ϕ =⇒ 0≤r≤1 . Hence

 

z = z
 1 ≤ z ≤ 2

Z2π Z1 Z2
3π
I= dϕ dr r2 zrdz = .
4
0 0 1
4. (a) Transform to cylindrical coordinate:
π π
 
− ≤ϕ≤
 x = r cos ϕ

 

  2
 2
y = r sin ϕ =⇒ 0 ≤ r ≤ 2 cos ϕ

 

z = z
 0 ≤ z ≤ a

Then
Zπ/2 cos ϕ
2Z Za
16a2
I= dϕ dr zrrdz = .
9
−π/2 0 0
(b) Transform to cylindrical coordinate:

 
 x = r cos ϕ 0 ≤ ϕ ≤ 2π

 

 
y = r sin ϕ =⇒ 0≤r≤a

 
 √
z = z
 0 ≤ z ≤ a2 − r 2

Hence √
Z2π Za a2 −r 2
2πa5
Z
I= dϕ dr zrrdz = .
15
0 0 0
(c) Transform to generalized cylindrical coordinate

 
 x = ar cos ϕ 0 ≤ ϕ ≤ 2π

 

 
y = ar sin ϕ =⇒ | J | = abr, 0≤r≤1 .
  √
z = bz′ 0 ≤ z ′ ≤ 1 − r 2

 

Hence √
Z2π Z1 Z1−r2
2πab2
I= dϕ dr bz′ rabrdz′ = .
15
0 0 0
5. Transform to cylindrical coordinate

 
 x = r sin ϕ 0 ≤ ϕ ≤ 2π

 

 
z = r sin ϕ =⇒ | J | = r, 0≤r≤h .

 

y = y
 r ≤ y ≤ h

Hence
Z2π Zh Zh
πh4
I= dϕ dr yrdr = .
4
0 0 r
6. Transform to generalized spherical coordinate

 
 x = ar sin θ cos ϕ 0 ≤ ϕ ≤ 2π

 

 
y = br sin θ sin ϕ =⇒ | J | = abcr2 sin θ and 0 ≤ θ ≤ π .

 

z = cr cos θ
 0 ≤ r ≤ 1

Hence
Z2π Zπ Z1
4abcπ
I= dϕ dθ r2 .abc.r2 sin θdr = .
5
0 0 0
CHAPTER 2
I NTEGRALS THAT DEPEND ON A PARAMETER
§1. T HE DEFINITE INTEGRALS THAT DEPEND ON A

PARAMETER
1.1 Definition
Suppose that f ( x, y) is a function defined with x ∈ [ a, b] and y ∈ Y such that for each
y ∈ Y, fixed the function f ( x, y) is integrable in [ a, b].
Then
Zb
I (y) = f ( x, y)dx (2.1)
a
is a function that is defined on Y and called integral that depends on a parameter of the
function f ( x, y) on [ a, b].
1.2 Properties
1. The continuity and limitation under the integral sign
If function f ( x, y) is defined and continuous on the rectangle D = [ a, b] × [c, d] then
the integral I (y) is continuous on [c, d], i.e
Zb Zb Zb
lim I (y) = lim f ( x, y)dx = lim f ( x, y)dx = f ( x, yo )dx = I (yo )
y→yo ∈[c,d] y→yo y→yo
a a a
Z2
Example 1.1. Compute lim x2 cos xydx.
y →0
0
29
30 Chapter 2. Integrals that depend on a parameter
Solution. Let [c, d] be any interval that contains the point y = 0. Then the function
f ( x, y) = x2 cos xy is continuous on the rectangle D = [0, 2] × [c, d]. Therefore the
Z2
integral I (y) = x2 cos xydx is continuous on [c, d] and we have
0
Z2 Z2
2 8
lim I (y) = I (0) = x cos 0dx = x2 dx = .
y →0 3
0 0
2. Differentiation under the integral sign

Suppose that
i) f ( x, y) is defined on the rectangle D = [ a, b] × [c, d] and continuous with respect

to x ∈ [ a, b] for each y ∈ [c, d], fixed.
∂ f ( x, y)
ii) f ( x, y) has the partial derivative that is continuous on D.
∂y
Then the integral I (y) is differential function on [c, d] and
Zb
′ ∂f
I (y) = ( x, y)dx, y ∈ [c, d] (Leibniz’s rule).
∂y
a
Example 1.2. Compute the derivative with respect to the parameter of the integral
Zπ/2
I ( a) = ln a2 − sin2 x dx, ( a > 1).
0
Solution.h Thei function f ( a, x ) = ln a2 − sin2 x is continuous in the region a > 1

π
and x ∈ 0, and has partial derivative with respect to a:
2
∂f 2a
= 2 , a>1
∂a a − sin2 x
is continuous in that region.
Hence we can apply the Leibniz’s formula to obtain
Zπ/2 Zπ/2
2a dx
I ′ ( a) = dx = 2a .
a2 − sin2 x ( a2 − 1) + cos2 x
0 0
Change the variable t = tg x we have

+∞ √
dt 2 a2 − 1 +∞ π
Z
′
I ( a) = 2a 2 2 2
=√ . arctg t = √ .
a + ( a − 1) t a2 − 1 a 0 a2 − 1
0
2. The generalized integarls that depend on a parameter 31
3. Integration under the integral sign

If f ( x, y) is defined and continuous on the rectangle D = [ a, b] × [c, d] then
Zd Zd Zb Zb Zd
I (y)dy = dy f ( x, y)dx = dx f ( x, y)dy.
c c a a c
Example 1.3. By integrating under the integral sign, compute the integral
Z1
1 xb − x a

I ( a, b) = sin ln , a > 0, b > 0.
x ln x
0
Solution. Assume that a < b. Firstly, we note that
Zb
xb − x a
= x y dy, 0 < a < b.
ln x
a
Therefore we can rewrite
Z1 Z1 Zb
xb − x a

1 y 1
I ( a, b) = sin ln dx = dx x sin dy.
x ln x x
0 0 a

1
Let f ( x, y) = x y sin ln .
x
1
Because lim f ( x, y) = lim ln x y sin
= 0, so that we can add the value f (0, y) = 0
x →0 x →0 x
such that the function f ( x, y) is continuous on the rectangle [0, 1] × [ a, b].
Therefor we can change the order of integration and get
Zb Z1
y 1
I ( a, b) = dy x sin sin dx.
x
a 0
Change the variable x = e−t , we have

Zb +∞
Z
I ( a, b) = dy e−t(1+y) sin tdt,
a 0
where
+∞
−(1 + y) sin t − cos t −t(1+y) +∞ 1
Z
e−t(1+y) sin tdt = 2
e = .
1 + (1 + y ) 0 1 + (1 + y )2
0
In conclusion we have
Zb
dy b−a
I ( a, b) = 2
= arctg(b + 1) − arctg( a + 1) = arctg .
1 + (1 + y ) 1 + (1 + a)(1 + b)
a
§2. T HE GENERALIZED INTEGARLS THAT DEPEND ON A

PARAMETER
2.1 The uniformly convergent integrals

Consider the integral
+∞
Z
I (y) = f ( x, y)dx, y ∈ Y. (2.2)
a
We say that I (y) is uniformly convergent if:
+∞
Z
i) For each y ∈ Y, fixed the integral f ( x, y)dx is convergent.
a
ii) ∀ε > 0, ∃ Ao = Ao (ε) > a (only depends on ε) such that ∀ A > Ao we have

Z+∞

f ( x, y)dx < ε, ∀y ∈ Y.
A
Some uniformly convergent criteria
1. Cauchy’s criterion
The necessary and sufficient condition for the integral (2.2) is uniformly convergent
on Y is:
∀ε > 0, ∃ Ao = Ao (ε) such that for all A′ , A′′ > Ao we have
′′
ZA

f ( x, y)dx < ε, ∀y ∈ Y.

′
A
2. Weierstrass’ criterion
Suppose that ϕ( x ) is nonnegative function on [ a, +∞) such that
| f ( x, y)| ≤ ϕ( x ), ∀ x ∈ [ a, +∞), ∀y ∈ Y.
+∞
Z +∞
Z
Then if the integral ϕ( x )dx is convergent then the integral f ( x, y)dx is uni-
a a
formly convergent on Y.
2. The generalized integarls that depend on a parameter 33
2.2 Properties
1. The continuity and limitation under the integral sign
Suppose that the function f ( x, y) is defined and continuous on [ a, +∞) × [c, d]. Then
if the integral (2.2) is uniformly convergent on [c, d] then I (y) is continuous on [c, d],
i.e
+∞
Z +∞
Z +∞
Z
lim I (y) = lim f ( x, y)dx = lim f ( x, y)dx = f ( x, yo )dx = I (yo )
y→yo ∈[c,d] y→yo y→yo
a a a
2. Differentiation under the integral sign

Suppose that
i) f ( x, y) is continuous with respect to x ∈ [ a, +∞).

∂ f ( x, y)
ii) f ( x, y) has the partial derivative that is continuous on [ a, +∞) × [c, d].
∂y
+∞
∂f
Z
iii) The integral ( x, y)dx is uniformly convergent on [c, d].
∂y
a
+∞
Z
Then the integral I (y) = f ( x, y)dx is differential on [c, d] and
a
+∞
∂f
Z
′
I (y) = ( x, y)dx, y ∈ [c, d].
∂y
a
3. Integration under the integral sign

If f ( x, y) is defined and continuous on the rectangle D = [ a, b] × [c, d] then
Zd Zd Zb Zb Zd
I (y)dy = dy f ( x, y)dx = dx f ( x, y)dy.
c c a a c
2.3 Euler’s integrals

1. The Gamma function Γ
+∞
Z
Γ( a) = x a−1 e− x dx, a > 0. (2.3)
0
The differentiation formula:
+∞
Z
Γ (k)
( a) = x a−1 e− x (ln x )k dx, k ∈ N
0
Some basic properties
i) Γ( a + 1) = aΓ( a)
ii) Γ(n + 1) = n!, n ∈ N
(2n − 1)!! √

1
iii) Γ n + = . π, n ∈ N
2 2n
π
iv) Γ( a).Γ(1 − a) = , 0 < a < 1.
sin πa
2. The Beta function B
Z1
B( a, b) = x a−1 (1 − x )b−1 dx, a > 0, b > 0. (2.4)
0
Some basic properties
i) B( a, b) = B(b, a) ∀ a > 0, b > 0

b−1
ii) B( a, b) = B( a, b − 1), a > 0, b > 1
a+b−1
π
iii) B( a, 1 − a) = , 0 < a < 1.
sin πa
§3. E XERCISES
Z1
y f (x)
Exercise 2.1. Consider the continuity of the integral I (y) = dx, where f ( x ) > 0
x 2 + y2
0
and is continuous on [0, 1].
Exercise 2.2. Compute the following integrals
Z1
xb − x a
1. dx, (0 < a < b);
ln x
0
Z∞ −αx
e − e− βx
2. dx, (α > 0, β > 0);
x
0
+∞
sin(bx ) − sin(cx )
Z
3. e−ax dx, ( a, b, c > 0);
x
0
+∞
Z
2
4. e− x cos(yx )dx.
0
4. Solutions 35
Zπ/2
Exercise 2.3. Express the integral sinm x cosn xdx through the function B(m, n) (m, n ∈
0
Z, m, n > 1).
Exercise 2.4. Compute the following integrals

Zπ/2
1. sin6 x cos4 xdx;
0
Za √ √
2. x2n a2 − x2 dx ( a > 0) (Suggest: put x = a t);
0
+∞
Z
2
3. x10 e− x dx;
0
+∞
dx
Z
4. .
1 + x3
0
§4. S OLUTIONS
y f (x)
Solution 2.1. • With y 6= 0, the function g( x, y) = is continuous on each
x 2 + y2
rectangle [0, 1] × [c, d] or [0, 1] × [−d, −c], 0 < c < d.
Because c can be arbitrarily small and d can be arbitrarily great so that I (y) is
continuous when y 6= 0.
• With y = 0.
Because f ( x ) > 0, ∀ x ∈ [0, 1] so that ∀m > 0 such that f ( x ) ≥ m > 0, ∀ x ∈ [0, 1].
Therefore ∀ε > 0 we have
Z1 Z1
ε f (x) mε 1
I (ε) = dx ≥ dx = m arctg
x 2 + ε2 x2
+ε 2 ε
0 0
Z1
−ε f ( x ) 1
I (−ε) = 2 2
dx ≤ m arctg
x +ε ε
0
1 π
=⇒ | I (ε) − I (−ε)| ≥ 2m arctg → 2m. khi ε → 0.
ε 2
Hence I (y) is interrupted at y = 0.
Solution 2.2. 1. Consider the function f ( x, y) = x y , we have f ( x, y) is continuous on

[0, 1] × [c, d].
Z1 Z1
1
The integral x y dx is uniformly convergent because xy ≤ xb ; x b dx = .
b+1
0 0
Hence
Z1 Zb Zb Z1
y b+1
I= dx x dy = dy x y dx = ln .
a+1
0 a a 0
e− xy
2. Put F ( x, y) = , we have
x
Zβ Zβ
e−αx − e− βx ′
= F ( x, α) − F ( x, β) = − Fy ( x, y)dy = e−yx dy
x
α α
Put f ( x, y) = e−yx , check the uniformly convergent conditions:
i) f ( x, y) is continuous on [0, +∞) × [α, β];

+∞
Z
ii) I (y) = e−yx dx is uniformly convergent on [α, β] as
0
e−yx ≤ e−αx , ∀( x, y) ∈ [0, +∞) × [α, β]

Z∞
1
and e−αx dx = - convergent.
α
0
Hence
Z∞ Zβ Zβ Z∞ Zβ
dy β
I= dx e−yx dy = dy e−yx dx = = ln .
y α
0 α α 0 α
e−ax sin yx
3. Put F ( x, y) = , we have
x
Zb
− ax sin bx − sin cx ′
e = F ( x, b) − F ( x, c) = Fy dy.
x
c
Put f ( x, y) = e ax cos yx, we have
i) f ( x, y) is continuous on [0, +∞) × [c, b];

Z∞
ii) e−ax cos yxdx is uniformly convergent on [c, b] because |e−ax cos yx | ≤ e−ax and
0
Z∞
the integral e−ax dx is convergent.
0
4. Solutions 37
Hence
+∞ Zb Zb Z∞ Zb
a b c
Z
− ax
I= dx cos yxdx = dy e cos yxdx = dy = arctg − arctg .
a2 + y2 a a
0 c c 0 c
√ 1
Solution 2.3. Put sin x = t, 0 < t ≤ 1 =⇒ cos xdx = √ dt.
2 t
Hence
Zπ/2 n−1/2 Z1
1 n−1/2
m 2
.t− /2 dt
m/2 1
I= sin x 1 − sin x cos xdx = t (1 − t )
2
0 0

1 m+1 n+1
= B , .
2 2 2
Solution 2.4. 1. Use the result of exercise 2.3 we have

7 5
Zπ/2 Γ .Γ
6 4 1 7 5 1 2 2 1 7 5
I= sin x cos xdx = B , = = B , =
2 2 2 2 Γ (6) 2 2 2
0

1 1 5!! √ 3!! √
Γ 3+ .Γ 2 +
1 2 2 1 23 . π. 22 π 3π
= = = .
2 Γ (5 + 1) 2 5! 512
√ adt
2. Put x = a t =⇒ dx = √ . Then
2 t
Z1
a2n+2 πa2n+2 (2n − 1)!!

2n n 1/2 1 −1/2 1 3
I= a t a (1 − t ) a t dt = B n+ , = .
2 2 2 2 2 (2n + 2)!!
0
√ dt
3. Put x = t, t ≥ 0 we have dx = √ and
2 t
+∞ √
5 −t 1 −1/2 1 11 9!! π
Z
I= t e t dt = Γ = .
2 2 2 26
0
1 −2/3
4. Put x3 = t =⇒ dx = t dt. Hence
3
+∞
1 t−2/3 dt

1 1 2 2π
Z
I= = B , = √ .
3 1+t 3 3 3 3 3
0
CHAPTER 3
L INE INTEGRAL
§1. L INE INTEGRAL OF THE FIRST KIND
1.1 Definition
Assume that f ( x, y) is a function of two variables which is defined in a plane curve AB.
Divide AB into n sub-curves ∆s1 , ∆s2 , . . . , ∆sn . Take an arbitrary point Mi in the curve ∆si .
n
The limit, if exists, of the sum ∑ f ( Mi )∆si when n → ∞ such that max d(∆si ) → 0, is
i =1 1≤ i ≤ n
called the line integral of the first kind (path integral) of the function f ( x, y) over the line
AB. It is denoted by
Z
f ( M )ds
AB
1.2 Calculation formulae

a) If AB is given by the equation y = y( x ), a ≤ x ≤ b, then
Z Zb q
f ( x, y)ds = f ( x, y( x )) 1 + y′2 ( x )dx (1)
AB a
b) If AB is given by the equation x = x (y), c ≤ y ≤ d, then
Z Zd q
f ( x, y)ds = f ( x ( y ), y ) 1 + x ′2 (y)dy (2)
AB c
39
40 Chapter 3. Line integral
c) If AB is given by the equation x = x (t), y = y(t), t1 ≤ t ≤ t2 , then
Z Zt2 q
f ( x, y)ds = f ( x (t), y(t)) x ′2 (t) + y′2 (t)dt (3)
AB t1
Example
Z 1.1. Calculate the following line integrals of the first kind
i) ( x + y)ds, where C is the circumference of the triangle OAB whose vertices are
C
O(0, 0), A(1, 0) and B(0, 1).
y
A
O x
Figure 3.1
Z Z Z Z
I= ( x + y)ds = ( x + y)ds + ( x + y)ds + ( x + y)ds
C OA AB BO
• In the line OA : y = 0, 0 ≤ x ≤ 1, hence
Z1
1
Z
( x + y)ds = ( x + 0)dx =
2
OA 0
• In the line AB : y = 1 − x, 0 ≤ x ≤ 1, hence
Z Z1 √ √
( x + y)ds = 2dx = 2
AB 0
• In the line BO : x = 0, 0 ≤ y ≤ 1, hence ds = dy
Z1
1
Z
( x + y)ds = ydy =
2
BO 0
√
We conclude
Z that I = 1 + 2.
ii) ( x − y)ds, where C is the circle x2 + y2 = 2x.
C √ √
Solution 1. We can divide C into twoZcurve C1 : yZ = 2x − x2 and C2 : y = − 2x − x2 ,
then apply the formula (1) to integrate ( x − y)ds, ( x − y)ds.
C1 C2
2. Line integral of the second kind 41
C1
b
O (1, 0) x
C2
Figure 3.2
Solution 2. We parameterize this circle by x = 1 + cos t, y = sin t, 0 ≤ t ≤ 2π.
Then x ′2 (t) + y′2 (t) = 1 and apply formula (3), we obtain
Z Z2π
( x − y)ds = (1 + cos t − sin t)dt = 2π
C 0
§2. L INE INTEGRAL OF THE SECOND KIND
2.1 Definition
Assume that P( x, y) and Q( x, y) are functions of two variables which are defined in
a plane curve AB. Divide AB into n sub-curves ∆s1 , ∆s2 , . . . , ∆sn whose initial points are
−−−−→ −−−−→
A0 ≡ A, A1 , . . . , An ≡ B. A vector Ai−1 Ai has coordinate Ai−1 Ai = ∆si = (∆xi , ∆yi ). Take
n
an arbitrary point Mi in the curve ∆si . The limit, if exists, of the sum ∑ [ P( Mi )∆xi +
i =1
Q( Mi )∆yi ] when n → ∞ such that max {∆xi , ∆yi } → 0, is called the line integral of the
1≤ i ≤ n
second kind of the functions P( x, y)dx + Q( x, y)dy over the line AB. It is denoted by
Z
Pdx + Qdy
AB

a) If AB is given by the equation y = y( x ); the initial and the end points correspond to
x = a and x = b respectively, then
Z Zb
Pdx + Qdy = [ P( x, y( x )) + Q( x, y( x ))y′ ( x )]dx (4)
AB a
If AB is given by the equation x = x (y); the initial and the end points correspond to
y = c and y = d respectively, then
Z Zd
Pdx + Qdy = [ P( x (y), y) x ′ (y) + Q( x (y), y)]dy (5)
AB c
If AB is given by the equation x = x (t), y = y(t); the initial and the end points
correspond to t = t1 and t = t2 respectively, then
Z Zt2
Pdx + Qdy = [ P( x (t), y(t)) x ′ (t) + Q( x (t), y(t))y′ (t)]dt (6)
AB t1
b) Green’s formula:
y
C
O x
Figure 3.3
Assume that the curve C is closed and restricts a domain D, and when going along
C, one will see the domain D on the left. Furthermore, suppose that the functions
P, Q together with their partial derivatives are continuous on D, then
∂Q ∂P
Z ZZ
Pdx + Qdy = − dxdy
∂x ∂y
C D
The direction of C defined as above is called positive direction. Inverse direction is

called negative one.
∂Q ∂P
Z
c) If = then the integral Pdx + Qdy does not depend on the path from A to
∂x ∂y
AB Z
B. We will choose a special path to calculate Pdx + Qdy.
AB
Example
Z 2.1. Calculate the following line integrals of the second kind
i) ( x2 + 2y)dx − ( x − y)dy, where C is the parabola x2 = 2y + 1 from A(1, 0) to B(−3, 5).
C
x2 − 1
y= , then dy = xdx.
2
y
B 4
A(1,0)
-3 O x
Figure 3.4
We have
Z−3h
x2 − 1 i
Z
2 2 2
( x + 2y)dx − ( x − y)dy = x + x − 1− x − x dx
2
C 1
Z−3
1 3 2 1 x4 2x3 x2
−3
= x + 2x − x − 2 dx = + − − 2x

2 2 4 3 2 1
1
8
=
3
Z
ii) ( x + y)dx + (y − x )dy, where C is the cycloid x = t − sin t, y = 1 − cos t, 0 ≤ t ≤ 2π
C
whose direction is the increasing direction of the parameter t.
dx = (1 − cos t)dt, dy = sin tdt, we have
Z2πh i
I= (t − sin t + 1 − cos t)(1 − cos t) + (1 − cos t − t + sin t) sin t dt
0
Z2πh i Z2πh i
= t − t cos t − t sin t + 2 − 2 cos t dt = t − t cos t − t sin t + 2 dt
0 0
t2 2π
= − t sin t − cos t + t cos t − sin t + 2t = 2π 2 + 6π

2 0
Example
Z
2.2. Calculate the following line integrals of the second kind
i) y2 dx − ( x2 y − x3 )dy, where C is the circle x2 + y2 = 4x with positive direction.
C
Because the circle is closed and the functions P = y2 , Q = x3 − x2 y are continuous in
R2 , we can apply Green’s formula

Z ZZ
2 2 3 2
y dx − ( x y − x )dy = 3x − 2xy − 2y dxdy
C x2 +y2 ≤4x
ZZ
=3 x2 dxdy
x2 +y2 ≤4x
(because −2xy − 2y is odd function

ZZ with respect to y and the domain of integration is
symmetric to the Ox axis, (−2xy − 2y)dxdy = 0).
x2 +y2 ≤4x
Set x = 2 + r cos ϕ, y = r sin ϕ then
Z Z2 Z2π
2
I =3 x dxdy = 3 dr (2 + r cos ϕ)2 rdrdϕ
x2 +y2 ≤4x 0 0
Z2 r3
=3 2π 4r + dr = 60π
2
0
Z p
ii) (y cos( xy) − 3x2 y)dx + ( x cos( xy) + 2x )dy, where C is the semi-circle x = 1 − y2 from
C
A(0, 1) to B(0, −1).
P = y cos( xy) − 3x2 y; Q = x cos( xy) + 2x.
y
A
C
O x
B
Figure 3.5
The integrating curve is not closed. Add to both sides of the observing integral the inte-
gral along the line BOA, we obtain
Z Z
I+ Pdx + Qdy = Pdx + Qdy
BOA L
where L = C ∪ BOA is a closed curve with negative direction which restricts the domain
D : x2 + y2 ≤ 1, x ≥ 0.
Apply Green’s formula to the right-hand side we have

Z ZZ
I+ Pdx + Qdy = − (2 + 3x2 )dxdy
BOA D
1
ZZ
=− (2 + 3x2 )dxdy
2
x 2 + y2 ≤1
1
ZZ
=− (4 + 3x2 + 3y2 )dxdy
4
x 2 + y2 ≤1
Z2π Z1
1 11π
=− dϕ (4 + 3r2 )rdr = −
4 8
0 0
Z
In the line BOA: x = 0 then dx = 0, Q = 0, so Pdx + Qdy = 0.
BOA
11π
In conclusion we have I = − .
8
2.3 Theorem of four equivalent propositions

Assume that D is a simply connected domain, P, Q together with their partial deriva-
tives are continuous functions in D. The four following propositions are equivalent
∂Q ∂P
1. = for all ( x, y) ∈ D.
∂x ∂y
Z
2. Pdx + Qdy = 0 for all closed curve L lying in D.
L
Z
3. Pdx + Qdy = 0 does not depend on the path from A to B, for all the paths L and
AB
A, B lying in D.
4. Pdx + Qdy is an exact integrand. That means there exists a function u( x, y) such
that du = Pdx + Qdy. u can be found by the following formulae
Zx Zy Zx Zy
u( x, y) = P( x, y0 )dx + Q( x, y)dy = P( x, y)dx + Q( x0 , y)dy
x0 y0 x0 y0
xdy − ydx
Z
Example 2.3. For which α the integral does not depend on the path from
( x 2 + y2 ) α
AB
A (0, −1) to B(0, 1) which lies in the plane x < 0. For this α calculate the line integral
xdy − ydx
Z
.
( x 2 + y2 ) α
AB
−y x
P= α
,Q = 2 .
( x2
2
+y ) ( x + y2 ) α
The integral does not depend on the path if and only if
∂Q ∂P
= for all ( x, y), x < 0
∂x ∂y
( x2 + y2 )α − 2αx2 ( x2 + y2 )α−1 ( x2 + y2 )α − 2αy2 ( x2 + y2 )α−1
⇔ = − ∀( x, y), x < 0
( x2 + y2 )2α ( x2 + y2 )2α
⇔α = 1
For α = 1, we choose a special path from A(0, −1) to B(0, 1), it is the circle x = cos t, y =
3π π
sin t, where t is from to .
2 2
π π
Z2 Z2
xdy − ydx cos td(sin t) − sin td(cos t)
Z
= = dt = −π
x 2 + y2 sin t2 + cos t2
AB 3π 3π
2 2
2.4 Area of a plane domain

Assume that D is the domain restricted by a closed curve C. The area of D is
1
I
S= ( xdy − ydx )
2
C
§3. E XERCISES
Exercise
Z 3.1. Calculate the following line integrals of the first kind
p
a) xy2 ds, where C is the curve x = 1 − y2 , −1 ≤ y ≤ 1.
C
Z p
b) x2 + y2 ds, where C is the circle x2 + y2 = ax.
ZC
4 4 2 2 2
c) ( x 3 + y 3 )ds, where C is the astroid x 3 + y 3 = a 3 .
CZ
d) xyds, where C is the hyperbol x = a cht, y = asht, (0 ≤ t ≤ 1).
ZC
e) ( x2 + y2 + z2 )ds, where C is the helix x = a cos t, y = a sin t, z = bt, (0 ≤ t ≤ 2π).
C
Exercise
Z 3.2. Calculate the following line integrals of the second kind
p
a) ydx + x2 dy, where C is the curve x = 1 − y2 from (0, 1) to (0, −1).
C
3. Exercises 47
Z
b) ( x2 + y2 )dx + ( x2 − y2 )dy, where C is the path y = 1 − |1 − x |, 0 ≤ x ≤ 2, whose
C
direction is increasing direction of the variable x.
( x + y)dx − ( x − y)dy
Z
c) , where C is the circle x2 + y2 = a2 , whose direction is
x 2 + y2
C
counter-clockwise. 
Z  x = 1 − cos t
d) ydx + (2 + 3x )dy, L is the curve , 0 ≤ t ≤ 2π, whose direction is
y = t − sin t
L
increasing direction of the parameter t.
Exercise 3.3. Use Green’s formula to calculate the following line integrals of the second
kind
a) xy2 dy − x2 ydx, where C is the circle x2 + y2 = a2 .
H
C h i
cos( x2 + y2 ) x4 dy + (y3 + 2y2 )dx .
H
b)
x 2 + y2 = π
H (sin x − y)dx + ( x + sin y)dy
c) 2 + y2
, where C is the circle x2 + y2 = 1 with positive direc-
C x
tion. Z
d) ey [(sin x − 1)dx + (1 − cos x )dy], where OA is the curve x = sin y from O(0, 0) to
OA
A(0, π ).
x3
Z
e) [ xy4 + x2 − ye xy ]dx + [ + xy2 − x − xe xy ]dy, where C is the semi-circle x2 + y2 =
3
C
1, y ≤ 0 from A(−1, 0) to B(1, 0).
f) e− x arcsin( xy)dx + ey arccos( xy)dy, where C is the curve | x | + |y| = 1.
H
C
H ( x + y)dx − ( x − y)dy
g) 2 + y2
for an arbitrary number a > 0, where Ca is the ellip ax2 +
Ca x
y2 = 1.
Exercise 3.4. Check out that the elements of the integration are exact integrands, then
calculate the following line integrals
(2,2π
Z )
y2 y y y y
a) 1 − 2 cos dx + sin + cos dy, along the curves lying in the plane x >
x x x x x
(1,π )
0.
(Z6,8)
xdx + ydy
b) p , along the curves lying in the plane y > 0.
x 2 + y2
(0,1)
(Z1,2)
ydx − xdy
c) along the curves which do not intersect with the axis Oy.
x2
(2,1)
(1,1,1
Z )
d) 2xdx − 3y2 dy − 4z3 dz.
(2,−1,0)
Exercise 3.5.
a) Find a, b such that the integral

Z
[ axy3 − 5y2 + y cos xy]dx + [3x2 y2 + bxy + x cos xy]dy
AB
does not depend on the line from A to B. Find the function u( x, y) such that the integrand
is du.
b) Find a function h( x ) such that the integral

Z
h( x )[( x sin y + y cos y)dx + ( x cos y − y sin y)dy]
AB
does not depend on the line from A to B. Use the found function h( x ) calculate the above
integral when the points A(0, π ) and B(1, 2π ).
c) Find a function h(y) such that the integral

Z
h(y)[y(2x + y3 )dx − x (2x − y3 )dy]
AB
does not depend on the line from A to B. Use the found function h(y) calculate the above
integral when the points A(0, 1) and B(−3, 2).
d) Find a function h( xy) such that the integral

Z
h( xy)[2y( x3 − y3 )dx + x (y3 − 4x3 )dy]
AB
does not depend on the line from A to B. Use the found function h(y) calculate the above
1
integral when the points A(1, 1) and B( , 2).
2
Exercise 3.6. Calculate the area of the following plane domain

a) The domain D is restricted by the vertical aixs Oy and the curve x = x (y) whose
representation is x = a(1 − cos t), y = a(t − sin t), 0 ≤ t ≤ 2π.
y2
b) The domain D is defined by x2 + ≤ 1, x ≥ 0.
4
4. Solution 49
§4. S OLUTION
Solution 3.1.
Z1
y dx 2
Z
a) x′ (y) = −p , then ds = , xy2 ds = y2 dy =
1 − y2 x 3
C −1
a a
b) x = (1 + cos t), y = sin t, 0 ≤ t ≤ 2π .
2 2
2π
Z p
a 2 Z
t
x2 + y2 ds = | cos |dt = 2a2
2 2
C 0
c) x = a cos3 t, y = a sin3 t, 0 ≤ t ≤ 2π
Z Z2π
4 4 7 7
( x + y )ds = 3a
3 3 3 (cos4 t + sin4 t)| sin t cos t|dt = 4a 3
C 0
Z1
sh2t √ a3 h
Z i
3
d) xyds = a2 a ch2tdt = (ch2) 2 − 1
2 6
C 0
Z2π √ √ 8
Z
e) ( x2 + y2 + z2 )ds = ( a2 + b2 t2 ) a2 + b2 dt = a2 + b2 2πa2 + π 3 b2
3
C 0
Solution 3.2.
p
a) x = 1 − y2 from (0, 1) to (0, −1).
π π
Set x = cos t, y = sin t, t is from to − , and
2 2
π
−2
3π − 10
Z Z
ydx + x2 dy = (− sin2 t + cos3 t)dt =
6
C π
2
Z1
2
b) OA : 0 ≤ x ≤ 1, y = x: 2x2 dx =
3
0
Z2
2
AB : 1 ≤ x ≤ 2, y = 2 − x: 2(2 − x )2 dx =
3
1
4
Z
( x2 + y2 )dx + ( x2 − y2 )dy =
3
C
c) x = a cos t, y = a sin t, t is from 2π to 0 (counter-clockwise direction).
Z0
( x + y)dx − ( x − y)dy
Z
= (−dt) = 2π
x 2 + y2
C 2π
Z Z2π
d) ydx + (2 + 3x )dy = [(t − sin t) sin t + (2 + 3 − 3 cos t)(1 − cos t)]dt = 10π
L 0
Solution 3.3.
Za Z2π
π 4
ZZ
xy2 dy − x2 ydx = ( x2 + y2 )dxdy = r3 dr
H
a) dϕ = a
C 2
x 2 + y2 ≤ a2 0 0
h i ZZ
x4 dy + (y3 + 2y2 )dx = − (4x3 − 3y2 − 4y)dxdy
H
b) Ib = −
x 2 + y2 = π
x 2 + y2 ≤ π
3 3
ZZ ZZ
Ib = 3 y2 dxdy = ( x2 + y2 )dxdy = π 5
2 4
x 2 + y2 ≤ π x 2 + y2 ≤ π
H ZZ
c) Ic = (sin x − y)dx + ( x + sin y)dy = 2 dxdy = 2π
C
x 2 + y2 ≤1
Z
ey [(sin x − 1)dx + (1 − cos x )dy] = [(sin x − 1)ey dx + ey (1 − cos x )dy] = J
H
d) Id +
C
AO
Zπ Z y
sin
1 π
J= dy ey dx = ( e + 1)
2
0 0 Z
AO : x = 0 then dx = 0, Q = 0, ey [(sin x − 1)dx + (1 − cos x )dy] = 0.
AO
1
Id = (eπ + 1)
2
x3
Z Z
e) Ie + [ xy4 + x2 − ye xy ]dx + [ + xy2 − x − xe xy ]dy = = J ( AmBOA has negative
3
BOA AmBOA
direction).
Z−1
x3 2
Z
BOA : y = 0, dy = 0: [ xy4 + x2 − ye xy ]dx + [ + xy2 − x − xe xy ]dy = x2 dx = −
3 3
ZZ BOA ZZ 1
J=− ( x2 + y2 − 1 − 3xy2 − 2x )dxdy = − ( x2 + y2 − 1)dxdy, ( D : x2 + y2 ≤ 1, y ≤ 0)
D D
Z2π Z1
π π 2
J= dϕ r (1 − r2 )dr = , Ie = +
4 4 3
π 0
−ey y − e− x x −(e x + e−x ) x

ZZ ZZ
f) I f = p dxdy = p dxdy = 0
1 − x 2 y2 1 − x 2 y2
| x |+|y|≤1 | x |+|y|≤1
(we use symmetric role of x, y in the domain then odd property with respect to x of the
−(e x + e−x ) x
function p )
1 − x 2 y2
4. Solution 51
g) ∃r > 0 : B = B(O, r ) ⊂ { ax2 + y2 ≤ 1}

( x + y)dx − ( x − y)dy
Z
= 0, so
x 2 + y2
Ca ∪ B −
( x + y)dx − ( x − y)dy
Z
Ig = , B(O, r ) : x = r cos ϕ, y = r sin ϕ, 0 ≤ ϕ ≤ 2π , then
x 2 + y2
B+
( x + y)dx − ( x − y)dy
Z
Ig = = −2π
x 2 + y2
B+
Solution 3.4.
∂Q ∂P 2y y y2 y
a) = = − 2 cos − 3 sin
∂x ∂y x x x x
Choose the path y = πx, 1 ≤ x ≤ 2, then I = 1.
∂Q ∂P − xy
b) = = 3
∂x ∂y ( x 2 + y2 ) 2
Choose the path AB, BC: A(0, 1); B(0, 8); C (6, 8), then I = 9.
∂Q ∂P 1
c) = = 2
∂x ∂y x
3
Choose the path AB, BC: A(2, 1); B(2, 2); C (1, 2), then I = − .
2
d) u = x2 − y3 − z4 satisfies du = 2xdx − 3y2 dy − 4z3 dz
I = u(2, −1, 0) − u(1, 1, 1) = 6
Solution 3.5.
a) a = 2, b = −10; u = x2 y3 − 5xy2 + sin xy + C
b) h′ ( x ) = h( x ); h( x ) = Ce x . Choose C = 1, h( x ) = e x .
(1,2π
Z ) (Z
1,π ) (1,2π
Z )
Ib = = + = (2π + 1)e
(0,π ) (0,π ) (1,π )
1 x2
c) h′ (y)y + 3h(y) = 0, h(y) = , u ( x, y ) = + xy
y3 y2
15
Ic = u(−3, 2) − u(0, 1) = −
4
1 2x y
d) h′ ( xy) xy + 3h( xy) = 0, h( xy) = 3 3 , u( x, y) = 2 + 2
x y y x
1 21
Id = u(1, 1) − u( , 2) = −
2 4
Solution 3.6.
1
Z
a) Sa = ( xdy − ydx )
2
C ∪ AO
1
Z
AO : x = 0, then: ( xdy − ydx ) = 0.
2
AO
The cycloids (C ) : x = a(1 − cos t), y = a(t − sin t), 0 ≤ t ≤ 2π ,
1
Z
Sa = ( xdy − ydx ) = 3πa2
2
C
1
Z
b) Sb = ( xdy − ydx )
2
E ∪ BOA
1
Z
BOA : x = 0, then: ( xdy − ydx ) = 0.
2
BOA
π π
The ellip x = cos t, y = 2 sin t, − ≤t≤ ,
2 2
1
Z
Sb = ( xdy − ydx ) = π
2
E
Fi
CHAPTER 4
S URFACE INTEGRAL
§1. S URFACE INTEGRAL OF THE FIRST KIND
1.1 Definition
Given a function f ( x, y, z) which defines in a surface S. Divide S into n sub-surfaces
∆S1 , ∆S2 , . . . , ∆Sn . In each ∆Si take an arbitrary point Mi . The limit, if exists, of the sum
n
∑ f ( Mi )∆Si when n → ∞ and max d(∆Si ) → 0 is called the surface integral of the first
i =1 1≤ i ≤ n
kind of the function f ( M ) in the surface S. This integral is denoted by
ZZ
f ( x, y, z)dS
S

Assume that S is the surface
z = z( x, y); (( x, y) ∈ D ⊂ R2 ),
where z( x, y) is a continuously differentiable function then

s
ZZ ZZ ∂z 2 ∂z 2
f ( x, y, z)dS = f ( x, y, z( x, y)) 1+ + dxdy
∂x ∂y
S D
ZZ
Example 1.1. Calculate the surface integral of the first kind I1 = ( x2 + y2 + z2 )dS,
S
where S is the sphere x2 + y2 + z2 = a2 .
53
54 Chapter 4. Surface integral
S1
O y
D : x 2 + y2 ≤ a2
S2
x
Figure 4.1
We divide S into two pieces S = S1 ∪ S2 ,

 
 z = p a2 − x 2 − y2  z=−
p
a2 − x 2 − y2
( S1 ) : ( S2 ) :
 x 2 + y2 ≤ a2  x 2 + y2 ≤ a2
q a
We can calculate that 1 + (z′x )2 + (z′y )2 = p , then
a2 − x 2 − y2
a
ZZ ZZ ZZ
I1 = + =2 a2 p dxdy
a2 − x 2 − y2
S1 S2 x 2 + y2 ≤ a2
Change into polar coordinate x = r cos ϕ, y = r sin ϕ, we have
Z2π Za
r
I1 = 2a3 dϕ √ dr = 4πa4
a2 − r2
0 0
§2. S URFACE INTEGRAL OF THE SECOND KIND
2.1 Definition
Given a function R( x, y, z) which defines in an oriented surface S. Divide S into n sub-
surfaces ∆S1 , ∆S2 , . . . , ∆Sn . In each ∆Si take an arbitrary point Mi . Denote by Di the area
of the orthogonal projection of ∆Si in the plane Oxy, its sign is (+) or (−) if the outnormal
vector at Mi make an acute or obtuse angle respectively with the positive direction of Oz.
n
The limit, if exists, of the sum ∑ R( Mi ) Di when n → ∞ and max d(∆Si ) → 0 is called the
i =1 1≤ i ≤ n
2. Surface integral of the second kind 55
surface integral of the second kind of the function R( M ) in the surface S with respect to
two variables ( x, y). This integral is denoted by
ZZ
R( x, y, z)dxdy
S
Similarly we define the surface integral of the second kind of the function P( x, y, z) with
respect to two variables (y, z) and of the function Q( x, y, z) with respect to two variables
(z, x ). In general, we consider the integral
ZZ
I= Pdydz + Qdzdx + Rdxdy
S

1. Assume that we want to calculate
ZZ
I1 = P( x, y, z)dydz,
S
where S is a surface given by the equation x = x (y, z); (y, z) ∈ D ⊂ Oyz. We have
ZZ
I1 = ε P( x (y, z), y, z)dydz
D
where ε = 1 if the angle between the outnormal vector and the positive direction of
the axis Ox is an acute angle, and ε = −1 if the angle between the outnormal vector
and the positive direction of the axis Ox is an obtuse angle.
2. We want to calculate
ZZ
I= P( x, y, z)dydz + Q( x, y, z)dzdx + R( x, y, z)dxdy
S
z −
→
n = (cos α, cos β, cos γ)
O y
x Figure 4.2
We find the unit outnormal vector −

→
n = (cos α, cos β, cos γ), then
ZZ
I= ( P cos α + Q cos β + R cos γ)dS,
S
where cos α, cos β, cos γ are the direction cosine of vector −

→
n.
3. Ostrogradsky’s formula: if S is a closed surface which restricts a volume V, and

P, Q, R are continuous together with their partial derivatives in V, then
∂P ∂Q ∂R
ZZZ
I= + + dxdydz
∂x ∂y ∂z
V
Example 2.1. Calculate the following surface integrals of the second kind
ZZ
a) I2 = ydzdx, where S is the outside of the sphere x2 + y2 + z2 = a2 , y ≤ 0.
S √
We rewrite y = − a2 − x2 − z2 .
y
D : x 2 + z2 ≤ a2
O x
S −
→
n ( x, y, z)
z
Figure 4.3
The outnormal vector makes an obtuse angle with the positive direction of the axis Oy,
then ε = −1. Hence
ZZ p
I2 = a2 − x2 − z2 dxdz
x 2 + z2 ≤ a2
Z2π Za p
2πa3
= dϕ a2 − r2 rdr =
3
0 0
dydz dzdx dxdy x 2 y2 z2

ZZ
b) I3 = + + , where S is the outside of the ellipsoid 2 + 2 + 2 = 1.
x y z a b c
S
The normal vector of the ellipsoid at the point M ( x, y, z) which points outwards is
s
−
→
2x 2y 2z x 2 y2 z2
n = 2 , 2 , 2 ; |− →n|=2 + 4 + 4;
a b c a4 b c
2. Surface integral of the second kind 57
and the unit normal vector is −

→
−
→ n
n0 = −→
|n|
We use the second formula and obtain
1 1 1 1
ZZ
I3 = + + dS
a2 b2 c2
q
x2 y2 z2
S a4
+ b4
+ c4
We divide S into two pieces S = S1 ∪ S2 ,

2 2 2 x 2 y2 z2
 x +y +z =1
 
 + 2+ 2 =1
( S1 ) : a2 b2 c2 ( S2 ) : a2 b c
z≥0 z≤0
 
For both integrals in S1 and S2 we have
x2 y2
a4
+ b4
1 + (z′x )2 + (z′y )2 = 1+c 2
x2 y2
1− a2
− b2
Hence
1 1 1 c
ZZ
I3 = 2 + + dxdy
a2 b2 c2
q
x2 y2
x2 y2 1− a2
− b2
+ 2 ≤1
a2 b
Z2πZ1
1 1 1 abcr
=2 + + √
a2 b2 c2 1 − r2
0 0
1 1 1
= 4πabc 2 + 2 + 2
a b c
In this example, although the surface is closed but we can not apply Ostrogradsky’s for-
1 1 1 x2 y2 z2
mula because the functions , , are not continuous in the domain 2 + 2 + 2 ≤ 1
x y z a b c
restricted by the ellipsoid.
ZZ
c) I4 = x3 dydz + y3 dzdx + z3 dxdy, where S is the outside of the sphere x2 + y2 + z2 = a2 .
S
We can apply Ostrogradsky’s theorem, then
ZZZ
I4 = 3 ( x2 + y2 + z2 )dxdydz
x 2 + y2 + z2 ≤ a2
Change into spherical coordinate x = r sin θ cos ϕ, y = r sin θ sin ϕ, z = r cos θ, we have
Z2π Zπ Za
12πa5
I4 = 3 dϕ sin θdθ r4 dr =
5
0 0 0
2.3 Stokes’ formula

−
→
n
z
L
S
O y
x
Figure 4.4
Assume that C is a closed, single, and piecewise smooth curve which restricts a piece-
wise smooth surface S; and P, Q, R are continuously differentiable functions in a domain
containing S. We have

cos α cos β cos γ

∂ ∂ ∂
I ZZ
Pdx + Qdy + Rdz = dS,

∂x ∂y ∂z

S S
P Q R
where cos α, cos β, cos γ are direction cosine of the normal vector of S; and if one stands
along the direction of S and follows C in its direction, then he will see S on the left.
Example 2.2. Using Stokes formula, calculate the following line integral of the second
kind
x z
a) I5 = (y − z)dx + (z − x )dy + ( x − y)dz, where C is the ellip x2 + y2 = a2 , + = 1,
H
C a h
(a > 0, h > 0), in the anticlockwise direction if one see from the positive direction of the
axis Ox.
z
−
→
n
O y
a
Figure 4.5
x
3. Exercises 59
x z
C is given, we choose S to be the surface + = 1, x2 + y2 ≤ a2 which is restricted by
a h
C. The outnormal vector of S makes an acute angle with the positive direction of Oz axis.
1 1 h a
Hence the normal vector is , 0, , then the unit outnormal vector is √ , 0, √ .
a h a2 + h2 a2 + h2
Apply Stokes’ formula we have
2( a + h )
ZZ
I5 = − √ dS
a2 + h2
S
x
For S, we substitute z = h 1 − and x2 + y2 ≤ a2 , then
a
s
2( a + h ) h2
ZZ
I5 = − √ 1+ dxdy = −2πa( a + h)
a2 + h2 a2
x 2 + y2 ≤ a2
ydx + zdy + xdz, where C is the circle x2 + y2 + z2 = a2 , x + y + z = 0), in the

H
b) I6 =
C
anticlockwise direction if one see from the positive direction of the axis Ox.
x
−
→
n
O C z
y Figure 4.6
We choose S to be the surface x2 + y2 + z2 ≤ a2 , x + y + z = 0. The direction of C leads to
the fact that the outnormal vector of S make an acute angle with the positive direction
1 1 1
Ox-axis, then the unit outnormal vector is −
→n = √ ,√ ,√ .
3 3 3
Apply Stokes’ formula, we have
ZZ √ √
I6 = − 3dS = −π 3a2
S
§3. E XERCISES
ZZ 4.1. Calculate the following surface integral of the first kind
Exercise
a) ( x + y + z)dS, where S is the semi-sphere x2 + y2 + z2 = a2 , z ≥ 0.
S
ZZ p
b) ( x2 + y2 )dS, where S is the boundary of the object x2 + y2 ≤ z ≤ 1.
ZZS
dS
c) , where S is the surface x + y + z = 1, x ≥ 0, y ≥ 0, z ≥ 0.
(1 + x + y )2
ZSZ p
d) ( xy + yz + zx )dS, where S is the cone z = x2 + y2 intersected by the cylinder
S
x2 + y2 = 2ax, ( a > 0).
ZZ 4.2. Calculate the following surface integral of the second kind

Exercise
a) z( x2 + y2 )dxdy, where S is the semi-sphere x2 + y2 + z2 = 1, z ≥ 0, which points
S
outwards.
y2
ZZ
b) ydzdx + z2 dxdy, where S is the inside of the ellipsoid x2 + + z2 = 1, x ≥ 0, y ≥
4
S
0, z ≥ Z0.
Z
c) x2 dydz + y2 dzdx + z2 dxdy, where S is the outside of the sphere ( x − a)2 + (y −
S
b)2 + (ZZz − c)2 = R2 .
d) (y − z)dydz + (z − x )dzdx + ( x − y)dxdy, where S is the outside of the cone x2 +
S
y2 = zZ2Z, 0 ≤ z ≤ h.
e) x2 dydz + y2 dzdx + z2 dxdy, where S is the outward boundary of the parallelepiped
S
0 ≤ x Z≤
Z a, 0 ≤ y ≤ a, 0 ≤ z ≤ a.
f) y2 zdxdy + xzdydz + x2 ydxdz, where S is the outside of the domain x ≥ 0, y ≥
S
0, x2 +ZZy2 ≤ 1, 0 ≤ z ≤ x2 + y2 .
g) ( x − y + z)dydz + (y − z + x )dxdz + (z − x + y)dxdy where S is the outside of the
S
ZZ | x − y + z| + |y − z + x | + |z − x + y| = 1.
surface
h) x3 dydz + y2 dzdx + zdxdy where S is the boundary of the cylinder x2 + y2 ≤ 1, − h ≤
S
z ≤ h, (h is a positive constant), which points outwards.
y3 z3 y2
ZZ
i) x3 dydz + dzdx + dxdy, where S is the boundary of the semi-ellipsoid x2 + +
2 3 2
S
z2
≤ 1, z ≥ 0 which points outwards.
3
§4. S OLUTION
Solution 4.1.
4. Solution 61
p
a) z = a2 − x2 − y2 ; x2 + y2 ≤ a2 , then
a
ZZ q
Ia = ( x + y + a2 − x 2 − y2 ) p dxdy = πa3
a2 − x2 − y2
x 2 + y2 ≤ a2
π
ZZ
b) z = 1: I1 = ( x2 + y2 )dxdy =
2
x 2 + y2 ≤1
√
p q √ ZZ √ π 2
z= x2 + y2 , 1 + (z′x )2 + (z′y )2 = 2, I2 = ( x2 + y2 ) 2dxdy =
2
x 2 + y2 ≤1
π √
Ib = (1 + 2)
2
c) z = 1 − x − y, 0 ≤ x ≤ 1, 0 ≤ y ≤ 1 − x
Z1 1Z− x √
3 √ 1
Ic = dx 2
dy = 3(ln 2 − )
(1 + x + y ) 2
0 0
p
d) z = x2 + y2 , then
ZZ q √
Id = xy + ( x + y) x 2 + y2 2dxdy
x2 +y2 ≤2ax
√ ZZ q
= 2 x x2 + y2 dxdy
x2 +y2 ≤2ax
(we subtract some functions which are odd with respect to√the variable y because the
64a4 2
domain is symmetric to the axis Ox). We conclude Id =
15
Solution 4.2.
p
a) z = 1 − x2 − y2 ; ( x, y) : x2 + y2 ≤ 1; ε = 1,
4π
ZZ q
Ia = ( x2 + y2 ) 1 − x2 − y2 dxdy =
15
x 2 + y2 ≤1
ZZ √
b) I1 = ydzdx; where S : y = 2 1 − x2 − z2 ; ( x, z) ∈ D1 = { x2 + z2 ≤ 1, x, z ≥ 0}; ε = 1
S
2π
ZZ p
I1 = 2 1 − x2 − z2 dxdz =
3
D1
r
y2 y2
ZZ
I2 = z2 dxdy; where S : z = 1 − x2 − ; ( x, y) ∈ D2 = { x2 + ≤ 1, x, y ≥ 0}; ε = 1
4 4
S
s
y2 π
ZZ
I2 = 1 − x2 − dxdy =
4 4
D2
7π
Ic =
12
x − a y − b z − c
c) −
→
n = ; ;
R R R
x − a 2 y − b 2 z − c 2 8πR3
ZZ
Ic = x + y + z dS = ( a + b + c)
R R R 3
S
p
(S : z = c ± R2 − ( x − a )2 − ( y − b )2 ; R2 ≥ ( x − a )2 + ( y − b )2 )
x y −1
d) −
→n = √ , √ ,√ (it makes an obtuse angle with the positive direction of Oz axis).
z 2 z 2 2
x y −1 1
ZZ ZZ
Id = √ (y − z) + √ (z − x ) + √ ( x − y) dS = √ (y − x )dS,
z 2 z 2 2 2
S S
S : x 2 + y2 = z2 , 0 ≤ z ≤ h. ZZ
Id = (y − x )dxdy = 0
x 2 + y2 ≤ h2
Za Za Za
e) Ie = 2 dx dy ( x + y + z)dz = 3a4
0 0 0
ZZZ
f) I f = (y2 + z + x2 )dxdydz; V : x ≥ 0, y ≥ 0, x2 + y2 ≤ 1, 0 ≤ z ≤ x2 + y2 .
V
Change into cylinderal coordinate, x = r cos ϕ; y = r sin ϕ, z = z;
π 2
Z2 Z1 Zr
π
If = dϕ dr (r2 + z)rdz =
8
0 0 0
ZZZ
g) Ig = 3dxdydz = 1; V : | x − y + z| + |y − z + x | + |z − x + y| ≤ 1.
V
ZZZ
h) Ih = (3x2 + 2y + 1)dxdydz; V : x2 + y2 ≤ 1; −h ≤ z ≤ h.
V
1 7πh
ZZZ ZZZ
Ih = (3x2 + 1)dxdydz = (3x2 + 3y2 + 2)dxdydz =
2 2
V V
2 2
 x2 + y + z ≤ 1

y2 z2
ZZZ
i) Ii = 3 x2 + + dV , where V : 2 3
2 3
z≥0

V
√ √
Change into spherical coordinate, x = r sin θ cos ϕ; y = r sin θ sin ϕ 2, z = r cos θ 3;
2π Z1 √
√ Z Z
π 4 6 6π
Ii = 3 6 dϕ sin θdθ r dr =
2 5
0 0 0
CHAPTER 5
F IELD THEORY
§1. S CALAR FIELD
A scalar field in R3 is a function u : R3 → R, u( x, y, z) ∈ R.

Its gradient vector is
∂u ∂u ∂u
grad u = , ,
∂x ∂y ∂z
At each point of the scalar field, gradient vector of u is the direction in which u varies
fastest.
−
→
Assume that l = (cos α, cos β, cos γ) is a unit vector. The derivatives of u with respect
−
→
to l direction is
∂u −
→ ∂u ∂u ∂u
−→ = grad u. l = ∂x cos α + ∂y cos β + ∂z cos γ
∂ l
Example 1.1. Given a scalar field u( x, y, z) = xy2 z3 + 2xy − z2 . Calculate grad u(1, 1, 1)
and the derivatives of u with respect to the direction grad v(1, 1, 1) at the same point,
where v( x, y, z) = z2 sin(πxy).

grad u = y2 z3 + 2y, 2xyz3 + 2x, 3xy2 z2 − 2z
⇒ grad u(1, 1, 1) = (3, 4, 1)

grad v = z2 πy cos(πxy), z2 πx cos(πxy), 2z sin(πxy)
−
→ 1 1
⇒ grad v(1, 1, 1) = (−π, −π, 0), l = − √ , − √ , 0
2 2
∂u 1 1 7
−→ (1, 1, 1) = − √ .3 − √ .4 + 0.1 = − √
∂ l 2 2 2
63
64 Chapter 5. Field theory
§2. V ECTOR FIELD
A vector field in R3 is a vector function

−
→ −
→ −
→ −
→
F ( x, y, z) = P( x, y, z) i + Q( x, y, z) j + R( x, y, z) k ,
−→ −→ − →
where i , j , k are orthonormal basis of R3 .
Divergence of this vector field is a scalar field
−
→ ∂P ∂Q ∂R
div F = + +
∂x ∂y ∂y
Rotation vector of this vector field is a vector field

−
→ − → − →
i j k
−
→ ∂

∂ ∂ ∂R ∂Q ∂P ∂R ∂Q ∂P
rot F = = − ; − ; −
∂x ∂y ∂z ∂y ∂z ∂z ∂x ∂x ∂y

P Q R
−
→
Flow field of a vector field F through an oriented surface S is
ZZ
Pdydz + Qdzdx + Rdxdy
S
Circulation of a vector field along a directed curve L is

Z
Pdx + Qdy + Rdz
L
We often use Ostrogradsky and Stokes formulae to calculate flow field and circulation of
−
→
F.
Example 2.1.
−
→
a) Calculate the flow field of the vector field F = ( x, y, z) through the base of the cone
V : x2 + y2 ≤ z2 , 0 ≤ z ≤ h in the outward direction.
−
→
The flow field of F is ZZ
I= xdydz + ydzdx + zdxdy
S+
where S : z = h, x2 + y2 ≤ h2
ZZ ZZ
I= xdydz + ydzdx + zdxdy = hdxdy = πh3
S x 2 + y2 ≤ h2
3. Exercises 65
−
→
b) Calculate the circulation of the vector field F = (−y, x, 1) along the circle C : x2 + y2 =
1, z = 0.
The circulation is
Z Z
J= −ydx + xdy − dz = −ydx + xdy
C x 2 + y2 =1
Apply Green’s theorem we have

ZZ
J=2 2dxdy = 2π
x 2 + y2 ≤1
−
→ −
→
A vector field F is a potential vector if there exists a scalar field u such that grad u = F .
−
→
u is called the potential function of F .
−
→ −
→ − →
Necessary and sufficient condition for F to be a potential field is rot F = 0 . Then we
find u by one of the following formulae
( x,y,z
Z )
u( x, y, z) = Pdx + Qdy + Rdz + C
( x0 ,y0 ,z0 )
Zx Zy Zz
= P( x, y0 , z0 )dx + Q( x, y, z0 )dy + R( x, y, z)dz + C
x0 y0 z0
Zx Zy Zz
= P( x, y, z)dx + Q( x0 , y, z)dy + R( x0 , y0 , z)dz + C
x0 y0 z0
...
−
→ −→ −→ −
→
Example 2.2. Examine that F = cos( x2 + 2y2 − 3z2 )( x i + 2y j − 3z k ) is a potential
field, and find its potential function.
−→
It is easy to check that rotF = 0 . We choose x0 = y0 = z0 =, the potential function is
Zx Zy Zz
2 2 2
u( x, y, z) = x cos x dx + 2y cos( x + 2y )dy − 3z cos( x2 + 2y2 − 3z2 )dz + C
0 0 0
1
= sin( x2 + 2y2 − 3z2 ) + C
2
§3. E XERCISES
p
Exercise 5.1. Let u = ln(1 + x2 + y2 + z2 ) be a scalar field and a point A(1, 2, −2) in R3 .
−→
Calculate the derivatives of u with respect to the direction OA at A. For which direction
−
→ ∂u
l that − ( A ) reaches its maximum?

→
∂ l
Exercise 5.2. Given a field u = e− xyz ( x2 − y). Calculate the derivatives of u with respect
to the direction grad u(0, 1, 2) at the point A(1, −1, 0).
x
Exercise 5.3. Find the angle between two gradient vectors of the scalar field u =
x 2 + y2 + z2
at the points A(1, 2, 2) and B(−3, 1, 0).
−
→
Exercise 5.4. Calculate the flow field of the vector field F = ( x, y, z) through the bound-
p
ary of the cone z ≤ 1 − x2 + y2 , 0 ≤ z ≤ 1.
−
→ p −
→ −
→
Exercise 5.5. Calculate the flow field of the vector field F = x2 y2 + z2 i + x3 z j +
−
→ y2 + z2
x2 y k through the outside boundary of the domain x2 + ≤ 1, x ≥ 0.
4
−
→ −
→ −
→
Exercise 5.6. Calculate the circulation of the vector field F = x (y + z) i + y( x + z) j +
−
→
z( x + y) k along the intersection curve of the sphere x2 + y2 + z2 = R2 , z ≥ 0 and the
cylinder x2 + y2 + y = 0.
Exercise 5.7. Prove that the following vector fields are potential fields and find their
potential function
−
→ −
→ −
→ −
→
a) F = yz(4x3 + y3 + z3 ) i + zx ( x3 + 4y3 + z3 ) j + xy( x3 + y3 + 4z3 ) k .
−
→ −
→ −→ −→
b) F = yz(2x + y + z) i + zx ( x + 2y + z) j + xy( x + y + 2z) k .
−
→ −→ −
→ −
→
c) F = ( x + y) i + ( x + z) j + (y + z) k .
§4. S OLUTION
p
Solution 5.1. u = ln(1 + x 2 + y2 + z2 ).
1 1 2 2
grad u = p p ( x, y, z) ⇒ grad u(1, 2, −2) = , ,−
(1 + x 2 + y2 + z2 ) x 2 + y2 + z2 12 12 12
−→ −
→ 1 2 2
OA = (1, 2, −2) ⇒ l = , , −
3 3 3
∂u 1 ∂u −
→
−→ = 4 ; − → reaches its maximum if and only if l = grad u( A) = (1, 2, −2).
∂OA ∂ l
4. Solution 67
Solution 5.2. u = − xyz ( x2 − y )

e
grad u = e− xyz − yz( x2 − y) + 2x, − xz( x2 − y) − 1, − xy( x2 − y) ; ⇒ grad u( A) = (2, −1, 2)
−
→ 2 1 ∂u √
grad u(0, 1, 2) = (2, −1, 0) ⇒ l = √ , − √ , 0 ; − → = 5
5 5 ∂ l
x
Solution 5.3. u =
x 2 + y2 + z2
1
2 2 2

grad u = 2 − x + y + z ; −2xy; −2xz
( x + y2 + z2 )2
1 1 8 8
grad u( A) = (7, −4, −4); grad u( B) = (−8, 6, 0) ⇒ cos ϕ = − ; ϕ = arccos −
81 100 9 9
Solution 5.4. ZZ ZZZ
I= xdydz + ydzdx + zdxdy = 3dxdydz
S V
p
V : 0 ≤ z ≤ 1− x 2 + y2 ; x 2 y2
+ ≤ 1, V is the cone whose base is a circle of radial R = 1
and whose altitude is h = 1, then I = π .
Solution 5.5.
ZZ q ZZZ q
2 3 2
I= x y2 + z2 dydz + x zdzdx + x ydxdy = 2x y2 + z2 dxdydz
S V
Set x = r cos θ, y = 2r sin ϕ sin θ, z = 2r cos ϕ sin θ , | J | = 4r2 sin θ

π
Z2 Z2π Z1
2 32π
I = 16 sin θ cos θdθ dϕ r4 dr =
15
0 0 0
Solution 5.6. We choose S : x2 + y2 + z2 = R2 ; x2 + y2 + y ≤ 0. Theoutnormal vector

−
→ x y z
makes an acute angle with the positive direction of Oz axis, then n = , , .
R R R
−→ −→→
rot F = (z − y), ( x − z), ( x − y) ; rot F .−

n = 0.
Apply Stokes’ formula, the circulation is
−→→
Z ZZ
I= x (y + z)dydz + y( x + z)dzdx + z( x + y)dxdy = rot F .−
n dS = 0
C S
Solution 5.7.
−→
a) rot F = 0
Zx Zy Zz
x0 = y0 = z0 = 0: u = 0dx + 0dy + xy( x3 + y3 + 4z3 )dz = xyz( x3 + y3 + z3 ) + C.
0 0 0
−→
b) rot F = 0
x0 = y0 = z0 = 0: u = xyz( x + y + z) + C.
−→
c) rot F = 0
Zx Zx Zz
x2 z2
x0 = y0 = z0 = 0: u = xdx + 0dy + (y + z)dz = + xy + yz + + C.
2 2
0 0 0
CHAPTER 6
S ERIES
§1. N UMBER SERIES
1.1 Definition
A number series is the expression
+∞
u1 + u2 + . . . + u n + . . . = : ∑ un
n =1
The n-th partial sum of this series is denoted by

n
Sn : = ∑ uk
k =1
+∞
If Sn tends to a finite value S when n → ∞ then ∑ un is said to be converged and its sum
n =1
+∞
is S; if not the series ∑ un is said to be diverged.
n =1
+∞ +∞
A series ∑ un is absolutely convergent if ∑ |un | is convergent. If a series is absolutely
n =1 n =1
convergent then it is convergent.
+∞ +∞
A series ∑ un is semiconvergent if it is convergent but ∑ |un | is divergent.
n =1 n =1
+∞
Example 1.1. For a 6= 0, consider the sequence ∑ aqn−1 .
n =1
If q = 1, Sn = na → ∞ when n → ∞.
qn − 1
For q 6= 1, Sn = a .
q−1
69
70 Chapter 6. Series
lim Sn exists if and only if lim qn = 0, or equivalently |q| < 1.

n→∞ n→∞
+∞ a
If |q| < 1, then the series ∑ aqn−1 is convergent whose sum is .
n =1 q−1
+∞
If |q| > 1, then the series ∑ aqn−1 is divergent.
n =1
+∞
In conclusion the series ∑ aqn−1 is convergent if and only if |q| < 1 and its sum is
n =1
a
.
q−1
∞ 1
Example 1.2. Find the sum of the series ∑ .
n=1 (3n − 2)(3n + 1)
We have
n n
1 1 1 1
Sn = ∑ (3k − 2)(3k + 1) = ∑ 3 3k − 2 3k + 1
−
k =1 k =1
1 1
= −
3 3(3n + 1)
1 1
so lim Sn = . This series is convergent and its sum is .
n→∞ 3 3
1.2 Convergent criterion

+∞
A series ∑ un does not change its convergent property if we add or substract a finite
n =1
+∞ +∞
number of terms of it. That means ∑ un and ∑ un have the same convergent property
n =1 n = n0
for a finite number n0 . Hence in the following criterion, if the statement is true for n = 1,
one should understand that n = n0 also takes effect.
1. Necessary condition
+∞
A number series ∑ un is convergent then lim un = 0.
n =1 n→∞
+∞ 1
Example 1.3. ∑ n ln 1 + is divergent because when n → ∞
n =1 n
1
un = n ln 1 + →1
n
We often use this criteria to prove divergence of a series.
2. Cauchy criteria
1. Number series 71
+∞
A number series ∑ un is convergent if and only if for an arbitrary number ε > 0,
n =1
there exists an integer N0 > 0 such that for all n ≥ N0 , p ≥ 0, we have
| u n +1 + u n +2 + . . . + u n + p | < ε
In our range of exercises we do not often use this criteria.
3. Comparison criterion
+∞ +∞
Assume that ∑ un and ∑ vn are two positive series.
n =1 n =1
Comparison criteria 1. Furthermore assume that un ≤ vn for all n ≥ 1.
+∞ +∞
If ∑ vn is convergent then ∑ un is convergent.
n =1 n =1
+∞ +∞
If ∑ un is divergent then ∑ vn is divergent.
n =1 n =1
un
Comparison criteria 2. Furthermore assume that lim =k
n→∞ vn
+∞ +∞
If 0 < k < +∞ then ∑ un and ∑ vn have the same properties of convergence or
n =1 n =1
divergence.
+∞ +∞ +∞
If k = 0 then ∑ un is convergent when ∑ vn is convergent; and ∑ vn is divergent
n =1 n =1 n =1
+∞
when ∑ un is divergent.
n =1
vn
If k = ∞, or equivalently lim = 0, we return to the case k = 0.
n→∞ un
1
Note that we often choose s to be un or vn , and keep in mind that the Riemann
n
+∞ 1
series ∑ s is convergent if and only if s > 1.
n =1 n
Example 1.4.
∞ π
a) Consider the series ∑ arctg
n =1 2n
π π
This is a positive series, and as n → ∞, we have arctg n ∼ n , and the series
2 2
∞ π ∞ 1 n
∑ n =π ∑ is convegent. Then the original series is convergent too.
n =1 2 n =1 2
√ √
∞ n+1− n−1
b) Consider the series ∑
n =1 nα
√ √
n+1− n−1 2 1
When n → ∞: α
= √ √ ∼ 1
, then we have
n ( n + 1 + n − 1) n α
n +2
α
1 1
If α > : the series is convergent; if α ≤ , the series is divergent.
2 2
∞ √
c) Consider the series ∑ e− n.
n =1
We know two important limits
an
i) lim = +∞, ( a > 1, ∀α), or nα ≤ en when n is large enough.
n→∞ nα
n
ii) lim β = +∞, (∀ β), or lnβ n ≤ n when n is large enough.
n→∞ ln n
√ √
We use the first limits: ( n)α ≤ e n when n is large enough, or equivalently,
√ α ∞ 1
e− n ≤ n− 2 , for large enough n and for all α. Choose α = 4, and the series ∑ 2 is
n =1 n
convergent; then the original series is convergent.
4. Cauchy criteria
p
n
We calculate l = lim | u n |.
n→∞
If l > 1 then the series is divergent.
If l < 1 then the series is absolutely convergent, and is convergent.
5. D’Alembert criteria
u
n +1
We calculate l = lim .
n→∞ un
If l > 1 then the series is divergent.
If l < 1 then the series is absolutely convergent, and is convergent.
Example 1.5.
∞ n2 + 5
a) ∑ has
n =1 3n
u
n +1 ( n + 1)2 + 5 1
lim = lim 2
= <1
n→∞ un n → ∞ 3( n + 5) 3
then this series is convergent due to D’Alembert criteria.
∞ an
b) ∑ 2 n
(0 < | a| 6= 1) has
n =1 (1 − a )
p
n
| a|n 1
q
n
l = lim |un | = lim =
n→∞ n → ∞ |1 − a2 | |1 − a2 |
√ 1
If 0 < | a| < 2 then l = > 1, this series is divergent due to Cauchy criteria.
|1 − a2 |
1. Number series 73
√ 1
If | a| > 2 then l = < 1, this series is convergent.
−1 a2
√ √
If | a| = 2 then lim |un | = lim n 2 = +∞, this series is divergent due to necessary
n→∞ n→∞
condition.
6. Integral criteria
Assume that f ( x ) is a positive continuous function which decreases in the interval

+∞
Z
[1, +∞) and tends to 0 as x → +∞. Then the infinite integral f ( x )dx and the
1
∞
series ∑ un , where un = f (n), have the same convergence or divergence property.
n =1
∞ 1
Example 1.6. Consider the convergence property of the series ∑
n =2 n ln2 n
1
We define f ( x ) = , for x ≥ 2. This function satisfies all the conditions in the
x ln2 x
integral criteria.
Z∞
dx 1 ∞ 1
=− =
x ln2 x ln x 2 ln 2
2
This integral is convergent then the observing series is convergent.
7. Leibnitz’s criteria for alternate series

∞
An alternate series is the series ∑ (−1)n un , where un > 0 for all n.
n =1
If un is an decreasing sequence which tends to 0 when n → ∞, then the alternate

∞
series ∑ (−1)n un is convergent.
n =1
(−1)n ∞
Example 1.7. We consider the series ∑ α
, where α > 0.
n =1 n
It is an alternate series which has
lim nα = +∞ for α > 0

n→∞
1
and α is an decreasing sequence. Use Leibnitz’s rule, we conclude that this series
n
is convergent.
1.3 Exercises
Exercise 6.1. Find the sum of the following series
∞ ∞ √ √ √ ∞
n2 1
a) ∑ n! b) ∑ ( n + 2 − 2 n + 1 + n) c) ∑ 4n2 −1
n =1 n =1 n =1
∞ ∞
1 1
d) ∑ n(n + 1)(n + 2)
e) ∑ arctg 1 + n + n2
n =1 n =1
Exercise 6.2. Prove that the following series are divergent

∞ ∞ n n2 ∞
n2 − 2n + 1 2n
a) ∑ 5n2 + (−1)n √n b) ∑ n+1 c) ∑ arctg
n
n =1 n =1 n =1
Exercise 6.3. Use comparison, D’Alembert, Cauchy and integral criterion, consider the
convergence of the following series
∞ ∞ ∞
1 n+1 p p 2
a) ∑ √ ln
n n−1
b) ∑( n4 + 2n + 1 − n4 + an) c) ∑ n3 e − n
n =1 n =1 n =1
∞ ∞ 2 2 2 ∞
n2 +5 ln 2 + ln 3 + . . . + ln n an n!
d) ∑ 3n
e) ∑ nα
f) ∑ nn , ( a 6= e)
n =1 n =1 n =1
∞ ∞ ∞ √
(2n − 1)!! 1 1 n2 n 2n
g) ∑ 22n (n − 1)!
h) ∑ n 1 −
5 n
i) ∑ n
4n − 3
n =1 n =1 n =1
∞ ∞ ∞
n + a n2 1 1
j) ∑ n+b
k) ∑ n ln n(ln ln n)2 l) ∑ n ln p , ( p > 0)
n
n =1 n =3 n =3
Exercise 6.4. Use Leibnitz criteria to consider the convergent property of the following
series
∞ ∞ √ ∞
(−1)n ln n n −1 n 5n + 3
a) ∑ b) ∑ (−1) c) ∑ (−1)n 2
n =2 n n =2 n+e n =1
n +n
Exercise 6.5. Consider the convergent property of the following series

∞ ∞ ∞
ln n 1 2n
a) ∑ n α ; ( α > 1) b) ∑ (ln n) p ; ( p > 0) c) ∑ n + 2n
n =1 n =1 n =1
∞
1 n + 1 ∞ p ∞ √
d) ∑ − ln e) ∑ sin(π n2 + a2 ); a ∈ R f) ∑ sin[π (2 + 3) n ]
n =1
n n n =1 n =1
∞ ∞ ∞
1 a n3 (n!)2
g) ∑ nα (ln n)β , (α, β > 0) h) ∑ cos
n
;a ∈ R i) ∑ n2
n =3 n =1 n =1 2
1. Number series 75
1.4 Solution
Solution 6.1.
n2 n 1 1
a) = = + ,n ≥ 2
n! ( n − 1) ! ( n − 2) ! ( n − 1) !
∞ n2 ∞ 1 ∞ 1
∑ = ∑ + ∑ = 2e
n=1 n! n =1 ( n − 1 ) ! n =2 ( n − 2 ) !
1 1
b) an = √ √ −√ √
n+2+ n+1 n+1+ n
1 1 1 √
Sn = √ √ −√ ⇒ S = −√ = 1− 2
n+2+ n+1 2+1 2+1
1 1 1 1
c) 2
= −
4n − 1 2 2n − 1 2n + 1
1 1 1
Sn = 1− ⇒S=
2 2n + 1 2
1 1 1 1 1 1 1 1
d) = − = −2 +
n(n + 1)(n + 2) 2 n(n + 1) (n + 1)(n + 2) 2 n n+1 n+2
1 1 1 1 1
Sn = 1− − + ⇒S=
2 2 n+1 n+2 4
1 ( n + 1) − n
e) arctg 2
= arctg = arctg(n + 1) − arctg n
1+n+n 1 + ( n + 1) n
π
Sn = arctg(n + 1) − arctg 1 ⇒ S =
4
Solution 6.2. All of these series do not satisfy necessary conditions
n2 − 2n + 1 1
a) lim √ = 6= 0
n→∞ 5n2 + (−1)n n 5
n n2 1
b) lim = e− 2 6= 0
n→∞ n + 1
2n π
c) lim arctg = 6= 0
n→∞ n 2
Solution 6.3.
1 n+1 1 2 2
a) √ ln ∼√ ∼ 3 (n → ∞), the series is convergent.
n n−1 nn−1 n2
√ √ (2 − a ) n + 1 (2 − a ) n + 1
b) n4 + 2n + 1 − n4 + an = √ √ ∼
n4 + 2n + 1 + n4 + an 2n2
1
If a = 2, un ∼ 2 the series is convergent.
2n
a−2
If a 6= 2, un ∼ , the series is divergent.
2n
2 1
c) For sufficiently large number n: n3 e−n ≤ 2 , the series is convergent.
n
a 2
( n + 1) + 5 3 n 1
d) lim n+1 = lim = < 1, the series is convergent.
n→∞ an n→∞ 3n +1 n2 + 5 3
ln2 n 1
e) α > 2: an ≤ α−1 ≤ 1+ε ; (0 < ε < α − 2), the series is convergent.
n n
ln2 2
α ≤ 2: an ≥ α−1 , the series is divergent.
n
a n +1 a a
f) lim = lim n = . The series is convergent if a < e, is divergent if a > e.
n→∞ an n→∞ e
1+ 1
n
a n +1 (2n + 1)!! 22n (n − 1)! 1

g) lim = lim 2(n+1) = < 1, the series is convergent.
n→∞ an n→∞ 2 n! (2n − 1)!! 2
√ 1 1 n 1
h) lim n an = lim 1 − = < 1, the series is convergent.
n→∞ 5 n→∞ n 5e
√ √n
n 2 1
i) lim an = lim n
n = < 1, the series is convergent.
n→∞ n→∞ 4n − 3 16
√ n + a n
j) lim n an = lim = e a−b .
n→∞ n→∞ n + b
If a > b then e a−b > 1 and the series is divergent.
If a < b then e a−b < 1 and the series is convergent.
If a = b, an = 1 does not satisfy the necessary condition, the series is divergent.
1
k) f ( x ) = ,x ≥ 3
x ln x (ln ln x )2
Z∞
1 ∞
f ( x )dx = − < +∞, the series is convergent.
ln ln x 3
3
1
l) f ( x ) = ,x ≥ 2
x ln p x  ∞
Z∞ ln ln x if p = 1


2
f ( x )dx = (ln x )1− p ∞
2

 if p 6= 1
1− p 2

The series is convergent if p > 1, is divergent if 0 < p ≤ 1.
Solution 6.4.
ln n ln n
a) lim = 0; an = is decreasing as n → ∞ because
n→∞ n n
ln x ′ 1 − ln x
f (x) = ; f (x) = < 0, ∀ x ≥ 3
x x2
The series is convergent.
√ √
n n
b) lim = 0; an = is decreasing as n → ∞ because
n→∞ n + e n+e
√
x ′ e−x
f (x) = ; f (x) = √ < 0, ∀ x ≥ 3
x+e 2 x ( x + e )2
The series is convergent.
2. Function series 77
∞ 5n + 3 ∞ (−1)n ∞ (−1)n
c) ∑ (−1)n = 2 ∑ + 3 ∑ , this series is convergent because both se-
n =1 n2 + n n =1 n + 1 n =1 n
ries in the righthand-side are convergent.
Solution 6.5.
ln n 1
a) We choose 0 < ε < α − 1, when n is large enough α ≤ α−ε , α − ε > 1 so the series is
n n
convergent.
1 1
b) For an arbitrary 1 > ε > 0, we have for a large enough number n: p ≥ ε , the
(ln n) n
series is divergent.
2n 2n
c) ∼ as n → ∞
n + 2n 2n
a 2 ( n + 1 ) 2n 1
lim n+1 = lim n + 1
= < 1, the series is convergent.
n→∞ an n→∞ 2 2n 2
x2
d) ln(1 + x ) = x − + o ( x2 ) as x → 0, so
2
1 1
n − ln 1 + ∼ 2 , as n → ∞
n 2n
the series is convergent.
√ √ a2 π
e) sin(π n2 + a2 ) = (−1)n sin(π n2 + a2 − nπ ) = (−1)n sin √
n2 + a2 + n
2
a π n a2 π o
0 < √ < π, ∀n, when n is large enough sin √ is a positive
n2 + a2 + n n2 + a2 + n
sequence which converges to 0. The original series is convergent.
√ √
f) {Sn }, Sn = (2 + 3)n + (2 − 3)n satisfy Sn+2 = 4Sn+1 − Sn , for all n ≥ 0.
By induction prove that Sn is divisible by 4, then it is even for all n.
√ √ √
Hence sin[π (2 + 3)n ] = − sin[π (2 − 3)n ] ∼ −π (2 − 3)n as n → ∞.
∞ √ √
∑ π (2 − 3)n is convergent because 0 < π (2 − 3) < 1, the original series is conver-
n =0
gent.
1 1
g) α > 1: ≤ where 0 < ε < α − 1, the series is convergent.
nα (ln n) β nα−ε
1 1
0 < α < 1: ≥ where 0 < ε < 1 − α, the series is divergent.
nα (ln n) β nα+ε
α = 1, see (1.3.l).
Summary the series is convergent if and only if α > 1 or α = 1, β > 1; and is divergent
if 0 < α < 1 or α = 1, 0 < β ≤ 1.
√ a n2 a2
h) lim n an = lim cos = e− 2 < 1, the series is convergent.
n→∞ n→∞ n
a n +1 ( n + 1)2
i) lim = lim = 0, the series is convergent.
n→∞ an n→∞ 22n+1
§2. F UNCTION SERIES
2.1 Function sequence

Assume that f 1 , f 2 , . . . , f n , . . . is a sequence of functions defined in a set X ⊂ R. x0 ∈ X
is called a convergent point of the above sequence if { f n ( x0 )} is a convergent sequence in
R.
A sequence { f n } is called uniformly convergent in a set X to a function f , denoted by
X
f n ⇒ f , if for an arbitrary positive number ε > 0, there exists a number n0 ∈ N such that
for all n ≥ n0 , we have
| f n ( x ) − f ( x )| < ε, ∀ x ∈ X
The number n0 depends only on ε, does not depend on x.
xn
Example 2.1. The sequence f n ( x ) = is uniformly convergent in [0, 1] to the function
n
f ( x ) = 0 because
xn 1
| f n ( x ) − f ( x )| = ≤
n n
h1i
for all x ∈ [0, 1]. Then for an arbitrary ε > 0 we can choose n0 = + 1.
ε
X
Proposition: f n ⇒ f if and only if lim max | f n ( x ) − f ( x )| = 0.
n→∞ x ∈ X
2.2 Function series

Definition
∞
A function series is ∑ un ( x ), where un ( x ), n ≥ 1, are functions defined in a set X ⊂ R.
n =1
Denote by Sn ( x ) the n−th partial sum of the above function series.
∞
The function series ∑ un ( x ) is called to converge at a point x0 if the sequence {Sn ( x0 )}
n =1
converges, is called to converge in a set X if {Sn ( x )} converges for every point x ∈ X. The
∞
set of all convergent points of ∑ un ( x ) is called the domain of convergence. The limits S
n =1
of the sequence {Sn } is called the sum of the function series.
1 ∞
Example 2.2. We consider the function series ∑ x
. It is convergent if and only if x > 1.
n =1 n
Then the domain of convergence of this series is (1, +∞).
∞
The function series ∑ un ( x ) is called to converge uniformly to a function S in a set X if
n =1
the sequence {Sn } converges uniformly to S in X.
Convergence criterion
To prove uniform convergence, we often use the following criterion.
∞
1. ∑ un ( x ) is uniformly convergent in X to a function S( x ) if and only if
n =1
lim max |Sn ( x ) − S( x )| = 0

n→∞ x ∈ X
2. Weierstrass’ criteria: If for all x ∈ X, we have
|un ( x )| ≤ an , ∀n ≥ 1
∞ ∞
and the number series ∑ an is convergent, then the function series ∑ un ( x ) con-
n =1 n =1
verges uniformly in X.
Example 2.3.
∞ cos nx
i) The function series ∑ 2 2
converge uniformly in R due to Weierstrass’ criteria.
n =1 n + x
Indeed, cos nx 1 1
≤ 2 ≤ 2 , ∀x ∈ R

2 2 2
n +x n +x n
∞1
and the series ∑ 2
is convergent.
n =1 n
(−1)n−1
∞
ii) Consider the function series ∑ 2
.
n =1 n + x
For each x ∈ R, the corresponding number series is convergent due to Leibnitz’s crite-
ria. Denote by S( x ), x ∈ R the sum of the number series. For all x ∈ R, we have
1 1
|S( x ) − Sn ( x )| ≤ ≤
x2 +n+1 n+1
then
1
0 ≤ lim max |Sn ( x ) − S( x )| ≤ lim =0
n→∞ x ∈ X n→∞ n + 1
∞ (−1)n−1
Hence lim max |Sn ( x ) − S( x )| = 0 and ∑ 2
converges uniformly in R.
n→∞ x ∈ X n =1 n + x
Properties of uniformly convergent function series

∞
Given a function series ∑ un .
n =1
∞
If {un }, n ∈ N, are continuous functions in the interval [ a, b] and ∑ un converges uni-
n =1
formly in [ a, b] to S( x ), then S( x ) is continuous in [ a, b], then is integrable in this interval
and
Zb Zb h ∞ i ∞ Zb
S( x )dx = ∑ un ( x) dx = ∑ un ( x)dx
a a n =1 n =1 a
If {un }, n ∈ N, are continuous functions together with their derivatives in the interval
∞ ∞
( a, b) and ∑ un is convergent to S( x ), ∑ u′n converges uniformly in ( a, b), then S( x ) is
n =1 n =1
differentiable in ( a, b) and
h ∞ i′ ∞
S′ ( x ) = ∑ un ( x ) = ∑ u′n ( x)
n =1 n =1
2.3 Power series

Power series, radial of convergence, domain of convergence
A power series is a function series of the following form
∞
∑ a n x n = a0 + a1 x + a2 x 2 + . . . + a n x n + . . .
n =1
∞
Radial of convergence of a power series is a number such that ∑ an x n is absolutely con-
n =1
a | x | < R and is p
vergent when divergent when | x | > R.
n +1
If lim = ρ (or lim n | an | = ρ), then the radial of convergence is determined by
n→∞ an n→∞

1
if 0 < ρ < ∞


ρ


R= 0 if ρ = ∞



∞ if ρ = 0

The domain of convergence contains (− R, R), together with the end points x = R or
x = − R if the power series converges at x = ± R respectively.
Properties of a power series
∞
The power series ∑ un converges uniformly in every closed interval [ a, b] ⊂ (− R, R).
n =1
∞
The sum of the power series ∑ un is continuous in its domain of convergence. We can
n =1
integrate or differentiate each terms of this series:
∞ ′ ∞
∑ un = ∑ an nxn−1
n =1 n =1
Zb ∞ ∞ Zb
∑ un dx = ∑ an x n dx
a n =1 n =1 a
for all closed intervals [ a, b] ⊂ (− R, R).

∞
If ∑ un ( x ) also converges at x = ± R, then
n =1
∞
∑ un (± R) = x→±
lim S( x )
R
n =1
∞ xn
Example 2.4. Consider the series ∑
√ n =1 n
R = lim n n = 1, then radial of convergence is R = 1.
n→∞
∞ 1
x = 1, ∑ is divergent.
n =1 n
∞ (−1)n
x = −1, ∑ is convergent.
n =1 n
So the domain of convergence is [−1, 1).
We can calculte the sum S( x ) of the observing series. In (−1, 1), S( x ) is differentiable
and
∞
1
S ( x ) = ∑ x n −1 =
′
n =1
1−x
Hence
Zx Zx
′ dt
S ( x ) = S (0) + S (t)dt = = −ln(1 − x )
1−t
0 0
Because the series also converges at x = −1, then
∞
(−1)n
−ln2 = S(−1) = lim S( x ) =
x →−1
∑ n
n =1
As in the above example, we often use differentiation and integration to find the sum of
a power series. We differentiate or integrate a series to obtain a new series, which we can
use some fundamental series expansions to calculate this. Here are some fundamental
series.
∞
1
= ∑ x n = 1 + x + x2 + . . . + x n + . . . (| x | < 1)
1−x n =0
∞
1
= ∑ (−1)n x n = 1 − x + x2 + . . . + (−1)n x n + . . . (| x | < 1)
1+x n =0
∞
(−1)n x2n+1 x3 (−1)n x2n+1
sin x = ∑ = x − + . . . + + . . . ( x ∈ R)
n=0 (2n + 1) ! 3! (2n + 1)!
∞
(−1)n x2n x2 x4 (−1)n x2n
cos x = ∑ (2n)! = 1 −
2!
+
4!
+ . . . +
( 2n ) !
+ . . . ; ( x ∈ R)
n =0
∞
xn x2 xn
ex = ∑ n! = 1 + x + + . . . + + . . . ( x ∈ R)
n =0 2! n!
√ 3 n+1
Example 2.5. Find the sum of the series 1 + 2+ +...+ √ +...
2 ( 2) n
∞
We consider the function series ∑ (n + 1) x n =: f ( x ).
n =0
Radial of convergence is R = 1, then for x ∈ (−1, 1), we can integrate each term of this
series in the interval [0, x ]:
Zx Zx ∞ ∞ Zx
n
f (t)dt = ∑ ( n + 1) t dt = ∑ (n + 1)tn dt
0 0 n =0 n =0 0
∞
n +1 x
= ∑x =
1−x
, x ∈ (−1, 1)
n =0
x ′ 1
Hence f ( x ) = = . The finding sum is
1−x (1 − x )2
√ 3 n+1 1 √
1+ 2+ +...+ √ + . . . = f √ = 2(3 + 2 2)
2 ( 2) n 2
2.4 Exercises
Exercise 6.6. Find the domain of convergence of the following series
∞ lnn ( x + n1 ) ∞
(−1)n+1 ∞
(n + x )n
a) ∑ √ b) ∑ c) ∑ x +n
n =1 x−e n =1
1 + n2 x n =0 n
∞ ∞ ∞
cos nx xn 1
d) ∑ 2nx
e) ∑ 1 + x2n
f) ∑ tgn ( x + n )
n =1 n =1 n =1
Exercise 6.7. Examine the uniform convergence of the following function sequences and
function series
a) f n ( x ) = x n − x n+1 in the interval [0, 1].
x
b) f n ( x ) = sin in the interval [0, 1] and in R.
n
∞
c) ∑ (1 − x ) x n in the interval [0, 1].
n =1
∞ x2
d) ∑ ln 1 + 2 in the interval [− a, a], ( a > 0).
n =1 n ln n
∞ x
e) ∑ 2n sin n in the interval [− a, a], ( a > 0).
n =1 3
∞ 2x + 1 n
f) ∑ an in [−1, 1], (| a| < 1).
n =1 x+2
Exercise 6.8. Find the convergent radial and domain of convergence of the following ex-
ponential series
∞ ∞ ∞
xn n
1 n2 n
a) ∑ √ b) ∑x ln(n + 1) c) ∑ 1+ x
n =1 n n =1 n =1
n
∞ ∞ ∞
2n .n! 2n x3n+1 2n−1 .x n−1
d) ∑ (2n)!
x e) ∑ (−2)n n + 1 f) ∑ √
2 3n −1
n =1 n =1 n=1 (2n − 1)
Exercise 6.9. Find the sum of the following series

∞ ∞ ∞
x n +1 (−1)n−1 x2n x4n+1
a) ∑ (2n)!!
b) ∑ 2n − 1
c) ∑
n =1 n =1 n=0 4n + 1
∞ ∞ ∞
(−1)n
d) ∑ n 2 x n −1 e) ∑ n ( n + 2) x n f) ∑ 4n2 − 1
n =1 n =1 n =1
2.5 Solution
Solution 6.6.
√ 1
a) Domain of determination: x > e. n an = ln( x + ) → ln x > 1 as n → ∞ then the series
n
is divergent at x > e. The domain of convergence is ∅.
b) x = 0, | an | = 1, the series is divergent.
∞ (−1)n+1 1 ∞ (−1)n+1
x 6= 0, ∑ 2
= ∑ .
n =1 1 + n x x n=1 1x + n2
1
For each x, { 1 } is a positive decreasing sequence when n is large enough which
+ n 2
x
tends to 0, so the series is convergent. The domain of convergence is R∗ .
n + x n 1 1
c) an = x
∼ e x x , so the series is convergent iff x > 1. Domain of convergence is
n n n
(1, +∞).
cos nx 1 ∞ 1 n
d) x > 0: nx ≤ nx , the series is ∑ is convergent because 2x > 1.

2 2 2 x
n =1
x ≤ 0, if the series is convergent at x then necessary condition leads to
cos nx
lim = 0 ⇒ lim cos nx = 0,
n→∞ 2nx n→∞
this is impossible. Domain of convergence is (0, +∞).

| x |n 1 n 1
e) | x | > 1: | an | = 2n
∼ as n → ∞; < 1 so the series is convergent.
1+x |x| |x|
| x |n
| x | < 1: | an | = ∼ | x |n as n → ∞; | x | < 1 so the series is convergent.
1 + x2n
1
| x | = 1, | an | = 9 0, the series is divergent. Domain of convergence is R\{±1}.
2
p
n
1
f) | an | = tg x + → tg x as n → ∞.
n
π π
If tg x < 1 ⇔ − + kπ < x < + kπ , the series is convergent.
4 4
π
If tg x = 1 ⇔ x = ± + kπ : an → e±2 6= 0 as n → ∞, the series is divergent.
4
If tg x > 1, the series is divergent.
π π
Domain of convergence: − + kπ, + kπ ; (k ∈ Z).
4 4
Solution 6.7.
nn 1
a) f ( x ) = 0, ∀ x ∈ [0, 1]; | f n ( x ) − f ( x )| = x n (1 − x ) ≤ n + 1
≤ → 0 as n → ∞.
( n + 1) n+1
[0,1]
fn ⇒ f .
x
b) f ( x ) = 0, ∀ x ∈ R; | f n ( x ) − f ( x )| = sin

n
x x
1 [0,1]
For all x ∈ [0, 1]: sin ≤ ≤ → 0 as n → ∞. f n ⇒ f .

n n n
x n
For x ∈ R: max sin ≥ sin = sin 1 9 0 as n → ∞. f n does not converge uniformly

x ∈R n n
to f in [0, 1].
c) Sn ( x ) = x − x n+1 → x if 0 ≤ x < 1, and Sn (1) → 0 as n → ∞. The function f ( x ) = 0 if
x = 1; f ( x ) = x if 0 ≤ x < 1 is not continuous in [0, 1] then Sn does not converge uniformly.
The series does not converge uniformly too.
d) ln(1 + x ) ≤ x, ∀ x ≥ 0; then
x2 x2 a2
ln 1 + ≤ ≤ ; ∀ x ∈ [− a, a]
n ln2 n n ln2 n n ln2 n
∞ 1
a2 ∑ is convergent then use Weierstrass’ criteria, the series converges uniformly
n ln2 n
n =1
in [− a, a].
x 2 n ∞ 2 n
e) 2n sin n ≤ a ; ∑ is convergent then use Weierstrass’ criteria, the series

3 3 n =1 3
converges uniformly in [− a, a].
2x + 1 2x + 1 n ∞
f) ∈ [−1, 1] for all x ∈ [−1, 1], then an ≤ | a|n , the series ∑ | a|n is con-

x+2 x+2 n =1
vergent then the function series converges uniformly in [−1, 1].
Solution 6.8.
a) R = 1, domain of convergence is [−1, 1).
b) R = 1, domain of convergence is (−1, 1).
1 1 1
c) R = , domain of convergence is − , .
e e e
3. Fourier series 85
d) R = +∞, domain of convergence is (−∞, +∞).

1 1 1 i
e) R = √
3
, domain of convergence is − √
3
, √ .
2 2 32
√ h √3 √3 i
3
f) R = , domain of convergence is − , .
2 2 2
Solution 6.9.
x
a) S( x ) = x (e 2 − 1), ∀ x ∈ R.
b) R = 1, ∀ x ∈ (−1, 1): S( x ) = x arctg x.
1 1+x 1
c) R = 1, ∀ x ∈ (−1, 1): S( x ) = ln + arctg x.
4 1−x 2
1+x
d) R = 1, ∀ x ∈ (−1, 1): S( x ) =
(1 − x )3
x2 − 3x
e) R = 1, ∀ x ∈ (−1, 1): S( x ) =
( x − 1)3
∞ (−1)n x2n 1 x2 + 1
f) S( x ) = ∑ = − arctg x
n=1 (2n + 1)(2n − 1) 2 2x
∞ (−1)n 1 π
∑ 2
= lim S( x ) = −
n=1 4n − 1 x →1 2 4
§3. F OURIER SERIES
3.1 Decomposition theorem

Assume that f ( x ) is a 2π periodic funtion and integrable in the closed interval [−π, π ].
Its Fourier series is a trigonometric series
∞
a0
+ ∑ an cos nx + bn sin nx
2 n =1
whose coefficients are

Zπ Zπ
1 1
a0 = f ( x )dx; an = f ( x ) cos nxdx, n ≥ 1
π π
−π −π
Zπ
1
bn = f ( x ) sin nxdx, n ≥ 1
π
−π
Theorem 6.1. Assume that f : R → R is a 2π periodic function that satisfies one of the
following conditions:
i) f is piecewise continuous function and its derivatives is piecewise continuous;
ii) f is piecewise monotonous and is bounded.
Then, the Fourier series of f ( x ) converges at every points and its sum S( x ) coincides
with f ( x ) at every continuous points of f . At discontinuous point c of f ( x ), we have
f ( c + 0) + f ( c − 0)
S(c) =
2
If f ( x ) is an odd function then an = 0, ∀n ≥ 0,
Zπ
2
bn = f ( x ) sin nxdx, n ≥ 1
π
0
If f ( x ) is an even function then bn = 0, ∀n ≥ 1,

Zπ Zπ
2 2
a0 = f ( x )dx; an = f ( x ) cos nxdx, n ≥ 1
π π
0 0
Example 3.1. Find the Fourier series of the 2π periodic function f ( x ), f ( x ) = x for x ∈
(−π, π ).
f ( x ) is bounded and is an increasing function in every intervals (−π + 2kπ, π + 2kπ ),
then it can be decomposed into Fourier series. We calculate the coefficients.
Because f ( x ) = x in (−π, π ) is an odd function then
Zπ
1
an = f ( x ) cos nxdx = 0, n ≥ 0
π
−π
Zπ π Zπ cos nx i
2 2h x
bn = x sin nxdx = − cos nx + dx

π π n 0 n
0 0
2
= (−1)n+1 ; (n ≥ 1)
n
Hence for x 6= (2n + 1)π,
∞
(−1)n+1
f (x) = 2 ∑ n
sin nx
n =1
Theorem 6.2. If f ( x ) is 2l periodic function which also satisfies one of the conditions
mentioned in the above theorem in the interval [−l, l ], then f ( x ) can be decomposed into
Fourier series
∞
a0 nπx nπx
+ ∑ an cos + bn sin
2 n =1
l l
whose coefficients are
Zl
1
a0 = f ( x )dx
l
−l
Zl
1 nπx
an = f ( x ) cos dx, n ≥ 1
l l
−l
Zl
1 nπx
bn = f ( x ) sin dx, n ≥ 1
l l
−l
Example 3.2. Find the Fourier series of the 2l periodic function f ( x ) that f ( x ) = x in
( a, a + 2l ).
f ( x ) is bounded and is an increasing function in every intervals ( a + 2kl, a + 2(k + 1)l ),
then it can be decomposed into Fourier series. We calculate the coefficients
Because the integral of a periodic function in every interval whose length is equal to
the periodic is the same then
Zl aZ+2l aZ+2l
1 1 1
a0 = f ( x )dx = f ( x )dx = xdx = 2( a + l )
l l l
−l a a
aZ+2l aZ+2l
1 nπx 1 xl nπx a+2l l nπx
an = x cos dx = sin − sin dx
l l l nπ l a nπ l

a a
1 2l 2 πna l2
nπx a+2l
= sin + 2 2 cos
l nπ l n π l a

2l πna
= sin ,n ≥ 1
nπ l
aZ+2l aZ+2l
1 nπx 1 − xl nπx a+2l l nπx
bn = x sin dx = cos + cos dx
l l l nπ l a nπ l

a a
−2l πna
= cos ,n ≥ 1
nπ l
Hence for x 6= a + 2nl,
∞
2l 1 nπa nπx nπa nπx
f ( x ) =( a + l ) +
π ∑ n sin l cos l − cos l sin l
n =1
∞
2l 1 nπ
=a + l +
π ∑ n
sin
l
(a − x)
n =1
Now we consider a function which satisfies one of the conditions in the first theorem in a
closed [ a, b]. To expand f ( x ) into a Fourier series, we construct a periodic function g( x )
whose periodic is larger or equal b − a such that
g( x ) = f ( x ), ∀ x ∈ [ a, b]
If g( x ) can be expanded into Fourier series then its sum coincides with g( x ), and also
f ( x ), at every continuous points in [ a, b]. If g( x ) is an odd function then its Fourier series
is sum of sine functions. If g( x ) is an even function then its Fourier series is sum of cosine
functions.
Example 3.3. Find the expansion of the function f ( x ) = x for 0 < x < 2 into Fourier
series of cosine functions and into Fourier series of sine functions.
To expand f ( x ) into Fourier series of cosine functions we construct an even function
g( x ), which is 4 periodic function and g( x ) = x for 0 < x < 2, l = 2. g( x ) is even then
bn = 0, n ≥ 1
Z2
a0 = xdx = 2;
0
Z2 Z2
nπx 2x nπx 2 2 nπx
an = x cos dx = sin − sin dx
2 nπ 2 0 nπ 2
0 0

4 0 if n is even
= 2 2 (cos nπ − 1) = 8
n π − if n is odd
n2 π 2
Hence for 0 < x < 2,
∞
8 1 (2n + 1)πx
f (x) = 1 −
π2 ∑ 2
cos
n=0 (2n + 1) 2
To expand f ( x ) into Fourier series of sine functions we construct an odd function h( x ),

which is 4 periodic function and h( x ) = x for 0 < x < 2, l = 2. h( x ) is odd then an = 0, n ≥ 0
Z2 Z2
nπx 2x nπx 2 2 nπx
bn = x sin dx = − cos + cos dx
2 nπ 2 0 nπ 2
0 0
4 (−1)n+1
=− cos nπ = 4
nπ nπ
Hence for 0 < x < 2,
∞
4 (−1)n+1 nπx
f (x) =
π ∑ n sin 2
n =1
3.2 Exercises
Exercise 6.10. Find the Fourier expansion of the following functions
a) f ( x ) is a periodic function with T = 2π and f ( x ) = | x | in the interval [−π, π ].
π−x
b) f ( x ) is a periodic function with T = 2π, and f ( x ) = in the interval (0, 2π ).
2
c) f ( x ) is a periodic function with T = 2π and f ( x ) = sin ax in the interval (−π, π ),
a 6= Z.
Exercise 6.11. Decompose the following functions into Fourier series of cosine and sine
functions 
0 if 0 ≤ x ≤ h
a) f ( x ) = in the interval [0, π ].
1 if h < x ≤ π

x


 if 0 ≤ x ≤ 1
b) f ( x ) = 1 if 1 < x < 2 in the interval (0, 3).


3 − x if 2 ≤ x ≤ 3

3.3 Solution
Solution 6.10.
π 4 ∞ cos(2n + 1) x
a) f ( x ) = − ∑ , ∀ x ∈ [−π, π ].
2 π n=0 (2n + 1)2
∞ sin nx
b) f ( x ) = ∑ , ∀ x ∈ (0, 2π ).
n =1 n
2 sin aπ ∞ (−1)n n
c) f ( x ) = ∑ 2 2
sin nx, ∀ x ∈ (−π, π ).
π n =1 a − n
Solution 6.11.
2 ∞ cos(nh) + (−1)n+1 π−h 2 ∞ sin(nh)
a) f ( x ) = ∑ sin nx and f ( x ) = − ∑ cos nx
π n =1 n π π n =1 n
2 3 ∞ 1 2nπ 2nπx
b) f ( x ) = + ∑ 2 cos − 1 cos and
3 π n =1 n 3 3
9 ∞ 1 nπ 2nπ 2nπx
f ( x ) = 2 ∑ 2 sin + sin sin
π n =1 n 3 3 3

Bai Giang Math4 PDF

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Bai Giang Math4 PDF

Загружено:

Авторское право:

Доступные форматы

H ANOI U NIVERSITY OF S CIENCE AND T ECHNOLOGY

S CHOOL OF A PPLIED M ATHEMATICS AND I NFORMATICS

N GUYEN T HI T HU H UONG AND T RAN M INH T OAN

M ULTIPLE I NTEGRAL , I NTEGRAL THAT DEPENDS ON A PARAMETER ,

Summary, Examples, Exercises and Solutions

Chapter 1 . Multiple Integral . . . . . . . . . . . . . . . . . . . . . . . . 5

3.1 Decomposition theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

§1. D OUBLE I NTEGRAL

1.1 Calculation of a double integral in Cartesian

1. (The Corollary of Fubini’s theorem)

Example 1.1. Calculate the double integral

where D = [0, 1] × [0, 2].

Example 1.3. Interchange the order of the following integrals:

From above figure, we have

1.2 Change of variables in double integrals, polar

y = x, y = 2x, xy = 1, xy = 3 ( x > 0).

Solution. 1. We have D = [0, 1] × [0, 1] (Figure 1.5).

To transform to polar coordinate, we put x = r cos ϕ; y = r sin ϕ.

Note: Some regions used frequently in polar coordinate

v) If D = x2 + y2 = −2ax, a > 0 = ( x + a)2 + y2 = a2 then

1.3 Applications of double integrals

2. To compute the area of a curved surface

3. To compute the volume of a object

Exercise 1.2. Calculate the following integrals

Exercise 1.4. Calculate these following integrals:

3. Divide D into two regions D = D1 ∪ D2 , where

4. Divide D into two regions D = D1 ∪ D2 , where

5. Note that D is axisymetricZZand the function f ( x, y) = | x | + |y| is even with respect

Note: The domain D is Ox-axisymetric and the function f ( x, y) = xy is odd with

I= dϕ r cos ϕ (r sin ϕ)2 rdr = 0.

2. Divide D into two regions D = D1 ∪ D2 , where

Solution 1.5. Divide the domain D into two subdomain D = D1 ∪ D2 , where

Solution 1.6. We have

Solution 1.8. We have

Solution 1.9. We have V = 2V1 , where

§2. T RIPLE I NTEGRAL

2.1 Calculation of a triple integral in Cartesian

where f ( x, y, z) is three-variables function that is continuous on V.

I= dxdy  f ( x, y, z)dz . (1.11)

Solution. V is tetrahedron bounded by two planes z = 0 and z = 1 − x − y, ( x, y) ∈ D,

2.2 Change of variables in triple integrals

Suppose that the following conditions are satisfied:

ii) The vecto-valued mapping Φ : V ′ → V is one-to-one.

iii) The Jacobian determinant

2.3 Calculate the triple integrals in cylindrical

x = r cos ϕ, y = r sin ϕ, z = z. (1.13)

and therefore we have the tranformation formula

Note: In some cases we use the generalized cyclindrical coordinate

Solution. Transform to cylindrical coordinate : x = r cos ϕ, y = r sin ϕ, z = z, 0 ≤ ϕ ≤ 2π.

We note that the paraboloid x2 + y2 = 2z cuts the plane x2 + y2 = 4 by the circle x2 + y2 =

2.4 Calculate the triple integrals in spherical

To get a one-to-one mapping we keep r > 0, 0 ≤ ϕ < 2π and 0 ≤ θ < π.

Therefore we have the tranformation formula

(a) V is the region bounded by the cylinder x2 + y2 = 2x and the planes z = 0, z =

(b) V is a half of the sphere x2 + y2 + z2 ≤ a2 , z ≥ 0 ( a > 0).

4. (a) Transform to cylindrical coordinate:

(b) Transform to cylindrical coordinate:

(c) Transform to generalized cylindrical coordinate

5. Transform to cylindrical coordinate

6. Transform to generalized spherical coordinate

§1. T HE DEFINITE INTEGRALS THAT DEPEND ON A

2. Differentiation under the integral sign

i) f ( x, y) is defined on the rectangle D = [ a, b] × [c, d] and continuous with respect

dydz dzdx dxdy x 2 y2 z2