LECTURE MATERIAL(LASU)

COURSE NAME:MATHEMATICAL METHODS 1

COURSE CODE: MAT 251

Y
By:
Fabass

October 14, 2020

EM
Course Outline

1. Techniques of integration

• integral of simple functions

• integration by simple substitution
AD
• use of trigonometric functions
• integration of rational functions
• integration by parts

2. Reduction formula

• definite and indefinite integrals

• Walli’s formulae
AC

3. Leibniz theorem and application

4. Partial and Total Derivatives

• Application to classification of critical points of functions and of two variables

5. Co-ordinate systems: Cartesian, polar,cylindrical and spherical coordinate system

• Transformation from one co-ordinate system to another

6. Wronskian: Linear independence/dependence

7. Evaluation of line,surface and volume integrals

A

8. Green’s and Stoke’s theorem: Application

9. Scalar and Vector Potentials

BT

10. Laplace’s and Poisson’s equations

1
1 Techniques of integration
The integral of a function depends solely on the nature of the function and how much information
that is know about it. Evaluation of integration problems requires a good understanding of suitable
techniques ranging from simple to complicated to efficiently solve the problem.
In this course, the following techniques:
1. Use of fundamental rules

Y
2. Simple Substitution

3. Use of trigonometric functions

4. Use of partial fractions

EM
5. integration by parts
are considered for the purpose of this topic. It is assumed that the student already have a background
in MAT 101: Basic Algebra, MAT 111: Trigonometry and MAT 112: Elementary Calculus.
Fundamental Rules
1. The integral of a sum of a finite number of functions is the sum of their separate integrals i.e.
n n Z
Z X ! 
X
fi dx = fi dx
AD i i

2. A constant factor may be brought outside the integral sign i.e.

Z Z
kf (x)dx = k f (x)dx , k = constant

3. The addition of a constant to the variable makes no difference to the form of the result. i.e.
Z Z
f (x ± a) dx ≡ f (X) dx, X = x ± a
AC

4. Multiplying the variable by a constant makes no difference to the forms of the NOTE: NOT
COMPLETE

5. The integral of a rational expression whose numerator is the derivative of its denominator is the
logarithm of the denominator
Examples
1.
√
√
Z Z Z Z
x2 + sin x + x dx = x2 dx + sin xdx + x dx
A

1 2 3
= x3 − cos x + x 2 + C
3 3
2. Z Z Z
6 cos x − 4x2 dx = x2 dx


cos x dx − 4
BT

4
= 6 sin x − x3 + C
3
3. Z
dx 1 1
2
= tan−1 (x − 7) + C
(x − 7) + 9 3 3
4. Z
1 1 1
cos x dx = 1 sin x
3 3
3
1
= 3 sin x + C
3

2
 5. Z Z
dx dx
p = p dx
(9 − 4x2 ) 32 − (2x)2
1 2
= arcsin x + C
2 3
6. Z Z
dx dx
√ = p dx
12x − 9x2 2 − (3x − 2)2
2

Y
1 1
= arcsin (3x − 2) + C
3 2
7.
3x2
Z

EM
dx = loge x3 + 1 + C


3
x +1
8.
1
Z Z

dx x
= dx
x loge x loge x
= loge (loge x) + C

9. Z Z
sin x
tan x dx = dx
cos x
AD =−
Z
− sin x
cos x
dx

= − loge cos x + C

10.  
Z Z 1
dx 1+x2
= dx
(1 + x2 ) tan−1 x tan−1 x
= loge (tan−1 x) + C
AC

Integration by simple substitution Suppose we have an expression in the integral, such that this
expression or part of it is replaced by a variable or other suitable function, then the given integral could
be reduced to a simpler form that may be easier to handle. In such a case, the suitable substitution
should be employed to help in obtaining the solution of the integral with just some little knowledge
of the fundamental rule.
Example
By using a suitable substitution obtain the solution of the following integrals.

1. Z
3
I= x2 e−x dx
A

Solution

du
Let u = −x2 , =⇒ = −3x2 , substituting
dx
BT

Z  
u 1 du
I= e − dx
3 dx
Z
1 du
=− eu dx
3 dx
Z
1
=− eu dx
3
1
= − eu + C, since u = −x3 .
3
1 −x3
=− e +C
3

3
 It is more of usual practice to make the substitution u = φ(x), differentiate it to obtain du
dx = φ0 (x)
and then replace dx by φdu du dx
0 (x) . i.e dx = du = du du.
dx

2. Z p
3
I= (x − 1) x2 − 2x + 3 dx.

Solution
du

Y
Let u = x2 − 2x + 3 , =⇒ = 2x − 2 and
dx
√
Z
du
I = (x − 1)( 3 u) ·
(2x − 2)
√
Z
du

EM
= (x − 1) 3 u ·
2(x − 1)
√
Z
1 3
= u du
2
1 3 4
= · u3 + C since u = x2 − 2x + 3
2 4
3 4
= (x2 − 3x + 3) 3 + C
8
3. Z π
4

Solution
AD
Consider the definite integral
π
6

Z
cot x dx

I= cotx dx
Z
cos x
I= dx
sin x
du
Let u = sin x, =⇒ dx = cos x and substitute gives
AC

Z
cos x du
I= ·
u cos x
Z
du
=
u
= loge u + C
= loge (sin x) + C

Hence,
Z π
4 π
cot x dx = [loge (sin x)] π4
A

π 6
6
  
1 1
= loge sin π − loge sin π)
4 6
   
1 1
BT

= loge √ − loge
2 2
1 ! √
2
= loge 1
2
√
= loge 2
An alternative approach when dealing with definite integrals is to change the limits of integration
for the variable x into corresponding limits for variable u. Using the previous example, with
u = sin x
π π 1
when x = , u = sin =
6 6 2

4
 π π 1
when x = , u = sin = √
4 4 2
Z π Z 1
4 2 du
cot x dx =
π √1 u
6 2
1
√
= [loge u] 1 2
√ 2
= loge 2

Y
NOTE: Sometime, it may be necessary to use more than one substitution.
Exercise: Evaluate the following integrals.
1.
1
e2u

EM
Z
du
0 3e2u + 2
2. Z 2
x
dx
1 x4 − 2x2 + 10
3. Z e2
1
sin(loge x)dx
e1 x

2 Reduction formulaAD
When considering certain integral whose integrand is a function of some arbitrary parameters,it usu-
ally convenient to obtain a relation between the integral and parameters involved. Obtaining the said
relation usually involves the application of some technique of integration by parts, to the integral and
the resulting relation is called a reduced formula, hence the phrase ”reduction formula”.
Reduction formula provides a systematic approach to the evaluation of the integral problem.We shall
illustrate this concept by considering some well known integral problems and obtaining their appro-
priate reduction formula.
AC

sinn x dx
R
2.1 Reduction formula for
Considering the integral Z
In = sinn x dx (1)

where n is positive, a reduction formula can obtain by applying the integration by parts technique to
the problem. Now
Z Z
In = sinn x dx = sinn−1 x · sin x dx
Z Z  Z 
n−1 n n−2
A

= sin x · sin x dx − (n − 1) sin x · cos x · sin x dx dx

Z
= − sinn−1 x cos x + (n − 1) sinn−2 x cos2 x dx
Z
BT

n−1
= − sin x cos x + (n − 1) sinn−2 x · (1 − sin2 x) dx
Z Z
n−1 n−2
= −sin x cos x + (n − 1) sin x dx − (n − 1) sinn x dx

= − sinn−1 x cos x + (n − 1)In−2 − (n − 1)In

=⇒ (1 + n − 1)In = − sinn−1 x cos x + (n − 1)In−2
1
− sinn−1 x cos x + (n − 1)In−2


In = (2)
n
Equation (2) is the reduction formula obtained for problem (1).
Problem (1) can now be systematically evaluated with the help of (2)

5
 R
2.2 Reduction formula for cosn x dx
Given the integral. Z
Jn = cosn x dx , n ∈ Z+ (3)
Z Z
Jn = cos x dx = cosn−1 x cos x dx ,
n
Using integration by parts
Z Z  Z 
n−1 n−2

Y
= cos x · cos x dx − (n − 1) cos x · (− sin x) cos x dx dx
Z
= cosn−1 x sin x + (n − 1) cosn−2 x sin2 x dx
Z
n−1
= cos x sin x + (n − 1) cosn−2 x · (1 − cos2 x) dx

EM
Z Z
n−1 n−2
= cos x sin x + (n − 1) cos x − (n − 1) cosn x dx

= cosn−1 x sin x + (n − 1)Jn−2 − (n − 1)Jn

∴ (1 + n − 1)Jn = cosn−1 x sin x + (n − 1)Jn−2

1
cosn−1 x sin x + (n − 1)Jn−2


Jn = (4)
n
Equation (4) is the reduction formula obtain for problem (3).
AD
2.3 Reduction formula for tann xdx
Suppose Z
Kn = tann x dx n > 2, then
Z
Kn = tann−2 x tan2 x dx
Z
tann−2 x · (sec2 x − 1) dx
AC

=
Z Z
n−2
= tan x sec x dx − tann−2 x dx
2

Z
= tann−2 x sec2 x − Kn−2 Using integration by substitution

tann−1 x
= − Kn−2
n−1
tann−1 x
Kn = − Kn−2 (5)
n−1
A

R
2.4 Reduction formula for secn x dx
Let Z
secn x dx ,
BT

Qn = n>2
Z
∴ Qn = secn−2 x sec2 x dx , integrating by parts,

6
 Z Z  Z 
n−2 2 n−3 2
Qn = sec x · sec x dx − (n − 2) sec x · sec x tan x sec x dx dx
Z
= secn−2 x tan x − (n − 2) secn−2 x tan2 x dx
Z
n−2
= sec x tan x − (n − 2) secn−2 x · (sec2 x − 1) dx
Z Z
n−2
= sec x tan x − (n − 2) sec x + (n − 2) secn−2 x dx
n

Y
= secn−2 x tan x − (n − 2)Qn + (n − 2)Qn−2
∴ (1 + n − 2)Qn = secn−2 x tan x + (n − 2)Qn−2
1
secn−2 x tan x + (n − 2)Qn−2

EM


Qn = (6)
n−1
NOTE:To obtain a reduction formula, the integral need not necessary involved trigonometric func-
tions.
Example:
Obtain a reduction formula for the integral
Z
xn ex dx

Solution AD Z
Let Tn = xn ex dx , then using integration by parts
Z Z  Z 
n x n−1 x
= Tn = x · e dx − nx e dx dx
Z
= xn ex − n xn−1 ex dx

= xn ex − nTn−1

Tn = xn ex − nTn−1 (7)
AC

Which is the required reduction formula

sinm x cosn x dx
R
2.5 Reduction formula for
Z
sinm x cosn x dx ,
Let Pm,n = m, n are integers
Z Z
Now, Pm,n = sin x cos x = sinm x cosn−1 x cos x dx Using integration by parts
m n

Z
m n−1
Pm,n = sin x cos x · cos x dx−
A

Z  Z 
n−2 m m−1 n−1


(n − 1) cos x · (− sin x)(sin x) + m sin x cos x · cos x cos x dx dx
Z
m+1 n−1
(n − 1) cosn−2 x sinm x sin2 x · (−1) + m sinm x cosn x dx dx
 
Pm,n = sin x cos x+
BT

Z Z
m+1 n−1 n−2
= sin x cos x + (n − 1) cos x(1 − cos x) sin x dx − m sinm x cosn x dx
2 m

Z Z Z
m+1 n−1 m n−2
= sin x cos x + (n − 1) sin x cos x − (n − 1) sin x cos x dx − m sinm x cosn x dx
m n

Z Z
m+1 n−1 m n−2
= sin x cos x + (n − 1) sin x cos x dx − (n + m − 1) sinm x cosn x dx

= sinm+1 x cosn−1 x + (n − 1)Pm,n−2 − (m + n − 1)Pm,n

∴ (1 + m + n − 1)Pm,n = sinm+1 x cosn−1 x + (n − 1)Pm,n−2
(m + n)Pm,n = sinm+1 x cosn−1 x + (n − 1)Pm,n−2

7
 1  m+1
x cosn−1 x + (n − 1)Pm,n−2

Pm,n = sin (8)
m+n
Exercises:

1. Find the reduction formula for Z

cotn x dx

and

Y
Z
cscn x dx

2. Evaluate the integrals

EM
(a) Z
sin5 x dx

(b) Z
tan6 x dx

(c) Z
sec4 x dx

3. Evaluate

(a)
AD Z
sin4 x cos5 x dx

(b) Z
sin2 x cos4 x dx
AC

4. Using integration by parts or otherwise, evaluate

(a) Z
x3 (loge x)2 dx

(b) Z
x sin x cos x dx

(c) Z
x−3 cos x dx
A

5. Obtain a reduction formula for Z

xn loge x dx
BT

6. Find the reduction formula for Z

xn cos nx dx

8
2.6 Definite integrals
R
Given the integral f (x) dx, the value of this integral is unbound, since the value of the variable x is
not restricted. In order words if the values of x are valid within a certain range, say between (a,b),
then we evaluate the integral between a ≤ x ≤ b i.e
Z b
f (x) dx
a

Y
The result of which is evaluated as:
Z b
f (x) dx = [F (x) + C]ba = F (b) − F (a)
a

EM
Examples:

(a)
2 2 2
3x6
Z  
5 1 6 1 1 1 71
3x dx = = x = (26 ) − (16 ) = 32 − =
1 6 1 2 1 2 2 2 2

(b)
Z π  π    
2 1 2 1 1 1 1
sin 2x dx = − cos 2x + C = − cos π − − cos 0 = + = 1
0 2 0 2 2 2 2

(c)

0
AD
x e
Z ∞ 
1 2 −3x 2 −3x
2 −3x
dx = − x e
3
− xe
9
− e
2 −3x ∞
27


0
=
2
27

(d) By using the reduction formula for secn x dx, evaluate the definite integral
R

Z π
4
sec6 x dx
0
AC

Solution
The reduction formula for secn x dx is
R

1
secn−2 x tan x + (n − 2)Qn−2


Qn =
n−1

secn x dx
R
where In =
Z π
4
In = sec6 x dx
0
1 4 π 4 4
A

= sec x tan x + 4I4 04 = + I4

5 5 5
1 2 π 4 2 2
I4 = sec x tan x + 2I2 04 = = + I2
3 5 3 3
Z π π
4
sec2 x dx tan x 04 = 1
BT


=
0

Using backward substitution, we have

 
4 4 2 2
I6 = + +
5 5 3 3
 
4 4 4 4 16 12 16 28
= + = + = + =
5 5 3 5 15 15 15 15

9
 (e) If Z e
In = x (ln x)n dx
1

where n ∈ Z+ , prove that

1 1
In = e2 − nIn−1
2 2
Hence evaluate In

Y
Solution
Z e
In = x (ln x)n dx
1
Z Z  Z 
n n−1 1
= (ln x) · x − n(ln x) · x dx dx
x

EM
 Z e
1 2 1
= x (ln x)n − n x(ln x)n−1 dx
2 2 1
 Z e   Z 
1 2 1 n−1 1 n−1
= e − n x(ln x) dx + n x(ln x) dx
2 2 1 x=e 2 x=1
1 2 1
= e − nIn−1
2 2
Now, to evaluate I3 , we use the above result as follows.

AD 1 1
I3 = e2 − (3)I2
2 2
1 1
I2 = e2 − (2)I1
2 2
1 1
I1 = e2 − (1)I0
Z2 e
2 Z e
1 1
I0 = x(ln x)0 dx = x dx = e2 −
1 1 2 2
   
1 2 1 1 2 1 1 2 1 1 2 1
In = e − (3) e − e − e −
2 2 2 2 2 2 2 2
AC

1 2

= e +3
8

2.7 Wallis’ formula

If the integrals considered for reduction formulae are taken as definite integrals with limit from 0 and
1
2 π with positive values for m and n, then we have the following results which are due to Wallis.

1. Z π
2 1 π
In = sinn x dx =
− sinn−1 x cos x + (n − 1)In−2 02
0 n
A

1h π  π  i n−1
= − sinn−1 cos + sinn−1 (0) cos(0) + In−2
2 2 2 n
n−1
= In−2 , n≥2
n
BT

now, successive application of this will reduce In to either I1 , if n is odd or I0 if n is even.

Thus for In , we have
Z π π
2
I1 = sin x dx = [− cos x]02 = 1
0
and for I0 , we have,
Z π
2 π π
I0 = dx = [x]02 =
0 2
Hence for integer n,
(n − 1)(n − 3)(n − 5) · · · (5)(3)(1)  π 
In = (1)
n(n − 2)(n − 4) · · · 6 · 4 · 2 2

10
 and for integer n,
(n − 1)(n − 3)(n − 5) · · · 6 · 4 · 2
In = (2)
n(n − 2)(n − 4) · · · 7 · 5 · 3
It is easy to verify that (1) and (2) also hold for
Z π
2
Jn = cosn x dx .
0

Y
Equation (1) and (2) are called Wallis’ formulae for both In and Jn

2. Z π
2
Pm,n = sinm x cosn x dx

EM
0
1  m+1 π
= sin x cosn−1 x + (n − 1)Pm,n−2 02
m+n
n−1
= Pm,n−2
m+n
n−1 n−3
= · Pm,n−4
m+n m+n−2
(n − 1)(n − 3)(n − 5)(n − 7) · · ·
Pm,n = (Pm,0 or Pm,1 ) (3)
(m + n)(m + n − 2)(m + n − 4) · · ·
AD
(Pm,0 if n is even or Pm,1 if n is odd) depending on n

If n is even, Pm,0 is equivalent to Im which is

(m − 1)(m − 3)(m − 5) · · · 5 · 3 · 1  π 
if m is even
m(m − 2)(m − 4) · · · 6 · 4 · 2 2

(m − 1)(m − 3)(m − 5) · · · 6 · 4 · 2
if m is odd
AC

m(m − 2)(m − 4) · · · 7 · 5 · 3
Now if n is even and m is also even, (3) becomes

(n − 1)(n − 3)(n − 5) · · · 3 · 1 · (m − 1)(m − 3)(m − 5) · · · 5 · 3 · 1 π 

(4)
(m + n)(m + n − 2)(m + n − 4) · · · (m + 2) · (m) · (m − 2) · (m − 4) · · · 6 · 4 · 2 2

Which is the Wallis’ formula for Pm,n when both m and are even.

In a similar manner, it is easy to show that Wallis’ formula for Pm,n n when both m,n are
not even can be written as
A

(n − 1)(n − 3)(n − 5) · · · (m − 1)(m − 3)(m − 5) · · ·

Pm,n =
(m + n)(m + n − 2)(m + n − 4) · · · 5 · 3 · 1
BT

Example:

1. By Wallis’ formula, evaluate

Z π
2
sin8 x dx
0
Solution
Since the exponent of sin x is 8, then n = 8 and it is even
Z π
2
i.e sin8 x dx = I8
0

11
 8−1 8−3 8−5 8−7
I8 = · · · I0
8 8−2 8−4 8−6
7 · 5 · 3 · 1 π 
=
8·6·4·2 2
35
= π
256
2. Evaluate Z π
2 1

Y
sin3 x cos5 x dx Ans:
0 24

3. Evaluate Z π
2 3π
sin6 x cos4 x dx Ans:
512

EM
0

4. Evaluate Z π
256
sin9 x cos4 x dx Ans:
0 15015

3 Leibnitz theorem and Application

n-th derivative of a product of two functions
If y = u(x)v(x), then
AD dy dv du
=u + v , i.e (1)
dx dx dx
dv du
y 0 = uv 0 + vu0 , where v 0 = and u0 =
dx dx
Now,
y 00 = uv 00 + v 0 u0 + vu00 + u0 v 0
y 00 = u00 v + 2u0 v 0 + uv 00 (2)
Differentiating (2) and collecting like terms, we obtain
AC

y 000 = u000 v + 3u00 v 0 + 3u0 v 00 + uv 000 (3)

A further differentiation would give

y (4) = u(4) v + 4u(3) v (1) + 6u(2) v (2) + 4u(1) v (3) + uv (4)

Note that in each case, the ”superscript” of u decreases regularly by 1 and the superscript of v increases
regularly by 1. The numerical coefficient are the normal binomial coefficients. Indeed (u(x)v(x))(n)
can be obtain by expanding (u + v)(n) using the binomial theorem where the ”powers” are interpreted
as derivatives can therefore be written as
A

y (n) = u(n) v +n C1 u(n−1) v (1) +n C2 u(n−2) v (2) + · · · +n Cn−1 u(1) v (n−1) + uv (n)

Where
n n!
Cr =
BT

r!(n − r)!

3.1 Theorem: (Leibnitz)

If y = u(x)v(x), then
n
X
(n) n
y = Cr u(n−r) v (r) where u(0) ≡ 0
r=0

12
3.2 Choice of function for u and v
To efficiently apply the Leibnitz theorem to the product y = uv, the function taken as

(a) u is the one whose nth derivative can easily be obtained

(b) v is the one whose derivatives reduce to zero after a few number of differentiation.

Example:

Y
Obtain the nth derivative of
y = x3 e2x
Solution
Given

EM
y = x3 e2x
Choose
v = x3 ; reason: fourth derivative is zero
u = e2x ; reason: u(n) = 2n e2x
Using Leibnitz theorem:
n
X
y (n) = n
Cr u(n−r) v (r)
r=0

u = e2x , u(n) = 2n e2x

AD
Now, v = x3 ; v (1) = 3x(2) ; v (2) = 6x; v (3) = 6; v (4) = 0; v (5) = 0; · · ·

y (n) =n C0 u(n) v (0) +n C1 u(n−1) v (1) +n C2 u(n−2) v (2) +n C3 u(n−3) v (3) +n C4 u(n−4) v (4) + · · ·
n(n − 1) (n−2) 2x n(n − 1)(n − 2) (
= 2n e2x · x3 + n2(n−1) e2x · 3x2 + 3 e · 6x + 2 n − 3)e2x · 6 + 0 + 0
2! 3
= 2(n−3) · e2x 23 · x3 + 3n · 22 · x2 + n(n − 1)6x + n(n − 1)(n − 2)
 

= e2x · 2(n−3) 8x3 + 12nx2 + 6n(n − 1)x + n(n − 1)(n − 2)

 
AC

4 Partial,Higher and Total Derivatives

4.1 Partial Derivatives
Recall that if F : X ⊆ < 7→ < is a scalar-valued function of one variable, then the derivative of F at a
number a ∈ X is  
0 F (a + h)F (a)
F (a) = lim (1)
h→0 h
Moreover,F is said to be differentiable at a precisely when the limit in equation (1) exist.
A

Definition: The partial derivative of f with respect to xi is the (ordinary) derivative of the partial
function of f with respect to xi .
Remark: A partial function is obtained by allowing only one variable to vary, while all the other
variables are held fixed.
BT

Standard notations for the partial derivative of f with respect to xi are:

∂f
, ∆xi f (x1 , · · · , xn ) f xi (x1 , · · · , xn )
∂xi
Symbolically, we have  
∂f f (x1 , · · · , xn ) − f (x1 , · · · , xn )
= lim
∂xi h→0 h
By definition, the partial derivatives is the (instantaneous) rate of change of f when all variable, except
the specified one, are held fixed.
Note: for the most,partial derivatives are easy to compute, once you become adept at treating variables

13
like constants.
Example:
If f (x, y) = x3 + x2 y 3 − 2y 2 , find fx (2, 1) and fy (2, 1)
Solution Holding y constant and differentiating w.r.t x, we get
fx (x, y) = 3x2 + 2xy 3
and so,

Y
fx (2, 1) = 3 · 22 + 2 · 2 · 13 = 12 + 4 = 16
Similarly, holding x constant and differentiating w.r.t to y, we get
fy (x, y) = 3x2 y 2 − 4y and
fy (2, 1) = 3 · 22 · 12 − 4 · 1 = 12 − 4 = 8

EM
Remark:
In the case where f is a (Scalar-valued) function of two variables, we can understand
∂f
(a, b)
∂x
geometrically as the slope at the point (a, b, f (a, b)) of the curve obtained intersecting the surface
Z = f (x, y) with the plane y = b.Similarly,
∂f
(a, b)
∂y

Example:
AD
is the slope at (a, b, f (a, b)) of the curve formed by the intersection of Z = f (x, y) and x = a

If f (x, y) = x2 y + cos(x + y)
∂f ∂f
then = 2xy − sin(x + y) and = x2 − sin(x + y)
∂x ∂y
Example:
If x2 y 00 + xy 0 = 0 , show that
2 (n+2) (n+1)
x y + (2n + 1)xy + (n2 + 1)y (n) = 0
AC

Solution
Given x2 y 00 + xy 0 + y = 0 (2)
differentiating (1) n times, using the Leibnitz theorem yields
  
(n+2) 2 (n+1) n(n − 1) (n) 
y x + ny 2x + y · 2 + 0 + 0 + y (n+1) x + ny (n) · 1 + 0 + 0 + · · · + y (n)
2!
Simplifying
x2 y (n+2) + (2n + 1)xy (n+1) + (n2 + 1)y (n) = 0 as required
Exercise:
A

(i)
2 +x
If y = ex , Show that,y 00 = y 0 (2x + 1) + 2y and hence prove that
(n+2) (n+1)
y = (2x + 1)y + 2(n + 1)y (n)
BT

(ii)
y = x3 (2x + 1)2 , find y (4)
(iii) π  π
If y = (2x − π)4 sin , evaluate y (6) when x =
2 2
2 2
Similarly, if u(x, y) = xy/(x + y ), then, from the quotient rule of ordinary calculus, we have
(x2 + y 2 )y − xy(2x) y(y 2 − x2 )
ux (x, y) = =
(x2 + y 2 )2 (x2 + y 2 )2
(x2 + y 2 )x − xy(2y) x(x2 − y 2 )
and ux (x, y) = =
(x2 + y 2 )2 (x2 + y 2 )

14
4.2 Higher Derivatives
If f is a function of two variables, then its partial derivatives fx , & f are also function of two variables.
So, we can consider their partial derivatives (fx )x , (fx )y , (fy )x and (fy )y which are called second partial
derivatives of f . if z = f (x, y), then
∂2z ∂2z
 
∂ ∂z
(fx )x = fxx = ; (f )
y x = f yx = =
∂x2 ∂x ∂y ∂x∂y

Y
∂2z ∂2z
 
∂ ∂z
(fx )y = fxy = ; (fy )y = fyy = =
∂y∂x ∂y ∂y ∂y 2
NOTE:

EM
fxy = fyx {Alexis clairant: 1713-1765}

Exercises:
1. Calculate
fxyz if f (x, y, z) = sin(3x + yz)

2. Find the first partial derivatives of the following:

(i)
f (x, y) = x2 y 2 (x4 + y 4 )
(ii)

(iii)
AD f (x, y) = ln(x2 + y 2 )

y
z = xx
3. Verify the Clairant’s theorem that uxy = uyx on the following
(i)
u = x5 y 4 − 3x2 y 3 + 2z 2
AC

(ii)
u = sin2 x cos y
(iii)
u = sin−1 (xy 2 )

4.3 Total Derivatives

Definition: Let f : X ⊆ <n 7→ < and let a ∈ <. The incremented change in f , denoted by ∆f , is
∆f = f (a + h) − f (a)
A

The total differential of f , denoted by df (a, h), is

∂f ∂f ∂f
df (a, h) = (a)h1 + (a)h2 + · · · + (a)hn
∂x1 ∂x2 ∂xn
BT

The significance of the differential is that for h ≈ 0

∆f = df
Note: We have abbreviated df (a, h) by df .
Sometimes, h1 is replaced by the expression ∆x1 or dx1 to emphasize that it represents a change in
the independent variable, in which case we write
∂f ∂f ∂f
df = dx1 + dx2 + · + dxn
∂x1 ∂x2 ∂xn
n
X ∂f
= dxj
∂xj
j=1

15
Remark: We’ve suppressed the evaluation of the partial derivatives at a, as is customary.
Example:

Suppose
f (x, y, z) = sin(xyz) = + cos(xyz) . Then
∂f ∂f ∂f
df = dx + dy + dz
∂x ∂y ∂z

Y
= yz{cos(xyz) − sin(xyz)}dx + xz{cos(xyz) − sin(xyz)}dy + xy{cos(xyz) − sin(xyz)}dz
= {cos(xyz) − sin(xyz)}{yz dx + xz dy + xy dz}
Example:
Let f (x, y) = x − y + 2x2 + xy 2 . Then for (a, b) = (2, 1),we have that the increment is

EM
∆f = f (2 + ∆x , −1 + ∆x) − f (2, 1)
= 2 + ∆x − (−1 + ∆y) + 2(2 + ∆x)2 + (2 + ∆x)(−1 + ∆y)2 − 13
= 10∆x − 5∆y + 2(∆x)2 − 2∆x∆y + 2(∆y)2 + ∆x(∆y)2
On the other hand,
df ((2, −1), (∆x, ∆y)) = fx (2, −1)∆x + fy (2, −1)∆y
= (1 + 4x + y 2 )|(2,−1) ∆x + (−1 + 2xy)|(2,−1) ∆y
AD = 10∆x − 5∆y

We see that df consist of exactly the terms od ∆f that are linear in ∆x and ∆y (i.e appear to first
power only). This will always be the case, of course,since that is the nature of the first-order Taylor
approximation. Use of the differential approximation is often sufficient in practice,
for when ∆y and ∆x are small, Higher powers of them will be small enough to make virtually negligible
contribution to ∆f . For example, if ∆x & ∆y are both 0.01,then
df = 0.1 − 0.05 = 0.05, and
∆f = (0.1 − 0.05) + 0.0002 − 0.0002 + 0.0002 + 0.000001
AC

= 0.05 + 0.000201 = 0.050201.

Thus, the values of df and ∆f are the same to three decimal places
Example:
A wooden rectangular block is to be manufactured with dimensions 3inch × 4inch × 6inch.Suppose
that the possible error in measuring each dimensions of the block is the same. We use differentials
to estimate how accurately we must measure the dimensions so that the resulting calculated error in
volume is no more than 0.1
Solution
Let the dimensions of the block be denoted by x(≈ 3 inch), y(≈ 4 inch), and z(≈ 6 inch). Then the
A

volume of block is.

V = xyz and V ≈ 3 · 4 · 6 = 72 inch3
The error in calculated volume is ∆V , which is approximated by the total differential dV . Thus
BT

∆V ≈ dV = Vx (3, 4, 6)∆x + Vy (3, 4, 6)∆y + Vz (3, 4, 6)∆z

= 24∆x + 18∆y + 12∆z
If the error n measuring each dimension is ∈, then we have

∆x = ∆y = ∆z =∈

Therefore,
dV = 24∆x + 18∆y + 12∆z = 24 ∈ +18 ∈ +12 ∈ = 54 ∈
To ensure (approximately) that |∆V | ≤ 0.1, we demand

|∆V | = |54 ∈ | ≤ 0.1

16
Hence,
0.1
|∈|≤ = 0.0019 in
54
So the measurements in each dimension must be accurate to within 0.0019inch.
Example:
The formula for the volume of a cylinder of radius r and height h is V (r, h) = πr2 h. If the dimensions
are changed by small amounts ∆r and ∆h, then the resulting change ∆V in volume is approximated
by the differential change∆V . That is

Y
∂V ∂V
∆V ≈ dV = df = ∆r + ∆h = 2πrh∆r + πr2 ∆h
∂r ∂h
Suppose the cylinder in actually a soft drink can, so that it has approximately dimensions of r = 1

EM
inch and h = 5 inch. Then
dV = π(10∆r + ∆h)
This statement show that, for these particular values of rand h, the volume is approximately 10 times
more sensitive to change in radius than change in height.i.e, if the radius is changed by an amount ∈,
then the height must be changed by roughly10∈ to keep the volume constant (i.e to make ∆V zero).
We used the word ”approximate” because our analysis arises from considering the differential change
dV , rather than the actual incremental change ∆V .

This soft drink can example has real application to product marketing strategies. Because the volume
AD
is so much more sensitive to change in radius than height, it is possible to make a can appear larger
than standard by decreasing its radius slightly (little enough so as to be hardly noticeable) and in-
creasing the height so no change in volume occurs.
Remark: Differentials and sensitivity analysis are great tools for industries.

Exercises:
1. Determine the total differential of the following functions

(i)
AC

f (x, y) = x2 y 3
(ii)
f (x, y, z) = x2 + 3y 2 − 2x3
(iii)
f (x, y, z) = cos(xyz)
(iv)
f (x, yz) = ex cos y + ey sin z
(v)
A

1
f (x, yz) = √
xyz
2. Near the point (1,-2,1), is the function
BT

g(x, y, z) = x3 − 2xy + x2 z + 7z most sensitive to changes in x, y, or z ?

3. To which entry in the matrix is the value of the determinant

2 3
most sensitive
−1 5

4. To estimate the volume of a cone of radius approximately 2m and height approximately 6m, how
accurately should the radius and height be measured so that the error in the calculated volume
estimated does not exceed 0.2m3 ? Assume that the possible error in measuring the radius and
height are the same.

17
5 Coordinate Systems
It is assumed that you are comfortable with Cartesian (rectangular) coordinate for <2 and <3 . The
Cartesian coordinate system will continue to be of prime importance to us, but with time, we
nd it advantageous to use different coordinate systems. In <2 , polar coordinates are useful for
describing figures with circular symmetry. In <3 , there are two particularly valuable coordinate
systems besides Cartesian coordinate,Cylindrical and Spherical coordinates.

Y
5.1 Cartesian and Polar Coordinate on <2
A point P(x,y) on the Cartesian coordinate system is a location on the <2 -plane such that two
perpendicular lines (one cutting the x-axis and the other cutting the y-axis) intersect at P(x,y)

EM
AD
AC
A
BT

18
 Polar coordinate are defined by considering
different geometric information.For point P other
than the origin, we assign to P the polar co-
ordinates (r,θ), where r is the radius of the
circle on which P lies and θ is the angle be-
tween the positive-axis and the ray on which
P lies, (θ is measured as opening counterclock-
wise)

Y
EM
Figure 1: A point P(x,y) on the Cartesian coordi-
nate

AD
AC

Figure 2: Polar coordinate system

The basic conversions between polar and Cartesian coordinates are provided by the following relations.
(
x = r cos θ
Polar to Cartesian:
y = r sin θ
A

(
r2 = x2 + y 2
Cartesian to Polar:
tan θ = xy
BT

5.2 Cylindrical coordinates

Cylindrical coordinates on <3 are a ”naive” way of generating polar coordinates to three dimensions
in the sense that they are nothing but polar coordinate used in place of the x- and y-coordinates.
To locate a point P on a cylindrical coordinate system, it is enough to give the radius of the cylinder,
the circumferential angle θ around the cylinder and the vertical position z along the cylinder, i.e,
P(r,θ,z)

19
 Y
EM
Figure 3: Cylindrical coordinate system

The following relations provide a way of converting from and to Cartesian to and from cylindrical
coordinates 
x = r cos θ

Cylindrical to Cartesian: y = r sin θ

z =z

AD 
r
 2 = x2 + y 2
Cartesian to Cylindrical: tan θ = xy


z =z

5.3 Spherical Coordinates

Roughly speaking, the spherical coordinates of P are given by specifying the radius ρ of thee sphere
containing P and the latitude & longitude readings of P along this sphere. More precisely, the spherical
AC

coordinates (ρ,ϕ,θ) of P are defined as follows:

ρ is the distance from P to the origin,
ϕ is the angle between the positive z-axis and the ray through the origin and P,
θ is thee angle between the positive x-axis and the ray made by dropping a perpendicular from P to
the xy-plane. Note: It is standard practice to impose the following restrictions on the range of values
for the individual coordinates.

ρ ≥ 0, 0 ≤ ϕ ≤ π, 0 ≤ θ < 2π

The basic relations between spherical coordinates and both cylindrical and Cartesian coordinates are
as follows
A

Spherical/cylindrical  
2 = r2 + z 2
r = ρ sin ϕ
 ρ

θ =θ tan ϕ = zr
 
z = ρ cos ϕ θ =θ
 
BT

Spherical/Cartesian

2 = x2 + y 2 + z 2

x = ρ sin ϕ cos θ ρ √
 

x2 +y 2
y = ρ sin ϕ sin θ tan ϕ = z
 

z = ρ cos ϕ tan θ = y

z

20
6 Wronskian: Linear Dependence
Definition: A set of n-given functions y1 (x); y2 (x), · · · yn (x) is said to be linearly dependent if there
exist n constants c1 , c2 , · · · cn not all zero such that

c1 y1 (x) + c2 y2 (x) + · · · + cn yn (x) = 0

and the set is linearly independent otherwise, i.e when

Y
c1 = c2 = · · · cn is equal to zero.

6.1 The Wronskian

EM
The set of n functions y1 (x); y2 (x), · · · yn (x) assumed to be at least (n-1) times differentiable over an
interval a ≤ x ≤ bhas associated with it a determinant of a matrix called the wronskian of the function.
This is given by
y1 y2 ······ yn

y10 y20 ······ yn0

w(x) =
.. .. .. ..
. . . .

AD (n−1) (n−2) (n−1)

y1 y2 ··· yn
The set of functions y1 ; y2 , · · · yn is said to be linearly independent over an interval if and only if the
determinant w(x) is not equal to zero on the interval.
Example:
Consider the functions 1, cos x, sin x
Solution
1 cos x sin x

w(x) = 0 − sin x cos x =1

AC

0 − cos x − sin x
∴ since w(x) 6= 0, it is linearly independent.
Exercises:

1. Verify whether the following functions are linearly independent

(a)
ex , e2x , e3x
(b)
A

ex , cos x, sin x
(c)
1, sin2 x
BT

(d)
cos(wt − β), cos wt, sin wt, (required as f(t))

2. Let (
1 + x3 , x ≤ 0
y1 (x) =
1 , x≥0
(
1 , x≤0
y2 (x) 3
1+x , x≥0
n
y3 (x) = 3 + x3 ∀x

21
 (a) show that y 00 are continuous for all x each of y1 , y2 , y3
(b) show that the wronskian of y1 , y2 , y3 is zero for all x
(c) Prove that y1 , y2 , y3 are linearly independent over the interval −1 ≤ x ≤ 1

7 Line Integrals
If a field exists in the x-y plane, producing a force F on a particle at the point K, them F can be

Y
resolved into two components

EM
AD Figure 4: Line integrals components

Ft along the tangent to the curve AB at K

Fn along the normal to the curve AB at K

The work done in moving the particle through a small distance δs from K to L along the curve
is then approximately Ft δs. So the total work done in moving a particle along the curve from A to B
is given by
AC

X Z
lim Ft δs = Ft δs, from A to B
δs→0
R R
This is normally written Ft δs where A and B are the end points of the curve, or as c Ft ds where
the curve c connecting A and B is defined.

Such an integral thus formed is called a line integral since integration is carried out along the
path of the particular curve c joining A and B
Z Z
Ft ds = Ft ds
AB c
A

Where c is the curve y = f (x) between A(x, y) and B(x2 , y2 )

BT

22
7.1 alternative form of a line integral
It is often convenient to integrate with respect to x or y than to take arc length as the variable. If Ft
has a component P in the x-direction and Q in the y-direction, then the work done from K to L can
be stated as
P δx + Qδy

Z Z

Y
Ft ds = (P dx + Qdy)
AB AB

In general then, the line integral can be expressed as

Z Z

EM
I= Ft ds = (P dx + Qdy)
c c

where c is the prescribed curve and Ft , P and Q are func-

tions of x and y

Example: R R
Evaluate

I=
Z

c
AD
{(x2 + 2y) dx + xy dy}
Figure 5: I = c Ft ds

from O(0, 0) to B(1, 4)along the curve y = 4x2

= c (P dx + Qdy)

In this case, c is the curve y y = 4x2

dy = 8x
Now
Z 1
AC

I= (9x2 + 32x4 ) dx = 9.4 {substitute for y in the integral and apply the limits}
0

7.2 Properties of line integrals

1. Z Z
F ds = (P dx + Qdy)
c c

2. Z Z Z Z
F ds = − F ds and (P dx + Qdy) = − (P dx + Qdy)
AB BA AB BA
A

i.e the sign of a line integral is reversed when the direction of the integration along the path is
reversed.

3. If the path of integration c joining A to B is divided into two parts AK and KB, then Ic =
IAB = IAK + IKB
BT

4. In all cases, the function y = f (x) that described the path of the integration involved must be
continuous and single-valued.

5. If the function y = f (x) that described the path of integration c is not single-valued for part of
its extent, the part is divided into two sections.

y = f1 (X) from AtoK

y = f2 (x) from KtoB

23
 Y
EM
Figure 6: Properties of line integrals

Example:
Evaluate Z
I= (x + y) dx from A(0, 1) to B(0, −1)along the
c

AD semi-circle x2 + y 2 = 1 for x ≥ 0
The function f(x) = y that describes the path of integration c is not single-valued. For any value of
1
x, y = ±(1 − x2 ) 2 .Therefore, we divide c into two part

(i)
1
y = (1 − x2 ) 2 from A to K

(ii)
1
y = −(1 − x2 ) 2 from K to B
AC

Now, Z
I= (P dx + Qdy) , but Q = 0
c
in this case Z Z 1 1
Z 0 1
2
I= P dx = (x + (1 − x ) ) dx +
2 (x − (1 − x2 ) 2 ) dx
c 0 1
Z 1 1
= 2(1 − x2 ) 2 dx
0
Z 1 1
=2 (1 − x2 ) 2 dx
A

0
Z π
2
= (1 − cos 2θ) dθ
0
π
=
BT

24
 Definition: A region is said to be simply connected if a
Figure 7: Example path joining A and B can be deformed to coincide any other
line joining A and B without going outside the region. In
other words, a region is simply connected if any closed path
on the region can be contracted to a single point without
leaving the region.

7.3 Line integral around a closed curve

Y
An integral to be evaluated around a closed curve on the
positive (counterclockwise) direction is usually denoted by
the symbol I

EM
Note: Unless otherwise required, we always proceed around the curve in a counterclockwise direction.
With a closed curve, the y-values on the path c cannot be single-value. Therefore, we divide the path
into two or more parts and treat each separate.
Example:

Evaluate the line integral I

AD I= {x2 dx − 2xy dy}
c
where c comprises the sides of the triangle joining
O(0, 0), A(1, 0) and B(0, 1)
1
Ans: I = −
3
Example:

Evaluate I
y dx
AC

c:x2 +y 2 =4

Ans: − 4π
Example:

Evaluate
I
I = {xy dx+(1+y 2 ) dy} where c is the boundary of the rectangle joining A(1,0),B(3,0),C(3,2) and D(1,2)
c

Ans: − 8
A

7.4 Line Integral in Space

Suppose that C is a smooth space curve given by the parametric equations
BT

x = x(t), y = y(t) and z = z(t), t ∈ [a, b]

or by a vector equation
r(t) = x(t)i + y(t)j + z(t)k.
If f is a function of three variables that is continuous on some region containing C, then we de
ne the line integral of f along C as

Z Z b  2  2  2 ! 12
dx dy dz
f (x, y, z) ds = f (x(t), y(t), z(t)) · + +
c a dt dt dt

25
Example:

Evaluate Z
y sin z ds ,
c
where c is the circular helix given by the equations

x = cos t, y = sin t, z = t, 0 ≤ t ≤ 2π

Y
Solution
s 2 2 2
Z Z 2π  
dx dy dz
y sin z ds = (sin t) sin t · + + dt
c 0 dt dt dt

EM
Z 2π p
= sin2 t sin2 t + cos2 t + 1 dt
0
√ Z 2π 1
= 2 (1 − cos 2t) dt
0 2
√  2π
2 1
= t − sin 2t
2 2 0
√
= 2π

7.5
AD
Dependence of line integral on the path of integration
Integration along two separate paths joining the same two end point does not necessary give identical
results
Example:
Evaluate Z
I = {xy 2 + 2x3 ydy} between O(0,0) and A(2,4)
c

(a) along c1 , i.e y = x2

AC

(b) along c2 , i.e y = 2x

(c) along c3 , i.e x = 0, from (0,0) to (0,4) and y = 4 from (0,4) to (2,4)

7.6 Exact differential

If P dx + Qdy is an exact differential where P and Q and their first derivatives are
finite and continuous inside the simply connected region R, then

(i)
∂P ∂Q
=
A

∂y ∂x

(ii)
Z
BT

I= (P dx + Qdy) is independent of the path of integration where c lies entirely within R.

c

(iii) I
I= {P dx + Qdy} is zero when c is a closed curve lying entirely within R.
c

Note: The above cane be extended to an equation in three variables x, y, z i.e P dx + Qdy + Rdz is
exact if
∂P ∂Q ∂P ∂R ∂Q ∂R
= , = , =
∂y ∂x ∂z ∂x ∂z ∂y

26
If it is exact, then.
Z
(P dx + Qdy + Rdz) is independent of the path of integration
c
I
{P dx + Qdy + Rdz} is zero when c is closed curve
c
Example:

Y
Show that Z
I= {(3x2 sin y + 2 sin 2x + y 3 ) dx + (x3 cos y + 3xy 2 ) dy}
c
is independent of the path of integration and evaluate it from A(0, 0) to B( π2 , π)

EM
7.7 Double Integral
Definition: The double integral of f on R, denoted by
ZZ  ZZ ZZ 
f dA or by f (x, y) dA or by (x, y) dx dyf ,
R R R

is the limit of the Riemann sum S ass the dimension ∆xi and ∆yi of the sub-rectangles Rij all approach
zero, i.e  
ZZ n
AD
provided of course, that thisRR
f dA =

limit exists.
lim
R

X
f (cij )∆xi ∆yi 
all ∆xi ,∆yi → 0
i,j=1

Note: The double integral R f dA can be used to compute the ”net volume” under the graph of
z = f (x, y) from a geometric point of view.

7.8 Fubini’s Theorem

Let f be bounded on R = [a, b] × [c, d], and assume that the set S of discontinuities of f on R has zero
AC

area. If every line parallel to the coordinate axes meets S in at most finitely many points, then
ZZ Z bZ d Z dZ b
f dA = f (x, y) dx dy = f (x, y) dx dy
R a c c a

Example:

Consider ZZ
x dA , Where R = [−2, 2] × [−1, 3]
R
ZZ Z 2 Z 3 Z 2 y=3 Z 2
A

= x dy dx = xy dx = 4x dx = 0
R −2 −1 −2 y=−1 −2

Now,
Z 3 Z 2 Z 3 x=2 Z 3
1 2
x dx dy = x = (2 − 2) dy = 0
BT

−1 −2 −1 2 x=−2 −1

7.9 Definition
We say that D is an elementary region in the plane if it can
be described as a subset of <2 of one of the following three
types
Type 1:

D = {(x, y) | γ(x) ≤ y ≤ δx, a ≤ x ≤ b}

where γ and δ are continuous on [a,b]

27
 Type 2:

D = {(x, y) | α(y) ≤ x ≤ β(y), c ≤ y ≤ d}

Where α and β are continuous on [c,d]

Y
EM
Figure 9: D = {(x, y) | α(y) ≤ x ≤ β(y), c ≤ y ≤ d}

Type 3:
AD
D is of both type 1 and type 2
AC

Figure 10: (a) The unit disk is a type 3 region (b) The unit disk as a type 1 region (c) The unit disk
as a type 2 region

7.10 Theorem
A

Let D be an elementary region in <2 and f a continuous function on D

(i) If D is of type 1, then

ZZ Z bZ δ(x)
BT

f dA = f (x, y) dy dx
D a α(x)

(ii) If D is of type 2, then

ZZ Z d Z β(y)
f dA = f (x, y) dx dy
D c α(y)

28
 Example:
Let D be the region bounded by the parabolas y = 3x2 ,
y = 4 − x2 and the y-axis. Since the parabolas intersect at
the point (1,3). Since D is a type 1 elementary region, we
may use the last theorem with f (x, y) = x2 y to find that
ZZ Z 1 Z 4−x2
2
x y dA = x2 y dy dx

Y
D 0 3x2

The limits for the first (inside) integration come from the
y-values of the top and bottom boundary curves of D.
The limits for second (outside) integration are the con-

EM
stant x-values that corresponds to the straight left and
right side of D. the evaluation itself is fairly straight for-
ward.

Figure 11: Example

1 Z 4−x2 1 y=4−x2
x2 y 2
Z Z
2
x y dy dx = dx
0 3x2 0 2 y=3x2
AD =

=
Z 1

1 1 2
Z

2 0
0
x2
2
{(4 − x2 )2 − (3x2 )2 } dx

x (16 − 8x2 + x4 − 9x4 ) dx

Z 1
= (8x2 − 4x4 − 4x6 ) dx
0
8 4 4 136
= − − =
3 3 7 105
AC

Example:
Let D be the region shown below (having a rectangle border)
A

Figure 12: Example

BT

Consider ZZ
(1 − x − y) dA.
D
Note that D is a type 3 elementary region, so there should be two ways to evaluate the double integral

29
considering D as a type 1 elementary region.
ZZ Z 1Z 1−x
(1 − x − y) dA = (1 − x − y) dy dx
D 0 0
1  y=1−x
y2
Z 
= y − xy − dx
0 2 y=0
Z 1
(1 − x)2

= (1 − x) − x(1 − x) − dx

Y
0 2
Z 1 1
(1 − x)2 1 1
= dx = − (1 − x)3 =
0 2 6 0 6
We can also consider D as a type 2 elementary region to obtain

EM
ZZ Z 1 Z 1−y
(1 − x − y) dA = (1 − x − y) dx dy
D 0 0

We leave it to you to check explicitly that this iterated integral also has a value of 61
Note: Of course not all regions in the plane are elementary, including even some relatively simple
ones. To integrate continuous functions over such regions, the best advice is to attempt to subdivide
the region into
finitely many of elementary type.
Example: AD
Let D be the annular region between the two concentric circles of radii 1 and 2 as shown below. Then
D is not an elementary region, but wee can break D up into four sub-regions that are of elementary
type.
AC

Figure 13: Example

We may compute the double integral as the sum of the integral over sub-regions. That is
A

ZZ ZZ ZZ ZZ ZZ
f dA = f dA + f dA + f dA + f dA
D D1 D2 D3 D4

For the type 1 sub-regions, we have

BT

√ √
ZZ Z 3 Z 4−x2
f dA = √ f (x, y) dy dx
D1 − 3 1

and √
ZZ Z 3 Z −1
f dA = √ √ f (x, y) dy dx
D3 − 3 4−x2
For the type 2 sub-regions, we have
ZZ Z 1 Z √4−y2
f dA = √ f (x, y) dx dy
D2 −1 1−y 2

30
and √
ZZ Z 1 Z − 1−y 2
f dA = √ f (x, y) dx dy
D4 −1 − 1−y 2

The difficulty of evaluating each of the preceding four iterated integral then depends on the complexity
of the integrand.
Exercises:

Y
1. Evaluate the given integration. In addition, sketch the regions D that are determined by the
limits of integration.

(a)
Z 2 Z y2
y dx dy

EM
0 0

(b)
Z 3 Z 2x+1
xy dy dx
−1 x

(c)
Z π Z sin x
y cos x dy dx
0 0

2. Evaluate AD ZZ
(x2 + y 2 ) dA
D
where d is the region in the first quadrant bounded by y = x, y = 3x, and xy = 3

3. Set up an appropriate iterated integral to

nd the area of the region bounded by the graph at y = x3 − x and y = ax2 for x ≥ 0 (a =
constant).
AC

8 Surface integrals
The surface integral of φ(x, y, z) over the surface S is given by

(a) Z
I= φ(x, y, z) ds
s
and

(b) s
∂z 2
Z Z   2
∂z
I= φ(x, y, z) 1 + + dx dy ; where z = f (x, y)
A

R ∂x ∂y
R
Note: When φ(x, y, z) = 1, then I = s ds gives the area of the surface S
s 2 2
BT

Z Z Z  
∂z ∂z
S= ds = 1+ + dx dy
s R ∂x ∂y

Example:
√
Find the area of the surface z = x2 + y2 over the region bounded by x2 + y 2 = 1
Solution
Since Area of a surface S is v
Z Z p
S= 1 + (∂x)2 + (∂y)2 dx dy
R

31
 ∂z x ∂z y
Now ∂x = z and ∂y = z
s 2 2 s
x2 y2 √
 
∂z ∂z
1+ + = 1+ 2 2
+ 2 2
= 2
∂x ∂y x +y x +y

√ Z Z
S= 2 dx, dy = 2 × area of the region R
R

Y
Since R is bounded by x2 + y2 = 1, i.e a circle the origin and radius 1, then, area of the region
R is π
√ Z Z
S= 2 dx dy = 2π
R

EM
9 Parameterized Surfaces
Definition: Let DD a region in <2 that consists of a connected open set, possibly together with
some or all of its boundary points. A parameterized surface in <3 is a continuous function.

X : D ⊆ <2 7→ <3

that is one-one on d,except possibly along δD. We refer to the image X(D) as the underlying surface
of x (or the surface parameterized by and denote it by S).
Note: The restriction on the region D and map x of the above definition are meant to ensure that D
AD
is a two-dimensional subset of <2 with a two-dimensional image. If we write the component functions
of X, then, for (s, t) ∈ D
X(s, t) = (x(s, t), y(s, t), z(s, t)) (1)
and the underlying surface can be described by the parametric equations

x = x(s, t)

y = y(s, t) (s, t) ∈ D

z = z(s, t)

AC

If we write the component functions of X as (1), we find that

i j k

∂x ∂y ∂z
Ts × Tt = ∂s ∂s ∂s

∂x ∂y ∂z
∂t ∂t ∂t

     
∂y ∂z ∂y ∂z ∂x ∂z ∂x ∂z ∂x ∂y ∂x ∂y
= − i+ − j+ − k
A

∂s ∂t ∂t ∂s ∂t ∂s ∂s ∂t ∂s ∂t ∂t ∂s

Using the notation of Jacobian, we obtain

∂(y, z) ∂(x, z) ∂(x, y)

BT

N (s, t) = Ts × Tt = i+ j+ k
∂(s, t) ∂(s, t) ∂(s, t)

10 Triple Integrals
The tripe integral of f on B, denoted by ZZZ
f dV
B
Or by ZZZ
f (x, y, z) dV
B

32
Or by ZZZ
f (x, y, z) dx dy dz
B
is the limit of the Riemann sum S as the dimensions ∆xi , ∆yj and ∆zk of the sub-boxes Bijk all
approach zero,1.e
ZZZ Xn
f dV = lim f (Cijk )∆xi ∆yj ∆zk
B all∆xi ,∆yj ,∆zk 7→ 0
i,j,k=1

Y
provided that the limit exists
Theorem: (Fubini’s Theorem)
Let f be bounded on B = [a, b] × [c, d] × [p, q] and assume that the set S of discontinuities of f has
zero volume, if every line parallel to the coordinate axes meets. S in at most

EM
finitely many points, then
ZZZ Z bZ dZ q Z bZ q Z d
f dV = f (x, y, z) dx dy dz = df dy dz dx
B a c p a p c
Z dZ bZ q Z dZ q Z b
= f dz dx dy = f dx dz dy
c a p c p a
Z q Z bZ d Z q Z dZ b
= f dy dx dz = f dx dy dz
p a c p c a

11
AD
Vector Surface Integrals
Let X : D 7→ <3 be a smooth parameterized surface,a when D is a bounded region in the plane, and
let F (x, y, z) be a continuous
RR vector field whose domain includes S = X(D). The vector surface integral
F along X, denoted by X F · ds is
ZZ ZZ
F · ds = F (X(s, t)) · N (s, t) ds dt , where X is the helioid
X D
AC

0 ≤ t ≤ 2πX(s, t) = (s cos t, sin t, t), 0 ≤ s ≤ 1,

Where have
∂(y, z) ∂(x, z) ∂(x, y)
N (s, t) = i− j+ k
∂(s, t) ∂(s, t) ∂(s, t)
sin t s cos t cos t − sin t cos t − sin t
= i− j+
0 1 0 1 sin t s cos t
= sin ti − cos tj + sk

ZZ Z 2π Z 1
F · ds = F (X(s, t)) · N (s, t) ds dt
A

X 0 0
Z 2π Z 1
= (s cos ti + s sin tj + (t − 2s sin t)k ) ds dt
0 0
Z 2π Z 1
BT

= (st − 2s2 sin t) ds dt = π 2

0 0

x = 3 cos s

S1 (lateral cylindrical surface) : y = 3 sin s , 0 ≤ s ≤ 2π

z =t 0 ≤ t ≤ 15


x = s cos t

S2 (bottom disk) : y = s sin t 0≤s≤3

z =0 0 ≤ t ≤ 2π


33
and 
x
 = s cos t
S3 (top disk) : y = s sin t , 0≤s≤3

z = 15 0 ≤ t ≤ 2π


Ans: 675π+0+135π = 810π

Remark:
If a surface S is given by the graph of z = g(x, y),whereg is of class C 1 on some region D in <2 , then

Y
S is parameterized by
X(x, y) = (x, y, g(x, y)) with(x, y) ∈ D. then
N (x, y) = −gx i + gy j + k

EM
so that ZZ ZZ q
f ds = f (x, y, g(x, y)) gx2 + gy2 + 1 dx dy
X D
Example:
Suppose S is the graph of the portion of the paraboloid
z = 4 − x2 − y 2 , where (x, y) varies throughout the disk

D = {(x, y) ∈ <2 x2 + y 2 ≤ 4}

To calculate
AD ZZ
(4−z) ds ,
X
where X(x, y) = (x, y, 4−x2 −y 2 )

Figure 14: Example

AC

Example:
Evaluate ZZ
z 3 ds where X : [0, 2π] × [0, π] 7→ <3
X
is the parameterized sphere of radius a

Solution
X(s, t) = (a cos s sin t , a sin s sin t , a cos t)
s
∂(x, y) 2 ∂(x, z) 2 ∂(y, z) 2
    
kN (s, t)k = + +
A

∂(s, t) ∂(s, t) ∂(s, t)

q
= a4 sin2 t cos2 t + sin2 s sin4 t + cos2 s sin4 t


p
= a2 sin2 t cos2 t + sin4 t
BT

= a2 sin t
Hence ZZ Z π Z 2π
3
z ds = (a cos t)3 a sin t ds dt
X 0 0
Z π
5
=a 2π cos3 t sin t dt
0
 π   
5 1 4 5 1 1
= 2πa − cos t = 2πa − − − =0
4 0 4 4

34
Note: To define and evaluate scalar surface integral over piece-wise smooth parameterized sur- faces,
simply calculate the surface integral over each smooth piece and add the results.
Example:

Let S be the closed cylinder of radius 3 with axis along the z-top face at z = 15 and bottom face at
z = 0. The S is a piece-wise smooth surface. From the
figure, S is the union of the three smooth pa-
rameterized surface S1 , S2 , S3 described next.

Y
Calculate ZZ
z ds
s

EM
Note: To write the integral
AD
Figure 15: Example
Z Z
F (x, y) dx dy
R

in terms of the new variables u and v, where x = f (u, v) and y = g(u, v), we for x and y in F (x, y)
and replace dxdy with
∂(x,y)
∂(u,v) du dv

The integral then becomes. Z Z

∂(x,y)
F (f (u, v), g(u, v)) ∂(u,v) du dv
R
AC

For transformation in three dimensions If x = f (u, v, w); y = g(u, v, w); z = h(u, v, w)

∂x ∂y ∂z
∂u ∂u ∂u

∂(x, y, z) ∂x ∂y ∂z
Then J(u, v, w) = = ∂v ∂v ∂v
∂(u, v, w)
∂x ∂y ∂z
∂w ∂w ∂w

and the element of volume dv = dxdydz becomes

dv = |J(u, v, w)| du dv dw
A

Also ZZZ
F (x, y, z) dx dy dz is transformed into
BT

ZZZZ
∂(x, y, z)
G(u, v, w) du dv dw
∂(u, v, w)
In particular we have
ZZ ZZ p
(4 − z) ds = (4 − (4 − x2 − y 2 )) 4x2 + 4y 2 + 1 dx dy
X Z ZD p
= (x2 + y 2 ) 4x2 + 4y 2 + 1 dx dy
D

35
To integrate, we switch to polar coordinates, i.e let x = r cos θ and y = r sin θ, where 0 ≤ r ≤ 2,
0 ≤ θ2π The integral becomes
Z 2π Z 2 p Z 2 Z 2π p
2 2
r 4r + 1 r dr dθ = r3 4r2 + 1 dθ dr
0 0 0 0
Z 2 p
= 2π r3 4r2 + 1 dr by Fubini’s theorem
0

Y
1
Now, let 2r = tan u, i.e r = 2 tan u so that
1
dr = sec2 u du
2
The previous integral transform to

EM
Z tan−1 4
1 p 1
2π tan3 u tan2 u + 1 · sec2 u du
0 8 2
Z tan−1 4
π
= tan3 u sec3 u du
8 0
−1
π tan 4
Z
= tan2 u sec 2u · (sec u tan u du)
8 0
−1
π tan 4
Z
=AD (sec2 u − 1) sec2 u sec u tan u du
8 0
√
Now let w = sec u so dw = sec u tan u du, Hence, when u = 0,w = 1 and whenu = tan−1 4, w = 17,
thus, the u-integral becomes
Z √17 √ !
π 2 2 391 17 + 1
(w − 1)w dw = π
8 1 60

Example: Find the surface area of the torus with,

AC


x
 = (a + b cos t) cos s
y = (a + b cos t) sin t, 0 ≤ s, t ≤ 2π ; a > b > 0 Ans: 4π 2 ab

z = b sin tz


12 Surface integrals for parameterized surfaces

Definition: Scalar surface integral
Let X :7→ <3 be a smooth parameterized surface, where D ⊂ <2 is a bounded region. Let f be a
continuous function
RR whose domain include S = X(D). Then the scalar surface integral of F along x
denoted by X f ds is
A

ZZ ZZ
f ds = f (X(s, t)) kTs × Tt k ds dt
D Z ZD
= f (X(s, t)) kN (s, t)k ds dt
BT

D
Note: If f identically 1 on all of X(D), then.
ZZ ZZ ZZZ
f ds = 1 ds = kTs × Tt k ds dt = surface area of X(D), otherwise
X X D
s
∂(x, y) 2 ∂(x, z) 2 ∂(y, z) 2
ZZ ZZ     
f ds = f (x(s, t), y(s, t), z(s, t)) + + ds dt
X D ∂(s, t) ∂(s, t) ∂(s, t)
Now,
" 2  2  2 # 21
∂(x, y) ∂(x, z) ∂(y, z)
kN (s, t)k = + +
∂(s, t) ∂(s, t) ∂(s, t)

36
Hence, the surface area is
ZZ ZZ
kTs × Tt k ds dt = kN (s, t)k
D D
s
∂(x, y) 2 ∂(x, z) 2 ∂(y, z) 2
ZZ     
= + + ds dt
D ∂(s, t) ∂(s, t) ∂(s, t)

Note:

Y
Ts : tangent vector to the s-coordinate curve X(s0 , t0 ) at (s0 , t0 )
Tt : tangent vector to the t-coordinate curve X(s0 ; t0 ) at (s0 , t0 )
N = Ts × Tt : standard normal vector arising from the parameterization X

EM
∂X
Ts (s0 , t0 ) = (s0 , t0 )
∂s 
∂x ∂y ∂z
= i+ j+ k
∂s ∂s ∂s (s,t)=(s0 ,t0 )

∂X
Tt (s0 , t0 ) = (s0 , t0 )
∂t 
∂x ∂y ∂z
= i+ j+ k
∂t ∂t ∂t (s,t)=(s0 ,t0 )

12.1 Jacobian
AD ∂(f1 , f2 , · · · , fr )
=
∂f1
∂x1
..
···
..
.
∂f1
∂xn
..
∂(x1 , x2 , · · · , xn ) . .
∂fr ∂fr
∂xn ··· ∂xn

Example:
Calculate the surface area of a sphere of radius a described by
AC

X(s, t) = (a cos s sin t, a sin s sin t, a cos t), 0 ≤ s ≤ 2π, 0≤t≤π

Ans: 4πa2
Example: Evaluate
ZZZ
(x2 ey + xyz) dv where B = [−2, 3] × [0, 1] × [05]
B

175 125
Ans: 3 (e − 1) + 8
A
BT

37
13 Green’s Theorem
Let D be a closed, bounded region in <2 Whose boundaryC = ∂D consists of finitely many simple
closed curves.Orient the curves of C so that D is on the left as one traverses C.
Let F (x, y) = M (x, y)i + N (x, y)j be a vector field of class C 1 throughout D. Then
I ZZ  
∂N ∂M
M dx + N dy = − dx dy
c D ∂x ∂y

Y
Example:
Let F = xyi + y 2 j and let D be the first quadrant region bounded by the line y = x and the parabola
y = x2 . We verify Green’s Theorem in this case.

EM
The region D and its boundary are shown below.
∂D is oriented counterclockwise, the orientation
stipulated by Green’s theorem. To calculate
I I
F · ds = xy dx + y 2 dy
∂D ∂D

we need to parameterize the two C1 piece of ∂D

separately
AD
Figure 16: Example
( (
x =t x =1−t
C1 : and C2 :
y = t2 0≤t≤1 y =1−t 0≤t≤1
AC

Hence I Z Z
2
xy dx + y dy = (xy dx + y dy) + (xy dx + y 2 dy)
∂D c1 c2
Z1 Z 1
2 4
= (t · t + t · 2t) dt + {(1 − t)2 + (1 − t)2 }(−dt)
0 0
1
=−
12
on the other hand
ZZ   Z 1Z x Z 1
∂ 2 ∂ 1
(y ) + (xy) dx dy = −x dx dy = −x(x − x2 ) dx = −
D ∂x ∂y 0 x2 0 12
A

The limit interval and the double integral agree.

Note: If D is any region to which Green’s theorem can be applied, then orienting ∂D appropriately,
we have I ZZ
1 1
BT

−y dx + x dy = 2 dx dy = area of D
2 2 D

38
 Example:
Use Green’s theorem to compute the area of the ellipse

x2 y 2
+ 2 =1
a2 b
Solution The ellipse itself may be parameterized counterclockwise by.
(

Y
x = a cos t
y = b sin t 0 ≤ t ≤ 2π
I
1
Now, required area = −y dx + x dy
2 ∂D

EM
1 2π
Z
= −b sin t(−a sin t dt) + a cos t(b cos t dt)
2 0
1 2π
Z
= (ab sin2 t + ab cos2 t) dt
2 0
1 2π
Z
= ab dt
2 0
= πab
Exercise: AD
1. Use Green’s theorem to calculate the line integral
I
y 2 dx + x2 dy
c

where C is the path formed by the square with vertices (0,0),(1,0),(0,1) and (1,1) oriented
counterclockwise

2. Use Green’s theorem to show that the area of a circle with radius r is πr
AC

3. Use Green’s theorem to

nd the area enclosed by the hypocycloid

X(t) = (a cos3 t + a sin3 t), 0 ≤ t ≤ 2π

A
BT

39
 14 Stoke’s Theorem
Let S be a bounded piece-wise smooth, oriented surface in <3 . Suppose that ∂D consist of
finitely many piece-wise C 1 simple, closed curves each of which oriented consistently with S. Let
F be a vector field of class C 1 whose domain includes S. Then
ZZ I Z
∇ × F · ds = f · ds = (∇ × F ) · n̂ ds where n̂
s ∂s

Y
Example:
Let S by the paraboloid z = 9 − x2 − y 2 defined over the disk in the xy-plane of radius 3 (i.e S
is defined for z ≥ 0 only). Then ∂Sconsist of the circle
n
C : (x, y, z)|x2 + y 2 = 9 , z=0

EM
Orient S with the upward-pointing unit normal vector n. we verify Stoke’s theorem for vector
field
F = (2z − y)i + (x + z)j + (3x − 2y)k
We calculate
i j k

∂ ∂ ∂
∇×F = ∂x ∂y ∂z = −3i − j + 2k

AD 2z − y x + z 3x − 2y
An upward-pointing normal vector N is given by N = ∇S
∇S
Note: n = |∇S|
N = 2xi + 2yj + k
(This vector may, of course be normalized to give an orientation normal n) Therefore with
n
D = (x, y)|x2 + y 2 ≤ 3
ZZ ZZ
∇ × F · ds = (−3i − j + 2k) · (2xi + 2yj + k) dx dy
s Z ZD
AC

= (−6x − 2y + 2) dx dy
Z ZD ZZ ZZ
= −6x dx dy − 2y dx dy + 2 dx dy
D D D
By the symmetry of D and the fact that −6x and 2y are odd functions, we have that the first
two double integrals are zero. The last double integral gives twice the area of D . Thus,
ZZ
∇ × F · ds = 2 · π 32 = 18π


3
On the other hand, we may parameterize the boundary of S as

x = 3 cos t
A


y = 3 sin t, ≤ t ≤ 2π This parameterization yields the orientation desired for ∂S

z =0


Then
BT

I Z 2π
F · ds = F (X(t)) · X 0 (t) dt
∂S 0
Z 2π
= (0 − 3 sin t, 3 cos t + 0, 9 cos t − 6 sin t) · (−3 sin t, 3 cos t, 0) dt
0
Z 2π
= (9 sin2 t = 9 cos2 t) dt
0
Z 2π
= 9 dt
0
= 18π

40