Download as pdf or txt
Download as pdf or txt
You are on page 1of 105

lOMoARcPSD|39572262

Aplied maths 2 - applied

Applied maths (University of Gondar)

Scan to open on Studocu

Studocu is not sponsored or endorsed by any college or university


Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)
lOMoARcPSD|39572262

Ambo University
Department of Mathematics

Lecture Note

Applied Mathematics II

March 24, 2016

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

Contents

1 Sequence and Series 5

1.1 Definition and types of sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

1.1.1 Types of sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

1.2 Convergence properties of sequences . . . . . . . . . . . . . . . . . . . . . . . . . 8

1.2.1 Monotonic and Bounded Sequence . . . . . . . . . . . . . . . . . . . . . 14

1.3 Subsequence and limit points . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

1.4 Definition of infinite series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

1.5 Convergence and divergence, properties of convergent series . . . . . . . . . . . . 24

1.6 Nonnegative term series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

1.7 Tests of convergence (integral, comparison, ratio and root tests) . . . . . . . . . 30

1.7.1 The Integral Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

1.7.2 Comparison Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

1.7.3 Limit Comparison Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

1.8 Alternating series and alternating series test . . . . . . . . . . . . . . . . . . . . 37

1.8.1 Alternating Series Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

1.9 Absolute and conditional convergence . . . . . . . . . . . . . . . . . . . . . . . 40

1.9.1 Ratio Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

1.9.2 Root Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

1.10 Generalized convergence tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

2 Power Series 48

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

2.2 Definition of power series at any x0 and x0 = 0 . . . . . . . . . . . . . . . . . . 49

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

CONTENTS 2

2.3 Convergence and divergence, radius and interval of convergence . . . . . . . . . 51

2.4 Representations of Functions as Power Series . . . . . . . . . . . . . . . . . . . . 55

2.5 Algebraic operations on convergent power series . . . . . . . . . . . . . . . . . . 57

2.6 Differentiation and integration of power series . . . . . . . . . . . . . . . . . . . 58

2.7 Taylor series; Taylor polynomial and application . . . . . . . . . . . . . . . . . . 60

2.7.1 Polynomial Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . 61

2.7.2 Defined by Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

2.7.3 Indeterminate Forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

3 Differential Calculus of Function of Several Variables 67

3.1 Notations, examples, level curves and graphs . . . . . . . . . . . . . . . . . . . 67

3.1.1 Notations and Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

3.1.2 Level curves and graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

3.2 Limits and continuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

3.2.1 Limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

3.2.2 Continuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

3.3 Partial derivatives; tangent lines, higher order partial derivatives. . . . . . . . . 76

3.3.1 A geometric interpretation of partial derivatives . . . . . . . . . . . . . . 78

3.3.2 Higher order partial derivatives . . . . . . . . . . . . . . . . . . . . . . . 78

3.4 The chain rule, implicit differentiation . . . . . . . . . . . . . . . . . . . . . . . 80

3.4.1 Chain rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

3.4.2 Implicit differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

3.4.3 Differentiability of functions of several variables . . . . . . . . . . . . . . 83

3.5 Directional derivatives and gradients . . . . . . . . . . . . . . . . . . . . . . . . 85

3.5.1 The Gradient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

3.6 Tangent planes and Total differential . . . . . . . . . . . . . . . . . . . . . . . 88

3.6.1 Tangent Planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

3.7 Tangent plane approximations of values of a function . . . . . . . . . . . . . . . 93

3.8 Relative extrema of functions of two variables . . . . . . . . . . . . . . . . . . . 94

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

CONTENTS 3

3.9 Largest and smallest values of a function on a given set . . . . . . . . . . . . . 98

3.10 Extreme values under constraint conditions: Lagrange’s multiplier . . . . . . . 99

4 Multiple Integrals 103

4.1 Double integrals and their evaluation by iterated integrals . . . . . . . . . . . . 103

4.2 Iterated integrated . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

4.3 Change of variable in double integrals . . . . . . . . . . . . . . . . . . . . . . . . 110

4.4 Double Integrals in Polar coordinates . . . . . . . . . . . . . . . . . . . . . . . . 112

4.5 Application of Double Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

4.5.1 Surface Area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

4.5.2 Triple Integrals in Cartesian coordinates . . . . . . . . . . . . . . . . . . 117

4.6 Triple integrals in cylindrical and spherical coordinates . . . . . . . . . . . . . . 120

4.6.1 Triple Integrals in Cylindrical Coordinates . . . . . . . . . . . . . . . . . 120

4.6.2 Triple Integrals in Spherical Coordinates . . . . . . . . . . . . . . . . . . 123

4.7 Application: Volume, center of mass of solid region . . . . . . . . . . . . . . . . 127

4.7.1 Some of the applications of multiple integrals . . . . . . . . . . . . . . . 128

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

Chapter 1

Sequence and Series

1.1 Definition and types of sequence

A sequence is a list of numbers


a1 , a2 , a3 , . . . , an , . . .
in a given order. A formal definition of sequence can be given as follows:
Definition 1.1.1. A sequence {an } is a function whose domain is the set of integers. The
functional values a1 , a2 , a3 , . . . , an , . . . are the terms of the sequence, and the term an is called
the nth term of the sequence.
Note 1.1.1. • The sequence {a1 , a2 , a3 , . . . , an , . . . } is also denoted by {an }, {an }∞
n=1 .

• Sometimes it is convenient to begin a sequence with ak . In this case the sequence is


{an }∞
n=k , and its terms are ak , ak+1 , ak+2 , . . . , an , . . . .

Definition 1.1.2. A sequence is said to be a constant sequence if Sn = k(constant) for every


n ∈ N.
Example 1.1.1. Let Sn = 2 for every n, is a constant sequence. That is, Sn = {2, 2, 2, 2, . . . }
Definition 1.1.3. A Recursively-defined sequence is a sequence where the first term(s)
are given and the next term is given in terms of the previous terms.
Example 1.1.2. List the first five terms of the sequence.
 ∞  
n 1 2 3 4 5
A. n+1
= , , , , ,...
2 3 4 5 6
n=1
 ∞  √ 
3 1
B. cos( nπ
6
) = 1, 2 , 2 , 0, . . .
n=0
∞
√ √ √
  
C. n−3 = 0, 1, 2, 3, . . .
n=3
Exercise 1.1.1. List the first five terms of the sequence

A. {(−1)n n+1
3n
}

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.1 Definition and types of sequence 5

B. { n cos(nπ)
2n−1
}
n
C. { (−π)
5n
}
D. a1 = 2, a2 = 4, an+1 = 2an − an−1 for n ≥ 2.

Graphing a sequence
To graph the sequence {an } we plot the points (n, an ) as n ranges over all possible values on a
graph.
Example 1.1.3. Graph the sequence { n+1 }∞ . The first few points on the graph are then,
n2 n=1
3 4 5
(1, 2), (2, ), (3, ), (4, ), . . . .
4 9 16
The graph is then,

Now let’s look at some special sequences, and their rules.

1.1.1 Types of sequences

Finite and infinite sequence


If the sequence goes on forever it is called an infinite sequence, otherwise it is a finite sequence.
Example 1.1.4. Some examples of finite and infinite sequence.

a) {1, 2, 3, 4, . . . } is a very simple sequence (and it is an infinite sequence)


b) {1, 3, 5, 7} is the sequence of the first 4 odd numbers (and is a finite sequence)

Arithmetic Sequence
In an arithmetic sequence the difference between one term and the next is a constant. In other
words, you just add some value each time . . . on to infinity. In general you could write an
arithmetic sequence like this:
{a, a + d, a + 2d, a + 3d, . . . }
Where:

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.2 Convergence properties of sequences 6

• a is the first term, and

• d is the difference between the terms (Usually called the ”common difference”)

And you can make the rule by:

xn = a + (n − 1)d

Example 1.1.5. Consider the sequence 1, 4, 7, 10, 13, 16, 19, 22, 25,. . .
This sequence has a difference of 3 between each number. Then its rule is xn = 3n − 2.

Geometric sequence
In a geometric sequence each term is found by multiplying the previous term by a constant.

Example 1.1.6. Consider the sequence 2, 4, 8, 16, 32, 64, 128, 256,. . .
This sequence has a factor of 2 between each number. Its Rule is xn = 2n .

In General you could write a geometric sequence like this:

{a, ar, ar2 , ar3 , . . . }.

Where:

• a is the first term, and

• r is the factor between the terms (called the ”common ratio”)

And the rule is: xn = ar(n−1) .

Note 1.1.2. r should not be 0. Because when r = 0 , you get the sequence {a, 0, 0, . . . } which
is not geometric.

1.2 Convergence properties of sequences

Let the sequence an = { n12 }, we see that if n is taken to be a large positive integer, then the
nth term of the sequence is small. That is as n becomes very large, { n12 } becomes very small
or very close to zero. This sequence behaves nicely in the sense that he terms approach some
fixed real number( in this case zero) as n becomes large. In this case we say that the sequence
converges to zero. With this introduction, we define convergence of a sequence.

Definition 1.2.1. A sequence {an } is said to converge to the real number A if and only if for
each ε > 0 there exists a positive real number N = N (ε) > 0 such that |an − A| < ε for all
n > N. The sequence {an } is said to be convergent if and only if there is a real number A such
that {an } converges to A. The sequence {an } is said to be divergent or to diverge if and only if
it is not convergent.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.2 Convergence properties of sequences 7

Example 1.2.1. Show that the sequence { n12 }∞n=1 converges to 0.


Solution: Here, we must show that we can make n12 as close to 0 as desired, just by making
n sufficiently large. So, given any ε > 0, we must find N sufficiently large so that for every
n > N,
1 1
2
− 0 < ε, or 2 < ε (1.1)
n n
Since n2 and ε are positive, we can divide both sides of (1.1) by ε and multiply by n2 , to obtain
1
< n2 .
ε
Taking square roots gives us r
1
< n.
ε
q
1
Working backwards now, observe that if we choose N to be an integer with N = ε
, then
1
n > N implies that n2
< ε, as desired.

So just how do we find the limits of sequences? We will use the following theorems and
facts. The following theorem is basically telling us that we take the limits of sequences much
like we take the limit of functions. In fact, in most cases well not even really use this theorem
by explicitly writing down a function. We will more often just treat the limit as if it were a
limit of a function.

Theorem 1.2.1. Given the sequence {an } if we have a function f (x) such that f (n) = an and
lim f (x) = L, then lim an = L.
x→∞ n→∞

In this section we continue our analysis of the convergence and divergence of sequences. Since
sequences are functions, we may add, subtract, multiply, and divide sequences just as we do for
functions on Applied Mathematics I. Rules for computing the limits of combination of sequences
are analogous to the rules for limits of combinations of functions. We present these rules now.

Theorem 1.2.2. Suppose lim an = L and lim bn = M and that c is a constant. Then
n→∞ n→∞

1. lim can = cL
n→∞

2. lim (an ± bn ) = L ± M
n→∞

3. lim an bn = LM
n→∞

4. lim ( abnn ) = L
M
, provided that bn 6= 0 and and M 6= 0.
n→∞

5. lim (an )p = Lp , if p > 0 and an > 0.


n→∞
lim an
6. lim ean = en→∞
n→∞

Definition 1.2.2. Let {an }∞ ∞


n=m be a sequence. A number L is the limit of {an }n=m if for every
ε > 0 there is an integer N such that if n > N , then |an − L| < ε.

Theorem 1.2.3. If a sequence of real numbers {an }∞


n=m has a limit, then this limit is unique.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.2 Convergence properties of sequences 8

 
n
Example 1.2.2. Find the limit of the sequence en
.

Solution. The expression enn is an indeterminate form of type ∞/∞ as n → ∞, so L’Hopital’s


rule is called for. However, we cannot apply this rule directly to enn because the functions n
and en are only defined at the positive integers, and hence are not differentiable functions. To
circumvent this problem, we will replace n by x, and apply L’Hopital’s rule to the function exx .
This yields
x 1
lim x = lim x = 0
x→∞ e x→∞ e
using Theorem 1.2.1 we can conclude that
n
lim =0
n→∞ en

Example 1.2.3. Determine if the following sequences converge or diverge. If the sequence
converges determine its limit.
( )∞
3n2 −1
A. 10n+5n2
n=1

Solution. In this case all we need to do is recall the method that was developed in Applied
Mathematics I to deal with the limits of rational functions.[Please if you do not remember
go to Applied Mathematics I and read about Limit] To do a limit in this form all we need
to do is factor from the numerator and denominator the largest power of n, cancel and
then take the limit.

3n2 − 1 n2 (3 − n12 ) 3 − n12 lim 3 − n12 3−0 3


n→∞
lim 2
= lim 2 10 = lim 10 = 10 = = .
n→∞ 10n + 5n n→∞ n ( + 5) n→∞ n + 5 lim n + 5 0+5 5
n n→∞

So the sequence converges and its limit is 35 . ◭


( )∞
e2n
B. n
n=1

Solution. We will need to be careful with this one. We will need to use L’Hospital’s Rule
on this sequence. The problem is that L’Hospital’s Rule only works on functions and not
on sequences. Normally this would be a problem, but we’ve got Theorem 1.2.1 from above
2x 2n
to help us out. Let’s define f (x) = ex and note that, f (n) = en . Theorem 1.2.1 says that
all we need to do is take the limit of the function.
e2n e2x 2e2x
lim = lim f (x) = lim = lim =∞
n→∞ n x→∞ x→∞ x x→∞ 1

So, the sequence in this part diverges (to ∞ ). ◭

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.2 Convergence properties of sequences 9

C. {(−1)n }

Solution. For this sequence all that we need to do is acknowledge that lim (−1)n doesn’t
n→∞
exist to get that the sequence is divergent. If youre not convinced that this limit doesnt
exist write down the first few terms of the sequence.
 ∞
n
(−1)
n=0
In order for a limit to exist the terms must be settling down towards a specific value and
these clearly will never do that. ◭
Theorem 1.2.4 (Squeezing Theorem). If there exists some integer N such that an ≤ bn ≤ cn
for all n ≥ N and
lim an = lim cn = L, then lim bn = L.
n→∞ n→∞ n→∞
.
( )∞
cos
√n
Example 1.2.4. Find the limit of the sequences n
.
n=1

Solution. From the well known property of Cosine, we recall −1 ≤ cos(n) ≤ 1, ∀n.
−1 cos n 1
⇒ √ ≤ √ ≤ √ as n ≥ 1.
n n n
−1
Since ⇒ √ n
→ 0 and ⇒ √1n → 0 as n → 0, by Squeezing theorem cos
√n
n
→ 0.
Therefore, lim cos
√ n = 0.
n

n→∞

Definition 1.2.3. If the sequence {an } diverges but does not diverge to +∞ or does not diverge
to −∞, then the sequence {an } is said to oscillate or to be an oscillating sequence.
Example 1.2.5. {(−1)n } and {sin(nπ/4)} are oscillating sequences.

The next theorem is convenient for sequences that alternate in signs and note that it will
only work if the sequence has a limit of zero.
Theorem 1.2.5. If lim |an | = 0, then lim an = 0.
n→∞ n→∞
n
Example 1.2.6. Determine if the sequence { (−1)
n
}∞
n=1 converge or diverge. If the sequence
converges determine its limit.

Solution. We will need to use Theorem 1.2.5 on this problem.


(−1)n 1
lim = lim = 0
n→∞ n n→∞ n
Therefore, since the limit of the sequence terms with absolute value bars on them goes to zero
we know by Theorem 1.2.5 that,
(−1)n
lim =0
n→∞ n
Which also means that the sequence converges to a value of zero. ◭

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.2 Convergence properties of sequences 10

Example 1.2.7. Show that if −1 < r < 1 then lim rn = 0.


n→∞

1 1
Solution. If r = 0, the result is trivial, so suppose otherwise. Then |r|
> 1, and so |r|
=1+p
for some p > 0. By Binomial Formula
1
n
= (1 + p)n = 1 + pn + positive terms ≥ pn
|r|
Thus,
1
0 ≤ |r|n ≤
pn
1 1
Since lim = lim 1 = 0, it follows from the Squeezing Theorem that lim |r|n = 0, or
n→∞ pn p n→∞ n n→∞
equivalently, lim |rn | = 0. By Theorem 1.2.5, lim rn = 0. ◭
n→∞ n→∞

Theorem 1.2.6. If limn→∞ an = L and the function f is continuous at L , then

lim f (an ) = f ( lim an ) = f (L).


n→∞ n→∞
( )
1
Example 1.2.8. Show using the logarithmic function that a n , for any a > 0 converges to
1.

1
Solution. Let xn = a n . We have
1
⇒ ln(xn ) = ln a
n
⇒ lim ln(xn ) = 0, using Theorem 1.2.6
n→∞

⇒ lim eln(xn ) = e0 = 1.
n→∞

1
Hence, lim a n = 1. ◭

Exercise 1.2.1. 1. Find the limit of the given sequence and prove that the sequence converges
to that limit.

A. {1/2n } C. {1 − (1/n)} E. {1 + ((−1)n / n)}
B. {1/n} D. {(−1)n /n2 } F. {n/(1 + n2 )}
 n 
a
2. Show that 1+ n converges to ea , for any number a.
 ∞
sin
√n
3. Find the limit of the sequence n
.
n=1
 ∞
sin2 n
4. Find the limit of the sequence n
.
n=1
 ∞
ln n
5. Show that the sequence n
converges to 0.
n=1
 ∞
2n3 +3
6. Determine the limit of 3n3 −4
.
n=1

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.2 Convergence properties of sequences 11

1.2.1 Monotonic and Bounded Sequence

We now need to take a look at some more terminology and definitions for sequences.

Definition 1.2.4. Given any sequence {an } we have the following.

1. We call the sequence {an } is increasing if an ≤ an+1 for every n.

2. We call the sequence {an } is decreasing if an ≥ an+1 for every n.

3. We call the sequence {an } is strictly increasing if an < an+1 for every n.

4. We call the sequence {an } is strictly decreasing if an > an+1 for every n.

5. If {an } is an increasing sequence or {an } is a decreasing sequence we call it is monotonic


sequence.

Testing for Monotonicity:


In order for a sequence to be strictly increasing all pairs of successive terms an and an+1 , must
satisfy an+1 − an > 0. More generally, a monotone sequence can be classified as follows

Difference between successive terms Classifications


an+1 − an > 0 Strictly increasing
an+1 − an < 0 strictly decreasing
an+1 − an ≥ 0 Increasing
an+1 − an ≤ 0 decreasing

Example 1.2.9. Show that the sequence defined by {an } = { n!1 } is monotonic strictly decreas-
ing.

Solution.
1 1
an+1 − an = −
(n + 1)! n!
1 − (n + 1)
=
(n + 1)!
n
=− < 0, thus an+1 < an , ∀n ≥ 1.
(n + 1)!
Therefore, {an } is decreasing sequence. ◭

Exercise 1.2.2. Use an+1 − an to show that the given sequence {an } is strictly increasing or
strictly decreasing.
 ∞  ∞  ∞
1 n n
A. n
B. 2n+1
C. 4n−1
n=1 n=1 n=1
 ∞  ∞  ∞
1
D. 1 − n E. n − 2n F. n − n 2

n=1 n=1 n=1

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.2 Convergence properties of sequences 12

If an and an+1 are successive terms in a strictly increasing sequence, then an < an+1 . If the
terms in the sequence are all positive, then we can divide both sides of the inequality by an to
obtain 1 < an+1
an
. More generally, monotone sequences with positive terms can be classified as
follows:

Ratio of successive terms Conclusion


an+1 /an > 1 Strictly increasing
an+1 /an < 1 strictly decreasing
an+1 /an ≥ 1 Increasing
an+1 /an ≤ 1 decreasing
 
n
Example 1.2.10. Show that the sequence {an } = n+1 is strictly increasing by by examining
the ratio of the successive terms.
n n+1
an = , and an+1 =
n+1 n+2
Thus,
an+1 (n + 1)/(n + 2) n+1 n+1 n2 + 2n + 1
= = . = (1.2)
an (n)/(n + 1) n+2 n n2 + 2n
Since the numerator in(1.2)exceeds the denominator, it follows that an+1 /an > 1 for x ≥ 1.
This proves that the sequence is strictly increasing.

Exercise 1.2.3. Use an+1 /an to show that the given sequence {an } is strictly increasing or
strictly decreasing.
 ∞  ∞  ∞
n 2n −1
A. 2n+1
B. 2n +1
C. ne
n=1 n=1 n=1
 ∞  ∞  ∞
10n nn
D. 2n
! E. n!
F. n − n2
n=1 n=1 n=1

If f (n) = an is the nth term of a sequence, and if f is differentiable for x ≥ 1, then we have
the following results:

Derivative of f for x ≥ 1
Conclusion for the sequence an = f (n)
f ′ (x) > 0 Strictly increasing
f ′ (x) < 0 strictly decreasing
f ′ (x) ≥ 0 Increasing
f ′ (x) ≤ 0 decreasing
 
n
Example 1.2.11. Consider the sequence an = n+1 . Let

x
f (x) =
x+1

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.2 Convergence properties of sequences 13

so that the nth term in the given sequence is an = f (n). The function f is increasing for x ≥ 1
since
(x + 1)(1) − x(1) 1
f ′ (x) = 2
= > 0.
(x + 1) (x + 1)2
Thus,
an = f (n) < f (n + 1) = an+1
which proves that the given sequence is Strictly increasing.
Exercise 1.2.4. Use differentiation to show that the sequence is strictly increasing or strictly
decreasing.
 ∞  ∞  ∞
n
A. 2n+1
B. 3 − n1 C. 1
n+ln n
n=1 n=1 n=1
 ∞  ∞  ∞
ln(n+2)
D. ne−2n E. n+2
−1
F. tan n
n=1 n=1 n=1

We now develop a method for determining the convergence of a sequence without direct ap-
plication of the definition of convergence. We begin by defining some special types of sequences.
Definition 1.2.5. Given any sequence {an } we have the following.

1. If there exists a number m such that m ≤ an for every n we say the sequence is bounded
below. The number m is sometimes called a lower bound for the sequence.
2. If there exists a number M such that m ≥ an for every n we say the sequence is bounded
above. The number M is sometimes called an upper bound for the sequence.
3. If the sequence is both bounded below and bounded above we call the sequence is bounded.
Definition 1.2.6. If A is a lower bound of a sequence {an } and if A has the property that
for every lower bound C of {an }, C ≤ A, then A is called the greatest lower bound of the
sequence. Similarly, if B is an upper bound of a sequence {an } and if B has the property that
for every upper bound D of {an }, B ≤ D , then B is called the least upper bound of the
sequence.

The Axiom of Completeness


Every nonempty set of real numbers which has a lower bound has a great- est lower bound.
Also, every set of real numbers which has an upper bound has a least upper bound.

Theorem 1.2.7. Every bounded monotonic sequence is convergent

Proof. We prove the theorem for the case when the monotonic sequence is increasing. Let the
sequence be {an }. Because {an } is bounded, there is an upper bound for the sequence. By the
axiom of completeness , {an } has a least upper bound, which we call B. Then if ǫ is a positive
number, B − ǫ cannot be an upper bound of the sequence because B − ǫ < B and B is the least
upper bound of the sequence. So for some positive integer N

B − ǫ < aN (1.3)

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.2 Convergence properties of sequences 14

Because B is the least upper bound of {an }, by Definition 1.2.5 it follows that

an ≤ B, for every positive integer n. (1.4)

Because {an } is an increasing sequence, we have from Definition 1.2.4(1)

an ≤ an+1 , for every positive integer n


and so
aN ≤ an , whenever n ≥ N (1.5)
From 1.3, 1.4, and 1.5 it follows that

B − ǫ < aN ≤ an ≤ B < B + ǫ whenever n ≥ N

from which we get


B − ǫ < an < B + ǫ, whenever n ≥ N
which can be written as
|an − B| < ǫ whenever n ≥ N.
This implies that lim an = B. Therefore, the sequence {an } is convergent.
n→∞
2n
Example 1.2.12. Investigate the convergence of the sequence n!
.

2n
Solution. First, note that we do not know how to compute lim . This has the indeterminate
n→∞ n!
form ∞/∞, but we cannot use L’Hopital’s Rule here directly or indirectly. (Why not?) The
n
graph of 2n! suggests that the sequence converges to 0. To confirm this suspicion, we first show
that the sequence is monotonic. We have
2n+1
an+1 (n+1)! 2n+1 n!
= 2n = · n
an n!
(n + 1)! 2
n
2(2 )n! 2
= = ≤ 1, for n ≥ 1 (1.6)
(n + 1)n!2n n+1
We can make a slightly stronger statement, though. Since we have established that the sequence
Multiplying both sides of (1.6) by an > 0
gives us an+1 ≤ an for all n and so, the 2n
n n!
sequence is decreasing. Next, since the se- 2 2
quence is decreasing, we have that 4 0.666667
2n 21 6 0.088889
|an | = ≤ = 2, 8 0.006349
n! 1!
10 0.000282
for n ≥ 1 (i.e., the sequence is bounded by
12 0.0000086
2). Since the sequence is both bounded and
monotonic, it must be convergent, by The- 14 1.88 × 10−7
orem 1.2.7. We display a number of terms 16 3.13 × 10−9
of the sequence in the accompanying table, 18 4.09 × 10−11
from which it appears that the sequence is 20 4.31 × 10−13
converging to approximately 0.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.3 Subsequence and limit points 15

is decreasing and convergent, we have from our computations that

0 ≤ an ≤ a20 ≈ 4.31 × 10−13 , for n ≥ 20.

Further, the limit L must also satisfy the inequality

0 ≤ L ≤ 4.31 × 10−13 .

We can confirm that the limit is 0, as follows. From (1.6),


2
L = lim an+1 = lim an ,
n→∞ n→∞ n+1
So that
2
L = lim lim an = 0 · L = 0.
n→∞ n + 1 n→∞

Theorem 1.2.8. Every convergent monotonic sequence is bounded.

Exercise 1.2.5. Determine if the following sequences are monotonic and/or bounded.
 ∞  ∞
2 n
A. −n D. n+1
n=1 n=1
 ∞  ∞
(−1)n
B. (−1)n+1 E. n
n=1 n=1
 ∞  ∞
2
C. n2
F. sin( nπ
4
)
n=1 n=1

1.3 Subsequence and limit points

We now introduce the concept of a subsequence of a sequence.

Definition 1.3.1. Let {an } be a sequence , and let {nk } be a sequence of positive integers such
that nk < nk+1 for each k, that is, {nk } = {nk }∞
k=1 is a strictly increasing sequence. Then the

sequence {ank } = {ank }k=1 is called a subsequence of the sequence {ank }.

Let’s have an example of a subsequence of a sequence.

Example 1.3.1. Consider the sequence { n1 }. If we let nk = 2k for each positive integer k, the
corresponding subsequence of { n1 } is { 2k
1 1
} = { 2n }. Furthermore, if we let {nk } be any strictly
increasing sequence of positive integers, then the sequence { n1k } is a subsequence of the sequence
{ n1 }.

Next, lets have some results on convergence related to subsequences.

Note 1.3.1. If {an } is a sequence, then {an } is trivially a subsequence of itself.

Theorem 1.3.1. If the sequence {an } converges to A, then every subsequence of the sequence
{an } also converges to A.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.3 Subsequence and limit points 16

Corollary 1.3.1. If the sequence {an } has two subsequences which converge to different limits,
then the sequence {an } is divergent.

Example 1.3.2. Consider the sequence {an } defined by the equation an = (−1)n for each
positive integer n. Then the subsequence a2n is defined by a2n = 1 for each positive integer n,
and the subsequence a2n−1 is defined by a2n−1 = −1 for each positive integer n. then, we have
a2n → 1 and a2n−1 → −1. Hence Corollary 1.3.1 implies that the sequence {(−1)n } diverges.

Corollary 1.3.2. If the sequence {an } has a subsequence which is divergent, then the sequence
{an } is divergent.

Example 1.3.3. Let an = sin( nπ 4


) for each positive integer n, and let nk = 4k − 2 for each
positive integer k. Corollary 1.3.2 implies that the sequence {an } = sin( nπ
4
) is divergent since
this sequence has the divergent subsequence

ank = {sin[(4k − 2)π/4]} = {(−1)k }.

Definition 1.3.2. Let {an } be a sequence. The number a is called a limit point of {an } if
and only if there is a subsequence of {an } converging to a.

Note that: limit point of a sequence and limit of a sequence are not the same. The following
theorem points out the relationship between the limit points of a sequence and the convergence
of the sequence.

Theorem 1.3.2. Let {an } be a sequence. The point a is a limit point of {an } if and only if for
each pair of positive numbers ε and N there is a k > N such that

|ak − a| < ε.

Corollary 1.3.3. Let {an } be a sequence. The point a is a limit point of {an } if and only if
for each ε > 0
|an − a| < ε
for infinitely many values of n.
 
n n+1
Example 1.3.4. Consider the sequence (−1) n . -1 and 1 are the limit points of the
 
n n+1
sequence (−1) n .

Exercise 1.3.1. Find the limit points of the given sequences.

A. 1, 3, −1, 1, 3, −1, . . .

B. {cos(nπ/2)}

C. {(1 − 1/n)}

D. {(1 − 1/n)}{sin(nπ/2)}

E. {n}

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.4 Definition of infinite series 17

1.4 Definition of infinite series

Definition 1.4.1. Given a sequence of numbers{an }, an expression of the form

a1 + a2 + a3 + · · · + an + · · ·

is an infinite series. The number an is the nth term of the series. The sequence {sn } defined by

s1 = a1
s2 = a1 + a2
s3 = a1 + a2 + a3
..
.
n
X
sn = a1 + a2 + a3 + · · · an = ak
k=1
..
.

is the sequence of partial sums of the series, the number sn being the nth partial sum.

Example 1.4.1. Find the first four elements of the sequence of partial sums sn and find a
formula for sn in terms of n.

Solution.

X
A. n
n=1

s1 =1, s2 = 1 + 2 = 3, s3 = 1 + 2 + 3 = 6, s4 = 1 + 2 + 3 + 4 = 10
1
sn = n(n + 1)
2


X 1
B.
n=1
(2n − 1)(2n + 1)
 
1 1 1
s1 = 1− =
2 3 3
   
1 1 1 1 1 1 2
s2 = 1− + − = (1 − ) =
2 3 3 5 2 5 5
     
1 1 1 1 1 1 1 1 3
s3 = 1− + − + − = (1 − ) =
2 3 3 5 5 7 2 7 7
       
1 1 1 1 1 1 1 1 1 1 4
s4 = 1− + − + − + − = (1 − ) =
2 3 3 5 5 7 7 9 2 9 9
       
1 1 1 1 1 1 1 1
sn = 1− + − + ··· + − + −
2 3 3 5 2n − 3 2n − 1 2n − 1 2n + 1
1 1 n
= (1 − )=
2 2n + 1 2n + 1

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.5 Convergence and divergence, properties of convergent series 18


X n
C. ln
n=1
n+1
∞ ∞
X n X
ln = [ln n − ln n + 1],
n=1
n + 1 n=1

s1 = ln 1 − ln 2 = − ln 2,
s2 = (ln 1 − ln 2) + (ln 2 − ln 3) = − ln 3
s3 = (ln 1 − ln 2) + (ln 2 − ln 3) + (ln 3 − ln 4) = − ln 4
s4 = (ln 1 − ln 2) + (ln 2 − ln 3) + (ln 3 − ln 4) + (ln 4 − ln 5) = − ln 5
sn = (ln 1 − ln 2) + (ln 2 − ln 3) + · · · + (ln(n − 1) − ln n) + (ln n − ln(n + 1)) = − ln(n + 1)

1.5 Convergence and divergence, properties of convergent series

Definition 1.5.1. Let {sn } be the sequence of partial sums of the series

a1 + a2 + · · · + an + · · ·

If the sequence {sn } converges to a limit S, then the series is said converge to S, and S is
called the sum of the series. We denote this by writing

X
S= an .
n=1

If the sequence of partial sums diverges, then the series is said to diverge. A divergent series
has no sum.

Remark 1.5.1. Sometimes it will be desirable to start the summation index in an infinite series
at n = 0 rather than n = 1, in which case we will view a0 as the zeroth term and s0 = a0 as
the zeroth partial sum. It can be proved that changing the starting value for the index has no
effect on the convergence or divergence of an infinite series.

Example 1.5.1. Determine whether the series converges. If the series converges, find its sum.
∞ ∞
X X 1
A. n B.
n=1 n=1
(2n − 1)(2n + 1)

X n
C. ln
n=1
n+1

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.5 Convergence and divergence, properties of convergent series 19

Exercise 1.5.1. Determine whether the series converges. If the series converges, find its sum.
∞ ∞
X 2n + 1 X 2
A. B.
n=1
n (n + 1)2
2
n=0
5 n−1

∞ ∞
X 2n−1 X 2
C. , D.
n=1
3n n=1
(4n − 3)(4n + 1)

Geometric Series
Geometric series play an important role in mathematical analysis. They also arise frequently
in the field of finance. The convergence or divergence of a geometric series is easily established.
Definition 1.5.2. A series of the form

X
arn−1 = a + ar + ar2 + ar3 + · · · + arn−1 + · · · a 6= 0
n=1

is called a geometric series with common ratio r.

The following theorem tells us the conditions under which a geometric series is convergent.
Theorem 1.5.1. If |r| < 1 , then the geometric series

X
arn−1 = a + ar + ar2 + ar3 + · · · + arn−1 + · · · a 6= 0
n=1

a
arn−1 =
P
converges, and its sum is 1−r
. The series diverges if |r| ≥ 1.
n=1


 n−1
1 1 1 1 1
. Here a = 1 and r = 21 . since
P
Example 1.5.2. 1. 1 + 2
+ 4
+ 8
+ ··· = 2
= 1− 21
n=1
|r| < 1, the series converges.

 n−1
e2 e3 e
. Here a = π and r = − πe
P
2. π − e + π
− π2
+ ··· = π − π
n=1
∞ n−1
π2

X e π
π − =  = .
π e π+e
n=1 1− π

e
The series converges since since − π
< 1.

1 3
∞ √
( 2)n−1 . This series diverges to ∞ since a = 1 > 0 and
P
3. 1 + 2 2 + 2 + 2 2 + · · · =
√ n=1
r = 2 > 1.

(−1)n−1 . This series diverges since r = −1.
P
4. 1 − 1 + 1 − 1 + 1 − · · · =
n=1

5. Letx = 0.323232 · · · = 0.32; then


∞ n−1
322 32 3

32 X 32 1 32 1 32
x= + 2
+ + ··· = = 1 =
100 100 100 n=1
100 100 100 1 − 100 99

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.5 Convergence and divergence, properties of convergent series 20

Telescoping Series
It is now time to look at the second of the three series in this section. In this portion we are
going to look at a series that is called a telescoping series. The name in this case comes from
what happens with the partial sums and is best shown in an example.

1
P
Example 1.5.3. Show that the series n(n+1)
, is convergent, and find its sum.
n=1

Solution. This is not a geometric series, so we go back to the definition of a convergent series
and compute the partial sums.
n
X 1 1 1 1 1
sn = = + + + ··· +
i=1
i(i + 1) 1·2 2·3 3·4 n(n + 1)

We can simplify this expression if we use the partial fraction decomposition


1 1 1
= −
i(i + 1) i i+1
Thus we have
n n
X 1 X 1 1
sn = = −
i=1
i(i + 1) n=1 i i+1
       
1 1 1 1 1 1 1
= 1− + − + − + ··· + −
2 2 3 3 4 n−1 n
1
=1−
n+1
And here,  
1
lim 1 − = 1 − 0 = 1.
n→∞ n+1
Therefore, the given series is convergent and


X 1
= 1.
n=1
n(n + 1)

Harmonic Series
The series ∞
X 1
n=1
n
is called the Harmonic series. The harmonic series is divergent.
If sn is the partial sum of the Harmonic series, then
1 1
sn = 1 + + + · · · = sum of the areas of rectangles shaded in Figure 1.1
2 3
1
> area under y = from x = 1 to x = n + 1
x
n+1
dx
Z
= = ln(n + 1)
x
1

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.5 Convergence and divergence, properties of convergent series 21

Figure 1.1: A partial sum of the partial series

Now lim ln(n + 1). Therefore, lim sn = ∞ and


n→∞ n→∞


X 1 1 1
= 1 + + + ··· diverges to infinity
n=1
n 2 3

Exercise 1.5.2. Determine whether the series converges. If the series converges, find its sum.
∞ ∞  
X 1 X 1 1
A. B. +
n=1
n+2 n=1
2n 2n
∞ ∞
X 3 X π
C. n
, D. tann
n=1
2 n=1
6
E. 0.272727 . . . D. 0.234234234 . . .

The next theorem tells us that the terms of a convergent series must ultimately approach
zero.
P∞
Theorem 1.5.2. If the series an is convergent, then lim an = 0.
n=1 n→∞

Remark 1.5.2. This gives a necessary condition for the convergence of a series; it is not
P1
sufficient. For example we have seen that n
is divergent, even though n1 → 0 as n → ∞.

Theorem 1.5.3 (The Test for Divergence). If lim an does not exist or if lim an 6= 0, then
n→∞ n→∞

P
the series an is divergent.
n=1

Example 1.5.4. Show that the following series are divergent


∞ ∞
X n X
A. B. (−1n )n sin(1/n)
n=2
2n − 1 n=0


n n
P
Solution. A. 2n−1
diverges to infinity since lim = 1/2 > 0
n=2 n→∞ 2n−1

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.6 Nonnegative term series 22


(−1n )n sin(1/n) diverges since
P
B.
n=0

1 sin 1 sin x
lim (−1n )n sin( ) = lim 1 n = lim+ = 1 6= 0.
n→∞ n n→∞
n
x→0 x

The following properties of series are immediate consequences of the corresponding properties
of the limits of sequences.
P∞ P∞ ∞
P
Theorem 1.5.4. Let an and bn be two infinite series, and c is a constant. If an and
n=1 n=1 n=1

P
bn converges with sum a and b respectively, then
n=1


X
A. (an ± bn ) converges to a ± b
n=1
X∞
B. can converges to ca
n=1

X
C. If bn is divergent ,
n=1

X
a) (an ± bn ) is divergent
n=1
X∞
b) can is divergent.
n=1

Example 1.5.5. Find the sum of the series



X 1 + 2n+1
.
n=1
3n

Solution. The given series is the sum of two geometric series,


∞ ∞ 1
X 1 X 1 1 3
n
= n−1
= = 1/2
n=1
3 n=1
3 3 1 − 1/3
and ∞ ∞ 4
X 2n+1 X 4 2 n−1 3
= ( ) = = 4.
n=1
3n n=1
3 3 1 − 2/3
1 9
Thus its sum is 2
+4= 2
by Theorem 4.6.1(a). ◭

1.6 Nonnegative term series

If every term in a series is nonnegative, we may sometimes determine whether or not the series
is convergent by comparing it with a series of nonnegative terms that we know to be convergent
or that we know to be divergent.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.7 Tests of convergence (integral, comparison, ratio and root tests) 23

1.7 Tests of convergence (integral, comparison, ratio and root tests)

In this and the next two sections we will develop several tests for determining the convergence
or divergence of an infinite series by examining the nth term an of the series. These tests will
confirm the convergence of a series without yielding a value for its sum. From the practical
point of view, however, this is all that is required. Once it has been ascertained that a series
is convergent, we can approximate its sum to any degree of accuracy desired by adding up the
terms of its nth partial sum sn , provided that n is chosen large enough.

1.7.1 The Integral Test


P
The Integral Test ties the convergence or divergence of an infinite series an to the convergence
n=1
R∞
or divergence of the improper integral f (x)dx, where f (n) = an .
1

Theorem 1.7.1 (The Integral Test). Suppose that f is a continuous, positive, and decreasing
function on [1, ∞) . If f (n) = an for n ≥ 1, then

X Z∞
an and f (x)dx
n=1 1

either both converge or both diverge.

Example 1.7.1. Determine if the following series is convergent or divergent.


∞ ∞
X 1 X 2
A. B. ne−n
n=2
n ln n n=0
∞ ∞
X 1 X 1
C. , p = constant D.
n=1
np n=1
n2

Solution. A.
ZM M
dx
lim = lim ln(ln x) = lim {ln(ln M ) − ln(ln 2)} = ∞
M →∞ x ln x M →∞ 2
M →∞
2

and the series diverges.

B.
ZM M
dx 1 −x2 1 1 2 1
lim −x 2 = lim − e = lim { e−1 − e−M } = e−1
M →∞ xe M →∞ 2 1
M →∞ 2 2 2
2

and the series converges.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.7 Tests of convergence (integral, comparison, ratio and root tests) 24

C. Consider
ZM ZM M
dx x1−p
−p M 1−p − 1
= x dx = =
xp 1−p 1 1−p
1 1

where p 6= 1.

– If p < 1,
M 1−p − 1
lim =∞
M →∞ 1−p
, so that the integral and thus the series diverges.
– If p > 1,
M 1−p − 1 1
=
lim
M →∞ 1−p p−1
, so that the integral and thus the series converges.
– If p = 1,
ZM ZM
dx dx
= = ln M
xp x
1 1

and lim ln M = ∞, so that the integral and thus the series diverges.
M →∞

D. Since p = 2 > 1, then the integral and thus the series converges.

Exercise 1.7.1. Use the integral test to determine the convergence of the following series.
∞ ∞
X
2 −n3
X 1
A. 7n e B. √
n=1 n=1
2n + 9
∞ 3 ∞
X [ln n] X 4n
C. D.
n=1
n n=1
n2 + 9

P-series ∞
1
P
If k > 0 then np
converges if p > 1 and diverges if p ≤ 1. Sometimes the series in this fact
n=k
are called p-series and so this fact is sometimes called the p-series test.

Example 1.7.2. Determine if the following series is convergent or divergent.


∞ ∞
X 1 X 1
A. B. √
n=4
n7 n=0
n
∞ ∞
X 1 X 1
C. √ D.
n=1
3
n n=1
n2

Solution. A. In this case p = 7 > 1 and so by this fact the series is convergent.
1
B. For this series p = 2
≤ 1 and so the series is divergent by the fact.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.7 Tests of convergence (integral, comparison, ratio and root tests) 25

1
C. For this series p = 3
≤ 1 and so the series is divergent by the fact.

D. For this series p = 2 > 1 and so the series is convergent by the fact.

It is important to note before leaving this section that in order to use the Integral Test the
series terms MUST be positive. If they are negative then the test doesn’t work. Also remember
that the test only determines the convergence of a series and does NOT give the value of the
series.

1.7.2 Comparison Test

The first part says that if we have a series whose terms are smaller than those of a known
convergent series, then our series is also convergent. The second part says that if we start with
a series whose terms are larger than those of a known divergent series, then it too is divergent.
P P
Theorem 1.7.2 (The Comparison Test). Suppose that an and bn are series with positive
terms.

P P
a. If bn is convergent and an ≤ bn for all n, then an is also convergent.
P P
b. If bn is divergent and an ≥ bn for all n, then an is also divergent.

Example 1.7.3. Determine if the following series is convergent or divergent.


∞ ∞
X n X n2 + 2
A. B.
n=1
n − cos2 n
2
n=1
n4 + 5
∞ ∞
X 5 X 1
C. 2
D.
n=1
2n + 4n + 3 n=2
ln n

Solution. A. Since the cosine term in the denominator doesn’t get too large we can assume
that the series terms will behave like,
n 1
2
=
n n
. which, as a series, will diverge. So, from this we can guess that the series will probably
diverge and so we’ll need to find a smaller series that will also diverge.
n n 1
> 2 = .
n2 2
− cos (n) n n
Then, since

X 1
n=1
n
diverges (it’s harmonic and the p-series test) by the Comparison Test our original series
must also diverge.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.7 Tests of convergence (integral, comparison, ratio and root tests) 26

B. The series terms should behave pretty much like


n2 1
4
= 2
n n
which will converge as a series. Therefore, we will need to find a larger series which also
converges. So,
n2 + 2 n2 + 2
< .
n4 + 5 n4
Then,
∞ ∞ ∞ ∞ ∞
X n2 + 2 X n2 X 1 X 1 X 1
4
= 4
+ 4
= 2
+
n=1
n n=1
n n=1
n n=1
n n=1
n4
As shown, we can write the series as a sum of two series and both of these series are
convergent by the p-series test. Therefore, since each of these series are convergent we
know that the sum,
n2 + 2
n4
is also a convergent series. Recall that the sum of two convergent series will also be
convergent. Now, since the terms of this series are larger than the terms of the original
series we know that the original series must also be convergent by the Comparison Test.
C. For large n the dominant term in the denominator is 2n2 so we compare the given series
P 5
with the series 2n2
. Observe that
5 5
< 2
2n2 + 4n + 3 2n
because the left side has a bigger denominator. (In the notation of the Comparison Test,
an is the left side and bn is the right side.) We know that
∞ ∞
X 5 5X 1
=
n=1
2n2 2 n=1 n2
is convergent because it’s a constant times a p-series with p = 2 > 1. Therefore


X 5
n=1
2n2
is convergent by part (i) of the Comparison Test.

1
> n1 . Since 1
P
D. For n = 2, 3, 4, . . . , we have 0 < ln n < n. Thus ln n n
diverges to infinity(it
n=2

1
P
is a harmonic series), so does ln n
by harmonic series.
n=2

The terms of the series being tested must be smaller than those of a convergent series or
larger than those of a divergent series. If the terms are larger than the terms of a convergent
series or smaller than those of a divergent series, then the Comparison Test doesn’t apply.
Consider, for instance, the series

X 1
n=1
2n − 1

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.7 Tests of convergence (integral, comparison, ratio and root tests) 27

The inequality
1 1
> n
−1 2n
2
bn = ( 21 )n is convergent and
P
is useless as far as the Comparison Test is concerned because
P 1
an > bn . Nonetheless, we have the feeling that ought to be convergent because it is
P 1 2n −1
very similar to the convergent geometric series ( 2 )n . In such cases the following test can be
used.

Exercise 1.7.2. Use the integral test to determine the convergence of the following series.
∞ ∞
X n X 5n
A. 3
B.
n=1
n +1 n=1
7n + 1
∞ ∞
X 1 X n2
C. D.
n=1
n + n2 + 5
3
n=1
n3 − 5

1.7.3 Limit Comparison Test

an
P P
Theorem 1.7.3. Let an and bn be non negative series. Suppose lim = L, where L is
n→∞ bn
a positive number. If L is positive (i.e. L > 0 ) and is finite (i.e. L < ∞ ), then either both
series converge or both series diverge.

Example 1.7.4. Determine if the following series is convergent or divergent.


∞ ∞
X 1 X n+5
A. √ B.
n=1
1+ n n=1
n3 − 2n + 3
∞ 2 ∞
X 2n + 3n X 1
C. √ D.
n=1
5 + n5 n=2
2n −1

Solution. A. The terms of the series decrease √1 . Observe that


n

1√ √
1+ n n 1
L= lim = lim √ = lim √ = 1.
n→∞ √1 n→∞ 1 + n n→∞ (1/ n) + 1
n

∞ ∞
√1 1√
P P
Since the p−series n
diverges to ∞ (p = 1/2) so does the series 1+ n
by Limit
n=1 n=1
Comparison Test.

B. For large n the terms behave like n/n3 so let us compare the series with the p− series

1
P
n2
, which we know converges.
n=1

n+5
n3 −2n+3 n3 + 5n2
L= lim 1 = lim 3 = 1.
n→∞ n→∞ n − 2n + 3
n2

n+5
P
Since L < ∞, the series n3 −2n+3
also converges by Limit Comparison Test.
n=1

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.8 Alternating series and alternating series test 28

C. The dominant part of the numerator is 2n2 and the dominant part of the denominator is
√ 5
n5 = n 2 . This suggests taking
2n2 + 3n 2n2 2
an = √ , bn = 5 = 1
5 + n5 n 2 n2
1 5 3
an 2n2 + 3n n 2 2n 2 + 3n 2 2+ 3 2+0
lim = lim √ = lim √ = lim q n = √ =1
n→∞ bn n→∞ 5 + n5 2 n→∞ 2 5 + n5 n→∞ 5 2 0 + 1
2 +1 n5
1
bn = 2 1/n1/2 is divergent ( p-series with p =
P P
Since 2
< 1 ), the given series diverges
by the Limit Comparison Test.

D. We use the Limit Comparison Test with


1 1
an = , bn =
2n −1 2n
and obtain
1
an 2n −1 2n 1
lim = lim 1 = lim = lim =1>0
n→∞ bn n→∞ n→∞ 2n − 1 n→∞ 1 − 1/2n
2n

1/2n is a convergent geometric series, the given series con-


P
Since this limit exists and
verges by the Limit Comparison Test.

Exercise 1.7.3. Determine if the following series is convergent or divergent.


∞ ∞
X 1 X n
A. 2
B.
n=1
n +1 n=1
4n2 −3
∞ ∞
X n+2 X 3n
C. √ D.
n=1
(n + 1) n + 3 n=2
n5n

1.8 Alternating series and alternating series test

The convergence tests that we have looked at so far apply only to series with positive terms.
In this section and the next we learn how to deal with series whose terms are not necessarily
positive. Of particular importance are alternating series, whose terms alternate in sign. The
last two tests that we looked at for series convergence have required that all the terms in the
series be positive. Of course there are many series out there that have negative terms in them
and so we now need to start looking at tests for these kinds of series.

Definition 1.8.1. An alternating series is a series whose terms are alternately positive and
negative.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.8 Alternating series and alternating series test 29

Example 1.8.1.

X (−1)n−1 1 1 1 1 1 1
A. =1− + − + − + + ···
n=1
n 2 3 4 5 6 7

X n 1 2 3 4 5
B. (−1)n = − + − + − + ···
n=1
n+1 2 3 4 5 6

We see from these examples that the nth term of an alternating series is of the form

an = (−1)n−1 bn or an = (−1)n bn .

Where bn is a positive number (In fact, bn = |an |.)) The following test says that if the terms of
an alternating series decrease toward 0 in absolute value, then the series converges.

1.8.1 Alternating Series Test

an and either an = (−1)n bn or an = (−1)n+1 bn where bn ≥ 0,


P
Suppose that we have a series
for all n. Then if,

1. lim bn = 0
n→∞

2. bn+1 ≤ bn for all n.

then the series converges.

Example 1.8.2. Determine if the following series is convergent or divergent.


∞ ∞
X (−1)n+1 X (−1)n n2
A. B.
n=1
n n=1
n2 + 5
∞ √ ∞
X n X cos(nπ)
C. (−1)n−3 D. √
n=0
n+4 n=1
n

Solution. A. First, identify the bn for the test.


∞ ∞
X (−1)n+1 X 1 1
= (−1)n+1 , bn =
n=1
n n=1
n n

Now, all that we need to do is run through the two conditions in the test.
1
lim bn = lim =0
n→∞ n→∞ n

and
1 1
> bn = = bn+1 .
n n+1
Both conditions are met and so by the Alternating Series Test the series must converge.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.9 Absolute and conditional convergence 30

B. First, identify the bn for the test.


∞ ∞
X (−1)n n2 X n2 n2
= (−1)n , bn =
n=1
n2 + 5 n=1
n2 + 5 n2 + 5
Lets check the conditions.
n2
lim bn = lim = 1.
n→∞ n→∞ n2 + 5

So, the first condition isn’t met and so there is no reason to check the second. Since this
condition isn’t met well need to use another test to check convergence. In these cases
where the first condition isn’t met it is usually best to use the divergence test.
(−1)n n2 n2
 
n
lim = lim (−1) lim
n→∞ n2 + 5 n→∞ n→∞ n2 + 5
 
= lim (−1)n (1)
n→∞

= lim (−1)n doesn’t exist


n→∞
This limit doesn’t exist and so by the Divergence Test this series diverges.

n
C. Let bn = .
n+4 √
n
lim bn = lim =0
n→∞ n+4 n→∞
Let’s start with the following function and its derivative.

x 4−x
f (x) = , f ′ (x) = √
x+4 2 x(x + 4)2
Now, there are three critical points for this function, x = −4, x = 0, and x = 4. The first
is outside the bound of our series so we wont need to worry about that one. Using the
test points, √
′ 3 ′ 5
f (1) = f (5) =
50 810
and so we can see that the function in increasing on 0 ≤ x ≤ 4 and decreasing on x ≥ 4.
Therefore, f (n) = bn since we know as well that the bn are also increasing on 0 ≤ n ≤ 4 and
decreasing on n ≥ 4. The bn are then eventually decreasing and so the second condition
is met. Both conditions are met and so by the Alternating Series Test the series must be
convergent.
D. To check for the convergence or divergence of this series use the fact cos(nπ) = (−1)n and
then ∞ ∞
X cos(nπ) X (−1)n
√ = √
n=1
n n=1
n

1.9 Absolute and conditional convergence

We have convergence tests for series with positive terms and for alternating series. But what if
the signs of the terms switch back and forth irregularly? We will see that the idea of absolute
convergence sometimes helps in such cases.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.9 Absolute and conditional convergence 31

P
Definition 1.9.1. A series an is called absolutely convergent if the series of absolute
P
values |an | is convergent.
P
Notice that if an is a series with positive terms, then |an | = an and so absolute convergence
is the same as convergence in this case.
Example 1.9.1. The series

X (−1)n−1 1 1 1
=1− + 2 − 2 + ...
n=1
n2 2 2 3 4
is absolutely convergent because

X (−1)n−1 1 1 1 1
2
= 2
= 1 + 2
+ 2
+ 2
+ ...
n=1
n n 2 3 4

is a convergent p-series ( p = 2).


Example 1.9.2. We know that the alternating harmonic series

X (−1)n−1 1 1 1
=1− + − + ...
n=1
n 2 3 4
is convergent, but it is not absolutely convergent because the corresponding series of absolute
values is ∞
X (−1)n−1 1 1 1 1
= = 1 − + − + ...
n=1
n n 2 3 4
which is the harmonic series ( p−series with p = 1) and is therefore divergent.
P
Definition 1.9.2. A series an is called conditionally convergent if it is convergent but
not absolutely convergent.

Example 1.9.2 above shows that the alternating harmonic series is conditionally convergent.
Thus, it is possible for a series to be convergent but not absolutely convergent. However, the
next theorem shows that absolute convergence implies convergence. We also have the following
fact about absolute convergence.
P
Theorem 1.9.1. If an is absolutely convergent, then it is also convergent.

1.9.1 Ratio Test

In this section we are going to take a look at a test that we can use to see if a series is absolutely
convergent or not. Recall that if a series is absolutely convergent then we will also know that
it’s convergent and so we will more often than not use it to simply determine the convergence
of a series. Before proceeding with the test let’s do a quick reminder of factorials. This test
will be particularly useful for series that contain factorials. If n is an integer such that n ≥ 0,
then n factorial is defined as,

n! = n(n − 1)(n − 2)(n − 3) . . . 3 · 2 · 1 if n ≥ 1.


0! = 1! = 1 by definition

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.9 Absolute and conditional convergence 32

Also, when dealing with factorials we need to be very careful with parenthesis. For (2n)! 6= 2n!
as we can see
(2n)! = (2n)(2n − 1)(2n − 2)(2n − 3)(2n − 4) · · · 3 · 2 · 1
2n! = 2[(n)(n − 1)(n − 2) · · · 3 · 2 · 1]
P
Theorem 1.9.2 (Ratio Test). Suppose we have the series an . Define
an+1
r = lim
n→∞ an
Then,

1. If r < 1 the series is absolutely convergent (and hence convergent).


2. If r > 1 the series is divergent.
3. If r = 1 the series may be divergent, convergent, or absolutely convergent.
Example 1.9.3. Determine if the series is convergent or divergent.
∞ ∞
X n3 X n!
A. (−1)n n B. (−1)n
n=1
3 n=1
5n
∞ ∞
X 1 X 1
C. (−1)n D. (−10)n
n=0
n2 + 1 n=1
42n+1 (n + 1)

Solution. A. We use the Ratio Test with


n3
an = (−1)n
3n
(n + 1)3
⇒ an+1 = (−1)n+1 n+1
3
3
(−1)n+1 (n+1)
3
(n + 1)3 3n

an+1 3 n+1 1 n+1 1
r = lim = lim n 3 = lim n+1 3
= lim = <1
n→∞ an n→∞ n
(−1) 3n n→∞ 3 n n→∞ 3 n 3
Thus, by the Ratio Test, the given series is absolutely convergent and therefore convergent.
B. Here is the limit.
an+1 (n + 1)!5n 1 (n + 1)! n+1
r = lim = lim n+1
= lim = lim =∞>1
n→∞ an n→∞ 5 n! n→∞ 5 n! n→∞ 5
So, by the Ratio Test this series diverges.
C. Lets first get r,
an+1 (−1)(n+1) (n+1)1 2 +1 n2 + 1
r = lim = lim = lim =1
n→∞ an n→∞ (−1)n n21+1 n→∞ (n + 1)2 + 1

So, as implied earlier we get r =1 which means the ratio test is no good for determining
the convergence of this series. We will need to resort to another test for this series. This
series is an alternating series and so lets check the two conditions from that test.
1
lim =0
n→∞ n2 +1

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.9 Absolute and conditional convergence 33

1 1
bn = > = bn+1 .
n2 +1 (n + 1)2 + 1
The two conditions are met and so by the Alternating Series Test this series is convergent.

D. Check that this series is convergent.


 
an+1 P
Raabe’s test. Let lim n 1 − an
= L. Then the series an
n→∞

(a) converges (absolutely) if L > 1.

(b) diverges or converges conditionally if L < 1.

If L = 1 the test fails. This test is often used when the ratio tests fails.
Gauss’ test
If an+1 L cn
P
an
= 1 − n
+ n 2 , where |c n | < P for all n > N , then the series an

(a) converges (absolutely) if L > 1.

(b) diverges or converges conditionally if L ≤ 1.

This test is often used when Raabe’s test fails.

Exercise 1.9.1. Use the ratio test to determine the convergence of the following series.
∞ ∞
X 2n X 1
A. B.
n=1
n! n=1
n2 +1
∞ ∞
X X 5n
C. ne−n D.
n=0 n=0
2n + 3 n

1.9.2 Root Test

This is the last test for series convergence that we’re going to be looking at. As with the Ratio
Test this test will also tell whether a series is absolutely convergent or not rather than simple
convergence. This test is convenient to apply when nth powers occur.
P
Theorem 1.9.3 (Root Test). Suppose we have the series an . Define
p 1
r = lim n |an | = lim |an | n
n→∞ n→∞

Then,

1. If r < 1 the series is absolutely convergent (and hence convergent).

2. If r > 1 the series is divergent.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.10 Generalized convergence tests 34

3. If r = 1 the series may be divergent, convergent, or absolutely convergent.

As with the ratio test, if we get r = 1 the root test will tell us nothing and we’ll need to use
another test to determine the convergence of the series. As an exercise for the reader please
1
show that lim n n = 1.

Example 1.9.4. Determine if the series is convergent or divergent.


∞ ∞
X nn X 2n + 3 n
A. B. ( )
n=1
31+2n n=1
3n + 2
∞ ∞
X 2n X 2n
C. D.
n=1
n n=1
n2

Solution. A. Here is the limit of the sequence


1
nn n n
r = lim = lim 1 =∞>1
n→∞ 31+2n 3 n
+2

So, by the Root Test this series is divergent

B. Now lets find the value of r


1
2n + 3 n n 2n + 3 2
r = lim = lim = <1
n→∞ 3n + 2 n→∞ 3n + 2 3
Hence, by the Root Test this series is convergent.

C. Similarly,
1
2n n 2
r = lim = lim 1
n→∞ n n→∞ nn
1
Now let’s compute lim n n = 1 Hence,
n→∞

1
2n n 2
r = lim = lim 1 =2>1
n→∞ n n→∞ nn
Therefore, by the root test the given series diverges.

Note: if we get r = 1 from the ratio test then the root test will also give r = 1 and so there
isn’t any reason to try the root test on anything that gives r = 1 on the ratio test.

1.10 Generalized convergence tests

By combining absolute convergence tests with tests for non negative series, we obtain conver-
gence tests that apply to any series, non negative or not.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

1.10 Generalized convergence tests 35

P
Theorem 1.10.1 (Generalized convergence tests). Let be an a series.


P ∞
P
a) Generalized comparison test: If |an | ≤ |bn | and |bn | converges, then an converges
n=1 n=1
(absolutely).

an
b) Generalized Limit comparison test: If lim bn
= L, where L is a positive number, and if

P
|an | ≤ |bn | converges then an converges (absolutely).
n=1

an
c) Generalized ratio test: Suppose that an 6= 0 for n ≥ 1 and lim = r(possibly ∞) that
n→∞ an+1


P
i) If r < 1, then an converges (absolutely).
n=1
P∞
ii) If r > 1, then an diverges
n=1
iii) If r = 1, no conclusion can be drawn.
p
d) Generalized root test: Suppose that lim n |an | = r (possibly ∞).
n→∞

P
i) If r < 1, then an converges (absolutely).
n=1
P∞
ii) If r > 1, then an diverges
n=1
iii) If r = 1, no conclusion can be drawn.

xn x2 x3
P
Example 1.10.1. Show that n
= x+ 2
+ 3
+ · · · converges absolutely for |x| < 1.
n=1
converges conditionally for x = −1, and diverges for x = 1 and for |x| > 1.

Solution. If x = 0, the series obviously converges. If x 6= 0, then


xn+1 /(n + 1) n
lim n
= |x| lim = |x|
n→∞ x /n n→∞ n+1
Therefore the Generalized Ratio Test implies that the given series converges for |x| < 1 and
diverges |x| > 1. ◭
xn
Example 1.10.2. Show that lim for all x.
n→∞ n!

xn
Solution. If x = 0, then the limit is 0. If x 6= 0, let an = n!
. Then

an+1 xn+1 /(n + 1)! 1


r = lim = lim n
= lim |x| = 0.
n→∞ an n→∞ x /n! n→∞ n + 1

Since r < 1, the series converges for all x . ◭

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

Chapter 2

Power Series

2.1 Introduction

Until now, we have discussed a series with constant terms. Now, We extend our discussion
of series to the case where the terms of the series are functions of the variable x. Pay close
attention, as the primary reason for studying series is that we can use them to represent
functions. This opens up numerous possibilities for us. For instance,

• From approximating the values of transcendental functions to calculating derivatives and


integrals of such functions
• Developing the mathematical tools needed to investigate the convergence of Taylor and
Maclaurin series.
• Defining functions as convergent series produces an explosion of new functions available
to us, including many important functions, such as the Bessel functions.

2.2 Definition of power series at any x0 and x0 = 0

Definition 2.2.1. Let x be a variable. A power series in x or a power series centered


at the origin is a series of the form

X
an xn = a0 + a1 x + a2 x2 + a3 x3 + · · · + an xn + · · ·
n=0

where the an ’s are constants and are called the coefficients of the series.
More generally, a power series in (x − x0 ) , where is x0 constant, is a series of the form

X
an (x − x0 )n = a0 + a1 (x − x0 ) + a2 (x − x0 )2 + a3 (x − x0 )3 + · · · + an (x − x0 )n + · · ·
n=0

Remark 2.2.1. 1) A power series in (x − c) is also called a power series centered at c,


or a power series about c . Thus, a power series in x is just a series centered at the
origin.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

2.2 Definition of power series at any x0 and x0 = 0 37

2) To simplify the notation used for a power series, we have adopted the convention that
(x − c)0 = 1, even when x = c.

We can view a power series as a function defined by the rule



X
f (x) = an (x − c)n
n=0

The domain of f is the set of all x for which the power series converges, and the range of f
comprises the sums of the series obtained by allowing x to take on all values in the domain of
f . If a function f is defined in this manner, we say that f is represented by the power series

an (x − c)n .
P
n=0

Example 2.2.1. Determine the values of x for which the following power series converges.
∞ ∞ ∞
X n n X (x − 3)n X
A. n+1
x B. C. n!xn
n=0
3 n=0
n n=0

Solution. A. Using the generalized Ratio Test, we have


an+1 (n + 1)xn+1 3n+1
lim = lim ·
n→∞ an n→∞ 3n+2 nxn
(n + 1)|x| |x| n+1
= lim = lim
n→∞ 3n 3 n→∞ n
|x|
= <1 = |x| ⇔ −3 < x < 3
3
So, the series converges absolutely for −3 < x < 3 and diverges for |x| > 3 (i.e., for x > 3
or x < −3). Since the Ratio Test gives no conclusion for the endpoints x = ±3, we must
test these separately. For x = 3, we have the series
∞ ∞ ∞
X n n X n n Xn
x = 3 =
n=0
3n+1 n=0
3n+1 n=0
3

Since lim n3 = ∞ = 6 0 the series diverges by the test for divergence. The series diverges
n→∞
when x = −3, for the same reason. Thus, the power series converges for all x in the
interval (−3, 3) and diverges for all x outside this interval.

B. Using the generalized Ratio Test, we have


an+1 (x − 3)n+1 n
lim = lim ·
n→∞ an n→∞ n+1 (x − 3)n
n
= lim (x − 3)
n→∞ n + 1
n
= |x − 3| lim
n→∞ n + 1

= |x − 3| < 1 ⇐⇒ 2 < x < 4

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

2.3 Convergence and divergence, radius and interval of convergence 38

Hence, the series converges absolutely for 2 < x < 4 and diverges for |x − 3| > 1 (i.e., for
x < 2 or x > 4). Since the Ratio Test gives no conclusion for the endpoints x = 2 and
x = 4 , we must test these separately.

(−1)n n1 which is converge series.
P
. For x = 2, we have the series
n=0

(−1)n n1 which is divergent series.
P
. For x = 2, we have the series
n=0
Therefore,The given power series converges for all x in the interval [2, 4) and diverges for
all x outside this interval.

C. For x 6= 0 and using the generalized Ratio Test, we have


an+1 (n + 1)!xn+1
lim = lim
n→∞ an n→∞ n!xn

= lim (n + 1)x
n→∞

= |x| lim n + 1
n→∞

=∞

Hence,by generalized Ratio Test the series diverges when x 6= 0. Thus, given series con-
verges for x = 0.

Exercise 2.2.1. Determine the values of x for which the following power series converges.
∞ ∞ ∞
X xn X xn X 2 n xn
A. B. C. (−1)n+1
n=0
(n + 1)2n n=1
n! n=0
n3n

2.3 Convergence and divergence, radius and interval of convergence

The following theorem, which we state without proof, tells us that the domain of a power series
is always an interval with x = c as its center. In the extreme cases the domain consists of
the infinite interval (−∞, ∞) or just the point x = c, which may be regarded as a degenerate
interval.

an (x − c)n , exactly one of the following is true:
P
Theorem 2.3.1. Given a power series
n=0

1. The series converges only at x = c .

2. The series converges for all x.

3. There is a number R > 0 such that the series converges for |x − c| < R and diverges for
|x − c| > R.

• The number R referred to in Theorem 2.3.1 is called the radius of convergence of the
power series. The radius of convergence is R = 0 in case (1) and ∞ in case (2).

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

2.3 Convergence and divergence, radius and interval of convergence 39

• The set of all values for which the power series converges is called the interval of con-
vergence of the power series.

Theorem 2.3.1 tells us that the interval of convergence of a power series centred at c is one of
the following

• just the single point c

• the interval (−∞, ∞)

• a finite interval centered at c

[c − R, c + R], or (c − R, c + R), or [c − R, c + R), or (c − R, c + R]

The radius of convergence, R, and found by using the ratio test on the power series: If
an+1 (x − c)n+1
 
an+1
ρ = lim = lim |x − c|
n→∞ an (x − c)n n→∞ an

an (x − c)n converges absolutely when ρ < 1, that is, where
P
exists, then the series
n=0

an+1
|x − c| < R = 1/ lim
n→∞ an
The series diverges if |x − c| > R.


an+1
an (x − c)n
P
Remark 2.3.1. Suppose that L = lim exists or ∞. Then the power series
n→∞ an n=0
has radius of convergence R = 1/L. (If L = ∞, then R = 0; if L = 0, then R = ∞.)

an xn converges absolutely on the interior of its interval of


P
Theorem 2.3.2. A power series
convergence.

Example 2.3.1. Determine the center, radius and interval of convergence of the following
power series
∞ ∞ ∞ ∞
X
n
X xn X
n
X (−1)n xn
A. x B. C. n!x D.
n=0 n=0
n! n=0 n=1
3n (n + 1)

Solution. A. By apply generalized ratio test , We have


an+1 xn+1
L = lim = lim
n→∞ an n→∞ xn
= lim |x|
n→∞

= |x|

Hence, the series is absolute converge if L = |x| < 1 and diverges if L = |x| > 1. The test
is inconclusive if |x| = 1 ( i.e if x = −1 and x = 1). so that we will have to investigate

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

2.3 Convergence and divergence, radius and interval of convergence 40

convergence at these points separately. At these points the series becomes


X∞
1n = 1 + 1 + 1 + · · · f or x = 1
n=0

X
(−1)n = 1 − 1 + 1 − · · · f or x = −1
n=0

These both series diverge. Thus the given series converges in the interval (−1, 1).
Therefore,the interval of convergence for the given power series is (−1, 1), and the radius
of convergence is R = 1 .
B. Applying the generalized ratio test for absolute convergence, we have
an+1 x(n+1)! n!
L = lim = lim ·
n→∞ an n→∞ (n + 1)! xk

x
= lim
n→∞ n + 1

=0
Since L = 0 < 1 for all x,the series converges absolutely for all x.
Therefore,the interval of convergence is (−∞, +∞) and Radius of convergence is R= +∞
c. If x 6= 0, then the generalized ratio test for absolute convergence yields
an+1 (n + 1)!xn+1
L = lim = lim
n→∞ an n→∞ n!xk

= lim (n + 1)x
n→∞

= +∞
Hence, the series diverges for all non zero values of x. Therefore,the interval of convergence
is single point x = 0 and the radius of convergence is R = 0
D. Since |(−1)k | = |(−1)k+1 | = 1 and using generalized ratio test for absolute convergence we
obtain
an+1 xn+1 3n (n + 1)
L = lim = lim n+1 ·
n→∞ an n→∞ 3 xn
 
|x| k + 1
= lim ·
n→∞ 3 k+2
|x| k+1
= lim
3 n→∞ k + 2
|x|
=
3
Hence, by the generalized ratio test for absolute convergence implies that the given series
converges absolutely if |x| < 3 and diverges if |x| > 3. The test is inconclusive if |x| = 3
( i.e if x = −3 and x = 3). so that we will have to investigate convergence at these points
separately.
Substituting x = −3 in the given series yields
∞ ∞
X (−1)n (−3)n X 1
n (n + 1)
=
n=1
3 n=1
n+1

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

2.4 Representations of Functions as Power Series 41

which is divergent harmonic series 1 + 12 + 31 + 14 + · · · .


Substituting x = 3 in the given series yields
∞ ∞
X (−1)n (3)n X (−1)n 1 1 1
= =1− + − + ··· .
n=1
3n (n + 1) n=1
n+1 2 3 4

which is the conditionally convergent alternating harmonic series.Thus,the interval of con-


vergence for the given series (−3, 3] and the radius convergence is R = 3

Exercise 2.3.1. 1. Determine the centre, radius and interval of convergence of the following
power series
∞ ∞
X (x − 2)n X (−1)n 2n xn
A. B. ( √
n=1
n2 3n n=0
n+1
∞ ∞
X n(x + 2)n X (−1)n xn
C. D.
n=0
3n+1 n=0
3n (n + 1)

2. Show that if p is a positive integer, then the power series



X (pn)!
xn
n=0
(n!)p
p
has a radius of convergence of 1/p

2.4 Representations of Functions as Power Series

In this section we learn how to represent certain types of functions as sums of power series
by manipulating geometric series or by differentiating or integrating such a series.You might
wonder why we would ever want to express a known function as a sum of infinitely many
terms. We will see later that this strategy is useful for integrating functions that don’t have
elementary antiderivatives, for solving differential equations, and for approximating functions
by polynomials.
Recall that the geometric series which is given by

X a
axn = provided that |x| < 1
n=0
1−x

Now, if we take a = 1 the series becomes



X 1
xn = provided that |x| < 1
n=0
1−x

1
xn provided that |x| < 1.
P
So that we can represent the function f (x) = 1−x
with power series
n=0
That means

1 X
f (x) = = xn provided that |x| < 1 (2.1)
1 − x n=0

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

2.5 Algebraic operations on convergent power series 42

Remark 2.4.1. The radius of convergence of this power series is R = 1 and the interval of
convergence is |x| < 1.
Example 2.4.1. Find a power series representation for the following function and determine
its interval of convergence.
1 2x2 x
A. g(x) = B. f (x) = C. h(x) =
1 + x3 1 + x3 5−x
1 1
Solution. A. By rewriting g(x) = 1+x3
= 1−(−x3 )
and substituting x = −x3 in equation 2.1
,we have
1 1
g(x) = 3
=
1+x 1 − (−x3 )
X∞
= (−x3 )n
n=0

X
= (−1)n (x3n ) provided that | − x3 | < 1 ⇔ |x| < 1
n=0

1
(−1)n (x3n ) with
P
Therefore, the function g(x) = 1+x3
can be represented by power series
n=0
the interval of convergence is |x| < 1.

B. Exercise!
x 1 1 x
C. By rewriting h(x) = 5−x
= 5
· 1− x5
and substituting x = 5
in equation 2.1 ,we have

x 1 1
g(x) = = ·
5−x 5 1 − x5

1 X  x n x
= provided that | | < 1 ⇔ |x| < 5
5 n=0 5 5

x 1 x n
P 
Therefore, the function h(x) = 5−x
can be represented by power series 5 5
with the
n=0
interval of convergence |x| < 5.

2.5 Algebraic operations on convergent power series

∞ ∞
an xn and bn xn be two power series with radius of convergence Ra
P P
Theorem 2.5.1. Let
=0 n=0
and Rb respectively, and let c be a constant. Then,

can xn has radius of convergence Ra , and
P
i.
n=0

X ∞
X
n
(can )x = c an xn
n=0 n=0

wherever the series on the right converges.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

2.6 Differentiation and integration of power series 43


(an + bn )xn has radius of convergence R at least as large as the smaller of Ra and
P
ii.
n=0
Rb (i.e R ≥ min{Ra , Rb }), and
X∞ ∞
X ∞
X
(an + bn )xn = an xn + bn x n
n=0 n=0 n=0

wherever both series on the right converge.


∞  ∞ 
n n
P P
iii. an x · bn x has radius of convergence R at least as large as the smaller of Ra
n=0 n=0
and Rb (i.e R ≥ min{Ra , Rb }), and

! ∞
! ∞
X X X
an xn · bn x n = cn xn
n=0 n=0 n=0

where

X
cn = a0 bn + a1 bn−1 + ... + an b0 = aj bn−j
j=0
∞ ∞ ∞
cn xn is called Cauchy product of the series an xn and bn x n .
P P P
The series
n=0 n=0 n=0
1
Example 2.5.1. Find the power series representation of the function f (x) = by using
(1 − x)2
Cauchy product of series.

1
xn holds for −1 < x < 1. We can determine a power series
P
Solution. We know that 1−x
=
n=0
1
representation for (1−x) 2 by taking the Cauchy product of this series with itself.

Since an = bn = 1 for n = 0, 1, 2, ..., we have


X∞
cn = 1 = n + 1 and
n=0

1 X
f (x) = 2
= (n + 1)xn provided that |x| < 1
(1 − x) n=0

2.6 Differentiation and integration of power series

Theorem 2.6.1. Suppose that f (x) the sum of the power series on an interval I; that is,
X∞
f (x) = an xn = a0 + a1 x + a2 x2 + a3 x3 + · · ·
n=0

then, if x is interior to I,

1. Differentiation of power series



X ∞
X
′ n
f (x) = Dx (an x ) = n(an xn−1 )
n=0 n=1

= a1 + 2a2 x + 3a3 x2 + · · ·

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

2.6 Differentiation and integration of power series 44

2. Integration of power series


Z x ∞ Z x ∞
X
n
X an n+1 1 1 1
f (t)dt = an t dt = x = a0 x + a1 x2 + a2 x3 + a3 x4 + · · ·
0 n=0 0 n=0
n+1 2 3 4

Example 2.6.1. Apply Theorem 2.6.1 to the geometric series


1
= 1 + x + x2 + x3 + · · ·
1−x
to obtain formulas for two new series.

Solution. Differentiating term by term yields


1
= 1 + 2x + 3x2 + 4x3 + · · · , −1 < x < 1.
(1 − x)2
Integration term by term gives
Z x Z x Z x Z x Z x
1 2
dt = 1dt + tdt + t dt + t3 dt + · · ·
0 1−t 0 0 0 0

That is,
x2 x3
− ln(1 − x) = x + + + · · · , −1 < x < 1
2 3
If we replace x by −x in the later and multiply both sides by -11, we obtain
x2 x3 x4
ln(1 + x) = x − + − + ··· , −1 < x < 1
2 3 4

Example 2.6.2. Use the geometric series to find a power series representation for tan−1 x.

Solution. Substitute −t2 for x in the geometric series. Since 0 ≤ t2 < 1 whenever −1 < t < 1,
we obtain
1
= 1 − t2 + t4 − t6 + t8 − · · · (−1 < t < 1)
1 + t2
Now integrate from 0 to x, where |x| < 1:
Z x Z x
−1 1
tan x = 2
dt = (1 − t2 + t4 − t6 + t8 − · · · )dt
0 1+t 0
3 5
x x x7 x9
=x− + − + + ···
3 5 7 9
∞ 2n+1
n x
X
= (−1) (−1 < x < 1)
n=0
2n + 1

2
Exercise 2.6.1. 1. Find a power series representation for 2xex given that a series repre-
∞ n
x
sentation for ex is ex =
P
n!
in (−∞, ∞).
n=0

2. Find a power series representation for cos x.


3.
4.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

2.7 Taylor series; Taylor polynomial and application 45

2.7 Taylor series; Taylor polynomial and application


an (x − c)n has a positive radius of convergence R, then the sum of the
P
If a power series
n=0
series defines a function f (x) on the interval (c − R, c + R). We say that the power series is a
representation of f (x) on that interval. What relationship exists between the function f (x) and
the coefficients a0 , a1 , a2 , . . . of the power series? The following theorem answers this question

Theorem 2.7.1. Suppose the series



X
f (x) = an (x − c)n == a0 + a1 (x − c) + a2 (x − c)2 + a3 (x − c)3 + · · ·
n=0

converges to f (x) for c − R < x < c + R, where R > 0. Then

f (k) (c)
ak =
k!
for k = 0, 1, 2, 3, . . . .

Definition 2.7.1 (Taylor and Maclaurin series). If f (x) has derivatives of all orders at
x = c(i.e.,if f (k) (c) exists for k = 0, 1, 2, 3, . . . ), then the series

X f (n) (c)
(x − c)n
n=0
n!
f ′′ (c) f (3) (c)
= f (c) + f ′ (c)(x − c) + (x − c)2 + (x − c)3 + · · ·
2! 3!

is called the Taylor series of f about c (or the Taylor series of f in powers of x − c). If
c = 0,the Taylor series becomes
f ′′ (0) 2 f (3) (0) 3
f (x) = f (0) + f ′ (0)x + x + x + ···
2! 3!
and this series is called Maclaurin series.

2.7.1 Polynomial Approximation

The nth partial sum of the Taylor series o centred at c,


n
X f (k) (c)
(x − c)k
k=0
k!
f ′′ (c) f (3) (c)
= f (c) + f ′ (c)(x − c) + (x − c)2 + (x − c)3 + · · ·
2! 3!

is called the nth -degree Taylor polynomial of f at c. If c = 0, we have the nth -degree
Maclaurin polynomial of f . First, notice that the partial sums of a Taylor series (like those

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

2.7 Taylor series; Taylor polynomial and application 46

for any power series) are simply polynomials. We define


n
X f (k) (c)
Pn (x) = (x − c)k
k=0
k!
f ′′ (c) f (n) (c)
= f (c) + f ′ (c)(x − c) + (x − c)2 + · · · + (x − c)n
2! n!
Theorem 2.7.2. If f has derivatives up to order n + 1 in an interval I containing c, then for
each x in I, there exists a number z between x and c such that
f ′′ (c) f (3) (c) f (n) (c)
f (x) = f (c) + f ′ (c)(x − c) + (x − c)2 + (x − c)3 + · · · + (x − c)n + Rn (x)
2! 3! n!
(2.2)
= Pn (x) + Rn (x)

where
f (n+1) (z)
Rn (x) = (x − c)n+1 .
(n + 1)!

n
f (k) (c)
− c)k which is a
P
Given f (x), x0 and n. Let Rn (x) = f (x) − Pn (x) = f (x) − k!
(x
k=0
remainder (error) made in approximating f (x) by Pn (x).

Example 2.7.1. For f (x) = ex , find the Taylor polynomial of degree n expanded about x = 0.
Then determine a) Pn (1) b) P5 (1).

′′
Solution. f (x) = ex , f ′ (x) = ex , f (x) = ex , f (3) (x) = ex and f (n) (x) = ex for all n. Thus,
f ′ (0) f ′′ (0) 2 f (3) (0) 3 f (n) (0) n
Pn (x) = f (0) + x+ x + x + ··· + x
1! 1! 3! n!
n
x2 x3 xn X xk
=1+x+ + + ··· + =
2! 3! n! k=0
k!

As a result
n
1 1 1 1 X 1
Pn (1) = 1 + + + + ... + = and
1! 2! 3! n! k=0 k!
n
1 1 1 1 1 X 1
P5 (1) = 1 + + + + + =
1! 2! 3! 4! 5! k=0 5!

Theorem 2.7.3 (Taylor’s Inequality). If |f (n+1) (x)| ≤ M for |x − x0 | ≤ d , then the remainder
Rn (x) of the Taylor series satisfies the inequality

M
Rn (x) ≤ |x − x0 |n+1 for |x − x0 | ≤ d.
(n + 1)!

We now turn our attention to finding the conditions under which the function has a power
series representation.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

2.7 Taylor series; Taylor polynomial and application 47

Theorem 2.7.4. Suppose that f has derivatives of all order on an interval I containing x0 and
that Rn (x) is the Taylor remainder of f at c. If

lim Rn (x)
n→∞

for every x in I, then f (x) is represented by the Taylor series of f at c ; that is,

X f (n) (c)
f (x) = (x − c)n
n=0
n!

Theorem 2.7.5. If is any real number, then


|x|n
lim = 0.
n→∞ n!


xn
of the function f (x) = ex does represent
P
Example 2.7.2. Show that the Maclaurin series n!
n=0
f.

Solution. We use Taylor’s Theorem with c = 0 . Since f (n+1) (x) = ex , we see that

f (n+1) (z) n+1 ez


Rn (x) = x = xn+1
(n + 1)! (n + 1)!
where z is a number between 0 and x. If x > 0, then ez < ex , since the function f (x) = ex is
increasing. Therefore,
ex
0 < Rn (x) < xn+1
(n + 1)!
By Theorem 2.7.5,
ex xn+1
lim xn+1 = ex lim =0
n→∞ (n + 1)! n→∞ (n + 1)!

so the Squeeze Theorem implies that

lim Rn (x) = 0.
n→∞

If x < 0, then z < 0, and hence ez < e0 = 1. Therefore,


xn+1
0 < |Rn (x)| <
(n + 1)!
and once again, the Squeeze Theorem implies that

lim Rn (x) = 0.
n→∞

It follows from Theorem 2.7.4 that the Maclaurin series of represents the function f (x) = ex
for all x 6= 0. Finally, the series represents f at x = 0, since f (0) = e0 = 1, and this is also the
value of the sum of the series at 0. ◭

P n x2n+1
Example 2.7.3. Show that the Maclaurin series (−1) (2n+1)! of the function f (x) = sin x
n=0
does represent f .

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

2.7 Taylor series; Taylor polynomial and application 48

Solution. Using Taylor’s Theorem with c = 0, we have


f (n+1) (z) n+1
Rn (x) = x
(n + 1)!

where z is a number between 0 and x. But f (n+1) (x) is either ± sin x or ± cos x for any
n(n = 0, 1, 2, . . . ). Therefore, |f (n+1) (z)| ≤ 1 , so

f (n+1) (z)
|x|n+1
|Rn (x)| = |x|n+1 ≤
(n + 1)! (n + 1)!
By Theorem 2.7.5,
|x|n+1
lim =0
n→∞ (n + 1)!

so the Squeeze Theorem implies that

lim Rn (x) = 0.
n→∞

It follows from Theorem 2.7.4 that



X n x2n+1
f (x) = (−1)
n=0
(2n + 1)!

as was to be shown. ◭

2.7.2 Defined by Integrals

Many functions that can be expressed as simple combinations of elementary functions cannot be
antidifferentiated by elementary techniques; their antiderivatives are not simple combinations
of elementary functions. We can, however, often find the Taylor series for the antiderivatives
of such functions and hence approximate their definite integrals.

Example 2.7.4. Find the Maclaurin series for


Z x
2
E(x) = e−t dt,
0

and use it to evaluate E(1) correct to 3 decimal places.

Solution. The Maclaurin series for E(x) is given by


Z x
t4 t6 t8

2
E(x) = 1 − t + − + − · · · dt
0 2! 3! 4!
3 5 x
t7 t9
 
t t
= t− + − + − ···
3 5 × 2! 7 × 3! 9 × 4! 0
3 5 7 9 ∞
x x x x X x2n+1
=x− + − + − ··· = (−1)n ,
3 5 × 2! 7 × 3! 9 × 4! n=0
(2n + 1)n!

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

2.7 Taylor series; Taylor polynomial and application 49

2
and is valid for all x because the series for e−t is valid for all t. Therefore,
1 1 1
E(1) = 1 − + − + ···
3 5 × 2! 7 × 3!
1 1 1 (−1)n−1
≈1− + − + ···
3 5 × 2! 7 × 3! (2n − 1)(n − 1)!
We stopped with the nth term. Again, the alternating series test assures us that the error in this
approximation does not exceed the first omitted term, so it will be less than 0.0005, provided
(2n + 1)n! > 2, 000. Since 13 × 6! == 9, 360, n = 6 will do. Thus,
1 1 1 1 1
E(1) ≈ 1 − + − + − ≈ 0.747,
3 10 42 216 1, 320
rounded to 3 decimal places. ◭

2.7.3 Indeterminate Forms

In this section we will see how Maclaurin polynomials could be used for evaluating the limits
of indeterminate forms. Here are two examples, using the series directly and keeping enough
terms to allow cancellation of the [0/0] factors.
Example 2.7.5. Evaluate
x − sin x (e2x − 1)(ln(1 + x3 ))
A. lim B. lim
x→0 x3 x→0 (1 − cos 3x)2

Solution. A.
 
x3 x5
x− x− 3!
+ 5!
− ···
x − sin x
lim = lim
x→0 x3 x→0 x3
x3 x5
3!
− 5!
+ ···
= lim
x→0 x3
1 x2 1 1
= lim − + ··· = =
x→0 3! 5! 3! 6
B.
  
(2x)2 (2x)3 3 x6
1 + (2x) + + + ··· − 1 x − + ···
(e2x − 1)(ln(1 + x3 )) 2! 3! 2
lim = lim 2
(1 − cos 3x)2
 
x→0 x→0
(3x)2 (3x)4
1− 1− 2!
+ 4!
− ···
2 + 2x + · · ·
= lim  2
x→0
9 2 34 4
2
x − 4!
x + ···
2 + 2x + · · · 2 8
= lim 2 = 9 2 =
(2) 81

x→0 4
9
2
− 34! x2 + · · ·

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

2.7 Taylor series; Taylor polynomial and application 50

Exercise 2.7.1. Use Taylor series to find the following limits


sin x3 − x3 sin x tan−1 x − x
a) lim b) lim c) lim
x→0 x9 x→0 x x→0 x3

ex − 1 ln 1 + x − sin 2x
d) lim e) lim
x→0 x x→0 x

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

Chapter 3

Differential Calculus of Function of


Several Variables

3.1 Notations, examples, level curves and graphs

3.1.1 Notations and Examples

There are many familiar formulas in which a given variable depends on two or more other
variables. For example, the area A of a triangle depends on the base length b, and height h by
the formula A = 12 bh : the volume V of a rectangular box depends on the length l, the width
w, and the height h by the formula V = lwh and the arithmetic average x of n real numbers,
depends on those numbers by the formula
x1 + x2 + · · · + xn
x=
n
Thus, we say that

• A is a function of the two variables b and h

• V is a function of the three variables l, w, and h;

• x is a function of the n variables .

Definition 3.1.1. A function f of two variables, x and y, is a rule that assigns a unique real
number f (x, y) to each point (x, y) in some set D in the xy-plane. The set D is the domain of
f and its range is the set of values that f takes on, that is, {f (x, y)|(x, y) ∈ D}.

Example 3.1.1. Let f (x, y) = 3x2 y − 1. Find f (1, 4), f (0, 9), f (t2 , t), f (ab, 9b), and the
natural domain of f .

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.1 Notations, examples, level curves and graphs 52

Solution. By substitution

f (1, 4) = 3(1)2 4 − 1 = 5

f (0, 9) = 3(0)2 9 − 1 = −1
√ √
f (t2 , t) = 3(t2 )2 t − 1 = 3t4 t − 1
√ √
f (ab, 9b) = 3(ab)2 9b − 1 = 9a2 b2 b − 1

The domain of functions of several variables consists of all points in space or (in plane for
functions of two variables) for which the formula is meaningful.
p
Example 3.1.2. f (x, y) = 9 − 4x2 − y 2 The domain of f is the region in the xy plane
bounded by the ellipse 4x2 + y 2 = 9, because the square root is defined only for nonnegative
numbers.
p
Example 3.1.3. gz(x, y, ) = x2 + y 2 + z 2 The domain of g consists of all points (x, y, z) in
space, because x2 + y 2 + z 2 ≥ 0 for all (x, y, z).

The following are some of functions of several variables with there domains and representa-
tions

1. f (x, y) = xy, for x ≥ 0 and y ≥ 0, Area of rectangle.


2. f (x, y, z) = xyz, for x ≥ 0, y ≥ 0 and z ≥ 0, Area of rectangular parallelepiped.
3. f (x, y, z) = 2xy + 2yx + 2xz for surface area of a rectangular parallelepiped for x ≥ 0,
y ≥ 0 and z ≥ 0.

Combinations of Functions of Several Variables


The sum, product, and quotient of two functions f and g of several variables are defined as
follows.

• (f + g)(x, y) = f (x, y) + g(x, y)


• (f − g)(x, y) = f (x, y) − g(x, y)
• (f g)(x, y) = f (x, y)g(x, y)
 
• fg (x, y) = fg(x,y)
(x,y)

The formula for functions of thee variables is analogous. The domain of f + g, f − g , and f g
consists of all points simultaneously in the domain of f and in the domain of g, whereas the
domain of fg consists of all points simultaneously in the domain of f and in the domain of g at
which g does not assume the value 0. A function f of two variables x and y is a polynomial
function if it is a sum of function of the form cxm y n , where c is a number and m and n are
nonnegative integers. A rational function is, as with functions of one variable, the quotient
of two polynomials functions. Similar terminology is used for polynomial and rational function
of three variables.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.2 Limits and continuity 53

3.1.2 Level curves and graphs

The graph of a function of two variables is the collection of points (x, y, f (x, y)) for which
(x.y) is in the domain of f . It is customary to let z = f (x, y) the graph of f consists of all
points such that z = f (x, y).
In sketching the graph of the function f of two variables, it is often helpful to determine the
intersection of the graph of f with planes of the form z = c. We call each such intersection the
trace of the graph of f in the plane z = c. Thus the trace of the graph of f in the plane z=c
is the collection of points (x, y, c) such that c = f (x, y).

Definition 3.1.2. The set of points (x, y) in the xy plane such that f (x, y) = c is called a
level curve of f .

Example 3.1.4. Describe the graph of the function in an xyz coordinate system and the level
curves associated with them.

a. f (x, y) = 1 − x − 12 y.

Solution. By definition, the graph of the given function is the graph of the equation
z = 1−x− 21 y which is a plane. A triangular portion of the plane can be sketched by plotting
the intersection with the coordinates axes and joining them ((0, 0, 1), (0, 2, 1) and (1, 0, 0))
with line segment. For any value of c the level curve f (x, y) = c is the line with equation
x − 12 y = 1 − c. ◭

b. f (x, y) = 1 − x2 − y 2

Solution. By definition, the graph of the given function is the graph of the equation

z = 1 − x2 − y 2 . (3.1)

In this case the level curve f (x, y) = c is a circle if c ≤ 1. ◭

Exercise 3.1.1. Identify the type of the level curve f (x, y) = c if,

a) f (x, y) = x2 + 4y 2 ; c = 1, 4

b) f (x, y) = x2 − y; c = −2, 2

c) f (x, y) = x2 − y 2 ; c = −1, 0, 1

3.2 Limits and continuity

3.2.1 Limit

If D is a set of points in 2 space, then a point (a, b) is called an interior point of D if there is
some circular disk with positive radius, centered at (a, b), and containing only points in D.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.2 Limits and continuity 54

(0, 0, 1)
y
(0, 2, 0)
y
y
(1, 0, 0)

Figure 3.2:

Figure 3.1:

Definition 3.2.1. Let f be defined throughout a set containing a disk centered at (a, b) except
possibly at (a, b) itself, and let L be a number. Then L is the limit of f at (a, b) if for every
ǫ > 0 there is a δ > 0 such that
p
if 0 < (x − a)2 + (y − b)2 < δ, then |f (x, y) − L| < ǫ
In this case we write
lim f (x, y) = L
(x,y)→(a,b)

and say that lim f (x, y) exists. Similarly let f be defined throughout a set containing a
(x,y)→(a,b)
ball centered at (a, b, c) except possibly at (a, b, c) it self. Then L is the limit of f at (a, b, c)
if for every ǫ > 0 there exists a δ > 0 such that
p
if 0 < (x − a)2 + (y − b)2 + (z − c)2 < δ, then |f (x, y, z) − L| < ǫ

In this case we write


lim f (x, y, z) = L
(x,y,z)→(a,b,c)

and say that lim f (x, y, z) exists.


(x,y,z)→(a,b,c)

Example 3.2.1. Show that lim x = a and lim y=b


(x,y)→(a,b) (x,y)→(a,b)

Solution. Let ǫ > 0, Observe that


p p
(x − a)2 ≤ (x − a)2 + (y − b)2

Therefore, if we let δ = ε it follows that


p
0 < (x − a)2 + (y − b)2 < δ,

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.2 Limits and continuity 55

p
then |x − a| = (x − a)2 < δ < ε. This proves that lim x = a . The second limit is
(x,y)→(a,b)
established analogously. ◭

Exercise 3.2.1. Show that lim y = b.


(x,y,z)→(a,b,c)

Theorem 3.2.1. If lim f (x, y) and lim g(x, y) exist, then


(x,y)→(a,b) (x,y)→(a,b)

i. lim (f ± g) (x, y) = lim f (x, y) ± lim g(x, y).


(x,y)→(a,b) (x,y)→(a,b) (x,y)→(a,b)

ii. lim (f g) (x, y) = lim f (x, y) lim g(x, y)


(x,y)→(a,b) (x,y)→(a,b) (x,y)→(a,b)

  lim f (x,y)
f (x,y)→(a,b)
iii. lim g
(x, y) = lim g(x,y)
provided that lim g(x, y) 6= 0.
(x,y)→(a,b) (x,y)→(a,b) (x,y)→(a,b)

iv. The limit of a polynomial always exists and is found simply by substitution. lim Pn (x, y) =
(x,y)→(a,b)
Pn (a, b)
Example 3.2.2. Show that
x3 + y 3 7 x3 + y 3
lim 2 2
= and lim = 0.
(x,y)→(−1,2) x + y 5 (x,y)→(0,0) x2 + y 2

Solution. Since lim x = −1 and lim y = 2 The product formula yields lim x3 =
(x,y)→(−1,2) (x,y)→(−1,2) (x,y)→(−1,2)
3
−1 and lim y =8
(x,y)→(−1,2)
2
lim x = 1 and lim y 2 = 4 Hence the sum and the quotient formulas combined to
(x,y)→(−1,2) (x,y)→(−1,2)
yield
lim (x3 + y 3 )
x3 + y 3 (x,y)→(−1,2)
lim =
(x,y)→(−1,2) x2 + y 2 lim (x2 + y 2 )
(x,y)→(−1,2)

lim x3 + lim y3
(x,y)→(−1,2) (x,y)→(−1,2)
=
lim x2 + lim y2
(x,y)→(−1,2) (x,y)→(−1,2)
−1 + 8 7
= = .
1+4 5
We cannot use quite the same procedure to verify the second limit, since lim (x2 + y 2 ) = 0.
(x,y)→(0,0)
So, the quotient formula does not apply. To verify the second limit we will show first that
x3
lim =0 (3.2)
(x,y)→(0,0) x2 + y 2

For this limit we observe that


x3 x3
0≤ ≤ = |x|
x2 + y 2 x2
Since lim x = 0 by Example 3.2.1, a version of the squeezing theorem for functions of two
(x,y)→(0,0)
variable yields 3.2. A similar argument shows that
y3
lim =0 (3.3)
(x,y)→(0,0) x2 + y 2

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.2 Limits and continuity 56

we can use the sum formula to combine the limits in 3.2 and 3.3:
x3 + y 3 x3 y3
lim = lim + lim =0+0=0
(x,y)→(0,0) x2 + y 2 (x,y)→(0,0) x2 + y 2 (x,y)→(0,0) x2 + y 2


2x2 y+3xy
Exercise 3.2.2. 1. Evaluate lim 2 .
(x,y)→(2,1) 5xy +3y

2. Evaluate lim ln xy .
(x,y)→(e,1)

Disproving limits:

• For a limit to exist, the function must approach that limit for every possible path of (x, y)
approaching (a, b). Thus, it is usually very hard to prove a limit exists, and easier to show
a limit does not exist.

• So, if a function f (x, y) approaches L1 as (x, y) approaches (a, b) along a path P1 and
f (x, y) approaches L2 6= L1 as (x, y) approaches (a, b) along a different path P2 , then
lim f (x, y) does not exist.
(x,y)→(a,b)

• Some simple paths to try are the lines along x = a, y = b, or any other line through the
point.
y
Example 3.2.3. Show the following limit does not exist: lim
(x,y)→(1,0) x+y−1

Solution. The point (x, y) may approach (0, 0) in infinitely many paths and if the limit exits
it is independent of the path followed. Among these paths we can take the y-axes and the line
y = x − 1.

Case 1: when (x, y) may approach (0, 0) along y-axis


y y
lim = lim = 0.
(x,y)→(1,0) x + y − 1 y→0 y − 1

Case2: when (x, y) may approach (0, 0) along the line y = x − 1


y x−1 1
lim = lim = .
(x,y)→(1,0) x + y − 1 x→1 x + x − 1 − 1 2

Thus the limit is not the same in different paths and by uniqueness of limit lim y
(x,y)→(1,0) x+y−1
doesn’t exist. ◭
x2 −y 2
Example 3.2.4. Show that lim 2 2 does not exist.
(x,y)→(0,0) x +y

x2 −y 2
Solution. Let f (x, y) = x2 +y 2
. First let’s approach (0, 0) along the x-axis. Then y = 0 gives
x2
f (x, 0) = x2
= 1 for all x 6= 0,so

f (x, y) → 1 as (x, y) → (0, 0) along the x-axis.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.3 Partial derivatives; tangent lines, higher order partial derivatives. 57

y2
We now approach along the y-axis by putting x = 0. The f (0, y) = y2
= −1 for all y 6= 0, so

f (x, y) → −1 as (x, y) → (0, 0) along the y-axis.

Since f has two different limits a long two different lines, the given limit does not exist. ◭

Exercise 3.2.3. Show the following limit does not exist

1. lim xy
(x,y)→(0,0)
x2 +y 2

2. lim
xy 2
(x,y)→(0,0)
x2 +y 4

3.2.2 Continuity

Definition 3.2.2. a) A function f of two variables is continuous at (a, b) if

lim f (x, y) = f (a, b)


(x,y)→(a,b)

b) A function f of three variables is continuous at (a, b, c) if

lim f (x, y, z) = f (a, b, c)


(x,y,z)→(a,b,c)

c) A function of several variables is continuous if it is continuous at each point in its


domain.
x2 −y 2
Example 3.2.5. Where is the function f (x, y) = x2 +y 2
continuous?

Solution. The function f is discontinuous at (0, 0) because it is not defined there. Since f
is a rational function, it is continuous on its domain, which is the set D = {(x, y)|(x, y) 6=
(0, 0)}. ◭

Example 3.2.6. Let 


 x2 −y2 if (x, y) 6= (0, 0)
x2 +y 2
g(x) =
0 if (x, y) = (0, 0)
g is defined at (0, 0) but g is still discontinuous there because lim g(x, y) does not exist.
(x,y)→(0,0)

3.3 Partial derivatives; tangent lines, higher order partial deriva-


tives.

Definition 3.3.1. Let f be a function of two variables and let (x0 , y0 ) be in the domain of f .
The partial derivative of f with respect to x at (x0 , y0 ) is defined by
f (x0 + h, y0 ) − f (x0 , y0 )
fx (x0 , y0 ) = lim
h→0 h

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.3 Partial derivatives; tangent lines, higher order partial derivatives. 58

provided that the limit exists. The partial derivative of f with respect to y at is defined by
f (x0 , y0 + h) − f (x0 , y0 )
fy (x0 , y0 ) = lim
h→0 h
provided that the limit exists.

The functions fx and fy that arise through partial differentiation and are defined by
f (x + h, y) − f (x, y) f (x, y + h) − f (x, y)
fx (x, y) = lim and fy (x, y) = lim
h→0 h h→0 h
∂f ∂f
are called the partial derivatives of f and are frequently denoted ∂x
and ∂y
.
∂f ∂f
Example 3.3.1. Let f (x, y) = x2 + 3xy + y − 1. Find the values of ∂x
= fx and ∂y
= fy at
the point (4, −5).

∂f
Solution. To find ∂x
= fx , we treat y as a constant and differentiate f with respect to x.
∂f ∂f 2
= (x + 3xy + y − 1) = 2x + 3y
∂x ∂x
The value of ∂f∂x
at (4, −5) is 2(4) + 3(−5) = −7. To find ∂f
∂y
, we treat x as a constant and
differentiate f with respect to y:
∂f ∂ 2
= (x + 3xy + y − 1) = 3x + 1.
∂y ∂y
∂f
The value of ∂y
at (4, −5) is 3(4) + 1 = 13. ◭

Example 3.3.2. Find the partial derivatives of

f (x, y) = 2x3 y 2 + 2y + 4x.

Solution. Treating y as a constant and differentiating with respect to x, we obtain

fx (x, y) = 6x2 y 2 + 4.

Treating x as a constant and differentiating with respect to y, we obtain

fx (x, y) = 4x3 y + 2.

Exercise 3.3.1. a) Let f (x, y) = 24xy − 6x2 y . Find fx and fy at (1, 2).
∂z ∂z
b) Let z = x2 cosy. Find ∂x
and ∂y
.

Theorem 3.3.1. If f and g have partial derivatives, then

(f ± g)x = fx + gx and (f ± g)y = fy + gy


(f g)x = fx g + f gx and (f g)y = fy g + f gy
   
f fx g − f gx f fy g − f gy
= and =
g x g2 g y g2

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.3 Partial derivatives; tangent lines, higher order partial derivatives. 59

Example 3.3.3. Let


x3 y − xy 3
f (x, y) = .
x2 + y 2
Find fx and fy

Solution. Since f is a rational function, fx (x, y) and fy (x, y) are defined for all x and y such
that the denominator x2 + y 2 is not 0, that is, for all points except the origin. To find fx (x, y)
we hold y constant and use the quotient rule for partial derivatives to differentiate with respect
to x:
(3x2 y − y 3 )(x2 + y 2 ) − (x3 y − xy 3 )2x x4 y + 4x2 y 3 − y 5
fx (x, y) = =
(x2 + y 2 )2 (x2 + y 2 )2

3.3.1 A geometric interpretation of partial derivatives

When we hold y equal to a constant y = y0 , z = f (x, y) becomes the function z = f (x, y0 ) of x,


whose graph is the intersection of the surface z = f (x, y) with the vertical plane y = y0 (Figure
4). The x-derivative fx (x0 , y0 ) is the slope in the positive x-direction of the tangent line to this
curve at x = x0 . Similarly, when we hold x equal to a constant x0 , z = f (x, y) becomes the

function z = f (x0 , y) of y, whose graph is the intersection of the surface with the plane x = x0
(Figure 5), and the y-derivative fy (x0 , y0 ) is the slope in the positive y-direction of the tangent
line to this curve at y = y0 .

3.3.2 Higher order partial derivatives

If f is a function of two variables, then its partial derivatives fx and fy are also functions of
two variables, so we can consider their partial derivatives (fx )x , (fx )y , (fy )x and (fy )y which are
called the second partial derivatives of f . If z = f (x, y), we use the following notation:

∂ ∂f ∂2f ∂2z

1. (fx )x = fxx = f11 = ∂x ∂x
= ∂x2
= ∂x2
 
∂ ∂f ∂2f ∂2z
2. (fx )y = fxy = f12 = ∂x ∂y
= ∂x∂y
= ∂x∂y

∂ ∂f ∂2f ∂2z

3. (fy )x = fyx = f21 = ∂y ∂x
= ∂y∂x
= ∂y∂x
 
∂ ∂f ∂2f ∂2z
4. (fy )y = fyy = f22 = ∂y ∂y
= ∂y 2
= ∂y 2

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.3 Partial derivatives; tangent lines, higher order partial derivatives. 60

Example 3.3.4. Find all the second order derivatives for f (x, y) = cos(2x) − x2 e5y + 3y 2

Solution. We should first need the first order derivatives


fx (x, y) = −2 sin(2x) − 2xe5y
fy (x, y) = −5x2 e5y + 6y
Now, let’s get the second order derivatives.
fxx (x, y) = −4 cos(2x) − 2e5y
fxy (x, y) = −10xe5y
fyx (x, y) = −10xe5y
fyy (x, y) = −25x2 e5y + 6

Example 3.3.5. Let f (x, y) = sin(xy 2 ). Find all the second partial derivatives of f .

Solution. The first partial derivatives are given by


fx (x, y) = y 2 cos(xy 2 ) and fy (x, y) = 2xy cos(xy 2 )
We obtain the second partials by computing the partial derivatives of the first partials:
fxx (x, y) = −4y 4 sin(xy 2 )
fxy (x, y) = 2y cos(xy 2 ) − 2xy 3 sin(xy 2 )
fyx (x, y) = 2y cos(xy 2 ) − 2xy 3 sin(xy 2 )
fyy (x, y) = 2y cos(xy 2 ) − 4x2 y 2 sin(xy 2 ).


Exercise 3.3.2. Let (
x3 y−xy 3
x2 +y 2
for (x, y) 6= (0, 0)
f (x, y) =
0 for (x, y) = (0, 0)
Show that fxy (0, 0) 6= fyx (0, 0)
Theorem 3.3.2. Let f be a function of two variables, and assume that fxy (x, y) and fyx (x, y)
are continuous at (a, b). Then
fxy (a, b) = fyx (a, b).
2 y2
Exercise 3.3.3. 1. Verify the Theorem 3.3.2 for f (x, y) = xe−x .
2. Compute all four second-order partial derivatives for
(a) f (x, y) = 3x2 + 4y 2 − 2x2 y + 7xy 2 − 9
(b) f (x, y) = cos(x) sin(y) − exy
3. Compute
(a) fxyx and fxyxy for f (x, y) = cos(x) sin(y) − exy .

(b) fx , fxy and fxyz for f (x, y, z) = xyz + x2 y 3 z 4 .

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.4 The chain rule, implicit differentiation 61

3.4 The chain rule, implicit differentiation

3.4.1 Chain rule

The chain rule for functions of one variable provides a formula for differentiating the composite
of two functions f and g. If u = f (x) and y = g(u) = g(f (x)), the chain rule can be expressed
in the Leibniz notation as
dy dy du
=
dx du dx
Composites involving functions of several variables have their own versions of the chain rule,
which involve derivatives and partial derivatives.
Theorem 3.4.1 (The chain rule(case 1)). Suppose that z = f (x, y) is a differentiable function
of x and y,where x = g(t) and y = h(t) are both differentiable functions of t. Then z is a
differentiable function of t and
dz ∂z dx ∂z dy
= + (3.4)
dt ∂x dt ∂y dt
dz
Example 3.4.1. Let z = x2 ey , x = sin t, and y = t3 . Find dt
.

dz ∂z dx ∂z dy 3
Solution. Using 3.4, we find that dt
= ∂x dt
+ ∂y dt
= 2xey cos t+x2 ey (3t2 ) = 2(sin t)e(t ) cos t+
3
3(sin2 t)e(t ) t2 ◭

Example 3.4.2. The pressure P (in kilopascal), volume V (in liter), and temperature T (in
Kelvin) of a mole of an ideal gas are related by the equation P V = 8.31T . Find the rate at
which the pressure is changing when the temperature is 300k and increasing at a rate of 0.1k/s
and the volume is 100L and increasing at a rate of 0.2L/s.

Solution. If t represents the time elapsed in seconds, then at given instant we have T =
300, dT
dt
= 0.1, V = 100, dV
dt
= 0.2. Since P = 8.31T
V
the chain rule gives
dP ∂P dT ∂P dV 8.31 dT 8.31 dV 8.31 8.31(300)
= + = − 2 = (0.1) − (0.2) = −0.04155
dt ∂T dt ∂V dt V dt V dt 100 1002
The pressure is decreasing at a rate of about 0.042kpa/s. ◭

Theorem 3.4.2 (The chain rule(case 2)). Suppose that z = f (x, y) is a differentiable func-
tion of x and y, where x = g1 (s, t) and y = g2 (s, t). Then z = f (g1 (s, t), g2 (s, t)), and
∂z ∂z ∂x ∂z ∂y
= + (3.5)
∂s ∂x ∂s ∂y ∂s
∂z ∂z ∂x ∂z ∂y
= + (3.6)
∂t ∂x ∂t ∂y ∂t
∂z ∂z
Example 3.4.3. Let z = x ln y, x = u2 + v 2 , and y = u2 − v 2 . Find ∂u
and ∂v
.

Solution. Using 3.5, we find that


 
∂z ∂z ∂x ∂z ∂y x
= + = (ln y)(2u) + (2u)
∂u ∂x ∂u ∂y ∂u y
2 2 u2 + v 2
= 2u ln(u − v ) + 2u 2
u − v2

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.4 The chain rule, implicit differentiation 62

∂x s
∂s
∂z
∂z ∂z ∂x ∂z ∂y ∂x x
= + ∂x
∂s ∂x ∂s ∂y ∂s ∂t
t
z ∂y s
∂z ∂z ∂x ∂z ∂y ∂s
= + y
∂t ∂x ∂t ∂y ∂t ∂z
∂y ∂y
∂t t

Example 3.4.4. Suppose that w = f (x, y), x = g(u, v), y = h(u, v), u = j(t), and v = k(t).
Find a formula for dw
dt
.

Example 3.4.5. Let z = f (u − v, v − u). show that


∂z ∂z
+ =0
∂u ∂v
Example 3.4.6. Suppose the volume V of a canonical sand pile grows at a rate of 4 cubic
inches per second and the radius r of the circular base grows at rate of e−r inches per second.
Find the rate at which the height h of the cone is growing at the instant when the volume is 60
cubic inches and the radius is 6 inches.
dw
Exercise 3.4.1. 1. Let w = x cos yz 2 , x = sin t, y = t2 , and z = et . Find dt
∂u
2. If u = x4 y + y 2 z 2 ,where x = rset , y = rs2 e(−t) ,and z = r2 s sin t find the value of ∂s
when
r = 2, s = 1, t = 0.

3. Let w = x + y 2 z 3 , x = 1 + u2 v 2 , y = uv, and z = 3u. Find ∂w ∂u
and ∂w
∂v
.

3.4.2 Implicit differentiation

Through the chain rule we can more completely describe the process of implicit differentiation
which you have studied in applied I.
dw
Example 3.4.7. Suppose that w = f (x, y) and y = g(x). Find a formula for dt

Solution. The appropriate diagram is as follows: It follows that


dw ∂w ∂w dy
= + (3.7)
dx ∂x ∂y dx
(In this formula, ∂w
∂x
refers to the partial derivative of z = w(x, y) with respect to x, whereas
dw
dx
refers to the derivative of z = f (x, g(x)) with respect to x.) ◭

Now suppose that f is a function of two variables that has partial derivatives, and assume
that the equation f (x, y) = 0 defines a differentiable function y = g(x) of x, so that f (x, g(x)) =
0. If w = f (x, y), then by assumption,
dw d d
= (f (x, g(x))) = (0)
dx dx dx

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.4 The chain rule, implicit differentiation 63

Therefore by 3.7,
dw ∂w ∂w dy
0= = +
dx ∂x ∂y dx
∂w dy
Finally, if ∂y
6= 0, then by solving for dx
, we obtain
−∂w
dy ∂x
= ∂w
(3.8)
dx ∂y

dy
Example 3.4.8. Let x3 + y 3 = 2xy. Find dx
.

Solution. Let w = x3 + y 3 − 2xy. Then


∂w ∂w
= 3x2 − 2y and = 3y 2 − 2x
∂x ∂y
and 3.8 implies that
−∂w
dy ∂x −(3x2 − 2y)
= ∂w
=
dx ∂y
3y 2 − 2x

3.4.3 Differentiability of functions of several variables

Recall that in the case of a function of a single variable, a function f (x) is differentiable only if
it is continuous; but that continuity does not guarantee differentiability. Intuitively, continuity
of f (x) requires that its graph be a continuous curve; and differentiability requires also that
there is always a unique tangent vector to the graph of f (x). In other words, a function f (x)
is differentiable if and only if its graph is a smooth continuous curve with no sharp corners (a
sharp corner would be a place where there would be two possible tangent vectors).
If we try to extend this graphical picture of differentiability to functions of two or more vari-
ables, it would be natural to think of a differentiable function of several variables as one whose
graph is a smooth continuous surface, with no sharp peaks or folds. Because for such a surface
it would always be possible to associate a unique tangent plane at a given point. However,
”differentiability” in this sense turns out to be a much stronger condition than the mere exis-
tence of partial derivatives. For the existence of partial derivatives at a point x0 requires only
a smooth approach to the point f (x0 ) along the direction of the coordinate axes. We have seen
examples of functions that are discontinuous even though

lim (x, 0) = lim (0, y) both exist.


x→0 y→0

(x−y) 2
∂f ∂f
For example, the function f (x, y) = (x 2 +y 2 ) has this property, and in fact both ∂x and ∂y exist

and are continuous functions at the point (0, 0). With this sort of phenomenon in mind we give
the following definition of differentiability.
Theorem 3.4.3. Let f be a function having partial derivatives throughout a set containing a
disk centered at (x0 , y0 ). If fx and fy are continuous at (x0 , y0 ) then there are functions ε1 and
ε2 of two variables such that

f (x, y)−f (x0 , y0 ) = fx (x0 , y0 )(x−x0 )+fy (x0 , y0 )(y−y0 )+ε1 (x, y)(x−x0 )+ε2 (x, y)(y−y0 ) for (x, y) in D

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.5 Directional derivatives and gradients 64

where lim ε1 (x, y) = 0 and lim ε2 (x, y) = 0.


(x,y)→(x0 ,y0 ) (x,y)→(x0 ,y0 )

Definition 3.4.1. A function f of two variables is differentiable at (x0 , y0 ) if there exists a


disk D centered at (x0 , y0 ) and functions ε1 and ε2 of two variables such that

f (x, y)−f (x0 , y0 ) = fx (x0 , y0 )(x−x0 )+fy (x0 , y0 )(y−y0 )+ε1 (x, y)(x−x0 )+ε2 (x, y)(y−y0 ) for (x, y) in D
(3.9)
where lim ε1 (x, y) = 0 and lim ε2 (x, y) = 0.
(x,y)→(x0 ,y0 ) (x,y)→(x0 ,y0 )

3.5 Directional derivatives and gradients

Definition 3.5.1. Let f be a function defined on a set containing a disk D centered at (x0 , y0 )
and let u = a1 i + a2 j be a unit vector. Then the directional derivative of f at (x0 , y0 ) in the
direction of u, denoted Du f (x0 , y0 ), is defined by
f (x0 + ha1 , y0 + ha2 ) − f (x0 , y0 )
Du f (x0 , y0 ) = lim
h→0 h
provided this limit exists.

Theorem 3.5.1. Let f be differentiable at (x0 , y0 ). Then f has a directional derivative at


(x0 , y0 ) in every direction. Moreover, if u = a1 i + a2 j is a unit vector, then

Du f (x0 , y0 ) = fx (x0 , y0 )a1 + fy (x0 , y0 )a2 . (3.10)

Example 3.5.1. Let f (x, y) = 6 − 3x2 − y 2 , and let u = √1 i − √1 j. Find Du f (1, 2).
2 2

Solution. Notice that u is a unit vector. First we calculate the partial derivatives of f :

fx (x, y) = −6x and fy (x, y) = −2y.

Therefore by 3.11
   
1 1
Du f (1, 2) = fx (1, 2) √ + fy (1, 2) √
2 2

   
1 −1
= (−6) √ + (−4) √ =− 2
2 2

Remark 3.5.1. The directional derivative in the direction of an arbitrary non-zero vector a is
1
defined to be Du f (x0 , y0 ), where u = kak a

Example 3.5.2. Let f (x, y) = xy 2 and let a = i − 2j. Find the directional derivative at (−3, 1)
in the direction of a.
p √
Solution. In this case kak = 12 + (−2)2 = 5, so we will find Du f (−3, 1), where
1 1 2
u= a= √ i− √ j
kak 5 5

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.5 Directional derivatives and gradients 65

Since fx (x, y) = y 2 and fy (x, y) = 2xy it follows from 3.11 that


   
1 1
Du f (−3, 1) = fx (−3, 1) √ + fy (−3, 1) √
5 5

   
1 −2 13
= (1) √ + (−6) √ = 5.
5 5 5

Note 1. Let u = a1 i + a2 j + a3 k be a unit vector in space. The directional derivative


Du f (x0 , y0 , z0 ) is defined by
f (x0 + ha1 , y0 + ha2 , z0 + ha3 ) − f (x0 , y0 , z0 )
Du f (x0 , y0 , z0 ) = lim
h→0 h
provided that the limit exists. Moreover if f is differentiable at (x0 , y0 , z0 ). Then

Du f (x0 , y0 , z0 ) = fx (x0 , y0 , z0 )a1 + fy (x0 , y0 , z0 )a2 + fz (x0 , y0 , z0 )a3 . (3.11)

3.5.1 The Gradient

Definition 3.5.2. a. Let f be a function of two variables that has partial derivatives at
(x0 , y0 ). Then the gradient of f at (x0 , y0 ), denoted grad f (x0 , y0 ) or ∇f (x0 , y0 ) is de-
fined by
grad f (x0 , y0 ) = ∇f (x0 , y0 ) = fx (x0 , y0 )i + fy (x0 , y0 )j

b. Let f be a function of three variables that has partial derivatives at (x0 , y0 , z0 ). Then the
gradient of f at (x0 , y0 , z0 ), which is denoted grad (x0 , y0 , z0 ) or ∇f (x0 , y0 , z0 ) is defined
by

grad f (x0 , y0 , z0 ) = ∇f (x0 , y0 , z0 ) = fx (x0 , y0 , z0 )i + fy (x0 , y0 , z0 )j + fz (x0 , y0 , z0 )k

Example 3.5.3. Find the gradient of the function f (x, y) = x2 y 3 − 4y at the point (2, −1).

Solution. We first compute the partial derivatives at (2, −1).

fx (x, y) = 2xy 2 and fy (x, y) = 3x2 y 2 − 4.

Hence fx (2, −1) = −4 and fy (2, −1) = 8. Therefore, ∇f (2, −1) = −4i + 8j. ◭

Note 2.
Du f (x, y) = ∇f (x, y) · u (3.12)
Example 3.5.4. If f (x, y, z) = x sin yz,

a. Find the gradient of f and

b. Find the directional derivative of f at (1, 3, 0) in the direction of v = i + 2j − k.

Solution. a. The gradient of f is

∇f (x, y, z) = hfx (x, y, z), fy (x, y, z), fz (x, y, z)i = hsin yz, xz cos yz, xy cos yzi = h0, 0, 3i

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.6 Tangent planes and Total differential 66

b. The unit vector in the direction of v = i + 2j − k is u = √1 i + √2 j − √1 k. Therefore,


6 6 6
  r
1 2 1 −1 3
Du f (1, 3, 0) = ∇f (1, 3, 0) · u = 3k · √ i + √ j − √ k = 3 √ =−
6 6 6 6 2

Theorem 3.5.2. Suppose f is differentiable function of two or three variables. The maximum
value of the directional derivative Du f (x) is |∇f (x)| and it occurs when u has the same direction
as the gradient vector ∇f (x).

Proof. We know that Du f = ∇f · u. Thus, Du f = ∇f · u = |∇f |kuk cos θ = k∇f k cos θ where
θ is the angle between ∇f and u. The maximum value of cos θ is 1 and this occurs when θ = 0.
Therefore the maximum value of Du f is k∇f k and it occurs when θ = 0, that is, when u has
the same direction ∇f .

Example 3.5.5. a. If f (x, y) = xey , find the rate of change of f at the point P (2, 0) in the
direction from P to Q(1/2, 2).

b. In what direction does f have the maximum rate of change? What is this maximum rate
of change?

Solution. a. We first compute the gradient vector:

∇f (x, y) = hfx , fy i = hey , xey i ⇒ f (2, 0) = h1, 2i.

The unit vector in the direction of P~Q = ( −3 2


, 2) is a u = h −3 , 4 i, so the rate of change of
5 5
f in the direction from P to Q is
   
−3 4 −3 4
Du f (2, 0) = ∇f (2, 0) · u = h1, 2i · h , i = 1 +2 =1
5 5 5 5

b. By the above Theorem f increases fastest in the direction of the gradient vector ∇f (2, 0) =

h1, 2i. The maximum rate of change is h∇f (2, 0)i = kh1, 2ik = 5.

3.6 Tangent planes and Total differential

3.6.1 Tangent Planes

Example 3.6.1. Assuming that the curve x2 − xy + 3y 2 = 5 is smooth, find a unit vector that
is perpendicular to the curve at (1, −1).

Definition 3.6.1. Let f be differentiable at a point (x0 , y0 , z0 ) on a level surface S of f . If


gradf (x0 , y0 , z0 ) 6= 0, then the plane through (x0 , y0 , z0 ) whose normal is gradf (x0 , y0 , z0 ) is the
plane tangent to S at (x0 , y0 , z0 ), and gradf (x0 , y0 , z0 ) is normal to S

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.6 Tangent planes and Total differential 67

The Gradient as a Normal Vector


Suppose a smooth curve C is a level curve of a func- grad f (x0 , y0 )
tion f that is differentiable at a point (x0 , y0 ) on C. as y C
a point (x, y) moves along C, the value f (x, y) is con- f (x, y) = c
u
stant by hypothesis and hence does not change. This
suggests that at (x0 , y0 ) the rate of change of f in the
(x0 , y0 )
direction of a unit vector u tangent to C is 0, that is,
Du f (x0 , y0 ) = 0 (Figure 3.3). However, 3.12 implies
that Du f (x0 , y0 ) = 0 for any unit vector u perpendic-
ular to gradf (x0 , y0 ). combining these results, we can
reasonably conjecture that gradf (x0 , y0 ) is perpendicu-
lar, or equivalently, normal to C at (x0 , y0 )(Figure 3.3)

Theorem 3.6.1. Let f be differentiable at (x0 , y0 ),


and let f (x0 , y0 ) = c. Also let C be a level curve Du f (x0 , y0 ) = 0
f (x, y) = c that passes through (x0 , y0 ). If C is smooth
and gradf (x0 , y0 ) 6= 0, then gradf (x0 , y0 ) is normal to Figure 3.3:
C at (x0 , y0 ).

tangent plane gradf (x0 , y0 , z0 )

Now let f be a function of three variables that is differen-


tiable at a point (x0 , y0 , z0 ), and let f (x0 , y0 , z0 ) = c. If
C is any smooth curve on the level surface f (x, y, z) = c
and if C passes through (x0 , y0 , z0 ), then an argument
tangent vector
similar to the one used in above Theorem shows that if
gradf (x0 , y0 , z0 ) 6= 0, then gradf (x0 , y0 , z0 ) is perpen-
dicular to the tangent vector of C at x0 , y0 , z0 (Figure
3.4). Because gradf (x0 , y0 , z0 ) is perpendicular to the
tangent vector to any such curve C at (x0 , y0 , z0 ), all
such tangents lie in a single plane called a tangent
plane. (x0 , y0 , z0 )
f (x, y, z) = c

Figure 3.4:

Since
gradf (x0 , y0 , z0 ) = fx (x0 , y0 , z0 )~i + fy (x0 , y0 , z0 )~j + fz (x0 , y0 , z0 )~k
and Since gradf (x0 , y0 , z0 ) is normal to the tangent plane at (x0 , y0 , z0 ), an equation of tangent
plane is

fx (x0 , y0 , z0 )(x − x0 ) + fy (x0 , y0 , z0 )(y − y0 ) + fz (x0 , y0 , z0 )(z − z0 ) = 0

Example 3.6.2. Find an equation of the plane tangent to the sphere x2 + y 2 + z 2 = 4 at the

point (−1, 1, 2).

Solution. The sphere is the level surface f (x, y, z) = 4, where f (x, y, z) = x2 + y 2 + z 2 . The

partials of f at (−1, 1, 2) are given by
√ √ √ √
fx (−1, 1, 2) = −2, fy (−1, 1, 2) = 2, fz (−1, 1, 2) = 2 2

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.6 Tangent planes and Total differential 68


Therefore an equation of the plane tangent at (−1, 1, 2) is
√ √
−2(x − (−1)) + 2(y − 1) + 2 2(z − 2) = 0

or equivalently,

−x + y + 2z = 4

Suppose f is a function of two variables that is differentiable at (x, y0 ), and let

g(x, y, z) = f (x, y) − z

Then the graph of f is the level surface g(x, y, z) = 0. Accordingly, we define the plane tangent
to the graph of f at (x0 , y0 , f (x0 , y0 )) to be the plane tangent to the level surface g(x, y, z) at
(x0 , y0 , f (x0 , y0 )). Since

gradg(x0 , y, z0 ) = fx (x0 , y0 )~i + fy (x0 , y0 )~j − ~k

an equation of the tangent plane of f at (x0 , y0 , f (x0 , y0 )) is

fx (x0 , y0 )(x − x0 ) + fy (x0 , y0 )(y − y0 ) − [z − f (x0 , y0 )] = 0 (3.13)

or equivalently,

z = f (x0 , y0 ) + fx (x0 , y0 )(x − x0 ) + fy (x0 , y0 )(y − y0 ) = 0 (3.14)

and the vector gradg(x0 , y, z0 ) = fx (x0 , y0 )~i + fy (x0 , y0 )~j − ~k is said to be normal to the graph
of f at (x0 , y0 , f (x0 , y0 )).

Example 3.6.3. Find the tangent plane to the elliptic paraboloid z = 2x2 + y 2 at the point
(1, 1, 3).

Solution. Let f (x, y) = 2x2 + y 2 . Then fx (x, y) = 4x and fy (x, y) = 2y. which implies

fx (1, 1) = 4, fy (1, 1) = 2

Then (by definition) the equation of the tangent plane at (1, 1, 3) is

z − 3 = 4(x − 1) + 2(y − 1) ⇒ z = 4x + 2y − 3

Example 3.6.4. Let f (x, y) = 6−3x2 −y 2 . Find a vector normal to the graph of f at (1, 2, −1),
and find an equation of the plane tangent to the graph of f at (1, 2, −1).

Differentials

If f is a function of two variables, we can replace (x0 , y0 ) by any point (x, y) in the domain of
f at which f is differentiable and the linear approximation is transformed into

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.6 Tangent planes and Total differential 69

f (x + h, y + k) − f (x, y) ≈ fx (x, y)h + fy (x, y)k (3.15)

The number fx (x, y)h + fy (x, y)k on the right side of (3.15) is usually called the differential
(or total differential) of f (at (x, y) with increments h and k) and is denoted df . Thus

f = fx (x, y)h + fy (x, y)k. (3.16)

Of course, df depends on x, y, h and k, even though they are not indicated in the notation df .
If g1 (x, y) = x and g2 (x, y) = y, then the differential dg1 is denoted by dx, and the differential
dg2 is denoted by dy. Since

(g1 )x (x, y) = 1, (g1 )y (x, y) = 0, (g2 )x (x, y) = 0, (g2 )y (x, y) = 1

We have dx = dg1 = 1.h + 0.k = h and dy = dg2 = 0.h + 1.k = k Therefore we can write (3.16)
as
∂f ∂f
df = fx (x, y)dx + fy (x, y)dy or df = dx + dy
∂x ∂y
Example 3.6.5. Let f (x, y) = xy 2 + y sin x. Find df .

∂f ∂f
Solution. Since ∂x
= y 2 + y cos x and ∂x
= 2xy + sin x. Thus df = (y 2 + y cos x)dx + (2xy +
sin x)dy ◭

Note 3. If f is a function of three variables that is differentiable at (x0 , y0 , z0 ), then the dif-
ferential df is defined by df = fx (x, y, z)h + fy (x, y, z)k + fz (x, y, zl The more usual form for
df is

df = fx (x, y, z)dx + fy (x, y, z) + fz (x, y, z)dz


∂f ∂f ∂f
df = dx + dy + dz
∂x ∂y ∂z
Example 3.6.6. Let f (x, y, z) = x2 ln(y − z). Find df .

∂f x2 ∂f −x2 x2
Solution. Since ∂x
= 2x ln(y − z), ∂f
∂y
= ,
y−z ∂z
= y−z
= z−y
. Thus, df = 2x ln(y − z)dx +
x2 x2
y−z
dy + z−y
dz. ◭

Example 3.6.7. Find the tangent plane to the elliptic paraboloid z = 2x2 + y 2 at the point
(1, 1, 3).

Solution. Let f (x, y) = 2x2 + y 2 . Then fx (x, y) = 4x and fy (x, y) = 2y. which implies

fx (1, 1) = 4, fy (1, 1) = 2

Then (by definition) the equation of the tangent plane at (1, 1, 3) is

z − 3 = 4(x − 1) + 2(y − 1) ⇒ z = 4x + 2y − 3

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.7 Tangent plane approximations of values of a function 70

3.7 Tangent plane approximations of values of a function

In example 3.6.7 the tangent line L(x, y) = 4x + 2y − 3 is a good approximation to f (x, y)


when (x, y) is near (1, 1). The function L is called the linearization of f at (1, 1) and the
approximation
f (x, y) ≈ 4x + 2y − 3
is called the linear approximation or tangent plane approximation of f at (1, 1).

Definition 3.7.1. An equation of the tangent plane to the graph of a function f of two variables
at the point (a, b, f (a, b)) is z = f (a, b)+fx (a, b)(x−a)+fy (a, b)(y −b) is called the linearization
of f at (a, b) and the approximation

f (x, y) ≈ f (a, b) + fx (a, b)(x − a) + fy (a, b)(y − b)

is called the linear approximation or the tangent plane approximation of f at (a, b).

Theorem 3.7.1. If the partial derivatives fx and fy exist near (a, b) and are continuous at
(a, b), then f is differentiable at (a, b).

Example 3.7.1. Show that f (x, y) = xex y is differentiable at (1, 0) and find its linearization
there. Then use it to approximate f (1.1, −0.1).

Solution. The partial derivatives are fx (x, y) = ex y + xyex y, fy (x, y) = x2 ex y, fx (1, 0) = 1


and fy (1, 0) = 1. Both fx and fy are continuous functions, so f is differentiable by the above
theorem. The linearization is L(x, y) = f (1, 0) + fx (1, 0)(x − 1) + fy (1, 0)(y − 0) = x + y. The
corresponding linear approximation is xex y ≈ x + y.So,

f (1.1, −0.1) ≈ 1.1 − 0.1 = 1

Compare this with the actual value of f (1.1, −0.1) = (1.1e)( − 0.1) ≈ 0.98542. ◭

3.8 Relative extrema of functions of two variables

Definition 3.8.1. Let f be a function of two variables, and let R be a set contained in the
domain of f . Then f has an absolute maximum (respectively, an absolute minimum value) on R
at (x0 , y0 ) if f (x, y) ≤ f (x0 , y0 )(respectively, f (x, y) ≥ f (x0 , y0 )) for all (x, y) in R. If R is the
domain of f , we say that f has an absolute maximum value (respectively, an absolute minimum
value)at (x0 , y0 ). Furthermore, f has a relative maximum value (respectively, a relative value)
at (x0 , y0 ) if there is a disk D centered at (x0 , y0 ) and contained in the domain of f such that
f (x, y) ≤ f (x0 , y0 ) (respectively, f (x, y) ≥ f (x0 , y0 )) for all (x, y) in D.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.8 Relative extrema of functions of two variables 71

Relative Maximums

Relative Minimums

Theorem 3.8.1. Let f have a relative extreme value at (x0 , y0 ). If f has partial derivatives at
(x0 , y0 ), then fx (x0 , y0 ) = fy (x0 , y0 ) = 0.

Definition 3.8.2. The point (a, b) is a critical point (or a stationary point) of f (x, y).
Provided one of the following is true,

1. ∇f (a, b) = 0 (this is equivalent to saying that fx (a, b) = 0 and fy (a, b) = 0 )

2. fx (a, b) and /or fy (a, b) doesn’t exist.

Example 3.8.1. Let f (x, y) = 3 − x2 + 2x − y 2 − 4y. Find all critical points of f .

Solution. The partial derivatives of f exist at every point in the domain of f , so relative
extreme values can occur only at points at which both partial derivatives are 0. The partial
derivatives are
fx (x, y) = −2x + 2 and fy (x, y) = −2y − 4.
Therefore fx (x, y) = 0 only if x = 1,and fy (x, y) = 0 only if y = −2. This means that
fx (x, y) = fy (x, y) only if (x, y) = (1, −2), and thus (1, −2) is the only critical point of f . ◭
p
Example 3.8.2. Let f (x, y) = (x2 + y 2 ). Determine all critical points and all relative ex-
treme values of f .

x y
Solution. We find that fx (x, y) = √ and fy (x, y) = x2 +y 2
. Since the partial derivatives
x2 +y 2
exist at all points except the origin, and since the origin is in the domain of f , the origin is a
critical point of f . Because fx (x, y) = 0 only if x = 0 and fy (x, y) = 0 only if y = 0, there is no
point (x, y) such that fx (x, y) = fy (x, y) = 0. Consequently (0, 0) is the only critical point of f .
Since f (0, 0) = 0 and f (x, y) ≤ 0 for all (x, y), it follows that 0 is the only (relative) minimum
value of f , and there is no relative maximum value. ◭

Theorem 3.8.2. If the point (a, b) is a relative extrema of the function f (x, y), then (a, b) is
also a critical point of f (x, y) and ∇f (a, b) = 0.

Note that this does NOT say that all critical points are relative extrema. It only says that
relative extrema will be critical points of the function. To see this let’s consider the function

f (x, y) = xy.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.8 Relative extrema of functions of two variables 72

The two first order partial derivatives are,

fx (x, y) = y, fy (x, y) = x

The only point that will make both of these derivatives zero at the same time is (0, 0) and
also (0, 0) is a critical point for the function. But if we start at the origin and move into either
of the quadrants where both x and y are the same sign the function increases. However, if we
start at the origin and move into either of the quadrants where x and y have the opposite sign
then the function decreases. In other words, no matter what region you take about the origin
there will be points larger than f (0, 0) = 0 and points smaller than f (0, 0) = 0. Therefore,
there is no way that (0, 0) can be a relative extrema. Critical points that exhibit this kind of
behaviour are called saddle points.

Exercise 3.8.1. 1. Let f (x, y) = y 2 − x2 . Show that the origin is the only critical point but
that f (0, 0) is not a relative extreme value of f .

2. Let f (x, y) = sin x + sin y for 0 ≤ x ≤ π2 , 0 ≤ y ≤ π2 find all critical points and determine
whether each critical point yields a relative maximum value, a relative minimum value or
a saddle point

Definition 3.8.3. If f is a function for which fx (x0 , y0 ) = fy (x0 , y0 ) = 0, we say that f has
a saddle point at (x0 , y0 ) if there is a disk centered at (x0 , y0 ) such that the following condition
holds: f assume its maximum value on one diameter of the disk only at (x0 , y0 ), and assumes its
minimum value on another diameter of the disk only at (x0 , y0 ). For example, if f (x, y) = y 2 −x2
then f has a saddle point at the origin.

Theorem 3.8.3 (Second Partial Test). Assume f has a critical point at (x0 , y0 ) and that f
has continuous second partial derivatives in a disk centered at (x0 , y0 ). Let

D(x0 , y0 ) = fxx (x0 , y0 )fyy (x0 , y0 ) − [fxy (x0 , y0 )]2

a. If D(x0 , y0 ) > 0 and fxx (x0 , y0 ) < 0 (or fyy (x0 , y0 ) < 0), then f has a relative maximum
value at (x0 , y0 ).

b. If D(x0 , y0 ) > 0 and fxx (x0 , y0 ) > 0 (or fyy (x0 , y0 ) > 0), then f has a relative maximum
value at (x0 , y0 ).

c. If D(x0 , y0 ) < 0 , then f has a saddle point at (x0 , y0 ).

Finally, if D(x0 , y0 ) = 0, then f may or may not have a relative extreme value at (x0 , y0 ). The
expression D(x0 , y0 ) in the second partial test is called the discriminant of f at (x0 , y0 ). It can
also be given in determinant form:

fxx (x0 , y0 ) fxy (x0 , y0 )


D(x0 , y0 ) =
fxy (x0 , y0 ) fyy (x0 , y0 )

Definition 3.8.4. A critical point (x0 , y0 ) is said to be degenerate if D(x0 , y0 ) = 0.

Example 3.8.3. Let f (x, y) = x2 − 2xy + 31 y 3 − 3y Using the second partial derivative test,
determine at which points f has relative extreme values and at which points f has saddle points.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.9 Largest and smallest values of a function on a given set 73

Solution. We find that fx (x, y) = 2x−2y and fy (x, y) = −2x+y 2 −3 Observe that fx (x, y) = 0
if x = y and fy (x, y) = 0 if −2x + y 2 − 3 = 0. Thus (x, y) is a critical point if x = y
and −2x + y 2 − 3 = 0. By substituting y for x we can transform the second equation into
y 2 −2y−3 = 0. The two solutions of this equation are y = 3 and y = −1. Thus the critical points
of f are (3, 3) and (−1, 1). For the second partials of f we find that fxx (x, y) = 2, fyy (x, y) = 2y,
and fxy (x, y) = −2. Therefore

D(3, 3) = fxx (3, 3)fyy (3, 3) − [fxy (3, 3)]2 = (2)(6) − 22 = 8 > 0

and

D(−1, 1) = fxx (−1, −1)fyy (−1, −1) − [fxy (−1, −1)]2 = (2)(−2) − (−2)2 = −8 < 0.

Since D(3, 3) > 0 and fxx (3, 3) = 2 > 0, f has a relative minimum value at (3, 3). Since
D(−1, −1) < 0, f has a saddle point at (−1, −1).

3.9 Largest and smallest values of a function on a given set

Definition 3.9.1. 1. A region in R2 is called closed if it includes its boundary. A region is


called open if it doesn’t include any of its boundary points.

2. A region in R2 is called bounded if it can be completely contained in a disk. In other


words, a region will be bounded if it is finite.

Theorem 3.9.1 (Extreme value theorem). If f is continuous in some closed, bounded set D in
R2 , then there are points in D, (x1 , y1 ) and (x2 , y2 ) so that f (x1 , y1 ) is the absolute maximum
and (x2 , y2 ) is the absolute minimum of the function in D.

Steps to find extreme value

1. Find all the critical points of f that lie in the region D and compute the values of f at
these points.

2. Find the extreme values of f on the boundary.

3. The largest and smallest values found in the first two steps are the absolute minimum and
the absolute maximum of the function.

Solution. By the Extreme value Theorem f has extreme values on R. Since fx (x, y) = y − 2x
and fy (x, y) = x it follows that fx (x, y) = 0 if y = 2x and fy (x, y) = 0 if x = 0. Thus the
only critical point of f is (0, 0), which happens to be a boundary point of R. Therefore the
extreme values of f on R must occur on the boundary of R, which is composed of the four line
segments l1 , l2 , l3 and l4 . On l1 , x = 0 and 0 ≤ y ≤ 1 and since f (0, y) = 0, the maximum and
the minimum values of f on l1 are both 0. On l2 , y = 0 and 0 ≤ x ≤ 1, and since f (x, 0) = −x2

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.10 Extreme values under constraint conditions: Lagrange’s multiplier 74

y
Let f (x, y) = xy−x2 , and let R be
l4
Example 3.9.1. the square region shown in Figure 1
below. Find extreme values of f l1 l3
0 l2 1 x

the maximum value of f on l2 is 0 and the minimum value is −1. On l3 , x = 1 and 0 ≤ y ≤ 1,


and since f (1, y) = y − 1 the maximum value of f On l3 is 0 and the minimum value is -1. On
l4 , y = 1 and 0 ≤ x ≤ 1, and since f (x, 1) = x − x2 the maximum value of f on l4 is 1/4 and
the minimum value is 0. By comparing the extreme values of f on l1 , l2 , l3 and l4 , we conclude
that the maximum value of f on R is 1/4 and the minimum value is -1. ◭

3.10 Extreme values under constraint conditions: Lagrange’s


multiplier

Let us consider the problem of finding an extreme value of a function f of two variables
subject to a constraint of the form g(x, y) = c; that is we seek an extreme value of f on the
level curve g(x, y) = c (rather than on the entire domain of f ). If f has an extreme value on
the level curve at the point (x0 , y0 ), then under certain conditions there exists a number λ
such that

∇f (x0 , y0 ) = λ∇g(x0 , y0 ).

Theorem 3.10.1. Let f and g be differentiable at (x0 , y0 ). Let C be the level curve g(x, y) = k
that contains (x0 , y0 ), Assume that C is smooth, and that (x0 , y0 ) is not an end point of the
curve. If ∇f (x0 , y0 ) and if f has an extreme value on C at (x0 , y0 ), then there is a number λ
such that
∇f (x0 , y0 ) = λ∇g(x0 , y0 )
is called a Lagrange multiplier for f and g.

The method of determining extreme values by means of Lagrange multipliers proceed as


follows:

1. Assume that f has an extreme value on the level curve g(x, y) = k

2. Solve the equations Constraint g(x, y) = k


(
fx (x, y) = λgx (x, y)
∇f (x, y) = λ∇g(x, y) =
fy (x, y) = λgy (x, y)

3. Calculate the values of f at each point (x, y) that arises in step 2, and at each end point
(if any) of the curve. If f has maximum value in the level curve g(x, y) = k it would be

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.10 Extreme values under constraint conditions: Lagrange’s multiplier 75

the largest of the values computed; if f has a minimum value on the level curve, it will be
the smallest of the values computed.

Example 3.10.1. Let f (x, y) = x2 +y 3 , Find the extreme values of f on the ellipse x2 +2y 2 = 1.
Solution: Let g(x, y) = x2 + 2y 2 So the constraint is g(x, y) = x2 + 2y 2 = 1. Since ∇f (x, y) =
2xi + 12y 2 j and ∇g(x, y) = 2xi + 4yj The equation we will use to find x and y are constraint
x2 + 2y 2 = 1 (
2x = λ2x
∇f (x0 , y0 ) = λ∇g(x0 , y0 ) = 2
12y = λ4y
By (3) either x = 0 or λ = 1 . If x = 0, then the (2) implies 2y 2 = 1 so that y = √12 or
y = − √12 . If λ = 1 then (3) becomes 12y 2 = 4y, which means that y = 0 or y = 13 . By (2) If
 2 √
y = 0, then x + 2(0) = 1 , so x = 1 or x = −1. If y = 3 , then x + 2 31
2 2 1 2
= 1 , so x = 37

7
or x = − 3
Thus the only possible extreme values of occur at (0, √12 ), (0, − √12 ), (1, 0), (−1, 0),
√ √ √ √ √
( 37 , 13 ), and ( −3 7 , 31 ). f (0, √12 ) = 2, f (1, 0) = 1 = f (−1, 0), f ( 37 , 31 ) = ( −3 7 , 31 ) = 25
27
. We
√ 1

conclude that the maximum value 2 of f occurs at (0, √2 ) and the minimum value of f = − 2
occurs at (0, − √12 ).

The Lagrange Method for Functions of Three variables

Next we consider the problem of finding extreme values of a function of three variables
subject to a constraint of the form g(x, y, z) = c. By an argument similar to that used for
functions of two variables, It is possible to show that if f has such an extreme value at
(x0 , y0 , z0 ), then ∇f (x0 , y0 , z0 ) and ∇g(x0 , y0 , z0 ), if not 0, are both normal to the level surface
g(x, y, z) = c at (x0 , y0 , z0 ), and hence are parallel to each other. Thus there is a number λ,
again called a lagrange multiplier, such that

∇f (x0 , y0 , z0 ) = λ∇g(x0 , y0 , z0 ).

To find the extreme values of f subject to the constraint g(x, y, z) = c, we follow the same
approach as in steps 1-3 for functions of two variables:

1. Assume that f has an extreme value on the level surface g(x, y, z) = c

2. Solve the equations

Constraint g(x, y, z) = c

 fx (x, y, z) = λgx (x, y, z)

∇f (x, y, z) = λ∇g(x, y, z) fy (x, y, z) = λgy (x, y, z)

 f (x, y, z) = λg (x, y, z)
z z

3. Calculate the values of f at each point (x, y, z) that arises in step 2, and at each end point
(if any) of the curve. If f has maximum value in the level curve g(x, y, z) = c it would be
the largest of the values computed; if f has a minimum value on the level curve, it will be
the smallest of the values computed.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

3.10 Extreme values under constraint conditions: Lagrange’s multiplier 76

Example 3.10.2. Let V (x, y, z) = xyz for x ≥ 0, y ≥ 0, and z ≥ 0. Find the maximum value
of V subject to the constraint 2x + 2y + z = 108.

Solution. Let g(x, y, z) = 2x + 2y + z, so the constraint is g(x, y, z) = 2x + 2y + z = 108.


Because
∇V (x, y, z) = yz~i + xz~j + xy~k and ∇g(x, y, z) = 2~i + 2~j + ~k
the equations we will use to find x, y and z are

Constraint 2x + 2y + z = 108 (3.17)



 yz = 2λ

∇f (x, y, z) = λ∇g(x, y, z) xz = 2λ (3.18)

 xy = λ

First we solve for λ in terms of x, y and z in


yz xz
λ= = = xy (3.19)
2 2
Since V (x, y, z) = 0 if x, y or z is 0, and since 0 is obviously not the maximum value of V
subject to (3.17), we can assume that x, y and z are different from 0. Then (3.19) tells us that
x = y and z = 2y. Substituting for x and z in (3.17) yields

2y + 2y + 2y = 108, so that y = 18

Thus x = 18 and z = 36, and therefore V (18, 18, 36) is the only possible extreme value of V
subject to the constraint. Since we are assuming that V has a maximum value subject to the
constraint, we conclude that V (18, 18, 36) = 11, 664 is that value. ◭

Exercise 3.10.1. 1. Let f (x, y) = 3x2 + 2y 2 − 4y + 1. Find the extreme values of f on the
disk x2 + y 2 ≤ 16.

2. Find the minimum distance from a point on the surface xy + 2xz = 5 5 to the origin.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

Chapter 4

Multiple Integrals

4.1 Double integrals and their evaluation by iterated integrals

Recall that the definition of definite integral of function of one variable.

Definition 4.1.1. For any function f defined on the interval [a, b] and kP k (the norm of the
partition) defined as the maximum of all the intervals on [a, b] (i.e., kP k = max{∆xi }), the
definite integral of f on [a, b] is:
Z b n
X
f (x) dx = lim f (ci )∆xi ,
a kP k→0
i=1

Provided the limit exists and is the same for all values of the evaluation points ci ∈ [xi−1 , xi ]
for i = 1, 2, . . . , n. In this case, we say f is integrable on [a, b].

In this section we want to integrate a function of two variables, f (x, y). With functions of one
variable we integrated over an interval ( i.e. a one-dimensional space) and so it makes some
sense then that when integrating a function of two variables we will integrate over a region of
R2 (two dimensional space).
We will start out by assuming that the region in R2 is a rectangle which we will denote as
follows,

R = [a, b] × [c, d]

This means that the ranges for x and y are a ≤ x ≤ b, c ≤ y ≤ d. Also, we will initially
assume that f (x, y) ≥ 0 although this doesn’t really have to be the case. Let’s start out with
the graph of the surface S give by graphing f (x, y) over the rectangle R.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.1 Double integrals and their evaluation by iterated integrals 78

Now, just like with functions of one variable let’s not worry about integrals quite yet. Let’s
first ask what the volume of the region under S (and above the xy-plane of course) is. We will
first approximate the volume much as we approximated the area above. We will first divide
up a ≤ x ≤ b into n subintervals and divide up c ≤ y ≤ d into m subintervals. This will divide
up R into a series of smaller rectangles and from each of these we will choose a point (x∗i , yj∗ ).
Here is a sketch of this set up.

Now, over each of these smaller rectangles we will construct a box whose height is given by
f (x∗i , yj∗ ). Here is a sketch of that.

Each of the rectangles has a base area of ∆A and a height of f (xi ∗ , yj ∗ ) so the volume of each
of these boxes is f (xi ∗ , yj ∗ )∆A . The volume under the surface S is then approximately,
n X
X m
V = f (xi ∗ , yj ∗ )∆A
i=1 j=1

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.2 Iterated integrated 79

We will have a double sum since we will need to add up volumes in both the x and y
directions.
To get a better estimation of the volume we will take n and m larger and larger and to get the
exact volume we will need to take the limit as both n and m go to infinity. In other words,
n X
X m
V = lim f (xi ∗ , yj ∗ )∆A
n,m→∞
i=1 j=1

Now, this should look familiar. This looks a lot like the definition of the integral of a function
of single variable. In fact this is also the definition of a double integral, or more exactly an
integral of a function of two variables over a rectangle. Here is the official definition of a
double integral of a function of two variables over a rectangular region R as well as the
notation that we’ll use for it.

Definition 4.1.2 (Double Integral of a Function of Two Variables over a Rectangular Region).
For any function f (x, y) defined on the rectangle R = {(x, y)|a ≤ x ≤ b, c ≤ y ≤ d}, divide
a ≤ x ≤ b into n subintervals and divide c ≤ y ≤ d into m subintervals each of the rectangles
RR
has a base area ∆A, the double integral of f over R is denoted by f (x, y)dA defined as:
R

ZZ X m
n X
f (x, y)dA = lim f (xi ∗ , yj ∗ )∆A
n,m→∞
R i=1 j=1

provided the limit exists and is the same for all choices of the evaluation points (xi ∗ , yj ∗ ) ∈ R
for i = 1, 2, , n. In this case, we say f is integrable over R. If f is non-negative and integrable
on R, then the volume V of the solid region between the graph of f and R is given by
ZZ
V = f (x, y)dA.
R

4.2 Iterated integrated

Definition 4.2.1. A. A plane region R is said to be vertically simple if there are two
continuous functions g1 and g2 on an interval [a, b] such that g1 (x) ≤ g2 (x) for a ≤ x ≤ b
and such that R is the region between the graph of g1 and g2 on [a, b]. In this case we say
that R is the vertically simple region between the graphs of g1 and g2 on [a, b] see figure
4.1.

B. A plane region R is horizontally simple if there are two continuous functions h1 and
h2 on an interval [c, d] such that h1 (y) ≤ h2 (y) for c ≤ y ≤ d and such that R is the region
between the graph h1 and h2 on [c, d]. In this case we say that R is the horizontally simple
region between the graphs of h1 and h2 on [c, d] see figure 4.2.

C. A plane region R is simple if it is both vertically simple and horizontally simple.

Example 4.2.1. Let R be the region between the graphs of y = x2 and y = x + 6.


Show that r is simple.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.2 Iterated integrated 80

Figure 4.1: Vertical simple region Figure 4.2: Horizontal simple region

Solution. First let us find the intersection points of the graphs of y = x2 and y = x + 6.
Observe that y = x2 and y = x + 6 if and Only if x2 = x + 6. Which means that x = −2
or x = 4. The graph intersects at the point (−2, 4) and (3, 9). Therefore, if g1 (x) = x2 and
g2 (x) = x + 6 , then g1 ≤ g2 on [−2, 3] , so R is the vertically simple region between the graphs
of g1 and g2 . To prove R is horizontally simple, we notice that R is composed of two portions,
one for which 0 ≤ y ≤ 4 and the other for which 4 ≤ y ≤ 9. Thus (x, y) is in R provided that
√ √
either 0 ≤ y ≤ 4 and − y ≤ x ≤ y. Or consequently
( √
− y, for 0 ≤ y ≤ 4
h1 (y) =
y − 6, 4 ≤ y ≤ 9

h2 (y) = y. Then, h1 ≤ h2 on [0, 9], so R is horizontally simple region between the graphs of h1
and h2 on [0, 9]. Since R is both vertically and horizontally simple, R is simple by definition. ◭

To compute a double integral, we can think of taking thin slices of the solid parallel to the yz
plane and integrating the function while holding x constant to find A(x) (i.e., perform a
partial integration), and then integrating the volume of all the thin slices to find the total
volume. This is called an iterated integral. Likewise, we could take thin slices parallel to
the xz plane to find A(y), and then integrate the volume of those slices to find total volume.
In either case, we get the same answer.

Theorem 4.2.1. Let f be continuous on a region R in the plane

a If R is vertically simple region between the graph of g1 and g2 on [a, b], then f is integrable
on R, and
ZZ Z b Z g2 (x)
f (x, y)dA = f (x, y) dx dy
a g1 (x)
R

b If R is horizontally simple region between the graph of h1 and h2 on [c, d] , then f is

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.2 Iterated integrated 81

integrable on R, and
ZZ Z bZ h2 (x)
f (x, y)dA = f (x, y) dy dx
a h1 (x)
R

Example 4.2.2. Let R be a rectangular region bounded by the lines x = −1, x = 2, y = 0 and
y = 2. Find ZZ
x2 ydA
R

Solution. The graph of R is vertically simple region between the graphs of y = 0 and y = 2
for −1 ≤ x ≤ 2. Therefore,
ZZ Z2 Z2
f (x, y)dA = x2 y dy dx
R −1 0

R2
To evaluate the iterated integral, we first compute x2 y dy for each x in −1 ≤ x ≤ 2 We obtain
0

Z2 Z2
2 2 1
x y dy = x y dy = x2 ( y 2 )(2 − 0) = 2x2
2
0 0

Because x is held constant when we integrate with respect to y. We conclude


ZZ Z2 Z2 Z2
f (x, y)dA = x2 y dy dx = 2x2 dx = 6
R −1 0 −1

6x+2y 2 dA, where R is the region enclosed by the parabola x = y 2


RR
Example 4.2.3. Evaluate
R
and the line x + y = 2.

Solution. The upper boundary changes form at x = 1. The left boundary is the same through-

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.2 Iterated integrated 82

out R. The right boundary is the same throughout R. Therefore choose horizontal strips.
Z1 Z2−y Z1
x=y2
(6x + 2y 2 ) dx dy =
 2
I= 3x + 2xy 2 x=2−y dy
−2 y 2 −2

Z1
(3(2 − y)2 + 2(2 − y)y 2 ) − (3y 4 + 2y 4 ) dy

=
−2
Z1
(12 − 12y + 3y 2 ) + (4y 2 − 2y 3 ) − 5y 4 dy

=
−2
Z1
12 − 12y + 7y 2 − 2y 3 − 5y 4 dy

=
−2
7 1
=[12y − 6y 2 + y 3 − y 4 − y 5 ]1−2
3 2
7 1 56
=(12 − 6 + − − 1) − (−24 − 24 − − 8 + 32)
3 2 3
99
=
2

The area A of a plane region R by


ZZ
A= 1dA
R

where R is the region between the graphs of two continuous functions g1 and g2 on [a, b] such
that g1 ≤ g2
ZZ Zb gZ2 (x) Zb
A= 1dA = 1 dy dx = (g1 − g2 ) dx.
R a g1 (x) a

Example 4.2.4. Find the area of the region in the xy-plane bounded by the curves y = x3 and

y = x.

Solution.
√ 
ZZ Z x 
Z1   Z1


2 3
 2
2 2
3

A= 1dA = dy dx = x − x dx = x 2 = (1 − 0) = .

3 
 3 1 3 3
R 0 x 0

Theorem 4.2.2 (Linear Combinations of Double Integrals). Let the function f (x, y) and g(x, y)
be integrable over the region R, and let c be any constant. Then the following holds:

RR RR
i cf (x, y)dA = c f (x, y)dA
R R

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.3 Change of variable in double integrals 83

RR RR RR
ii (f (x, y) + g(x, y)dA = f (x, y))dA + f (x, y)dA
R R R

iii . if R = R1 ∪ R2 , where R1 and R2 are non overlapping regions, then .


ZZ ZZ ZZ
f (x, y)dA = f (x, y)dA + f (x, y)dA
R R1 R2

4.3 Change of variable in double integrals

In evaluating a multiple integral over a region R, it is often convenient to use the coordinate
other than rectangular, such as the polar coordinates, cylindrical coordinates, spherical
coordinates, etc.
If we let (u, v) be curvilinear coordinates of points in a plane, there will be a set of
transformation x = f (x, y), y = g(x, y) mapping points (x, y) of the xy-plane into points (u, v)

of the uv-plane. In such case the region R of the xy-plane is mapped into a region R of the
uv-plane. We the have
∂(x, y)
ZZ ZZ
F (x, y) dx dy = G(u, v)| |dudv (4.1)
R R ∂(u, v)

where G(u, v) = F {f (u, v), g(u, v)} and


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

is the Jacobian of x and y with respect to u and v. The result 4.1 corresponds to change of
variables for double integrals.
∂(x,y)
Example 4.3.1. If u = x2 − y 2 and v = 2xy, find ∂(u,v)
in terms of u and v.

Solution.
∂u ∂u
∂(u, v)
= ∂x
∂v
∂y
∂v
= 4(x2 + y 2 )
∂(x, y) ∂x ∂y

From the identity (x2 + y 2 )2 = (x2 − y 2 )2 + (2xy)2 we have



(x2 + y 2 )2 = u2 + v 2 and x2 + y 2 = u2 + v 2 .
∂(x,y) 1
Then ∂(u,v) = ∂(u,v) = 4(x21+y2 ) = 4√u12 +v2 ◭
∂(x,y)

Example 4.3.2. a Let x = r cos θ and y = r sin θ. Find the value of ∂(x,y)
∂(r,θ)
.
RR p
b Evaluate x2 + y 2 dx dy, where R is the region in the xy-plane bounded by x2 + y 2 = 4
R
and x2 + y 2 = 9.

∂(x,y) xr xθ cos θ −r sin θ


Solution. a ∂(r,θ)
= = = r cos2 θ + r sin2 θ = r.
y r yθ sin θ r cos θ

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.4 Double Integrals in Polar coordinates 84

b By transforming the rectangular coordinates (x, y) in to polar coordinates (r, θ) and using
4.1 we have
Z2π Z3 Z2π 3 3 Z2π
∂(x, y) r 19 38π
ZZ p ZZ
2 2 2
x + y dx dy = r dr dθ = r dr dθ = dθ = dθ = .
∂(r, θ) 3 2 3 3
R R′ 0 2 0 0

Exercise 4.3.1. 1. Evaluate the following iterated integrals.


R1 Rx 2
R1 Rx3 y
a. (1 + x ) dy dx c. e x dy dx
0 0 0 0
Rπ sin
Rx
ex sin y dx dy
RR
b. y dy dx d.
0 0

2. Compute the double integral of the function f (x, y) = x2 + y 2 over the region bounded by
the curves y = 1 − x2 and y = x2 − 1 in the xy-plane.

3. Compute the double integral of the function f (x, y) = 6 − x + 2y over the region bounded
by the curves x = y 2 and y = 2x in the xy-plane.

4. Find the volume V of the solid region D bounded above by the paraboloid z = 4 − x2 − y 2
and below by the xy-plane(hint V = (4 − x2 − y 2 )dA)
RR
R

4.4 Double Integrals in Polar coordinates

To this point we have seen quite a few double integrals. However, in every case we’ve seen to
this point the region R could be easily described in terms of simple functions in Cartesian
coordinates. In this section we want to look at some regions that are much easier to describe
in terms of polar coordinates. For instance, we might have a region that is a disk, ring, or a
portion of a disk or ring. In these cases using Cartesian coordinates could be somewhat
cumbersome.
So, if we could convert our double integral formula into one involving polar coordinates we
would be in pretty good shape. The problem is that we can’t just convert the dx and the dy
into a dr and a dθ . In computing double integrals to this point we have been using the fact
that dA = dx dy and this really does require Cartesian coordinates to use. Once we’ve moved
into polar coordinates dA 6= dr dθ and so we’re going to need to determine just what dA is
under polar coordinates. After some manipulations it is found that dA = r dr dθ.Here is a
sketch of some region using polar coordinates.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.4 Double Integrals in Polar coordinates 85

So, our general region will be defined by inequalities,

α≤θ≤β
h1 (θ) ≤ r ≤ h2 (θ)

Theorem 4.4.1. Suppose that h1 and h2 are continuous on [α, β], where 0 ≤ α ≤ β ≤ 2π, and
that 0 ≤ h1 (θ) ≤ h2 (θ) for α ≤ θ ≤ β, Let R be the region between the polar graphs of r = h1 (θ)
and r = h2 (θ) for α ≤ θ ≤ β and if f is continuous on R, then

ZZ Zβ hZ2 (θ)
f (x, y)dA = f (r cos θ, r sin θ)r dr dθ
R α h! (θ)

In the event f is non-negative on R, the volume V of the region between the graph of f and R
is given by
Zβ hZ2 (θ)
V = f (r cos θ, r sin θ)r dr dθ
α h1 (θ)

and the area A of R is given by


Zβ hZ2 (θ)
A= r dr dθ
α h1 (θ)

Example 4.4.1. Evaluate the integral by converting into polar coordinates,


ZZ
2xydA
R

, R is the portion of the region between the circles of radius 2 and radius 5 centred at the origin
that lies in the first quadrant.

Solution. First let’s get R in terms of polar coordinates. The circle of radius 2 is given by
r = 2 and the circle of radius 5 is given by r = 5 . We want the region between them so we
will have the following inequality for r.

2≤r≤5

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.4 Double Integrals in Polar coordinates 86

Also, since we only want the portion that is in the first quadrant we get the following range of
θ’s. 0 ≤ θ ≤ π/2. Now that we’ve got these we can do the integral.
π
ZZ Z2 Z5
2xydA = 2r cos θr sin θr dr dθ
R 0 2
π
Z Z5 2

= 2r3 cos θ sin θ dr dθ


0 2
π
2
5
r4
Z  
= 2 cos θ sin θ dθ
4 2
0
π
Z2
609
= cosθsinθ dθ
2
0
609
=
4

Example 4.4.2. Let D be the solid region bounded above by the paraboloid z = 4 − x2 − y 2 and
below by the xy plane. Find the volume V of D.

Solution. The region R over which the integral is to be taken is bounded by the circle (Inter-
section of the surface with plane), whose equation in polar coordinate is r = 2: therefore D
can be described as the region between the graph of the paraboloid and the disk, As a result
ZZ Z2π Z2
2 2
4 − (r cos θ)2 − (r sin θ)2 r dr dθ
 
4−x −y dA =
R 0 0
Z2π Z2
4r − r3 dr dθ

=
0 0
Z2π 2
r4
2
= (2r − ) dθ
4 0
0
Z2π
=4 dθ = 8π
0

Exercise 4.4.1. 1 Determine the volume of the region that lies under the sphere, x2 + y 2 +
z 2 = 1 above the plane z = 0 and inside the cylinder x2 + y 2 = 5.

2 Determine the area of the region that lies inside r = 3 + sin θ and outside r = 2.

3 Find the volume of the region that lies inside the z = x2 + y 2 and below the plane z = 16.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.5 Application of Double Integrals 87

4.5 Application of Double Integrals

Double integrals have different applications in solving practical problems. Some of its
applications are used

i to find area of a plane figure and volume of certain solids(We have already seen.)

ii to find surface area of three dimensional surfaces(We will see in the next subsection), e.t.c.

4.5.1 Surface Area

Definition 4.5.1. Let R be a vertically or horizontally simple region, and let f have continuous
partial derivatives on R. If G is the graph of f on R, then the surface area S of G is defined
by
ZZ q
S= [fx ]2 + [fy ]2 + 1dA. (4.2)
R

Example 4.5.1. Find the surface area S of the portion of the paraboloid

z = 2 − x2 − y 2

that lies above the xy-plane.

Solution. The given surface lies over the region R in the xy-plane bounded by the circle
x2 + y 2 = 2. If f (x, y) = 2 − x2 − y 2 , then fx (x, y) = −2x and fy (x, y) = −2y. By equation 4.2,
ZZ q ZZ p
S= 2 2
[fx ] + [fy ] + 1dA = 4x2 + 4y 2 + 1dA
R R

This double integral can easily be evaluated by using polar coordinates.



Z2 Z 2√
13
ZZ p
S= 4(x2 + y 2 ) + 1dA = 4r2 + 1r dr dθ = π
3
R 0 0

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.5 Application of Double Integrals 88

p
Example 4.5.2. Find the surface area S of the frustum of the cone Z = x2 + y 2 with
minimum and maximum radii 1 and 2 respectively.

Solution. The given surface lies over the region R in the xy-plane bounded by the annulus
12 ≤ x2 + y 2 ≤ 22 . If f (x, y) = x2 + y 2 , then fx (x, y) = √ x2 2 and fy (x, y) = √ 2y 2 .
p
x +y x +y
So that
ZZ s
x2 y2
ZZ q
S= [fx ]2 + [fy ]2 + 1dA = + + 1dA
x2 + y 2 x2 + y 2
R
ZRZ √ √ ZZ √ √ √
= 2dA = 2 dA = 2A(R) = 2π(22 − 12 ) = 3 2π
R R



Exercise 4.5.1. 1. Find the surface area of that portion of the surface z = 4 − x2 that lies
above the rectangle R in the xy plane whose coordinates satisfy 0 ≤ x ≤ 10 and 0 ≤ y ≤ 40
2. Find the surface area of the portion of the paraboloid x2 + y 2 = z below the plane z = 1.
3. Find the area of the portion of the plane Z = x + 3y that lies inside the elliptical cylinder
2 2
with equation x4 + y9 = 1.

These are not the only applications of double integrals. The other applications related center
of gravity of a body, mass of a body and moment of inertia are mentioned at the end of the
next chapter.

4.5.2 Triple Integrals in Cartesian coordinates

Now that we know how to integrate over a two-dimensional region we need to move on to
integrating over a three-dimensional region. We used a double integral to integrate over a two
dimensional region and so it shouldn’t be too surprising that we’ll use a triple integral to
integrate over a three dimensional region. The notation for the general triple integrals is,
ZZZ
f (x, y, z)dV.
D

Let’s start simple by integrating over the box, D = [a, b] × [c, d] × [r, s]. Note that when using
this notation we list the x’s first, the y’s second and the z’s third. The triple integral in this
case is,
Zr Z d Z b
f (x, y, z) dx dy dz.
c a
s

Note that we integrated with respect to x first, then y, and finally z here, but in fact there is
no reason to the integrals in this order. There are 6 different possible orders to do the integral
in and which order you do the integral in will depend upon the function and the order that
you feel will be the easiest. We will get the same answer regardless of the order however. This
can be stated as a theorem:

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.5 Application of Double Integrals 89

Theorem 4.5.1 (Fubini’s Theorem). Let D be the rectangular box dened by the inequalities

a ≤ x ≤ b, c ≤ y ≤ d, r ≤ z ≤ s

If f is continuous on the region D, then


ZZZ Zb Z d Z s
f (x, y, z)dV = f (x, y, z) dz dy dx
c r
D a

Moreover, the iterated integral on the right can be replaced with any of the ve other iterated
integrals that result by altering the order of integration.

Example 4.5.3. Evaluate the following integral,


ZZZ
8xyzdV,
D

where D = [2, 3] × [1, 2] × [0, 1]

Solution. Just to make the point that order doesn’t matter let’s use a different order from
that listed above. We’ll do the integral in the following order.
ZZZ Z2 Z 2 Z 0
xyzdV = xyz dz dx dy
3 1
D 1
Z2 Z 2 1
2
= 4xyz dx dy
3 0
1
Z2 Z 2 1
= 4xy dx dy
3 0
1
Z2 2
2
= 2x y dx dy
3
1
Z2 2
= 10y dx dy = 15
3
1

Theorem 4.5.2. Let D be the solid region between the graphs of two continuous functions F1
and F2 on vertically or horizontally simple region R in the plane, and let f be continuous on
D. Then ZZZ ZZ Z F2 (x,y)
f (x, y, z)dV = ( f (x, y, z) dz)dA
F1 (x,y)
D R
F2R
(x,y)
We evaluate f (x, y, z) dz by integrating with respect to z while both x and y are held fixed,
F1 (x,y)
thus obtaining a number depending on x and y.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.5 Application of Double Integrals 90

If R is the vertically simple region between the graphs of g1 (x) and g2 (x) on [a, b], we
evaluate the double integral over R by using
ZZZ Zb Z g2 (x) Z F2 (x,y)
f (x, y, z)dV = (( f (x, y, z) dz) dy) dx.
g1 (x) F1 (x,y)
D a

In contrast, if R is the horizontally simple region between the graphs of h1 (y) and h2 (y)
on [c, d], we evaluate the double integral over R by using
ZZZ Zc Z h2 (y) Z F2 (x,y)
f (x, y, z)dV = (( f (x, y, z) dz) dx) dy.
h1 (y) F1 (x,y)
D d

Example 4.5.4. Let R be the triangular region in the xy plane between the graphs of y = 0
and y = x for 0 ≤ x ≤ 1 , and let D be the solid region between the graphs of the surfaces
z = −y 2 and z = x2 for all (x, y) in R. Evaluate
ZZZ
(x + 1)dV.
D

Solution. By the above discussions we have,


ZZZ Z1 Z x Z x2
(x + 1)dV = [ ( (x + 1) dz) dy] dx
0 −y 2
D 0
Z1 Z x x2
= (x + 1)z dy dx
0 −y 2
0
Z1 Z x
= (x + 1)(x2 + y 2 ) dy dx
0
0
Z1 x
xy 3 ) y3
= (x3 y + + x2 y + dx
3 3 0
0
Z1
4 3
= (x4 + x3 ) dx =
3 5
0

Evaluation of triple integrals as they are may sometimes be difficult or even possible. In this
case changing one coordinate system to another may facilitate the process. If (u, v, w) are
curvlinear coordinates in three dimensions, there will be a set of transformation equations
x = f (u, v, w), y = g(u, v, w) and z = h(u, v, w) and we can write:
∂(x, y, z)
ZZZ ZZZ
F (x, y, z) dx dy dz = G(u, v, w) dudvdw (4.3)
∂(u, v, w)
D D′

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.6 Triple integrals in cylindrical and spherical coordinates 91

where G(u, v, w) = F {f (u, v, w), g(u, v, w), h(u, v, w)} and


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

is the Jacobian of x, y and z with respect to u, v and w. The result (4.3)corresponds to


change of variables for triple integrals.

Exercise 4.5.2. 1. Evaluate the following


√ πiterated integrals.
R1 Rx x+y R 6 Ry Ry
(1 + y 2 z cos xz) dx dz dy
R
A. (z − 2x − y) dz dy dx B.
−1 0 x−y 0 0 0
π π
R 3 R1 Ry
ln
2
2 2 sin
Rz
(z 2 + 1)ey dz dy dx
RR
C. D. (sin y) dx dy dz
0 0 0 0 0 0
√1 π
R13 Re R x R2 cos
Rz cos
R yz
E. z(ln x)2 dz dx dy F. (x cos(yz) dx dy dz
−15 1 0 − π2 − cos z − cos yz

ey dV , where D is the solid region bounded by the planes y = 1,


RRR
2. Evaluate the integral
D
z = 0, y = x, y = −x and z = y.
RRR
3. Evaluate the integral zydV , where D is the solid region in the first octant bounded
D p
above by the planes z = 1 and below by the cone z = x2 + y 2 .

Now we are going to see cylindrical and spherical coordinates as special cases of the topic
change of variables in triple integrals.

4.6 Triple integrals in cylindrical and spherical coordinates

4.6.1 Triple Integrals in Cylindrical Coordinates

Let (x, y, z) be the rectangular coordinates of a point P in space. If (r, θ) is a set of polar
coordinates for the point (x, y), then we call (r, θ, z) a set of cylindrical coordinates for P as
shown in the figure 4.3.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.6 Triple integrals in cylindrical and spherical coordinates 92

P (x, y, z) = P (r, θ, z)
Given the rectangular coordinates (x, y, z) of
a point P , we can determine a set of cylindri-
cal coordinates for P with aid of the formulas
x2 + y 2 = r2 and tan θ = xy (if x 6= 0). Con-
versely, from any set (r, θ, z) of cylindrical
coordinates of a point P we can determine
r
the rectangular coordinates (x, y, z) of P by θ
y
the formulas x = r cos θ and y = r sin θ.

Figure 4.3:

Let see equations of common solids in rectangular and cylindrical coordinates.

Surface Rectangular Cylindrical


Cylinder x 2 + y 2 = a2 r=a
Sphere x2 + y 2 + z 2 = a2 r2 + z 2 = a2
Double circular cone x 2 + y 2 = a2 z 2 r = az or z = a cot φ
Circular paraboloid x2 + y 2 = az r2 = az

As you have seen in previous section,


ZZZ Z Z Z F2 (x,y)
f (x, y, z)dV = ( f (x, y, z) dz)dA
F1 (x,y)
D R

for a solid region between the graphs of two continuous functions F1 and F2 on a vertically or
Horizontally simple region R in the xy-plane and for a continuous function f on D.
Theorem 4.6.1. Let D be a solid region between the graphs of F1 and F2 on R, where R is
the plane region between the polar graphs of h1 and h2 on [α, β], with 0 ≤ β − α ≤ 2π and
0 ≤ h1 (θ) ≤ h2 (θ) for α ≤ θ ≤ β. If f is continuous on D, then
ZZZ Zα Z h2 (θ) Z F2 (r cos θ,r sin θ)
f (x, y, z)dV = f (r cos θ, r sin θ, z)r dz dr dθ.
h1 (θ) F1 (r cos θ,r sin θ)
D β

Integration by means of cylindrical coordinates is especially effective when expressions


containing x2 + y 2 appear in the integrand or in the limits of integration and the region over
which the integration is taken is easily described by polar coordinates.
Example 4.6.1. Let D be the solid region bounded above by the plane y + z = 4, below by the
xy- plane and on the sides by the cylinder x2 + y 2 = 16 as shown in the figure 4.4. Evaluate
ZZZ p
x2 + y 2 dV.
R

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.6 Triple integrals in cylindrical and spherical coordinates 93

Figure 4.4:

Solution. Observe that D is a solid region between the graphs of z = 0 and z = 4 − y on R,


where R is the disk x2 + y 2 ≤ 16. In polar coordinate R is the region between the polar graphs
of r = 0 and r = 4 for 0 ≤ θ ≤ 2π. Consequently, in cylindrical coordinates D is the solid
region between the graphs z = 0 and z = 4 − r sin θ for (r, θ) in R. Then by Theorem 4.6.1 we
have
ZZZ p Z2π Z4 4−r
Z sin θ Z2π Z4 4−r sin θ
x2 + y 2 dV = r · r dz dr dθ = r2 z dr dθ
0
R 0 0 0 0 0
Z2π Z4 Z2π  4
4 3 r4

2 3
= (4r − r sin θ) dr dθ = r − sin θ dθ
3 4 0
0 0 0
Z2π     2π
256 256 512
= − 64 sin θ dθ = θ + 64 cos θ = π
3 3 0 3
0



R3 R 2
9−x R1
Example 4.6.2. Evaluate y 2 dz dy dx.

−3 − 9−x2 (x2 +y 2 )2

√ √
Solution. The limit of integrations −3 and 3 in the first integral and − 9 − x2 and 9 − x2
on the second integral tells us that those two integrals are taken over the region of the disk

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.6 Triple integrals in cylindrical and spherical coordinates 94

x2 + y 2 ≤ 9 or r ≤ 3, in the xy- plane. It follows that



Z3 Z9−x2 Z1 Z0 Z3 Z1
2
y dz dy dx = (r sin θ)2 r dz dr dθ

−3 − 9−x2 (x2 +y 2 )2 2π 0 r4

Z0 Z3  1 Z0 Z3  
3 2 3 7
= r sin θz dr dθ = r −r sin2 θ dr dθ]
r4
2π 0 2π 0
Z2π  3
r4 r8 6399 2π 2
Z
2
= − sin θ dθ = − sin θ dθ
4 8 0 8 0
0
2π
6399 2π

6399 sin 2θ 6399
Z
=− (1 − cos 2θ) dθ = − θ− =− π
16 0 16 2 0 8

Exercise 4.6.1. Evaluate the following integrals by changing into cylindrical coordinates.

(x2 + y 2 )dV , where D is the solid region bounded by the cylinder x2 + y 2 = 1 and the
RRR
1.
D
planes z = 0 and z = 4.

(xz)dV , where D is the portion of the ball portionof theballx2 + y 2 + z 2 ≤ 4 in the first
RRR
2.
D
octant.
RRR 2
3. (y z)dV , where D is the solid region bounded above by the sphere x2 + y 2 + z 2 = 4.
D

4.6.2 Triple Integrals in Spherical Coordinates

Given a point P with rectangular coordinates (x, y, z). Let ρ = |OP | be the distance between
the origin O and P , and let φ be the angle measured (downward) from the positive z− axis to
−→
OP as shown in the figure 4.5. Also let θ be the same angle as in the case of cylindrical
coordinates, namely the angle from the positive x− axis to OQ, where Q is the projection of
P onto the xy plane(θ is measured in counter clockwise direction as shown on the xy-plane
from the side of the positive z− axis).

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.6 Triple integrals in cylindrical and spherical coordinates 95

P (x, y, z) = P (ρ, φ, θ)
Then the point P is said to have spherical
coordinates ρ, θ and φ, and we write P =
ρ
(ρ, θ, φ) as well as P = (x, y, z). Where 0 ≤
φ
r ≤ ∞, 0 ≤ θ ≤ 2π and 0 ≤ φ ≤ π. From
trigonometry we found that r = ρ sin φ and
z = ρ cos φ . These equations along with the
r
polar coordinated formulae x = r cos θ and θ
y
y = r sin θ.

Figure 4.5:

Yield the following formulae for converting from spherical coordinates to rectangular
coordinates: x = r cos θ = ρ sin φ cos θ, y = r sin θ = ρ sin φ sin θ and z = ρcosφ. As you can
easily verify that x2 + y 2 + z 2 = ρ2 ⇒ ρ = x2 + y 2 + z 2 , tan θ = xy , cos φ = √ 2 z 2 2 ,
p
x +y +z
(x2 + y 2 + z 2 6= 0).

Theorem 4.6.2 (Triple integrals in spherical coordinates). If f (x, y, z) is continuous on a solid


region D, then
ZZZ ZZZ
f (x, y, z)dV = f (ρ sin φ cos θ, ρ sin φ sin θ, ρcosφ)ρ2 sin φ dρ dφ dθ.
D D′

z 2 dV , where D is a solid region x2 + y 2 + z 2 ≤ 1.


RRR
Example 4.6.3. Evaluate
D

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.6 Triple integrals in cylindrical and spherical coordinates 96

Solution. By using spherical coordinates


ZZZ Z2π Zπ Z1
2
z dV = (ρ cos φ)2 ρ2 sin φ dρ dφ dθ
D 0 0 0
Z2π  Zπ  Z1 Zπ  1
ρ5
 
4 2 2
= ρ dρ (cos φ) ρ sin φ dφ dθ = 2π cos2 φ sin φ dφ
5 0
0 0 0 0
Zπ  
2π 2 2π 2 4
= cos φ sin φ dφ = = π
5 5 3 15
0

RRR p
Example 4.6.4. Use spherical coordinates and evaluate x2 + y 2 + z 2 dV , where D is the
D
ball x2 + y 2 + z 2 ≤ a2 (a 6= 0)

Solution. Using the transformation equations x = ρ sin φ cos θ, y = ρ sin φ sin θ, z = ρ cos φ, ρ2 =
x2 + y 2 + z 2 and dV = ρ2 sin φ dρ dφdθ, we have
ZZZ p Z2π Zπ Za
x2 + y 2 + z 2 dV = (ρ · ρ2 sin φ dρ dφ dθ
D 0 0 0
Z2π  Zπ  Za   Z2π  Zπ  4 a  
3 ρ
= ρ dρ sin φ dφ dθ = sin φ dφ dθ
4 0
0 0 0 0 0
2π Zπ
a4 sin φ
Z    
= dφ dθ
4
0 0
Z2π π Z2π
a4 a4
=− (cos φ) dθ = dθ = πa4
4 0 2
0 0


(x2 + y 2 )dV , where D is a solid region
RRR
Exercise 4.6.2. 1. Evaluate the triple integral
q D

bounded above by the unit sphere x + y + z = 1 and below by the cone z = − 13 (x2 + y 2 )
2 2 2

using spherical coordinates.


dV
, where D is the bounded solid region between the cylinder x2 +y 2 = 4
RRR
2. Evaluate x2 +y 2 +z 2
D
and the nappes of the cone x2 + y 2 = z 2 using spherical coordinates.
dV 2 2 2
RRR
3. Evaluate 2 2 2
3 , where D is the spherical shell 1 ≤ x +y +z ≤ e using spherical
D (x +y +z ) 2
coordinates.
4. Use spherical coordinates to evaluate
√ √
2 2
Z2 Z4−x2 4−x Z −y p
z 2 x2 + y 2 + z 2 dz dy dx.

−2 − 4−x2 0

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.7 Application: Volume, center of mass of solid region 97

5. Use either cylindrical or spherical coordinates to evaluate the integral


√ √
2 2
Z1 Z1−x2 1−xZ −y √
2 2 2 3
e−( x +y +z ) 2 dz dy dx.
−1 0 0

4.7 Application: Volume, center of mass of solid region

Suppose f (x, y, z) = 1. The volume of the solid D is defined to be


ZZZ
V = dV.

Example 4.7.1. Use triple integral to find the volume of the solid enclosed between the cylinder
x2 + y 2 = 16 and the planes z = 0 and y + z = 4.

Solution. The solid D and its projection R on the the xy- are shown in Figure 4.6. The
lower surface of the solid is the plane z = 0, and the upper surface is the plane y + z = 4, or
equivalently, z = 4 − y. Thus,
ZZZ Z Z  Z4−y  ZZ
V = dv = dz dA = (4 − y)dA
D R 0 R
Z2π Z4
= (4 − r sin θ)r dr dθ (by using cylindrical coordinates)
0 0
Z2π  Z4 Z2π  4
r3
 
2 2
= (4r − r sin θ) dr dθ = 2r − θ dθ
sin 0
0 0 0
Z2π     2π
64 64
= 32 − sin θ dθ = 32θ + cos θ
3 3 0
0
 
64 64
= 64π + − = 64π
3 3

Figure 4.6:

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.7 Application: Volume, center of mass of solid region 98

Example 4.7.2. Find the volume V of the region in the first octant bounded by the planes
z = 10 + x + y, y = 2 − x, y = x, z = 0 and x = 0.

Solution. The region R of the solid bounded in the xy-plane is as shown in the fig—. The line
y = 2 − x and y = x intersect at (1, 1). Hence, x ≤ y ≤ 2 − x as 0 ≤ x ≤ 1. Therefore,
ZZZ Z1 Z2−x 10+x+y
Z Z1  Z2−x 10+x+y
Z  
V = dv = dx dy dz = dz dy dx
D x=0 y=x z=0 x=0 y=x z=0
Z1  Z2−x Z1 1
x3

34
= (10x + x + y) dy dx = {20 − 20x − 2x2 } dx = 2{11x − 5x2 − } = .
3 0 4
0 x 0

4.7.1 Some of the applications of multiple integrals

Some of the applications of multiple integral are used to

• evaluate the area of a region enclosed by a plane curves


• calculate area of curved surfaces
• find the volume of a solid region
• calculate a mass of a lamina
• calculate a mass of a solid figure
• calculate center of gravity of a plane lamina and a solid
• calculate moment of inertia of a plane lamina and a solid

Now let see some of its applications:

1. Calculation of Mass
(a) For a plane lamina, if the surface density at the point P (x, y) be ρ = f (x, y), then the
elementary mass at P is equal to ρ dx dy. therefore, the total mass of the lamina is
ZZ
ρ dx dy

with integrals embracing the whole area of the lamina. In polar coordinates, taking
ρ = φ(r, θ) at the point P (r, θ), the total mass of the lamina is
ZZ
ρ dr dθ

(b) For a solid if the density at a point P (x, y, z) be ρ = f (x, y, z), then the total mass of
the solid is ZZ
ρ dx dy dz
with appropriate limit of integration.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.7 Application: Volume, center of mass of solid region 99

Figure 4.7:

2. Center of Gravity

(a) To find the center of gravity (x, y) of a plane lamina, take the element of mass ρ dx dy
at the point P (x, y), then
RR RR
xρ dx dy yρ dx dy
x = RR , y = RR
ρ dx dy ρ dx dy

(b) To find Center of Gravity (x, y, z) of a solid, take an element of mass ρ dx dy dz


enclosing the point P (x, y, z). Then,
RRR RRR RRR
xρ dx dy dz yρ dx dy dz zρ dx dy dz
x = RRR , y = RRR , z = RRR
ρ dx dy dz ρ dx dy dz ρ dx dy dz
Example 4.7.3. Find the mass of the tetrahedron bounded by the coordinate planes and
the plane xa + yb + zc = 1, with variable density ρ(x, y, z) = µxyz.

Solution. Elementary mass at point P (x, y, z) is µxyz dx dy dz. Therefore, the whole
RRR
mass is equal to µxyz dx dy dz, the integrals
 embracing
 the whole volume of OABC(Fig
4.7). The limits for Z are from 0 to Z = c 1 − xa − yb . The limits for y are from 0 to
 
x
y = b 1 − a and the limits of x are from 0 to a. Hence the required mass is

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.7 Application: Volume, center of mass of solid region 100

 
c 1− x
a
− yb  
x x
Za b(1−
Z a) Za b(1−
Z a) c 1− x − yb
xyz 2
Z a
µxyz dz dy dx = µ dy dx
2 z=0
0 0 0 0 0
x
Za b(1−
Z a)
c2 x y
=µ xy · (1 − − )2 dy dx
2 a b
0 0
x
Za b(1−
Z a) 2
µc2 x y2 y3
  
x
= x 1− y−2 1− + 2 dy dx
α a a b b
0 0
Za  2
µc2 x y2 x y3 y4
 
= x 1− −2 1− + dx
α a 2 a 3b 4b2
0
Za  4 4 4 
µc2 b2 2b2 b2
  
x x x
= 1− − 1− + 1− dx
α 2 a 3 a 4 a
0
Za  4
µb2 c2 x µa2 b2 c2
= x 1− dx = .
24 a 720
0

Example 4.7.4. Find the volume and centroid (center of gravity) of the uniform “ice-
Cream Cone” D bounded by the cone φ = π6 and the sphere ρ = 2a cos φ of radius a. The
sphere and the part of the cone within it are shown in figure–.

π
Solution. The ice-cream cone is described by the inequalities 0 ≤ θ ≤ 2π, 0 ≤ φ ≤ 6
and
0 ≤ ρ ≤ 2a cos φ. Its volume is given by
π π
Z2π Z6 2aZcos φ Z2π Z6
2 8 3
V = ρ sin φ dρ dφ dθ = a sin φ dφdθ
3
0 0 0 0 0
  π6
16 1 7
= a3 π − cos4 φ = πa3
3 4 0 12
Now for the centroid, it is clear by symmetry that x = y = 0. We may also assume that
D has a constant density δ, so that the mass of D is numerically equal to volume times its
7
density equal to 12 πa3 δ. Because Z = ρ cos φ, the Z-coordinates of the centroid of D is

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.7 Application: Volume, center of mass of solid region 101

RRR RRR
ρzdV zdV
1
ZZZ
D D
Z = RRR = RRR = zdV
ρdV dV V
D D D
π
Z2π Z 6 2aZcos φ
12
= ρ3 cos φ sin φ dρ dφ dθ
7πa3
0 0 0
π
Z2π Z6   π6
48 5 96 1 6 37
= a cos φ sin φ dφ dθ = a − cos φ = a
7 7 6 0 28
0 0
 
37
Hence the centroid of the ice-cream is located at the point 0, 0, 28
a . ◭

3. Moment of Inertia

If a particle of mass m of a body be at a dis-


tance r from a given line, them mr2 is called
the moment of inertia of the particle about
the given line and the sum of similar expres-
sions taken for all the particles of the body,
mr2 is called the moment of inertial of
P
i.e
the body about the given line. (fig4.8) If M
be the total mass of the body and we write
its moment of inertia is equal to M k 2 , then k
is called the radius of gyration of the body
about the axis.

Figure 4.8:

(a) Moment of inertia of a plane lamina Consider an elementary mass ρ dx dy at the point
P (x, y) of a plane area A so that it’s Moment of Inertia about x- axis is equal to
ρ dx dyy 2 .
∴ M.I of the lamina about the x− axis, i.e. Ix is given by
ZZ
Ix = ρy 2 dx dy.
R

Similarly, M.I of the lamina about the y− axis, i.e. Iy is given by


ZZ
Iy = ρx2 dx dy.
R

Also, M.I of the lamina about an axis perpendicular to xy-plane, i.e. Iz is given by
ZZ
Iz = ρ(x2 + b2 ) dx dy.
R

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.7 Application: Volume, center of mass of solid region 102

(b) Moment of inertia of a solid Consider elementary mass ρ dx dy dz enclosing a point


P (x, y, z) of a solid from volume V . Distance of P from the x- axis is equal to
p
y 2 + z 2 . Therefore, M.I. of this element about the x− axis is equal to ρ dx dy dz(y 2 +
z 2 ). Thus M.I of this solid about the x− axis is
ZZZ
Ix = ρ(y 2 + z 2 ) dx dy dz.
D

Similarly, M.I about the y− axis is


ZZZ
Iy = ρ(x2 + z 2 ) dx dy dz.
D

M.I about the z− axis is


ZZZ
Iz = ρ(x2 + y 2 ) dx dy dz.
D

Example 4.7.5. A solid ball D with constant density δ is bounded by the spherical surface
with equation ρ = a. Use spherical coordinates to compute its volume V and its moment
of inertia Iz around the z− axis.

Solution. The points of the ball are described by the inequalities 0 ≤ ρ ≤ a, 0 ≤ φ ≤ π


and 0 ≤ θ ≤ 2π. Volume of the ball is given by
Z2π Zπ Za Z2π Zπ
1
ZZZ
V = dV = ρ2 sin φ dρ dφ dθ = a3 sin φ dφ dθ
3
D 0 0 0 0 0
Z2π Z2π
1 π 2 4
= a3 dθ = a3 dθ = πa3 .

− cos φ 0
3 3 3
0 0

The distance from a typical point (ρ, φ, θ) of the sphere to the z- axis is r = ρ sin φ, so the
moment of inertial of the sphere around that axis is
ZZZ Z2π Zπ Za
2
Iz = r δdV = δρ4 sin3 φ dρ dφ dθ
D 0 0 0
Z2π Za Zπ
1 2 2
= δ5 sin φ dφ dθ = πδa5
3
sin3 φ dφ = ma2
5 5 5
0 0 0

where m = 43 πa3 δ. ◭

Exercise 4.7.1. (a) Find the volume bounded by the cylinder x2 + y 2 = 4 and the planes
y + z = 4 and z = 0.
(b) Find the mass of a lamina in the form of the cardioid r = a(1 + cos θ).
(c) Find the centroid of a loop of the lemniscate r2 = a2 cos 2θ.
(d) If the density at any point of the solid octant of the ellipsoid ( xa )2 + ( yb )2 + ( zc )2 = 1
various as xyz, find the coordinates of the center of gravity of the solid.

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)


lOMoARcPSD|39572262

4.7 Application: Volume, center of mass of solid region 103

(e) Using double integrals, find the moment of inertial about the x− axis of the area
enclosed by the lines x = 0, y = 0, xa + yb = 1.
(f ) Find the moment of inertia about the z-axis of a homogeneous tetrahedron bounded by
the planes x = 0, y = 0, z = x + y and z = 1.
(g) A lamina is shaped like the circular sector R in the first quadrat bounded by the co-
ordinate axes and an arc of the unit circle x2 + y 2 = 1. Find the mass and center of
gravity (x, y) of the lamina its density function is ρ(x, y) = x2 + y 2 .

Downloaded by Abdii Kumsaa (kumsaaabdii.bus@gmail.com)

You might also like