Lecture Notes 230

Calculus II Lecture Notes
David M. McClendon
Department of Mathematics
Ferris State University
2016 edition
2016
c David M. McClendon
1
Contents
Contents 2
1 Review of Calculus I 5
1.1 Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.2 Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.3 The definite integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.4 Homework exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2 Basic integration techniques 24

2.1 Integrals to memorize . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.2 Rewriting the integrand . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.3 Elementary u-substitutions . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.4 More complicated u-substitutions . . . . . . . . . . . . . . . . . . . . . 35
3 Intermediate integration techniques 44

3.1 Integration by parts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.2 Partial fractions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.3 Summary of integration techniques . . . . . . . . . . . . . . . . . . . . 56
3.4 Mathematica commands for integration . . . . . . . . . . . . . . . . . . 56
4 Improper integrals 61
4.1 Boundedness vs. unboundedness . . . . . . . . . . . . . . . . . . . . . 61
4.2 Horizontally unbounded regions . . . . . . . . . . . . . . . . . . . . . 63
4.3 Vertically unbounded regions . . . . . . . . . . . . . . . . . . . . . . . 66
4.4 Theoretical approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2
Contents
5 Applications of the integral 80

5.1 Area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.2 Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.3 General principles behind all applications of integration . . . . . . . . 96
5.4 Arc length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
5.5 One-dimensional moments and centers of mass . . . . . . . . . . . . . 100
5.6 Two-dimensional moments and centers of mass . . . . . . . . . . . . . 105
5.7 Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
6 Parametric equations 130

6.1 Introduction to parametric equations . . . . . . . . . . . . . . . . . . . 130
6.2 Parametric equations of common graphs . . . . . . . . . . . . . . . . . 134
6.3 Calculus with parametric equations . . . . . . . . . . . . . . . . . . . . 143
6.4 Transformations on parametric equations . . . . . . . . . . . . . . . . 153
7 Introduction to infinite series 162

7.1 Motivation and big-picture questions . . . . . . . . . . . . . . . . . . . 162
7.2 Convergence and divergence . . . . . . . . . . . . . . . . . . . . . . . 166
7.3 Σ-notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
7.4 Elementary properties of convergence and divergence . . . . . . . . . 172
7.5 Changing indices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
8 Geometric series and the Ratio Test 180

8.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
8.2 The Geometric Series Test . . . . . . . . . . . . . . . . . . . . . . . . . 183
8.3 Applications of geometric series . . . . . . . . . . . . . . . . . . . . . . 189
8.4 The Ratio Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
9 Convergence tests for positive series 204

9.1 Classifying series according to sign . . . . . . . . . . . . . . . . . . . . 204
9.2 The Integral Test; harmonic and p-series . . . . . . . . . . . . . . . . . 206
9.3 The nth -term Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
9.4 The Comparison Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
10 Absolute and conditional convergence 218

10.1 Alternating series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
10.2 The triangle inequality . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
10.3 Absolute and conditional convergence . . . . . . . . . . . . . . . . . . 224
3
Contents
10.4 Rearrangement of infinite series . . . . . . . . . . . . . . . . . . . . . . 226

10.5 Summary of classification techniques . . . . . . . . . . . . . . . . . . . 229
10.6 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
11 Taylor series 236

11.1 Uniqueness of power series . . . . . . . . . . . . . . . . . . . . . . . . 236
11.2 Applications of Taylor series . . . . . . . . . . . . . . . . . . . . . . . . 245
11.3 General theory of power series . . . . . . . . . . . . . . . . . . . . . . 252
Index 263
4
Chapter 1
Review of Calculus I
Big picture issues

1. What is calculus?
2. What are some typical kinds of problems you learn how to solve / mathe-
matical procedures you learn in Calculus I?
3. What is the difference between a math problem that is a calculus problem

and a math problem that is NOT a calculus problem?
5
Most of Calculus I theory in one chart
CALCULUS
OBJECT
MOTIVATING
PROBLEM
APPROXIMATION
OF THE
SOLUTION
HOW THE
APPROXIMATION
IMPROVES
THEORETICAL
SOLUTION
HOW THE
THEORETICAL
SOLUTION
IS COMPUTED
IN PRACTICE
APPLICATIONS
6
1.1. Limits
What is the difference between a math problem that is a calculus problem and a
math problem that is NOT a calculus problem?
A calculus problem is one that contains a limit; a non-calculus problem con-
tains no limit.
What is calculus?
Calculus is the study of limits.
DERIVATIVE - APPLICATIONS

*

LIMIT - INTEGRAL - APPLICATIONS
HH
HH
H
j
H
? - ?
1.1 Limits
Recall: To say
lim f (x) = L
x→a
means that as x gets closer and closer to a, then f (x) gets closer and closer to L.
This suggests that the graph of f looks like one of the following three pictures:
The graph on the left is “continuous” at a; the other two graphs are not. More
precisely,
7
1.1. Limits
Definition 1.1 A function f : R → R is continuous at a if lim f (x) = f (a). f is

x→a
called continuous if it is continuous at every point in its domain.
Theorem 1.2 Any function which is the quotient of functions made up of powers of
x, sines, cosines, arcsines, arctangents, exponentials and/or logarithms is continuous
everywhere except where the denominator is zero.
This theorem suggests that to evaluate most limits, you should start by plug-
ging in a for x. If you get a number, that is usually the answer.
Example: Evaluate these limits:
1. lim (3 sin 2x)

x→π/3
2. lim x+3
x−2
x→4
Example: What about these:

x−5
3. lim (x−2) 2
x→2
2 −3x−10
4. lim x x−5
x→5
When you plug in a for x and you get 0 in a denominator, you have to work a
bit harder to evaluate a limit.
nonzero
• When you get 0
:
x−5
Example 3 above: lim (x−2) 2
x→2
8
1.1. Limits
• When you get 00 :
2 −3x−10
Example 4 above: lim x x−5
x→5
Techniques for dealing with 00 :

– Factor and cancel (as in Example 4)
– Conjugate square roots
– Clear denominators of “inside” fractions
– L’Hôpital’s Rule
2 −3x−10
Example 4 revisited: lim x x−5
x→5
Theorem 1.3 (L’Hôpital’s Rule) Suppose f and g are differentiable functions. Sup-
pose also that either
lim f (x) =lim

x→a x→a
g(x) = 0 or x→a
lim f (x) =lim
x→a
g(x) = ±∞.
Then:
f (x) L f 0 (x)
lim =lim .
x→a g(x) x→a g 0 (x)
Limits at infinity
We can also take a limit of a function as x → ∞. To say lim f (x) = L means that
x→∞
as x grows larger and larger without bound, then f (x) approaches L. Graphically,
this means f has a horizontal asymptote (HA) y = L:
9
1.1. Limits
Although ∞ is not a number, it can be manipulated in some ways as if it is a

number, so one can evaluate infinite limits by “plugging in ∞ for x” and applying
the following arithmetic rules for ∞ (the cs represent arbitrary constants):
(
∞ ∞ if c > 0
∞±c=∞ ∞·∞=∞ ∞ =∞ c·∞=
−∞ if c < 0
√
(
∞ c ∞ if c > 0
∞=∞ ln ∞ = ∞ e =∞ ∞ =
0 if c < 0
c ∞ c
=0 = ±∞ = ±∞ so long as c 6= 0
∞ 0 0
In the last two situations, you need to analyze the sign carefully to determine
whether the answer is ∞ or −∞.
Warning: The following expressions are indeterminate forms, i.e. they can
work out to be different things depending on the context (these are evaluated using
L’Hôpital’s Rule):
0 ∞
0·∞ ∞−∞ 1∞ ∞0 00
0 ∞
lim lnxx .
Example 5: Evaluate the limit x→∞
3
lim 4x
Example 6: Evaluate the limit x→∞ +2x+1
2x3 +x2 +2
.
The last example generalizes into the following principle which has many use-
ful applications:
10
1.2. Derivatives
Theorem 1.4 (Limits at infinity for rational functions) Suppose f is a rational

function, i.e. has form
am xm + am−1 xm−1 + am−2 xm−2 + ... + a2 x2 + a1 x + a0

f (x) = .
bn xn + bn−1 xn−1 + bn−2 xn−2 + ... + b2 x2 + b1 x + b0
Then:
1. If m < n (i.e. largest power in numerator < largest power in denominator),

then lim f (x) = 0.
x→∞
2. If m > n (i.e. largest power in numerator > largest power in denominator),

then lim f (x) = ±∞.
x→∞
3. If m = n (i.e. largest powers in numerator and denominator are equal), then

lim f (x) = abm .
x→∞ n
1.2 Derivatives
Definition 1.5 (Limit definition of the derivative) Let f : R → R be a function
and let x be in the domain of f . If the limit
f (x + h) − f (x)
lim
h→0 h
exists and is finite, say that f is differentiable at x. In this case, we call the value of
df dy
this limit the derivative of f and denote it by f 0 (x) or dx or dx .
Differentiable functions are smooth, i.e. are continuous and do not have sharp
corners, vertical tangencies or cusps.
Assuming it exists, the derivative f 0 (x) computes:

1. the slope of the line tangent to f at x;
2. the slope of the graph of f at x;
3. the instantaneous rate of change of y with respect to x;
4. the instantaneous velocity at time x (if f is the position of the object at time
x).
11
1.2. Derivatives
We do not compute derivatives using the limit definition. We use differentiation

rules, which consist of a list of functions whose derivative we memorize and a list
of rules which tell us how to differentiate more complicated functions made up of
pieces whose derivatives we know.
Functions whose derivatives we memorize:
• text
–
–
–
–
Differentiation rules for more complicated functions:
• Constant Multiple Rule: (cf )0 (x) = c f 0 (x)
• Sum Rule: (f + g)0 (x) = f 0 (x) + g 0 (x)
• Difference Rule: (f − g)0 (x) = f 0 (x) − g 0 (x)
12
1.2. Derivatives
Example: Suppose f (x) = 3x6 sin x. Find f 0 (x).
4 dy
Example: Suppose y = x2
− 2 ln x + 9e2x . Find dx
.
Example: Find the slope of the line tangent to the graph of f (x) = 3 cos x+2 sin x
at x = π4 .
Solution: The slope of the tangent line at π

4
is given by f 0 ( π4 ). By usual rules,
f 0 (x) = 3 · − sin x + 2 cos x

π π π

f0 = −3 sin + 2 cos
4 √4 √4
2 2 −1
= −3 · +2· =√ .
2 2 2
Derivatives have many applications. The most important (for Math 230 pur-
poses) is that given a differentiable function f , you can approximate values of f
near a using the tangent line to f at a:
13
1.3. The definite integral
Definition 1.6 Given a differentiable function f and a number a at which f is differ-

entiable, the tangent line to f at a is the line whose equation is
y = f (a) + f 0 (a)(x − a) (a.k.a. L(x) = f (a) + f 0 (a)(x − a)).
For values of x near a, f (x) ≈ L(x); approximating f (x) via this procedure is called
linear approximation.
You can also conclude many things about the graph of f from looking at its
derivative and its higher-order derivatives, as defined below:
Definition 1.7 Let f : R → R be a function.
• The zeroth derivative of f , sometimes denoted f (0) , is just the function f itself.
dy
• The first derivative of f , sometimes denoted f (1) or dx
, is just f 0 .
d2 y
• The second derivative of f , denoted f 00 or f (2) or dx2
, is the derivative of f 0 :
f 00 = (f 0 )0 .
dn y
• More generally, the nth derivative of f , denoted f (n) or dxn
,is the derivative of
f (n−1) :
f (n) = ((((f 0 )0 ) · · ·0 )0
The first derivative of a function measures its tone (i.e. whether it is increasing
or decreasing). The second derivative of a function measures its concavity (i.e.
whether its graph is cupped upward or downward). In Math 230, we will see
what the higher-order derivatives of f have to do with the function.
1.3 The definite integral

Definition 1.8 Given function f : [a, b] → R, the definite integral of f from a to b
is Z b n
X
f (x) dx = lim f (ck )∆xk ,
a ||P||→0
k=1
where the expression inside the limit is a Riemann sum for f .
14
Note: In Math 230, the limit above always exists (but it doesn’t always exist for
crazy functions f ... take Math 430 for more on that).
The definite integral of a function is a number which is supposed to give the

signed area of the region between the graph of f and the x-axis. Area above the
x-axis is counted as positive area; area below the x-axis is counted as negative area.
As with derivatives, we do not compute integrals with this definition. We use
the following important circle of ideas:
Definition 1.9 Given function f , an antiderivative of f is a function F such that

F0 = f.
Ex: F (x) = sin x is an antiderivative of f (x) = cos x.
Every continuous function has an antiderivative (although you may not be able
to write its formula down); any two antiderivatives of the same function must dif-
fer by a constant (so if you know one antiderivative, you know them all by adding
a +C to the one you know).
Definition 1.10 Given function f , the indefinite integral of f , denoted

Z
f (x) dx,
is the set of all antiderivatives of f .

R
Ex: cos x dx = sin x + C.
Theorem 1.11 (Fundamental Theorem of Calculus Part II) Let f be continuous

on [a, b]. Suppose F is any antiderivative of f . Then
Z b
f (x) dx = F (x)|ba = F (b) − F (a).
a
Despite the similar notation, f (x) dx and ab f (x) dx are very different objects.
R R
The first object is a set of functions; the second object is a number.
15
Exercise: Without looking anything up, evaluate each of the following indefi-
nite integrals, if they are “easy” to solve. If the integral isn’t easily solved, put a
“?”. The idea here is that an integral which is “easy” should use facts from alge-
bra regarding exponents, logarithms, etc. and the rules memorized in Calculus I.
“Easy” integrals do not use sophisticated calculus techniques or trig identities.
R R
1. 0 dx = 22. cot x dx =
R
2. 1 dx = sin2 x dx =
R
23.
R
3. 5 dx = 24.
R
cos2 x dx =
x4 dx =
R
4. 25.
R
sec2 x dx =
R
5. x dx = 26.
R
csc2 x dx =
R 1
6. dx =
tan2 x dx =
R
x 27.
R 1
7. x3
dx =
cot2 x dx =
R
28.
x−5 dx =
R
8. R
29. sec x tan x dx =
R 1
9. x2 +1
dx = R
30. sec x cot x dx =
R 1
10. x3 +1
dx = R
R √ 31. csc x tan x dx =
11. x dx = R
R √ 32. csc x cot x dx =
12. 3 x dx =
R √
R
3
33. csc x tan x dx =
13. x2 dx =
R
R 2/5
34. csc x sec x dx =
14. x dx =
R x
R −1/2 35. e dx =
15. x dx =
R x2
R 1 36. e dx =
16. √ 5 x
dx =
R x
17.
R
sin x dx = 37. 2 dx =
R
18.
R
cos x dx = 38. ln x dx =
R
19.
R
tan x dx = 39. log10 x dx =
R
40. arctan x dx =
R
20. sec x dx =
R R
21. csc x dx = 41. arcsin x dx =
16
Theorem 1.12 (Integration Rules to Memorize)

Z
0 dx = C
Z
M dx = M x + C
Z
xn+1
xn dx = +C (so long as n 6= −1)
n+1
Z
1
dx = ln |x| + C (I don’t care so much about the | |)
Z x
sin x dx = − cos x + C
Z
cos x dx = sin x + C
Z
sec2 x dx = tan x + C
Z
csc2 x dx = − cot x + C
Z
sec x tan x dx = sec x + C
Z
csc x cot x dx = − csc x + C
Z
ex dx = ex + C
Z
1 Z
1
dx = arctan x + C (also √ dx = arcsin x + C)
x2 + 1 1 − x2
We know a little about how to integrate functions whose pieces are integrals
we memorize, but we don’t know as much about how to compute integrals as we
know about how to compute derivatives:
Theorem 1.13 (Linearity of Integration) Suppose f and g are integrable functions.

Then:
Z b Z b Z b
[f (x) + g(x)] dx = f (x) dx + g(x) dx;
a a a
Z b Z b Z b
[f (x) − g(x)] dx = f (x) dx − g(x) dx;
a a a
Z b Z b
[k · f (x)] dx = k f (x) dx for any constant k
a a
(and the same rules hold for indefinite integrals as well).
17
NOTE: Integration is not multiplicative nor divisive:

Z Z Z
f (x)g(x) dx 6= f (x) dx · g(x) dx
! R
Z
f (x) f (x) dx
dx 6= R
g(x) g(x) dx
18
1.4. Homework exercises
1.4 Homework exercises

In problems 1-22, evaluate the following limits. If they are equal to ∞ or −∞,
say so; if the limits fail to exist and are not infinite, say so.
1. 12.
x−1 e2x − 1
lim 2 lim
x→1 x + 1 x→0 x
2. 13.
lim 7 sin x
x→4 lim
x→0 x
3. 14.
x2 − 9
lim 2 lim ln x
x→3 x + x − 12 x→∞
4. 15.
x+2 lim ln x
lim+ 2 x→0+
x→2 x − 4
5. 16.
lim e−x lim ex
x→∞ x→0
6. 17.
1
x
+ 13 lim e1/x
lim x→0+
x→−3 1 + 1
x+2
18.
7. lim sin x
x x→∞
lim 2
x→∞
19.
8. lim arctan x
3x x→∞
lim
x→∞ x2 + 2
20.
9. lim e−1/x
−2 x→0+
lim
x→∞ x 21.
10. (
x2 x ≤ 1
2
4x − 3x + 1 lim f (x) where f (x) = .
lim x→1+ 5x x > 1
x→∞ 3 − 2x2
11. 22.
4x2 |x|
lim x lim
x→∞ e − 2 x→0 x
dy
23. Find dx
if y = x3 − 3x4 + 2.
24. Find f 0 (x) if f (x) = x1 .
19
25. Find the derivative of f (x) = 5x + 7.

26. If f (x) = 2x4 − 3x2 + 5x + 7, find f 0 (−1).
27. Find f 0 (8) if f (x) = 4
3 x.
√
28. If the position of an object at time t is given by f (t) = (3 ln t)2 , find the velocity
and acceleration of the object at time t = e.
ex
29. If f (x) = 3 ln x + 5
, find f 00 (x).
x2 −1
30. If f (x) = x
, find f 0 (x).
√ 1 dy
31. Let y = 3 x − x3/7 + 3 2.
√
x
Find dx
.
sin x df
32. Let f (x) = 7ex − 4
. Find dx
.
d
33. Find dx
(2 tan x − 3 cos x).
34. Find the derivative of f (x) = 4 cot x + 2 sec x − 1.
35. Find f 0 (x) if f (x) = e4x .
36. Find h0 (x) if h(x) = x
tan x
.
√
37. Find the derivative of f (x) = sin x.
dy
38. Find dx
if y = 2x4 sin x.
39. Find the second derivative of f (x) = cos(3x2 ).
40. Find the ninth derivative of f (x) = 4 sin x.
x2 −3x+4
41. Find f 0 (x) if f (x) = 2x3 −x+1
.
42. Find the derivative of f (x) = x ln(3x2 + 1).
43. Find the derivative of f (x) = 14 arctan x2 .
44. Find the zeroth derivative of f (x) = 8x7 + 2 cot 4x.
45. Find the slope of the line tangent to y = 2x3 − 3x when x = 2.
46. Find the equation of the line tangent to y = 4x(x2 − 3)4 when x = 2.
dy
47. Find dx
if y = xex .
48. Find f 0 (0) if f (x) = |x|.
49. Let y = 4x2 . Find dy.
20
50. Let y = 2 sin 3x. Find dy.
In problems 51-60, evaluate the given integral.

R1 R (x−1)2
51. 0 2x3 dx 56. x
dx
(3x2 − 2x7 + 5) dx
R
52.
R
57. 0 dx
R √x 5
R π/2
53. 3
+ x4
dx 58. 0 3 cos x dx
6 sec2 x dx
R 5 R
54. x
dx 59.
R −6
(4ex + 1) dx
R
55. 60. x2 +1
dx
R
61. Compute (sec x tan x + csc x cot x) dx.
R √
62. Compute 03 9 − x2 dx.
Hint: Think about the shape of the graph of the integrand.
R4
63. Compute −3 |x| dx.
Hint: Think about the shape of the graph of the integrand.
64. Find the derivative of f (x) = x2 sin x. What does your answer tell you about
the value of Z π
2x sin x + x2 cos x dx?
0
Rx
65. Find the derivative of the function F (x) = 2 3t2 dt.
Hint: Use the part of the Fundamental Theorem of Calculus not mentioned
in these notes.
66. Find the area under the graph of the function f (x) = √8 from x = 1 to x = 9.
x
67. Suppose that an object’s velocity at time t is given by v(t) = 4 sin t. Find the
displacement of the object between times 0 and 2π 3
.
68. Suppose that the acceleration of an object is given by a(t) = 48t. If the velocity
of the object at time 0 is 12 and the position of the object at time 0 is 10, what
is the position of the object at time 3?
21
Answers
1. 0 25. 5
2. 7 26. 3
−1
3. 6 27. 12
7
18
28. v(e) = ; a(e) = 0
4. ∞ e
−3 ex
29. x2
+ 5
5. 0
x2 +1
1
30. x2
6. 9
31. 3
√
2 x
− 37 x−4/7 − 23 x−5/3
7. ∞ cos x
32. 7ex − 4
8. 0
33. 2 sec2 x + 3 sin x
9. 0 34. −4 csc2 x + 2 sec x tan x
10. −2 35. 4e4x
tan x−x sec2 x
11. 0 36. tan2 x
37. √1 cos x
12. 2 2 sin x
13. 1 38. 8x3 sin x + 2x4 cos x
14. ∞ 39. −36x2 cos(3x2 ) − 6 sin(3x2 )

40. 4 cos x
15. −∞
(2x−3)(2x3 −x+1)−(6x2 −1)(x2 −3x+4)
41. (2x3 −x+1)2
16. 1
6x2
42. ln(3x2 + 1) + 3x2 +1
17. ∞
1 1
43. 8 (x/2)2 +1
18. DNE
π
44. 8x7 + 2 cot 4x
19. 2
45. 21
20. 0
46. y = 64 + 1152(x − 2)
21. 5
47. ex + xex
22. DNE 48. DNE
23. 3x2 − 12x3 49. dy = 8x dx
−1
24. x2
50. dy = 6 cos 3x dx
22
1
51. 2
61. sec x − csc x + C
52. x3 − 41 x8 + 5x + C 62. 9π
4
53. 2 3/2
9
x − 35 x−3 + C 63. 25
2
54. 5 ln x + C
64. f 0 (x) = 2x sin x + x2 cos x; it tells
55. 4ex + x + C you that the value of the integral
π
1 2
is [x2 sin x]0 = 0.
56. 2
x − 2x + ln x + C
65. 3x2
57. C
58. 3 66. 32
59. 6 tan x + C 67. 6
60. −6 arctan x + C 68. 110
23
Chapter 2
Basic integration techniques
Recall: we know from Calculus I that to evaluate an integral like

Z b
f (x) dx
a
we use the Fundamental Theorem of Calculus, which says that

Z b
f (x) dx = [F (x)]ba = F (b) − F (a)
a
where F is any antiderivative of f , i.e. F 0 = f . This means that we need to learn

various methods to compute antiderivatives (so that we can evaluate integrals).
R
Recall also that the indefinite integral of a function f , denoted f (x) dx, is the
set of all antiderivatives of f . From Calculus I we know that if F is any single an-
tiderivative of f , then all antiderivatives of f are given by F (x) + C where C is an
arbitrary constant.
In the next two chapters we will discuss various methods to compute antideriva-
tives/integrals. The general idea is that there are a variety of techniques which
work in various circumstances; one needs to learn which technique to use in which
situation.
24
2.1. Integrals to memorize
2.1 Integrals to memorize

In Calculus I you learn many derivative rules, such as
d 2 d
(x ) = 2x and (sin x) = cos x.
dx dx
Every derivative rule you learn is also an integration rule; for example the above
two rules translate as
and
We also learn the following rules, called the linearity rules for integration, which
allow us to split integrals into pieces:
Theorem 2.1 (Linearity of Integration) Suppose f and g are integrable functions.

Then:
Z b Z b Z b
[f (x) + g(x)] dx = f (x) dx + g(x) dx;
a a a
Z b Z b Z b
[f (x) − g(x)] dx = f (x) dx − g(x) dx;
a a a
Z b Z b
[k · f (x)] dx = k f (x) dx for any constant k;
a a
Z Z Z
[f (x) + g(x)] dx = f (x) dx + g(x) dx;
Z Z Z
[f (x) − g(x)] dx = f (x) dx − g(x) dx;
Z Z
[k · f (x)] dx = k f (x) dx for any constant k.
WARNING: Integration is not multiplicative nor divisive:

Z Z Z
f (x)g(x) dx 6= f (x) dx · g(x) dx
! R
Z
f (x) f (x) dx
dx 6= R
g(x) g(x) dx
25
Example 1: Evaluate the following integral:

Z
−1
csc2 x dx
4

Z 1
x

2 − + x9 dx
0 3

!
Z
4 8
√ − dx
3
x x

Z
4 sec2 x − 2x4 + 7ex dx
26
Here is a list of integration rules (discussed in Chapter 1) which, together with

linearity, allows you to do most easy integrals:
Theorem 2.2 (Integration Rules to Memorize)

Z
0 dx = C
Z
M dx = M x + C
Z
n xn+1
x dx = + C (so long as n 6= −1)
n+1
Z
1
dx = ln |x| + C (I don’t care so much about the | |)
Z x
sin x dx = − cos x + C
Z
cos x dx = sin x + C
Z
sec2 x dx = tan x + C
Z
csc2 x dx = − cot x + C
Z
sec x tan x dx = sec x + C
Z
csc x cot x dx = − csc x + C
Z
ex dx = ex + C
Z
1
dx = arctan x + C
x2 + 1
Z
1
√ dx = arcsin x + C
1 − x2
27
2.2. Rewriting the integrand
2.2 Rewriting the integrand

It is often useful to rewrite the integrand using algebra, a log rule or a trig iden-
tity.

Z
tan2 x dx

Z
(x2 − 1)2
dx
x

Z
ln(2x ) dx
28
2.3. Elementary u-substitutions
Useful log/exp rules and trig identities for rewriting integrands
ln(AC) = ln A + ln C tan2 x = sec2 x − 1

A

ln = ln A − ln C cot2 x = csc2 x − 1
C
sin x
ln(AC ) = C ln A tan x =
cos x
ln A cos x
logb (A) = cot x =
ln b sin x
1 + cos 2x
AB = eB ln A 2
cos x =
2
2 1 − cos 2x
sin x =
2
2.3 Elementary u-substitutions

In Calculus I you learn how to compute integrals using a u−substitution:
Theorem 2.3 (Integration by u−substitution - Indefinite Integrals)

Z Z
0
f (g(x)) · g (x) dx = f (u) du
by setting u = g(x).
Theorem 2.4 (Integration by u−substitution - Definite Integrals)

Z b Z g(b)
0
f (g(x)) · g (x) dx = f (u) du
a g(a)
by setting u = g(x).
Notice that for this formula to work, the integrand must have the form
f 0 (g(x)) · g 0 (x).
That means that for an elementary u−substitution to be a valid integration tech-
nique, the integrand must consist of two terms multiplied together, where one
29
term is the derivative of part of the other term.

Z
2
xe−x dx

Z
x
dx
(x2 + 1)2
30

Z
7x5 sec2 x6 dx

Z π/4
ecos x sin x dx
0
31
R
Example 12: We know that cos x dx = sin x + C. Evaluate the following inte-
grals: Z
cos 2x dx
−5
Z
cos x dx
3
Z
cos (7x − 4) dx
Z
cos (3 − x) dx
32
The technique in Example 12 is completely general, and it is a good idea (to

save time) to remember the pattern. I call this kind of reasoning the Linear Re-
placement Principle:
Theorem 2.5 (Linear Replacement Principle) Suppose you know

Z
f (x) dx = F (x) + C.
Then for any constants m and b,

Z
1
f (mx + b) dx = F (mx + b) + C.
m

Z
e6x dx

x
Z
sin dx
2

Z
(3x − 2)13 dx

Z
csc2 4x dx
33
Sometimes you need to rewrite an integral before using a u−substitution:

Z
tan x dx

Z
sin2 x dx
The following “power reducing identities” are helpful to evaluate sin2 x dx

R
Note:
2
R
and cos x dx:
1 − cos 2x 1 + cos 2x
sin2 x = cos2 x =
2 2
34
2.4. More complicated u-substitutions
2.4 More complicated u-substitutions

We have seen how to perform u−substitution when the integral consists of two
terms multiplied together. In some cases, one can also perform u−substitution
when the integral consists of three terms multiplied together, so long as one term
is the derivative of part of one of the other terms.

Z √
x3 x2 + 1 dx
To execute a more complicated u−substitution, we think of the integrand as

being three terms multiplied together as follows:
35

Z √
x2 x + 3 dx
Examples 19 and 20 above illustrate a general principle. To evaluate integrals

of the form Z
xn (mx ± b)q dx
where n is an integer but q is not a positive integer, use the u−substitution u =

mx ± b.
36

Z
x+5
dx
2x + 1

Z
x2 + 4
dx
x+2
37
Sometimes you can use trig identities to rewrite an integrand before using a
complicated u−substitution:

Z
sin4 x cos3 x dx
Integrals which contain powers of tangent and secant, or contain powers of

cosecant and cotangent, are handled similarly.
38
Here are two last examples which illustrates how if you are just willing to try
something, sometimes things work out:

Z
1
√ dx
x x−1
39

Z q √
9− x dx
40

In Problems 1-33, compute the following integrals by hand:
1. 12. 24.
Z
2
Z
sin x Z 1 √
t2 t − dt √ dx x x + 1 dx
t cos x
0
2. 13. Z
Z 4 √ e7−3x dx 25.
(3 − x) x dx Z
0
14. x2 (x + 3)8 dx
3. Z
1
Z dx
(1 + 2x ) dx 2 2 7x + 1 26.
Z
15. Z cos2 3x dx
4. sin 4x dx
2
1
Z
x 1+ dx 16. 27.
x
2
Z Z
5. cos x dx sin2 4x dx
Z π/3 3
3 cot2 x dx 17.
π/4 Z 28. Z
2 sec2 3x dx sin5 x dx
6.
Z
(x − 7)2 18. 29.
dx
Z
x/2
x e dx Z
cos3 x sin x dx
7. 19.
Z √ 2
Z
x x3 − 1 dx csc πx cot πx dx 30.
Z
8. 20. sin4 2x cos 2x dx
Z
x2 + 1 Z
3x + 2
dx dx
x3 + 3x − 2 5x − 1 31.
9. 21. Z
1 + sin x
2x
Z
1 dx
Z
e dx cos x
dx (4x + 5)3
e2x + 1
32.
10. 22. Z
1
Z 3 2x
e +1
Z 3
x2 dz
dx dx cos z − 1
0 e2x 2 x−1
11. 23. 33.
2x ln x2
Z Z Z
cot x dx dx dx
x−4 x
41
R
34. a) Evaluate the integral sin x cos x dx by performing the u−substitution
u = sin x.
b) Evaluate the integral of part (a) by performing the u−substitution u =
cos x.
c) Reconcile the (apparently) different answers you get to parts (a) and (b)
of this question.
Answers
Note: with indefinite integrals, answers can sometimes vary a little bit.

1 4 3 2
1. 4
t − t2 + C 16. 2
sin 3
x +C
16
2. 5 17. 2
tan 3x + C
3
4 5
3. 5
x + 43 x3 + x + C 18. 2ex/2 + C
1 2
4. 2
+ 2x + ln x + C
x 19. −1
csc πx + C
π
√
5. 3 − 3 − π4 20. 3
5
x + 13
25
ln(5x − 1) + C
1 2 −1
6. 2
x − 14x + 49 ln x + C 21. 8
(4x + 5)−2 + C
2
7. 9
(x3 − 1)3/2 + C 22. 7
2
+ ln 2
1
8. 3
ln(x3 + 3x − 2) + C 23. 2x + 8 ln(x − 4) + C
1
√
9. 2
ln(e2x + 1) + C 4
24. 15 4
+ 15 2
10. 7
2
− 12 e−6 25. 1
11
(x + 3)11 − 53 (x + 3)10 + (x + 3)9 + C
1 1
11. ln(sin x) + C 26. 2
x + 12
sin 6x + C
√ 1 1
12. −2 cos x + C 27. 2
x − 16
sin 8x + C
−1 7−3x −1
13. 3
e +C 28. 5
cos5 x + 32 cos3 x − cos x + C
1 −1
14. 7
ln(7x + 1) + C 29. 4
cos4 x + C
−1 1
15. 4
cos 4x + C 30. 10
sin5 2x + C
31. − ln(1 − sin x) + C

Hint: First, rewrite the integrand by multiplying through the numerator and
denominator by (1 − sin x).
32. csc z + cot z + C
42
Hint: First, rewrite the integrand by multiplying through the numerator and
denominator by (cos z + 1).
33. ln2 x + C
Hint: First, rewrite ln x2 as 2 ln x.
34. (a) 21 sin2 x + C; (b) −1

2
cos2 x + D; (c) Set the two answers equal to one an-
other; rewriting this equation gives 21 = D − C so since C and D are arbitary
constants, as long as they are related by 12 = D − C, the answers reconcile.
43
Chapter 3
Intermediate integration
techniques
3.1 Integration by parts

Motivation:
d 7
x sec x =
dx
More generally, the Product Rule tells us
(f g)0 (x) =
Rewrite the Product Rule as follows:
44
3.1. Integration by parts
Theorem 3.1 (Integration by Parts (IBP) Formula)

Z Z
u dv = uv − v du.
R
The IBP formula
R
is used (usually) Rto find integrals of the form u dv (you choose
u and dv) where v du is easier than u dv.
R
Example 1: x sin x dx
Thought Process:
Solution:
x2 ex dx
R
Example 2:
Solution:
45
More generally, to handle integrals like xn f (x) dx where f (x) is exponential,

R
sine or cosine, you would (at least theoretically) use integration by parts n times.
R1 2 x
Example 3: x e dx
0
Solution:
R
Example 4: ln x dx
Solution:
R R
(The integrals arctan x dx and arcsin x dx are similar.)
Parts vs. u-Substitutions

Z Z
x sin x dx vs. sin x cos x dx
46
How to choose u and dv

Suppose you know you want to use integration by parts. How do you decide
which part of the integral should be u and which should be dv?
R R
General principle: The integral v du should be easier than u dv. So think
ahead!
Specific guidelines:
HIGH PRIORITY u (choose these to be u if possible):
MEDIUM PRIORITY u:
LOW PRIORITY u (avoid choosing these as u if possible):
Example: For each of the following integrals, decide if integration by parts is

an appropriate method to try. If so, write what you would choose for u and dv:
3x2 arctan x dx
R
(a)
R
(b) x ln x dx
R
(c) 2x sin 4x dx
R 3x
(d) e cos 2x dx
R ln x
(e) x
dx
R ln x
(f) x2
dx
4x3 e2x dx
R
(g)
R x2 +3x−3
(h) 2x+3
dx
R 2x+3
(i) x2 +3x−3
dx
47
3.2. Partial fractions
3.2 Partial fractions

Motivation: Consider the following expression:
2 1
+
x−3 x−1
This is equal to
2(x − 1) 1(x − 3) [2x − 2] + [x − 3] 3x − 5
+ = = 2 .
(x − 3)(x − 1) (x − 1)(x − 3) (x − 1)(x − 3) x − 4x + 3
This means Z
2 1 3x − 5
Z
+ dx = dx
x−3 x−1 x2 − 4x + 3
3x−5 2 1
Goal: Rewrite expressions like x2 −4x+3
as x−3
+ x−1
.
Definition 3.2 A polynomial in x is an expression of the form a0 + a1 x + a2 x2 +

... + an xn where a0 , a1 , ..., an are constants. The degree of a polynomial is the highest
power of x that appears. A polynomial is called irreducible if it cannot be factored
into two polynomials, both of lower degree.
Examples:
• x4 − 3x + 1 is a degree 4 polynomial
• x2 − 3x9 − 7x + 5x6 is a degree 9 polynomial
7 √
• x x+2 , x, x1 , sin x, 4xπ , ... are not polynomials
• x2 + A is an irreducible polynomial if A > 0
• x2 − 4 is not irreducible
• Ax ± B is irreducible
Theorem 3.3 (Fundamental Theorem of Algebra) No polynomial of degree ≥ 3

is irreducible.
In the same way that every whole number factors into a product of primes, we
have:
48
Theorem 3.4 Every polynomial factors into a product of irreducibles.
A precise statement of the problem: Given a rational function p(x)

q(x)
(a rational
function is the quotient of any two polynomials) with degree(p) < degree(q), write
p(x)
q(x)
as
p(x) p1 (x) p2 (x) pn (x)
= + + ... +
q(x) q1 (x) q2 (x) qn (x)
where:
• degree(pj ) < degree(qj ) for all j; and
• every qj (x) is a power of an irreducible polynomial, i.e.
qj (x) = ax ± b or qj (x) = (ax ± b)r or
qj (x) = (ax2 ± bx ± c)r where ax2 ± bx ± c doesn’t factor

p(x)
This procedure is called splitting q(x)
into partial fractions or finding the par-
p(x)
tial fraction decomposition of q(x)
.
Example: Find the partial fraction decomposition of

1
.
x2 −1
STEP 1: Factor the denominator completely
STEP 2: “Guess” the form of the decomposition
STEP 3: Write the equation from step 2 so that it has common denominator
STEP 4: Clear the denominator from the equation obtained in step 3
49
STEP 5: Find the unknowns A, B, C, ... (two methods)
STEP 6: Write answer
Remark: There is no calculus in the partial fraction decomposition itself (you

do no limit, derivative or integral in steps 1-6 above). This is really an algebraic
technique, whose most common use is to rewrite integrands.
50
Example: Evaluate the following integral:

Z
3x2 + 10x − 24
dx
x3 + 2x2 − 8x
51
Example: Find the partial fraction decomposition of
2x3 − x2 − x + 1
x2 (x − 1)2
52
Question: In a partial fraction problem, what should your “guess” be in Step

2?
It is hard to write down a “one-size fits all” statement; here are some examples:
p(x)
q(x)
GUESSED FORM OF DECOMPOSITION
6x−1
(x−5)(x+2)
−4
x(x−3)(x+4)
2
(x−4)2 (x+6)
7
x4 (x−1)3 (x+2)
x−1
(x2 +5)(x−2)
x3 −3x+4
(x2 +1)3 x2 (x−7)(x2 +6)
General guidelines:
1. Make sure the denominator is factored completely first.
2. For every linear term (x ± a)r in the denominator, the guessed form needs r
terms of the form
A B C const
+ + + ... +
x ± a (x ± a)2 (x ± a)3 (x ± a)r
3. For every unfactorable quadratic (x2 +ax+b)r in the denominator, the guessed
form needs r terms of the form
Ax + B Cx + D Ex + F const x + const
+ 2 + 2 + ... +
x2 + ax + b (x + ax + b)2 (x + ax + b) 3 (x2 + ax + b)r
53
Remark: Partial fractions is a useful technique to evaluate integrals under the

following circumstances:
1. The integrand must be a rational function (i.e. contain only nonnegative
powers of x and no trig/exp/log functions).
2. The degree of the numerator must be less than the degree of the denominator.
3. The denominator must be easily factored.
4. The numerator must be “sufficiently complicated”.
Example: Z
x
dx
(x − 2)3
Solution # 1: Use partial fractions:
x A B C
= + +
(x − 2)3 x − 2 (x − 2) 2 (x − 2)3
x A(x − 2)2 B(x − 2) C
= + +
(x − 2)3 (x − 2)3 (x − 2)3 (x − 2)3
x = A(x − 2)2 + B(x − 2) + C
x = A(x2 − 4x + 4) + Bx − 2B + C
x = Ax2 + (B − 4A)x + (4A − 2B + C)
Thus A = 0, B − 4A = 1 and 4A − 2B + C = 0. That means B = 1 and C = 2. So

!
Z
x Z
1 2 −1 −1
dx = + dx = + + C.
(x − 2)3 (x − 2)2 (x − 2)3 x − 2 (x − 2)2
Solution # 2:
54
Example: Evaluate the following integral:

Z
5x2 + 20x + 6
dx
x3 + 2x2 + x
Elementary integral formulas useful in conjunction with partial

fractions
Each term in a partial fraction decomposition usually works out to be something
like one of these terms below:
Z
1
dx = ln |x ± a| + C
x±a
Z
1 1
dx = ln |ax ± b| + C
ax ± b a
1 −1 1
Z
n
dx = + C (if n 6= 1)
(x − a) n − 1 (x − a)n−1
Z
1
dx = arctan x + C
x2 + 1
Z
1 1 x
2 2
dx = arctan + C
x +a a a
55
3.3. Summary of integration techniques
3.3 Summary of integration techniques

1. Check to see if the integral is one you can “just write the answer to”.
2. If the integrand contains terms which are added or subtracted, consider using
linearity rules to split the integrand into pieces you can “just write the answer
to”.
3. See if you can rewrite the integrand using algebra, log rules or a trig identity.
4. If the integrand is a function you can integrate by hand with a mx ± b instead

of an x, use the Linear Replacement Principle.
5. If the integrand contains terms which are multiplied together, where one
part is the derivative of something in the other part (up to a constant), use
a u−substitution.
6. If the integrand has terms multiplied together which are unrelated, try inte-
gration by parts.
7. If the integrand is a rational function where the denominator factors (and the
degree of the denominator is greater than the degree of the numerator), use
partial fractions to decompose the integrand.
8. If all else fails, use Mathematica (see below).
3.4 Mathematica commands for integration

To find the indefinite integral of a function:
Integrate[f unction, x]
To find the exact value of a definite integral of a function from a to b:

Integrate[f unction, {x,a,b}]
To find a decimal approximation to a definite integral of a function from a to b:

NIntegrate[f unction, {x,a,b}]
To find a partial fraction decomposition (without integrating):

Apart[expression]
56

In Problems 1-10, evaluate the following integrals by hand.
1. Z 6. Z
−3x
xe dx x cos x dx
2. 7. Z π/2
Z
4x
dx x sin 2x dx
ex 0
3. Z 8. Z 1
3 x4
x e dx 2xe−x/2 dx
0
4. Z 9. Z
x ln(x + 5) dx arctan x dx
5. Z 10. Z
−2
x ln x dx x arctan x dx
In Problems 11-19, determine whether it is better to use parts or a u−substitution

to evaluate the integral. If parts are better, write “PARTS” and give your choices of
u and dv. Otherwise, write “u−SUB” and give your choice of u. You do not have
to evaluate these integrals.
11. 16.
Z
x arctan x
√
Z
dx dx
2 + 3x x2 + 1
12. Z
x2 cos x dx 17. Z
1
dx
13. Z x ln x
x ln2 x dx
18. Z
14. Z 3x2 e2x−1 dx
arcsin x dx
15. Z 19. Z
sin x cos x dx ex cos(3x) dx
In Problems 20-23, write the “guessed” form of the partial fraction decomposi-
tion of the given expression. You do not need to solve for the constants.
1 1 A B
Example: Solution: = +
x2 − 1 x2 − 1 x−1 x+1
57
4 3x
20. x2 −3x
22. x2 +4x−5
7x+2 2
21. x3 −5x2 −24x
23. 3x2 +9x−54
In Problems 24-27, evaluate the integrals by hand.
24. 26.
Z
1
Z
5−x
dx dx
x2 − 16 2x2 + x − 1
25. 27.
Z
3 Z
5x2 − 12x − 12
dx dx
x +x−2
2 x3 − 4x
In Problems 28-33, write the “guessed” form of the partial fraction decomposi-
tion of the given expression. You do not need to solve for the constants.
2x+3 1
28. x(x2 +1)
31. (x−2)2 (x2 +4)2
4 x+1
29. (x−3)(x+2)2 32. x3 (x2 +x+1)(x−7)
5−x2 7
30. x3 (x−1)2
33. (x+4)3
In Problems 34-37, evaluate the integrals by hand.
34. 36.
Z
4x2 + 2x − 1
Z
x2 − x + 9
dx dx
x3 + x2 (x2 + 9)2
35. 37.
Z
x2
Z 5
x−1
dx dx
x4 − 2x2 − 8 1 x2 (x + 1)
integral 2x−3
R
38. a) Evaluate the (x−1)3
dx by performing a u−substitution.
b) Evaluate the integral from part (a) via partial fractions.
c) Verify that the answers you get to parts (a) and (b) are equal.
d) Evaluate the integral using Mathematica. Which method do you think
Mathematica used to compute the integral?
39. In the “Elementary integral formulas used with partial fractions” subsection
on p. 54, we asserted that
Z
1 1 x
dx = arctan + C.
x2 +a 2 a a
Show that this formula is correct. Hint: Rewrite the integrand by factoring
out a2 from the denominator and then use the Linear Replacement Principle.
58
Answers
−1
1. 3
xe−3x − 19 e−3x + C
2. −4xe−x − 4e−x + C
1 x4
3. 4
e +C
−1 3
4. 4
x + 52 x + 12 x2 ln(x + 5) − 25
2
ln(x + 5) + C
−1
5. x
− x1 ln x + C
6. x sin x + cos x + C
π
7. 4
8. 8 − 12e−1/2
9. x arctan x − 21 ln(x2 + 1)
1 2 −1
10. 2
x arctan x + 2
x + 12 arctan x + C
11. u-SUB; u = 2 + 3x
12. PARTS; u = x2 ; dv = cos x dx
13. PARTS; u = ln2 x; dv = x dx
14. PARTS; u = arcsin x; dv = dx
15. u-SUB; u = sin x or u = cos x
16. u-SUB; u = arctan x
17. u-SUB; u = ln x
18. PARTS; u = 3x2 ; dv = e2x−1 dx
19. PARTS; u = ex ; dv = cos 3x dx (or vice versa)

4 A B
20. x2 −3x
= x
+ x−3
7x+2 A B C
21. x3 −5x2 −24x
= x
+ x−8
+ x+3
3x A B
22. x2 +4x−5
= x+5
+ x−1
2 A B
23. 3x2 +9x−54
= x+6
+ 3x−9
1
24. 8
ln(x − 4) − 81 ln(x + 4) + C
59
25. ln(x − 1) − ln(x + 2) + C

3
26. 2
ln(2x − 1) − 2 ln(x + 1) + C
−1
27. 4
ln x + 18 ln(x − 2) + 18 ln(x + 2) + C
A Bx+C
28. x
+ x2 +1
A B C
29. x−3
+ x+2
+ (x+2)2
A B C D E
30. x
+ x2
+ x3
+ x−1
+ (x−1)2
A B Cx+D Ex+F
31. x−2
+ (x−2)2
+ x2 +4
+ (x2 +4)2
A B C Dx+E F
32. x
+ x2
+ x3
+ x2 +x+1
+ x−7
A B C
33. x+4
+ (x+4)2
+ (x+4)3
1
34. x
+ 3 ln x + ln(x + 1) +C
√
35. 6
2
arctan √x2 + 16 ln(x − 2) − 61 ln(x + 2) + C
36. 1
2
(x2 + 9)−1 + 13 arctan x3 + C
−4
37. 5
+ ln 25
9
38. a) −2(x − 1)−1 + 12 (x − 1)−2 + C

b) 1
2
(x − 1)−2 − 2(x − 1)−1 + C
c) They are clearly equal.
d) It isn’t clear, because Mathematica simplifies the answer so much.
39. Follow the hint:

Z
1 Z
1 1
2 2
dx = 2
· x2 dx
x +a a a2 + 1
1 Z 1
= 2 2 dx
a x
+ 1
a
" #
1 1 x
= 2 · 1 arctan + C (by the Lin. Repl. Principle)
a a
a
1 x
= 2 · a arctan + C
a a
1 x
= arctan + C.
a a
60
Chapter 4
Improper integrals
4.1 Boundedness vs. unboundedness

Rb
Recall Let f be continuous and let a, b ∈ R. Then a f (x) dx gives the area under
f from x = a to x = b.
Since f is continuous on [a, b], it must be that the graph of f encloses a bounded
region (in the sense that one can draw a box around it):
Rb
This guarantees that a f (x) dx < ∞.
New question: Can you compute the area of an unbounded region?
A typical unbounded region has infinite area:
61
4.1. Boundedness vs. unboundedness
But some unbounded regions have finite area! Here is an example:
Notice that in the previous picture, the region obtained was unbounded, but
in the direction it was unbounded, the width/thickness of the region decreases
to zero. Another situation where the width/thickness of an unbounded region
decreases to zero is when the region is described by a graph with an asymptote:
“Horizontally unbounded region” “Vertically unbounded region”
Better question: How are such unbounded regions described?
62
4.2. Horizontally unbounded regions
4.2 Horizontally unbounded regions

Definition 4.1 Improper integrals for horizontally unbounded
R∞
regions] Let f (x) be
continuous on [a, ∞). Define the improper integral a f (x) dx to be
Z ∞ Z b
f (x) dx = lim f (x) dx.
a b→∞ a
this limit exists and is equal to a finite number L, we write a∞ f (x) dx = L nd say
R
If
R∞
Ra∞
f (x) dx converges (to L). If this limit does not exist or is equal to ±∞, we say
a f (x) dx diverges.
A picture to explain the definition:
Remark: If a∞ f (x) dx converges for a nonnegative function f (i.e. f (x) ≥ 0 for

R
all x), it must be the case that

lim f (x) = 0.
x→∞
The converse of this statement is false, as is seen in the following example:
Example: Compute the following improper integral:

Z ∞
1
dx.
1 x
3
1 b
63
Example: Compute the following improper integral:

Z ∞
1
dx.
1 x2
3
1 b
Now we generalize the results of the preceding two examples:
Example: Let p > 0 be a constant. Compute the following improper integral:

Z ∞
1
dx.
1 xp
From the previous two examples we know:

When p = 1, the integral diverges (i.e. the area under the curve is infinite).
When p = 2, the integral converges to 1 (i.e. the area under the curve is 1).
64
Solution for p 6= 1, p > 0:
The results of the previous problem are summarized in the following theorem,
which should be memorized:
Theorem 4.2 (Convergence/divergence of p-integrals) The improper integral

Z ∞
1
dx
1 xp
1
converges to p−1
if p > 1, but diverges if p ≤ 1.
65
4.3. Vertically unbounded regions
Example: Determine whether or not the following improper integral converges

or diverges: Z ∞
1
dx
e x ln3 x
4.3 Vertically unbounded regions

If the region under consideration is unbounded in a vertical direction, one
needs to be more careful in rewriting the improper integral as a limit. There are
three situations; the first situation is when the vertical asymptote is on the right-
hand edge of the region whose area you want to compute:
66
Definition 4.3 (Improper integrals for vertically unbounded regions I) (VA on

right-hand edge of region) Let f (x) be continuous on [a, c) and suppose lim− f (x) =
Rc x→c
±∞. Define the improper integral a f (x) dx to be
Z c Z b
f (x) dx = lim− f (x) dx.
a b→c a
this limit exists and is equal to a finite number L, we write ac f (x) dx = L and say
R
If
Rc
Rac
a b c
The next situation is when the vertical asymptote is on the left-hand edge of the
region whose are you want to compute:
Definition 4.4 (Improper integrals for vertically unbounded regions II) (VA on
the left-hand edge of the region) Let f (x) beR continuous on (a, c] and suppose
lim+ f (x) = ±∞. Define the improper integral ac f (x) dx to be
x→a
Z c Z c
f (x) dx = lim+ f (x) dx.
a b→a b
this limit exists and is equal to a finite number L, we write ac f (x) dx = L and say
R
If
Rc
Rac
67
a b c
The last situation is more complicated. If a region is bounded by a graph with

a vertical asymptote in the middle of the region, then the improper integral splits
into two parts which must be studied separately:
Definition 4.5 (Improper integrals for vertically unbounded regions III) (VA
in middle of region) Let f (x) be continuous on [a, c] except at a single point b ∈
(a, c). Suppose lim+ f (x) = lim− f (x) = ∞ or lim+ f (x) = lim− f (x) = −∞.
x→b
Rx→b x→b x→b
the improper integral ac f (x) dx is said to converge only if ab f (x) dx and
R
Then
Rc
b f (x) dx converge (in the sense of the previous definitions). In this case we set
Z c Z b Z c
f (x) dx = f (x) dx + f (x) dx
a a b
where the two integrals on the right side are defined as in the preceding definitions.
a b c
Key concept: Must consider the two regions above separately.
68

or diverges:
Z π/2
tan x dx
0
69

or diverges: Z ∞
1
√ dx
0 x(x + 1)
70
4.4. Theoretical approaches
4.4 Theoretical approaches

Frequently in mathematics we encounter quantities (such as improper inte-
grals) where it is just as important (if not moreso) to determine whether or not the
quantity converges than it is to determine what number the quantity converges to.
It is useful to know some “tricks” based on theoretical concepts which help tell us
whether or not an improper integral converges.
The first “trick” is that when integating over a horizontally unbounded region,
the starting index of the integral doesn’t affect whether or not the integral con-
verges.
Theorem 4.6 (Starting index is irrelevant to convergence/divergence) Suppose

a < b and f is continuous on [a, b]. Then:
R∞ R∞
1. if a f (x) dx converges, so does b f (x) dx;
R∞ R∞
2. if a f (x) dx diverges, so does b f (x) dx.
A picture to explain:
Note: This theorem does not say that

Z ∞ Z ∞
f (x) dx = f (x) dx.
a b
If f is positive and a < b, then (assuming these integrals converge) it is clear that
Z ∞ Z ∞
f (x) dx > f (x) dx
a b
because you are integrating a positive function over a larger region.
71
The next set of results describe how one can “split up” improper integrals
whose integrands are composed of terms added/subtracted together:
R∞
Theorem 4.7 (Linearity of improper integrals I) Suppose
R∞ a f (x) dx = L and
a g(x) dx = M . Then:
R∞
1. a [f (x) + g(x)] dx = L + M ;
R∞
2. a [f (x) − g(x)] dx = L − M ;
R∞
3. a kf (x) dx = kL for any constant k.
R∞
Theorem 4.8 (Linearity of improper integrals II) Suppose
R∞ a f (x) dx = L but
a g(x) dx diverges. Then:
R∞
1. a [f (x) + g(x)] dx diverges;
R∞
2. a [f (x) − g(x)] dx diverges;
R∞
3. a kg(x) dx diverges, for any constant k 6= 0.
4.9 (Linearity of improper integrals III) Suppose that both a∞ f (x) dx

R
Theorem
R∞
and a g(x) dx diverge. Then you Rknow nothing about whether or not the improper
integrals a∞ [f (x) + g(x)] dx and a∞ [f (x) − g(x)] dx converge or diverge.
R
The preceding three theorems can be summarized in the following three “morals”,
which we will see again and which hold throughout mathematics:
1.
2.
3.
72
Example: Determine whether or not the following improper integrals converge

or diverge:
R∞ 3 2

(a) 1
√
x
+ x3
dx
R∞ 1 3

(b) 6 5x2
− x3/2
dx
Our last theoretical result is based on a very important and very general math-
ematical idea of reasoning by way of inequalities:
Theorem 4.10 (Comparison Test for Improper Integrals) Suppose that
0 ≤ f (x) ≤ g(x)
for all x ≥ a. Then:

R∞ R∞
1. if a g(x) dx converges, so does a f (x) dx;
R∞ R∞
2. if a f (x) dx diverges, so does a g(x) dx.
A picture to explain:
73

or diverges: Z ∞
3
5
dx
1 x + 2x + 2

or diverges: Z ∞
2
√ dx
12 x−7
74

or diverges:
Z ∞
sin x2 + 3
dx
2 x8
General principles to help construct inequalities

1. Addition in the denominator. Suppose the integrand is of the form

4+?
where , 4 and ? are all positive quantities. Then, one can start with one of
these two inequalities

≤ or ≤
4+? 4 4+? ?
and try to apply the Comparison Test.
(The reason these inequalities are valid is that removing positive terms from
the denominator makes the denominator smaller, which makes the entire
fraction bigger.)
75
2. Subtraction in the denominator. Suppose the integrand is of the form

4−?
where , 4 and ? are all positive quantities. Then, one can start with this
inequality:

≥
4−? 4
(The reason this inequality is valid is that removing negative terms from the
denominator makes the denominator bigger, which makes the entire fraction
smaller.)
3. Terms containing sines or cosines. Suppose the integrand contains some

expression of the form cos or sin where is some expression. Then, one
can start with the inequality
−1 ≤ cos ≤ 1 ( or − 1 ≤ sin ≤ 1)
76

In Problems 1-6, evaluate the improper integral (if the integral diverges, say
so).
1. Z ∞ 4. Z ∞
−4 2
x dx dx
1 1 x2 +1
2. Z ∞ 5.
1 Z ∞
ln x
√ dx dx
8
3
x 1 x
3. 6. Z ∞
Z ∞
−x
2
xe dx dx
0 3 (x + 3)5/3
R∞ 1
7. In class we showed that 1 x dx diverges. Have Mathematica evaluate this
integral. Does Mathematica recognize that this integral diverges?
8. Use Mathematica to find the values of
Z ∞
xn e−x dx
0
for n = 1, 2, 3, 4 and 5. Find a pattern in these answers to make a guess as

R ∞ n −x
to what the value of 0 x e dx is for an arbitrary whole number n (your
answer should be in terms of n).
9. For each of the following improper integrals, write them as a limit or a sum
of limits:
R4 2
• Example: 0 x−1 dx
Solution: lim− 0b x−1

R 2 R4 2
• dx + lim+ B x−1 dx
b→1 B→1
R ∞ −x
• Example: 2 e dx
• Solution: lim 2b e−x dx
R
b→∞
Then, sketch a picture (similar to those on pages 66-67) which reflects how
the limit is being used to evaluate the improper integral.
R2 1
a) 1 x ln x dx
R1 x
b) 0 x2 −1 dx
R∞ 1
c) 0 xex dx
R5 1
d) 3 (x−5)2 dx
R5 1
e) 3 (x−3)2 dx
77
In Problems 10-15, evaluate the improper integral (if the integral diverges, say
so).
10. 13.
Z 1
1
Z 4
1
dx √ dx
0 x2 2 4−x
11. 14. Z ∞
Z 1 1
x ln x dx dx
0 1 x ln x
12. 15.
Z e
2
Z 3
1
ln x dx dx
0 1 x2 −1
In Problems 16-25, determine whether the improper integral converges or di-
verges (you do not necessarily need to evaluate the integral if it converges). Note:
the intent here is for you to use theory, not to perform lots of computations.
16. Z ∞ 21.
5 Z ∞
cos(πx) + 3
dx dx
1 x8 1 x4
17. Z ∞ 22. Z ∞
2 3 2

4
dx − 2 dx
6 x +1 1 x x
18. 23. Z ∞
Z ∞
1 2 5

√ dx 3
+ 4 dx
3 x−1 2 x x
19. 24. Z ∞
Z ∞ sin(2x + 1) + 4
4x −3/2
dx √ dx
2 3 x
20. Z ∞ 25. Z ∞
1 1
dx dx
1 x
e +x 8 2x − 1
78
Answers
1
1. 3
5. diverges
2. diverges 6. 31/3
22/3
3. 1
7. Yes, Mathematica recognizes that this
4. π/2 integral diverges.
R ∞ n −x
8. 1; 2; 6; 24; 120. In general, 0 x e dx = n!.
R2 1 R2 1
9. a) 1 x ln x dx = lim+ b x ln x dx
b→1
(Picture looks like the one on page 7, with the asymptote at x = 1)
R1 x Rb x
b) 0 x2 −1 dx = lim− 0 x2 −1 dx
b→1
R∞ 1 R1 1 RB 1
c) 0 xex
dx = lim+ b xex dx + lim 1 xex
dx
b→0 B→∞
(Picture looks like the example on page 10, with the asymptotes at x = 0
and y = 0)
R5 1 Rb 1
d) 3 (x−5)2 dx = lim− 3 (x−5)2 dx
b→5
R5 1 R5 1
e) 3 (x−3)2 dx = lim+ b (x−3)2 dx
b→3
10. diverges
−1
11. 4
12. 0 19. converges

√
13. 2 2 20. converges
14. diverges 21. converges
15. diverges 22. diverges
16. converges 23. converges
17. converges 24. diverges
79
Chapter 5
Applications of the integral
5.1 Area
Definite integrals were invented for the purpose of finding areas of regions.
However, there is a slight difference between computing an integral and finding
an area, because integrals can be negative (whereas areas are never negative). For-
mally speaking, integrals compute signed areas.
For example, suppose we are asked to compute the area between the graphs of
f and g between x = a and x = b where f and g are as in the following picture:
This area is
80
5.1. Area
Here is another example. Here f is the solid line and g is the dashed line; the
problem is to compute the total area enclosed by the graphs:
a d b
Example 1: Find the area between the graphs of f (x) = x2 +2 and g(x) = −x−3
on [−2, 3].
81
5.1. Area
Example 2: Find the area enclosed by the graphs of f (x) = 2 − x2 and g(x) = x.
82
5.1. Area
Example 3: Find the area enclosed by the graphs of f (x) = x3 + x2 − 6x and

g(x) = −x3 + 5x2 .
Solution: Start by finding the intersection points of the graphs:
x3 + x2 − 6x = −x3 + 5x2
2x3 − 4x2 − 6x = 0
2x(x2 − 2x − 3) = 0
2x(x + 1)(x − 3) = 0
⇒ x = 0, x = −1, x = 3
83
5.1. Area
Example 4: Find the area of the region which is above the x−axis, below y = 4,
to the left of y = (x − 1)2 and to the right of y = 2x. (The region is pictured below.)
4
-1 1 2 3 4
-1
84
5.1. Area
Example 4 again:
-1 1 2 3 4
-1
85
5.1. Area
Example 5: Write an integral

√with respect to the variable x which gives the area
1
enclosed by the graphs of y = x and y = 6 x. Write another integral with respect
to y that gives the same area.
Solution: Start by finding the intersection points of the graphs:

√ 1
x= x
√ 6
6 x=x
√
(6 x)2 = x2
36x = x2
0 = x2 − 36x
0 = x(x − 36)
⇒ x = 0, x = 36
86
5.2. Volume
5.2 Volume
Recall that in the previous section we talked about area, which is a 2-dimensional
problem:
a x b
A way to think about the computation of these areas is as follows:
1. Let x run from a to b.
2. At each x, think of an infinitely narrow rectangle. These rectangles are called

cross-sections for the region.
3. Find the length of this rectangle (call the length L(x)).
• the length depends on x

• the length is usually f (x) − g(x) where f is the top function and g is the
bottom function
4. The rectangles’ width is dx, so the area of the narrow rectangle is
5. “Add up” the areas of these rectangles to get the area of the region:
Z b
A= L(x) dx
a
87
5.2. Volume
We can repeat the reasoning on the previous page to compute volumes of 3-

dimensional solids:
1. Let x run from a to b.
2. At each x, think of an infinitely thin plane figure perpendicular to the x−axis.

These rectangles are called cross sections for the solid.
3. Find the area of this rectangle (call the length A(x)).
• the area depends on x

• use geometry formulas (area of circle, square, etc.)
• you could have to use an integral to find A(x) (especially in Math 320)
4. The thickness of the cross-sections is dx, so the volume of each cross-section

is
5. “Add up” the volunes of these cross-sections to get the volume of the solid:
Z b
V = A(x) dx
a
88
5.2. Volume
Rb
General volume formula: V = a A(x) dx where A(x) = area of the cross-
section at x
Example 1: Consider a solid whose base consists of the region between the y-
axis and the right-half of the circle x2 +y 2 = 16. Cross-sections of the solids parallel
to the y-axis are squares with bases in the xy-plane. Find the volume of the solid.
Solution: To get started, let’s draw some pictures:
BASE 3-D PICTURE CROSS-SECTION

4
-1 1 2 3 4 5
-2
-4
89
5.2. Volume
The next example illustrates a method called the disc method (volume), which
is one of three methods used to calculate volumes of solids obtained by revolving
a region around some axis.
Example 2: The region R between the graph of f (x) = x4/3 + 1 and the x−axis
on [0, 2] is revolved around the x−axis to produce a solid. Find the volume of the
solid.
BASE 3-D PICTURE CROSS-SECTION
-1 1 2 3
-1
In general, if you take the region between f and the x−axis between x = a and
x = b and revolve that region around the x−axis to produce a solid, the volume of
the solid generated is
Z b
V = π[f (x)]2 dx.
a
This formula is called the disc method because cross-sections are circles/disks.
90
5.2. Volume
Example 3: Find the volume of a solid obtained by revolving the region en-
closed by the graphs of f (x) = x2 and g(x) = x about the line y = 3.
In general, if a region is revolved around a line to produce a solid with a hole

(like a donut), the cross-sections are washers and the volume is:
Z b
V = πR2 − πr2 dx.
a
This formula is called the washer method because cross-sections are washer-shaped.
91
5.2. Volume
You can do y−integration with volume as well:

√ 1
Example 4: The region between y = x and y = 4
x is revolved around the
y−axis to produce a solid. Find its volume.
92
5.2. Volume
Example 5: A solid is formed by revolving the region enclosed by the graphs

of y = 13 x3 + 2x − 73 , x = 3 and the x−axis about the y−axis. Find its volume.
15
10
-1 1 2 3
First attempt at at solution: use washer method, integrate with respect to y.
Problem with this method:
We need a new method, which is called the shell method or cookie-cutter

method.
We will integrate “inside-out” with respect to x, and take cylindrical cross-
sections.
93
5.2. Volume
√
Example 6: The region enclosed by y = x, y = 2 and x = 0 is revolved around
the line y = −1 to produce a solid. Find its volume.
The various methods of finding volumes of solids formed by revolution of a

region about a horizontal or vertical line are summarized in the chart on the next
page.
94
5.2. Volume
95
SOLID COMES FROM REVOLUTION SOLID COMES FROM REVOLUTION
ABOUT A VERTICAL LINE x = L ABOUT A HORIZONTAL LINE y = L
shell method cross-section at x: disk/washer method cross-section at x:
INTEGRATE
WITH
RESPECT
TO x
a x b a x b
(Equations r = |x − L| R = |g(x) − L|
need to be h = |g(x) − f (x)| r = |f (x) − L|
solved for y SA(x) = 2πrh A(x) = πR2 − πr2
in terms
of x)
Rb Rb
⇒V = a 2π |x − L| |g(x) − f (x)| dx ⇒V = a [π(g(x) − L)2 − π(f (x) − L)2 ] dx
disk/washer method cross-section at x: shell method cross-section at x:
INTEGRATE
WITH
RESPECT
TO y
(Equations R = |f −1 (y) − L| r = |y − L|
need to be r = |g −1 (y) − L| h = |f −1 (y) − g −1 (y)|
solved for x A(y) = πR2 − πr2 A(y) = 2πrh
in terms
of y)
Rd Rd
⇒V = c [π(f −1 (y) − L)2 − π(g −1 (y) − L)2 ] dy ⇒V = c 2π|y − L| |f −1 (y) − g −1 (y)| dy
5.3. General principles behind all applications of integration
5.3 General principles behind all applications of integration

Before turning to our next application of integration, something more general.
All applications of integration are based on the same principle:
Theorem 5.1 (General principle of applications of integration) Suppose that A

and Q are two quantities such that if quantity Q is constant, then A is obtained from
Q by multiplying Q by the change in another quantity x, i.e.
A = Q · ∆x when Q is constant.
Then, when Q becomes a quantity that depends on x, we have

Z b
A= Q(x) dx.
a
Here are some examples, most of which we have seen before, where this gen-
eral principle is at work:
A Q formula when Q is constant integral formula

d = v · ∆t d = ab v(t) dt
R
displacement d velocity v
(t = time)
Rb
area A cross-sectional A = h · ∆x A= a h(x) dx
height h (x = length)
Rb
volume V cross-sectional V = A · ∆x V = a A(x) dx
area A (x = length)
Rb
work W force F W = F · ∆x W = a F (x) dx
(x = distance)
In the next few sections, we will use this general principle to develop formulas
which give more applications of integration.
96
5.4. Arc length
5.4 Arc length

Goal: Find the length of curves which are pieces of the graph of a function
y = f (x).
What exactly does one mean by length?

10
0 1 2 3 4 5
Problem: Find the length s of the curve y = f (x) between x = a and x = b.
Remark: in mathematics, the letter s is used for arc length. No one knows why.
Solution: Let’s suppose first that the slope of y = f (x) is constant (so that the
slope is going to be like the Q of the chart on the previous page). That means we
are assuming f (x) is...
Thus
q
s= (∆x)2 + (∆y)2
q
= (∆x)2 + (m∆x)2
q
= (∆x)2 (1 + m2 )
√
= 1 + m2 ∆x.
97
5.4. Arc length
On the previous page, we found that when the slope m is constant, the arc
length from x = a to x = b is
√
s = 1 + m2 ∆x.
That means that if the slope of y = f (x) is nonconstant (and depends on x), the arc
length from x = a to x = b is
s=
(by applying the general principle of integration applications from the previous
section). We have derived the following theorem:
Theorem 5.2 (Arc length formula) Let f be a differentiable function on [a, b]. Then
the length of the graph of y = f (x) from x = a to x = b is
Z bq
s= 1 + [f 0 (x)]2 dx.
a
Example 1: Write an integral which computes the length of the curve y = sin x
from x = 0 to x = π/2.
Example 2: Write an integral which computes the length of the curve y = ln x

from x = 1 to x = 5.
98
5.4. Arc length
Example 3: Find the length of the curve y = (4 − x2/3 )3/2 from x = 1 to x = 3.
Note: Observe that in order to get an integral that you can actually work out,
you have to start with a weird function. This is your first exposure to the notion
that integrals representing lengths of curves are harder to compute exactly than those
representing areas or volumes of regions. In fact there is an entire branch of math
research devoted to studying integrals which represent the lengths of curves; this
is the theory of elliptic integrals.
99
5.5. One-dimensional moments and centers of mass
5.5 One-dimensional moments and centers of mass

This packet deals with the computation of the center of mass of a physical sys-
tem.
Question: What is meant by “center of mass”?
Imagine you had to balance a tabletop on the top of a pyramid:
The center of mass of the table is the place where you would have to put the
point of the pyramid to balance it.
If the tabletop is circular and made of a uniform material, this is easy. You just
put the center of the table on top of the pyramid. But if the tabletop isn’t round
(or otherwise symmetric) and/or if the tabletop has nonuniform density (i.e. it is
made from a porous material with varying amounts of air pockets in it), it isn’t so
easy to see immediately where the center of mass is. We want to figure out how to
find the center of mass in these situations.
The reason for wanting to know how to compute a center of mass is that in a
physics problem, if you are studying the motion of a large object (like a comet or a
space station orbiting the earth), you can treat the object as a single point (which is
easier to study mathematically than a large blob) so long as the point you use is
the center of mass of the object.
100
Discrete masses along a line

We will start by considering the problem of finding the center of mass for a one-
dimensional physical system (that is, where the mass is distributed along a line,
i.e. instead of a tabletop we are thinking of a pencil or metal bar or stick, etc.).
Put some individual masses m1 , m2 , ..., mn at certain points x1 , x2 , ..., xn along

a number line, where the number line is supported at some point x called the
fulcrum:
Physics tells us that each mass exerts torque on the system, causing the system
to want to rotate about the fulcrum. Masses on opposite sides of the fulcrum exert
torque in opposite directions.
Physics ⇒ torque created by j th mass = (force) · (distance from fulcrum)

= (weight) · (distance from fulcrum)
= (mass)(acceleration)(distance from fulcrum)
= mj (9.8) (xj − x)
Adding the torque for each of the masses in the system gives total torque T , where
n
X
T = mj (9.8)(xi − x)
j=1
In our picture above,
T =
101
so if the fulcrum is at x = 8,
T =
If the fulcrum is at x = 6,
T =
If the fulcrum is at x = 7,
T =
When the total torque is zero, we say the system is at equilibrium and the cor-
responding location of the fulcrum is called the center of mass of the system and
is denoted x. (In our example x = 7.)
Now let’s do the same example without using numbers. Suppose individual
masses m1 , m2 , ..., mn at certain points x1 , x2 , ..., xn along a number line. What is x?
We have proven:
102
Theorem 5.3 (Center of mass for discrete masses along a line) Suppose discrete
masses m1 , m2 , ..., mn are located at respective points x1 , x2 , ..., xn along a number
line. Then:
1. The moment about the origin of this system is

n
X
M0 = mj xj .
j=1
2. The total mass of this system is

n
X
M= mj .
j=1
3. The center of mass of the system is
M0
x= .
M
Continuous mass along a line

Consider a wire (or bar or stick) situated along the x−axis with varying density
δ(x) (sometimes density is denoted by ρ(x)).
Question: What is x?
Suppose first that the density δ is constant. Then the total mass is
M = (density) · (length) = δ · ∆x.
So considering the general principle of applications of integration, if the density is

nonconstant then the total mass is
M=
103
n
Notice the similarity to the discrete case, where M was mj . Essentially, the
P
j=1
summation is replaced by an integral and the masses mj are replaced by the den-
sity function δ(x). In the same way, the moment about the origin should be
M0 =
Summarizing, we have
Theorem 5.4 (Center of mass for continuously distributed mass on a line) If a

mass is distributed along a line, running from x = a to x = b, such that the density of
the mass at position x is δ(x), then:
1. The moment about the origin of this mass is

Z b
M0 = x δ(x) dx.
a
2. The total mass of this mass is

Z b
M= δ(x) dx.
a
3. The center of mass of this system is

b R
M0 a x δ(x) dx
x= = Rb .
M a δ(x) dx
Example: Find the center of mass of a thin rod of length 2 units, whose density
x units from the left edge of the rod is δ(x) = x2 + x.
104
5.6. Two-dimensional moments and centers of mass
5.6 Two-dimensional moments and centers of mass

Discrete masses in a plane
Now, we start considering two-dimensional objects (like slabs and tabletops). We
start with point masses in a plane; this situation is similar to the situation in Sec-
tion I above except that you need to consider x− and y−coordinates separately.
The center of mass will be a point (x, y).
Consider a single mass m located at point (x, y).
(1) Consider the x−axis as an axis of rotation:
⇒ Moment about x−axis of this particle is Mx = m y.
(2) Now, consider the y−axis as an axis of rotation:
⇒ Moment about y−axis of this particle is My = m x.
Now, suppose you have a bunch of masses in a plane:
105
We conclude:
Theorem 5.5 (Center of mass for discrete masses in a plane) Suppose that dis-
crete masses m1 , m2 , ..., mn are located at respective points (x1 , y1 ), (x2 , y2 ), ..., (xn , yn )
in a plane. Then:
1. The moment about the x−axis of this system is

n
X
Mx = mj yj .
j=1
2. The moment about the y−axis of this system is

n
X
My = m j xj .
j=1

n
X
M= mj .
j=1
4. The center of mass of the system is (x, y) where

My Mx
x= and y= .
M M
Continuous mass in a plane

My Mx
The formulas x = M
, y= M
still hold. But what are Mx , My and M ?
Setup:
d
a x b
106
Question: What is the center of mass (x, y) for a (flat) slab in the plane, possibly
with nonconstant density?
Note: We assume in Math 230 that the density is either constant, or depends
only on x (i.e. the density is δ(x), not δ(x, y)). Otherwise one would need machin-
ery learned in Math 320.
To compute Mx , My and M , we take the formulas from the previous discussion

and adapt them in the same way that we adapted the formulas for discrete masses
along a line to get to the formulas for continuous mass along a line. This gives:
Theorem 5.6 (Center of mass for continuously distributed mass in a plane) If

a planar region runs from x = a to x = b and has height l(x) at point x, and if the
density of this region at point (x, y) is δ(x), then:

Z b
My = x l(x) δ(x) dx.
a

Z b
M= l(x) δ(x) dx.
a
My Mx
x= and y= .
M M
Notice there is no formula for Mx in the above theorem. This is because, when
you generalize the formula for Mx obtained earlier in this section, you get
Z d
Mx = y l(y) δ(y) dy
c
and this is a problem because we are assuming the density depends on x, not y.
However, this problem is fixable:
107
Theorem 5.7 (Center of mass for continuously distributed mass in a plane) If

R is the region lying above function g and below function f for a ≤ x ≤ b:
and the density of R at point (x, y) is given by function δ(x), then:

Z b
My = x [f (x) − g(x)] δ(x) dx.
a
2. The moment about the x−axis of this system is

Z b
1
Mx = [f (x)]2 − [g(x)]2 δ(x) dx.
a 2

Z b
M= [f (x) − g(x)] δ(x) dx.
a
My Mx
x= and y= .
M M
We already have seen statements (1), (3) and (4) in the previous theorem. To
see why (2) is true, think of the formula Mx as a volume computation (which looks
like something obtained by the washer method without the π but with a 12 ). If you
compute the same volume with shells, you get the integral
Z d
y l(y) δ(y) dy
c
which is the formula we already knew for Mx .
108
Example: Find the center of mass of the region enclosed by y = 4x − x2 and the
x−axis, where the density function is δ(x) = x + 1.
4
-1 1 2 3 4 5
-1
109
Example: Find the center of mass of the region bounded by y = 9 − x2 and the
x−axis, assuming that the density is constant.
10
-3 -2 -1 1 2 3
Notice that this answer did not depend on the actual value of δ. So we may as
well have assumed that the density of the region is 1. This situation arises fairly
often, so we invent the following terminology:
Definition 5.8 The centroid of a region is the center of mass of that region, assuming
it has constant density. (To compute a centroid, just assume δ = 1 and use the formulas
already discussed.)
110
Example: Write integrals which can be used to find the centroid of the region
enclosed by y = x and y = x2 − 6.
4
-4 -2 2 4
-2
-4
-6
-8
111
5.7. Probability
5.7 Probability
Finite probability
A motivating example: Suppose you spin a spinner and it stops at a random loca-
tion. You either win or lose money depending on where the spinner lands:
' $
@
@
@ +10
@ +4
−8 @
@A
@Av
A
AU
−2
& %
We represent such a situation mathematically by something called a random

variable:
Definition 5.9 A random variable (r.v) is a quantity X which depends on the re-
sult of some random experiment.
In the spinner example, X is the amount you win/lose (i.e. is the number on
which the spinner lands). To describe X, we need to know two things about it:
1.
2.
This information can be conveyed by means of a table:
x f (x) = P (X = x)
10
-8
-2
112
5.7. Probability
Note: The range of X are the numbers in the left-hand column, and the proba-
bility of each value of X is in the right-hand column.
Notice also that f is a function, so we can visualize f by means of a graph:

f (x)
6
1/2 v
1/4 v
1/8 v v
- x
−8 −2 4 10
Example: In the context of this spinner, let X be the number spun. What is the
probability that X ≥ 0 (in symbols, we are asking for P (X ≥ 0))?
Solution:
The above computation illustrates the following general concept:
Theorem 5.10 Let X be a random variable taking only the values x1 , x2 , ..., xn , with
respective probabilities f (x1 ), f (x2 ), ..., f (xn ). Then for any set E, the probability that
X is in E is X
P (E) = P (X ∈ E) = f (xj ).
xj ∈E
In other words, for random variables that only take finitely many different val-
ues, probabilities are computed via addition.
113
5.7. Probability
Expected value
A new question: Suppose you spin the spinner 800 times. How much would you
expect to be ahead or behind after the 800 spins?
This example generalizes in the following formula:
Definition 5.11 Let X be a random variable taking only the values x1 , x2 , ..., xn , with
respective probabilities f (x1 ), f (x2 ), ..., f (xn ). The expected value of X, denoted
EX, is
n
X
EX = xj f (xj ).
j=1
The expected value of a random variable is the amount you would expect the r.v.
to average if you repeat the experiment over and over.
114
5.7. Probability
Continuous random variables

The random variable X coming from the spinner in the preceding section is called
finite-valued because there are only finitely many possible values of X (namely
−8, −2, 4 and 10).
In the real world, random variables are not necessarily finite-valued. For exam-
ple, suppose you want to consider the amount of time until some particular event
occurs. Here are some particular examples:
• Let X = amount of time until your cell phone rings next
• Let X = amount of time until a machine part fails
• Let X = amount of time until a customer is involved in a traffic accident
(Actuaries working for insurance companies are particularly interested in things

like the third example.)
In all these examples, X is not finite-valued because X could be 1, 5, 100, 3.7,

√
19
3
, π, etc. In this setting X takes values in an interval (the interval of values for X
in these examples is [0, ∞) but it could be something different in general).
Example: Suppose X is the amount of time until your cell phone rings. What
is the probability that X = 34 minutes?
Example: Suppose X is the amount of time until your cell phone rings. What
is the probability that X ≥ 34 minutes?
Definition 5.12 A random variable X is called continuous if it takes values in an

interval, and if the probability that X takes any one particular value is zero.
Recall that we described finite-valued random variables by using a chart, or a

function f , or a graph. How can we describe a continuous random variable?
115
5.7. Probability
Example: Choose a real number from the interval from the interval [0, 4] “uni-
formly” (this means that all real numbers should be “relatively equally likely" to
be chosen). What is the probability that X ∈ [2, 3]?
Solution: Think geometrically:
There is a more useful way to reinterpret the same solution:
This allows us to consider situations where all numbers are not relatively equally
likely. Suppose f : R → R is some function whose graph is like this:
0 2 3 7 8 10 12
The function f above is called a density function for the random variable X.
More generally:
116
5.7. Probability
Definition 5.13 Let X be a continuous random variable. Then a function f : R → R

is called a density function for X if
Z b
P (X ∈ [a, b]) = f (x) dx for all a ≤ b.
a
The definition implies three things that must be true about the density function
of any random variable:
1.
2.
3.
The key principle is that a continuous random variable X is completely de-

scribed by specifying its density function f , and probabilities associated to X are
computed with integrals of the density function.
In particular, for any number b,

Z b
P (X = b) = P (X ∈ [b, b]) = f (x) dx = 0
b
and
Z b
P (X ∈ (a, b]) = P (X ∈ [a, b]) − P (X = a) = P (X ∈ [a, b]) − 0 = f (x) dx
a
so for a continuous r.v. X,
P (a < X < b), P (a ≤ X < b), P (a ≤ X ≤ b) and P (a < X ≤ b)
are all the same and all equal to

Z b
f (x) dx.
a
117
5.7. Probability
Expected value of a continuous random variable

Question: Given continuous r.v. X with density function f , what is EX?
Answer: Recall that if X was finite-valued,

n
X
EX = xj f (xj ).
j=1
This formula generalizes as an integral:
Definition 5.14 Let X be a continuous random variable with density function f .

Then the expected value of X, denoted EX, is
Z ∞
EX = x f (x) dx.
−∞
Example: The amount of time (measured in months) until a driver causes an

accident is a continuous random variable whose density function is
(
25 −3
12
x if x ∈ [1, 5]
f (x) =
0 otherwise
Find the probability that the driver will cause an accident within two months, and
find the expected amount of time until the driver causes an accident.
118
5.7. Probability
Example: Let X be a continuous random variable with density function

(
ce−2x if x ≥ 0
0 if x < 0
where c is a constant.
(a) Find c.
(b) Find P (3 < X < 5).
(c) Find P (X ∈ [−4, 1)).
(d) Find P (X = 7).
(e) Find the expected value of X.
119

1. Find the area of the region enclosed by the graph of f (x) = x2 − 6x and the
x−axis.
2. Find the area of the region enclosed by the graphs of f (x) = x2 + 2x + 1 and
g(x) = 2x + 5.
3. Find the area of the region enclosed by the graphs of f (x) = x2 − 4x + 3 and
g(x) = −x2 + 2x + 3.
4. Find the area of the region enclosed by the graphs of f (x) = (x − 1)3 and
g(x) = x − 1.
√
5. Find the area of the region enclosed by the graphs of f (x) = 3x + 1 and
g(x) = x + 1.
6. Find the area of the region lying between the graphs of f (x) = 2 sin x and
g(x) = tan x from x = −π
3
to x = π3 .
7. Find the area of the region lying between the graphs of f (x) = cos x and
g(x) = 2 − cos x from x = 0 to x = 2π.
8. Let R be the region of the xy−plane lying above the x−axis, below the line
1 2
y = 32 x − 4, and above the curve y = 15 x − 32 x + 1. Find the area of R.
9. a) Write down the equation of a circle of radius r, centered at the origin.

b) Write down an expression (in terms of one or more integrals) which will
give the area of this circle.
c) Use Mathematica to evaluate the expression you got in part (b).
d) Evaluate the integral from part (b) by hand. Hint: You need to rewrite
the integrand and then use the substitution u = arcsin xr .
10. An ellipse (centered at the origin) is the graph of an equation of the form
2
x2
a2
+ yb2 = 1. Such an ellipse looks like an oval passing through the points
(a, 0), (−a, 0), (0, b) and (0, −b).
a) Write down an expression (in terms of one or more integrals) which will
give the area of this ellipse.
b) Use Mathematica to evaluate the expression you got in part (a).
11. Consider the quadrilateral whose vertices are (in order) (0, 2), (4, 7), (6, 6)
and (2, 0) . Write down an expression involving integrals which gives the
area of this quadrilateral, and evaluate the expression to find the area of the
quadrilateral (you can use Mathematica or do the integrals by hand).
120
√
12. Let R be the region enclosed by the graphs of y = x and y = x.
a) Sketch a picture of R, clearly indicating which graph is which.

b) Write a formula involving one or more integrals with respect to the vari-
able x which gives the area of R.
c) Write a formula involving one or more integrals with respect to the vari-
able y which gives the area of R.
d) Use Mathematica to evaluate the integrals from parts (b) and (c) (or do
them by hand), and verify that you get the same answer.
13. Let R be the region enclosed by the graphs of x = 0, 3x + y = 6 and y = x2 − 4.
a) Sketch a picture of R, clearly indicating which graph is which.

able x which gives the area of R.
c) Write a formula involving one or more integrals with respect to the vari-
able y which gives the area of R.
d) Use Mathematica to evaluate the integrals from parts (b) and (c) (or do
them by hand), and verify that you get the same answer.
14. Consider a solid whose base is the triangle in the xy−plane whose vertices
are (0, 0), (0, 2) and (6, 0), such that cross-sections of the solid are squares
parallel to the y−axis. Find the volume of the solid.
15. Consider a solid whose base is the region in the xy−plane lying above y = x
and below y = 2 − x2 , such that cross-sections of the solid parallel to the
y−axis are semicircles whose diameter lies in the xy−plane. Find the volume
of the solid.
16. Consider a solid whose base is the interior of the circle x2 + y 2 = 16 in the
xy−plane. If cross-sections of the solid parallel to the y−axis are rectangles
which are twice as high as they are wide, find the volume of the solid.
17. Find the volume of the solid obtained by revolving the region below y = sec x
and above the x−axis from x = 0 to x = π4 around the x−axis.
18. Find the volume of the solid obtained by revolving the region below y =
2 − x2 and above y = 0 around the x−axis.
In Problems
√ 19-22, let R be the region in the xy−plane bounded by the graphs
y = x, y = 0 and x = 4.
19. Let S1 be the solid obtained by revolving R around the x−axis.
121
a) Write a formula involving one or more integrals with respect to the vari-
able x which gives the volume of S1 .
able y which gives the volume of S1 .
c) Evaluate the integrals from parts (a) and (b) using Mathematica, and ver-
ify that you get the same answer.
20. Let S2 be the solid obtained by revolving R around the y−axis.
c) Evaluate the integrals from parts (a) and (b) using Mathematica, and ver-
ify that you get the same answer.
21. Let S3 be the solid obtained by revolving R around the line x = 6.
22. Let S4 be the solid obtained by revolving R around the line y = −3.
In Problems 23-35, let Q be the region in the xy−plane lying to the right of the
y−axis, above the curve y = 2x2 and below the line y = 8.
23. Let S1 be the solid obtained by revolving Q around the x−axis.
24. Let S2 be the solid obtained by revolving R around the line x = −4.
122
25. Let S4 be the solid obtained by revolving R around the line y = 8.
26. Use calculus to show that the volume of a sphere of radius r is 34 πr3 . Hint:
You can visualize such a sphere by first thinking of a circle of radius r in
the plane, and building a solid over the base where the cross-sections of the
solid parallel to the y−axis are themselves circles. Write down an integral
which gives the volume, and evaluate the integral by hand (it’s not hard) to
get 34 πr3 .
27. Suppose a bird flies in a straight line so that its velocity at time t is 50t2 − 20t
mi/hr. Find the distance the bird travels between times t = 0 and t = 1.
28. If the density of a metal rod is constant, then the mass of the metal rod is
equal to the density times the length of the rod. Suppose a metal rod has a
nonconstant density, where the density x units from the left edge of the rod
is 2x2 + x + 1 grams/unit. Find the mass of the rod, if the wire is 3 units long.
29. If the force applied to an object is constant, then the work done by that force
is equal to the force times the distance the object moves. Suppose a force is
applied to an object where the force at position x is 2x + 3 Newtons. Find the
work done in moving the object from position 5 meters to position 10 meters.
30. If the current i in an electric circuit is constant, then the charge q that builds
up in the circuit is equal to the current times the amount of time that the
current is in the circuit. Find the charge in an electric circuit if a current of
sin t + 2 cos t + 3 is applied to the circuit from time 0 to time π.
31. Find (by hand) the length of the curve y = 23 x3/2 + 1 from x = 0 to x = 1.
32. Find (by hand) the length of the curve y = 2x3/2 − 2 from x = 0 to x = 8.
33. Write an integral which gives the length of the curve y = ln(sin x) between
x = π4 and x = π2 , and use Mathematica to evaluate the integral.
√3
34. Write an integral which gives the length of the curve y = x2 between the
points (1, 1) and (8, 4), and use Mathematica to evaluate the integral.
123
35. Write an integral which gives the length of the curve y = x1 from x = 1 to
x = 3, and use Mathematica to find a decimal approximation to this integral.
36. Write an integral which gives the length of the curve y = sin x from x = 0 to
x = π, and use Mathematica to find a decimal approximation to this integral.
37. Find the total length (you can use Mathematica to do the integral) of the curve
x2/3 + y 2/3 = 4 (the graph of this curve is found below (it extends 8 units from
the origin to the left, right, up and down):
-8 8
-8
38. Suppose a system consists of four objects, with respective masses 7, 4, 3 and
8, located along an axis with respective positions −3, −2, 5 and 6. Find the
total mass of the system, the moment about the origin of the system, and the
center of mass of the system.
39. Suppose a system consists of three objects located along an axis. The masses
of the three objects are 6, 5 and 3, and the respective positions of the objects
are −5, 1 and 3.
40. Suppose two siblings sit on the ends of a see-saw; a brother, who weighs 75
pounds, and a sister, who weighs 50 pounds. If the seesaw is 12 feet long,
how far from the brother should the fulcrum be placed so that the see-saw is
balanced?
41. A metal rod of length 10cm has density δ(x) = 1 + x2 g/cm at a distance x cm
from the left end of the rod. Find the mass of the rod, and find how far from
the left end of the rod the center of mass is.
42. A rod with density δ(x) = 2 + sin x kg/unit is positioned along the positive
x−axis, with its left end at x = 0 and its right end at x = 3π 4
. Find the
x−coordinate of the center of mass of the rod.
43. Suppose a system consists of three objects in a plane. The first mass is 5g,
located at (2, 2); the second mass is 1g, located at (−3, 1), and the last mass is
3g, located at (1, −4). Find the total mass of the system, the moments about
the x− and y− axis of the system, and the center of mass of the system.
44. Find the center of mass of a system which consists of the following five
masses:
124
mass 3 4 2 1 6
position (−2, −3) (5, 5) (7, 1) (0, 0) (−3, 0)
√
45. Let R be the region in the xy−plane above the x−axis, below y = x and to
the left of x = 9.
a) Assuming that R has constant density, find the center of mass of R.

b) Assuming that the density of any point (x, y) in R is δ(x) = x + 2, find
the center of mass of R.
46. Let R be the region in the first quadrant between the graphs of y = x2 and
y = x3 .
a) Assuming that R has constant density, find the center of mass of R.

b) Assuming that the density of any point (x, y) in R is δ(x) = 2 − x2 , find
the center of mass of R.
47. Find the centroid of the region bounded by the graphs of y = xe−x , y = 0 and
x = 5. (Use Mathematica to evaluate the integrals.)
48. Find the centroid of a semicircle of radius r (a picture is shown below, but
you have to figure out what graph is the top of the semicircle, and how far to
the left or right it goes).
49. Find the centroid of a circle of radius r centered at the origin. Hint: there is a
more clever approach than doing a bunch of integrals.
50. Consider the region in the xy−plane bounded by the graphs of y = x2 and
y = b, where b > 0 is a positive constant. What is the x−coordinate of the
center of mass of this region? Is the y−coordinate of the center of mass of this
region greater than, or less than 2b ? Explain your reasoning.
51. Let X be a random variable taking values in the interval [0, 1]. If the density
function of X is f (x) = cx2 + x, find the value of c.
52. Let X be a random variable taking values in the interval [0, 2]. If the density
function of X is f (x) = ce−x/2 , find the value of c.
125
53. Let X be a random variable taking values in [0, 4] whose density function is
f (x) = 18 x.
a) Find the probability that X > 2.

b) Find the probability that X = 3.
c) Find the probability that X ≤ 1.
d) Which is more likely, that X = 1 or X = 3?
e) Which is more likely, that X is very close to 1 or that X is very close to
3?
f) Find a number b such that P (X ≥ b) = 32 .
54. Let X be a random variable taking values in [0, ∞) whose density function is
f (x) = ce−3x .
a) Find c.
b) Find the probability that X > 10.
c) Find the probability that 3 ≤ X ≤ 4.
55. Suppose that the lifespan of a certain organism is a random variable X whose
density function is f (x) = 4xe−2x (where x ∈ (0, ∞) is measured in years).
a) Find the probability that the organism lives for less than one year.
b) Find the expected lifespan of the organism.
56. Suppose that the lifetime (in months) of a light bulb is a random variable X
1
with density function f (x) = 81 e− 8 x (again, X takes values in (0, ∞).
a) Find the probability that the light bulb lasts more than four months.
b) Find the expected lifetime of the lightbulb.
57. Find the expected value of a random variable whose density function is f (x) =
1
2
sin x for 0 ≤ x ≤ π.
58. Let X be a continuous random variable taking values in some interval. The
median of X is a number m such that P (X ≤ m) = 21 . In each of these
examples, find the median of X, if X has the given density function:
a) f (x) = 4x3 for 0 ≤ x ≤ 1

b) f (x) = 2e−2x for 0 ≤ x
126
Answers
3
1. 36 5. 2
32
2. 3
6. 2 − ln 4
3. 9 7. 4π
√
1 103−4 10
4. 2
8. 9
9. a) x2 + y 2 = r2
R √
b) A = 4 0r r2 − x2 dx
c) πr2
d) πr2
Ra q x2
10. a) A = 4 0 b 1− a2
dx
b) πab
R2h 5 i R4h 5 i
11. A = ( x + 2) − (2 − x)
0 4
dx + 2 ( x + 2) − ( 3 x − 3)
4 2
dx
R h i
+ 6 ( −1 x + 9) − ( 3 x − 3)
4 2 2
dx = 18
√
12. a) The curved graph on the top is y = x; the diagonal line is y = x:
1.0
0.8
0.6
0.4
0.2
0.0 0.2 0.4 0.6 0.8 1.0
R1 √ R1 1
b) 0 ( x − x) dx c) 0 (y − y 2 ) dy d) 6
13. a) The line on the left is x = 0, the line in the upper right is 3x + y = 6 and
the curve is y = x2 − 4.
6
0.5 1.0 1.5 2.0
-2
-4
R2 2 34
b) 0 [(6 − 3x) − (x − 4)] dx d) 3
R0 √
y + 4 dy + 06 (2 − y3 ) dy
R
c) −4
127
R6
14. V = 0 [2 − 13 x]2 dx = 8
R 1 1 2−x2 −x 2 81
15. V = −2 2π 2
dx = 80
π
R4 √ 2
2048
16. V = −4 8 16 − x2 dx = 3
R π/4
17. V = 0 π sec2 x dx = π
R √2 √
18. V = √ π(2 − x2 )2 dx = 64 2 π
− 2 15
R4
19. a) V = 0 πx dx
R2
b) V = 0 2πy(4 − y 2 ) dy
c) 8π
R4 √
20. a) V = 0 2πx x dx
R2
b) V = 0 [π42 − π(y 2 )2 ] dy
128
c) 5
π
R4 √
21. a) V = 0 2π(6 − x) x dx
R2
b) V = [π(6 − y 2 )2 − π(6 − 4)2 ] dy
0
√
a) V = 04 [π( x + 3)2 − π(0 + 3)2 ] dx
R
22.
R2
b) V = 0 2π(y + 3)[4 − y 2 ] dy
R2
23. a) V = 0 [π(8)2 − π(2x2 )2 ] dx
R8 q
y
b) V = 0 2πy 2
dy
R2
24. a) V = 0 2π(x + 4)[8 − 2x2 ] dx
R8h q
y
i
b) V = 0 π( 2
+ 4)2 − π(0 + 4)2 dy
R2
25. a) V = 0 π(8 − 2x2 )2 dx
R8 q
y
b) V =0 2π(8 − y)[ 2 ] dy
Rr √
26. V = −r π( r2 − x2 )2 dx = 43 πr3
20
27. 3
mi
51
28. 2
g
29. 90 Nm
30. 2 + 3π
128
−2+2(2)3/2
31. 3
−2+2(73)3/2
32. 27
R π/2 √
33. s = π/4 1 + cot2 x dx = − ln(tan π8 )
√ √
r h i2
R8 1 −1/2 80 10−13 13
34. s = 1 1+ 2
x dx = 27
R3q
35. s = 1 + [−x−2 ]2 dx ≈ 2.14662
1
R √
36. s = 0π 1 + cos2 x dx ≈ 3.8202
37. 48
I have not found the answers to problems 38-50 on moments and centers of
mass.
3
51. c = 2
e
52. c = 2e−2
3
53. a) P (X > 2) = 4
b) P (X = 3) = 0
1
c) P (X ≤ 1) = 16
d) They are equally likely (the probability of both is zero).
e) It is more likely that X is very close to 3 since f (3) > f (1).
f) b = √4 .
3
54. a) c = 3
b) P (X > 10) = e−30
c) P (3 ≤ X ≤ 4) = e−9 − e−12 .
55. a) P (X < 1) = 1 − 3e−2
b) EX = 1 year
56. a) P (X > 4) = e−1/2
b) EX = 8 months
π
57. EX = 2
58. a) m = 2−1/4
b) m = 21 ln 2
129
Chapter 6
Parametric equations
6.1 Introduction to parametric equations

In Calculus I, we mostly consider functions where y is a function of x:
y = f (x)
We represent such a function pictorially by sketching a graph; for example the

function f (x) = x2 − 2 has graph
4
-3 -2 -1 1 2 3
-1
-2
-3
which reflects tables of values like

x -2 -1 0 1 2
y = f (t) 2 -1 -2 -1 2
130
6.1. Introduction to parametric equations
One thing we know is that the graph of any function y = f (x) must pass the
Vertical Line Test: any vertical line must hit the graph of y = f (x) in at most one
point.
10 10
5 5
-10 -5 5 10
-6 -4 -2 2 4
-5 -5
-10 -10
Suppose we consider some curve which does not pass the Vertical Line Test. Is
there any way to represent such a curve as a function (and subsequently compute
derivatives, integrals and their applications)?
The answer is yes; the method is to think of both the x− and y− coordinates of
functions of a third variable, which is usually called t. The resulting set of equa-
tions (
x = f (t)
y = g(t)
are called parametric equations for the curve. As an example, consider the set of
parametric equations (
x = t2 − 1
y = t − t2
To obtain a graph of this set of equations, choose some values of t and solve for the
corresponding x and y. Plot the points x and y to obtain the graph. In our example,
we have
t -2 -1 0 1 2
x = f (t)
y = g(t)
(x, y)
131
-5 -4 -3 -2 -1 1 2 3 4 5
-1
-2
-3
-4
-5
-6
-7
-8
Note: For each t, the point (f (t), g(t)) should be thought of as the location of an
object at time t, and the curve is the path the object traces out as it moves. Thus
the graph of a set of parametric equations indicates a direction of motion along the
curve (from negative t to positive t) as well as the curve itself.
Our goal in this chapter is to analyze the motion of an object whose position at
time t is given by parametric equations like these. In particular, we would like to
know:
132
Converting parametric equations to Cartesian equations

First method: Solve one of the two parametric equations for t (or an expression of
t), and plug into the other equation.
Example: Find Cartesian equations equivalent to the parametric equations

(
x = 3t − 5
y = t cos t

(
x = 7 cos2 t − 3
y = 8 cos t + 1
Second method: Use a trig identity to combine expressions of x or y.

(
x = 4 sin t
y = 5 cos t
133
6.2. Parametric equations of common graphs
6.2 Parametric equations of common graphs

Circles
Recall that the Cartesian equation of a circle centered at (h, k) with radius r is
(x − h)2 + (y − k)2 = r2 .
For example, here is the graph of (x − 3)2 + (y + 2)2 = 16:

2
-1 1 2 3 4 5 6 7
-1
-2
-3
-4
-5
-6
To deduce parametric equations of a circle, we start with a trig identity which

resembles the equation of the circle:
Theorem 6.1 (Parametric equations of a circle) The circle centered at (h, k) with
radius r, whose Cartesian equation is (x−h)2 +(y−k)2 = r2 , has parametric equations
(
x = r cos t + h
.
y = r sin t + k
These parametric equations trace out the circle counterclockwise, with t = 0

corresponding to the right-most point on the circle and t generally corresponding
to the angle of the point on the circle, measured from the positive x−direction. (See
the picture on the next page.)
134
Example: Find parametric equations for the circle (x + 2)2 + y 2 = 7, where the
circle is traced out counterclockwise.
Question: What if you want the circle to be traced out clockwise? (Later)
Ellipses
An ellipse is the graph of any Cartesian equation of the form
(x − h)2 (y − k)2
+ = 1.
a2 b2
(If a = b, then the ellipse is actually a circle, so circles are in fact special kinds of
ellipses.)
The graph of such an ellipse looks like this:
135
By an argument similar to the one that produced parametric equations for a

circle, we can find parametric equations of an ellipse:
Theorem 6.2 (Parametric equations of an ellipse) The ellipse centered at (h, k)

2 2
whose Cartesian equation is (x−h)
a2
+ (y−k)
b2
= 1 has parametric equations
(
x = a cos t + h
.
y = b sin t + k
As with circles, these parametric equations trace out the ellipse counterclock-
wise, with t = 0 corresponding to the right-most point on the circle and t generally
corresponding to the angle of the point on the ellipse, measured from the positive
x−direction.
Example: Sketch the graph of the ellipse whose parametric equations are x =
2 cos t − 1, y = 3 sin t + 5.
10
-6 -5 -4 -3 -2 -1 1 2 3 4 5
-1
136
Hyperbolas
A hyperbola is the graph of any Cartesian equation of the form
(x − h)2 (y − k)2 (y − k)2 (x − h)2

− =1 or − =1
a2 b2 b2 a2
The graph of such a hyperbola looks like one of two things:
Left-right hyperbolas Up-down hyperbolas
(x − h)2 (y − k)2 (y − k)2 (x − h)2

− =1 − =1
a2 b2 b2 a2
“x2 term positive, y 2 term negative” “y 2 term positive, x2 term negative”
To obtain parametric equations for a hyperbola, one mimics the procedure that
was done for ellipses. The difference is that to obtain the minus sign in the equa-
tion, one starts not by thinking of
cos2 t + sin2 t = 1
but by thinking of
sec2 t − tan2 t = 1.
This eventually yields the following:
Theorem 6.3 (Parametric equations of a left-right hyperbola) If a hyperbola has

2 2
Cartesian equation (x−h)
a2
− (y−k)
b2
(
x = a sec t + h
.
y = b tan t + k
137
Theorem 6.4 (Parametric equations of an up-down hyperbola) If a hyperbola has

2 2
Cartesian equation (y−k)
b2
− (x−h)
a2
(
x = a tan t + h
.
y = b sec t + k
Remark: There are other parametric equations for hyperbolas, which we do not
mention here.
Remark: I am not mentioning the direction the hyperbolas are traced out by
these parametric equations here. Ask me outside of class if you want to know how
this works.
Example: Sketch the graph of the hyperbola whose parametric equations are
x = sec t + 2, y = 4 tan t.
6
-6 -5 -4 -3 -2 -1 1 2 3 4 5 6
-1
-2
-3
-4
-5
Example: Sketch the graph of the hyperbola whose parametric equations are
x = 3 tan t − 1, y = sec t.
6
-6 -5 -4 -3 -2 -1 1 2 3 4 5 6
-1
-2
-3
-4
-5
138
Lines and segments

Suppose we have a line passing through the points (x0 , y0 ) and (x1 , y1 ). To write
parametric equations for this line, we will try to write x and y as linear functions
of t, where t = 0 corresponds to (x0 , y0 ) and t = 1 corresponds to (x1 , y1 ).
First, for the formula for x. When t = 0, x = x0 and when t = 1, x = x1 . The

linear relationship between x and t therefore has slope
and therefore has equation
Now, for the formula for y. When t = 0, y = y0 and when t = 1, y = y1 . The

linear relationship between y and t therefore has slope
and therefore has equation
To summarize:
Theorem 6.5 (Parametric equations of a line) The parametric equations of the line
passing through points (x0 , y0 ) and (x1 , y1 ) are
(
x = x0 + (x1 − x0 )t
y = y0 + (y1 − y0 )t
where t ranges from −∞ to ∞.
These parametric equations trace out the line in the direction moving from
(x0 , y0 ) to (x1 , y1 ), with t = 0 corresponding to (x0 , y0 ) and t = 1 corresponding
to (x1 , y1 ).
The parametric equations of a segment are similar. A segment is the portion of

a line lying between two points on that line. To write the parametric equations of
139
the segment between (x0 , y0 ) and (x1 , y1 ), use the same parametric equations as the
line through those points and truncate the values of t:
Theorem 6.6 (Parametric equations of a segment) The parametric equations of

the segment connecting the points (x0 , y0 ) and (x1 , y1 ) are
(
x = x0 + (x1 − x0 )t
.
y = y0 + (y1 − y0 )t
where t ranges from 0 to 1.
These parametric equations trace out the segment in the direction moving from
(x0 , y0 ) to (x1 , y1 ), with t = 0 corresponding to (x0 , y0 ) and t = 1 corresponding to
(x1 , y1 ).
Example: Find parametric equations for the line y = 2x − 3. (There will be an

easier method of doing this presented later.)
Example: Find parametric equations for the line x = −2, which trace out the
line downward.
Example: Find parametric equations for the segment connecting the points
(−3, 2) and (4, −5).
140
Example: Sketch the graph of the curve whose parametric equations are
(
x=3−t
, where t ranges from 0 to 1.
y = −2 + 3t
6
-6 -5 -4 -3 -2 -1 1 2 3 4 5 6
-1
-2
-3
-4
-5
Parametric equations of Cartesian functions

Suppose you are given a function y = h(x). It is easy to write parametric equations
for such a function.
Theorem 6.7 (Parametric equations of y = h(x)) If a function has Cartesian equa-

tion y = h(x), then it can be described by the parametric equations
(
x=t
.
y = h(t)
These parametric equations trace out the curve y = h(x) from left to right,
where the value of t is the same as the value of x.
141
Example: Find parametric equations for the curve y = x2 − 3. Sketch the

graph of this function, and indicate the points on the graph corresponding to
t = −2, −1, 0, 1, 2.
6
-6 -5 -4 -3 -2 -1 1 2 3 4 5 6
-1
-2
-3
-4
-5
Parametric equations of common functions


 x = r cos t + h
Circle center (h, k) radius r
y = r sin t + k




 x = a cos t + h
Ellipse center (h, k)
y = b sin t + k




 x = a sec t + h
Hyperbola center (h, k) opening left/right
y = b tan t + k




 x = a tan t + h
Hyperbola center (h, k) opening up/down
y = b sec t + k




 x = x0 + (x1 − x0 )t
Line through (x0 , y0 ) and (x1 , y1 )
y = y0 + (y1 − y0 )t




 x=t
Cartesian function y = h(x)
y = h(t)


142
6.3. Calculus with parametric equations
6.3 Calculus with parametric equations

Two-dimensional motion; vectors
The most important application of parametric equations is to describe the motion
of an object moving in two-dimensions (in Calculus III you learn how to study
three-dimensional motion) as time passes.
Suppose an object’s position in an xy−plane at time t is (x, y) where x and y are

given by the parametric equations
(
x = f (t)
.
y = g(t)
Question: How do you describe the velocity, speed, acceleration of the object?
To describe the physical motion of such an object, we need the notion of a two-
dimensional vector. The idea of a vector is made more precise in linear algebra,
but for now you should think of a two-dimensional vector as a mathematical ob-
ject sitting in an xy−plane which has two (and only two) attributes: a length and
a direction. (In linear algebra you will learn that vectors are actually much more
abstract than the ones we think about here.)
One good pictorial representation of a vector is an arrow (warning: the notion

of thinking of a vector as an arrow should be discarded if/when you take linear
algebra); we think of the vector as going from the tail of the arrow to the head.
143
Note: A vector is really more like a “floating” arrow in that what is relevant
is how far to the left/right and up/down the vector points, not where the vector
starts and ends. For example, the two vectors
are the same vector since they point in the same direction and have the same
length. In practice, the place where you locate a vector depends on the context
of the problem under consideration.
To represent a two-dimensional vector with numbers, we make two measure-

ments. First, we measure the directed horizontal distance from the tail of the arrow
to the head, and second, we measure the directed vertical distance from the tail of
the arrow to the head. These two numbers are recorded as an ordered pair, and
labelled with a letter. For example:
When writing by hand, you usually put an arrow over a letter to emphasize
that the letter represents a vector rather than a number; when typed, a letter is put
in boldface to indicate that it is a vector.
The length of a vector comes from the Pythagorean Theorem / Distance √ 2 For-
mula: given vector v = → −
v = (x, y), the length of v, denoted ||v||, is x + y 2 . In
q √ √
particular, the length of (3, −4) is ||(3, −4)|| = 32 + (−4)2 = 9 + 16 = 25 = 5.
One example of a vector is the position of an object in the setting described

above. We think of the object’s position as the vector p(t) = (f (t), g(t)), and think
of this vector pictorially as starting at the origin and ending at (f (t), g(t)). In this
setting the head of the vector is located at the actual position of the object.
Motivated by what we learned in Calculus I (namely that the first derivative of

position is velocity, and the second derivative of position is acceleration), we make
the following definitions:
144
Definition 6.8 Suppose an object is moving in an xy−plane such that its position at
time t is the point/vector p = (x, y) where
(
x = f (t)
.
y = g(t)
Then the velocity of the object at time t is the vector
v(t) = (f 0 (t), g 0 (t)).
The speed of the object at time t is the length of the velocity vector:
q
s(t) = [f 0 (t)]2 + [g 0 (t)]2 .
The acceleration of the object at time t is the vector
a(t) = (f 00 (t), g 00 (t)).
Note: The velocity of an object is a vector (it indicates a speed and a direction)
and should be drawn such that the tail of the vector is at the point (f (t), g(t)). The
velocity vector will always point in a direction tangent to the position curve, and
point in the same direction the curve is being traced out.
However, the speed of an object is a scalar (a.k.a. number). The speed is the
length of the velocity vector; the longer the velocity vector is, the faster the object
is moving.
The acceleration of an object is also a vector which should be drawn such that
the tail of the vector is at the point (f (t), g(t)). The acceleration vector will always
point to the inside of the position curve, looking forward in time from (f (t), g(t)).
(Another way of saying this is that the curve will “bend” towards the acceleration
vector.)
145
Example: Suppose a bug is crawling around an xy−plane such that its position
at time t is (x, y) where x = 2t3 + t and y = 1 − t + t2 − t3 . Find the velocity and
acceleration of the particle at time t = 1.
Example: Suppose at time t = 6 the position vector of a particle is (5, 2), the
velocity vector of the particle is (−2, 3) and the acceleration vector of the particle
is (−4, 1). Find the speed of the particle at time t = 6 and sketch a graph of the
particles motion at times close to t = 6, indicating the velocity and acceleration
vectors appropriately.
6
-6 -5 -4 -3 -2 -1 1 2 3 4 5 6
-1
-2
-3
-4
-5
146
Arc length of parametric curves

Suppose an object’s position at time t is given by some parametric equations
(
x = f (t)
.
y = g(t)
Between two times t = a and t = b, the object will follow some path. A natural
question to ask is what the length of this path. Equivalently, we want to know how
far the object travels between times t = a and t = b.
Suppose for now that the speed of the object is constant. Then the distance the
object travels over a length ∆t of time is
Note: It doesn’t make sense to multiply velocity by time because
By the general principle of application of integrals, that means that if the speed
is nonconstant, the distance travelled by the object between times t = a and t = b
is
Theorem 6.9 (Arc length formula) If x = f (t); y = g(t) are the parametric equa-
tions of some curve, then the length of that curve between times t = a and t = b
is v
u dx 2
Z bu ! !2
dy
s= t
+ dt.
a dt dt
Example: Use the arc length formula to verify that the circumference of a circle
of radius r is 2πr.
147
Tangent lines to parametric curves

Suppose you are given some parametric equations
(
x = f (t)
y = g(t)
and you want to know the slope of the graph of these parametric equations at some
value of t. We know from Calculus I that the slope is given by “the” derivative.
But now there are lots of things which are derivatives:
As an example, for the parametric equations

(
x = t − sin t
y = 2t cos t
we have
dy
But what is the derivative dx
?
To figure this, we use the Chain Rule which says
dy dy dx
= ·
dt dx dt
dy
Solving for dx
, we obtain
148
Theorem 6.10 (Slope of line tangent to curve given by parametric equations)

The slope of the line tangent to the curve whose parametric equations are
(
x = f (t)
y = g(t)
is
dy
dy dt g 0 (t)
= dx = .
dx dt
f 0 (t)
Example: Find the equation of the line tangent to the ellipse

(
x = 3 cos t
y = 2 sin t − 1
when t = π4 .
Higher-order derivatives
In the previous section we discussed how to find the slope of a curve whose para-
metric equations are given. Now, we discuss how to find the concavity of such a
curve. The concavity is given by
d2 y
,
dx2
the second derivative of y with respect to x.
d2 y
Example: Find dx2
for the curve whose parametric equations are
(
x = t2 − 1
y = t2 + t
149
Incorrect solution: Start by computing the first derivative:

dy
dy dt
= dx =
dx dt
Now differentiate this to obtain the second derivative:

d2 y
=
dx2
The solution above is incorrect. The problem is asking us to compute
What we actually computed was
To fix this, we again use the Chain Rule. We know (by the Chain Rule) that for
any quantity A,
d dA dA dx
(A) = = ·
dt dt dx dt
so we obtain
dA d
dA dt
(A)
= dx = dt dx .
dx dt dt
dy
Substituting dx
in for A in the above equation, we obtain

! d dy
d2 y d dy dt dx
= = dx .
dx2 dx dx dt
We have established:
150
Theorem 6.11 (Concavity of curve given by parametric equations) The concav-

ity of the line tangent to the curve whose parametric equations are
(
x = f (t)
y = g(t)
is
2 d dy
dy dt dx
= dx .
d2 x dt
d2 y
Earlier example, revisited: Find dx2
for the curve whose parametric equations
are (
x = t2 − 1
y = t2 + t
151
Summary of calculus formulas for parametric equations
Position (usually given) p(t) = (x, y) = (f (t), g(t))
Velocity v(t) = (f 0 (t), g 0 (t))
q
Speed (length of the velocity vector) [f 0 (t)]2 + [g 0 (t)]2
Acceleration a(t) = (f 00 (t), g 00 (t))
Rbq
Arc length from t = a to t = b s= a [f 0 (t)]2 + [g 0 (t)]2 dt
dy
dy g 0 (t)
Slope of tangent line dx
= dt
dx = f 0 (t)
dt
d dy
d2 y dt ( dx )
Concavity dx2
= dx
dt
Position, velocity and acceleration are vector quantities; the others are numbers
(a.k.a. scalars).
152
6.4. Transformations on parametric equations
6.4 Transformations on parametric equations

Non-uniqueness of parametric equations / time-changes
A curve never has only one set of parametric equations which represent it. In
particular, suppose some curve is the graph of
(
x = f (t)
(∗)
y = g(t)
Then, the same curve is also the graph of

(
x = f (kt)
(#)
y = g(kt)
for any constant k 6= 0. In particular:
• If k > 0, the equations (#) trace the curve out in the same direction as (∗).
• If k < 0, the equations (#) trace the curve out in the opposite direction as (∗).
• If |k| < 1, the equations (#) trace the curve out more slowly than (∗); the
closer |k| is to zero, the more slowly the curve is traced out.
• If |k| > 1, the equations (#) trace the curve out more quickly than (∗); the
larger |k| is, the faster the curve is traced out.
• If |k| = 1, the equations (#) trace the curve out at the same rate as (∗) (in the
same direction if k = 1; in the opposite direction if k = −1).
153
6.4. Transformations on parametric equations
Shifts and stretches

The way that the graph of some set of parametric equation is shifted or stretched
by simple algebra manipulations is summarized in this theorem:
Theorem 6.12 (Transformations on graphs of parametric equations) If C is some

curve which is the graph of (
x = f (t)
,
y = g(t)
then:
(
x = a f (t)
1. The graph of is the curve C, stretched by a factor of a units hori-
y = g(t)
zonally. If a < 0, the curve is reflected through the y−axis.
(
x = f (t)
2. The graph of is the curve C, stretched by a factor of b units verti-
y = b g(t)
cally. If b < 0, the curve is reflected through the x−axis.
(
x = f (t) + h
3. The graph of is the curve C, shifted by h units horizontally (to
y = g(t)
the right if h > 0 and to the left if h < 0).
(
x = f (t)
4. The graph of is the curve C, shifted by k units vertically (up if
y = g(t) + k
k > 0 and down if k < 0).
154

In Problems 1-15, write parametric equations for the indicated curve or Carte-
sian equation.
1. y = 2x − 3 6. y = x2 − x + 1
2. (x + 3)2 + (y − 1)2 = 25 7. x2 + (y − 2)2 = 8

x2 y2 8. y = sin x
3. 16
− 121
=1
(x−1)2 (y+4)2
4. ey = 2x2 + 3 9. 9
+ 6
=1
(x+2)2 y2
5. 5
+ (y − 6)2 = 1 10. 4
− (x + 1)2 = 1
11. The circle of radius 8 centered at (−3, 2).
12. The line passing through (3, −1) and (−2, −3).
13. The ellipse passing through (−2, −6), (−2, 10), (0, 2) and (−4, 2).
14. The ellipse centered at (4, −1) which passes through (12, −1) and (4, −4).
15. The line passing through (−3, 5) and (2, −1).
16. Write parametric equations which trace out the circle (x − 3)2 + (y + 4)2 = 25
in a counterclockwise direction.
(x−1)2 2
17. Write parametric equations which trace out the ellipse 6
+ (y+3)
36
= 1 in a
counterclockwise direction.
(y−2)2
18. Write parametric equations which trace out the ellipse x2 + 16
= 1 in a
counterclockwise direction.
In Problems 19-29, identify the parametric equations as being equations of a line,

circle, ellipse, hyperbola, or “other”. Then sketch a graph of the parametric equa-
tions (by hand) and give the Cartesian equation corresponding to the given para-
metric equations.
19. x = 3 cos t − 1, y = 4 sin t 24. x = 3 sin t, y = 3 cos t + 2

20. x = t, y = (t − 2)2 25. x = 3 tan t + 4, y = 3 sec t − 2
21. x = 2 cos t + 3, y = 2 sin t − 4 26. x = 5 cos t − 3, y = 2 sin t
22. x = t − 1, y = t + 3 27. x = 2t − 3, y = 5t − 1
23. x = 4 sec t, y = 2 tan t + 1 28. x = t, y = cos t
155
29. x = 17 t + 72 , y = 74 t − 3
7
In problems 30 to 35, assume that an object’s position (x, y) is described by the set
of parametric equations. Find the velocity and acceleration of the object at time t.
√
30. x = 2t2 + 3, y = t sin 2t 33. x = t2 + 1, y = 2t
31. x = 3et − 2e−t , y = 2e3t + 7e−2t 34. x = 3t−2 , y = t2 − t
32. x = 4 cos 2t, y = 2 sin 3t 35. x = 4 tan 2t + 2, y = 3 cot 4t − 1
36. Suppose an insect is crawling around a plane so that its position at time t is
given by x = 2t + 1 + 2et , y = 7t − 3 − 4et . Find the speed of the insect at time
t, and the speed of the insect at time 1.
37. Suppose a planet is orbiting a star in such a way that its position in some
plane at time t is x = r cos t + h, y = r sin t + k. Show that the planet’s speed
is constant.
38. Suppose an ant is crawling around on an xy−plane so that its position at time
t is given by x = t + sin πt, y = t3 − 2t.
a) What is the position of the ant at time t = 1?
b) What is the ant’s velocity at t = 1?
c) What is the ant’s speed at t = 1?
d) What is the ant’s acceleration at t = 1?
e) Sketch a picture of the ant’s motion at times near t = 1, indicating the
velocity and acceleration vectors appropriately.
f) Write an integral which will compute the distance the ant crawls be-
tween times t = 1 and t = 3. Use Mathematica to compute a decimal
approximation to this integral.
39. Suppose an airplane is flying (at constant height) so that its position on an
xy−grid at time t is (x, y), where x = t sin πt and y = t cos πt.
a) Find a point in the plane that the plane flies over twice. At what times
does the plane fly over this point?
b) Find the position of the airplane at time t = 1.
c) Find the velocity of the airplane when t = 1.
d) Find the speed of the airplane when t = 1.
e) Sketch a picture of the plane’s motion at times near t = 1, indicating the
velocity and acceleration vectors appropriately.
156
f) Write an integral which will compute the distance the plane flies be-
tween times t = 1 and t = 2. Use Mathematica to compute a decimal
approximation to this integral.
40. Use Mathematica to find a decimal approximation to the length of the curve
whose parametric equations are x = ln t, y = t + 1 from t = 1 to t = 6.
41. Find the exact length of the curve whose parametric equations are x = e−t cos t,
y = e−t sin t from t = 0 to t = π2 .
42. Write an integral which will find the length of a curve whose parametric
2
equations are x = t −1
t
, y = 3t cos t from t = 1 to t = 2. Use Mathematica to
find a decimal approximation to this integral.
dy
43. Let x = 2 + sec t and y = 1 + 2 tan t. Find dx
in terms of t.
dy d2 y
44. Let x = 2 cot t and y = 2 sin2 t. Find dx
and dx2
in terms of t.
45. Find√the slope
√ of the line tangent to the graph of the parametric equations
x = t, y = t − 1 when t = 2.
46. Find the equation of the line tangent to the graph of the parametric equations
x = cos3 t, y = sin3 t when t = π4 .
47. Consider the graph of the parametric equations x = cos t + t sin t, y = sin t −
t cos t.
a) Find all t in [0, 2π] where the line tangent to the graph is horizontal.
b) Find all t in [0, 2π] where the line tangent to the graph is vertical.
48. Consider the graph of the parametric equations x = 1 − t, y = t3 − 3t.
a) Find all t where the line tangent to the graph is horizontal.
b) Find all t where the line tangent to the graph is vertical.
dy d2 y
49. Suppose x = 2t and y = 3t − 1. Find dx
and dx2
for arbitrary t, and when
t = 3.
dy d2 y
50. Suppose x = 2 cos t and y = 2 sin t. Find dx
and dx2
for arbitrary t, and when
t = π3 .
51. Write parametric equations which trace out the circle (x − 1)2 + (y + 8)2 = 9
in a clockwise direction.
2
52. Write two sets of parametric equations which trace out the ellipse (x+4)
18
+
(y−1)2
36
= 1 in a counterclockwise direction, where the speed of the second set
of equations is twice that of the first.
157
2
53. Write parametric equations which trace out the ellipse (x + 3)2 + (y−1)
49
= 1 in
a clockwise direction.
54. Suppose the graph of some set of unknown parametric equations x = f (t),
y = g(t) is given below (the direction of motion is from the upper right to the
lower right):
6
-6 -5 -4 -3 -2 -1 1 2 3 4 5 6
-1
-2
-3
-4
-5
-6
Sketch the graphs of each of the following sets of parametric equations, in-
cluding the direction of motion:
a) x = f (2t),y = g(2t)
b) x = f (−t), y = g(−t)
c) x = f (t) − 2, y = g(t)
d) x = f (t) + 1, y = g(t) − 3
e) x = 2f (t), y = g(t) − 1
f) x = −f (t), y = g(t)
g) x = −2f (t), y = −g(t)
h) x = 21 f (t) − 2, y = 2g(t) + 1
i) x = f (t + 3), y = g(t + 3)
j) x = g(t), y = f (t)
158
Answers
1. x = t, y = 2t − 3 18. x = cos t, y = 4 sin t + 2

2. x = 5 cos t − 3, y = 5 sin t + 1 (x+1)2 y2
19. ellipse 9
+ 16
=1
3. x = 4 sec t, y = 11 tan t 20. other; y = (x − 2)2
4. x = t, y = ln(2t2 + 3) 21. circle (x − 3)2 + (y + 4)2 = 4
√
5. x = 5 cos t − 2, y = sin t + 6
22. line y = x + 4
6. x = t, y = t2 − t + 1 x2 (y−1)2
√ √ 23. hyperbola 16
− 4
=1
7. x = 2 2 cos t, y = 2 2 sin t + 2
24. circle x2 + (y − 2)2 = 9
8. x = t, y = sin t
(y+2)2 (x−4)2
√ 25. hyperbola 9
− 9
=1
9. x = 3 cos t + 1, y = 6 sin t − 4
(x+3)2 y2
10. x = sec t − 1, y = 2 tan t 26. ellipse 25
+ 4
=1
√ 27. line y = 52 x + 13
11. x = 2 2 cos t − 3, 2
√
y = 2 2 sin t + 2 28. other y = cos x
12. x = 3 − 5t, y = −1 − 2t 29. line y = 4x − 11
7
13. x = −2 + 2 cos t, 30. →

−
v (t) = (4t, sin 2t + 2t cos 2t)
y = 2 + 8 sin t →
−
a (t) = (4, 4 cos 2t − 4t sin 2t)
14. x = 4 + 8 cos t,y = −1 + 3 sin t
31. →
−
v (t) = (3et + 2e−t , 6e3t − 14e−2t )
15. x = −3 + 5t, y = 5 − 6t →
−
a (t) = (3et − 2e−t , 18e3t + 28e−2t )
16. x = 5 cos t + 3, y = 5 sin t − 4 32. →
−
v (t) = (−8 sin 2t, 6 cos 3t);
√ →
−
17. x = 6 cos t + 1, y = 6 sin t − 3 a (t) = (−16 cos 2t, −18 sin 2t)
33. →
−
v (t) = (t(t2 + 1)−1/2 , 2);
→
−
a (t) = ((t2 + 1)−1/2 , −t2 (t2 + 1)−3/2 , 0)
34. →
−
v (t) = (−6e−3 , 2t − 1);
→
−
a (t) = (18e−4 , 2)
35. →
−
v (t) = (8 sec2 2t, −12 csc2 4t)
→
−
a (t) = (32 sec2 2t sec 2t, 96 csc2 4t cot 4t)
159
q
36. s = (2 + 2et )2 + (7 − 4et )2 ; s(1) = 5
37. s = r (which does not depend on t)
38. a) (1, −1)

b) (1 − π, 1)
q
c) (1 − π)2 + 1
d) (0, 6)
e) Picture omitted
R3q
f) 1 (1 + π cos πt)2 + (3t2 − 2)2 dt ≈ 22.7476
39. a) The plane flies over the point (1/2, 0) at times t = 1/2 and t = −1/2.
b) (0, −1).
c) (−π, −1)
√
d) π 2 + 1
e) Picture omitted
R2q
f) 1 [πt cos πt + sin πt]2 + [cos πt − πt sin πt]2 dt ≈ 4.82124
40. 5.38402
√
41. 2(1 − e−π/2 )
rh i2
t2 +1
+ [3 cos x − 3x sin x]2 dt ≈ 4.4799
R2
42. s = 1 t2
dy
43. dx
= 2 csc t
dy d2 y
44. dx
= −4 cos 2t sin2 t sin 2t; dx2
= 2 sin4 t(1 + 2 cos 2t + 3 cos 4t)
√
45. 2
1 1
46. y = −(x − √
2 2
) + √
2 2
47. a) t = π, t = 2π
b) t = π2 , t = 3π
2
48. a) t = −1, t = 1
b) There is no such t.

dy dy d2 y d2 y
49. dx
= 32 ; dx t=3
= 32 ; dx2
= 0; dx2 t=3
=0

dy dy −1 d2 y −1 d2 y
50. dx
= − cot t; dx t=π/3
= √ ;
3 dx2
= 2
csc3 t; dx2 t=π/3
= −4.
160
51. x = 3 cos(−t) + 1, y = 3 sin(−t) − 8

√
52. First set: x = 18 cos t − 4, y = 6 sin t + 1
√
Second set: x = 18 cos(2t) − 4, y = 6 sin(2t) + 1
53. x = cos(−t) − 3, y = 7 sin(−t) + 1
54. a) same as the given graph

b) same as given graph, but direction of motion reversed
c) shift the given graph left two units
d) shift the given graph to the right one unit and down three units
e) stretch the graph horizontally by a factor of 2 and shift down one unit
f) reflect the graph through the y−axis
g) reflect the graph through both axes, and then stretch horizontally by a
factor of 2
h) stretch vertically by factor of 2, shrink horizontally by a factor of 12 , then
shift left two units and up one unit
i) same as the given graph
j) reflect the original graph through the diagonal line y = x
161
Chapter 7
Introduction to infinite series
7.1 Motivation and big-picture questions

Consider a race of 100m between me and Usain Bolt. Assume Bolt runs 10m/s and
that I run 8m/s. However, I get a head start of 10 m.
Here is an argument that attempts to demonstrate that I will win this race:
162
7.1. Motivation and big-picture questions
Here is an argument which attempts to demonstrate that I will not win the race:
How do you reconcile these two arguments?
General questions with this setup

Given an infinite list of numbers a1 , a2 , a3 , ...
1. Classification problem: Can you add a1 + a2 + a3 + .... and get a
finite number for the answer?
2. Computation problem: If so, what is the numerical value of
a1 + a2 + a3 + ...?
Examples:
1 + 1 + 1 + 1 + 1 + 1 + ...
1 + 12 + 41 + 18 + 1
16
+ ...
163
1 + 12 + 31 + 14 + 15 + 61 + ....
1 − 21 + 13 − 14 + 51 − 16 + ...
An interesting example:
1 − 1 + 1 − 1 + 1 − 1 + 1 − 1 + 1 − 1 + ...
Solution #1:
Solution #2:
Solution #3:
Which (if any) of these solutions is correct?
This interesting example illustrates a major problem with trying to study infi-
nite series. Irrespective of which of these solutions is correct, what we know is that
they cannot all be correct. This means that we definitely cannot legally rearrange
or regroup terms when adding up infinitely many numbers.
General questions with this setup (revised)

1. Classification problem: Can you add a1 + a2 + a3 + .... and get a
finite number for the answer?
2. Computation problem: If so, what is the numerical value of
a1 + a2 + a3 + ...?
3. Rearrangement problem: When, if ever, can you legally rearrange or
regroup the terms of the sum a1 + a2 + a3 + ...?
164
More on why this question is hard

Formally speaking, addition is a binary operation, i.e. it has two inputs and one
output:
Suppose you wanted to add six numbers together, i.e.
3+8+4+7+2+9
The reason this works is that you can add any two numbers at a time and get
the same answer (this is called the associative property). For example,
3+8+4+7+2+9
But, with an infinite list of numbers there are two problems with this approach:
1.
2.
Therefore we need a mechanism to add infinite lists of numbers which goes

beyond basic arithmetic. It turns out that calculus can be used to address this
problem.
165
7.2. Convergence and divergence
7.2 Convergence and divergence

Question: What does calculus have to do with adding infinite lists?
As discussed earlier, adding up an infinite list of numbers is hard because
1. when adding the numbers two at a time, you never run out of numbers to
add; and
2. the associative property fails (i.e. regrouping and rearranging terms of an

infinite sum is generally illegal).
In calculus, we have encountered two other problems which are difficult to

solve:
HOW THE
APPROXIMATION APPROXI-
MOTIVATING OF THE MATION THEORETICAL
PROBLEM SOLUTION IMPROVES SOLUTION
Find the
slope of the
line tangent
to function
f at x
Find the
area under
the graph of
function f
from a to b
Find the
sum of
an infinite
list of numbers
a1 + a2 + a3 + ...
166
Definition 7.1 Given an infinite series a1 + a2 + a3 + a4 + ..., and given any index
N , the N th partial sum of the series, denoted SN , is
SN = a1 + a2 + a3 + ... + aN .
Note: SN is always defined, since it is a sum of finitely many numbers.
Remark: The indexing of an infinite series does not always start with the index
1. In general, the N th partial sum of an infinite series is the sum of all terms in the
series whose index is ≤ N . For example, if your series is
a5 + a6 + a7 + a8 + a9 + ...,
then the seventh partial sum of this series is
In particular, to get SN , you add up terms until you get to index N (you don’t
necessarily add up N terms).
Example: Find the second, fourth and fifth partial sums of each of the following
infinite series (assume that each series starts with a1 ):
1 − 1 + 1 − 1 + 1 − 1 + 1 − 1 + ...
1 1 1 1 1
1− + − + − + ...
2 3 4 5 6
Solution:
1 1
S2 = 1 − = .
2 2
1 1 1 7
S4 = 1 − + − = .
2 3 4 12
1 1 1 1 47
S5 = 1 − + − + = .
2 3 4 5 60
167
Definition 7.2 Let a1 + a2 + a3 + a4 + ... be an infinite series. For each N , let

SN = a1 + a2 + ... + aN be the N th partial sum of the series. Then:
1. If L is a real number such that lim SN = L, then we say that the infinite series
N →∞
a1 + a2 + a3 + ... converges (to L) and write a1 + a2 + a3 + ... = L. In this
setting L is called the sum of the series.
2. If lim SN = ±∞ or if lim SN DNE, then we say that the infinite series

N →∞ N →∞
a1 + a2 + a3 + ... diverges.
“Big picture” questions with infinite series (revised)

1. Classification problem: Does the infinite series a1 + a2 + a3 + ....
converge or diverge?
2. Computation problem: If the infinite series a1 + a2 + a3 + ... converges,
what is its sum?
regroup the terms of the infinite series a1 + a2 + a3 + ... without affecting
whether or not the series converges and without affecting the sum of the
series?
Before addressing these questions, we turn to issues with notation.
168
7.3. Σ-notation
7.3 Σ-notation
Writing a1 + a2 + a3 + ... over and over again is annoying. We shorthand this
expression by writing
∞
X
an
n=1
In this notation:
• n is the
• 1 is the
• ∞ is the
• the individual numbers a1 , a2 , a3 , ... are called the
• the number an is the
Example:
∞
X n
n=1 n + 3
Note: We will see that some of the things we want to say about series do not
depend on the starting index of the series. In this setting, we will just write an
P
to represent the series. If you see an (without the upper and lower indices indi-
P
cated), this means one of two things:
1. The starting index was given earlier in the problem and is being omitted
solely for the sake of brevity (while this is “legal”, it is not recommended
that you do this), or
2. Some property of the series is being described which does not depend on the
starting index of the series (so the starting index is irrelevant to the context).
169
7.3. Σ-notation
Example: For each series, write the series out with + signs, and give the second
term, the third partial sum, and the ninth term of the series:
∞
X 1
n=1 n
∞
X (−1)n
2
n=3 n + 2
∞
X
n
n=0
170
7.3. Σ-notation
Example: Write each of the following series in Σ notation:

5 5 5 5 5
(a) + + + + +...
7 8 9 10 11
2 2 2 2 2
(b) + + + + +...
3 7 11 15 19
7 10 13 16 19
(c) + + + + +...
8 16 25 26 27
1 1 1 1 1
(d) 1− + − + − +...
5 9 13 17 21
171
7.4. Elementary properties of convergence and divergence
7.4 Elementary properties of convergence and divergence

First, we restate the definition of convergence using Σ−notation:
Definition 7.3 Let an be an infinite series. For each N , let SN be the N th partial
P
sum of the series; this is defined to be the sum of all the an for which n ≤ N . Then:
1. If L is a real number such that lim SN = L, then we say the infinite series
N →∞
a1 + a2 + a3 + ... converges (to L) and write an = L. In this setting L is
P
called the sum of the series.
2. If lim SN = ±∞ or if lim SN DNE, then we say the infinite series

P
an
N →∞ N →∞
diverges.
Important: There is a big difference between saying “ an converges” and say-

P
ing “an converges”.
In particular, you should never omit the Σ when describing whether or not
an infinite series converges.
Now, we list some elementary results about convergence of series. They should
remind you of similar results for improper integrals (there is a good reason for this,
as we will see later).
Theorem 7.4 (Linearity I) Suppose an is an infinite series that converges to L

P
and bn is an infinite series that converges to M . Then:

P
1. The series (an + bn ) converges to L + M .

P
2. The series (an − bn ) converges to L − M .

P
3. For any constant k, the series (k an ) converges to kL.

P
172
Theorem 7.5 (Linearity II) Suppose an is an infinite series that converges to L

P
and bn is an infinite series that diverges. Then:

P
1. The series (an + bn ) diverges.

P
2. The series (an − bn ) diverges.

P
3. For any constant k 6= 0, the series (k bn ) diverges.

P
Theorem 7.6 (Linearity III) Suppose an is an infinite series that diverges and
P
bn is an infinite series that diverges. Then:

P
1. The series (an + bn ) might converge or diverge.

P
2. The series (an − bn ) might converge or diverge.

P
Essentially, the content of these theorems can be summed up in the following

phrases:
Linearity Part I: “convergent ± convergent = convergent”

“constant times convergent = convergent”
Linearity Part II: “convergent ± divergent = divergent"
“nonzero constant times divergent = divergent”
Linearity Part III: “divergent ± divergent = unknown".
These are the same maxims that applied to improper integrals.
∞
Theorem 7.7 (Starting Index is Irrelevant) Suppose an is an infinite series.
P
n=K
Then, so long as all the terms are defined, for any constant M we have:
∞ ∞
1. an converges if and only if an converges.
P P
n=M n=K
∞ ∞
2. an diverges if and only if an diverges.
P P
n=M n=K
∞
Note: Suppose that an converges. Then since starting index is irrelevant,
P
n=M
∞
then an converges as well. But in general, these two series do not converge to
P
n=K
the same sum.
173
Example: Earlier in this packet we found that

∞
X 1 1 1 1
n
= 1 + + + + ... = 2.
n=0 2 2 4 8
Use this fact to compute the following sums:

∞
X 1
(a) n
n=3 2
∞
X 4
(b) n
n=−1 2
∞
1 3
X
(c) n
+ n+1
n=2 2 2
174
7.5. Changing indices
7.5 Changing indices

Motivation: consider the following two series:
∞ ∞
X 1 X 1
n=2 n + 3 n=1 n + 4
A useful technique to master is the ability to change the starting index of a se-
ries.
Example: Rewrite the following series so that its starting index is n = 1:

∞
X (n − 1)4
n=3 3n
175
7.5. Changing indices

∞
X 4(n + 2)5
n=1 nn

∞
X (−1)n
4
n=0 n + 1
176

In problems 1-12, you are given an infinite series “written out”. Write the given
series in Σ−notation (keep in mind that there are multiple correct answers).
1. 1 − 31 + 51 − 71 + 19 − ... 7. 1
16
− 1
64
+ 1
44
− 1
45
+ ...
3 4 5 6 7
2. 25
+ 125
+ 625
+ 55
+ 56
+ ... 8. 1 + 1 + 12 + 61 + 1
24
+ 1
120
+ 1
720
+ ...
3. 1 + 31 + 91 + 1
27
+ 1
81
+ 1
35
+ ... 9. 1
14
+ 1
17
+ 1
20
+ 1
23
+ ...
1
4. −2 + 2 − 2 + 2 − 2 + 2 − 2 + 2... 10. 2
+ 24 + 83 + 4
16
+ 5
32
+ ...
4 7 10 13 16 19 1 1 1 1
5. 8
+ 15
+ 22
+ 29
+ 36
+ 43
+ ... 11. 3
− 2·32
+ 3·33
− 4·34
+ ...
6. − 29 − 2
25
− 2
49
− 2
81
− 2
121
− ... 12. 1
2
+ 16 + 1
12
+ 1
20
+ ...
In problems 13-20, you are given an infinite series in Σ−notation. For each series,
find the third term of the series (if it exists), find the ninth term of the series, and
find the fourth partial sum of the series.
∞ ∞
2n−1
13. 17.
P P
n
n!
n=1 n=0
∞ ∞
3
14. [1 + (−1)n ] 18. (−1)n 2n−1
P P
n=0 n=2
∞ ∞
1 1
15. 19.
P P
n n2 +n
n=4 n=1
∞ ∞
cos(πn) (−1)n
16. 20.
P P
n+1 n
n=1 n=1
In problems 21-28, you are given that some infinite series converges to some num-
ber. Then you are given a second infinite series which relates to the first series in
some way. Find the sum of the second series.
∞ ∞
1 1
21. Given = e, compute .
P P
n! n!
n=0 n=2
∞ ∞
1 π2 3
22. Given , compute .
P P
n2
= 6 n2
n=1 n=1
∞ ∞
(−1)n π 2n (−1)n 3π 2n
23. Given = −1, compute .
P P
(2n)! (2n)!
n=0 n=2
∞ ∞
e−2 2n 2n+3
24. Given = 1, compute .
P P
n! n!
n=0 n=3
177
∞ ∞
5n 5n
25. Given = e5 , compute .
P P
n! (n+2)!
n=0 n=2
∞ ∞
1 3
P P
2n 2n
n=0 n=−3
∞ n ∞ n+2
1 1
P P
n 2
n 2
n=0 n=2
28. Given 1 − 13 + 15 − 71 + 19 − ... = π4 , compute 2

7
− 29 + 2
11
− 2
13
+ 2
15
− ....
In problems 29-36, you are given an infinite series. Rewrite each series in Σ−notation
such that the initial term of each rewritten series corresponds to the given index.
∞
4
29. ; starting index 0
P
(n+1)(n+2)
n=3
∞
32n−1
P
n!
n=2
∞
(−1)n+1
P
n2n
n=1
∞
2n−1
P
(n−2)3 −n
n=5
∞
1
33. (−1)n e−3n ; starting index 0
P
n=1
34. 18 − 6 + 2 − 32 + 29 − 2
27
+ ...; starting index 3
35. 2 + 23 + 52 + 27 + 29 + ....; starting index 4

3 4 5 6 7 8
36. 8
− 11
+ 14
− 17
+ 20
− 23
+ ...; starting index 0
178
Answers
Note: there are multiple correct answers to numbers 1-12.
∞ ∞ ∞
(−1)n 3n+1 1
1. 5. 9.
P P P
2n+1 7n+1 3n+11
n=0 n=1 n=1
∞ ∞ ∞
n −2 n
2. 6. 10.
P P P
5n−1 (2n+1)2 2n
n=3 n=1 n=1
∞ ∞ n ∞
1 −1 (−1)n+1
3. 7. 11.
P P P
3n 4 n3n
n=0 n=2 n=1
∞ ∞ ∞
1 1
4. 2 · (−1)n 8. 12.
P P P
n! n(n+1)
n=1 n=0 n=1
13. third term is 53 ; ninth term is 17

9
; fourth partial sum is 71
12
.
14. third term is 0; ninth term is 0; fourth partial sum is 6.
15. third term is 0; ninth term is 19 ; fourth partial sum is 41 .

−1 −1 −13
16. third term is 4
; ninth term is 10
.
17. third term is 6; ninth term is 362880; fourth partial sum is 34.
−3 −3 29
18. third term is 5
; ninth term is 17
1 1 4
19. third term is 12
; ninth term is 90
−1 −1 −7
20. third term is 3
; ninth term is 9
π 26 ∞
21. e − 2 28. 2
− 15 33.
P 1
(−1)n+1 e−3n−3
n=0
π2 ∞
22. 4
29.
P
2
(n+4)(n+5) ∞ n
3 2 n=0 1
23. −6 + π 34. (−1)n+1 486
P
2 3
∞ n=3
32n−5
30.
P
24. 8e2 − 40 n=4
(n−2)!
∞
2
1 2 118 35.
P
25. 25
e − 75
∞
(−1)n 2n−7
31.
P
n=4
(n−1)2n−1
n=2
26. 48
∞ ∞
2n+7 n+3
32. 36. (−1)n 3n+8
P P
3
27. 8 n=1
(n+2)3 −n−4
n=0
179
Chapter 8
Geometric series and the Ratio

Test
8.1 Definitions
By far, the most important class of infinite series are geometric series:
Definition 8.1 A series an is called geometric if there exists a real number r such
P
that an+1 = ran for all n. The number r is called the common ratio of the series.
Why is r called the “common ratio”?
Consequence:
180
8.1. Definitions
Suppose the initial index of the geometric series is n = 0. Then, by repeated

application of the formula an+1 = ran , we see:
If we change notation and call the number a0 just “a”, we get the following im-
portant characterization of geometric series:
Theorem 8.2 (Characterization of geometric series) Every infinite geometric se-

ries can be written in the standard form
∞
arn
X
n=0
where a is the initial term of the series and r is the common ratio of the series.
In other words, every geometric series is the sum of a constant times all the
nonnegative powers of the common ratio.
If you are given a geometric series, you can always rewrite the series in the
standard form given in the theorem above, where the starting index is n = 0. To
study a geometric series, your first step should almost always be to rewrite the
series in this standard form. To do this, simply write the terms out and factor out
the initial term of the series.
181
8.1. Definitions
Example: For each series, determine if the series is geometric. If it is, write the
series in standard form and identify the common ratio of the series.
∞ n+1
X 7
(a) 3·
n=1 8
4 8 16 32
(b) 2− + − + −...
3 9 27 81
1 1 1
(c) 1+ + + +...
2 3 4
182
8.2. The Geometric Series Test
8.2 The Geometric Series Test

Question: Under what circumstances does a geometric series converge, and what
is the sum of a convergent geometric series?
Theorem 8.3 (Geometric Series Test) Consider a geometric series written in stan-
∞
dard form arn . Then:
P
n=0
1. The series converges if and only if |r| < 1 (or if a = 0).
2. The series diverges if and only if |r| ≥ 1.

a
Furthermore, if the series converges, its sum is 1−r
.
Proof of the Geometric Series Test: There are several cases to consider:
Case 1: a = 0
In this case every term of the series is 0 (since they are all of the form arn ).
Therefore the N th partial sum is
N N
arn =
X X
SN = 0=0
n=0 n=0
and since lim SN = lim 0 = 0, the series converges to 0.

N →∞ N →∞
Case 2: a 6= 0, r = 1
N N N
In this case, SN = arn = a1n =
P P P
a = a + a + ... + a = a(N + 1).
n=0 n=0 n=0
Therefore lim SN = lim a(N + 1) = ±∞ so the series an diverges.
P
N →∞ N →∞
Case 3: a 6= 0, r = −1
183
Case 4: a 6= 0, |r| =
6 1
184
Remark 1: Since you can always factor a constant a out of a series, the simplest
way to state the content of the Geometric Series Test is as follows:
∞
(
1
X
n = 1−r if |r| < 1
r
n=0
diverges if |r| ≥ 1
(If you learn nothing else from me this semester, learn this fact.)
Remark 2: The geometric series with a = 1 and r = −1 is the series
∞
1(−1)n = 1 − 1 + 1 − 1 + 1 − 1 + 1 − 1 + ...
X
n=0
which we encountered earlier in this chapter. According to the Geometric Series

Test, this series diverges (since |r| = | − 1| = 1), so none of the values for this sum
obtained earlier by various regrouping procedures are correct. This series diverges
and cannot be legally added.
Remark 3: In the context of proving the Geometric Series Test, we proved a

formula for the partial sums of a geometric series. We restate this as a theorem, as
this result will be used in examples that follow. In particular, this formula holds in
any situation where r 6= 1 (even if |r| > 1):
Theorem 8.4 (Finite Sum Formula for a Geometric Series) Consider a geomet-
∞
ric series written in standard form arn where r 6= 1. Then the N th partial sum
P
n=0
satisfies
N
X
n 2 a(1 − rN +1 )
3 N
ar = a[1 + r + r + r + ... + r ] = .
n=0 1−r
If a = 1, the above formula reduces to

N
1 − rN +1
rn = 1 + r + r2 + r3 + ... + rN =
X
.
n=0 1−r
Theorem 8.4 holds irrespective of whether |r| < 1 or |r| > 1.
185
Examples: For each finite or infinite geometric series, find the sum (if the series
diverges or is not geometric, say so):
2 2 2 2
(a) + + 5 + 7 + ...
3 27 3 3
1 1 1
(b) 4 − 2 + 1 − + − + ...
2 4 8
(c) 2 + 6 + 18 + 54 + ...
186
(d) 2+6+18+54+...+2·335
15 n
X 2
(e) 4
n=2 3
1 1 1 1
(f) 1− + − + −...
2 3 4 5
187
∞
−2 n
X
(g)
n=0 5
Solution: This geometric series is already in standard form.
−2
We have a = 1, r = 5
so this series converges to
a 1 1 5
= = = .
1−r 1 − (−2/5) 7/5 7
∞
X 2 · 15n+3
(h) 2n−1
n=1 7
∞
X 7(−1)n 32n−1
(i)
n=2 5 · 24n+3
188
8.3. Applications of geometric series
8.3 Applications of geometric series

Investments
Example: Suppose you invest $1000 each year in an account which earns 5% in-
terest each year. How much will you have in your account after 30 years (assume
that you make your annual deposit on January 1, and that “after 30 years” means
“on December 31 of the 30th year”)?
To answer this question, we first have to start with a side question. Suppose
you invest $P once into an account which pays interest rate R, compounded once
per time period. How much do you have after n time periods?
Answer to side question:
Back to the original problem:
189
Repeating decimals
Example: Write the repeating decimal .243737373737... as a fraction in lowest terms.
190
Pharmacokinetics
Example: Suppose you give a patient a 10 mg dose of a drug daily. If the patient’s
bodily functions remove 90% of whatever amount of that drug is in the patient’s
body,
(a) How much will be in the patient’s body after 14 days (just before he takes the
15th dose)?
(b) How much will end up in the patient’s body (just before he takes each dose)
if he takes the dose indefinitely?
191
Fractal geometry
Example: Consider the Koch snowflake, a figure constructed by the following
iterative process: first,
√
start with an equilateral triangle of side length 1 (the area of
3
such a triangle is 4 ).
Second, divide each side of the trangle into three segments of equal length and
attach an equilateral triangle to the middle of each segment. Then erase the middle
of each segment. After doing this, you get the following figure:
Third, repeat this procedure indefinitely. This means that at each stage, you
take each side of the figure, divide it into thirds, attach an equilateral triangle to the
middle third of each segment, and the erase the middle of each previous segment.
If you carry out this procedure, you get the following sequence of figures in the
next three steps:
Repeating this procedure infinitely many times produces a figure called the
Koch snowflake. The question is to find the perimeter of the Koch snowflake, and
find the area enclosed by the Koch snowflake.
192
SIDE
# OF LENGTH AREA
SIDES OF OF TOTAL ADDITIONAL
#∆S IN EACH ∆ EACH ∆ AREA PERIMETER
STAGE ADDED FIGURE ADDED ADDED ADDED ADDED
√ √
3 3
0 1 3 1 4 4
3
..
.
⇒ total area of Koch snowflake =
⇒ perimeter of Koch snowflake =
193
8.4. The Ratio Test
8.4 The Ratio Test

Recall our three “big picture” questions with infinite series:
“Big picture” questions with infinite series

Given an infinite series an , ...
P
1. Classification problem: Does the infinite series an converge or diverge?

P
2. Computation problem: If an converges, what is its sum?

P

regroup the terms of an without affecting whether or not the series
P
converges and without affecting the sum of the series?
In this section we start trying to solve the first of these questions - determining
whether or not a series converges.
Overall approach to the classification problem: we will develop a bunch of

“tests” which will tell us whether or not a series converges or diverges; the trick is
to figure out which test(s) to use on which series.
In this packet we develop our first test, based on reasoning coming from geo-
metric series.
Recall: a series an is geometric if there exists a common ratio r, i.e.

P
a1 a2 a3 a4
= = = = ... = r.
a0 a1 a2 a3
∞
All geometric series can be rewritten as arn ; by the GST we have
P
n=0
∞
(
a
X
n = 1−r if |r| < 1
ar .
n=0
diverges if |r| ≥ 1
Now let an be any (not necessarily geometric) infinite series. In this setting
P
a1 a2 a3 a4
, , , , etc.
a0 a1 a2 a3
are not likely to be the same number. However, it may be the case that these ratios
get closer and closer to some number ρ (this is the Greek letter “rho”), i.e.
an+1
lim
n→∞ a
= ρ.
n
194
8.4. The Ratio Test
This means that for large n, an+1 ≈ ρ so an should behave like a geometric series
P
an
with common ratio very, very close to ρ.
Consequences:
1. If ρ > 1, then an behaves like a geometric series with |r| > 1, i.e.
P
2. If ρ < 1, then an behaves like a geometric series with |r| < 1, i.e.
P
3. If ρ = 1, then an behaves like a geometric series with r ≈ 1.

P
The conclusion of all this logic is what is called the Ratio Test:
|an+1 |
Theorem 8.5 (Ratio Test) Suppose an is an infinite series and let ρ = lim .
P
n→∞ |an |
Then:
1. If ρ < 1, then an converges.

P
2. If ρ > 1, then an diverges.

P
3. If ρ = 1, or if ρ DNE, then this test tells you nothing.
Remarks:
1. Notice in this theorem we added absolute value signs; this is a computational

convenience that will help simplify some examples.
2. Since you are taking the limit of ratios which are positive, the value of ρ must
be nonnnegative. If you get ρ < 0, you have done something wrong (you
forgot the absolute values inside the limit).
Example 1: Determine whether or not the following series converges or di-

verges:
∞
X 5n
n
n=1 n7
195
8.4. The Ratio Test
Types of series for which the Ratio Test works well:
Types of series for which the Ratio Test will not work well:

verges:
∞
X 2n
n=0 n!
196
8.4. The Ratio Test

verges:
∞
X (−1)n n2 24n
n=0 32n
197
8.4. The Ratio Test

verges:
∞
X (2n)!
2
n=0 (n!)
198
8.4. The Ratio Test

verges:
∞
X (−1)n n!
n=0 nn
199
8.4. The Ratio Test
Remark: Notice that the evaluation of ρ in the Ratio Test often contains either a
simplification of exponents of the form
cn+1 cn 1
=c or = etc.
cn cn+1 c
or a simplification of factorials of the form
n! 1 (n + 1)! (2n)! 1
= or =n or = etc.
(n + 1)! n+1 n! (2n + 2)! (2n + 2)(2n + 1)
When simplifying factorial expressions, it is often useful to write out the terms
being multiplied to see how they will be cancelled. For example, the last equality
above comes from
(2n)! 2n(2n − 1)(2n − 2) · · · 3 · 2 · 1 1
= = .
(2n + 2)! (2n + 2)(2n + 1)2n(2n − 1)(2n − 2) · · · 3 · 2 · 1 (2n + 2)(2n + 1)
In general,

verges:
∞
X 2n
n=2 n − 1
2
200

In problems 1-16, find the sum of each finite or infinite series (if it converges; if the
series diverges, say so):
∞ n ∞ h n i
1 3 2
1. 9.
P P
5 2n
+ 3
n=1 n=0
∞
n+2
∞
−3 5 3
10. −
P
2.
P
6n (−1)n 3n 5
n=0 n=1
3. 9 − 92 + 49 − 89 + 9
− 9
+ ... 11. 80 + 40 + 20 + 10 + 5 + 52 + 54 + ...
16 32
∞
17 3·8n
12.
P
2
4.
P
52n−3
5n n=0
n=3
∞
2·32n−1
13.
P
11 n
3 5·24n+3
5.
P
n=2
42n
n=2
∞
14. 3−n
P
6. 2 + 4 + 8 + 16 + ... + 2100 n=0
∞ n ∞
−4 3n −5
7.
P
15.
P
3 6n
n=3 n=1
∞ ∞
6n−1
8. 16. 4n 5−n
P P
7n+1
n=2 n=0
17. Suppose the initial term of a geometric series is 3 and the series converges to
2. What is the common ratio of the series?
2
18. A geometric series with common ratio 3
converges to 8. What is the initial
term of the series?
19. Write the repeating decimal .71717171717... as a fraction in lowest terms.
20. Write the repeating decimal 1.314314314314... as a fraction in lowest terms.
23. A ball is dropped from a height of 15 ft onto a concrete slab. Each time the
ball bounces, it rebounds directly to 23 of its previous height. Find the total
distance the ball travels before it comes to rest.
24. Suppose a patient takes 25mg of a certain drug each day. If 80% of the drug
is excreted by bodily functions each day, how much of the drug will be in the
patient’s body immediately before she takes her 22nd dose of the medicine?
201
25. Suppose your drinking water contains poison, and as such you injest 2.3mg
of the poison each day. Although your body gets rid of 10% of the poison
in your body each day, when you accumulate 20mg of the poison in your
system you will be dead. How long do you have before you need to stop
drinking your water?
26. Suppose your drinking water contains poison, and as such you injest 2.3mg
of the poison each day. Suppose further that when you accumulate 250mg of
the poison in your system you will be dead. What percent of the poison in
your system does your body need to excrete daily in order to never die from
the poison? (Assume that you will live forever if the poison doesn’t kill you.)
27. In the figure below, the triangle indicated by the solid lines is an isoceles
right triangle whose height is 1 unit. All the dashed line segments drawn
are perpendicular to either the base of the triangle or the hypotenuse. If the
dashed line segments continue indefinitely, find the total length of the dashed
line segments.
In problems 28-38, Determine, with justification, whether or not the following

series converge or diverge.
∞ ∞
(−3)n n!
28. (−2)n 2−n 34.
P P
nn
n=1 n=1
∞
7n
29.
P
∞
n8 5n en
35.
P
n=4 n!
n=1
∞
(n!)3
30.
P
(3n)! ∞
n=0 (−1)n n!2n!
36.
P
∞ (3n)!
(−1)n 42n n=0
31.
P
7n
n=0
∞
n!(3n)!
37.
P
∞
nn 2
32. n=0 [(2n)!]
P
n3 n!
n=0
∞ ∞ h n i
n2014 2 n4 3n
33. 38.
P P
1.01n n!
+ 5n
n=0 n=1
202
Answers
1 1313
1. 4
20. 999
−18 634
2. 5 21. 2475
3. 6 22. 1319
9990
15
1 1
4. 50
1− 5 23. 75 ft
24. 6.25 mg
10
9 3
5. 208
1− 16
25. On the 19th day you will die.
6. 2(2100 − 1)
26. .92%
7. diverges √
6
27. 1 + 2
8. 49
28. diverges
9. 9
−179 29. diverges
10. 100
30. converges
11. 160
9375 31. diverges
12. 17
13. 27 32. diverges

2240
14. 3 33. converges

2
15. 0 34. diverges
16. 5 35. converges

−1 36. converges
17. r = 2
8 37. diverges
18. 3
71
19. 99
38. converges
203
Chapter 9
Convergence tests for positive

series
9.1 Classifying series according to sign

Recall our general questions with infinite series:
General questions with infinite series

Given an infinite series an :
P
1. Classification problem: Does the series converge or diverge?

2. Computation problem: If it converges, what is the sum?
regroup the terms of the series?
This chapter deals only with question 1 above: determining whether or not a
series converges.
We know two tests so far which tell us whether some series converge or di-
verge:
1.
2.
In general, you try to apply one of these tests if possible.
204
9.1. Classifying series according to sign
What do you do next, if these tests don’t apply? Next, look at the signs of the
individual terms of the series. This is because whether or not any of the rest of our
tests are valid depends in part on what the signs of the series are.
Definition 9.1 An infinite series is called positive if all its terms are positive; more
precisely an is positive if an ≥ 0 for all n.
P
Definition 9.2 An infinite series is called negative if all its terms are negative; more
precisely an is negative if an ≤ 0 for all n.
P
General technique to study

convergence/divergence of a negative series:
Given a negative series, factor out −1 from the series, i.e. write
(−an ) = −
P P
an .
The series an is a positive series; use one of the tests below
P
to determine whether or not it converges. By linearity, the

series (−an ) has the same behavior.
P
205
9.2. The Integral Test; harmonic and p-series
9.2 The Integral Test; harmonic and p-series

∞
Theorem 9.3 (Integral Test) Suppose an is a positive series and f is a function
P
n=0
with the following properties:
1. f (n) = an for all values of n; and
2. f is decreasing.
Then:
R∞ ∞
• if f (x) dx converges, then an converges, and
P
0
n=0
R∞ ∞
• if f (x) dx diverges, then an diverges.
P
0
n=0
Proof of the Integral Test:
0 1 2 3 4 5 6 7 8 9 10
206
Important remarks about the Integral Test

1. The Integral Test can only be applied to positive series.
2. The easiest way to check that a function f is decreasing is to show that the
derivative f 0 is negative.
3. If your series contains terms like n!, the Integral Test cannot be used (because
you cannot define a function with an x! in it whose domain contains non-
integers).
4. Although the statement of the Integral Test above uses the starting index 0 for
both the integral and the series, this is really not important (since the starting
index of a series is irrelevant).
5. Suppose f is a decreasing, positive function with f (n) = an for all n. It is im-

portant to know that while the Integral Test ensures that the series an and
P
R∞
the integral 0 f (x) dx either both converge or both diverge, in the situation
where both converge, the series and the integral do not generally converge
to the same number. For example,
Z ∞ ∞
X
f (x) dx = 3 does not imply an = 3
0 n=0
(it does mean that the sum converges to something).
6. The Integral Test should be used only as a last resort (other tests are better).
Example: Determine, with appropriate justification, whether the following se-

ries converges or diverges:
∞
X 1
n=2 n ln n
207
Recall that when studying improper integrals we came across the following
fact: Z ∞ (
1 converges if p > 1
dx
1 x p diverges if p ≤ 1
Putting this fact together with the Integral Test, we have:
Definition 9.4 An infinite series is called a p−series it is of the form

X 1
np
for a constant p > 0.
Theorem 9.5 (p-Series Test) If an is a p−series, then

P
1. an converges if p > 1.
P
2. an diverges if p ≤ 1.
P
∞
7
Example:
P
n6
n=5
∞
−8
Example:
P
√
n
n=2
∞
Example: 1 + 21 + 13 + ... = 1
P
n
n=1
The last example has a special name:
Definition 9.6 The harmonic series is the infinite series

∞
X 1 1 1 1 1
= 1 + + + + + ...
n=1 n 2 3 4 5
Definition 9.7 A harmonic series is any infinite series of the form

X A
Bn + C
where A, B and C are constants (with B 6= 0).
208
Theorem 9.8 If an is a harmonic series, then an diverges.

P P
Proof:
Z ∞ b
A A A A

= lim ln(Bx + C) = lim ln(Bb + C) − ln(B + C) = ±∞
1 Bx + C b→∞ B 1 b→∞ B B
A
so by the Integral Test, the series diverges.
P
Bn+C
Note:
1. p−series for which p = 1 are harmonic series.
2. Geometric series vs. p−series: if the variable of summation is n and p is a

constant, then
Example: Determine, with appropriate justification, whether or not the follow-

ing series converge or diverge:
∞
(a) 2n−7/2
P
n=3
∞
2
(b)
P
3n+4
n=3
∞
2 3
(c) −
P
n n3
n=3
∞
−2
(d)
P
√
n
n=3
209
9.3. The nth -term Test
∞
2n2
(e)
P
5n7
n=3
9.3 The nth-term Test

Idea:
Theorem 9.9 (nth Term Test) Suppose lim an 6= 0,

an is an infinite series. If n→∞
P
then an diverges.
P
Important:
lim |an | = 0. Therefore:

lim an = 0 if and only if n→∞
Note: n→∞
Theorem 9.10 (nth Term Test, restated) Suppose an is an infinite series. If lim
P
n→∞
|an | =
6 0 (this includes when this limit is ∞ or DNE), then an diverges.
P
210
9.3. The nth -term Test
Examples: Determine, with appropriate justification, whether or not the fol-

lowing series converge or diverge.
∞
X 1
(a)
n=2 n2 +1
∞
3 7
X
(b) n
− 2
n=1 2 n
∞
X n+1
(c)
n=0 n + 3
211
9.4. The Comparison Test
9.4 The Comparison Test

Based on the same logic we used with improper integrals, we get the following
test for series:
Theorem 9.11 (Comparison Test) Suppose 0 ≤ an ≤ bn for all n. Then:
1. If the infinite series an diverges, then bn diverges.

P P
2. If the infinite series bn converges, then an converges.

P P
Remarks:
1. The Comparison Test is only useful for positive series.
2. Be very careful with the logic!

Think of an as the “small series” and think of bn as the “big series”:
P P
What you can do with the Comparison Test:

• You can conclude that the small series converges.
• You can conclude that the big series diverges.
What you cannot do with the Comparison Test:
• The Comparison Test never allows you to conclude that the big series
converges.
• The Comparison Test never allows you to conclude that the small series
diverges.
Idea: Use this test with series that are similar to a “simpler” series (usually the
simpler series is a geometric or p−series). Reason as follows:
the the the the

given ≤ “simpler” “simpler” ≤ given
series series series series
Conclusion:
the “simpler” No conclusion can
By the Comparison
series be drawn from the
Test, the given series
converges Comparison Test
converges.
Conclusion:
the “simpler” No conclusion can
By the Comparison
series be drawn from the
Test, the given series
diverges Comparison Test
diverges.
212
Classes of series which suggest the use of the Comparison Test:

1. Series whose terms contain addition in the denominator of a fraction.
2. Series whose terms contain subtraction in the denominator of a fraction.
3. Series whose terms contain sines and cosines.
1. Addition in the denominator. If the terms have form

4+?
2. Subtraction in the denominator. If the terms have form

4−?
3. Series whose terms contain sines or cosines. If the terms contain some ex-
pression of the form cos or sin
Examples: Determine, with appropriate justification, whether or not the fol-

lowing series converge or diverge.
∞
X 5
(a)
n=1 n3 + 3n + 8
213
∞
1 1
X
(b) −
n=1 n n+4
∞
X 3 − cos n
(c)
n=1 n
∞
3n + 7n2
X
(d)
n=1 12n6
∞
X n
(e)
n=0 n4 +1
Solution: Notice 0 ≤ n4n+1 ≤ nn4 = n13 .

P 1
3 converges
nP
(it is a p-series with p = 3 > 1).
⇒ an converges by the Comparison Test.
214
∞
X 2
(f) √
n=10
3
n−3
∞
2
e−n
X
(g)
n=1
∞
X (−1)n+1
(h)
n=1 n
215

Determine whether the following series converge or diverge. You should state the
name of the test(s) you use and completely justify your reasoning, giving argu-
ments like those in the examples of this text.
∞ ∞
3 2n2 +3n−2
1. 12.
P P
n4 n2 +4n+1
n=1 n=1
∞ ∞
3 1 1
2. 13. −
P P
5n−3 k k+1
n=1 k=1
∞ ∞
−3 1
3. 14.
P P
√
2n n ln(2n )
n=1 n=4
∞ ∞
2+cos(3n)
4. e−1/n 15.
P P
n
n=2 n=1
∞ ∞
2n2 +3 1
16.
P
5.
P
5n2 −4 en +e−n
n=2 n=0
∞ ∞
4 2 3+cos(2n)
17.
P
6.
P
n5
+ n 4n2
n=1 n=0
∞ ∞
ln n 3+cos(2n)
18.
P
7.
P
√
n 43n
n=3 n=0
∞ ∞
3+2n
19.
P
8.
P
ln n 3n +4
n=1 n=2
∞ ∞
arctan n 4+sin(n2 +2n)
20.
P
9.
P
√ 3
n2 +1 n=1 n5 +1
n=1
∞ ∞ 2 −3n
(n − 1)−1/2 21. 5−n
P
10.
P
n=2 n=2
Hint: Change the starting index ∞

3
22.
P
4+sin4 (2k)
to n = 1. k=1
∞ ∞
2 6n
11. 3ne−n 23.
P P
62n +3
n=1 n=1
216
Answers
1. converges 13. converges

4. diverges
16. converges
5. diverges
17. converges
6. diverges
18. diverges
7. diverges
19. converges
8. diverges
20. converges
9. converges
217
Chapter 10
Absolute and conditional

convergence
10.1 Alternating series

Recall: we are trying to determine whether a given series an converges or
P
diverges.
Example: Consider the series

∞
X (−1)n+1 1 1 1 1 1
= 1 − + − + − + ...
n=1 n 2 3 4 5 6
This series is called the alternating harmonic series. Does it converge or diverge?
• This series is not geometric ⇒
• This series does not contain exponentials or factorials ⇒
• This series is not a p-series ⇒
• This series is neither positive nor negative ⇒
• lim |an | =
n→∞
218
10.1. Alternating series
Now what?
Back to the definition of convergence: to say an = L means

P
∞
(−1)n+1
Let’s compute the partial sums of the alternating harmonic series :
P
n
n=1
S 1 = a1 = 1
1
S 2 = a1 + a2 = 1 − =
2
1 1
S 3 = a1 + a2 + a3 = 1 − + =
2 3
S4 =
Sn
S1 =1
S3 =56
S5
S6
S4
S2 =12
n
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Observe:
219
Here is a theorem which generalizes the argument on the previous pages:
Theorem 10.1 (Alternating Series Test (AST)) Suppose an is an infinite series

P
such that:
1. an is alternating (i.e. the terms being added alternate between positive and
P
negative);
2. lim |an | = 0; and

n→∞
3. |an | ≥ |an+1 | for all n.
Then an converges.
P
Important:
How to classify an alternating series

1. Verify that the series is alternating.
2. If the Ratio Test is appropriate, use it.
lim |an |.
3. Otherwise, compute n→∞
• If this limit is nonzero, the series diverges by the nth -term Test.
• If this limit is zero, verify that |an | ≥ |an+1 | (it will in Math 230). Then
the series converges by the AST.
220
Example: Determine whether or not the following series converges or diverges:

X cos(πn)
n2 + 1
Example: Determine whether or not the following series converges or diverges:

∞
n
(−1)n+1
X
n=2 n+4
221
10.2. The triangle inequality
10.2 The triangle inequality

Motivating example: Does the following series converge or diverge?
∞
X sin(en ) −n
n
e
n=1 | sin(e )|
Problem: This series is not geometric, not a p−series, not positive (so the In-
tegral and Comparison Tests are no good) and not alternating (so the Alternating
Series Test is no good). The only test we know so far that we could use to study
this is the nth −term Test:
A new idea in the study of series is the following important concept about num-
bers:
Theorem 10.2 (Triangle Inequality for R) For all real numbers a and b, |a + b| ≤
|a| + |b|.
Proof of Triangle Inequality for R:
ab ≤ |ab|
2ab ≤ 2|ab|
a + 2ab + b2 ≤ |a|2 + 2|ab| + |b|2
2
(a + b)2 ≤ (|a| + |b|)2

√ √
(take of both sides; note z 2 = |z|)
|a + b| ≤ ||a| + |b|| = |a| + |b|.
Reason this is called the Triangle Inequality:
222
10.2. The triangle inequality
Theorem 10.3 (Generalized Triangle Inequality for R) For any finite list of real
numbers a1 , a2 , ..., an ,
n n
X X

aj ≤ |aj |.
j=1 j=1
Theorem 10.4 (Triangle Inequality for Infinite Series) Let an be an infinite

P
series. If |an | converges, then an also converges.

P P
(In this case, we have | an | ≤ |an |.)

P P
Application: Consider the series

∞
X sin(en ) −n
n
e
n=1 | sin(e )|
(this was the motivating example from earlier):
If we let this series be denoted by an , then

P

X sin(en ) X 1 n
−n −n −n
X X X
|an | = | sin(en )|
e = |±1| e = e =
e
Since |an | converges, so does an by the triangle inequality.

P P
Proof of the Triangle Inequality for Infinite Series: Let bn = an + |an |. Notice
that if an < 0, an +|an | = an −an = 0 and if an ≥ 0, then an +|an | = |an |+|an | = 2|an |.
Thus bn is always either 0 or 2|an |. In particular, we have
0 ≤ bn ≤ 2|an |.
Now we are given as a hypothesis of this theorem that |an | converges, so by

P
linearity 2|an | converges, and by the Comparison Test, bn converges. Last, we

P P
see X X X X
an = (bn − |an |) = bn − |an |
is the difference of two convergent series, hence an converges.
P
223
10.3. Absolute and conditional convergence
10.3 Absolute and conditional convergence

Definition 10.5 Let an be an infinite series. We say the series is absolutely con-
P
vergent (or that the series converges absolutely) if |an | converges.

P
Remarks:
• By the Triangle Inequality for Infinite Series, we know that if a series is abso-
lutely convergent, then it converges.
• If a series diverges, then it cannot converge absolutely (this is the contrapos-

itive of the immediate preceding statement).
• If a series an is positive, then there is no difference between |an | and

P P
an , so saying that a positive series converges is the same as saying that it

P
absolutely converges.
• If a series is negative, then |an | = − an so to say a negative series con-

P P
verges is the same as saying that it absolutely converges.
• Based on these observations, there are three possibilities for an infinite series
an :
P
1. The series an converges absolutely (i.e. |an | converges).

P P
2. The series an diverges.

P
3. Something else (which given the remarks above must be that an con-
P
verges but |an | diverges).

P
Definition 10.6 Let an be an infinite series. If an converges but |an | diverges,

P P P
then we say an is conditionally convergent (or that the series converges con-
P
ditionally).
Based on the ideas developed thus far, we can now create a Venn diagram
which incorporates the sign of a series together with whether the series converges
absolutely, converges conditionally, or diverges. This Venn diagram is on the next
page:
224
10.3. Absolute and conditional convergence
CONVERGENT SERIES
POSITIVE ALTERNATING
SERIES SERIES
COND’L.
CONV’T
ABSOLUTELY
CONVERGENT
NEITHER
POSITIVE NOR
ALTERNATING
ALL INFINITE SERIES
We can now refine one of our “big picture” questions with infinite series:
General questions with infinite series

Given an infinite series an :
P
1. Classification problem: Does the series converge absolutely, converge

conditionally, or diverge? (I.e. where on the above Venn diagram does it
belong?)
2. Computation problem: If it converges, what is the sum?
regroup the terms of the series?
For now, we turn our attention to the third question: the rearrangement prob-
lem.
225
10.4. Rearrangement of infinite series
10.4 Rearrangement of infinite series

Goal: Determine the circumstances under which the terms of an infinite series
can be legally rearranged without affecting the sum.
Motivating example: The alternating harmonic series

∞
1 1 1 1
(−1)n+1
X
= 1 − + − + ...
n=1 n 2 3 4
Does this series converge or diverge?
Does this series converge absolutely or converge conditionally?
Note: The alternating harmonic series is the prototypical example of a condi-

tionally convergent series.
∞
Let L = (−1)n+1 n1 = 1 − 21 + 13 − 14 + ....
P
n=1
Since the series alternates and has positive initial term, L ∈ [ 12 , 1].
Suppose it was legal to rearrange this series. Then:

1 1 1 1 1 1 1
L=1− + − + − + − + ...
2 3 4 5 6 7 8
1 1 1 1 1 1 1
=1− − + − − + − − ...
2 4 3 6 8 5 10
1 1 1 1 1 1 1 1

= 1− − + − − + − − + ...
2 4 3 6 8 5 10 12
=
226
Further investigation: Look only at the terms of the alternating harmonic series
which are positive:
∞
1 1 1 X 1
1 + + + + ... =
3 5 7 n=1 2n + 1
Now look only at the terms of the alternating harmonic series which are negative:
∞ ∞
1 1 1 1 1 X −1 −1 X 1
− − − − − − ... = =
2 4 6 8 10 n=1 2n 2 n=1 n
Consequence:
What happened when the series was rearranged?
Loosely speaking, the problem is that this series converges conditionally (as
opposed to absolutely). Any conditionally convergent series is comprised of an
infinite amount of “positive stuff” and an infinite amount of “negative stuff”, and
the series only converges because the terms cancel each other out in a very delicate
way. On the other hand, an absolutely convergent series (almost by definition) has
only a finite amount of “positive stuff” and a finite amount of “negative stuff”, so
no matter how you rearrange the series you always get the same thing. To sum-
marize:
Theorem 10.7 (Rearrangement Theorem) Suppose an is an infinite series.

P
1. If an converges conditionally, then the terms of that series can be rearranged

P
so that the rearranged series converges to any number you like! (The series can
also be rearranged so that the rearranged series diverges.)
2. If an converges absolutely to L, then no matter how the terms of the series are
P
regrouped or rearranged, the rearranged series still converges absolutely to L.
227
Additional remarks:
1. Geometric series always converge absolutely or diverge (they never converge

conditionally), since by the GST their convergence depends on |r| (rather
than just on r).
2. If a series is shown to converge by the Ratio Test, then the series converges
absolutely (since when evaluating ρ one takes absolute values of the terms in
the series).
In particular, for any conditionally convergent series, if you tried the Ratio
Test you would always get ρ = 1. (However, if ρ = 1 the series could con-
verge absolutely, converge conditionally or diverge.)
3. p-series converge absolutely when they converge (since they are positive).
4. If a series is positive or negative, it must converge absolutely if it converges.
5. Since the Integral Test and Comparison Test apply only to positive series, if
you use one of these to show that a series converges, then the series must
converge absolutely.
6. The Alternating Series Test tells you only that a series converges (not whether
the series converges absolutely or conditionally). To distinguish between
these situations, you need further analysis.
Putting together everything we know, we have the outline on the next page
which describes how to determine whether a given series is absolutely convergent,
conditionally convergent or divergent.
228
10.5. Summary of classification techniques
10.5 Summary of classification techniques

In general, follow these directions to classify most infinite series:
∞
1. If the series is geometric, then write in the standard form arn .
P
n=0
If |r| ≥ 1, the series diverges by the GST.

a
If |r| < 1, then the series converges absolutely to 1−r
by the GST.
2. If the series contains only multiplication/division and has at least one
|an+1 |
exponential or factorial term, try the Ratio Test: compute ρ =n→∞
lim |an |
.
If ρ > 1, the series diverges by the Ratio Test.
If ρ < 1, the series converges absolutely by the Ratio Test.
If ρ = 1, the Ratio Test is inconclusive.
lim |an |.
3. If the series is alternating, compute n→∞
If this limit is nonzero, the series diverges by the nth -term Test.
If this limit is zero, verify |an | ≥ |an+1 | and conclude that the series
converges by the AST. Then, consider the series |an |.
P
If |an | converges, then an converges absolutely.

P P
If |an | diverges, then an converges conditionally.

P P
4. If the series is a p-series, then by the p-series Test it converges absolutely if

p > 1 and diverges if p ≤ 1.
6 0, then the series diverges by the nth -term Test.
lim |an | =
5. If n→∞
6. If the series can be split, analyze the two pieces separately and use
linearity rules.
7. If the series is positive (or if it is negative, in which case you first factor
out a (−1) to leave a positive series):
(a) If the series contains addition or subtraction in the denominator, or
sines and/or cosines, try the Comparison Test.
(b) If the terms of the series look like a function you can integrate, try
the Integral Test (often necessary if the series has ln in it).
229
10.6. Examples
10.6 Examples
In each of the following examples, you are to classify the given series as ab-
solutely convergent, conditionally convergent, or divergent, giving appropriate
justification.
∞
X (−1)n nn
(a)
n=1 4n
230
10.6. Examples
∞
(−1)n n−1/3
X
(b)
n=1
∞
X 2n4
(c) 8 11
n=1 n + 3n + 12
231
10.6. Examples
∞
(−1)n
!
X 3
(d) 2
+ n
n=1 n 4
1 1 1 1 1
(e) 1− + − + − 5 +...
5 25 125 625 5
232

In problems 1-6, determine whether the following series converge or diverge. Com-
pletely justify your reasoning:
∞ ∞ n+1
(−1)
1. (−1)n 8n−3/4 4.
P P
√
3n+8
n=1 n=1
∞ ∞
(−1)n n2 4 cos(πn)
2. 5.
P P
3n 2n +3
n=2 n=3
∞ ∞
4
5n +3 (−1)n n2
3. (−1)n 20n4 +2n 6.
P P
2 +n+1 ln(n+4)
n=4 n=0
In problems 7-22, classify the following statements as true or false:
7. If a negative series converges, then it must converge absolutely.
8. If a series converges conditionally, then its terms can be legally rearranged

without affecting the sum.
9. If a series |an | diverges, then an must also diverge.

P P
10. If a series |an | diverges, then an cannot converge absolutely.

P P
11. If a series an diverges, then |an | must also diverge.

P P
12. If a series an converges, then |an | must also converge.

P P
13. If a series |an | converges, then an must also converge.

P P
14. It is possible for an alternating series to diverge.
15. It is possible for an alternating series to converge absolutely.
16. It is possible for an alternating series to converge conditionally.
17. It is possible for a positive series to diverge.
18. It is possible for a positive series to converge absolutely.
19. It is possible for a positive series to converge conditionally.
20. If lim |an | = 1, then an diverges.

P
n→∞
21. If n→∞
lim an = 0, then an converges.
P
lim |an | = 0, then

22. If n→∞ an converges.
P
233
In problems 23-34, determine whether the following series converge absolutely,

converge conditionally, or diverge. You should state the name of the test(s) you use
and completely justify your reasoning, giving arguments like those in the examples
of this chapter.
∞
23. (−1)n sin n
P
n=2
∞
n+1
24.
P
ln(2n−5)
n=4
∞
2 cos(πk)
25.
P
k2
k=1
∞ n
(−1)
26.
P
√
2 n+2
n=0
∞
27. (−1)n+1 n2n+1
P
n=1
∞
(−1)n 2n
28.
P
n3 +5
n=0
∞
(−1)n 42n
29.
P
7n n!
n=0
∞
30. 4(−1)n+1 n−3/5
P
n=1
∞ h
(−1)n
i
3
31.
P
n5
+ n
n=0
∞ 1
(−1)[ 2 n(n+1)]
32.
P
4n +n2
n=1
∞
n
33. (Challenge)
P
ln n+1
n=1
1 1 1 1
34. (Challenge) 1 + 1.1
+ 1.11
+ 1.111
+ 1.1111
+ ...
234
Answers
1. converges 18. True
2. converges 19. False
3. diverges 20. True
7. True 24. diverges
8. False 25. converges absolutely
9. False 26. converges conditionally
10. True 27. converges conditionally
11. True 28. converges absolutely
12. False 29. converges absolutely
15. True 32. converges absolutely
16. True 33. answer omitted
17. True 34. answer omitted
235
Chapter 11
Taylor series
11.1 Uniqueness of power series

Goal: Use the ideas of infinite series to study functions.
Main Problem: Given a function f , find numbers a0 , a1 , a2 , ... so that we can

write f as a series of the following form:
∞
an xn = a0 + a1 x + a2 x2 + a3 x3 + ...
X
f (x) =
n=0
Theoretical Solution: Suppose this can be done. Then by repeatedly differen-

tiating f , we get
f (x) = a0 + a1 x + a2 x2 + a3 x3 + a4 x4 + a5 x5 + ...
f 0 (x) = a1 + 2a2 x + 3a3 x2 + 4a4 x3 + 5a5 x4 + ...
f 00 (x) = 2a2 + 3 · 2a3 x + 4 · 3a4 x2 + 5 · 4a5 x3 + ...
f 000 (x) = 3 · 2a3 + 4 · 3 · 2a4 x + 5 · 4 · 3a5 x2 + ...
.. .
. = ..
f (n) (x) = n · (n − 1) · (n − 2) · · · 3 · 2 · 1an + (const)x + (const)x2 + ...
236
11.1. Uniqueness of power series
Plugging in x = 0 to all these expressions, we get
f (0) =
f 0 (0) =
f 00 (0) =
f 000 (0) =
.. ..
. = .
f (n) (0) =
Importantly, this gives us a formula for all the an :
We have proven the following:
Theorem 11.1 (Uniqueness of power series) Suppose f is a function which can

be differentiated over and over again at x = 0. Then if we write f as a power series of
the form
∞
an x n ,
X
f (x) =
n=0
then the coefficients an must satisfy
f (n) (0)
an = for all n.
n!
In other words, the only power series of the form an xn which can represent f is the
P
series
∞
X f (n) (0) n
x .
n=0 n!
This series is called the Taylor series (centered at 0) of f or the Maclaurin series
of f .
Corollary 11.2 Suppose you have two power series which represent the same function
of x, i.e.
∞ ∞
an x n = bn x n
X X
n=0 n=0
for all x in some open interval containing 0. Then, for every n, an = bn .
237
Note: There are functions which cannot be represented as a power series. How-
ever, we will not encounter those in Math 230 (so we will assume that every func-
tion f is equal to its Taylor series at all x for which the Taylor series converges).
Example 1: f (x) = ex .
Note that since the exponential function is its own derivative, we have f (n) (x) =
ex for all n and we know therefore that f (n) (0) = e0 = 1 for all n. Thus, for all n, we
have
an =
so the Taylor series of f is
One can show that this series converges for all x using the Ratio Test.
238
We have obtained the following formula (which must be memorized):
∞
xn x2 x3 x4
ex =
P
n!
=1+x+ 2
+ 3!
+ 4!
+ ...
n=0
Example 2: f (x) = sin x.
n f (n) (x) f (n) (0) an
Notice the pattern:
239
Thus the Taylor series of sin x is
This series can also be shown to converge for all x using the Ratio Test. We have
obtained the following formula (which must be memorized):
∞
(−1)n x2n+1 x3 x5 x7 x9
=x− − − ...
P
sin x = (2n+1)! 3!
+ 5! 7!
+ 9!
n=0
Example 3: f (x) = cos x
We could go through the procedure of Example 2 again and find a pattern with
the an . But there is a better way:
∞
(−1)n x2n x2 x4 x6 x8
=1− − − ...
P
cos x = (2n)! 2
+ 4! 6!
+ 8!
n=0
240
There is one other function whose Taylor series we “know” (even though we
didn’t call it a Taylor series at the time). Suppose we set an = 1 for all n. Then we
get
∞ ∞ ∞
an x n = 1xn = xn
X X X
n=0 n=0 n=0

1
This series is equal to 1−x
, but only when |x| < 1. Thus we have (remember
this):
∞
1
xn = 1 + x + x2 + x3 + ... for −1 < x < 1
P
1−x
=
n=0
This last series can be manipulated to find Taylor series of some other functions.
If you replace the x on both sides of the above equation with −x, we get
1
= 1 + x + x2 + x3 + ...
1−x
1
=
1+x
Now if we integrate both sides of the equation above, we get

1
= 1 − x + x2 − x3 + ...
1+x
ln(1 + x) =
We have derived:
∞
(−1)n+1 n
= x − 12 x2 + 13 x3 − 14 x4 + ... for −1 < x ≤ 1
P
ln(1 + x) = x
n
n=1
1
Now go back to the series for 1−x , and let’s do some different manipulations.
2
First, replace x with −x , and then integrate:
241
1
= 1 + x + x2 + x3 + ...
1−x
1
=
1 + x2
We have derived:
∞
(−1)n 2n+1
= x − 31 x3 + 15 x5 − 17 x7 + ... for −1 ≤ x ≤ 1
P
arctan x = 2n+1
x
n=0
242
To summarize, we have derived Taylor series representations for six common

functions.
You must know these cold.
The “big six” Taylor series
∞
xn x2 x3
ex = + ... (holds for all x)
P
n!
=1+x+ 2
+ 3!
n=0
∞
(−1)n x2n+1 x3 x5 x7
=x− − + ... (holds for all x)
P
sin x = (2n+1)! 3!
+ 5! 7!
n=0
∞
(−1)n x2n x2 x4 x6
=1− − (holds for all x)
P
cos x = (2n)! 2
+ 4! 6!
+ ...
n=0
∞
1
xn = 1 + x + x2 + x3 + ... (holds when x ∈ (−1, 1))
P
1−x
=
n=0
∞
(−1)n+1 n x2 x3 x4
=x− − (holds when x ∈ (−1, 1])
P
ln(1 + x) = x n 2
+ 3 4
+ ...
n=1
∞
(−1)n 2n+1 x3 x5 x7
=x− − + ... (holds when x ∈ [−1, 1])
P
arctan x = x
2n+1 3
+ 5 7
n=0
From these six Taylor series, one can obtain the Taylor series for lots of other
functions via manipulations:
243
Example: Find a power series representation (i.e. find the Taylor series) of
each of the following functions (give the answer in Σ−notation, and in written-out
form):
(a) f (x) = arctan x3
5
(b) f (x) = x2 e−3x
x
(c) f (x) = (1−x)2
(d) f (x) = x3 cos(2x)

∞
(−1)n x2n
Solution: Start with cos x = .
P
(2n)!
n=0
∞ ∞
(−1)n (2x)2n (−1)n 4n x2n
Replace x with 2x to get cos 2x = .
P P
(2n)!
= (2n)!
n=0 n=0
∞ n n 2n ∞
4 x (−1)n 4n x2n+3
Now multiply by x3 in front to get x3 cos 2x = x3 (−1)(2n)! .
P P
= (2n)!
n=0 n=0
(e) f (x) = 2 sin x4

∞
(−1)n x2n+1
Solution: Start with sin x = .
P
(2n+1)!
n=0
∞ ∞
(−1)n (x4 )2n+1 (−1)n x8n+4
Replace x with x4 to get sin x4 = .
P P
(2n+1)!
= (2n+1)!
n=0 n=0
∞ n 8n+4
x
Now multiply by 2 in front to get 2 sin x4 = 2 (−1) .
P
(2n+1)!
n=0
4
(f) f (x) = 3+5x
244
11.2. Applications of Taylor series
11.2 Applications of Taylor series

Evaluation of hard limits without L’Hôpital’s rule
Example: Evaluate the limit
6 sin x − 6x + x3
lim
x→0 2x5
Old solution using L’Hôpital’s Rule:
6 sin x − 6x + x3 0
lim 5
=
x→0 2x 0
2
L 6 cos x − 6 + 3x 0
= lim 4
=
x→0 10x 0
L −6 sin x + 6x 0
= lim =
x→0 40x3 0
L −6 cos x + 6 0
= lim =
x→0 120x2 0
L 6 sin x 0
= lim =
x→0 240x 0
L 6 cos x 6 1
= lim = = .
x→0 240 240 40
New solution using Taylor series:
245
Example: Evaluate the limit

arctan 3x − 3x
lim
x→0 x2
Approximation of function values and definite integrals

First, we introduce some language which describes the partial sums of a Taylor
series of a function.
Definition 11.3 Suppose f is a function which can be differentiated over and over
again at x = 0; the Taylor series of f is
∞
X f (n) (0) n
x .
n=0 n!
If we truncate this series at the N th power term, we obtain a partial sum of the Taylor
series called the N th Taylor polynomial (centered at 0) (a.k.a. Taylor polynomial
of order N ) of f . This polynomial is denoted PN (x).
Example: Let f (x) = ex . Then the Taylor series of f is

∞
X xn x2 x3
f (x) = =1+x+ + + ...
n=0 n! 2! 3!
246
and the Taylor polynomials of f (x) = ex are:
P0 (x) =
P1 (x) =
P2 (x) =
P3 (x) =
..
.
PN (x) =
Example: Let f (x) = sin x. Then the Taylor series of f is

∞
X (−1)n x2n+1 x3 x5 x7
f (x) = =x− + − + ...
n=0 (2n + 1)! 3! 5! 7!
x3 x5 x7
= 0 + x + 0x2 − + 0x4 + + 0x6 − + ...
3! 5! 7!
and the Taylor polynomials of f (x) = sin x are:
P0 (x) =
P1 (x) =
P2 (x) =
P3 (x) =
P4 (x) =
P5 (x) =
247
General properties of Taylor polynomials:
1. PN (x) is a polynomial of degree ≤ N ;
2. P0 (x) is the constant function of height f (0);
3. P1 (x) is the tangent line to f when x = 0;
4. PN (x) is the best N th degree polynomial approximation to f near 0.
The point of Taylor series and Taylor polynomials is that if a function f has
th
N Taylor polynomial PN , then PN (x) ≈ f (x) (this approximation improves as N
increases, but is pretty good even for small N ). For example, let’s look at the Taylor
polynomials for ex and sin x graphically:
8
-4 -2 2 4
-10 -5 5 10
-1
-2
-3
So to approximate the value of a function or a definite integral, we can replace

the function with its N th Taylor polynomial to get a good approximation.
248
Example: Approximate ln(1.2) using the fourth Taylor polynomial (a.k.a. Tay-
lor polynomial of order 4) for an appropriately chosen function.
Example: Approximate e.1 using a Taylor polynomial of order 3 for an appro-

priately chosen function.
R1
Example: Approximate 0 sin(x4 ) dx by replacing the integrand with its tenth
Taylor polynomial.
249
Higher-order derivatives
Example: Let f (x) = x20 cos(3x8 ). Find f (100) (0), the 100th derivative of f at zero.
Analysis of numerical series

P∞
Example: Determine whether or not the series n=1 e1/n − 1 converges or di-
verges.
250
Example: Find the sum of the following series:

1 1 1 1
1− + − + − ...
2 3 4 5

1 1 1 1 1
1+ + + + + 5 ...
2 4 · 2! 8 · 3! 16 · 4! 2 · 5!
2π 3 2π 5 2π 7
π− + − + ...
23 · 3! 25 · 5! 27 · 7!
251
11.3. General theory of power series
11.3 General theory of power series

In the last handout, we saw how to use certain series to study functions. In
particular, given a function f , we can define the Taylor series of f :
∞
X f (n) (0) n
x
n=0 n!
The partial sums of this series are denoted Pn (x) and called the Taylor polynomials
of f . Ideally, f (x) ≈ Pn (x), so f can be replaced by Pn in approximation problems.
Now, we want to study the general behavior of series of the form

∞
an x n ,
X
n=0
where the an are constants (which may or may not have anything to do with some
function f ). Such functions are called power series (centered at 0). In fact, we can
study more general objects:
Definition 11.4 A power series (in x) centered at x = a is a function of the form

∞
an (x − a)n
X
f (x) =
n=0
= a0 + a1 (x − a) + a2 (x − a)2 + a3 (x − a)3 + a4 (x − a)4 + ...
where the a0 , a1 , ... are real numbers. A power series (in x) is a power series centered
at x = 0, i.e. a power series is a function of the form
∞
an x n
X
f (x) =
n=0
= a0 + a1 x + a2 x2 + a3 x3 + a4 x4 + ...
The numbers a0 , a1 , a2 , ... are called the coefficients of the power series; the terms of
the series are a0 , a1 x, a2 x2 , ...
Note: When evaluating a power series, 00 is always taken to be 1 (even though

technically 00 is an indeterminate form).
Most important question regarding power series: What is the domain of a

power series?
252
Example 1: For what x does the following series converge?

∞
X (x − 6)n
n=1 n4n
Solution: Apply the Ratio Test (we expect ρ to depend on x):

∞
n!(x + 1)n
X
n=0
Solution: Again apply the Ratio Test:
253

∞
X xn
n=0 3n!
Solution: Again apply the Ratio Test:
Notice the pattern in the three preceding examples:
254
∞
Theorem 11.5 (Cauchy-Hadamard (C-H) Theorem) Let an (x−a)n be a power
P
n=0
series centered at a. There exists a quantity R (which is either a nonnegative real num-
ber or ∞) called the radius of convergence of the series such that:
1. If 0 < R < ∞, then:

a) the power series converges absolutely for x ∈ (a − R, a + R);
b) the power series diverges for x ∈ (−∞, a − R) or x ∈ (a + R, ∞);
c) anything can happen when x = a − R or x = a + R.
2. If R = 0, then the power series converges absolutely when x = a but diverges

for all other x.
3. If R = ∞, the power series converges absolutely for all x.
If you don’t like using the Ratio Test, you can find the value of R by the follow-
ing formula:
∞
Theorem 11.6 (Abel’s Formula) Let an (x − a)n be a power series centered at a.
P
n=0
Then, the radius of convergence of this power series is given by
|an |
R = lim
n→∞ |an+1 |
(assuming this limit exists).
Remarks:
1. Abel’s Formula uses only the coefficients of the power series and not the powers
of (x − a).
2. Abel’s Formula looks like the computation done in the Ratio Test, but the
fraction used to compute R is “upside-down” compared to what was used to
compute ρ in the Ratio Test.
3. Abel’s Formula has nothing to do with numerical series (those with no xn or

(x − a)n ); it is only useful for power series.
255
According to the C-H Theorem, the set of x for which a power series converges
is an interval; this interval is called the interval of convergence of that series and is
the domain of that series.
General procedure to find the interval of convergence of a power

∞
series an (x − a)n
P
n=0
1. Read off the value of a (where the power series is centered).
2. Compute the radius of convergence R using Abel’s Formula:
|an |
R = lim .
n→∞ |a
n+1 |
3. If R = 0, then the series converges absolutely for x = a and diverges for all
other x. The interval of convergence is [a, a] or just {a}.
4. If R = ∞, then the series converges absolutely for all x. The interval of

convergence is (−∞, ∞).
5. If 0 < R < ∞, then by the C-H Theorem:

• the series converges absolutely on (a − R, a + R);
• the series diverges on (−∞, a − R) and (a + R, ∞).
The interval of convergence in this case always runs from a − R to a + R;
what you have to determine is whether the endpoints a − R and a + R should
be included in the interval. To figure this out, plug each of these endpoints
in for x (do them one-by-one) and check whether the numerical series you
obtain converge or diverge.
Example: For each power series, find the radius of convergence and the set of
x for which the series converges.
∞
X (−1)n (x + 1)n
(a)
n=0 (2n)!
256
∞
3n (x − 2)n
X
(b)
n=0
∞
n n xn
X
(c)
n=2
257

In problems 1-21, find the Taylor series representation of the given function.
3
1. f (x) = 1−x
13. f (x) = ln(1 − x2 )
1 14. f (x) = x sin 3x
2. f (x) = 1−x2
x
3. f (x) = ln(1 + x) 15. f (x) = ex2
1
4. f (x) = (1−x)3 16. f (x) = x sin x2 − x3
1
Hint: Differentiate 1−x
twice. 17. f (x) = ex + e−x
5. f (x) = 1 Hint: Find the Taylor series of ex
1−x3
and e−x independently, and then add
2
6. f (x) = 2+5x
them term-by-term.
7. f (x) = −3
−2−x
18. f (x) = (x2 + 2) cos x
Hint: Find the Taylor series of x2 cos x
8. f (x) = e−x
and 2 cos x independently, and then
9. f (x) = cos 4x add them term-by-term.
10. f (x) = ln(1 − 2x) 19. f (x) = x arctan x2

1 2
11. f (x) = 1+2x
20. f (x) = x2 +1
1
12. f (x) = cos x2 21. f (x) = (1−x)2
For problems 22-32, evaluate the following limits without using L’Hopital’s Rule:
sin x arctan 6x2 −6x2
22. lim 28. lim x6
x→0 x x→0
cos x−1
23. lim x sin x8 −x8
x→0 29. lim x20
x→0
sin x3 −x3
24. lim x9
x→0 sin x8 −x8
30. lim x24
arctan x−x x→0
25. lim x3
x→0
sin x8 −x8
ex −1−x 31. lim x30
26. lim 2x2
x→0
x→0
2ex −2−2x−x2 arctan x9 −ln(x9 +1)

27. lim x3
32. lim x18
x→0 x→0
258
In problems 33-37, approximate each of the following numbers using the second
Taylor polynomial of an appropriately chosen function:
33. ln .8 35. e3/5
34. sin .3 36. arctan 61

√ √
37. 2 Hint: Here, the appropriate function is f (x) = x + 1. You will have to
figure out the second Taylor polynomial of f (x) by computing derivatives of
f at zero and using the definition of Taylor series.
In problems 38-42, approximate each of the following numbers using the fourth
Taylor polynomial of an appropriately chosen function:
√
38. sin 14 40. e
39. cos .2 41. arctan 12

√ √
42. 2 Hint: As in problem 37, the appropriate function is f (x) = x + 1.
R 1/2
43. Approximate 0 cos(4x2 ) dx by replacing the integrand with its fourth Tay-
lor polynomial.
R 1/2
44. Approximate 0 arctan x2 dx by replacing the integrand with its fourth Tay-
lor polynomial.
R1 3
45. Approximate −1 e−x dx by replacing the integrand with its fourth Taylor
polynomial.
46. Approximate 01 ln(2x2 + 1) dx by replacing the integrand with its fourth Tay-
R
lor polynomial.
R1 2
47. Approximate x sin(x2 ) dx
0 by replacing the integrand with its sixth Taylor
polynomial.
48. Approximate 01 x8 sin x dx by replacing the integrand with its twelfth Taylor
R
polynomial. Describe the integration technique that one would use to find
the exact value of this integral. (Isn’t using a Taylor polynomial better?)
In problems 49-55, find each higher-order derivative:
49. Find f (6) (0) if f (x) = sin x2 .
50. Find f (36) (0) if f (x) = cos x2 .
51. Find f (100) (0) if f (x) = 4 ln(2x2 + 1).
259
3
52. Find f (40) (0) if f (x) = e2x .
3
53. Find f (42) (0) if f (x) = e2x .
∞
xn
54. Find f (30) (0) if f (x) = .
P
(4n)!
n=0
∞
x4n+2
55. Find f (30) (0) if f (x) = .
P
(4n)!
n=0
56. Suppose the third-order Taylor polynomial (centered at 0) of some unknown

2
function f is given by P3 (x) = 2 − x − x3 + 2x3 . Find f (0), f 0 (0), f 00 (0) and
f 000 (0).
∞ √
57. Determine whether the series e1/ n − 1 converges or diverges.
P
n=1
∞
58. Determine whether the series cos( n1 ) − 1 converges or diverges.
P
n=1
∞
1
59. Determine whether the series ln 1 − converges or diverges.
P
n
n=1
In problems 60-69, find the sum of each of the following series (you may assume
without proof that the series converge):
60. 1 − 13 + 51 − 71 + 19 − 1
11
+ ...
2 2 2 2
61. 2 + 2 + 2!
+ 3!
+ 4!
+ 5!
+ ...
π π2 π4 π6
62. 4
− 42 2!
+ 44 4!
− 46 6!
+ ...
9 27 81 35
63. 1 − 3 + 2!
− 3!
+ 4!
− 5!
+ ...
e2 e3 e4
64. 1 + e + 2
+ 3!
+ 4!
+ ...
2π 2 2π 4 2π 6
65. 2 − 2!
+ 4!
− 6!
+ ...
1 1 1 1
66. 1 − 2!
+ 4!
− 6!
+ 8!
− ...
100 10000 106 108
67. 2!
− 4!
+ 6!
− 8!
+ ...
π2 π4 π6 π8
68. 1 − 22 3!
+ 24 5!
− 26 7!
+ 28 9!
− ...
1 1 1 1
69. 2
− 22 ·2
+ 23 ·3
− 24 ·4
+ ...
260
Answers
∞
1. 3xn = 3 + 3x + 3x2 + 3x3 + 3x4 + ...
P
n=0
∞
2. x2n = 1 + x2 + x4 + x6 + ...
P
n=0
∞
(−1)n+1 n x2 x3 x4 x5
3. =x− − − ...
P
x n 2
+ 3 4
+ 5
n=1
∞
4. (n + 1)(n + 2)xn = 2 + 6x + 12x2 + 20x3 + 30x4 + ...
P
n=0
∞
5. x3n = 1 + x3 + x6 + x9 + ...
P
n=0
∞ n 2 3 4
−5
6. xn = 1 − 52 x + 5
x2 − 5
x3 5
x4 − ...
P
2 2 2
+ 2
n=0
∞ n 2 3 4
3 −1 3
7. xn − 32 · 21 x + 3 1
x2 − 3 1
x3 + 3 1
x4 − ...
P
2 2
= 2 2 2 2 2 2 2
n=0
∞
(−1)n n x2 x3 x4
8. =1−x+ − − ...
P
x n! 2 3!
+ 4!
n=0
∞
(−1)n 42n 2n 42 2 44 4 46 6
9. =1− −
P
x
(2n)! 2!
x + 4!
x 6!
x + ...
n=0
∞
−2n n
10. = −2x − 42 x2 − 83 x3 − 16 4
− 32 5
− ...
P
x n 4
x 5
x
n=1
∞
11. (−2)n xn = 1 − 2x + 4x2 − 8x3 + 16x4 − ...
P
n=0
∞
(−1)n 4n x4 x8 x12
12. =1− −
P
x
(2n)! 2!
+ 4! 6!
+ ...
n=0
∞
−1 2n
13. = −x2 − 21 x4 − 13 x6 − 14 x8 15 x10 − 16 x12 + ...
P
x n
n=1
∞
(−1)n 32n+1 2n+2 3 2 33 4 35 6 37 8 39 10
14. − − − ...
P
x
(2n+1)!
= 1!
x 3!
x + 5!
x 7!
x + 9!
x
n=0
∞
(−1)n 2n+1 x5 x7 x9
15. = x − x3 + − − ...
P
x n! 2! 3!
+ 4!
n=0
∞
(−1)n 4n+3
16. = − 3!1 x7 + 5!1 x11 − 7!1 x15 + 9!1 x19 − ...
P
(2n+1)!
x
n=1
∞
2
17. x2n = 2 + x2 + 4!2 x4 + 6!2 x6 + ...
P
(2n)!
n=0
261
∞
1 2 2 1 2
18. 2+ (−1)n+1 − x2n+2 = 2 + 1 − x2 − − x4 + ...
P
(2n)! (2n+2)! 2! 2! 4!
n=0
∞
(−1)n 4n+3 x7 x11 x15
19. = x3 − −
P
x
2n+1 3
+ 5 7
+ ...
n=0
∞
20. 2(−1)n x2n = 2 − 2x2 + 2x4 − 2x6 + 2x8 − ...
P
n=0
∞
21. (n + 1)xn = 1 + 2x + 3x2 + 4x3 + 5x4 + ...
P
n=0
14701 30!
22. 1 39. 15000
54. 120!
211
23. 0 40. 128
55. 870
−1 11
24. 6 41. 24 56. f (0) = 2; f 0 (0) = −1;
−1 179
f 00 (0) = −2
3
; f 000 (0) = 12.
25. 3 42. 128
1 57. diverges
26. 4 43. 9
20
1 58. converges
27. 3 44. 1
24
59. diverges
28. −72 45. 2
π
29. 0 4
60. 4
46. 15
30. −1 1
61. 2e
6 47. 5 √
2
31. −∞ 2557 62. 2
48. 14040
;
32. 1
use parts 8 times to 63. e−3
2
33. −11 evaluate this integral. 64. ee

50
34. .3 49. −120 65. −2

−36!
35. 89 50. 18! 66. cos 1
50
1 −252 100!
36. 6
51. 50
67. cos 10 − 1
11 2
37. 8
52. 0 68. π
95 214 42!
38. 384
53. 14!
69. ln( 12 )
262
Index
PN (x), 246 center of mass, idea of, 100
nth Term Test, 210 center of mass, one-dimensional con-
p-Series Test, 208 tinuous, 104
p-series (definition), 208 center of mass, one-dimensional dis-
“big six” Taylor series, 243 crete, 103
center of mass, two-dimensional con-
Abel’s Formula, 255 tinuous, 107, 108
absolutely convergent, 224 center of mass, two-dimensional dis-
acceleration (parametric equations), 145 crete, 106
alternating series, 220 centroid, 110
Alternating Series Test, 220 Chain Rule, 12
antiderivative, 15 circle, 134
applications of integration, general prin- circle, parametric equations of, 134
ciple, 96 classification problem (series), 164, 168,
arc length, 97 194, 204
Arc length formula (Cartesian equa- classifying a series, methods for, 229
tion), 98 coefficients (of a power series), 252
Arc length formula (parametric equa- common ratio, 180
tions), 147 Comparison Test (series), 212
area, 80 Comparison Test for Improper Integrals,
associative property, 165 73
AST, 220 complicated u-substitutions, 35
bounded (region), 61 computation problem (series), 164, 168,
194, 204
C-H Theorem, 255 concavity, 14
calculus formulas for parametric equa- Concavity (of parametric equations),
tions, 152 151
Cartesian function, parametric equa- conditionally convergent, 224
tions of, 141 Constant Multiple Rule, 12
Cauchy-Hadamard Theorem, 255 continuous (function), 8
263
Index
continuous (random variable), 115 elementary u-substitution, 29

continuous at a, 8 ellipse, 135
converge ± converge = converge (se- ellipse, area of, 120
ries), 172 ellipse, parametric equations of, 136
converge ± converge = converge (in- elliptic integrals, 99
tegrals), 72 equilibrium (physical system), 102
converge ± diverge = diverge (series), expected value (finite-valued r.v.), 114
173 expected value, continuous random vari-
converge ± diverge = diverge (inte- able, 118
grals), 72
convergence/divergence, p-integrals, Finite Sum Formula for a Geometric
65 Series, 185
converges (improper integral), 63, 67, first derivative, 14
68 force, 96
converges (series), 168, 172 fulcrum, 101
converges absolutely, 224 Fundamental Theorem of Algebra, 48
converges conditionally, 224 Fundamental Theorem of Calculus, 15
cookie-cutter method (volume), 93 General principle of applications of in-
cross-section, 87 tegration, 96
decimals, repeating, 190 Generalized Triangle Inequality, 223
definite integral, 14 geometric series, 180
definition of the derivative, 11 Geometric Series Test, 183
definition of the integral, 14 geometric series, applications of, 189
degree (of a polynomial), 48 geometric series, characterization of,
density (one-dimensional), 103 181
density function (of a random variable), Geometric Series, Finite Sum Formula,
117 185
derivative, 11 guess (in partial fraction decomposi-
derivative, definition of, 11 tion), 53
derivatives, rules for computing, 12 harmonic series, 208
Difference Rule, 12 harmonic series, divergence of, 209
differentiation rules, 12 horizontal asymptote, 9
disc method, 90 horizontally unbounded (region), 63
displacement, 96 how to choose u and dv, 47
diverge ± diverge = unknown (series), how to classify an infinite series, 229
173 how to guess (partial fractions), 53
diverge ± diverge = unknown (inte- hyperbola, 137
grals), 72 hyperbola (left-right), parametric equa-
diverges (improper integral), 63, 67, tions of, 137
68 hyperbola (up-down), parametric equa-
diverges (series), 168, 172 tions of, 138
264
Index
improper integral, horizontally unbounded moment about the y-axis, continuous,

region, 63 107, 108
improper integral, vertically unbounded moment about the y-axis, discrete, 106
region, 66 moment about the origin, continuous,
improper integrals, linearity rules, 72 104
indefinite integral, 15 moment about the origin, discrete, 103
Integral Test, 206
integral, definite, 14 nth derivative, 14
integral, definition of, 14 Nth partial sum, 167
integral, indefinite, 15 Nth Taylor polynomial, 246
integral, linearity of, 17, 25 negative series, 205
integrand, rewriting the, 28 p-integral, 65
integration by parts formula, 45 p-integral Test, 65
integration rules to memorize, 17, 27 parametric equations, 131
interval of convergence (of power se- parametric equations of common func-
ries), computing, 256 tions, 142
investments, 189 parametric equations, list of calculus
irreducible (polynomial), 48 formulas, 152
irrelevance of starting index, 71 parametric equations, non-uniqueness,
Koch snowflake, 192 153
parametric equations, shifts, 153
L’Hôpital’s Rule, 9 parametric equations, transformations,
length (of a curve y = f (x)), 97 153
length (of a vector), 144 partial fraction decomposition, 49
limit, 7 partial fractions, useful integral formu-
limit definition of derivative, 11 las for, 55
limits at infinity, 9 partial sum, 167
limits at infinity for rational functions, parts vs. u-substitutions, 46
11 parts, integration by, 45
line, parametric equations of, 139 polynomial, 48
linear approximation, 14 position (parametric equations), 144
Linear Replacement Principle, 33 positive series, 205
linearity of improper integral, 72 Power Rule, 12
linearity of infinite series, 172 power series (centered at a), 252
Linearity of Integration, 17, 25 power series (in x), 252
Product Rule, 12
Maclaurin series, 237
median (of a random variable), 126 Quotient Rule, 12
moment about the x-axis, continuous,
108 radius of convergence, 255
moment about the x-axis, discrete, 106 random variable, 112
Ratio Test, 195
265
Index
rearrangement problem (series), 164, transformations on parametric equa-

168, 194, 204 tions, 153
Rearrangement Theorem, 227 Triangle Inequality, 222
repeating decimals, 190 Triangle Inequality for Infinite Series,
rewriting the integrand, 28 223
Riemann integral, 14 Triangle Inequality, Generalized, 223
two-dimensional vector, 143
scalar, 145
second derivative, 14 u and dv, how to choose, 47
segment, 139 u-substitution, elementary, 29
segment, parametric equations of, 140 u-substitutions, more complicated, 35
series, alternating, 220 uniqueness of power series, 237
series, geometric, 180
series, negative, 205 vector, 143
series, positive, 205 velocity (parametric equations), 145
shell method (volume), 93 vertically unbounded (region), 66
signed areas, 80 volume, 87
Slope of tangent line, parametric equa- volume formula, generic, 89
tions, 149 washer method (volume), 91
speed (parametric equations), 145 work, 96
starting index is irrelevant (integrals),
71 zeroth derivative, 14
starting index is irrelevant (series), 173
sum (of an infinite series), 168, 172
Sum Rule, 12
tangent line, 14
Taylor polynomial of order N , 246
Taylor series, definition, 237
Taylor series, list of, 243
terms (of a series), 252
tone, 14
total mass, one-dimensional continu-
ous, 104
total mass, one-dimensional discrete,
103
total mass, two-dimensional continu-
ous, 107
total mass, two-dimensional continu-
ous, 108
total mass, two-dimensional discrete,
106
266

Lecture Notes 230

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lecture Notes 230

Uploaded by

Copyright:

Available Formats

Calculus II Lecture Notes

2 Basic integration techniques 24

3 Intermediate integration techniques 44

5 Applications of the integral 80

6 Parametric equations 130

7 Introduction to infinite series 162

8 Geometric series and the Ratio Test 180

9 Convergence tests for positive series 204

10 Absolute and conditional convergence 218

10.4 Rearrangement of infinite series . . . . . . . . . . . . . . . . . . . . . . 226

11 Taylor series 236

Big picture issues

3. What is the difference between a math problem that is a calculus problem

Definition 1.1 A function f : R → R is continuous at a if lim f (x) = f (a). f is

Example: Evaluate these limits:

1. lim (3 sin 2x)

Example: What about these:

• When you get 00 :

Techniques for dealing with 00 :

lim f (x) =lim

Although ∞ is not a number, it can be manipulated in some ways as if it is a

Theorem 1.4 (Limits at infinity for rational functions) Suppose f is a rational

am xm + am−1 xm−1 + am−2 xm−2 + ... + a2 x2 + a1 x + a0

1. If m < n (i.e. largest power in numerator < largest power in denominator),

2. If m > n (i.e. largest power in numerator > largest power in denominator),

3. If m = n (i.e. largest powers in numerator and denominator are equal), then

Assuming it exists, the derivative f 0 (x) computes:

We do not compute derivatives using the limit definition. We use differentiation

Functions whose derivatives we memorize:

Differentiation rules for more complicated functions:

• Constant Multiple Rule: (cf )0 (x) = c f 0 (x)

• Sum Rule: (f + g)0 (x) = f 0 (x) + g 0 (x)

• Difference Rule: (f − g)0 (x) = f 0 (x) − g 0 (x)

Example: Suppose f (x) = 3x6 sin x. Find f 0 (x).

Solution: The slope of the tangent line at π

f 0 (x) = 3 · − sin x + 2 cos x

Definition 1.6 Given a differentiable function f and a number a at which f is differ-

y = f (a) + f 0 (a)(x − a) (a.k.a. L(x) = f (a) + f 0 (a)(x − a)).

Definition 1.7 Let f : R → R be a function.

1.3 The definite integral

where the expression inside the limit is a Riemann sum for f .

The definite integral of a function is a number which is supposed to give the

Definition 1.9 Given function f , an antiderivative of f is a function F such that

Ex: F (x) = sin x is an antiderivative of f (x) = cos x.

Definition 1.10 Given function f , the indefinite integral of f , denoted

is the set of all antiderivatives of f .

Theorem 1.11 (Fundamental Theorem of Calculus Part II) Let f be continuous

The first object is a set of functions; the second object is a number.

Theorem 1.12 (Integration Rules to Memorize)

Theorem 1.13 (Linearity of Integration) Suppose f and g are integrable functions.

(and the same rules hold for indefinite integrals as well).

NOTE: Integration is not multiplicative nor divisive:

1.4 Homework exercises

25. Find the derivative of f (x) = 5x + 7.

50. Let y = 2 sin 3x. Find dy.

In problems 51-60, evaluate the given integral.

13. 1 38. 8x3 sin x + 2x4 cos x

14. ∞ 39. −36x2 cos(3x2 ) − 6 sin(3x2 )

59. 6 tan x + C 67. 6

60. −6 arctan x + C 68. 110

Basic integration techniques

Recall: we know from Calculus I that to evaluate an integral like

we use the Fundamental Theorem of Calculus, which says that