Chapter 11

Applications of
Differentiation, II

Reference: “Calculus”, by James Stewart.

In this chapter, we continue investigating

Curve Sketching

Recall that in our discussion of straight lines they could have

• positive slope, or

• negative slope, or

• zero slope.

m>0 m=0 m<0

positive slope zero slope negative slope

1
2 CHAPTER 11. APPLICATIONS OF DIFFERENTIATION, II

In this section we will be using the derivative to discuss the slopes

of curves in general.
Recall that the slope of a function at a point is indicated by it
the derivative of the function at that point, so if f is a differentiable
function:

• If f 0 (x) > 0 for all x on an interval then we say that f is increas-

ing on that interval.

• If f 0 (x) < 0 for all x in an interval then we say that f is decreas-

ing on that interval.

For example, suppose a function f has the following graph:

y
D
A

−2 B
4 x

This function is:

• increasing on the sections CA and BD, and

• decreasing on the section AB.

At the points A and B, the function is neither increasing nor de-

creasing, and this is the case we will be mostly interested in. This is
such a useful property for us that we give it a name:

Definition. Let f be a function. A point a ∈ dom f is called a stationary point

if f 0 (a) = 0.
 3

Extreme points

The main reason that we are interested in stationary points in this

chapter, is that they can be used to find the extreme points in a func-
tion.

Definition. Let f be a function, and let c ∈ dom f .

• f has local maximum at c if there is an open interval (a, b) con- The plural of maximum is maxima.

taining c such that, for every x ∈ (a, b), f (x) ≤ f (c) if f (x)
exists.

• Similarly, f has a local minimum at c if there is an open interval The plural of minimum is minima.

(a, b) containing c such that, for every x ∈ (a, b), f (x) ≥ f (c) if
f (x) exists.

• c is called a (local) extremum (or extreme point) if there is ei- The plural of extremum is extrema.

ther a local maximum or local minimum at c.

Intuitively, these definitions mean that something is a local max-

imum if it is the biggest value nearby, and something is a local mini-
mum if it is the smallest value nearby. This is the reason for the term
local; what happens in the function overall is not a concern, only
what happens around the point c. Don’t be concerned about ’how
nearby is nearby?’ As long as there is any interval, not matter how
large or small, where the function takes either the smallest or largest
value, there is a local extreme point.

Example 1. Consider the following graph of y = f (x), where f : [−1, 4] → R.

4 CHAPTER 11. APPLICATIONS OF DIFFERENTIATION, II

local maximum
y
local maximum

−1 4 x

local minimum local minimum

f has local extreme points at the positions marked. However, f (−1)

is also the smallest value nearby to x = −1 so there is a local mini-
mum at x = −1. Similarly, f (4) is the largest value of f (x) nearby to
x = 4, so there is also a local maximum at x = 4.

The functions that we encounter in Mathematics 1 will always

have a local extremum at the end points of their domain, if those
end points are included in the domain. If the end points of the do-
main are not included, then it is not necessary for a function to have
extrema at all:

Example 2. Let f : (0, 1) → R be given by f (x) = x, and let g : (−1, 1) → R

be given by g(x) = x2 . Then g has a local minimum but no maximum
and f has no extreme points at all.

Proof. (Not Examinable:) To show that f has no extreme points, we

will use a technique called ‘proof by contradiction’. This technique
involves assuming the opposite of what we want to prove, and show-
ing that it cannot work.
So, suppose that f has a local minimum at c. By definition, there
is an interval (a, b) containing c so that f (c) ≤ f (x) for every x in
(a, b). We want to find a point d > 0 (so that d ∈ dom f ) which is
 5

c+a
closer to a than c. 2
would work, but it’s possible that a < 0 in
which case d could be negative. So let

c

if a ≤ 0
2
d=
 c+a if 0 < a

2

d is an element of (a, b) ∩ (0, 1), and

f (d) = d < c = f (c)

so there cannot be a local minimum at c.

This argument doesn’t work if the domain of f is closed, because
d and c could be the same value (in which case d = c = 0).
A similar argument works to show that f has no maximum, and
let g has no maximum. g has a local minimum at x = 0, though,
because 02 = 0 and

f (x) = x2 ≥ 02 = f (0)

for every x.

The question remains, however, how do we find extreme points

which do note occur at the end points of domains. For this, we need
to introduce a new definition:

Definition. Let f be function and let c ∈ dom f . Then we call c a

critical point if either

• f 0 (c) = 0, or

• f 0 (c) does not exist.

So all stationary points are critical points, and will a shape similar
to these when graphed (for example):
6 CHAPTER 11. APPLICATIONS OF DIFFERENTIATION, II

However, critical points also include points for which the deriva-
tive does not exist, such as in these examples:

The other thing to notice is that according to our definition of dif-

ferentiability, the derivative is never defined at the end point of a
domain.

Example 3. Find the x–values of all critical points of the function

given by
f (x) = x2 − 3x + 2.

Solution: Differentiating the given function gives f 0 (x) = 2x − 3.

(i) If f 0 (x) = 0 then

2x − 3 = 0
3
i.e. x =
2
3
Thus there is a critical point at x = .
2
(ii) We also need to consider whether there are any x–values for
which f 0 (x) does not exist. However, we know that the implied
domain for f 0 (x) = 2x − 3 is R, so there are no such values.
3
Thus the only critical point occurs when x = .
2
√
Example 4. Find the x–values of all critical points of f (x) = 3
x.
 7

1
Solution: Since f (x) = x 3 then

1 2 1
f 0 (x) = x− 3 = 2 .
3 3x 3

(i) If f 0 (x) = 0 then

1
2 = 0.
3x 3
⇒ 1 = 0,

Which is impossible. Thus there are no points where f 0 (x) = 0 .

(ii) Since f 0 (x) = 1

2 , f 0 (x) is not defined when x = 0 (since we can-
3x 3
not have 0 in the denominator of a fraction). Since 0 ∈ dom f ,
there is a critical point at x = 0.

So in this example the only critical point occurs when x = 0.

The link between critical points and extrema is established by

Fermat’s Theorem, which we state without proof.

Theorem 1 (Fermat’s Theorem). If f : [a, b] → R has a local extreme

point at c ∈ (a, b) then f has a critical point at x = c.

Thus, to find the local maxima and local minima of a function we

should look at the critical points of that function, including the end
points if there are any.

Example 5. Consider the function given by f (x) = x2 − 3x + 2.

 8 CHAPTER 11. APPLICATIONS OF DIFFERENTIATION, II

y y = x2 − 3x + 2
From the graph, we see
that there is a local mini-
3
mum at x = 2
. This occurs 2
at the critical point which
we found in Example 3.
1 2 x
( 32 , − 41 )

First Derivative Test

We know from Fermat’s theorem finding the critical points of a func-

tion tells us all the places that an extreme point could occur, but for
An interior critical point is one that interior critical points we need a method to establish, once we know
is not at an end point
a critical point, what type of extreme point it is (or if it’s something
else). The ‘type’ of a critical point is often referred to as the nature. If
the function is continuous, we can determine the nature of a critical
point by examining the first derivative, f 0 (x), using what we call the
first derivative test.
The first derivative test involves looking at the sign of the first
derivative on either side of a critical point, in exactly the same way
that we used sign diagrams for inequalities in chapter 2. There are
three possible outcomes:

Local Maximum

Suppose that a continuous function f has a critical point at x = c.

(That is, suppose that f 0 (c) = 0 or that f 0 (c) does not exist.)
If
 9

• f 0 (x) > 0 for all x immediately to the left of c, and

• f 0 (x) < 0 for all x immediately to the right of c

then f has a local maximum at x = c.

f 0 (x)
+ −
positive slope negative slope
c

f (x)

Local Minimum

Suppose that a continuous function f has a critical point at x = c.

If

• f 0 (x) < 0 for all x immediately to the left of c, and

• f 0 (x) > 0 for all x immediately to the right of c

then f has a local minimum at x = c.

f 0 (x)
− +
negative slope positive slope
c

f (x)

Stationary point of inflection

Suppose that a continuous function f has a stationary point at x = c.

(That is, suppose that f 0 (c) = 0.) If

• f 0 (x) > 0 for all x immediately to the left of c, and

 10 CHAPTER 11. APPLICATIONS OF DIFFERENTIATION, II

• f 0 (x) > 0 for all x immediately to the right of c

or if

• f 0 (x) < 0 for all x immediately to the left of c, and if

• f 0 (x) < 0 for all x immediately to the right of c

(i.e. neither the local maximum or minumum case occurs) then f has
A point of inflection is point a stationary point of inflection at x = c.
where the second derivative, f 00 (x),
changes from positive to negative
or negative to positive. Graphically, f 0 (x)
this means that the function changes
the direction it ’curls’. + 0 +
positive slope positive slope
c

f (x)

f 0 (x)
− 0 −
negative slope negative slope
c

f (x)

Example 6. Sketch the graph of y = x3 − x2 − x + 1 by finding

• the x and y–intercepts;

• the coordinates of all the critical points; and

• the nature of all the critical points.

Solution:
 11

x–intercepts:

Factorising gives y = (x − 1)2 (x + 1). So if y = 0 then x = 1, −1.

y–intercept:

If x = 0 then y = 03 − 02 − 0 + 1 = 1.

critical points:
dy dy
dx
= 3x2 − 2x − 1 = (3x + 1)(x − 1). dx
is defined everywhere
so there are no non-stationary critical points.
dy
dx
= 0 if x = − 31 , or 1.
If x = − 13 then y = − 27
1
− 19 + 13 + 1 = 32
27
.
If x = 1 then y = 1 − 1 − 1 + 1 = 0.

 
1 32
So the critical points are − , and (1, 0).
3 27

nature of the critical points:

By substituting into the first derivative we see that: Once again, this is exactly the same
process as when we used a sign dia-
gram.

• If x < − 13 , then f 0 (x) > 0.

• If x ∈ (− 13 , 1), then f 0 (x) < 0.

• If x > 1, then f 0 (x) > 0.

Diagrammatically,

f 0 (x)
+ 0 − 0 +

− 13 1

f (x)
 12 CHAPTER 11. APPLICATIONS OF DIFFERENTIATION, II

So by the first derivative test, there is a local maximum at (− 13 , 32

27
),
and a local minimum at (1, 0).

We now use this to sketch a graph of y = f (x).

y y = x3 − x2 − x + 1

(− 13 , 32
27 )

−1 1 x

In the next example, we see how to sketch a graph of y = f 0 (x),

when we are given a graph of y = f (x). The steps to be used in such
examples are as follows:

Step 1: Find the points where the slope of the given graph is zero.

At these points we have f 0 (x) = 0, and so we draw y = 0.

Step 2: Find any points where the given graph is not differentiable.

(For example, find any points for which the given graph is not
continuous, or where there is a sharp point.) At these points
f 0 (x) does not exist, and so the graph y = f 0 (x) has a hole in it,
and possibly a sudden jump.

Step 3: Consider each “section” of the given graph (between the points
found in Steps 1 and 2).
 13

– If the slope of the given graph is positive (i.e. if f 0 (x) > 0),
then draw positive y–values in that section.

– If the slope of the given graph is negative (i.e. if f 0 (x) < 0),
then draw negative y–values in that section.

Note: We will never be able to produce an exact graph; we shall

not be concerned with the exact shape of our final graph, as long as
the above three steps have been followed.

Example 7. The graph of y = f (x) is drawn below. Sketch y = f 0 (x).

y

y = f (x)

Solution: We first note that f 0 (x) = 0 whenever the tangent to y =

f (x) is horizontal. We mark the x–coordinates of these points on the
graph as a, b and c.
y

a b c x

y = f (x)

Note that :

• f 0 (x) = 0 ⇐⇒ x = a, b, c.
14 CHAPTER 11. APPLICATIONS OF DIFFERENTIATION, II

• f is increasing when x < a. Therefore f 0 (x) > 0 when x < a.

• f is decreasing on the interval (a, b). Therefore f 0 (x) < 0 when

x ∈ (a, b).

• f is inreasing on the interval (b, c). Therefore f 0 (x) > 0 when

x ∈ (b, c).

• f is decreasing when x > c. Therefore f 0 (x) < 0 when x > c.

Using this information, we sketch a graph of y = f 0 (x).

y

a b c x

y = f (x)

y = f 0 (x)

Global Maxima and Minima

Definition. Let f be a function and let c ∈ dom f

• f has a global maximum at c if f (c) ≥ f (x) for every x in the

function’s domain.

• f has a global minimum at c if f (c) ≤ f (x) for every x in the

function’s domain.
 15

• a global extremum at c (or global extreme point) at c if it has a

global maximum or global minimum at c.

The difference between a global extremum and a local extremum

point is that a local extremum is only concerned with the nearby
values of the function, while a global extremum is the biggest (or
smallest) value for the whole function. Every global extremum is
also a local extremum.
y
global (and local) maximum

local maximum

local minimum
x

global (and local) minimum

Just as there exist functions without local extrema, not every func-
tion necessarily has a global extremum. However the following the-
orem does hold (which we state without proof).

Theorem 2 (Extreme value theorem). Let f : [a, b] → R be a continuous

function. Then f achieves both a global minimum and a global maximum.

Finding global extreme points for continuous functions is usu-

ally straightforward. Since every global extremum is also a local ex-
tremum, we know from Fermat’s theorem that they must occur at
critical points (including the endpoints of the domain if included).
We don’t need to worry about whether the critical points are local
maxima or minima, we can just compare the values of the function
at the critical points to see which are the biggest/smallest.
 16 CHAPTER 11. APPLICATIONS OF DIFFERENTIATION, II

Example 8. Find the global extreme values of the function

f : [0, 2] → R where f (x) = x3 − 3x + 2.

Solution: In this example, the endpoints of the domain are x = 0 and

x = 2.
To find the interior critical points, we must first find f 0 (x):

f (x) = x3 − 3x + 2

⇒ f 0 (x) = 3x2 − 3

The derivative is defined everywhere, so the only critical points are

the stationary points. Set

f 0 (x) = 0 = 3(x2 − 1)

= 3(x + 1)(x − 1)

⇒ x = 1 or − 1

However, −1 is not in dom f so we shall ignore this value. Thus,

the only interior critical point is at x = 1.
To find the global extreme values, we just need to evaluate the
function at 0, 2, and −1.

Critical point: f (1) = 13 − 3 × 1 + 2 = 0.

Endpoints: f (0) = 03 − 3 × 0 + 2 = 2.
f (2) = 23 − 3 × 2 + 2 = 4.

So

Don’t get confused between the f has a global maximum value of 4, which occurs when x = 2,
maximum value of the function, and
the value of x where that maximum
occurs and

f has a global minimum value of 0, which occurs when x = 1.

 17

Optimization (Applied Max/Min Problems)

When using the tools described above to find extrema in applica-

tions, the most difficult part is usually in the set up of the question
and identification of the the problem. It is particularly important to
make sure that you identify the variables involved in the question.

Steps for Solving Applied Max/Min Problems:

1. Identify and name variables. For many questions it is helpful to

draw a diagram labelled by symbols representing the variables
relevant to the problem.

2. Write down a formula for the quantity to be maximized or minimized.

3. Write the formula found in Step 2 in terms of a single variable This

usually involves making use of

• geometric properties of the diagram, or

• information given in the question.

4. Find all critical points. That is, if f is the function to be maxi-

mized or minimized then find all x–values in the domain of
f such that f 0 (x) = 0 or f 0 (x) does not exist. Do not forget the
end points.

Step 5: Establish the nature of these critical points.

Determine the relevant extrema. Depending on the question,
that could involve using the first derivtaive test to identify local
extrema as well.
18 CHAPTER 11. APPLICATIONS OF DIFFERENTIATION, II

Step 6: Answer the question.

Note: Your answer should always be given in the context of
the question. An application question will usually need to be
answered by a sentence.

Examples

Example 9. Find two negative numbers whose product is 100 and

whose sum is a maximum.

Solution: A diagram is not very relevant in this example, but let x

and y be two negative numbers as in the question, hence we know

xy = 100 (11.1)

We want to maximize the sum of x and y, which we will call S.i.e.

S =x+y (11.2)

From equation 11.1

xy = 100 (11.3)
100
⇐⇒ y = (11.4)
x

which can be substituted into Equation 11.2 to give

100
S = x+
x
= x + 100x−1

We wish to maximise this function on the domain x ∈ (−∞, 0).

Notice that this domain does not include any endpoints. To find the
critical points, we differentiate S:

dS 100
=1− 2 ,
dx x
 19

This derivative exists for all x in the domain of S, thus any critical
dS
points must occur when = 0. So set
dx
dS 100
=1− 2 =0
dx x
x2 − 100
⇒ =0
x2
⇒ x2 − 100 = 0

∴ x = ±10

Since x was defined to be a negative number, the only critical

point occurs at -10.
Since this is a continuous function with only one critical point, if
that critical point will be a global maximum if and only if it is a local
maximum.

By the First Derivative

dS
Test, we can deduce that dx
+ 0 −
we have a local maxi- −10 0
mum (and also a global
S
maximum) at x = −10.

Substituting x = −10 into Equation 11.4, we see that y also equals

−10.
Thus the two required numbers are −10 and −10.

Example 10. A rectangular poster (printed on one side only) is to

contain 343 cm2 of printed matter on it, with a 4 cm margin along
the top, a 3 cm margin along the bottom, and a 2 cm margin along
the sides. Find the dimensions of the poster with the smallest area.

Solution: Let x and y be the width and height, respectively, of the

poster
20 CHAPTER 11. APPLICATIONS OF DIFFERENTIATION, II

4 cm

printed matter y
2 cm 2 cm

3 cm
x

We want to minimize

A = xy. (11.5)

From the diagram, we see that the printed region of the poster
has dimensions y − 7 cm and x − 4 cm. Thus we need x > 4, y > 7
and

(x − 4)(y − 7)
= 343 (11.6)
343
⇒y−7 = .
x−4
343
⇒y = + 7
x−4
7x + 315
= . (11.7)
x−4

Substituting Equation 11.7 into Equation 11.5 gives

x(7x + 315)
A =
x−4
2
7x + 315x
=
x−4

where x > 4.
 21

Step 4: Then, by the quotient rule for differentiation, we have

dA (14x + 315)(x − 4) − 1(7x2 + 315x)

=
dx (x − 4)2
14x2 + 259x − 1260 − 7x2 − 315x
=
(x − 4)2
7x2 − 56x − 1260
=
(x − 4)2
7(x − 18)(x + 10)
= .
(x − 4)2

This derivative exists for all x in the domain of the function A;

dA
thus any critical points must occur when = 0.
dx
dA
Solving = 0 gives x = 18 or x = −10. Since we know x > 4 we
dx
will only investigate the nature of the critical point at x = 18.
By the First Derivative
Test, we can deduce that dA
dx
we have a local mini- − 0 +

mum at x = 18. This local 4 18

minimum is also the global A

minimum.

Substituting x = 18 into Equation 11.6, we see that

7 × 18 + 315
y =
18 − 4
441
=
14
= 31.5

Thus the poster with smallest area should have length 18 cm

across the top, and sidelength 31.5 cm.