International Baccalaureate

MATHEMATICS
Analysis and Approaches (SL and HL)
Lecture Notes

Christos Nikolaidis

TOPIC 5
CALCULUS

5.1 THE LIMIT limf(x) – THE DERIVATIVE f (x) : A ROUGH IDEA! ….…….. 1

5.2 DERIVATIVES OF KNOWN FUNCTIONS – RULES ………………………………. 9

5.3 TANGENT LINE – NORMAL LINE AT SOME POINT x0 ………………………. 16

5.4 THE CHAIN RULE ………………………………..………………………………………………… 21

5.5 MONOTONY - MAX, MIN ……………………………………………………………………... 29

5.6 CONCAVITY - POINTS OF INFLECTION …………………….……………………….. 34

5.7 OPTIMISATION ……………………………….……………………………….……………………. 39

5.8 THE INDEFINITE INTEGRAL  f(x)dx …………………………………………………… 44

5.9 INTEGRATION BY SUBSTITUTION …………………………………...………….………. 50

b
5.10 THE DEFINITE INTEGRAL  f(x)dx - AREA BETWEEN CURVES ……..
a
56

5.11 KINEMATICS (DISPLACEMENT, VELOCITY, ACCELARATION) …………… 66

Only for HL

5.12 CONTINUITY AND DIFFERENTIABILITY ………………………………………………… 75

5.13 L’HÔPITAL’s RULE …………………………………………………………………………………... 84

5.14 IMPLICIT DIFFERENTIATION – MORE KINEMATICS ………………………….. 89

5.15 RATE OF CHANGE PROBLEMS ……………………………………………………………. 94

5.16 FURTHER INTEGRATION BY SUBSTITUTION ……………………………………… 99

5.17 INTEGRATION BY PARTS .……………………………………………………………………. 106

5.18 FURTHER AREAS BETWEEN CURVES - VOLUMES ………………………….. 112

5.19 DIFFERENTIAL EQUATIONS ………………………………………………………………… 117

5.20 MACLAURIN SERIES – EXTENSION OF BINOMIAL THEOREM ….…….. 128

December 2020
TOPIC 5: CALCULUS Christos Nikolaidis

5.1 THE LIMIT limf(x) – THE DERIVATIVE f ΄(x): A ROUGH IDEA

This paragraph may look very “technical”. Do not pay much attention on
your first reading. You may skip and proceed to paragraph 5.2; you will
realize that the derivative in practice is much easier than it appears here!

 THE LIMIT limf(x)

Consider the function f(x) = 2x+3
It is clear that when x=2, then f(2) = 7.

But what is the behavior of f(x) when x approaches 2?

approach x=2 approach x=2

from values less than 2 from values greater than 2
x f(x) x f(x)
1.8 6.6 2.2 7.4
1.9 6.8 2.1 7.2
1.99 6.98 2.01 7.02
1.999 6.998 2.001 7.002

that is that is
if x 2= then f(x)  7 if x 2+ then f(x)  7

Thus in general
if x 2 [x tends to 2]
then f(x)  7 [f(x) tends to 7]

1
TOPIC 5: CALCULUS Christos Nikolaidis

In order to express this fact we write

lim f(x)  7
x 2

and say that the limit of f(x), as x tends to 2, is 7.

Remark
In fact for the right column we write lim- f(x)  7
x 2

while for the right column we write lim f(x)  7

x 2

and these are called side limits. If the side limits are equal then

lim f(x)  7
x 2

In this example
lim f(x)  7  f(2)
x 2

lim f(x)  9  f(3) and so on!

x 3

The situation lim f(x)  f(a) occurs very often, however, this is not
x a

always the case (otherwise the limit would be nothing more than a
simple substitution!).

Let’ s see a case where the limit is not a simple substitution!

We will find informally (by using our GDC) the limit

sinx
lim .
x 0 x

Notice that the function is not defined at x=0. The graph looks like
y

x
-5 5

2
TOPIC 5: CALCULUS Christos Nikolaidis

Let’s approach x=0 by using our GDC:

approach x=0 approach x=0

from values less than 0 from values greater than 0
x f(x) x f(x)
-0.1 0.998334 0.1 0.998334
-0.01 0.999983 0.01 0.999983
-0.001 0.999999 0.001 0.999999

that is that is
if x 0= then f(x)  1 if x 0+ then f(x)  1

We say that the limit when x tends to 0 is 1 and we write

sinx
lim 1
x 0 x

Sometimes the limit can be +∞ or -∞.

1
Let us investigate informally (by using our GDC) the limit lim .
x 0 x
1
f(x)=
x

3
TOPIC 5: CALCULUS Christos Nikolaidis

approach x=0 approach x=0

from values less than 0 from values greater than 0
1 1
x f(x)= x f(x)=
x x
-0.1 -10 0.1 10
-0.001 -1000 0.001 1000
-0.000001 -1000000 0.000001 1000000

Here we only have side-limits:

1 1
lim =+  and lim_ =- 
x 0 x x 0 x

Remember:

If lim f(x)   or lim f(x)   we say x=a is a vertical asymptote.

x a x a

Thus, in our example, x=0 is a vertical asymptote.

We also define limits of the form lim f(x) or lim f(x) . Thus, we
x   x  

study the behavior of the function when x approaches +  or -  .

1 1
Let us find informally (by using our GDC) the limits lim , lim .
x   x x   x

x approaches -  x approaches + 
x f(x) x f(x)
-1000 -0.001 1000 0.001
-1000000 -0.000001 1000000 0.000001

Both limits are 0, namely

1 1
lim =0 and lim = 0 .
x   x x   x

Remember:
If lim f(x)  a or lim f(x)  a then y=a is a horizontal asymptote.
x   x - 

Thus, in our example, y=0 is a horizontal asymptote.

4
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 1
y
5
4
3
2
x 3
f(x)  1
x2 x
-2 -1 1 2 3 4 5 6
-1
-2
-3

We know that
 x=2 is a vertical asymptote.
The formal explanation is that lim f(x)   and lim f(x)  
x 2 x 2

 y=1 is a horizontal asymptote.

The formal explanation is that lim f(x)  1 and lim f(x)  1
x   x - 

Look at an interesting limit that provides the irrational number

e=2.7182818…

EXAMPLE 2
x
 1
Investigate informally (by using your GDC) the limit lim  1   .
x   x

x approaches + 
x f(x)
1000 2.7169239…
1000000 2.7182804…
1010 2.7182818…

The resulting limit is in fact the number e=2.7182818… That is,

x
 1
lim  1   = e .
x   x

5
TOPIC 5: CALCULUS Christos Nikolaidis

 RATE OF CHANGE (OR GRADIENT) IN A STRAIGHT LINE

Consider the line f(x)=2x+3.

y
8
7
6
5
4
3
2
1
x
-4 -3 -2 -1 1 2 3 4
-1
-2

Notice that
when x changes from 1 to 2
then y changes from 5 to 7

Hence, the corresponding rate of change is

Δy f(2)  f(1) 7 5
= = = 2
Δx 2 1 2 1

We understand that the rate of change between any two points on

the line is always the same.
For example,
when x changes from 0 to 2
then y changes from 3 to 7

Hence, the corresponding rate of change is still

Δy f(2)  f(0) 7 3
= = = 2
Δx 20 20

This common value is the gradient of the line.

Next, we will see that the gradient is not only defined for straight
lines but also for other curves.

6
TOPIC 5: CALCULUS Christos Nikolaidis

 RATE OF CHANGE ΙΝ Α CURVE

In a curve which is not a straight line, the rate of change between

any two points is not always the same.

For example, in f(x)=x2

Δy f(2)  f(1) 4 1
Rate of change from x=1 to x=2: = = 3
Δx 2 1 2 1
Δy f(3)  f(1) 91
Rate of change from x=1 to x=3: = = 4
Δx 31 31

However, we can measure the “instantaneous” rate of change at

any particular point on the curve.

We will estimate the rate of change near x=1 as follows:

Let A(1,1). We select a neighboring point B with

x-coordinate = 1+h (where h is very small)
y-coordinate =(1+h)2

As we move from point A to point B, the rate of change is

Δy (1  h) 2 - f(1) 1  2h  h 2 - 1 2h  h 2
    2h
Δx (1  h)- 1 h h

If we let h become very small, that is h0, the result will be the
rate of change at point A. But this is a limit!

7
TOPIC 5: CALCULUS Christos Nikolaidis

Therefore, the rate of change (or gradient) of f(x)=x2 at x=1 is

lim (2  h) =2
h 0

Let us apply the same technique for any point A(x,x2) and a
neighboring point B with
x-coordinate = x+h (where h is very small)
y-coordinate =(x+h)2
Then

Δy (x  h) 2  x 2 x 2  2xh  h 2  x 2 2xh  h 2
 = = = 2x  h
Δx h h h

and the rate of change at any x is

lim (2x  h) = 2x
h 0

Therefore, for f(x)=x2

the gradient at any point x is equal to 2x
We write
f (x)  2x
Thus for example,

gradient at x=-1: f (-1) =-2

gradient at x=0: f (0) =0
gradient at x=1: f (1) =2
gradient at x=2: f (2) =4
gradient at x=3: f (3) =6

The new function f (x) , which is derived from f(x), is called the
derivative of f. Thus

f (x)  DERIVATIVE = RATE OF CHANGE = GRADIENT at x .

8
TOPIC 5: CALCULUS Christos Nikolaidis

5.2 DERIVATIVES OF KNOWN FUNCTIONS - RULES

The derivative of a function f(x) is a new function denoted by f (x) .

As explained in the preceding section, f (x) indicates the rate of
change, or otherwise the gradient of f(x) at any particular point x.
We have seen for example that

for the function f(x)=x2,

the derivative is f (x) =2x.

We present the derivatives of the most common functions:

f(x) f ΄(x)
xn nxn-1
sinx cosx
cosx -sinx
ex ex
1
tanx
cos 2 x
1
lnx
x
1
x 2 x
c (constant) 0

Let us especially elaborate on the first formula (xn)΄ = nxn-1

f(x)= xn f ΄(x) =nxn-1

x10 10x9
x4 4x3
x3 3x2
x2 2x
x 1
1 0

9
TOPIC 5: CALCULUS Christos Nikolaidis

This formula also applies for –tive values of n:

f(x)= xn f ΄(x) =nxn-1

x-10 -10x-11
x-3 -3x-4
x-2 -2x-3
x-1 -x-2

It also applies for rational values of n:

f(x)= xn f ΄(x) =nxn-1

x6.4 6.4x5.4
3 1/2
x3/2 x
2
5 2/3
X5/3 x
3
1 -1/2
X1/2 x
2

EXAMPLE 1

 1  2
Show that (a)  2   3 (b)  x   1
x  x 2 x
Solution
1 2
(a) 2  x  2 , so the derivative is -2x-3 = 3
x x
1 1 1
(b) x  x 1/2 , so the derivative is x-1/2 = 1/2
=
2 2x 2 x

EXAMPLE 2
Let f(x)=x7. Find
(a) f(0), f(1), f(2)
(b) f ΄(x)
(c) f ΄(0), f ΄(1), f ΄(2)
(d) the rate of change of f(x) at x=2
(e) the gradient of f(x) at x=2

10
TOPIC 5: CALCULUS Christos Nikolaidis

Solution
(a) f(0)=0, f(1)=1, f(2)=128
(b) f ΄(x) = 7x6
(c) f ΄(0)=0, f ΄(1)=7.16=7, f ΄(2)=7.26=448
(d) it is f ΄(2) = 448
(e) it is f ΄(2) = 448

 SYMBOLS
If y=f(x), the derivative is denoted by the following symbols

dy d
y΄ or f ΄(x) or or f(x)
dx dx

The derivative at some specific value of x, say x=2, is denoted by

dy
f ΄(2) or
dx x  2

For example, if y=f(x)=x3, we can write

dy d 3
y ΄=3x2 or (x3)΄ = 3x2 or =3x2 or x =3x2
dx dx
Moreover,
dy
f ΄(2)=12 or =12
dx x  2
The procedure of finding the derivative is called differentiation.

 RULES OF DIFFERENTIATION

Rule (1): (f+g)΄ = f ΄ + g΄ (f-g)΄ = f ΄ - g΄

EXAMPLE 3
For f(x) = x5+x3 , f ΄(x) = 5x4+3x2
For g(x) = x5-x3, g ΄(x) = 5x4-3x2
1
For h(x)= x7+ex-lnx+sinx-x+5, h΄(x) = 7x6+ex- +cosx-1
x

11
TOPIC 5: CALCULUS Christos Nikolaidis

Rule (2): (af)΄ = af ΄ (a=constant number)

EXAMPLE 4
For f(x) = 3sinx, f ΄(x) = 3cosx
For g(x) = 7ex, g΄(x) = 7ex
For h(x) = 5x3, h΄(x) = 5(3x2)=15x2
1 10
For k(x) = 10lnx, k΄(x) = 10 =
x x

Let us combine Rules (1) and (2):

(af + bg)΄ = af ΄+ bg΄

EXAMPLE 5
For f(x) = 2x3-3x2+7x+5, f ΄(x) = 6x2-6x+7
3
For g(x) = 5x7+3lnx-7cosx, g΄(x) = 35x6+ +7sinx
x

NOTICE:
The differentiation rules above may also be expressed as follows

d d d
[f(x) ±g(x)] = f(x) ± g(x)
dx dx dx

d d
[af(x)] =a f(x)
dx dx

d d d
[af(x)  bg(x)] = a f(x) + b g(x)
dx dx dx

EXAMPLE 6
d
For f(x)=x3 and g(x)=sinx, [4f(x)  5g(x)] = 12x2+5cosx
dx

12
TOPIC 5: CALCULUS Christos Nikolaidis

Rule (3): (f .g)΄ = f ΄.g + f .g΄ (product rule)

Be careful !!!
If f(x) = x5sinx then f ΄(x) is not (5x4)(cosx)

We must follow the product rule above.

EXAMPLE 7

For f(x) = x5sinx, f ΄(x)=(x5)΄ sinx + x5 (sinx)΄ = 5x4sinx + x5cosx

1
For g(x) = xlnx, g΄(x)= (x)΄lnx + x(lnx)΄ = 1lnx + x = lnx +1
x

For, h(x) = x2ex, h΄(x) = 2xex+x2ex


Rule (4): f  f   g- f  g (quotient rule)
  =
 g g2
 

EXAMPLE 8

x3 (x 3 )sinx  x 3 (sinx) 3x 2 sinx  x 3 cosx

For f(x) = , f ΄(x)= =
sinx (sinx) 2 sin 2 x

x 3  5x (3x 2 - 5)(x 2  1)  (x 3 - 5x)2x

For g(x) = , g ΄(x) =
x2 1 (x 2  1)2

Sometimes, we can avoid the quotient rule. Look at the following

EXAMPLE 9
x 3  2x  1
For f(x) =
x
method A: The quotient rule gives

(3x 2 - 2)x  (x 3 - 2x  1)1 2x 3  1 1

f ΄(x) = 2
= 2
=2x- 2
x x x

13
TOPIC 5: CALCULUS Christos Nikolaidis

method B: we may modify f(x) by splitting into three fractions

x 3 2x 1
f(x) = - + = x2-2+x-1 , so that
x x x
1
f ΄(x) = 2x-x-2 =2x-
x2

 HIGHER DERIVATIVES

The second derivative f ΄΄(x) is in fact the derivative of f ΄(x).

The third derivative f ΄΄΄(x) is the derivative of f ΄΄(x). And so on!

EXAMPLE 10

 For f(x) = x5, f ΄(x) = 5x4 f ΄΄(x) = 20x3 f ΄΄΄(x) = 60x2

 For g(x) = sinx g΄(x) = cosx g΄΄(x) = -sinx g΄΄΄(x) = -cosx

 For h(x) = ex, h΄(x) = ex h΄΄(x) = ex h΄΄΄(x) = ex

Alternative notation:
d2y d2
f ΄΄(x) can also be written as or f(x)
dx 2 dx 2
d3y d3
f ΄΄΄(x) can also be written as or f(x)
dx 3 dx 3

EXAMPLE 11
1 3 3 2
Let f(x)= x - x +2x+7. Solve the equations
3 2

(a) f ΄(x) = 0 (b) f ΄΄(x) = 0

Solution

(a) f ΄(x) = x2-3x+2, hence f ΄(x) = 0  x2-3x+2=0  x=1 or x=2

(b) f ΄΄(x) = 2x-3, hence f ΄΄(x) = 0  2x-3=0  x=3/2

14
TOPIC 5: CALCULUS Christos Nikolaidis

 SOME EXTRA FUNCTIONS (only for HL)

The derivatives of any exponential and any logarithmic function:

f(x) f ΄(x)
ax ax .lna
1
logax
xlna

1
Notice: for a=e we obtain the particular cases, (ex)΄=ex and (lnx)΄=
x

EXAMPLE 12

For f(x) = 3x, f ΄(x) = 3xln3

7
For g(x) = 7log5x, g ΄(x) =
xln5

The derivatives of the known reciprocal trigonometric functions:

f(x) f ΄(x)
secx secxtanx
cscx -cscxcotx
cotx -csc2x

The derivatives of the inverse trigonometric functions:

f(x) f ΄(x)
1
arcsinx
1- x 2
1
arccosx 
1- x 2
1
arctanx
1  x2

15
TOPIC 5: CALCULUS Christos Nikolaidis

5.3 TANGENT LINE - NORMAL LINE AT SOME POINT x0

Remember: A straight line with

 gradient m
 passing through point (x0,y0)
has equation
y- y0 = m(x- x0)

For example, the line passing through A(1,2) with gradient m=3
has equation
y- 2 = 3(x- 1)

Consider a function y=f(x) and some point x0. Then we also know
y0=f(x0)

y0

x0

We know that the gradient of the graph at (x0,y0) is mT=f ΄(x0).

We define:

TANGENT LINE at x0: normal line tangent line

the line with gradient mT
which passes through (x0,y0)
y0

NORMAL LINE at x0:

the perpendicular line to the
tangent at (x0,y0).
1
Its gradient is mN= 
mT
x0

The point (x0,y0) is also known as point of contact.

16
TOPIC 5: CALCULUS Christos Nikolaidis

 METHODOLOGY

Given y=f(x)
and some point x=x0

We find the point of contact (x0,y0) since y0=f(x0)

f ΄(x)
1
mT=f ΄(x0) and so mN= 
mT

The equations of the lines are

TANGENT LINE NORMAL LINE

y-y0=mT(x-x0) y-y0= mN(x-x0)

EXAMPLE 1
Consider the function
f(x)=x2
Find the tangent line and the normal line at x=3

Solution
The point of contact is (3,9) (since f(3)=9)

The gradient function is f ΄(x)=2x. Thus

1
mT = f ΄(3)=6 and mN= 
6

The tangent line is y-9=6(x-3)

1
The normal line is y-9=  (x-3)
6

After performing the necessary operations we obtain the final

forms
Tangent line: y-9=6(x-3)  y  9  6x  18  y=6x-9

1 1 1 1 19
Normal line: y-9=  (x-3)  y  9  - x   y=  x+
6 6 2 6 2

17
TOPIC 5: CALCULUS Christos Nikolaidis

NOTICE

An alternative way to obtain the tangent and the normal lines is to

use the formulas:

TANGENT LINE NORMAL LINE

y  mT x  c y= m N x+c

The point (x0,y0) helps us to find the constant c.

In the previous example, since m=6 and the point is (3,9):

 Tangent line: y=6x+c

At (3,9) 18+c=9  c=-9, thus y=6x-9

1
 Normal line: y=  x+c
6
1 1 19 1 19
At (3,9)  3+c=9  c=9+  c= thus y=  x+
6 2 2 6 2

EXAMPLE 2

Consider the function f(x)=5x3-2x+1

Find the tangent lines to the curve which are parallel to the line
L: y=13x+8
Solution

A tangent line parallel to L must have gradient mT=13.

But f ΄(x) = 15x2-2, so
15x2-2=13  15x2=15  x2=1  x=1 or x=-1

Hence, we will have two parallel lines, at the points x=1 and x=-1,
with gradient m=13. The points of contact are (1,4) and (-1,-2)

 At (1,4), y-4=13(x-1)  y-4=13x-13  y=13x-9

 At (-1,-2) y+2=13(x+1)  y+2=13x+13  y=13x+11

18
TOPIC 5: CALCULUS Christos Nikolaidis

NOTICE

Suppose that the gradient at (x0,y0) is mT=0. Then

The tangent line is a horizontal line with equation y=y0

The normal line is a vertical line1 with equation x=x0

EXAMPLE 3
Consider the function f(x)=sinx
π
Find the tangent line and the normal line at x=
2
y

x
-π/2 π/2 π 3π/2 2π

-1

Solution
It is f ΄(x)=cosx
π π π
At x= , y=sin =1, thus the point of contact is ( ,1)
2 2 2
π
mT = cos =0 (mN is not defined)
2
Tangent line: the horizontal line y=1

π
Normal line: the vertical line x= (look at the graph above!)
2

For trickier questions, let’s have in mind the following observation.

At the point of contact between f(x) and a tangent y=mx+c
functions are equal: f(x)=y
derivatives are equal: f ΄(x)=m

1 A vertical line has no gradient. If it passes through (x0,y0), it has equation x=x0.

19
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 4
The line y  mx - 3 is tangent to the curve f(x)=x4-x. Find m

Solution
To simplify the process, two relations hold at the point of contact:
functions are equal: x4-x  mx - 3
derivatives are equal: 4x3-1  m
Hence,
x4-x  (4x 3 - 1)x - 3  x 4 - x  4x 4 - x - 3

 3x 4  3
 x4  1
 x  1
If x  1 then m=3
If x  1 then m=-5

A variation of the same problem is presented below.

EXAMPLE 5
Consider the function f(x)=x4-x. Find the tangent lines passing
through the point (0,-3) [notice that the point is not on the curve]

Solution
A line passing through the point (0,-3) with gradient m
has equation y  3  m(x - 0) , that is y  mx - 3
Thus, we follow exactly the same process as in example 4 above.

If x  1 then m=3 and the tangent line is y  3x - 3

If x  1 then m=-5 and the tangent line is y  -5x - 3

20
TOPIC 5: CALCULUS Christos Nikolaidis

5.4 THE CHAIN RULE

Consider the function

f(x) = (2x2+3)2

Let us find the derivative f ΄(x). We first expand

f(x) = (2x2+3)2=4x4+12x2+9

and then
f ΄(x) = 16x3+24x (*)
What about
f(x) = (2x2+3)10 ?

Do not attempt to expand, it will be laborious!

Fortunately, there is an alternative, much easier way, to find f ΄(x).

The method is called chain rule.

The function looks like u10 where u=(2x2+3)

The derivative of u10 is 10u9, that is 10(2x2+3)9. The only extra

thing we have to do in order to complete the “chain” is to multiply
by u΄ that is by 4x. Therefore,

f(x) = (2x2+3)10 [Set u=2x2+3, so u΄=4x]

f ΄(x) = 10 (2x2+3)9 (4x)

= 40x (2x2+3)9

Similarly,
f(x) = (2x2+3)2 [Set u=2x2+3, so u΄=4x]
f ΄(x) = 2(2x2+3)(4x)
=8x(2x2+3)

Notice that f ΄(x) = 16x3+24x [as found in (*) above!]

In general, look at the differentiation table again

21
TOPIC 5: CALCULUS Christos Nikolaidis

f(x) f ΄(x)
xn nxn-1
sinx cosx
cosx -sinx
1
tanx
cos 2 x
ex ex
1
lnx
x
1
x 2 x

The chain rule says:

If instead of x we have another quantity u (depending on x) then

we perform the derivative as before for u and multiply the result
by u΄.

f(x) f ΄(x)
un nun-1 .u΄
sinu cosu .u΄
cosu -sinu .u΄
1 .
tanu u΄
cos 2 u
eu eu .u΄
1 .
lnu u΄
u
1 .
u u΄
2 u

EXAMPLE 1
f(x) = sin(2x2+3) [Here u=2x2+3]
f ΄(x) = cos(2x2+3) (2x2+3)΄
= 4x cos(2x2+3)

22
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 2

f(x) = e5x+3 [Here u=5x+3]

f ΄(x) = e5x+3 5

= 5e5x+3

EXAMPLE 3

f(x) = esinx [Here u=sinx]

f ΄(x) = esinxcosx

EXAMPLE 4

f(x) =ln (x2+4) [Here u= x2+4]

1
f ΄(x) = 2
2x
x 4
2x
= 2
x 4

EXAMPLE 5

f(x) = 3x 2  5x  2 [Here u=3x2+5x+2]

1
f ΄(x) = (6x+5)
2 3x 2  5x  2
6x  5
=
2 3x 2  5x  2

NOTICE

In many examples of using the chain rule, u has the form u=ax+b,
so that u΄=a. Hence,

we simply differentiate for u and multiply the result by a.

23
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 6
Let us consider all the usual functions where instead of x we have
u=3x+7.
f(x) f ΄(x)

sin(3x+7) 3cos(3x+7)

cos(3x+7) -3sin(3x+7)

e3x+7 3e3x+7

3
ln(3x+7)
3x  7
3
3x  7 2 3x  7

(3x+7)5 15(3x+7)4

3
tan(3x+7) 2
cos (3x  7)

Next, instead of x we have u=5x.

f(x) f ΄(x)

sin(5x) 5cos(5x)

cos(5x) -5sin(5x)

e5x 5e5x

5 1
ln(5x) 
5x x
5
5x 2 5x
5
tan(5x) 2
cos (5x)

24
TOPIC 5: CALCULUS Christos Nikolaidis

Perhaps the most confusing case for the chain rule is the function
of the form sinnx (as well as cosnx or tannx). Remember that

sinnx means (sinx)n [so that u=sinx]

EXAMPLE 7
a) f(x) = sin3x [u=sinx]
f ΄(x) = 3sin2x cosx

b) f(x) = sin2x [u=sinx]

f ΄(x) = 2sinx cosx

c) f(x) = cos5x [u=cosx]

f ΄(x) = -5cos4x sinx

d) f(x) = tan2x [u=tanx]

1
f ΄(x) = 2tanx
cos 2 x

NOTICE
In fact, the chain rule refers to the derivative of a composition of
two functions

(f
g)(x) = f(g(x))

It says that

f(g(x)  f (g(x))g(x)
I admit that this definition is not so “elegant”! The best way to
learn the chain rule is to practice with a great deal of examples.

It is possible to have a “double chain”, that is a chain inside

another chain! It is in fact the derivative of the composition of
three functions. The following example is indicative.

25
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 8

a) Let f(x) = ln(sin(3x+1))

1
f ΄(x) = [sin(3x+1)]΄ [u =sin(3x+1)]
sin(3x  1)
1
= [3cos(3x+1)] [v = 3x+1]
sin(3x  1)
cos(3x  1)
=3
sin(3x  1)

b) Let f(x) = 3sin5(x2+1)

f ΄(x) =15 sin4(x2+1)[sin(x2+1)]΄ [u =sin(x2+1)]

= 15 sin4(x2+1)cos(x2+1)(2x) [v = x2+1]

= 30xsin 4 (x 2  1)cos(x 2  1)

c) Let f(x) = e sin(3x)

f ΄(x) = e sin(3x) [sin(3x)]΄ [u =sin(3x)]

= e sin(3x) [3cos(3x)] [v = 3x]

= 3 e sin(3x) cos(3x)

d) Let f(x) = sin 2 x  sin2x

1
f ΄(x) = (sin 2 x  sin2x)
2
2 sin x  sin2x
1
= (2sinxcosx  2cos2x)
2
2 sin x  sin2x

sinxcosx  cos2x
=
sin 2 x  sin2x

26
TOPIC 5: CALCULUS Christos Nikolaidis

 AN ALTERNATIVE WAY TO LOOK AT THE CHAIN RULE

Let y depend on u, and u depend on x:

We have the chain

y

The chain rule says

dy dy du
=
dx du dx

[Notice: it is easy to remember this formula as it looks like a

simplification of fractions!!!]

Look at again

y = (2x2+3)10

Then u=2x2+3

so that y = u10

The chain rule gives

dy dy du
=
dx du dx

= 10u9 (4x)

= 10(2x2+3)9(4x) [replace back u=2x2+3]

= 40x(2x2+3)9 [as found earlier]

EXAMPLE 9
dy
Let y= esinx. Find y΄ = by using the formula
dx
dy dy du
=
dx du dx
[it is in fact Example 3 again]

27
TOPIC 5: CALCULUS Christos Nikolaidis

Solution
Set u=sinx so that y =eu

dy dy du
=
dx du dx
= eu cosx
= esinx cosx [as in Example 3]

The following example also justifies the term “chain”!!!

EXAMPLE 10 (Mainly for HL)

dP
Let P = Q3 and Q=lnR. Find in terms of R
dR

We have the chain

P

(as P depends on Q and Q depends on R)

The chain rule gives
dP dP . dQ
=
dR dQ dR
1
= 3Q2 . R
1
= 3(lnR)2 R

28
TOPIC 5: CALCULUS Christos Nikolaidis

5.5 MONOTONY – MAX, MIN

 INCREASING – DECREASING FUNCTIONS (MONOTONY)

Consider the following graph

y=f(x)

a1 a2 a3 a4 a5 a6 a7

Let us make some observations:

 The domain of the graph is the interval [a1,a7]

The points x=a1 and x=a7 are called endpoints
 We say that we have a local max (or just max) at points:
x=a2 , x=a4, x=a7
(can you explain why?)
 We say that we have a local min (or just min) an points:
x=a1 , x=a3, x=a5
(can you explain why?)
All these points (max and min) are called turning points or
extreme values
 Notice that x=a6 is not a turning point (neither max nor min),
as around f(a6) you can find smaller as well as larger values.
 The function is increasing (goes up) in the interval (a1,a2)
 The function is decreasing (goes down) in the interval (a2,a3)
 The function is increasing (goes up) in the interval (a3,a4)
 The function is decreasing (goes down) in the interval (a4,a5)
 The function is increasing (goes up) in the interval (a5,a7)

29
TOPIC 5: CALCULUS Christos Nikolaidis

Remember though that

A +tive gradient means that the function is increasing (goes up)
A -tive gradient means that the function is decreasing (goes down).

But we know that

derivative = gradient
In other words

If f  (x)  0 then f is increasing ( )

If f  (x)  0 then f is decreasing ( )

Notice: The increasing or decreasing behavior of a function is also

known as monotony!

 TURNING POINTS: MAX - MIN

How can we find the turning points (max or min) of a function?

First of all, the end points could be extreme values like a1 ,a7 above.

As far as the interior points is concerned, observe that the gradient

at any turning point is 0 (the tangent lines at those values are
horizontal!). Therefore, in the graph above, at x =a2, a3, a4, a5
f ΄(x)=0

PROPOSITION:
If f(x) has a turning point (max or min) at some interior point x0
and f ΄(x0) exists, then
f ΄(x0)=0

Notice though that f ΄(a6)=0 although a6 is not a turning point.

Hence the inverse proposition is not true.

30
TOPIC 5: CALCULUS Christos Nikolaidis

Therefore, apart from the endpoints, the candidate points for

turning points (max/min) are the following

1 points x where f ΄(x)=0

2 points where f ΄(x) does not exist

These points are called stationary points

In order to verify whether such a stationary point x0 is a turning

point (max or min) we must perform the following test

FIRST DERIVATIVE TEST for x0

Check the sign of f ΄(x) to verify if the function is increasing or
decreasing just before and after x0 :

x x0 x x0
f ΄(x) + - f ΄(x) - +
Conclusion for f(x)   Conclusion for f(x)  
Conclusion for x0 max Conclusion for x0 min

If the sign does not change we have neither max nor min.

 METHODOLOGY
Given y=f(x)
Step 1 we find f ΄(x)
Step 2 we solve f ΄(x)=0 (say that roots are a,b,c)
Step 3 we construct a table as follows to perform the
first derivative test

x a b c
f ΄(x) + - + +
Conclusion for f(x)    
Conclusion for x0 max min nothing

31
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 1
Consider
1 3
f(x)= x -2x2+3x+5
3
We find
f ΄(x)=x2-4x+3
We solve
x2-4x+3=0
The solutions are x=1 and x=3
We construct the table

x 1 3
f ΄(x)=x2-4x+3 + - +
Conclusion for f(x)   
Conclusion for x0 max min

Therefore,
we have a max at x=1 [and the max value of f is f(1)=6.33]
we have a min at x=3 [and the min value of f is f(3)=5]

An alternative way to check if a stationary point is a max or a min

(no need to check monotony) is the following:

SECOND DERIVATIVE TEST for x0

Find f ΄΄(x) (if it exists!)

If f ΄΄(x0) > 0 then x0 is a min

If f ΄΄(x0) < 0 then x0 is a max

If f ΄΄(x0) = 0 we don’t get an answer. We go back to the first

derivative test.

32
TOPIC 5: CALCULUS Christos Nikolaidis

Let us use the same example as above

EXAMPLE 2
Consider
1 3
f(x)= x -2x2+3x+5
3
We found
f ΄(x)=x2-4x+3
and the stationary points x=1 and x=3
We find
f ΄΄(x)=2x-4
For x=1, f ΄΄(1)=-1<0, so we have a max at x=1
For x=3, f ΄΄(3)= 2>0, so we have a min at x=3

EXAMPLE 3
Consider
f(x)= (x-1)4
We find
f ΄(x)=4(x-1)3
There is only one stationary point at x=1.
We have
x 1
f ΄(x)=4(x-1)3 - +
Conclusion for f(x)  
Conclusion for x0 min

However, the second derivative test does not give an answer here.
We find
f ΄΄(x)= 12(x-1)2
since f ΄΄(1)= 0 (neither positive nor negative)

33
TOPIC 5: CALCULUS Christos Nikolaidis

5.6 CONCAVITY - POINTS OF INFLECTION

 CONCAVITY
Consider again the graph of the preceding section

y=f(x)

b1 b2 b3 b4 b5 b6 b7

Our concern now is different! It is to investigate the intervals

where the curve
looks like () : we say that the function is concave up
looks like () : we say that the function is concave down2

We observe that:
 The function is concave down () in the interval (b1,b2)
 The function is concave up () in the interval (b2,b3)
 The function is concave down () in the interval (b3,b4)
 The function is concave up () in the interval (b4,b5)
 The function is concave down () in the interval (b5,b6)
 The function is concave up () in the interval (b6,b7)

 The concavity changes at the points x=b2, b3, b4, b5, b6, b7
These points are called points of inflection or inflection points

It is easy to verify the concavity by using the second derivative

If f ΄΄(x) > 0 then f is concave up (  )

If f ΄΄(x) < 0 then f is concave down (  )

2
To be more formal, a function is concave up/down if the tangent line at each
point is under/above the curve.

34
TOPIC 5: CALCULUS Christos Nikolaidis

Short explanation (mainly for HL)

Look at the curve of the following concave up function f(x)

The gradient is -tive in the beginning, it is “less” -tive as we move

forward, it sometimes becomes 0 and then becomes +tive and
“more” +tive as we move forward. In other words the gradient
increases. That is, the function of the gradient f ΄(x) is increasing.
But we know that the derivative of an increasing function is +tive.
Hence, the derivative of f ΄(x), that is f ΄΄(x) is +tive!
Similarly, if the function f(x) is concave down, the second derivative
must be -tive!

 INFLECTION POINTS
How can we find the inflection points?
Since the concavity changes at an inflection point, the sign of f ΄΄(x)
changes from + to – or vice-versa. Therefore the second derivative
an inflection point must be 0.

PROPOSITION:
If f(x) has an inflection point at some point x0 and f ΄΄(x0) exists,
then
f ΄΄(x0)=0

Notice again that the equation f ΄΄(x)=0 gives us the candidates x0

for inflection points. In order to verify if x0 is indeed an inflection
point we must check the sign of f ΄΄(x) just before and after that
point.

35
TOPIC 5: CALCULUS Christos Nikolaidis

 METHODOLOGY
Given y=f(x)
Step 1 we find f ΄(x) and f ΄΄(x)
Step 2 we solve f ΄΄(x)=0 (say that roots are a,b,c)
Step 3 we construct a table as follows

x a b c
f ΄΄(x) + - + +
Conclusion for f(x)    
Conclusion for x0 i.p. i.p. nothing

EXAMPLE 1
Consider again
1 3
f(x)= x -2x2+3x+5
3
We find
f ΄(x)=x2-4x+3
f ΄΄(x)=2x- 4
We solve
2x- 4=0
The solution is x=2
We construct the table

x 2
f ΄΄(x)=2x-4 - +
Conclusion for f(x)  
Conclusion for x0 i.p.

Let us summarize the information we have for that particular

example
1 3
f(x)=
x -2x2+3x+5
3
in order to sketch the graph of this function.

36
TOPIC 5: CALCULUS Christos Nikolaidis

First of all, it helps to know the y-intercept and the x-intercepts,

that is the roots of f(x) (if possible):
y-intercept: for x=0, y=f(0)=5
x-intercepts: we must solve f(x)=0 (only by GDC as the degree is 3)

We consider a unified table containing the solutions of both f ΄(x)=0

and f ΄΄(x)=0:
x 1 2 3
f ΄(x) + - - +
f ΄΄(x) - - + +
Conclusion for f(x)    
   
Conclusion for x0 max i.p. min

The following table of values will also help

x 0 1 2 3
f(x) 5 6.33 5.66 5

y
7
max
6 i.p.

5
min
4

1
x
-3 -2 -1 1 2 3 4 5 6 7
-1

37
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 2
Consider the function
f(x)= x e x
Find possible maximum, minimum values and points of inflection.
Solution
We have
f ΄(x)= x e x + e x

Stationary points: x e x + e x =0  e x (x+1)=0  x=-1

We use table:
x -1
f ΄(x) - +

min

Furthermore,
f ΄΄(x)= x e x + e x + e x = x e x +2 e x

Then x e x +2 e x =0  e x (x+2)=0  x=-2

We use table:
x -2
f ΄΄(x) - +
 
i.p.

Notice: Look at the graph of this function at your GDC to confirm

that x=-1 gives a min and observe that x=-2 gives a point of
inflection.

38
TOPIC 5: CALCULUS Christos Nikolaidis

5.7 OPTIMISATION

In problems of optimization we have to construct a function in

terms of some variable x, and then we use derivatives to find the
“optimum” solution, that is the maximum or the minimum value
of the function.

EXAMPLE 1
Among all the rectangles of perimeter 20, find the one of the
maximum area.
Discussion
A rectangle of perimeter 20 may have dimensions
1x9 2x8 3x7 4x6 etc
The corresponding areas are
9 16 21 24 etc
Which is the one of the maximum area?
Solution

Let x be one of the sides (this will be our main variable).

If the other side is y, then
Perimeter = 20  2x +2y =20  y = 10-x

The function of optimisation is

Area: A = xy = x(10-x)=10x-x2

dA
We find  10  2x
dx
dA
Stationary points: = 0  10  2x = 0  x =5
dx
The 2nd derivative test is easier here: A΄΄ = -5.
At x=5 A΄΄ < 0, thus we have a maximum value there.

39
TOPIC 5: CALCULUS Christos Nikolaidis

Therefore, the rectangle of maximum area is the square (that is

when x = 5), and the maximum area is A = 25.

Let’s reverse the role of the perimeter and the area. Next we know
the area of the rectangle and we are looking for the minimum
perimeter.

EXAMPLE 2
Among all the rectangles of area 25, find the one of the minimum
perimeter.
Solution

Again, let x be one of the sides (this will be our main variable).
If the other side is y, then
25
Area = 25  xy=25  y =
x
The function of optimisation is
50
Perimeter: P = 2x+2y = 2x+
x
dP 50
We find = 2 2
dx x
dP 50
Stationary points: = 0  2  2 = 0  x 2 = 25  x = 5
dx x
100
The 2nd derivative test gives: P΄΄ = .
x3
For x=5, P΄΄ > 0 , thus we have a minimum value there.

Therefore, the rectangle of minimum perimeter is the square, (that

is when x = 5), and the minimum perimeter is A = 25.

40
TOPIC 5: CALCULUS Christos Nikolaidis

Sometimes, there is a different “cost” for each side.

EXAMPLE 3
We want to construct a rectangle fence for an area of 24m2, but
the cost for the material of the front side is 10$ per meter while
the cost for the material of the other 3 sides is 5$ per meter. Find
the cheapest solution!
Solution

Let x be the front side (this will be our main variable).

If the other side is y, then
24
Area = 24  xy=24  y =
x
The function of optimisation is
240
Cost: C = 10x+5x+2(5y) = 15x+10y = 15x 
x
dC 240
We find = 15  2
dx x
dC 240
Stationary points: =0  15  2 = 0  x 2 = 16  x = 4
dx x
480
The 2nd derivative test gives: C΄΄ = .
x3
For x=4, C΄΄ > 0 , thus we have a minimum value there.

Therefore, the best rectangle has dimensions 4 x 6 and the

minimum cost is C = 120$

This rationale applies to 3D shapes as well. For example, they may

give us a rectangular prism or a cylinder of a given volume and we
are looking for the optimum surface area, or vice-versa.

41
TOPIC 5: CALCULUS Christos Nikolaidis

Let’s see an example of a shape determined by the boundaries of a

given function!

EXAMPLE 4
Consider the region enclosed by y  9  x 2 and x-axis.
Find the rectangle of largest area inscribed within that region.
y
10

x
-4 -3 -2 -1 1 2 3 4

Discussion
There are two extreme cases:
 the height of the rectangle is 0, the width is 6. The area is 0.
 the height of the rectangle is 9, the width is 0. The area is 0.
Somewhere in between there is a rectangle of maximum area.
Solution
Key point: We call x the coordinate of the bottom right corner.
Then
Width = 2x
Height = y  9  x 2 (the function itself!)

Thus, the function of optimisation is

Area: A = 2x(9-x2) = 18x-2x3

dA
We find  18  6x 2
dx
dA
Stationary points: = 0  18  6x 2 = 0  x = 3
dx
The 2nd derivative test gives: A΄΄ = -12x.
At x= 3 A΄΄ < 0, thus we have a maximum value there.
Therefore, the rectangle of maximum area has dimensions 2 3 x6
and the maximum area is A = 12 3 .

42
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 5

5km
3km

C 4km B
B

A swimmer is at point A inside the sea, 3km away from the beach.
She wants to go to point B at the beach, which is 5 km away.
When she swims she covers 1 km in 8 minutes
When she runs she covers 1km in 5 minutes. Find
(a) the time she spends when she swims directly to B;
(b) the time she spends when she swims to C and then runs to B;
(c) the best time she can achieve if she swims first to some point
D between B and C and then runs to B
Solution
(a) T=TAB= 58 = 40 min
(b) T= TAC + TCB = 38 + 45 = 44 min
(c) Let D be between C and B and CD=x km. Then DB = 4-x km
Also AD2 = AC2 + CD2  AD = 9 x2
Therefore, the function of optimization is the total time

T= TAD + TDB = 8 9  x 2 + 5(4-x) = 8 9  x 2 -5x+20 min

dT 8x
We find  5
dx 9 x2
8x
Stationary points:  5  0  8x  5 9  x 2
2
9 x
15
 64x 2  25(9  x 2 )  39x 2  225  x   2.4 km
39
We can easily verify that this is a min (2nd derivative test).
Thus, the point D is 2.4km from C. The best time is T=38.7 min.

43
TOPIC 5: CALCULUS Christos Nikolaidis

5.8 THE INDEFINITE INTEGRAL  f(x)dx

 THE INDEFINITE INTEGRAL
Consider
F(x)=x2
The derivative is F ΄(x)=2x

The inverse problem: If they give us the result

f(x)=2x
can we find a function F(x), such that F ΄(x) = f(x) ?

Of course, the answer could be F(x)=x2. We say that F(x)=x2 is an

antiderivative of f(x)=2x. But it is not the only one!
Notice that
x2
x2+1
x2+2
x2+5
x2+2007
in general x2+c for any constant c

are also antiderivatives of f(x)=2x.

We say that x2+c is the indefinite integral of f(x)=2x and we use the
notation

 2xdx = x2+c
Hence,

if F ΄(x) = f(x) then  f(x)dx = F(x)+ c

For example,

 5x
4
since (x5)΄=5x4 we obtain dx = x5 + c
x5
 x dx =
4
we deduce that + c (why?)
5

44
TOPIC 5: CALCULUS Christos Nikolaidis

Therefore, we can easily obtain the following results

f(x)  f(x)dx
1 x +c
x2
x +c
2
x3
x2 +c
3
x4
x3 +c
4
x 11
x10 +c
11

In general,

x n 1
 x dx =
n
+c (if n  -1)
n 1

Notice also,

 5dx = 5x + c, since (5x)΄=5

 adx = ax + c, since (ax)΄=a (a=constant)

 cosxdx = sinx + c, since (sinx)΄=cosx

If we remember the derivatives of the basic functions we obtain the

following results:
f(x)  f(x)dx
x n 1
xn
n 1
a ax
cosx sinx
+c
sinx -cosx
ex ex
1
lnx
x

NOTICE: The action of finding the integral is called integration.

45
TOPIC 5: CALCULUS Christos Nikolaidis

x n 1
 x dx =
n
 REMARK FOR +c
n 1

The same formula applies for –tive values of n. For example,

x -4 1
 x dx =
-5
+c = +c
-4 - 4x 4

1 1
What about x 2
dx ? We know that
x 2
=x-2, so
-1
1 x 1
 x 2 dx =  x dx = - 1 +c = - x +c
-2

Also, the same formula applies for rational values of n:

x 8/5 5 8/5
 x dx =
3/5
+c = x +c
8/5 8

What about  x dx ? We know that x =x1/2 , so

x 3/2 2 3/2
 x dx =  x 1/2
dx = +c = x +c
3/2 3

1
 x dx =  x
-1
Notice that this formula does not apply for dx .

1
Only for this particular power we have the formula  x dx =lnx+c

1
 REMARK FOR  x dx (only for HL)

1
In fact  x dx = ln|x|+c .
1
Indeed: if x>0, then [ln|x|]΄ =[lnx]΄= ,
x
1 1
if x<0, then [ln|x|]΄ =[ [ln(-x)]΄= (-1)= .
-x x
1
That is why the antiderivative of is ln|x|+c
x

46
TOPIC 5: CALCULUS Christos Nikolaidis

 RULES OF INTEGRATION

 [f(x)  g(x)]dx =  f(x)dx   g(x)dx

 af(x)dx = a  f(x)dx

That is, if we have a “long” expression like

 [af(x)  bg(x)  ch(x)  dk(x)]dx

where a,b,c,d are constants, we do not “touch” a,b,c,d and
integrate only f(x), g(x), h(x), k(x):

a  f(x)dx  b  g(x)dx  c h(x)dx  d  k(x)dx

EXAMPLE 1

 [3x  5e x  2cosx]dx = 3  x 2 dx +5  e x dx -2  cosxdx

2

x3
= 3( )+ 5ex -2sinx + c
3
= 3 x 3 + 5ex -2sinx + c

In fact, the shaded step above is not necessary! We can proceed to

the next step by estimating directly the integrals of x2, ex, cosx

EXAMPLE 2
x5 x4 x3 x2
 [2x  8x  5x  7x  2]dx = 2
4 3 2
+8 -5 -7 +2x + c
5 4 3 2
2 5 5 7
= x +2 x 4 - x3 - x2 + 2x + c
5 3 2

 FIND the constant c

Sometimes, we are given an extra condition of the form f(a)=b in

order to find the value of the constant c. We find in fact, the
specific antiderivative which satisfies this condition.

47
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 3
Let f ΄(x)=6x2- 4x+5. Find f(x) if f(1)=8.

Clearly, f(x) is the integral of f ΄(x)=6x2- 4x+5. That is,

x3 x2
 [6x  4x  5]dx = 6
2
f(x)= -4 +5x+c = 2x3-2x2+5x+c
3 2
Next, we have to find the value of c:

f(1)=8, so 2(1)3-2(1)2+5(1)+c=8
so 5+c=8
so c=3
Therefore, f(x) = 2x3-2x2+5x+3

 TWO MORE BASIC INTEGRALS (only for HL)

In the IB Math HL formula booklet you will also find the formulas

1 1 x
a 2
x 2
dx  arctan  c
a a

1 x
 2
a x 2
dx  arcsin
a
c

1 x
Indeed, the derivative of f(x) = arctan is
a a
1 1 1 1 1 1
f ‘(x) = 2
  2 2 2 
a x a a a x a  x2
2
1 
a a2

x
Similarly, the derivative of f(x) = arcsin is
a
1 1 1 1 1
f ‘(x) =   
x
2 a a a2  x 2 a2  x 2
1 
a a2

Later on we will present a different proof of those two formulas.

48
TOPIC 5: CALCULUS Christos Nikolaidis

Let’s find some integrals of this form

EXAMPLE 4
1
1x 2
dx  arctanx  c

1 1 x
4x 2
dx 
2
arctan  c
2

5 5 x
 13  x 2
dx 
13
arctan
13
c

5 5 1 52 2x 5 2x
 9  4x 2
dx 
4  9 2
dx 
43
arctan
3
 c = arctan
6 3
c
x
4

Similarly

1
 1  x2
dx  arcsinx  c

1 x
 4x 2
dx  arcsin
2
c

5 x
 13  x 2
dx  5arcsin
13
c

5 5 1 5 2x
 9  4x 2
dx 
2  9 2
dx  arcsin
2 3
c
x
4

49
TOPIC 5: CALCULUS Christos Nikolaidis

5.9 INTEGRATION BY SUBSTITUTION

Before we present this method of integration, let us see a simple

case where we have (x+b) or (ax+b) instead of x in our function.

 THE LINEAR CASE:  f(x  b)dx or  f(ax  b)dx

Consider the integrals

 cos(x  5)dx
 cos(3x  5)dx
They look like  cosxdx but instead of x we have u=x+5 and u=3x+5
respectively.

If we have x+b instead of x,

we estimate the integral as before, but

we write x+b instead of x in the result.

If we have ax+b instead of x,

we estimate the integral as before, but

1
we write ax+b instead of x, and multiply the result by
a

Hence,

 cos(x  5)dx = sin(x+5) +c [u=x+5]

1
 cos(3x  5)dx = 3
sin(3x+5) +c [u=3x+5]

[(Indeed, if we differentiate the results

the derivative of sin(x+5) is cos(x+5)

1 1
the derivative of sin(3x+5) is cos(3x+5) .3 = cos(3x+5) ]
3 3
Similarly,
1
 cos(3x)dx = 3
sin(3x) +c [u=3x]

50
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 1

e 3x  2
 e dx
3x  2
= +c
3
e-5x  2
e
-5x  2
dx = +c
-5
e 10x
 e dx
10x
= +c
10

e
x 8
dx = ex+8 +c

cos(2x  1)
 sin(2x  1)dx = 
2
+c

cos2x
 sin2xdx = 
2
+c

sin(5 - x)
 cos(5 - x)dx =
-1
+c = -sin(5-x) +c

1 ln 7x  3
 7x  3 dx =
7
+c

1 ln 3- 7x 
 3- 7x dx =
-7
+c

1
 x  3 dx = ln(x-3) +c

(3x  5)6
 (3x  5) dx = 3  6 +c
5

EXAMPLE 2
Find the integral I   cos 2 xdx
by using the double angle formula cos2x  2cos 2 x  1
Solution
1  cos2x
If we solve for cos2x we get cos 2 x  .
2
Thus
1 1 sin2x  x sin2x
I
2  (1  cos2x)dx   x 
2 2 
c  
2 4
c

51
TOPIC 5: CALCULUS Christos Nikolaidis

 SOME DISCUSSION ABOUT THE DIFFERENTIAL dx

(you may ignore this page, it is extra information)

The quantity dx is called the differential of x.

It indicates a little change in x. (Compare with Δx=x2-x1)

For a function y=f(x), we know that

dy
 f (x)
dx
dy
Although is a symbol and not a fraction, it behaves like a
dx
Δy
fraction. Besides, it comes from a fraction .
Δx
Thus we can solve for dy and write

dy  f (x)dx

This gives a relation between the differentials dy and dx.

In fact, this relation indicates in what extend a little change in x

causes a little change in y.

For example, consider the function y=x2. Then

dy
=2x  dy=2xdx
dx

We can also solve for dx and write

dy
dx= .
2x

Similarly, for y=sinx

dy
=cosx  dy=cosxdx
dx
dy
or dx=
cosx

52
TOPIC 5: CALCULUS Christos Nikolaidis

 THE METHOD OF SUBSTITUTION (THE SIMPLE CASE)

In fact, we have already seen a simple version of substitution,

where an expression of the form u=ax+b appears in the integral.

For example, we have seen that

1
I=  cos(3x  5)dx  sin(3x  5)  c
3

Let us use this example to demonstrate analytically the method of

substitution.

 Let u=3x+5
du du
 Then  3  dx 
dx 3
 The integral becomes
du 1 1
I=  cosu   cosudu  sinu  c
3 3 3
1
Thus, I= sin(3x  5)  c
3

In u=ax+b we can directly write down the result as explained. But

for more difficult cases we follow this process.

In general, our target is to remove all x’s, as well as dx, from the
original integral and obtain a simpler integral in terms of u and du.

METHODOLOGY:
We select an appropriate substitution u=g(x)
[having in mind that g (x) must exist in the integral]

 Let u=g(x)
du du
 Then = g (x)  dx=
dx g(x)
 Express the initial integral in terms of u and du
 Calculate the new integral
 Replace u= g(x) back in the result

53
TOPIC 5: CALCULUS Christos Nikolaidis

Consider
I=  3x 2 (x 3  5)7 dx

We select u=x3+5 [since the derivative 3x2 exists inside the integral]

Let u=x3+5
du du
Then =3x2  dx=
dx 3x 2

The integral obtains the convenient form

du
I=  3x 2 u7
3x 2 
= u7 du

Hence,
u8
I= c
8
Finally, we replace back u=x3+5 to get

(x 3  5)8
I= c
8

Notice that
I=  x 2 (x 3  5)7 dx

can be treated in the same way. The derivative of u=x3+5 is 3x2.

We are happy that x2 exists inside the integral (we don’t mind for
the constant 3). Thus,

let u=x3+5
du du
then =3x2  dx=
dx 3x 2

The result is
du 1 7 1 u8 (x 3  5)8
I=  x 2 u7
3x 2 3 
= u du =  c = c
3 8 24

Sometimes, the substitution is given as a hint! See next example.

54
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 3
Find I=  x x 2  3dx by using the substitution u=x2+3.
Solution
du du
Let u=x2+3, then =2x  dx=
dx 2x
Thus,
3
du 1 1 2 2 1 3 1 3
I=  x u =  udu = u +c= u 2 +c= (x 2  3) 2 +c
2x 2 2 3 3 3

EXAMPLE 4
2x  cosx
Find I=  dx .
x 2  sinx
Solution
Notice that the derivative of the denominator is the numerator.
du du
Let u= x 2  sinx , then =2x+cosx dx=
dx 2x  cosx
Thus,
2x  cosx du 1
I=  =  du =lnu+c=ln (x 2  sinx)  c
u 2x  cosx u

If you get used to this simple case of substitution you may directly
write down the result. But unless you feel confident enough follow
the whole process!

EXAMPLE 5
(lnx) 2 1 lnx
Find I 1=  dx , I 2=  dx , I 3=  dx
x x(lnx) 2 x
Solution
For all three of them we let u=lnx [Why?].
1
We will obtain the integrals of u 2 , 2 , u respectively. Then
u
(lnx)3 1 2
I 1= c, I 2=   c, I 3= (lnx)3/2  c (check!)
3 lnx 3

55
TOPIC 5: CALCULUS Christos Nikolaidis

b
5.10 THE DEFINITE INTEGRAL  f(x)dx -AREA BETWEEN CURVES
a

 THE CALCULATION
Before we discuss the use of the definite integral, we will show that
its estimation is very easy!

If  f(x)dx = F(x) + c ,
 f(x)dx = F(x)
b b
then a
which means F(b)-F(a)
a

For example,
since  (2x  3)dx = x2 + 3x + c
we have 
0
4
  4
(2x  3)dx = x 2  3x 0 = (42+3.4)-(02+3.0) = 28

(2x  3)dx = x  3x 
3 3

2
and -1 =(9+9)-(1-3)= 18+2=20
1

Notice: the constant c is omitted in the definite integral! (why?)

 THE FORMAL DEFINITION

Consider the graph of the straight line f(x)=2x

y=2x

0 3

 The area below the line y=2x from x=0 up to x=3 is in fact the
36
area of the shaded triangle: 9
2
 
0
3
 
2xdx = x 2
3
0 =32-02 = 9

This is not an accident!

56
TOPIC 5: CALCULUS Christos Nikolaidis

In general, if f(x) is a continuous function and f(x)>0

(i.e. the graph is a continuous line above the x-axis)

y=f(x)

a b

then the shaded area under the curve, from x=a to x=b is given by
the definite integral

b
 f(x)dx
a

Let us see an example

EXAMPLE 1
Find the area under the curve y=x2, between the vertical lines x=1
and x=3

9 y=x2

1 3

3
3
2  x 3  33 1 3
It is  x dx =   =  = 9- 0.33 = 8.66
1
 3 1 3 3

57
TOPIC 5: CALCULUS Christos Nikolaidis

IMPORTANT REMARK
b
If f(x)<0 then  f(x)dx
a
is also negative! It is in fact the negative

value of the area between the line y=f(x) and the x-axis.

For example, for f(x)=-x2,

3
3
2 - x 3  33 1 3
0 - x dx = 
 3 0
 =-
3

3
= -9+ 0.33 = -8.66

Suppose that f(x) has both positive and negative values, say

y=f(x)

A
C

a b B c d

Then
d d
a
f(x)dx = A-B+C while  | f(x) | dx = A+B+C
a

Hence,
d
the integral 
a
f(x)dx does not estimate

the total area between the curve and the x-axis.

If they ask us to find the total shaded area, that is A+B+C, we

must consider
d
AREA =  | f(x) | dx
a

which in fact splits into

b c d
 f(x)dx -  f(x)dx + 
a b c
f(x)dx

58
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 2
Consider the function
f(x)=sinx, 0  x  2π

Find
π 2π 2π
(a)  0
f(x)dx (b) 
π
f(x)dx (c) 0
f(x)dx

(d) the total area between the curve and the x-axis within [0,2π]

Solution
π
(a)  sinxdx =  cosx ] 0π =-cosπ+cos0=1+1=2
0

2π
(b)  sinxdx =  cosx ] π2π =-cos2π+cosπ=-1-1=-2
π

2π
(c)  sinxdx =  cosx ] 02π =-cos2π+cos0=-1+1=0
0

[it is in fact the sum of the first two integrals above]

(d) The total area is given by

2π
A= 0
| sinx | dx

In practice, we find (a) and (b) and add the absolute values

A = 2+2= 4

Notice: your GDC in graph mode, estimates directly both

(c) the definite integral = 0 and (d) the total area = 4

59
TOPIC 5: CALCULUS Christos Nikolaidis

 PROPERTIES OF DEFINITE INTEGRALS

The rules of indefinite integrals are still valid here
b b b
  [f(x)  g(x)]dx =  f(x)dx +  g(x)dx
a a a
b b
  af(x)dx =a  f(x)dx
a a

Moreover,
b c c
 a
f(x)dx +  f(x)dx =  f(x)dx
b a
(use areas to explain why!)

[notice that the bounds a,b,c are consecutive]

a b
 b
f(x)dx =   f(x)dx
a
(accepted by definition)

 f ΄(x)dx = f(x)
b b
 a
(derivative and integral
a

are inverse to each other)

b b b
  f(x)dx
a
=  f(y)dy
a
=  f(t)dt
a
= …

[in other words, the variant x is irrelevant!!]

EXAMPLE 3
5
Suppose that  f(x)dx  10 . It is also given that f(0)=15, f(5)=3
0

Then we can also estimate the following integrals:

5
 
0
2f(x)dx = 20
0
  f(x)dx
5
= -10

0 1
 
5 2
f(x)dx = -5

 
0
5
(f(x)  4x)dx = 
5

0
5
f(x)dx +  4xdx = 10 + 2x 2
0
 
5
0 = 10+50 = 60
5 5 5
2f(x)dx +  1dx = 20 + x 0 = 20+5 = 25
5
 
0
(2f(x)  1)dx = 
0 0

2 5 5
 
0
f(x)dx  2
f(x)dx =  f(x)dx  10
0

= f(x)0 = f(5)-f(0) = 15- 3 = 12

5 5
  f ΄(x)dx
0

5 5
 
0
f(t)dt =  f(x)dx
0
= 10
8 5
 
3
f(x - 3)dx =  f(y)dy  10
0
by setting y=x-3, dy=dx.

60
TOPIC 5: CALCULUS Christos Nikolaidis

 SUBSTITUTION FOR DEFINITE INTEGRALS

If substitution is needed for definite integrals, it would be safe to

find the indefinite integral first and then proceed to the definite
integral. This is because the boundaries of the initial integral refer
to dx and not to du. Let us explain.

EXAMPLE 5
2 x
Find I=  2
dx ,
0 x 4
Solution

Method A: we find the indefinite integral first

du du
We use u=x2+4, thus  2xdx  dx= , so that
dx 2x
x x du 1 1 2
 x 2  4 dx =  u 2x = 2 lnu  c = 2 ln(x  4)  c
Therefore,
2
1  1 1 1 8 1
I=  ln(x 2  4) = ln8  ln4 = ln = ln2
2 0 2 2 2 4 2

Method B: we change the boundaries

du du
Again we use u=x2+4, thus  2xdx  dx=
dx 2x
We also change the boundaries of the definite integral using u=x2+4

x u
0 4
2 8

In this case it is not necessary to go back to x.

Namely,
8
2 x 8 x du 1  1 1 1
I=  2 dx =  =  lnu = ln8  ln4 = ln2
0 x 4 4 u 2x
2 4 2 2 2

61
TOPIC 5: CALCULUS Christos Nikolaidis

 AREA BETWEEN CURVES

Consider the graphs of two functions y=f(x) and y=g(x), such that
f(x)  g(x) for any x

y=f(x)

y=g(x)

a b

The area between the two curves from x=a to x=b is given by
b
 [f(x)- g(x)]dx
a

b
Indeed,  f(x)dx gives the area under y=f(x)
a
b
while  g(x)dx gives the area under y=g(x),
a

hence their difference gives the shaded area requested!

NOTICE
Even if we have only one function, for example f(x)=sinx

 in [0,π] we can think of the area between y=sinx and y=0

π π
Area =  [sinx - 0]dx = 
0 0
sinxdx

 in [π,2π] we can think of the area between y=0 and y=sinx

2π 2π
Area = 
π
[0 - sinx]dx =  
π
sinxdx

62
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 6
Find the shaded area below:

y=
1 x+4
2

y=x

2 6

It is
6 1 6 1
Area =  [( x  4)- x]dx =  (4 - x)dx
2 2 2 2
6
 x2 
= 4x  = (24-9)-(8-1) = 15-7 = 8
 4  2

In general, the area between two curves is given by

b
 | f(x)- g(x) | dx
a

Consider the following situation

y=f(x)

y=g(x)

a k b

b
The shaded area is given by A=  | f(x)- g(x) | dx . More explicitly,
a

k b
A=  [f(x)- g(x)]dx +  [g(x)- f(x)]dx
a k

63
TOPIC 5: CALCULUS Christos Nikolaidis

 METHODOLOGY
 Find the intersection points by solving the equation f(x)=g(x)
(in the picture above there is only one intersection point, k)
 Determine which function is above the other within each
interval
(in the picture above, f(x)  g(x) in [a,k] , g(x)  f(x) in [k,b] )
 Split the integral appropriately:
k b
(in the example above:  [f(x)- g(x)]dx
a
+  [g(x)- f(x)]dx
k
)

EXAMPLE 7
Find the area enclosed by the graphs f(x)=x2 and g(x)=x+2 in the
first quadrant

y=x+2

y=x2

 Intersection points:
f(x)=g(x)  x2=x+2
 x2-x-2=0
 x=-1, x=2
Since we are interested in the first quadrant only, we consider only
x=2.
 Within [0,2], x+2  x2
 The shaded area is
2
2
2
2
2 x2 x3 
0 [(x  2)- x ]dx  0 (x  2- x )dx =  2  2x  3 
0
8 10
= (2+4- )- 0 = = 3.33
3 3

64
TOPIC 5: CALCULUS Christos Nikolaidis

Mind carefully the region required.

EXAMPLE 8
Consider the curve
y x

We define two regions

A: among the curve y  x , x-axis and the line x=9.

B: among the curve y  x , y-axis and the line y=3.

The corresponding areas are given by

9 9 9
1/2 2  2
A   x dx   x dx   x 3/2   93/2  0  18
0 0 3 0 3

9 9 9

 
B   3  x dx   3- x  1/2
  2   2 
dx  3x  x 3/2    27  93/2   0  9
3 3
0 0  0  
or
B  ( Area of rectangle)  A  27 - 18  9

65
TOPIC 5: CALCULUS Christos Nikolaidis

5.11 KINEMATICS (DISPLACEMENT-VELOCITY-ACCELERATION)

Consider a body moving along a straight line. The body is free to

move forwards and backwards. The displacement s is the distance
|OA| from a fixed point O (also known as the origin).

O A

The displacement s is given as a function of time. For example

s=t2-4t+3
means that
at time t=0 the displacement is 3 units from the fixed point O
at time t=1 the displacement is 0, (the body goes back to point O)
at time t=2 the displacement is -1, (the body is before point O)
at time t=3 the displacement is 0, (at point O again)
at time t=4 the displacement is 3, (it is moving forward)

Then
ds
Velocity = rate of change of displacement: v=
dt

dv
Acceleration = rate of change of velocity: a=
dt

d2s
Notice also that a is the second derivative of s. a=
dt 2

Displacement Velocity Acceleration

s v a

derivative

66
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 1
Consider
s= t3-12t+15
representing the motion of a particle along a straight line, where
the displacement s is given in m (meters),
the time t is given in sec (seconds).
Then
ds dv
v==3t2-12 a= =6t
dt dt
Notice that v is measured in m/sec while a is measured in m/sec2.
For example,
at time t=1, s=4m
v=-9m/sec
a=6m/sec2
at time t=3, s=6m
v=15m/sec
a=18m/sec2
Notice also,

The body is stationary when the velocity is 0

Hence, let us solve

v=0
 3t2-12=0
 t2=4
 t=2sec

(notice that time is always positive, thus we reject t=-2)

At that time (t=2), s=-1m

v=0m/sec
a=12m/sec2

67
TOPIC 5: CALCULUS Christos Nikolaidis

NOTICE
If s>0, the body is to the right of the fixed point O.
If s<0, the body is to the left of the fixed point 0.
If s=0, the body is at the fixed point 0.

If v>0, the body is moving to the right

If v<0, the body is moving to the left
If v=0, the body is stationary (it changes direction)

If a>0, the body accelerates

If a<0, the body decelerates
If a=0, the velocity is stationary (probably maximum speed)

The motion may be given as a graph of s with respect to t

s(t)

The displacement s from the fixed point O is in fact the vertical

distance from the horizontal axis. This distance may be positive,
negative or zero.

68
TOPIC 5: CALCULUS Christos Nikolaidis

 GOING BACKWARDS (BY USING INTEGRATION)

If we are given the acceleration of a moving body we can find the

velocity and the displacement by using integration.

Displacement Velocity Acceleration

s v a
derivative

integral

However, in this opposite direction we must be given some extra

information in order to estimate any emerging constant c.

EXAMPLE 2
Let v=12t2-2t. Find the acceleration and the displacement, given
that the initial displacement is 5m.
Solution
dv
a= =24t-2 ms-1
dt
s=  vdt =  (12t 2 - 2t)dt =4t3-t2+c
but s=5, when t=0, thus c=5. Hence, s=4t3-t2+5 m

EXAMPLE 3
Let a=12t. Find the displacement, given that the moving body
starts from rest; the initial displacement is 5m.
Solution
v=  adt =  12tdt =6t2+c
but v=0, when t=0, thus c=0. Hence, v=6t2
s=  vdt =  6t 2 dt =2t3 +c
but s=5, when t=0, thus c=5. Hence, s=2t3 +5 m

We must distinguish between the displacement and the distance

travelled of a moving body.

69
TOPIC 5: CALCULUS Christos Nikolaidis

 DISPLACEMENT vs DISTANCE TRAVELLED

Let the velocity of a moving body be given in ms-1 by

v = 12 – 3t 2

while the initial displacement from a fixed point O is 0m.

The displacement of the body is given by

s=  vdt =12t-t3+c

Since s=0 when t=0, we obtain

s=12t-t3

Let us observe the velocity, the displacement, as well as the total

distance travelled at particular times:

Position Time Velocity Displacement Distance

t v s travelled
0 Initially, t=0 v=12 ms-1 s = 0m 0m
A At t=1 v=9 ms-1 s = 11m 11m
B At t=2 v=0 ms-1 s = 16m 16m
C At t=3 v= -15 ms-1 s = 9m 23m

In other words, the moving body

starts from O
moves forward in the first two seconds (positions A,B)
rests for a while at B (s=16m)
goes back to position C (s=9m)

O C A B

0 2 4 6 8 10 12 14 16 18 20m

Notice that at C, i.e. after 3 seconds, the distance travelled goes up

to 23m (16m forward from A to C + 7m backwards from C to D).

The mathematical tools are very clear: the indefinite integral gives
the displacement, the definite integral gives the distance travelled.

70
TOPIC 5: CALCULUS Christos Nikolaidis

Displacement s=  vdt
b
Distance travelled from t=a to t=b d=  v dt
a

In our example,

the distance travelled in the first two seconds is given by

2 2 2

d=  v dt   12 - 3t dt   (12 - 3t 2 )dt  12t  t 3
2
 2
0 =16m
0 0 0

the distance travelled in the first three seconds is given by

3 3 2 3
d=  v dt   12 - 3t dt   (12- 3t )dt -  (12- 3t 2 )dt =16+7=23m
2 2

0 0 0 2

NOTICE
If we draw the graph of v against t,

1 2 3

The distance travelled from t=0 to t=2 is given by the area above
the t-axis.

The distance travelled from t=0 to t=3 is given by the total area
between the curve and t-axis, from t=0 to t=3 (shaded area above)

Be careful, even when v is positive, the displacement and the

distance travelled do not always coincide.

71
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 4
Let

v  4t  t 2
Given that the initial displacement is 10m find the displacement
and the distance travelled a) after 3 sec b) after 6 sec.
Solution
 For the displacement
t3
s  2t 2  c
3
Since s=10 when t=0, we obtain
t3
s  2t 2   10
3
Therefore,
when t=3, s  19m
when t=6, s  10m

 For the distance travelled, in the first n seconds (in general)

n n
d   v dt   4t  t 2 dt
0 0

We need the stationary points first:

v  0  4t  t 2  0  t  0 or t  4

We observe that the velocity is positive in the first 4 seconds.

Hence, in the first 3 seconds

3 3
2 2 t3 
d   (4t  t )dt  2t    9m
0  3 0

while, in the first 6 seconds

4 6
32 32 64
d   (4t  t )dt   (4t  t 2 )dt 
2
   21.3 m
0 4
3 3 3

Notice The GDC gives directly the results

3 6
d   4t  t 2 dt  9 m d   4t  t 2 dt  21.3 m
0 0

72
TOPIC 5: CALCULUS Christos Nikolaidis

ONLY FOR

HL

73
TOPIC 5: CALCULUS Christos Nikolaidis

74
TOPIC 5: CALCULUS Christos Nikolaidis

5.12 CONTINUITY AND DIFFERENTIABILITY

 CONTINUITY
In paragraph 5.1 we have seen that

for f(x) = 2x+3, lim f(x)  7 = f(2)

x 2

sinx
for f(x)  lim f(x)  1 , but f(0) is not defined.
x x 0

sinx
If we observe the graph of f(x) 
x
y

x
-5 5

we notice a “discontinuity” at x=0.

It’s worth it to see another similar situation.

x2  4
f(x) 
x2

The function is not defined at x=2. However x=2 is not a vertical

asymptote. It is interesting to see what happens to f(x) as x 2.

x f(x)
1.9 3.9
1.99 3.99
1.999 3.999
2.001 4.001
2.002 4.002

It seems that lim f(x)  4 .

x 2

75
TOPIC 5: CALCULUS Christos Nikolaidis

Indeed, look at the graph of f (there is an “empty” point on it)

y
6

1
x
-3 -2 -1 1 2 3 4
-1

When x approaches 2, the value f(x) approaches 4. Thus

f(2) is not defined

but lim f(x)  4 .

x 2

In fact, we can simplify the function as

x 2  4 (x  2)(x  2)
f(x)    x  2, where x  2
x2 x2

That is why we obtain the graph of the straight line y=x+2 with
some “discontinuity” at x=2. Moreover,

x2  4
lim  lim (x  2)  4
x2 x2 x 2

We say that a function is continuous at x=a, when

 The value f(a) exists;

 The limit lim f(x) exists;

x a

 lim f(x)  f(a)

x a

For the continuity at any particular point we must check all three
presuppositions.

76
TOPIC 5: CALCULUS Christos Nikolaidis

Most of the known functions are continuous everywhere (since they

look like uninterrupted curves!). For example, lines, quadratics,
polynomials in general, exponentials are continuous functions.

Remember that at some x=a we may have different side limits

lim f(x) and lim f(x)

x a x a

If they are equal, say lim f(x) = lim f(x) = b, then we can say that
x a x a

limf(x) = b
xa

It is worthwhile to see the following examples of “step” functions to

further clarify the notions of limit and continuity.

EXAMPLE 1

2x  1, if x  2
Let f(x)  
7, if x  2

y
7

1
x
-3 -2 -1 1 2 3 4 5
-1

 limf(x)  5 [when x approaches 2,the value f(x) approaches 5];

x2

 f(2) =7.
 But lim f(x)  f(2)
x 2

Thus the function is not continuous at x=2.

77
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 2

Consider the function

y
7

x 2 , if x2 6
f(x)  
 5, if x2 5

1
x
-4 -3 -2 -1 1 2 3 4 5 6 7
-1

We can see that f(2)=4 but lim f(x) does not exist.
x 2

[In fact, only side limits exist: lim f(x)  4 and lim f(x)  5 ]
x 2 x 2

Therefore, the function is not continuous at x=2.

(thus the functions is not continuous in general).

EXAMPLE 3
Consider the function
y
7

x 2 , if x  2 6
f(x)  
 4, if x  2
5

1
x
-4 -3 -2 -1 1 2 3 4 5 6 7
-1

We can see that f(2)=4 and also lim f(x)  4 .

x 2

Since lim f(x)  f(2) the function is continuous at x=2.

x 2

(in fact the function is continuous everywhere).

78
TOPIC 5: CALCULUS Christos Nikolaidis

 THE FORMAL DEFINITION OF THE DERIVATIVE

Let y=f(x) be a continuous curve and A(x,f(x)) some point on it:

B
f (x+h)

A
f (x)
x
x x+h

We select a neighboring point B with

x-coordinate = x+h (where h is very small)
y-coordinate =f(x+h)

As we move from point A to point B, the rate of change is

Δy f(x  h)- f(x)


Δx h

If we let h become very small, that is h0, the result will be the
rate of change at point A, that is the derivative f (x) .

f(x  h)- f(x)

f (x) = lim
h 0 h

Let us apply this formula to the function f(x) = x2.

f(x  h)- f(x)
f (x)  lim
h 0 h

(x  h) 2  x 2
 lim
h 0 h

x 2  2xh  h 2  x 2
 lim
h 0 h

h(2x  h)
 lim
h 0 h

 lim (2x  h)  2x
h 0

79
TOPIC 5: CALCULUS Christos Nikolaidis

Therefore,

f (x)  2x

If we apply the definition for f(x)=xn we will find that

f(x  h)- f(x)
f (x)  lim  nx n-1 (exercise!)
h 0 h

EXAMPLE 4
Show from first principles (that is by using the formal definition),
that the derivative of the function
f(x)  x 3  2x
is
f (x)  3x 2  2 .
Solution
f(x  h)- f(x)
f (x)  lim
h 0 h

[(x  h)3  2(x  h)]  [ x 3  2x]

 lim
h 0 h

x 3  3x 2 h  3xh 2  h 3  2x  2h  x 3  2x
 lim
h 0 h

3x 2 h  3xh 2  h 3  2h
 lim
h 0 h

h(3x 2  3xh  h 2  2)
 lim
h 0 h

 lim(3x 2  3xh  h 2  2)
h 0

Now, we are able to set h=0 and obtain,

f (x)  3x 2  2

If f (x) exists we say that f(x) is differentiable at x. We also say that

f(x) differentiable if it f (x) exists for any point x of the domain.

80
TOPIC 5: CALCULUS Christos Nikolaidis

Notice
f differentiable at x  f continuous at x

The opposite is not necessarily true: that is, the function may be
continuous at some point x but not differentiable at this point.

Indeed, let us see again the step function of example 3

y
7

x 2 , if x  2 6
f(x)  
 4, if x  2 5

1
x
-4 -3 -2 -1 1 2 3 4 5 6 7
-1

The function is continuous at x=2 but it is not differentiable at x=2.

Informally speaking, this is because

There is a “corner” point: we cannot draw a tangent line at x=2

or otherwise,

The curve is continuous at x=2 but is not “smooth” at this point

But let us give a more formal proof.

Method 1 (without mentioning the definition from first principles)

At x=2 we find the “side” derivatives of f:

f- (x)  (x 2 )  2x , thus f- (2)  4

f (x)  (4)  0 , thus f  (2)  0

Since f- (2) and f (2) differ, f (2) does not exist.

81
TOPIC 5: CALCULUS Christos Nikolaidis

Method 2 (from first principles)

At x=2 we find the “side” derivatives of f:

f(2  h)- f(2) (2  h) 2 - 4 4  4h  h 2 - 4

f- (2)  lim-  lim-  lim-
h 0 h h 0 h h 0 h
4h  h 2
 lim-  lim- (4  h)  4
h 0 h h 0

while
f(2  h)- f(2) 4- 4
f  (2)  lim  lim  lim (0)  0
h 0 h h0 h h0

Since f- (2) and f (2) differ, f (2) does not exist.

Let us slightly modify this function:

EXAMPLE 5
 x2, if x  2
f(x)  
4x - 4, if x  2
We firstly show that f is continuous at x=2:

 lim f(x)  2 2  4
x 2

 lim f(x)  8  4  4
x 2 

 f(2)=4

Therefore lim f(x)  4  f(2) and the functions is continuous at x=2.

x2

We next show that f is differentiable at x=2:

 f- (x)  (x 2 )  2x , thus f- (2)  4

 f  (x)  (4x - 4)  4 , thus f  (2)  4

Therefore f (2)  4 and the functions is differentiable at x=2.

Notice: let’s also show that f  (2)  4 from first principles:

f(2  h)- f(2) [4(2  h)- 4]- 4 4h

f  (2)  lim  lim  lim ( )  lim (4)  4
h0 h h0 h h0 h h0

82
TOPIC 5: CALCULUS Christos Nikolaidis

We will see the same example in a different version:

EXAMPLE 5
 x2, if x  2

f(x)   a, if x  2
 bx  c, if x  2

Find a, b and c given that the function is (continuous and)
differentiable.
Solution
The function is continuous and differentiable for any x≠2.
We check continuity and differentiability at x=2.
Continuity:

 lim f(x)  2 2  4
x 2

 lim f(x)  2b  c
x 2 

 f(2)=a

Since f is continuous at x=2: 4  2b  c  a (1)

Differentiability:

 f- (x)  (x 2 )  2x , thus f- (2)  4

 f  (x)  (bx  c)  b , thus f (2)  b

Since f is differentiable at x=2: b4 (2)

(1) and (2) give

a=4, b=4, c=-4

83
TOPIC 5: CALCULUS Christos Nikolaidis

5.13 L’HÔPITAL’s RULE (for HL)

 A FIRST DISCUSSION
We know that
0 5
0 is not defined
5 0
However, when x tends to 0,
5
approaches either + or -
x
But what about
0
?
0

Consider a function of the form

f(x)
g(x)

f(x)→0 f(x) x- 5 0
if then tends to 0 e.g lim  0
g(x)→5 g(x) x 5 x 5

5
lim  +
f(x)→5 f(x) x 5 x- 5
if then tends to + or – e.g
g(x)→0 g(x) 5
lim  –
x 5 x- 5

But again, what happens when both f(x) and g(x) tend to 0?

The result could be anything: 0 or + or – or any real number!

0
For example, when x→0, all the functions below have the form ,
0
however

x3 x sinx 2sinx
lim  0, lim  , lim  1, lim 2 etc
x 0 x x 0 x 3 x 0 x x 0 x

(check these functions on your GDC near x=0).

0
That is why we say that is an indeterminate form.
0

Another indeterminate form is .


84
TOPIC 5: CALCULUS Christos Nikolaidis

f(x) 0 
 CASES WHERE IS NOT OF THE FORM or
g(x) 0 

Here we deal with limits of the form

f(x) f(x) f(x)

lim , lim , lim
x a g(x) x   g(x) x -  g(x)
0 
If the fraction is not of the form or we can easily find the
0 
result. For example,
1
5
lim
2x  3 5
 lim x 5 lim
x 3 0
 0
x 1 3x  5 8 x   1 7 x 3 2
x 3 9
7
x
x
2 e 2 0
lim 0 lim 0 [ and ]
x   3x  5 x -  x  3  

2x  7 2x  7  
lim   lim   [ and  ]
x   3 x   ex 3 0

2 2 2 -2
lim   lim   [ and  ]
x 3 (x  3) 2 x 3 (x  3) 2 0 
0

2 2 2 2
lim   lim   [ and  ]
x 3 x 3 x 3 x 3 0 
0

0 
 THE INDETERMINATE FORMS or
0 

In some cases the answer is easy. For example

 4x  7 4x  7 4x 2  7
: lim  2, lim  0, lim  
 x   2x  3 x   2x 2  3 x   2x  3

[remember the discussion about horizontal asymptotes]

0 x 2 - 3x x(x - 3)
: lim  lim  lim x  3
0 x 3 x- 3 x 3 x- 3 x 3

It is also known that

sinx 0
lim 1 (it is )
x 0 x 0

For more complicated cases the following theorem helps; we simply

need the derivatives of f(x) and g(x).

85
TOPIC 5: CALCULUS Christos Nikolaidis

L’ Hôpital’s rule:

f(x)→0 f(x)→±∞ f(x) f  (x)

If or lim  lim
g(x)→0 g(x)→±∞ g(x) g  (x)

f  (x)
provided that lim exists.
g  (x)

EXAMPLE 1

 
4x  7    (4x  7) 4
lim  lim  lim  2
x   2x  3 x   (2x  3) x   2

0 
   
0 
Remember to indicate the form above = :  or  .

Sometimes we need to apply L’Hôpital’s rule more than once.

EXAMPLE 2
 
2    
2x  4x  7  4x  4 4 2

lim 2
 lim  lim 
x   5x  3x  2 x   10x  3 x   10 5

EXAMPLE 3
0 0
   
ex  x  1  0  ex  1  0  ex 1
lim  lim  lim 
x 0 x2 x 0 2x x   2 2

EXAMPLE 4
3e 2x  2
Find the horizontal asymptotes of f(x) 
e 2x  1
 when x  

 
3e 2x  2    6e 2x 6
lim 2x  lim 2x
 lim  3 H.A. y=3
x   e 1 x   2e x   2


 when x - , the limit is not of the form since e 2x  0 .

3e 2x  2  2
lim   2 H.A. y=-2
x   e 2x  1 1

86
TOPIC 5: CALCULUS Christos Nikolaidis

 OTHER INDETERMINATE FORMS

The following are also indeterminate forms

0 , , 1

For example, look at the following 0   forms:

 1
lim x    lim 1  1
x 0  x  x 0

 1
lim x 2    lim x  0
x 0  x  x 0

 1  1
lim x  3   lim 2  
x 0  x  x 0 x

0 
More complicated forms can be transformed to the form or
0 
and thus answered by using L’Hôpital’s rule.

EXAMPLE 5
When x→0 then lnx→-∞. Thus lim (xlnx ) is of the form 0   .
x 0

But
1 
 
lnx 
lim (xlnx )  lim  lim x  lim (-x)  0
x 0 x 0 1 x 0 1 x 0
 2
x x

EXAMPLE 6
The limit lim
x  
 
x  1  2x is of the form    . But

lim  x  1  2x  lim   x  1  2x  x  1  2x 
x   x   x  1  2x

x  1- 2x 1- x
 lim  lim
x   x  1  2x x   x  1  2x

 
 -1
 lim  
x  1 1

2 x 1 2x

87
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 7
We will show that
x
 1
lim  1    e .
x   x
It has the indeterminate form 1  . But
x
x  1  1
xln  1 
 1 ln  1  
1    e 
x
e  x
 x
 1
Let’s find the limit of the exponent xln 1   ; it has the form 0   :
 x
1
 1 
 
 1  1   x2 
ln 1   0
  1  
 1 x 0 x 1
lim xln 1    lim   lim   lim 1
x    x x   1 x   1 x    1
 2 1  
x x  x
Hence
x  1
 1 xln  1 
lim  1   = lim e  x  =e1=e
x   x  x 

NOTICE
x
 a
 In a similar way we can show that lim  1   e .
a
x   x
sinx
 As we said earlier we accept the limit lim  1 as known.
x 0 x
0
Although it is of the form and the rule would give
0
0
 
sinx 0 (sinx) cosx
lim  lim  lim 1
x 0 x x 0 x x 0 1

we cannot use L’Hôpital ! This is because the proof of the fact

sinx   cosx (by first principles)

sinx
already uses the limit lim 1.
x 0 x

The proof of this limit in particular uses other trigonometric

techniques and it is beyond the scope of this course.

88
TOPIC 5: CALCULUS Christos Nikolaidis

5.14 IMPLICIT DIFFERENTIATION – MORE KINEMATICS (for HL)

We are very familiar with functions of the form y=f(x). However, in

many real applications, we are not given a clear function where y
is expressed in terms of x, but a more complicated relation which
involves x and y.

Consider for example the relation

x2+y2=1

Actually, this relation represents a circle of radius 1 on the

Cartesian plane:

-1 1

(i.e. the pairs (x,y) that satisfy this relation form that circle)

Obviously this is not the graph of a function (as a vertical line may
cross the graph at two points). However, if we solve for y we obtain

y2=1-x2  y=  1  x 2
that is, two different functions together:
y= 1  x 2 is the semicircle above the x-axis
y=  1  x 2 is the semicircle under the x-axis

The question is
dy
What is the derivative y΄= ?
dx

89
TOPIC 5: CALCULUS Christos Nikolaidis

 2x x
Case 1: If y= 1  x 2 the derivative is y΄ = =-
2 1  x2 1  x2
 2x x
Case 2: If y=  1  x 2 the derivative is y΄ =  =
2 1  x2 1  x2

x
We can observe that in both cases the result is equal to y΄=-
y

A more elegant way to obtain the same result is the following:

Consider again
x2+y2=1

Differentiate both sides with respect to x. Do have in mind though

that y is a function of x, so the derivative of y2 is not 2y but 2yy΄
(chain rule). Hence, the differentiation gives

x
2x+2yy΄= 0  2yy΄=-2x  y΄=-
y

Notice that it is not necessary to solve the initial equation for y

(sometimes this may be very difficult or even impossible!). We just
differentiate both sides with respect to x and then solve for y΄. This
process is known as implicit differentiation.

EXAMPLE 1
dy
Find y΄= if 2x 2  x 2 y3  y 2
dx
(notice that solving for y seems to be a nightmare!!!)
Solution
The implicit differentiation gives
4x+2xy3+x23y2y΄=2yy΄
Now we simply solve for y΄:
4x+2xy3=2y y΄-3x2y2y΄
 4x+2xy3=(2y-3x2y2)y΄
4x  2xy3
 y΄=
2y  3x 2 y 2

90
TOPIC 5: CALCULUS Christos Nikolaidis

Ok, you may complain that the result is not a clear expression of x
as usual but involves x and y as well!!! However this is not a
problem in general!

EXAMPLE 2
Consider again the equation of the circle
x2+y2=1
Find the tangent lines
1 3
a) at the point (x,y)= ( , )
2 2
(confirm that this point lies on the circle)
b) at the points of the circle with x=0
Solution
dy
First, we need the derivative y΄= :
dx
We have seen that
x
y΄=-
y
1 3 x 1/2 1 3
a) At (x,y)= ( , ) , it is m=y΄=- =- =- =-
2 2 y 3/2 3 3
Hence, the tangent line has the form
3
y =- x+c
3
But
3 1 3 3 3 4 3
- +c=  c= +  c=
3 2 2 2 6 3
So that the tangent line is
3 4 3
y=
x+
3 3
b) For x=0 we obtain two values for y:
x2+y2=1  y2=1  y=  1
 At (x,y)=(0,1), the gradient is m=y΄=0 and the tangent line is
y= 0x+c, or finally y=1
 At (x,y)=(0,-1), the gradient is m=y΄=0 and the tangent line is
y= 0x+c, or finally y=-1

91
TOPIC 5: CALCULUS Christos Nikolaidis

[Indeed, notice that A(0,1) and B(0,-1) are the highest and the
lowest points of the circle respectively

y
A y=1

x
-1 1
y=-1
B

Apparently, the tangent lines at those points are the horizontal

lines y=1 and y=-1]

EXAMPLE 3
Let
x2+y+xcosy=π.
Find the tangent and the normal lines at x=0.
Solution
Firstly, for x=0 we obtain y=π, so the given point is (0,π).
Implicit differentiation gives
2x  y΄  cosy  x( siny)y΄  0
 y΄  (xsiny)y΄  -2x  cosy
- 2x  cosy
 y΄ 
1  xsiny
At (x,y)=(0,π), the gradient is
m=y΄=1.
The tangent line has the form y=x+c. By using the point (0,π) we
find c=π and finally
y=x+π
The normal line has the form y=-x+c. By using the point (0,π) we
find c=π and finally
y=-x+π

92
TOPIC 5: CALCULUS Christos Nikolaidis

 MORE ON KINEMATICS

Remember the following relation

Displacement Velocity Acceleration

s v a
derivative

integral

In this situation, s,v,a are all given in terms of time t.

For example, given that

v=6t2
then
dv
a = =12t
dt
However, the velocity (v) is sometimes given as a function of s,
instead of t. For example, v=3s2

The following formula for acceleration may be derived (using the

chain rule)
dv dv ds dv
a   v
dt ds dt ds
Concentrate on
dv
a v
ds

EXAMPLE 4
Let v=3s2
Then
dv
a= v = (3s2)(6s)=18s3
ds

In fact, we use implicit differentiation on v=3s2 with respect to t:

dv ds
a= = (6s) = (6s) v=(6s) (3s2)=18s3
dt dt

93
TOPIC 5: CALCULUS Christos Nikolaidis

5.15 RATE OF CHANGE PROBLEMS (for HL)

In many applications we have quantities depending on time t. If A

is a function of t we know that
dA
= rate of change of A
dt

a) Suppose that two quantities A and B depend on each other. Any

change according to time in one of them implies a change in the
dA dB
other as well. The relation between and , the rates of
dt dt
change of the two quantities, can be found by implicit
differentiation with respect to time t.

For example
dA dB
If A=2B3 then  6B 2
dt dt

dA dB 1 dB
If A=2B+lnB then 2 
dt dt B dt

3 dA 3 dB
If sinA= then cosA  2
B dt B dt

b) Suppose now that three quantities A,B and C are related. The
dA dB dC
relation between the rates of change , and , can be
dt dt dt
found by implicit differentiation with respect to time t.

For example
dA dB dC
If A=2B3+4C2 then  6B 2  8C
dt dt dt

dA dB dC
If A=B2C3 then  2BC 3  3B 2 C 2
dt dt dt

dA dB dB dC
If A=B2+5BC then  2B  5C  5B
dt dt dt dt

dA dB 1 dC
If tanA=2B+lnC then sec 2 A 2 
dt dt C dt

94
TOPIC 5: CALCULUS Christos Nikolaidis

In problems involving rates of change we work as follows

Methodology
1. The problem usually refers to the rates of change of two
quantities (one rate is given, the other one is required).
Determine the quantities A and B
2. Find the relation between A and B
dA dB
3. Find the relation between and
dt dt
4. If the question refers to a specific moment (for particular
value of t, or A or B) the substitution takes place in end!
(in the relation of step 3, not in the relation of step 2)

Notice: We work similarly if three or more quantities are involved.

EXAMPLE 1
Consider an expanding square frame. If the side of the square
increases in rate 2ms-1 find the rate of change of its area, when
the side is 10m.
Solution
dx
x= side, =2m/sec
dt
x
dA
A= area, =?
dt
The relation between A and x is
A=x2
Hence,
dA dx dA
= 2x  = 4x
dt dt dt
Therefore, when x=10m
dA
=40m2/sec
dt x 10

95
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 2
Consider an expanding sphere. If the volume increases in rate
5cm3/sec find the rate of change of its radius r,
(i) when r = 3 cm
(ii) at the moment the sphere reaches the volume of 36π cm3
Solution
dV dr
=5cm3/sec, =?.
dt dt
The relation between V and r is given by
4
V= πr 3
3
Hence,
dV dr
=4πr2
dt dt
dr 5
 =
dt 4π r 2
(i) when r = 3

dr 5
= m/sec
dt r 3
36π
(ii) when V=36π, the original relation gives
4
36π= πr 3  r = 3cm.
3
Therefore, the answer is as in (i)

Have in mind that speed is also a rate of change. When a moving

body A has speed 5m/sec, that means that the distance x from
some fixed point O changes in rate 5m/sec

O x A
dx
If A is moving to the right, x increases so  5m/sec
dt
dx
If A is moving to the left, x decreases so  5m/sec
dt

96
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 3

Two cars, A and B, are traveling at 60 km/h and 70 km/h

respectively, on straight roads, as shown in the diagram below.

A
O x

y z

At a given instant both cars are 100 km away from O. Find, at

that instant, the rates of change

(a) of the distance between the cars.

ˆΒ
(b) of the angle θ  ΟΑ
Solution
dx dy dz dθ
 60km/h  70km/h ? ?
dt dt dt dt
π
When x  y  100 then z  100 2 and θ 
4
(a) Relation between x,y,z: z 2  x 2  y2
dz dx dy dz dx dy
Hence 2z  2x  2y z x y
dt dt dt dt dt dt
dz dz
 100 2  100(60)  100(70)   5 2km/h
dt dt
y
(b) Relation between x,y,θ : tanθ   y  xtanθ
x
dy dx dθ
Hence  tanθ  xsec 2 θ
dt dt dt
π π dθ
 -70  60tan  100sec 2
4 4 dt
dθ
  0.65 rad/h
dt

97
TOPIC 5: CALCULUS Christos Nikolaidis

Let us see a final example with three variables.

EXAMPLE 4
It is given that
1 2
A r h  2r 3
3
Find the rate of change of h when r=3 and h=6, under two
circumstances:
dA
a) h is always double of r and  30
dt
dA dh
b)  30 and 8
dt dt
Solution
a) Since h=2r, the original relation becomes
2 8
A  r 3  2r 3  r 3
3 3
Hence
dA dr
 8r
dt dt

Therefore, when r=3

dr dr 5
30  24  
dt dt 4

b) By implicit differentiation on the original relation we obtain

dA 2 dr 1 2 dh dr
 rh  r  6r 2
dt 3 dt 3 dt dt

Therefore, when r=3 and h=6,

dr dr
30  12  24  54
dt dt
dr
 6  66
dt
dr 1
 
dt 11

98
TOPIC 5: CALCULUS Christos Nikolaidis

5.16 FURTHER INTEGRATION BY SUBSTITUTION (for HL)

We have seen the simple case of substitution u=f(x), when the

derivative of f(x) is part of the integrand.

For example,

f  (x)
I=  dx = ln(f(x) + c
f(x)

Indeed,

Let u = f(x)
du du
Then  dx  f (x)  dx  f (x)
and

f  (x) du 1
I=    du  lnu  c = ln(f(x) + c
u f  (x) u

In this case we can omit the whole process and give directly the
result.

EXAMPLE 1
Find
cosx x
A=  dx , B=  dx
sinx  1 x 1
2

Solution
Since the derivative of u=sinx+1 is cosx
cosx
A=  dx  ln(sinx  1)
sinx  1

For B, the derivative of u=x2+1 is 2x.

f  (x)
We slightly modify the integral to obtain the form  f(x) dx
x 1 2x 1
B=  2
dx   2 dx  ln ( x 2  1)  c
x 1 2 x 1 2

99
TOPIC 5: CALCULUS Christos Nikolaidis

Similarly,

e
f(x)
 f  (x)dx = ef(x) + c

 sinf(x)  f  (x) dx =-cosf(x) + c

 cosf(x)  f  (x) dx =sinf(x) + c

f(x)n 1
 f(x)  f  (x) dx =
n
c
n1

etc.

EXAMPLE 2

 sinx e
cosx
dx = e cosx  c

2 3 1 3 1 3
x ex 5
dx =  3x 2 e x 5 dx = e x 5  c
3 3

1 1
 xsin(7x
2
 3)dx =  14xsin(7x 2  3)dx =  cos(7x 2  3)  c
14 14

5 2 5 5 (x 2  3)6 5
 5x(x  3) dx =
2 5
 2x(x  3) dx = c= (x 2  3)6  c
2 2 6 12
1
3 2 32 2 3/2
 3x x  3dx = 2  2x(x  3) 2 dx = 2 3 (x  3)  c = (x  3)  c
2 2 3/2

(lnx) 3 1 1 3 1 (lnx) 4 (lnx) 4

 2x dx =
2x
(lnx) dx =
2 4
 c =
8
c

Remark.
If you don’t feel confident to find directly the result, you may
always follow the detailed procedure of substitution.

100
TOPIC 5: CALCULUS Christos Nikolaidis

We are going to see two more cases of substitution.

CASE 1: We reuse u=f(x) to get rid of all remaining x’s.

Consider
I=  x 3 x 2  3dx
Let’s use the substitution
u=x2+3
(despite the fact that x3 is not the derivative of u!)
Then
du du
=2x  dx=
dx 2x
Thus,
du 1
I=  x 3 u = x
2
udu
2x 2

Well, the result still contains x2, but

u=x2+3x2=u-3

and thus
1
I=
2  (u  3) udu

Therefore,
1
2 
I = (u u  3 u )du
1 3 1
=  (u 2  3u 2 )du
2
1 2 52 2 3
= ( u  3 u 2 )+c
2 5 3
1 52 3
= u  u 2 +c
5
1 2 5 3
= (x  3) 2  (x 2  3) 2 +c
5

Characteristic examples of this case are integrals of the form

polynomial polynomial
 ax  b
dx ,  (ax  b) n
dx

where we let u=ax+b

101
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 3
x2
Find I=  dx
x2
Solution
Let u=x+2 (so that x=u-2)
du
Then = 1  dx=du
dx
Thus,
x2 (u - 2) 2 u 2  4u  4
I=  du =  du =  du
u u u
4 u2
=  (u  4  )du =  4u  4lnu  c
u 2
(x  2) 2
=  4(x  2)  4ln(x  2)  c
2

Another popular substitution of this kind is u=ex due to its simple

derivative. When you see rational expressions containing ex, think of
this substitution!

EXAMPLE 4
ex e 2x
Find I 1=  dx I 2=  dx
e 2x  4 ex  4
Solution
For both integrals:
du du
let u= ex, then = ex  dx= x
dx e
Then
u dx du 1 u 1 ex
I 1=  2
u  4 e x  u2  4 2
  arctan  c  arctan c
2 2 2
while
u 2 dx u u  4- 4 4
I 2=  x
 du   du   1  du
u4 e u4 u4 u4

 u  4ln | u  4 | c  e x  4ln | e x  4 | c

102
TOPIC 5: CALCULUS Christos Nikolaidis

 CASE 2: Let x=(expression of u) [instead of u=(expression of x)]

In this case, the substitution is not very obvious and it is usually

given in an exam!

We will see two characteristic substitutions of this kind.

EXAMPLE 5
dx x
For I=  2
, we use the substitution x=2tanu (so u= arctan )
x 4 2
dx
We have, = 2sec 2 u  dx= 2sec 2 u du
du
Thus,
2sec 2 udu 1 sec 2 udu 1 sec 2 udu 1
4tan 2 u  4 2  tan 2 u  1 2  sec 2 u
I=    =  du
2

1 1 x
= u  c = arctan  c
2 2 2

EXAMPLE 6
dx x
For I=  , we use the substitution x=2sinu (so u  arcsin )
4- x 2 2

dx
We have, =2cosu  dx=2cosudu
du
Thus,
2cosudu 2 cosudu cosu
I =  
cosu 
 = du
2
4 - 4sin u 2 2
1- sin u
x
= u  c = arcsin c
2

In general, the two formulas (of the formula booklet)

1 1 x 1 x
a 2
x 2
dx  arctan  c
a a  a2  x 2
dx  arcsin
a
c

can be shown by suing the substitutions x  atanu and x  asinu

respectively.

103
TOPIC 5: CALCULUS Christos Nikolaidis

We can see these two substitutions in similar cases:

see expression use substitution

a2  x 2 x  atanθ
a2  x 2 x  asinθ

EXAMPLE 7
Find I=  16- x 2 dx , by using the substitution x=4sinθ

Solution.
We have,

dx
=4cosθ  dx=4cosθdθ
du
Thus,

I=  16- 16sin 2θ 4cosθ d θ  16 1- sin 2 θ cos θdθ  16  cos 2θdθ

We use the double angle identity

cos2θ  1
cos 2θ 
2

Thus
 sin2θ 
I  8 (cos2θ  1)dθ =8   θ  +c =4sin2θ+8θ +c
 2 

But
x
θ = arcsin
4

and

4sin2θ = 8sinθcosθ = 2xcosθ

2 x2 x
= 2x 1- sin θ = 2x 1- = 16- x 2
16 2

Therefore,
x x
I= 16- x 2 +8arcsin +c
2 4

104
TOPIC 5: CALCULUS Christos Nikolaidis

Look at the three integrals below. Although of very similar form,

they have completely different results. The difference lies in the fact
that the three quadratics involved have two, exactly one, or no real
roots respectively.

EXAMPLE 8
Consider
2 2 2
I1   2
dx I2   2
dx I3   2
dx
x  4x  3 x  4x  4 x  4x  5

Hints:
2 1 1
For I1 it is given that 2
  . [easy to confirm!]
x  4x  3 x  3 x  1

For I2 notice that the quadratic is a perfect square.

For I3 use the vertex form of the quadratic: x 2  4x  5  ( x - 2) 2  1

Solution

 1 1 
I1     dx  ln x  3  ln x  1  c
 x 3 x 1
2 2
I2   2
dx   2(x - 2)- 2 dx   c
(x - 2) x- 2
2
I3   dx  2arctan(x  2)  c
(x - 2) 2  1

EXAMPLE 9
Use the vertex form of - x 2  4x  3 to find
2
I dx
- x 2  4x  3
Solution
The vertex form is –(x-2)2+1.

Thus
2
I dx  2arcsin(x  2)  c
1- (x - 2) 2

105
TOPIC 5: CALCULUS Christos Nikolaidis

5.17 INTEGRATION BY PARTS (for HL)

 DISCUSSION
In this paragraph we study integrals of the form

I=  (f  g )dx

Let’s make it clear from the very beginning that there is no

“product rule” for integrals. It is sometimes very difficult, and
more often impossible, to find the indefinite integral of a product.
However, we exploit the product rule for differentiation

(f  g)  f   g  f  g 
which gives
f   g  (f  g)  f  g 

If we apply integration on both sides we obtain the so-called

Integration by parts formula

 f 'gdx =f.g-  f  g' dx

This formula does not give an answer for any product but in some
cases the integral in the RHS is much easier than the original and
thus we obtain a result.

 THE METHOD
Consider
I=  xe x dx
Since (ex)΄=ex this integral can be expressed as
I=  (e x )xdx
The integration by parts formula gives

I = exx-  e x x' dx = exx-  e x dx = xex- ex+c

[you may easily confirm that the derivative of the result gives xex]

106
TOPIC 5: CALCULUS Christos Nikolaidis

In fact, the process is even quicker. We integrate one of the factors

and then we differentiate the other factor as follows

derivative

 f '  gdx = f . g -  f  g' dx

integral

For example, in I=  xe x dx we integrate e x and then differentiate x

 xe e
x x
dx = x ex - dx

and thus the final result is xex-ex+c

Notice
If we try to integrate the other factor, that is x , and then
differentiate e x , we obtain

x x2 x x2 x
 xe dx  2
e 
2
e dx

The result of course is not wrong, but it is not practical! The second
integral is worse than the original!

But how do we choose the function we integrate first?

We follow the priority list below

Priority (for integration)

1. e x , sinx, cosx
2. xn (or polynomials)
3. lnx, arctanx, arcsinx, arccosx

In the following easy examples we indicate by a double underscore

the factor we integrate and by a single underscore the function we
differentiate.

107
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 1

(a)  x cosxdx  xsinx   sinxdx  xsinx  cosx  c

e (2x  5)dx = ex( 2x  5 )-  2e x dx
x
(b)
= ex(2x+5)- 2ex + c
= 2xex + 3ex + c

x e x dx = x 2 ex-  2xe x dx
2
(a)
= x 2 ex- 2  x e x dx [we repeat the rule]
= x 2 ex- 2[ xe x   e x dx]
= x 2 ex- 2xe x +2 e x +c

1 3 1 1
(b) x
2
lnxdx = x lnx -  x 3 dx
3 3 x
1 3 1
=
3
x lnx-
3  x 2 dx
1 3 1 3
= x lnx- x +c
3 9

(e) I=  lnxdx = ?
Well, we do not see any product here! But this can be written as

 1 lnxdx
and 1=x0 is one of the factors mentioned in the priority list!
The integration by parts gives
1
I=  1 lnxdx =x lnx -  x  dx = xlnx-  dx = xlnx-x+c
x
1
[ Verification: the derivative of y=xlnx-x+c is y΄=lnx+x -1=lnx ]
x

In the following example we calculate the integral I   e x sinxdx .

Notice that both factors lie in the first priority. You may choose
any factor you like (I bet you will choose ex). The result is quite
interesting!

108
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 2
Find I =  e x sinxdx
Solution
We choose ez for integration:
I =  e x sinxdx =exsinx-  e x cosxdx [we carry on; choose again ex]

= exsinx- e x cosx   e x sinxdx 
= exsinx-excosx-  e x sinxdx
In the final expression we obtain again the original integral I, which
seems to lead to a dead-end. But in fact we have
I = exsinx+excosx- I
We solve for I and obtain
2I = exsinx+excosx
and finally
e x sinx  e x cosx
I= +c
2

 FURTHER OBSERVATIONS

In the priority list above we may have some variations of the

factors
 Instead of ex you may have eax, or any exponential ax
 Instead of sinx, cosx you may have sin(ax), cos(bx)
 Instead of lnx you may have any logarithm logax

EXAMPLE 3
1 2 3x 2
I =  x 2 e 3x dx  x e   xe 3x dx
3 3
1 2 1 1 
 x 2 e 3x   xe 3x   e 3x dx 
3 3 3 3 
1 2 2 3x
 x 2 e 3x  xe 3x  e c
3 9 27

109
TOPIC 5: CALCULUS Christos Nikolaidis

 CASES OF INTEGRATION BY PARTS

General Form
Examples Theoretical Questions
nZ+, a,bR
In   x n e x dx x Express In in terms of In-1
3 x 2
n
In,a   x e dx ax  x e dx ,  x e 2 dx Hence find I0 , I1 , I2 ,…
In   x n cosxdx 2
n x cosxdx Express In in terms of In-2
In   x sinxdx
In,a   x n cos(ax)dx
x
2
cos3xdx
n
In,a   x sin(ax)dx
lnx
Ia   x alnxdx  x lnxdx , x 5
dx Find a general formula for Ia

Ia,b   e ax sin(bx)dx
e
-x
sin2xdx Find a general formula for Ia,b
Ia,b   e ax cos(bx)dx
In   cosn xdx 2 3
Express In in terms of In-2
In   sin n xdx
 cos xdx ,  cos xdx
Hence find I2 , I4 and I3 , I5

Ia,b   sin(ax)cos(bx)dx  sin2xcos3xdx Find a general formula for Ia,b

In   x n arctanxdx
 arctanxdx ,  xarctanxdx ,  x
2
arctanxdx
In   x n arcsinxdx
2
In   x n arccosxdx  arcsinxdx ,  x arcsinxdx

 lnx 
n 2 3
In  dx  (lnx) dx ,  (lnx) dx

EXAMPLE 4
If In   cos n xdx , express In in terms of In- 2
In   cosxcos n-1 xdx  sinxcos n 1 x  (n  1) sin 2 xcos n- 2 xdx
 sinxcos n 1 x  (n  1)  (1 - cos 2 x)cos n- 2 xdx
 sinxcos n 1 x  (n  1)  (cos n- 2 x - cos n x)dx
 sinxcos n 1 x  (n  1)In- 2  (n  1)In
 In +(n-1) In  sinxcos n 1 x  (n  1)In- 2
 n In  sinxcos n 1 x  (n  1)In- 2
1 n 1
Thus In  sinxcos n 1 x  In- 2
n n

110
TOPIC 5: CALCULUS Christos Nikolaidis

 INTEGRATION BY PARTS FOR DEFINITE INTEGRALS

Again, as in the case of substitution, it would be safe to find the
indefinite integral first and then proceed to the definite integral!
Otherwise the integration by parts for definite integrals takes the
form

f '  gdx = f  g  a -
b b b
a  a
f  g' dx

EXAMPLE 5
Find
2
I=  e x (2x  5)dx
0

Solution
Method A: we find the indefinite integral first

e
x
(2x  5)dx =2xex+3ex+c [ example 1(b) above ]
Therefore,

I= 2xe x  3e x  2
0 =(4e2+3e2)-(0+3)=7e2-3

Method B: we keep the boundaries

2

I =  e x (2x  5)dx = e x  (2x  5) 0 -
0

2

2

0
e x  2dx

= (9e2-5) - 2e x  
2
0

=(9e2-5)- (2e2-2) = 7e2-3

111
TOPIC 5: CALCULUS Christos Nikolaidis

5.18 FURTHER AREAS BETWEEN CURVES - VOLUMES (for HL)

Consider the curve

y x

We found in paragraph 5.10 that

9 9
A x dx  18 and B   (3- x )dx  9
0 0

An alternative way to measure the area is to move in y-axis:

 consider the functions as x in terms of y

 let y move in y-axis (instead of x in x-axis).

Thus we have two formulas for the area

about x-axis about y-axis

b b
Area=  ydx Area=  xdy
a a

In our example above,

y  x  x  y2

and y ranges between 0 and 3:

3
3  y3 
B=  y dy     9  0  9
2

 3 0
0

3
3  y3 
A   (9- y )dy   9y 
2
  (27  9)  0  18
0
 3 0

112
TOPIC 5: CALCULUS Christos Nikolaidis

NOTICE
If y=f(x)  x=f-1(y) the alternative formula
b b
-1
Area=  xdy = f (y)dy
a a

can be written as
b
Area=  f -1 (x)dx
a

In other words, we can write the inverse function in terms of y

(followed by dy) or in terms of x as usual (followed by dx).

In our example above

3 3
 y 2 dy and  x 2 dx
0 0

are exactly the same.

EXAMPLE 1
Find the area among y=lnx, x-axis and the line x=e.

About x-axis:

A   lnxdx  xlnx  x 1  (elne  e)  (1ln1  1)  1

e e
1

About y-axis:
y=lnx x=ey
Thus
1

A   (e - e y )dy  ey  e y
0

1
0
 (e  e)  (0  1)  1

113
TOPIC 5: CALCULUS Christos Nikolaidis

 VOLUME OF REVOLUTION

a b

If we rotate the graph of y=f(x), 3600 about the x-axis, we will

obtain a 3-D shape as follows:

The volume of this shape is directly given by (the proof is omitted!)

b
V= π  y 2 dx
a

If we have two curves

y1=f1(x) and y2=f2(x)

with f1(x) ≥ f2(x) ≥ 0

the solid generated when the region between the two curves is
revolved 3600 about x-axis is given by

b 2 2
V= π  (y1 - y2 )dy
a

114
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 2
Consider the segment of the straight line
1
y= x , where 0  x  4 .
4
Find the volume of the cone generated by a 3600 rotation of this
segment.

b 4 1 2
V = π  y 2 dx = π  ( x) dx
a 0 4
4
π 4
2 π x3 
=
16  0
x dx =
16
 
 3 0
π 4 3 4π
= =
16 3 3

[Notice that the known formula for the volume of the cone of
height h=4 and radius r=1 gives:
1 4π
V= π r 2 h =
3 3
as expected!

If the rotation takes place about the vertical axis (y-axis), then we
solve y=f(x) for x, (hence x=f-1(y)), and the formula now is

b
V= π  x 2 dy
a

Notice that the solids generated when a curve is rotated in x-axis

or in y-axis are completely different.

115
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 3

If we rotate the parabola y=x2 about the y-axis we obtain the 3D

shape above! Find the volume if the height is 4.

We solve for x, so that x= y .

The volume is
4
b
2
4
2
4  y2 
V= π  x dy = π  x dy = π  ydy =π   =8π
 2 0
a 0 0

EXAMPLE 4
Consider the region between the curves y=2x2 and y=2x. The two
curves intersect at x=0 and x=1.

If the region is rotated 2π about x-axis, the volume generated is

1
1
2 4x3 x5   1 1  8π
V= π  (4x - 4x )dy = 4π    =4π    
0
3 5 0  3 5  15
If the region is rotated 2π about y-axis, the volume generated is
2
 y y2
2   y 2 y3   2 π
V= π    dy = π 
   =π 1   
 4 12  0
0 2 4 3 3
  

116
TOPIC 5: CALCULUS Christos Nikolaidis

5.19 DIFFERENTIAL EQUATIONS (for HL)

You already know some simple differential equations:

Find f(x) given that

f (x)  8x 3 , with f(1)  5

Or otherwise, express y in terms of x given that

dy
 8x 3 , with y=5 when x=1
dx

Clearly, by integration we find

y  2x 4  c

The condition y=5 when x=1 (or otherwise y(1)=5) gives

c=3

Thus the final answer is

y  2x 4  3

The terminology we use is as follows

dy
 8x 3 differential equation
dx

y(1)=5 boundary condition

y  2x 4  c general solution

y  2x 4  3 particular solution

Well, a differential equation relates a function y=f(x) with its

derivatives. Thus, it involves
x, y, y , y , etc.

Our task is to find the functions y=f(x) that satisfy this relation.

117
TOPIC 5: CALCULUS Christos Nikolaidis

For example,

xy  2y- y(x  2)  x 2  x

is a differential equation and

y  ex  x

is a particular solution, since

y  e x  1
y  e x
and

LHS = x(e x  1)  2e x  (e x  x)(x  2)

= xe x  x  2e x  xe x  2e x  x 2  2x
= x  x2
= RHS

Perhaps there are more solutions! As we said, our task is to find

the whole family of functions y=f(x) that satisfy the differential
equation.
This is a 2nd order differential equation as it involves the derivatives
up to y .
In general, the order of a D.E. is the number of the highest
derivative in the equation.

Here we only deal with 1st order D.E. that is our equations involve
only
dy
x, y and
dx
We will investigate only 3 particular cases of 1st order D.E.

 D.E. of separable variables

 Homogeneous D.E.
 Linear D.E. (with integrating factor)

We will also present a numerical method (approximation) for D.E.

 Euler’s method

118
TOPIC 5: CALCULUS Christos Nikolaidis

 D.E. OF SEPARABLE VARIABLES

A differential equation is of separable variables if we can separate

x’s from y’s and express the equation in the form
f(y)dy=g(x)dx
Then we integrate both parts and get the result

 f(y)dy   g(x)x

EXAMPLE 1
dy
Solve the D.E. =4xy2, given that y(1)=2.
dx
Solution
We separate the variables:
dy dy
=4xy2  =4xdx
dx y2
dy
 y 2
=  4xdx

1
  = 2x 2  c
y
1
 y = [general solution]
2x 2  c

The boundary condition gives

1 5
y(1)=2    2  -1=4+2c  c=-
2c 2
Therefore,
1
y
5
2x 2 
2
or equivalently
2
y [particular solution]
5 - 4x 2

dy
Even the simple example  8x 3 in the introduction can be seen
dx
as a D.E. of separable variables:

dy
 8x 3  dy=8x3dx   dy = 8x 3 dx  y  2x 4  c
dx

119
TOPIC 5: CALCULUS Christos Nikolaidis

The evaluation of the constant c can be done in an earlier step.

EXAMPLE 2
dy
Solve the D.E.  xy 2  x , given that y(0)=1.
dx
Solution
dy dy
 xy 2  x   x(y 2  1)
dx dx
dy
  xdx
y2  1

dy
 y 2
1
  xdx

x2
 arctany  c [general solution]
2
Since y(0)=1,
π
arctan1  c  c 
4

Therefore,
x2 π
arctany  
2 4
 x2 π 
 y  tan   [particular solution]
 2 4

A common problem in this category is to find a population P where

the rate of increase (or decrease) is proportional to the population
itself:
dP
 kP
dt
By separating variables

dP dP
 kdt     kdt  lnP  kt  c  P  e kt c  e kt e c
P P

By setting P0  e c (initial population) we obtain

P  P0 e kt

that is an exponential growth for the population in terms of time.

120
TOPIC 5: CALCULUS Christos Nikolaidis

 HOMOGENEOUS D.E.

A differential equation is called homogeneous if it can take the form

dy  y
= F  
dx x
y
[I.e. the RHS is a function of ]
x

It is not so difficult to recognize the homogeneous D.E.

For example,
2
dy x 2  xy  y 2 dy y y
 becomes  1    
dx x2 dx x x
3
dy x 2 y  y3 dy  y y
 becomes    
dx x3 dx x  x 
2
x 2  y2 y
1   
dy x 2  y2 dy x2 dy x
 becomes   
dx xy dx xy dx y
x2 x
3
x 3  2y 3  y
1  2 
dy x 3  2y 3 dy x3 dy x
 becomes   
dx x 2 y - 3x 3 dx x 2 y - 3x 3 dx y
-3
x3 x

Then we use the substitution

y
v  y  xv
x
dy dv
=v+x [by using product rule]
dx dx
dy  y
and the D.E. = F   takes the form
dx x
dv
v+x =F(v)
dx

which is always a D.E. of separable variables (v in terms of x)

[can you see why?].

121
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 3
Find the general solution of the D.E.
dy
x2  x 2  xy  y 2 .
dx
Solution
2
2 dy 2 2 dy x 2  xy  y 2 dy y y
x  x  xy  y   2
  1    
dx dx x dx x x

y
Let v  y  vx
x

dy dv
Then =v+x
dx dx

Thus

dv dv
v+x =1-v+v2  x =1-2v+v2
dx dx
dv
 x =(v-1)2
dx
dv dx
 2
=
(v - 1) x

dv dx
  (v - 1) 2
=
x
1
  = ln x +c
v- 1
1
 v- 1 = 
ln x  c
1
 v =1 
ln x  c
y
Finally, since v  ,
x

y 1  1 
=1   y= x  1 
x ln x  c  ln x  c 


122
TOPIC 5: CALCULUS Christos Nikolaidis

 1st ORDER LINEAR D.E. (WITH INTEGRATING FACTOR)

The last category contains the D.E. of the form

dy
 P(x)y  Q(x)
dx

where P(x) and Q(x) are functions of x only.

Methodology:

 we spot P(x) and Q(x)

we find the so called integrating factor   e 

P(x)dx
 (ignore +c)

 It holds y   Qdx

 We calculate the integral in the RHS and solve for y

We will demonstrate the procedure by using an easy example and

then we will explain why this method works!

EXAMPLE 4
Find the general solution of the D.E.
dy 2
 y  5x 2 .
dx x
Solution
2
P(x)  and Q(x)  5x 2 .
x

The integrating factor is ignore +c

2
 dx
  e
P(x)dx
 e x  e 2lnx  x 2
Then

y   Qdx  x2y   5x 4 dx  x 5  c

Therefore
c
y  x3 
x2

123
TOPIC 5: CALCULUS Christos Nikolaidis

Explanation:
dy
Differential equation:  P(x)y  Q(x)
dx

  e
P(x)dx
Integrating factor:

d  d   P(x)dx    P(x)dx 
Notice that  e   e P(x)  P(x)
dx dx    

If we multiply the D.E. by  we obtain

dy
  P(x)y  Q(x)
dx

But the LHS is the derivative of the product y [can you see why?]

Thus
d
(y)  Q(x)
dx
 y   Q(x)dx

Provided that the last integral is easy to find we can solve for y
and obtain the result.

Based on this explanation, let us provide a more analytical solution

for the D.E. in Example 4:
dy 2
 y  5x 2 .
dx x

The integrating factor is

2
 x dx
  e
P(x)dx
e  e 2lnx  x 2

We multiply the equation by x2:

dy d
x2  2xy  5x 4  (x 2 y)  5x 4
dx dx
Thus
c
x 2 y   5x 4 dx  x 5  c  y  x3 
x2

124
TOPIC 5: CALCULUS Christos Nikolaidis

 EULER’s METHOD: A NUMERICAL SOLUTION

Consider the D.E.

dy
 f(x, y) , y(x0)=y0
dx

where f(x,y) is an expression in terms of x and y.

By using a step h, we find consecutive points

(x0,y0), (x1,y1), (x2,y2), …

that approximate the solution.

We complete a table of the form

n xn yn
0 x0 y0
1
2

by using the recursive relations

xn+1 = xn + h
yn+1 = yn + hf(xn,yn)

We will demonstrate the procedure by using the D.E. of example 2

again and then we will explain the method!

EXAMPLE 5
dy
 xy 2  x , with y(0)=1.
dx
Find an approximation of y(1) using step h=0.2
Solution
The relation xn+1 = xn + h = xn + 0.2 gives directly the column of xn

n xn yn
0 0 1
1 0.2 1
2 0.4
2 0.6
4 0.8
5 1

125
TOPIC 5: CALCULUS Christos Nikolaidis

The relation
yn+1 = yn + hf(xn,yn) = yn + 0.2( x n yn2  x n )
gives
y1 = y0 + 0.2( x 0 y02  x 0 ) =1
y2 = y1 + 0.2( x 1 y12  x 1 ) =1.08
y3 = y2 + 0.2( x 2 y22  x 2 ) =
etc
We finally obtain
n xn yn
0 0 1
1 0.2 1
2 0.4 1.08
2 0.6 1.25331
4 0.8 1.56181
5 1 2.11209

Hence
for x=1, y  2.11209

Notice
 We can easily obtain the table above by using recursion in your
GDC.
For Casio FX we use

MENU
RECURSION
SET[F5]
Start=0,
End =100 (or more),
a 0= 0,
b0=1
EXIT
an+1 = an + 0.2
bn+1 = bn + 0.2( an bn2  an )
EXE

126
TOPIC 5: CALCULUS Christos Nikolaidis

 x2 π 
 The exact solution of the D.E. is y  tan   [see example 2]
 2 4

Thus for x=1, the actual value of y is

1 π
y= tan     3.40822
2 4

The approximation y  2.11209 we found above is not great!

However, we can improve the result by reducing the step value.

For h=0.01, (thus we need 100 steps) the GDC gives

for x=1, y  3.26945

which is much better.

Explanation of the method:

dy
 f(x, y) , y(x0)=y0
dx

actual value

approximation

The tangent line at (x0,y0) of the unknown solution curve is

y  y0  m(x  x 0 )
dy
But m   f(x 0 , y0 ) , so
dx (x
0 , y0 )

y  y 0  f(x 0 , y0 )(x  x 0 )  y  y0  (x - x 0 )f(x 0 , y0 )

For x=x1, we let h=x1-x0 and approximate the actual value of y by

y1  y0  hf(x 0 , y0 )
Similarly we get
y2  y1  hf(x 1 , y1 )
and so on.

127
TOPIC 5: CALCULUS Christos Nikolaidis

5.20 MACLAURIN SERIES – EXTN OF BINOMIAL THM (for HL)

Consider the infinite geometric series

1  x  x2  x3 …

We know that the series converges for -1<x<1 and the result is
1
S 
1x

In this paragraph we have the opposite task:

1
We are given a function, say f(x)  , and we wish to express it
1x
as an infinite series (which is called power series) of the form

a0  a1 x  a2 x 2  a3 x 3  …

(the power series looks like a polynomial of “infinite degree”).

 THE MACLAURIN SERIES

Suppose that a function f(x) has derivatives of every order near 0.
Then f(x) can be expressed as a power series as follows

f (0) 2 f (0) 3 
f (n) (0) n
f(x)  f(0)  f (0)x 
2!
x 
3!
x … [ =  n! x ]
n 0

This is known as Maclaurin series of the function.

1
For example, if f(x)  then
1x

f (n) (x) f (n) (0)

f(x)  (1  x)-1 1
f (x)  (1  x)-2 1
f (x)  2(1  x)-3 2
f (x)  6(1  x)-4 3!
etc

and thus
f(x)  1  x  x 2  x 3  …

It is in fact the geometric series in the beginning. Amazing, isn’t it?

128
TOPIC 5: CALCULUS Christos Nikolaidis

Explanation of the Maclaurin series:

We wish to express f(x) as

f(x)  a 0  a1 x  a 2 x 2  a3 x 3  a 4 x 4  …

Then
f (x)  a1  2a 2 x  3a 3 x 2  4a 4 x 3 …

f (x)  2a 2  3! a3 x  (3)(4)a 4 x 2  …

f (x)  3! a3  4! a 4 x  … etc

Therefore,

f(0)  a 0  a 0  f(0)
f (0)  a1  a1  f (0)
f (0)
f (0)  2a 2  a 2 
2
f (0)
f (0)  3!a 2  a3  etc
3!
In general
f (n) (0)
an 
n!
In fact, the partial sums of the Maclaurin series give good
approximations of f(x) near x=0:
a 0  a1 x is the tangent line of f(x) at x=0
a0  a1 x  a2 x 2 is the “best” quadratic that approximates f(x)
a0  a1 x  a2 x 2  a3 x 3 is the “best” cubic that approximates f(x)

1
For f(x)  (black curve) look at the approximations below:
1x

y  1  x [tangent] y  1  x  x2 y  1  x  x 2  x3

129
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 1
Find the Maclaurin series of the function f(x)=sinx up to the term in x5
Solution
f (n) (x) f (n) (0)
f(x) =sinx 0
f (x)  cosx 1
f (x)  -sinx 0
f (x)  -cosx -1
f (4) (x)  sinx 0
f (5) (x)  cosx 1

and thus

f (0) 2 f (0) 3
f(x)  f(0)  f (0)x  x  x …
2! 3!
gives
x3 x5
sinx  x   …
3! 5!

Notice: Use your GDC to compare the graph of sinx with the graphs
of the partial sums
x3 x3 x5
x , x  , etc
3! 3! 5!
The result is amazing!

Similarly we can obtain

x 2 x3
ex  1  x   …
2! 3!
x2 x4
cosx  1   …
2! 4!
x2 x3
ln(1  x)  x   …
2 3
x3 x5
arctanx  x   …
3 5

130
TOPIC 5: CALCULUS Christos Nikolaidis

Apart from the classical way to obtain the Maclaurin series we can
also apply simple substitution, products, differentiation and
integration on the already known series.

EXAMPLE 2
2
Find the Maclaurin series of the function f(x)  e x
Solution

Since

x x 2 x3
e 1x   …
2! 3!
we can replace x with x2 and obtain

2 x 4 x6
ex  1  x 2   …
2! 3!

EXAMPLE 3
Find the Maclaurin series of the function f(x)  e x sinx
Solution

We can multiply the series for e x and sinx.

 x 2 x3  x3 x5 
e x sinx   1  x    …  x    …
 2! 3!  3! 5! 

We can find gradually the constant term, the terms in x, the terms
in x2 etc:

x3 x3 x 4 x 4
e x sinx  x  x 2     …
3! 2! 3! 3!
x3
 x  x2  … [there is no x4]
3

Notice also

 if we differentiate the series for sinx (term by term) we obtain

the series of cosx.

 If we differentiate the series for ex we obtain ex itself.

131
TOPIC 5: CALCULUS Christos Nikolaidis

In general, we are able to differentiate a series term by term.

EXAMPLE 4
1
Find the Maclaurin series of the function f(x)  2
x 1
Solution
x3 x5
arctanx  x   …
3 5
1
Since (arctanx)  2
x 1

1  x3 x5 
2
  x    …  1  x 2  x 4  …
x 1  3 5 

We can also integrate term by term. It is more convenient to

consider definite integrals.

EXAMPLE 5
Find the Maclaurin series of cosx by integrating the series of
x3 x5
sinx  x   …
3! 5!
Solution

(a) By using indefinite integrals:

 x3 x5   x2 x4 
 sinxdx    3! 5! 
 x    … dx  - cosx   
 2! 4!
 …  c

For x=0 we obtain -1=c. Thus

 x2 x4  x2 x4
- cosx     …  1  cosx  1   …
 2! 4!  2! 4!

(b) By using definite integrals from 0 to x:

x x x
 x3 x5  x2 x4 
0 sinxdx  0  3! 5! 
 x    … dx  
- cosx x
0     …
 2! 4! 0
 x2 x4  x2 x4
 - cosx  1     …  0  cosx  1   …
 2! 4!  2! 4!

132
TOPIC 5: CALCULUS Christos Nikolaidis

 DIFFERENTIAL EQUATIONS AND MACLAURIN SERIES

Consider a differential equation of the form

dy
 F(x, y) with boundary condition y(0)=y0
dx

The analytical solution is not always easy or sometimes not possible

at all. However, we can easily find the Maclaurin series of the
solution y=f(x) (and thus a good approximation of the solution).

f (0) 2 f (0) 3
f(x)  f(0)  f (0)x  x  x …
2! 3!

But, f(0)= y0 [by the boundary condition]

dy
f (0) =  F(0, a) [by the D.E. itself]
dx x  0
ya

dy d2y
Implicit differentiation on gives and thus f (0) , and so on.
dx dx 2

EXAMPLE 6
Find the Maclaurin series up to x2 for the solution of the D.E.
dy
 x 2  y2 with y=3 when x=0
dx
Solution
Clearly f(0)=3
dy
f (0) =  0 2  32  9
dx x 0
y 3

d2y dy d2y
 2x  2y , thus f (0) =  2  0  2  3  9  54
dx 2 dx dx 2 x 0
y a

Therefore
54 2
x  …  3  9x  27x 2  …
y  3  9x 
2!
d2y
Notice. We can also express in terms of x and y only:
dx 2
d2y dy
2
 2x  2y  2x  2y(x 2  y 2 ) , thus f (0) =54
dx dx

133
TOPIC 5: CALCULUS Christos Nikolaidis

 THE EXTENTION OF THE BINOMIAL THEOREM

Remember that the binomial theorem gives

 n  n  n n 
(1  x) n =   +   x +   x 2  ⋯    x n
 0  1 2 n 

If we expand more the coefficients we obtain

n(n - 1) 2 n(n - 1)(n - 2) 3

(1  x) n = 1 +nx+ x + x ⋯+ xn
2 3!

This version allows us to use negative or even fractional values for

index n.

Thus for example,

- n(-n - 1) 2 - n(-n - 1)(-n - 2) 3

(1  x)-n = 1- nx + x + x ⋯
2 3!
but now we have infinitely many terms.

This is an alternative way to obtain some infinite power series.

For example
1
 (1  x) 1  1  x  x 2  x 3  …
1 x

Similarly,
1
 (1  x) 1  1  x  x 2  x 3  …
1 x

EXAMPLE 7
x
Find the Maclaurin series up to x4 for f(x) 
(1  x)3
Solution
 (-3)(-4) 2 (-3)(-4)(-5) 3 
f(x)  x(1  x)-3  x  1  3x  x  x  ⋯
 2 3! 
and finally
f(x)  x  3x 2  6x 3  10x 4  ⋯

134
TOPIC 5: CALCULUS Christos Nikolaidis

EXAMPLE 8
Find the Maclaurin series up to x3 for f(x)  1  x by using the
extension of the binomial theorem.
Solution
The formula
n(n - 1) 2 n(n - 1)(n - 2) 3
(1  x) n = 1 + nx + x + x ⋯
2 3!
gives
1 1 1 1 3
1 (- ) (- )(- )
1
f(x)  (1  x) 2
= 1 + x + 2 2 x2 + 2 2 2 x3  ⋯
2 2 6
and finally
1 1 1 3
f(x)  1 + x- x2 + x ⋯
2 8 16

EXAMPLE 9
Find the Maclaurin series up to x2 for f(x)  2x  3
2
by using the
extension of the binomial theorem.
Solution
2 2
  2x  1 2x 
f(x)  2x  3
2
 3 1    1  
  3  9 3 
The formula
n(n - 1) 2
(1  x) n = 1 + nx + x ⋯
2
gives
2
1  2x (-2)(-3)  2x  
f(x)  12     ⋯
9  3 2  3  

and finally
1 4 4 2
f(x)   x x ⋯
9 27 27

135