GRE 321: ENGIMEERING MATHEMATICS IV

COURSE CONTENT
1. Fourier series
- Euler coefficients
- Even and odd functions
- Sine and cosine functions with simple applications
2. Gamma, beta and probability functions
3. Simple linear programming
4. Elements of functions of several variables
5. Legendre and Bassel functions and their properties
6. Complex variables
- Differentiation and integration of complex functions
- Cauchy-Rieman equations and related theorems
7. Laplace and Fourier transforms
8. Element of probability, density and distribution functions
9. Statistics
- Regression and correlation
- Expectations and large sampling
- Test hypothesis and quality and quality control

FOURIER SERIES
Fourier series is a periodic phenomenon used to represent a periodic function as an infinite
trigonometrical series in a sine and cosine terms. Fourier series is applied in various
engineering fields and constitute a very important tools in solving problems that involve
Ordinary Differential Equations and Partial Differential Equations (ODEs and PDEs) related
to boundary and initial value problems of mechanics, heat flow and other fields.
Periodic Function
A function f(x) is said to be periodic if its function values repeat at regular intervals of the
independent variable. The regular interval between repetitions is the period of the
oscillations.
Example
1. The graph of y = sin x; where the values of y increases as x increases from 0 – 3600

The period is 3600 or 2π radians and the amplitude (maximum displacement from the
position of rest) is 1
2. Cosine graph of y = cos x is
A function f(x) is called a periodic function if f(x) is defined for all real x, and if there are
some positive number P, called a period of function of f(x), such that
𝑓(𝑥) = 𝑓(𝑥 + 𝑝), 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑟𝑒𝑎𝑙 𝑥
Such that for any integer n = 1, 2, 3, …
𝑓(𝑥) = 𝑓(𝑥 + 𝑛𝑝), 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑟𝑒𝑎𝑙 𝑥
Also, if two functions f(x) and g(x) are defined on the interval 𝑎 ≤ 𝑥 ≤ 𝑏 and have a period
p, then af(x) + bg(x) with any constant a and b also has the period p and
𝑎
∫ 𝑓(𝑥). 𝑔(𝑥) = 0
𝑏
Then, we say the two functions are orthogonal to each other on the interval 𝑎 ≤ 𝑥 ≤ 𝑏
The periodic functions of sin nx and cos nx where n = 0, 1, 2, 3, …. Forms an infinite collection
of periodic functions that are mutually orthogonal on the interval −𝜋 ≤ 𝑥 ≤ 𝜋and indeed on
any other interval of width 2π.
That is the periodic expansion of functions f(x) of period 2π of 1, cos x, sin x, cos 2x, sin 2x,
… cos nx, sin nx, gives the trigonometric series of the form
𝑎0 + 𝑎1 cos 𝑥 + 𝑏1 sin 𝑥 + 𝑎2 cos 2𝑥 + 𝑏2 sin 2𝑥 + ⋯
Giving
∞

𝑎0 + ∑(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥)

𝑛=1
Where 𝑎0 , 𝑎1 , 𝑏1 , 𝑎2 , 𝑏2 , … are constants, called the coefficients of the series. Each of the terms
has the period of 2π, thus, if the coefficients are such that the series converges, its sum will
be a function of period 2π.

Fourier series
The Fourier series of f(x) is
∞

𝑓(𝑥) = 𝑎0 + ∑(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥)

𝑛=1
Where 𝑎0 , 𝑎𝑛 , 𝑏𝑛 are constants called Fourier constant/coefficients of f(x) given by Euler
formulas

• Euler coefficients
1 𝜋
𝑎0 = ∫ 𝑓(𝑥)𝑑𝑥
2𝜋 −𝜋
1 𝜋
𝑎𝑛 = ∫ 𝑓(𝑥)𝑐𝑜𝑠𝑛𝑥𝑑𝑥
𝜋 −𝜋
1 𝜋
𝑏𝑛 = ∫ 𝑓(𝑥)𝑠𝑖𝑛𝑛𝑥𝑑𝑥
𝜋 −𝜋
For uniform notation, the above equations can be presented as
 ∞
𝑎0
𝑓(𝑥) = + ∑(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥)
2
𝑛=1
Where 𝑎0 , 𝑎𝑛 , 𝑏𝑛 are constants called Fourier constant/coefficients of f(x) given by Euler
formulas

• Euler coefficients
1 𝜋
𝑎0 = ∫ 𝑓(𝑥)𝑑𝑥
𝜋 −𝜋
1 𝜋
𝑎𝑛 = ∫ 𝑓(𝑥)𝑐𝑜𝑠𝑛𝑥𝑑𝑥
𝜋 −𝜋
1 𝜋
𝑏𝑛 = ∫ 𝑓(𝑥)𝑠𝑖𝑛𝑛𝑥𝑑𝑥
𝜋 −𝜋

• Even and odd functions

• Even Function
A function f(x) is aid to be even if
𝑓(−𝑥) = 𝑓(𝑥)
That is the function value for a particular negative value of x is the same with the
corresponding positive value of x. the graph of an even function is therefore symmetrical
about the y-axis. The Fourier series of even functions has only cosine terms

Even Function graphs symmetrical to y-axis

• Odd Function
A function f(x) is aid to be odd if
𝑓(−𝑥) = −𝑓(𝑥)
That is the function value for a particular negative value of x is numerically equal to the
corresponding positive value of x. the graph of an even function is therefore symmetrical
about the origin. The Fourier series of even functions has only sine terms

Odd Function graphs symmetrical to the origin

• Even and Odd Function

Functions exists which are neither even nor odd, those contain both sine and cosine terms
SINE AND COSINE FUNCTIONS WITH SIMPLE APPLICATIONS

• Functions with Period 2π

Examples
1. If the complimentary side of a sinusoidal voltage is represented by the wave form below,
establish the Fourier series to define the progression
f(x)

– 2π –π 0 π 2π x
Solution
To determine the working functions,
𝐹𝑜𝑟 𝑡ℎ𝑒 𝑟𝑎𝑛𝑔𝑒 − 𝜋 ≤ 𝑥 ≤ 0, 𝑓(𝑥) = 2
𝐹𝑜𝑟 𝑡ℎ𝑒 𝑟𝑎𝑛𝑔𝑒 0 ≤ 𝑥 ≤ 𝜋,
𝑑𝑦 𝑦2 − 𝑦1 0 − 2 2
= = =
𝑑𝑥 𝑥2 − 𝑥1 0 − 𝜋 𝜋
2
𝑖𝑒 𝑑𝑦 = 𝑑𝑥
𝜋
2
∴𝑦 = 𝑥+𝑐
𝜋
Choosing any of the boundaries, c = 0
2
∴ 𝑦 = 𝑥, 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑟𝑎𝑛𝑔𝑒 0 ≤ 𝑥 ≤ 𝜋
𝜋
Thus,
2, −𝜋 ≤ 𝑥 ≤ 0
𝑓(𝑥) = {2
𝑥, 0 ≤ 𝑥 ≤ 𝜋
𝜋

Using the Fourier series equation

∞
𝑎0
𝑓(𝑥) = + ∑(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥)
2
𝑛=1
We determine the Fourier coefficients from
1 𝜋
𝑎0 =
∫ 𝑓(𝑥)𝑑𝑥
𝜋 −𝜋
1 0 𝜋
2
= [∫ 2𝑑𝑥 + ∫ 𝑥𝑑𝑥 ]
𝜋 −𝜋 0 𝜋
𝜋
1 𝑥2
{[2𝑥]0−𝜋 + [ ] }
𝜋 𝜋 0
1 1
[2{0 − (−𝜋)}] + [𝜋 2 ]
𝜋 𝜋
1 3𝜋
[2𝜋 + 𝜋] =
𝜋 𝜋
𝑎0 = 3
 1 𝜋
𝑎𝑛 =∫ 𝑓(𝑥)𝑐𝑜𝑠𝑛𝑥𝑑𝑥
𝜋 −𝜋
0
1 2 𝜋
= [2 ∫ 𝑐𝑜𝑠𝑛𝑥𝑑𝑥 + ∫ 𝑥𝑐𝑜𝑠𝑛𝑥𝑑𝑥 ]
𝜋 −𝜋 𝜋 0
𝑈𝑠𝑖𝑛𝑔 𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑡𝑖𝑜𝑛 𝑏𝑦 𝑝𝑎𝑟𝑡𝑠 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑠𝑒𝑐𝑜𝑛𝑑 𝑝𝑎𝑟𝑡, ∫ 𝑢𝑑𝑣 = 𝑢𝑣 − ∫ 𝑣𝑑𝑢
𝑠𝑖𝑛𝑛𝑥
𝐿𝑒𝑡 𝑢 = 𝑥, 𝑑𝑢 = 𝑑𝑥 𝑎𝑛𝑑 𝑑𝑣 = 𝑐𝑜𝑠𝑛𝑥𝜋𝑑𝑥, 𝑣 =
𝑛
1 1 2 𝑠𝑖𝑛𝑛𝑥 𝜋 1 𝜋
∴ 𝑎𝑛 = {2 |𝑠𝑖𝑛𝑛𝑥|0−𝜋 + [|𝑥 | − ∫ 𝑠𝑖𝑛𝑛𝑥𝑑𝑥 ]}
𝜋 𝑛 𝜋 𝑛 0 𝑛 0
1 1 2 𝑠𝑖𝑛𝑛𝑥 𝜋 1
∴ 𝑎𝑛 = {2 |𝑠𝑖𝑛𝑛𝑥|0−𝜋 + [|𝑥 | + |𝑐𝑜𝑠𝑛𝑥|𝜋0 ]}
𝜋 𝑛 𝜋 𝑛 0 𝑛2

𝐴𝑙𝑙 𝑡ℎ𝑒 𝑠𝑖𝑛𝑛𝜋 𝑎𝑛𝑑 sin 0 = 0, 𝑡ℎ𝑒 𝑓𝑖𝑟𝑠𝑡 𝑎𝑛𝑑 𝑠𝑒𝑐𝑜𝑛𝑑 𝑡𝑒𝑟𝑚𝑠 𝑎𝑟𝑒 = 0
2
∴ 𝑎𝑛 = 2 2 [𝑐𝑜𝑠𝑛𝜋 − 1]
𝑛 𝜋
𝑇𝑎𝑘𝑖𝑛𝑔 𝑡ℎ𝑒 𝑣𝑎𝑙𝑢𝑒𝑠 𝑜𝑓 𝑛 𝑓𝑟𝑜𝑚 0 … . ∞
4
− 2 2 , 𝑓𝑜𝑟 𝑛 𝑜𝑑𝑑
𝑎𝑛 = { 𝑛 𝜋
0, 𝑓𝑜𝑟 𝑛 𝑒𝑣𝑒𝑛

1 𝜋
𝑏𝑛 =
∫ 𝑓(𝑥)𝑠𝑖𝑛𝑛𝑥𝑑𝑥
𝜋 −𝜋
0
1 2 𝜋
= [2 ∫ 𝑠𝑖𝑛𝑛𝑥𝑑𝑥 + ∫ 𝑥𝑠𝑖𝑛𝑛𝑥𝑑𝑥]
𝜋 −𝜋 𝜋 0
𝑈𝑠𝑖𝑛𝑔 𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑡𝑖𝑜𝑛 𝑏𝑦 𝑝𝑎𝑟𝑡𝑠 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑠𝑒𝑐𝑜𝑛𝑑 𝑝𝑎𝑟𝑡, ∫ 𝑢𝑑𝑣 = 𝑢𝑣 − ∫ 𝑣𝑑𝑢
𝑐𝑜𝑠𝑛𝑥
𝐿𝑒𝑡 𝑢 = 𝑥, 𝑑𝑢 = 𝑑𝑥 𝑎𝑛𝑑 𝑑𝑣 = 𝑠𝑖𝑛𝑛𝑥𝜋𝑑𝑥, 𝑣 = −
𝑛
1 1 2 𝑐𝑜𝑠𝑛𝑥 𝜋 1 𝜋
∴ 𝑏𝑛 = {−2 |𝑐𝑜𝑠𝑛𝑥|0−𝜋 + [−𝑥 | | + ∫ 𝑐𝑜𝑠𝑛𝑥𝑑𝑥 ]}
𝜋 𝑛 𝜋 𝑛 0 𝑛 0
1 1 2 𝑐𝑜𝑠𝑛𝑥 𝜋 1
∴ 𝑏𝑛 = {−2 |𝑐𝑜𝑠𝑛𝑥|0−𝜋 + [−𝑥 | | + |𝑠𝑖𝑛𝑛𝑥|𝜋0 ]}
𝜋 𝑛 𝜋 𝑛 0 𝑛2
𝐴𝑙𝑙 𝑡ℎ𝑒 𝑠𝑖𝑛𝑛𝜋 𝑎𝑛𝑑 sin 0 = 0, 𝑡ℎ𝑒 𝑡ℎ𝑖𝑟𝑑 𝑡𝑒𝑟𝑚 = 0
1 −2 2𝜋
∴ 𝑏𝑛 = { [𝑐𝑜𝑠0 − {cos(−𝑛𝜋)}] − [𝑐𝑜𝑠𝑛𝜋 − 𝑐𝑜𝑠0]}
𝜋 𝑛 𝑛𝜋
1 −2 2
∴ 𝑏𝑛 = { [𝑐𝑜𝑠0 − {cos(−𝑛𝜋)}] − [𝑐𝑜𝑠𝑛𝜋 − 𝑐𝑜𝑠0]}
𝜋 𝑛 𝑛
1 −2 2
∴ 𝑏𝑛 = { [1 + cosn 𝜋] − [𝑐𝑜𝑠𝑛𝜋 − 1]}
𝜋 𝑛 𝑛
−2
∴ 𝑏𝑛 = {[1 + cosn 𝜋] + [𝑐𝑜𝑠𝑛𝜋 − 1]}
𝜋𝑛
−2
∴ 𝑏𝑛 = (2𝑐𝑜𝑠𝑛𝜋)
𝜋𝑛
𝑐𝑜𝑠𝑛𝜋 = −1
4
∴ 𝑏𝑛 = −
𝑛𝜋
The Fourier series equation is given as:
3 4 1 1 4 1 1 1
𝑓(𝑥) = − 2 {𝑐𝑜𝑠𝑥 + 𝑐𝑜𝑠3𝑥 + 𝑐𝑜𝑠5𝑥 + ⋯ } − {𝑠𝑖𝑛𝑥 + 𝑠𝑖𝑛2𝑥 + 𝑠𝑖𝑛3𝑥 + 𝑠𝑖𝑛4𝑥 + ⋯ }
2 𝜋 9 25 𝜋 2 3 4
2. A mechanical vibration is represented by the periodic function below
[f(x) = f(x + 2π)]
y
π

– 2π –π 0 π 2π x
Use Fourier series to solve the pattern of the function

𝑑𝑦 𝑦2 − 𝑦1 0−𝜋
= =
𝑑𝑥 𝑥2 − 𝑥1 0 − (−𝜋)

𝑦 = −𝑥 + 𝑐
𝑦1 = 𝑥1 + 𝑐; 0 = 0 + 𝑐, 𝑐 = 0

−𝑥; −𝜋 < 𝑥 < 0

𝑓(𝑥) = { ; 𝑓(𝑥 + 2𝜋) = 𝑓(𝑥)
0; 0 < 𝑥 < 𝜋
Using the Fourier series equation as
𝑎
𝑓(𝑡) = 20 + ∑∞
𝑛=1(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥) We determine the Fourier coefficients from

1 𝜋 0 𝜋 0
1 20
𝑎0 = ∫ 𝑓(𝑥)𝑑𝑥 = ∫ −𝑥𝑑𝑥 + ∫ 0𝑑𝑥 = ∫ −𝑥𝑑𝑥 = − 𝑥 ]−𝜋
𝜋 −𝜋 −𝜋 0 −𝜋 2𝜋
1 𝜋2 𝜋
= − [0 − (−𝑥)2 ] = =
2𝜋 2𝜋 2
1 𝜋
𝑎𝑛 = ∫ 𝑓(𝑥)𝑐𝑜𝑠𝑛𝑥𝑑𝑥
𝜋 −𝜋
1 0
− ∫ −𝑥𝑐𝑜𝑠𝑛𝑥𝑑𝑥𝑑𝑥
𝜋 −𝜋
𝑈𝑠𝑖𝑛𝑔 𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑡𝑖𝑜𝑛 𝑏𝑦 𝑝𝑎𝑟𝑡𝑠, ∫ 𝑢𝑑𝑣 = 𝑢𝑣 − ∫ 𝑣𝑑𝑢
𝑠𝑖𝑛𝑛𝑥
𝐿𝑒𝑡 𝑢 = 𝑥, 𝑑𝑢 = 𝑑𝑥 𝑎𝑛𝑑 𝑑𝑣 = 𝑐𝑜𝑠𝑛𝑥𝑑𝑥, 𝑣 =
𝑛
1 𝑠𝑖𝑛𝑛𝑥 0 1 0
∴ 𝑎𝑛 = − [|𝑥 | − ∫ 𝑠𝑖𝑛𝑛𝑥𝑑𝑥 ]
𝜋 𝑛 −𝜋 𝑛 𝜋
𝑠𝑖𝑛𝑛𝜋 = 0 𝑎𝑛𝑑 sin 0 = 0, ∴ 𝑡ℎ𝑒 𝑤ℎ𝑜𝑙𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑓𝑖𝑟𝑠𝑡 𝑡𝑒𝑟𝑚 = 0
1 1
= − [0 + 2 |𝑐𝑜𝑠𝑛𝑥|0𝜋 ]
𝜋 𝑛
1
∴ 𝑎𝑛 = − 2 [𝑐𝑜𝑠0 − cos (𝑛𝜋)]
𝜋𝑛

6
Cos0 = 1

1
∴ 𝑎𝑛 = [1 − cos 𝑛𝜋]
𝜋𝑛2
𝑇𝑎𝑘𝑖𝑛𝑔 𝑡ℎ𝑒 𝑣𝑎𝑙𝑢𝑒𝑠 𝑜𝑓 𝑛 𝑓𝑟𝑜𝑚 1 … . ∞
0, 𝑓𝑜𝑟 𝑛 𝑒𝑣𝑒𝑛
𝑎𝑛 = { 2
− 2 , 𝑓𝑜𝑟 𝑛 𝑜𝑑𝑑
𝜋𝑛
𝟐 𝒎𝒂𝒓𝒌𝒔
1 𝜋
𝑏𝑛 = ∫ 𝑓(𝑥)𝑠𝑖𝑛𝑛𝑥𝑑𝑥
𝜋 −𝜋
1 0
− ∫ −𝑥𝑠𝑖𝑛𝑛𝑥𝑑𝑥𝑑𝑥
𝜋 −𝜋
𝑈𝑠𝑖𝑛𝑔 𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑡𝑖𝑜𝑛 𝑏𝑦 𝑝𝑎𝑟𝑡𝑠, ∫ 𝑢𝑑𝑣 = 𝑢𝑣 − ∫ 𝑣𝑑𝑢
𝑐𝑜𝑠𝑛𝑥
𝐿𝑒𝑡 𝑢 = 𝑥, 𝑑𝑢 = 𝑑𝑥 𝑎𝑛𝑑 𝑑𝑣 = 𝑠𝑖𝑛𝑛𝑥𝑑𝑥, 𝑣 = −
𝑛
1 𝑐𝑜𝑠𝑛𝑥 0 1 0
∴ 𝑏𝑛 = − [|−𝑥 | − ∫ 𝑐𝑜𝑠𝑛𝑥𝑑𝑥 ]
𝜋 𝑛 −𝜋 𝑛 𝜋
1 𝑐𝑜𝑠𝑛𝑥 0 1
∴ 𝑏𝑛 = − [|−𝑥 | + 2 |𝑠𝑖𝑛𝑛𝑥|0𝜋 ]
𝜋 𝑛 −𝜋 𝑛
𝑠𝑖𝑛𝑛𝜋 = 0 𝑎𝑛𝑑 sin 0 = 0, ∴ 𝑡ℎ𝑒 𝑤ℎ𝑜𝑙𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑠𝑒𝑐𝑜𝑛𝑑 𝑡𝑒𝑟𝑚 = 0
1
∴ 𝑏𝑛 = − [(0 − (−𝜋)((𝑐𝑜𝑠0 − cos (−𝑛𝜋)]
𝑛𝜋
1
∴ 𝑏𝑛 = − [𝜋(𝑐𝑜𝑠0 − cos (−𝑛𝜋)]
𝑛𝜋
1
∴ 𝑏𝑛 = − [𝜋(𝑐𝑜𝑠0 − cos (𝑛𝜋)]
𝑛𝜋
1
∴ 𝑏𝑛 = − [𝜋(1 − cos (𝑛𝜋)]
𝑛𝜋
𝑐𝑜𝑠𝑛𝜋 = −1
1
∴ 𝑏𝑛 = − [𝜋((1 − (−1)]
𝑛𝜋
1
∴ 𝑏𝑛 = − [𝜋𝑐𝑜𝑠𝑛𝜋 − 0]
𝑛𝜋
1
∴ 𝑏𝑛 = − [𝑐𝑜𝑠𝑛𝜋]
𝑛
𝑇𝑎𝑘𝑖𝑛𝑔 𝑡ℎ𝑒 𝑣𝑎𝑙𝑢𝑒𝑠 𝑜𝑓 𝑛 𝑓𝑟𝑜𝑚 1 … . ∞
1
− , 𝑓𝑜𝑟 𝑛 𝑒𝑣𝑒𝑛
𝑏𝑛 = { 𝑛
1
, 𝑓𝑜𝑟 𝑛 𝑜𝑑𝑑
𝑛

7
The Fourier series equation is given as:
1 2 1 1 1 1 1
𝑓(𝑡) = − {𝑐𝑜𝑠𝑥 + 𝑐𝑜𝑠3𝑥 + 𝑐𝑜𝑠𝑥+. . } + {𝑠𝑖𝑛𝑥 − 𝑠𝑖𝑛2𝑥 + 𝑠𝑖𝑛3𝑥 − 𝑠𝑖𝑛4𝑥+. . }
4 𝜋 9 25 2 3 4

Functions with Periods other than 2π

In practice, periodic functions other than 2π are encountered, such as Time intervals T that
𝑇 𝑇
can be defined over and interval 0 ≤ 𝑥 ≤ 𝑇, − 2 ≤ 𝑥 ≤ 2 , 𝑒𝑡𝑐, or period P = 2L, where L may
be the length of a violin string, length of rod in heat condition, etc.
The mathematical manipulation is to use the change of scale from periodic function of 2π to
the one being considered. For instance, if a violin string is plucked, the to and fro movement
from rest position completes a cycle in 2L.

𝜋
Taking the period 2π to be proportional to 2L, ie, 𝜋𝛼𝐿,so that 𝜋 = 𝑘𝐿. Thus, 𝑘 = 𝐿
Taking the change of scale as 𝑣 = 𝑘𝑥, with k such that the old period 𝑣 = 2𝜋, gives new
period x = 2L
𝜋𝑥
∴ 𝑣 = 𝑘𝑥 =
𝐿
So that the Fourier series becomes
∞
𝑎0 𝑛𝜋𝑥 𝑛𝜋𝑥
𝑓(𝑥) = + ∑ (𝑎𝑛 𝑐𝑜𝑠 + 𝑏𝑛 𝑠𝑖𝑛 )
2 𝐿 𝐿
𝑛=1
With the Fourier coefficients given by Euler formulas as
1 𝐿
𝑎0 = ∫ 𝑓(𝑥)𝑑𝑥
𝐿 −𝐿
2 𝐿 𝑛𝜋𝑥
𝑎𝑛 = ∫ 𝑓(𝑥)𝑐𝑜𝑠 𝑑𝑥
𝐿 −𝐿 𝐿
2 𝐿 𝑛𝜋𝑥
𝑏𝑛 = ∫ 𝑓(𝑥)𝑠𝑖𝑛 𝑑𝑥
𝐿 −𝐿 𝐿
Example

8
Sketch the graph of the function below and state what type of function it is and determine the Fourier
series expansion.
𝑥, 0≤𝑥≤1
𝑓(𝑥) = {
1 − 𝑥, 1≤𝑥≤2
Solution
To sketch the graph of the functions, we employ the equation of a straight line; y = mx + c
𝑓𝑜𝑟 𝑓(𝑥) = 𝑥; 𝑐 = 0; 𝑚 = 1; 𝑦1 = 𝑥1 ; 𝑦2 = 𝑥2
𝑑𝑦 𝑦2 − 𝑦1
= =1
𝑑𝑥 𝑥2 − 𝑥1
𝑑𝑦 𝑦2 − 𝑦1
= = 1; 𝑦2 − 𝑦1 = 1; ∴ 𝑦2 = 1 + 𝑦1 ; 𝑤ℎ𝑒𝑟𝑒 𝑦1 = 0; 𝑡ℎ𝑢𝑠, 𝑦2 = 1
𝑑𝑥 1−0
Slope = 1 at 0 ≤ 𝑥 ≤ 1 would be the line rising from x = 0 to x = 1 and y = 1
𝑓𝑜𝑟 𝑓(𝑥) = 1 − 𝑥; 𝑐 = 1; 𝑚 = −1; 𝑦1 = 1 − 𝑥1 ; 𝑦2 = 1 − 𝑥2
𝑑𝑦 𝑦2 − 𝑦1
= = −1
𝑑𝑥 𝑥2 − 𝑥1
𝑑𝑦 𝑦2 − 𝑦1
= = −1; 𝑦2 − 𝑦1 = −1; ∴ 𝑦2 = 𝑦1 − 1; 𝑤ℎ𝑒𝑟𝑒 𝑦1 = 0; 𝑡ℎ𝑢𝑠, 𝑦2 = −1
𝑑𝑥 2−1
Slope = 1 at 1 ≤ 𝑥 ≤ 2 would be the line dropping from x = 1to x = 2 and y = -1

f (x)

0 1 2 3 x

-1

Using the Fourier series equation as

∞
𝑎0 𝑛𝜋𝑥 𝑛𝜋𝑥
𝑓(𝑥) = + ∑ (𝑎𝑛 𝑐𝑜𝑠 + 𝑏𝑛 𝑠𝑖𝑛 )
2 𝐿 𝐿
𝑛=1
𝑥, 0≤𝑥≤1
𝑓(𝑥) = {
1 − 𝑥, 1≤𝑥≤2
We determine the Fourier coefficients from
Where the period P = 2L = 2, thus, L = 1
1 2
1 𝐿 1 1 2
𝑥2 𝑥2
𝑎0 = ∫ 𝑓(𝑥)𝑑𝑥 = [∫ 𝑥𝑑𝑥 + ∫ (1 − 𝑥)𝑑𝑥] = [| | + |𝑥|12 − | | ]
𝐿 −𝐿 1 0 1 2 0 2 1
1 1 1 1
[( − 0) + (2 − 1) − (2 − )] = + 1 − 2 + = 0
2 2 2 2
𝑎0 = 0

9
 2 𝐿 𝑛𝜋𝑥 2 1 𝑛𝜋𝑥 2
𝑛𝜋𝑥
𝑎𝑛 = ∫ 𝑓(𝑥)𝑐𝑜𝑠 𝑑𝑥 = [∫ 𝑥𝑐𝑜𝑠 𝑑𝑥 + ∫ (1 − 𝑥)𝑐𝑜𝑠 𝑑𝑥]
𝐿 −𝐿 𝐿 1 0 1 1 1
1 2
𝑎𝑛 = 2 [∫ 𝑥𝑐𝑜𝑠𝑛𝜋𝑥𝑑𝑥 + ∫ (1 − 𝑥)𝑐𝑜𝑠𝑛𝜋𝑥𝑑𝑥 ]
0 1

𝑈𝑠𝑖𝑛𝑔 𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑡𝑖𝑜𝑛 𝑏𝑦 𝑝𝑎𝑟𝑡𝑠, ∫ 𝑢𝑑𝑣 = 𝑢𝑣 − ∫ 𝑣𝑑𝑢

For the first part of the equation
1
𝐿𝑒𝑡 𝑢 = 𝑥, 𝑑𝑢 = 𝑑𝑥 𝑎𝑛𝑑 𝑑𝑣 = 𝑐𝑜𝑠𝑛𝜋𝑥𝑑𝑥, 𝑣 = 𝑠𝑖𝑛𝑛𝜋𝑥
𝑛𝜋
Taking the second part of the equation
1
𝐿𝑒𝑡 𝑢 = 1 − 𝑥, 𝑑𝑢 = −𝑑𝑥 𝑎𝑛𝑑 𝑑𝑣 = 𝑐𝑜𝑠𝑛𝜋𝑥𝑑𝑥, 𝑣 = 𝑠𝑖𝑛𝑛𝜋𝑥
𝑛𝜋
1 1 1 1 1 2
∴ [{ |𝑥𝑠𝑖𝑛𝑛𝜋𝑥|10 − ∫ 𝑠𝑖𝑛𝑛𝜋𝑥𝑑𝑥 } + { |(1 − 𝑥)𝑠𝑖𝑛𝑛𝜋𝑥|12 + ∫ 𝑠𝑖𝑛𝑛𝜋𝑥𝑑𝑥 }]
𝑛𝜋 𝑛𝜋 0 𝑛𝜋 𝑛𝜋 1
1 1 1
∴ [|𝑥𝑠𝑖𝑛𝑛𝜋𝑥|10 + |𝑐𝑜𝑠𝑛𝜋𝑥|10 |+(1 − 𝑥)𝑠𝑖𝑛𝑛𝜋𝑥|12 − |𝑐𝑜𝑠𝑛𝜋𝑥|12 ]
𝑛𝜋 𝑛𝜋 𝑛𝜋
1 1 1
∴ [(𝑠𝑖𝑛𝑛𝜋 − 0) + (𝑐𝑜𝑠𝑛𝜋 − 𝑐𝑜𝑠0) + (1 − 2)𝑠𝑖𝑛2𝑛𝜋 − (1 − 1)𝑠𝑖𝑛𝑛𝜋 − (𝑐𝑜𝑠2𝑛𝜋
𝑛𝜋 𝑛𝜋 𝑛𝜋
− 𝑐𝑜𝑠𝑛𝜋)]
𝑠𝑖𝑛𝑛𝜋 = 𝑠𝑖𝑛2𝑛𝜋 = 0, 𝑐𝑜𝑠0 = 1
1 1 1
∴ [ (𝑐𝑜𝑠𝑛𝜋 − 1) − (𝑐𝑜𝑠2𝑛𝜋 − 𝑐𝑜𝑠𝑛𝜋)]
𝑛𝜋 𝑛𝜋 𝑛𝜋
1
∴ 2 2 [𝑐𝑜𝑠𝑛𝜋 − 1 − 𝑐𝑜𝑠2𝑛𝜋 + 𝑐𝑜𝑠𝑛𝜋]
𝑛 𝜋
1
𝑇ℎ𝑢𝑠, 𝑎𝑛 = 2 2 [2𝑐𝑜𝑠𝑛𝜋 − 𝑐𝑜𝑠2𝑛𝜋 − 1]
𝑛 𝜋
𝑇𝑎𝑘𝑖𝑛𝑔 𝑡ℎ𝑒 𝑣𝑎𝑙𝑢𝑒𝑠 𝑜𝑓 𝑛 𝑓𝑟𝑜𝑚 1 … . ∞
1 −4
𝑛 = 1; 𝑎𝑛 = 2 2 [−2 − 1 − 1] = 2 2
𝑛 𝜋 𝑛 𝜋
1
𝑛 = 2; 𝑎𝑛 = 2 2 [2 − 1 − 1] = 0
𝑛 𝜋
1 −4
𝑛 = 3; 𝑎𝑛 = 2 2 [−2 − 1 − 1] = 2 2
𝑛 𝜋 𝑛 𝜋
1 −4
𝑛 = 4; 𝑎𝑛 = 2 2 [−2 − 1 − 1] = 2 2
𝑛 𝜋 𝑛 𝜋
0; 𝑓𝑜𝑟 𝑛 = 𝑜𝑑𝑑
𝑎𝑛 = { −4
; 𝑓𝑜𝑟 𝑛 = 𝑒𝑣𝑒𝑛
𝑛2 𝜋 2
2 𝐿 𝑛𝜋𝑥 2 1 𝑛𝜋𝑥 2
𝑛𝜋𝑥
𝑏𝑛 = ∫ 𝑓(𝑥)𝑠𝑖𝑛 𝑑𝑥 = [∫ 𝑥𝑠𝑖𝑛 𝑑𝑥 + ∫ (1 − 𝑥)𝑠𝑖𝑛 𝑑𝑥]
𝐿 −𝐿 𝐿 1 0 1 1 1

10
 1 2
𝑏𝑛 = 2 [∫ 𝑥𝑠𝑖𝑛𝑛𝜋𝑥𝑑𝑥 + ∫ (1 − 𝑥)𝑠𝑖𝑛𝑛𝜋𝑥𝑑𝑥 ]
0 1

𝑈𝑠𝑖𝑛𝑔 𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑡𝑖𝑜𝑛 𝑏𝑦 𝑝𝑎𝑟𝑡𝑠, ∫ 𝑢𝑑𝑣 = 𝑢𝑣 − ∫ 𝑣𝑑𝑢

For the first part of the equation
1
𝐿𝑒𝑡 𝑢 = 𝑥, 𝑑𝑢 = 𝑑𝑥 𝑎𝑛𝑑 𝑑𝑣 = 𝑠𝑖𝑛𝑛𝜋𝑥𝑑𝑥, 𝑣 = − 𝑐𝑜𝑠𝑛𝜋𝑥
𝑛𝜋
Taking the second part of the equation
1
𝐿𝑒𝑡 𝑢 = 1 − 𝑥, 𝑑𝑢 = −𝑑𝑥 𝑎𝑛𝑑 𝑑𝑣 = 𝑠𝑖𝑛𝑛𝜋𝑥𝑑𝑥, 𝑣 = − 𝑐𝑜𝑠𝑛𝜋𝑥
𝑛𝜋
1 1 1 1 1 2
∴ [{− |𝑥𝑐𝑜𝑠𝑛𝜋𝑥|10 + ∫ 𝑐𝑜𝑠𝑛𝜋𝑥𝑑𝑥 } + { |−(1 − 𝑥)𝑐𝑜𝑠𝑛𝜋𝑥|12 − ∫ 𝑐𝑜𝑠𝑛𝜋𝑥𝑑𝑥 }]
𝑛𝜋 𝑛𝜋 0 𝑛𝜋 𝑛𝜋 1
1 1 1
∴ [−|𝑥𝑐𝑜𝑠𝑛𝜋𝑥|10 + |𝑠𝑖𝑛𝑛𝜋𝑥|10 +|−(1 − 𝑥)𝑐𝑜𝑠𝑛𝜋𝑥|12 − |𝑠𝑖𝑛𝑛𝜋𝑥|12 ]
𝑛𝜋 𝑛𝜋 𝑛𝜋
1 1 1
∴ [−(𝑐𝑜𝑠𝑛𝜋 − 0) + (𝑠𝑖𝑛𝑛𝜋 − 𝑠𝑖𝑛0) − (1 − 2)𝑐𝑜𝑠2𝑛𝜋 + (1 − 1)𝑐𝑜𝑠𝑛𝜋 − (𝑠𝑖𝑛2𝑛𝜋
𝑛𝜋 𝑛𝜋 𝑛𝜋
− 𝑠𝑖𝑛𝑛𝜋)]

1 1 1
∴ [−(𝑐𝑜𝑠𝑛𝜋 − 0) + (𝑠𝑖𝑛𝑛𝜋 − 𝑠𝑖𝑛0) − (1 − 2)𝑐𝑜𝑠2𝑛𝜋 + (1 − 1)𝑐𝑜𝑠𝑛𝜋 − (𝑠𝑖𝑛2𝑛𝜋
𝑛𝜋 𝑛𝜋 𝑛𝜋
− 𝑠𝑖𝑛𝑛𝜋)]
𝑠𝑖𝑛𝑛𝜋 = 𝑠𝑖𝑛2𝑛𝜋 = 𝑠𝑖𝑛0 = 0, 𝑐𝑜𝑠0 = 1
1
∴ [−(𝑐𝑜𝑠𝑛𝜋) + 𝑐𝑜𝑠2𝑛𝜋]
𝑛𝜋
1
∴ 𝑏𝑛 = [𝑐𝑜𝑠2𝑛𝜋 − 𝑐𝑜𝑠𝑛𝜋]
𝑛𝜋
𝑇𝑎𝑘𝑖𝑛𝑔 𝑡ℎ𝑒 𝑣𝑎𝑙𝑢𝑒𝑠 𝑜𝑓 𝑛 𝑓𝑟𝑜𝑚 1 … . ∞
1 1 3
𝑛 = 1; 𝑏𝑛 = [2(1) − 1(−1)] = [2 + 1] =
𝑛𝜋 𝑛𝜋 𝑛𝜋
1 1 1
𝑛 = 2; 𝑏𝑛 = [2(1) − 1(1)] = [2 − 1] =
𝑛𝜋 𝑛𝜋 𝑛𝜋
1 1 3
𝑛 = 3; 𝑏𝑛 = [2(1) − 1(−1)] = [2 + 1] =
𝑛𝜋 𝑛𝜋 𝑛𝜋
1 1 1
𝑛 = 4; 𝑏𝑛 = [2(1) − 1(1)] = [2 − 1] =
𝑛𝜋 𝑛𝜋 𝑛𝜋
3
; 𝑓𝑜𝑟 𝑛 = 𝑜𝑑𝑑
𝑏𝑛 = { 𝑛𝜋
1
; 𝑓𝑜𝑟 𝑛 = 𝑒𝑣𝑒𝑛
𝑛𝜋
1 1 1 4 1 1 1
𝑓(𝑥) = {3𝑠𝑖𝑛𝑥 + 𝑠𝑖𝑛2𝑥 + 𝑠𝑖𝑛3𝑥 + 𝑠𝑖𝑛4𝑥 + ⋯ } − 2 { 𝑐𝑜𝑠2𝑥 + 𝑐𝑜𝑠4𝑥 + 𝑐𝑜𝑠6𝑥+. . }
𝜋 2 4 𝜋 4 16 36

11
Function with Period, T
𝑇 𝑇
If y = f(x) is defined over the interval 0 ≤ 𝑥 ≤ 𝑇, − 2 ≤ 𝑥 ≤ 2. This interval can be converted
to 2π by changing the unit of the independent variable. In most practical cases involving
physical oscillations, the independent variable is time (t) with the periodic interval normally
denoted by T, ie, f(t + T) = f(t). Here, a cycle is completed in T(s) and the frequency f(Hz).
1 𝑟𝑎𝑑 2𝜋 2𝜋
𝑓 = ; 𝑡ℎ𝑒 𝑎𝑛𝑔𝑢𝑙𝑎𝑟 𝑣𝑒𝑙 ( ),𝜔 = , 𝑠𝑜 𝑡ℎ𝑎𝑡 𝑇 =
𝑇 𝑠 𝑇 𝜔
The angle, x (rad) at any time t is 𝑥 = 𝜔𝑡
The Fourier series of the function, T is given as
∞ ∞
𝑎0 𝑎0 2𝑛𝜋𝑡 2𝑛𝜋𝑡
𝑓(𝑡) = + ∑(𝑎𝑛 𝑐𝑜𝑠𝑛𝜔𝑡 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝜔𝑡) = + ∑ (𝑎𝑛 𝑐𝑜𝑠 + 𝑏𝑛 𝑠𝑖𝑛 )
2 2 𝑇 𝑇
𝑛=1 𝑛=1
The Fourier coefficients are as follows
2𝜋
2 𝑇 𝜔 𝜔
𝑎0 = ∫ 𝑓(𝑡)𝑑𝑡 = ∫ 𝑓(𝑡)𝑑𝑡
𝑇 0 𝜋 0
2𝜋
2 𝑇 2𝑛𝜋𝑡 𝜔 𝜔
𝑎𝑛 = ∫ 𝑓(𝑡)𝑐𝑜𝑠 𝑑𝑡 = ∫ 𝑓(𝑡)𝑐𝑜𝑠𝑛𝜔𝑡𝑑𝑡
𝑇 0 𝑇 𝜋 0
2𝜋
2 𝑇 2𝑛𝜋𝑡 𝜔 𝜔
𝑏𝑛 = ∫ 𝑓(𝑡)𝑠𝑖𝑛 𝑑𝑡 = ∫ 𝑓(𝑡)𝑠𝑖𝑛𝑛𝜔𝑡𝑑𝑡
𝑇 0 𝑇 𝜋 0
Example
Sketch the graph of the function below and then determine the Fourier series expansion if f(t + 2) =
f(t)
(1 + 𝑡) , −1 ≤ 𝑡 ≤ 0
𝑓(𝑡) = {
0, 0≤𝑡≤1
Solution
The sketch
𝑓(𝑡) = 1 + 𝑡 ≡ 𝑦 = 𝑚𝑥 + 𝑐 𝑎𝑡 − 1 ≤ 𝑡 ≤ 0
Thus, the slope rises from -1 to 0 at 1 intercept on y-axis
𝑓(𝑡) = 0 𝑎𝑡 0 ≤ 𝑡 ≤ 1
That is, f (t) runs flat on the x-axis
f(t)

-2 -1 0 1 t
Using the Fourier series equation as

12
 ∞
𝑎0 2𝑛𝜋𝑡 2𝑛𝜋𝑡
𝑓(𝑡) = + ∑ (𝑎𝑛 𝑐𝑜𝑠 + 𝑏𝑛 𝑠𝑖𝑛 )
2 𝑇 𝑇
𝑛=1
The period = 2, that is, T = 2
We determine the Fourier coefficients from
2 𝑇 𝑇
𝑎0 = ∫ 𝑓(𝑡)𝑑𝑡 = ∫ 𝑓(𝑡)𝑑𝑡
𝑇 0 0
0
= ∫ (1 + 𝑡)𝑑𝑡
−1
0
𝑡2
[𝑡 + ]
2 −1
(−1)2
0 − [(−1) + ]
2
1
0 − [ − 1]
2
1
0 − [− ]
2
1
𝑎0 =
2
2 𝑇 2𝑛𝜋𝑡 0
𝑎𝑛 = ∫ 𝑓(𝑡)𝑐𝑜𝑠 𝑑𝑡 = ∫ 𝑓(𝑡)𝑐𝑜𝑠𝑛𝜋𝑡𝑑𝑡
𝑇 0 𝑇 −1

𝑈𝑠𝑖𝑛𝑔 𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑡𝑖𝑜𝑛 𝑏𝑦 𝑝𝑎𝑟𝑡𝑠, ∫ 𝑢𝑑𝑣 = 𝑢𝑣 − ∫ 𝑣𝑑𝑢

𝑠𝑖𝑛𝑛𝜋𝑡
𝐿𝑒𝑡 𝑢 = 1 + 𝑡, 𝑑𝑢 = 𝑑𝑡 𝑎𝑛𝑑 𝑑𝑣 = 𝑐𝑜𝑠𝑛𝜋𝑡𝑑𝑡, 𝑣 =
𝑛𝜋
𝑠𝑖𝑛𝑛𝜋𝑡 0 1 0
∴ 𝑎𝑛 = [(1 + 𝑡) | | − ∫ 𝑠𝑖𝑛𝑛𝜋𝑡𝑑𝑡]
𝑛𝜋 −1 𝑛𝜋 −1
𝑠𝑖𝑛𝑛𝜋𝑡 0 1
= [(1 + 𝑡) | | + 2 2 |𝑐𝑜𝑠𝑛𝜋𝑡|0−1 ]
𝑛𝜋 −1 𝑛 𝜋
𝑠𝑖𝑛𝑛𝜋 = 0 𝑎𝑛𝑑 sin 0 = 0, ∴ 𝑡ℎ𝑒 𝑤ℎ𝑜𝑙𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑓𝑖𝑟𝑠𝑡 𝑡𝑒𝑟𝑚 = 0
1
∴ 𝑎𝑛 = 2 2 [𝑐𝑜𝑠0 − cos (−𝑛𝜋)]
𝑛 𝜋
1
∴ 𝑎𝑛 = 2 2 [1 − cos 𝑛𝜋]
𝑛 𝜋
𝑇𝑎𝑘𝑖𝑛𝑔 𝑡ℎ𝑒 𝑣𝑎𝑙𝑢𝑒𝑠 𝑜𝑓 𝑛 𝑓𝑟𝑜𝑚 0 … . ∞
0, 𝑓𝑜𝑟 𝑛 𝑒𝑣𝑒𝑛
𝑎𝑛 = { 2
, 𝑓𝑜𝑟 𝑛 𝑜𝑑𝑑
𝑛2 𝜋 2
2 𝑇 2𝑛𝜋𝑡 0
𝑏𝑛 = ∫ 𝑓(𝑡)𝑠𝑖𝑛 𝑑𝑡 = ∫ 𝑓(𝑡)𝑠𝑖𝑛𝑛𝜋𝑡𝑑𝑡
𝑇 0 𝑇 −1

𝑈𝑠𝑖𝑛𝑔 𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑡𝑖𝑜𝑛 𝑏𝑦 𝑝𝑎𝑟𝑡𝑠, ∫ 𝑢𝑑𝑣 = 𝑢𝑣 − ∫ 𝑣𝑑𝑢

𝑐𝑜𝑠𝑛𝜋𝑡
𝐿𝑒𝑡 𝑢 = 1 + 𝑡, 𝑑𝑢 = 𝑑𝑡 𝑎𝑛𝑑 𝑑𝑣 = 𝑠𝑖𝑛𝑛𝜋𝑡𝑑𝑡, 𝑣 = −
𝑛𝜋

13
 𝑐𝑜𝑠𝑛𝜋𝑡 0 1 0
∴ 𝑏𝑛 = [−(1 + 𝑡) | | + ∫ 𝑐𝑜𝑠𝑛𝜋𝑡𝑑𝑡]
𝑛𝜋 −1 𝑛𝜋 −1
𝑐𝑜𝑠𝑛𝜋𝑡 0 1
= [−(1 + 𝑡) | | + 2 2 |𝑠𝑖𝑛𝑛𝜋𝑡|0−1 ]
𝑛𝜋 −1 𝑛 𝜋
𝑠𝑖𝑛𝑛𝜋 = 0 𝑎𝑛𝑑 sin 0 = 0, ∴ 𝑡ℎ𝑒 𝑤ℎ𝑜𝑙𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑠𝑒𝑐𝑜𝑛𝑑 𝑡𝑒𝑟𝑚 = 0
1
∴ 𝑏𝑛 = − [𝑐𝑜𝑠0]
𝑛𝜋
1
∴ 𝑏𝑛 = −
𝑛𝜋
The Fourier series equation is given as:
1 2 1 1 1 1 1 1
𝑓(𝑡) = + {𝑐𝑜𝑠𝜋𝑡 + 𝑐𝑜𝑠3𝜋𝑡 + 𝑐𝑜𝑠5𝑡+. . } − {𝑠𝑖𝑛𝜋𝑡 + 𝑠𝑖𝑛2𝜋𝑡 + 𝑠𝑖𝑛3𝜋𝑡 + 𝑠𝑖𝑛4𝜋𝑡+. . }
4 𝜋 9 25 𝜋 2 3 4

14