Differential Equations

ID 32002
2.1 Fourier Series

Eng.N.Sulaxshan
Lecturer in Mathematics

Department of Interdisciplinary Studies,

Faculty of Engineering,
South Eastern University of Sri Lanka
1
 Periodic Functions
If the value of each ordinate 𝑓(𝑡) repeat itself at equal intervals
in the abscissa, then 𝑓(𝑡) is said to be a periodic function.
If 𝑓 𝑡 = 𝑓 𝑡 + 𝑇 = 𝑓 𝑡 + 2𝑇 = ⋯ then 𝑇 is called the
period of the function 𝑓 𝑡 .
For example, 𝑠𝑖𝑛𝑥 = sin 𝑥 + 2ᴨ = sin 𝑥 + 4ᴨ = ⋯ so
sin 𝑥 is a periodic function with the period 2ᴨ, This is also
called sinusoidal periodic function.

2
 Fourier Series
Non-sinusoidal periodic function into a fundamental and its
harmonics.
A series of sines and cosines of an angle and its multiples of the
form
𝑎0
+ 𝑎1 𝑐𝑜𝑠𝑥 + 𝑎2 𝑐𝑜𝑠2𝑥 + 𝑎3 𝑐𝑜𝑠3𝑥 + ⋯ + 𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏1 𝑠𝑖𝑛𝑥
2
+ 𝑏2 𝑠𝑖𝑛2𝑥 + 𝑏3 𝑠𝑖𝑛3𝑥 + ⋯ + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥
∞
𝑎0
= + 𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥
2
𝑛=1
Is called the Fourier series. Where 𝑎𝑛 , 𝑏𝑛 are Fourier constants.
A periodic function 𝑓(𝑥) can be expanded in a Fourier series.
The series consists of the following.
1. A constant term 𝑎0 called DC component in electrical work
2. A component at the fundamental frequency determined by 𝑎1 , 𝑏1
3. A component of the harmonics (multiples of the fundamental
frequency) determined by 𝑎2 , 𝑎3 … . 𝑏2 , 𝑏3,…
3
 Dirichlet’s Conditions
Let a function 𝑓(𝑥) satisfy the following conditions:
a) 𝑓(𝑥) is defined in the interval 𝑐 < 𝑥 < 𝑐 + 2𝑙
b) 𝑓(𝑥) and 𝑓′(𝑥) are sectional continuous in 𝑐 < 𝑥 < 𝑐 + 2𝑙
c) 𝑓 𝑥 + 2𝑙 = 𝑓(𝑥), that is 𝑓(𝑥) is periodic with period 2𝑙
Then at every point of continuity, we have
𝑎0 ∞ 𝑛𝜋𝑥 𝑛𝜋𝑥
𝑓 𝑥 = + 𝑛=1 𝑎𝑛 𝑐𝑜𝑠 + 𝑏𝑛 𝑠𝑖𝑛 …..(1)
2 𝑙 𝑙
1 𝑐+2𝑙 𝑛𝜋𝑥
Where 𝑎𝑛 = 𝑐 𝑓(𝑥) 𝑐𝑜𝑠 𝑑𝑥 ,
𝑙 𝑙
1 𝑐+2𝑙 𝑛𝜋𝑥
𝑏𝑛 = 𝑐 𝑓(𝑥) 𝑠𝑖𝑛 𝑑𝑥
𝑙 𝑙
Point of discontinuity, the left side of (1) is replaced by
1
𝑓 𝑥 + 𝛿 + 𝑓(𝑥 − 𝛿)
2
The series (1) with coefficients given above is called the
Fourier series of 𝑓(𝑥) and for many problems 𝑐 = 0 𝑜𝑟 − 𝑙. In
case 𝑙 = π, 𝑓 𝑥 has period of 2𝜋 and above series and
coefficients can be simplified
4
 Dirichlet’s Conditions
Also function 𝑓(𝑥) should satisfy for Dirichlet’s conditions:
a) 𝑓(𝑥) is single-valued and bounded
b) Has at most a finite number of maxima and minima
c) Has only a finite number of discontinuous.

Advantages of Fourier series

Discontinuous function can be represented by Fourier series.
Although derivatives of the discontinuous functions do not
exist (This is not true for Taylor’s series)
The Fourier series is useful in expanding the periodic
functions since outside the closed interval, there exists a
periodic extension of the function
Expansion of an oscillating function by Fourier series all
modes of oscillation (fundamental and all overtones) which is
extremely useful in physics
Fourier series of a discontinuous function is not uniformly
5
convergent at all points
 Useful Integrals
Below we have a few results which we need in deriving
Fourier series. Given that m and n are positive integers or zero

𝑙 𝑛𝜋𝑥

−𝑙
𝑠𝑖𝑛 𝑑𝑥 = 0, 𝑖𝑓 𝑛 ≠ 0
𝑙
𝑙 𝑛𝜋𝑥

−𝑙
𝑐𝑜𝑠 𝑑𝑥 = 0, 𝑖𝑓 𝑛 ≠ 0
𝑙
𝑙 𝑚𝜋𝑥 𝑛𝜋𝑥

−𝑙
𝑠𝑖𝑛 . 𝑠𝑖𝑛 𝑑𝑥 = 0, 𝑖𝑓 𝑚≠𝑛
𝑙 𝑙
= 𝑙, 𝑖𝑓 𝑚 = 𝑛
= 0, 𝑖𝑓 𝑚 = 𝑛 = 0
𝑙 𝑚𝜋𝑥 𝑛𝜋𝑥

−𝑙
𝑐𝑜𝑠 . 𝑐𝑜𝑠 𝑑𝑥 = 0, 𝑖𝑓 𝑚 ≠ 𝑛
𝑙 𝑙
= 𝑙, 𝑖𝑓 𝑚 = 𝑛
= 2𝑙, 𝑖𝑓 𝑚 = 𝑛 = 0

6
 Useful Integrals
𝑙 𝑚𝜋𝑥 𝑛𝜋𝑥

−𝑙
𝑠𝑖𝑛 . 𝑐𝑜𝑠 𝑑𝑥 = 0, 𝑖𝑓 𝑚 ≠ 𝑛
𝑙 𝑙
= 0, 𝑖𝑓 𝑚 = 𝑛
𝑎𝑥 𝑒 𝑎𝑥
 𝑒 𝑠𝑖𝑛𝑏𝑥𝑑𝑥 = 2 2 (𝑎𝑠𝑖𝑛𝑏𝑥 − 𝑏𝑐𝑜𝑠𝑏𝑥)
𝑎 +𝑏
𝑎𝑥 𝑒 𝑎𝑥
 𝑒 𝑐𝑜𝑠𝑏𝑥𝑑𝑥 = 2 2 (𝑎𝑐𝑜𝑠𝑏𝑥 − 𝑏𝑠𝑖𝑛𝑏𝑥)
𝑎 +𝑏
The above results remain valid even when the limits of integrals
− 𝑙, 𝑙 are replaced by 𝑐, 𝑐 + 2𝑙 respectively.

Discontinuous Functions
At a point of discontinuity, Fourier series gives the value of
𝑓(𝑥) as the arithmetic mean of left and right limits
At a point of discontinuity,
1
𝑥 = 𝑐, 𝑓 𝑥 = 𝑓 𝑥 + 𝛿 + 𝑓(𝑥 − 𝛿)
2
7
Proof
𝑙
𝑙 𝑛𝜋𝑥 𝑛𝜋𝑥 𝑙
1. −𝑙
𝑠𝑖𝑛
𝑙
𝑑𝑥 = −𝑐𝑜𝑠 𝑙 . 𝑛𝜋
−𝑙
−𝑙
= 𝑐𝑜𝑠𝑛𝜋 − cos(−𝑛𝜋)
𝑛𝜋

=0 ; 𝑛≠0

𝑙
𝑙 𝑛𝜋𝑥 𝑛𝜋𝑥 𝑙
2. −𝑙
𝑐𝑜𝑠
𝑙
𝑑𝑥 = 𝑠𝑖𝑛 𝑙 . 𝑛𝜋
−𝑙

𝑙
= 𝑠𝑖𝑛𝑛𝜋 − sin(−𝑛𝜋)
𝑛𝜋
𝑙
= 0+0
𝑛𝜋
=0

8
 𝑙 𝑚𝜋𝑥 𝑛𝜋𝑥
3. −𝑙
𝑠𝑖𝑛
𝑙
𝑠𝑖𝑛
𝑙
𝑑𝑥

If 𝑚 ≠ 𝑛
𝑙
−1 𝜋𝑥 𝜋𝑥
= [cos 𝑚 + 𝑛 − cos 𝑚 − 𝑛 ]𝑑𝑥
−𝑙 2 𝑙 𝑙
𝑙
−1 (𝑚 + 𝑛)𝜋𝑥 𝑙 𝜋𝑥 𝑙
= 𝑠𝑖𝑛 . − sin 𝑚 − 𝑛 .
2 𝑙 (𝑚 + 𝑛)𝜋 𝑙 (𝑚 − 𝑛)𝜋 −𝑙
𝑙
−𝑙 1 𝜋𝑥 1 𝜋𝑥
= sin 𝑚 + 𝑛 − sin(𝑚 − 𝑛) )
2𝜋 𝑚 + 𝑛 𝑙 𝑚−𝑛 𝑙 −𝑙

−𝑙 1 1
= sin 𝑚 + 𝑛 𝜋 − sin 𝑚 − 𝑛 𝜋
2𝜋 𝑚 + 𝑛 𝑚−𝑛
−𝑙
= 0−0
2𝜋
=0

9
 𝑙 𝑚𝜋𝑥 𝑛𝜋𝑥
4. −𝑙
𝑐𝑜𝑠
𝑙
𝑐𝑜𝑠
𝑙
𝑑𝑥

If 𝑚 ≠ 𝑛
𝑙
1 𝜋𝑥 𝜋𝑥
= [cos 𝑚 + 𝑛 + cos 𝑚 − 𝑛 ]𝑑𝑥
−𝑙 2 𝑙 𝑙
𝑙
1 (𝑚 + 𝑛)𝜋𝑥 𝑙 𝜋𝑥 𝑙
= 𝑠𝑖𝑛 . + sin 𝑚 − 𝑛 .
2 𝑙 (𝑚 + 𝑛)𝜋 𝑙 (𝑚 − 𝑛)𝜋 −𝑙
𝑙
𝑙 1 𝜋𝑥 1 𝜋𝑥
= sin 𝑚 + 𝑛 + sin(𝑚 − 𝑛) )
2𝜋 𝑚 + 𝑛 𝑙 𝑚−𝑛 𝑙 −𝑙

𝑙 1 1
= sin 𝑚 + 𝑛 𝜋 + sin 𝑚 − 𝑛 𝜋
2𝜋 𝑚 + 𝑛 𝑚−𝑛
𝑙
= 0+0
2𝜋
=0

10
If 𝑚 ≠ 𝑛
𝑙
1 𝜋𝑥 𝜋𝑥
= [cos 𝑚 + 𝑛 + cos 𝑚 − 𝑛 ]𝑑𝑥
−𝑙 2 𝑙 𝑙
𝑙
1 (𝑚 + 𝑛)𝜋𝑥 𝑙 𝜋𝑥 𝑙
= 𝑠𝑖𝑛 . + sin 𝑚 − 𝑛 .
2 𝑙 (𝑚 + 𝑛)𝜋 𝑙 (𝑚 − 𝑛)𝜋 −𝑙
𝑙
𝑙 1 𝜋𝑥 1 𝜋𝑥
= sin 𝑚 + 𝑛 + sin(𝑚 − 𝑛) )
2𝜋 𝑚 + 𝑛 𝑙 𝑚−𝑛 𝑙 −𝑙

𝑙 1 1
= sin 𝑚 + 𝑛 𝜋 + sin 𝑚 − 𝑛 𝜋
2𝜋 𝑚 + 𝑛 𝑚−𝑛
𝑙
= 0+0
2𝜋
=0

11
If 𝑚 = 𝑛
𝑙
𝑚𝜋𝑥
𝑠𝑖𝑛2 𝑑𝑥
−𝑙 𝑙
𝑙
1 2𝑚𝜋𝑥
= [1 − 𝑐𝑜𝑠 ]𝑑𝑥
−𝑙 2 𝑙
𝑙
1 2𝑚𝜋𝑥 𝑙
= 𝑥 − 𝑠𝑖𝑛 .
2 𝑙 2𝑚𝜋 −𝑙

𝑙 𝑙
= (𝑙 − −𝑙) − (sin 2𝑚𝜋 − sin(−2𝑚𝜋))
2 2𝑚𝜋
𝑙 𝑙
= 2𝑙 − (0 + 0)
2 2𝑚𝜋
=𝑙
If 𝑚 = 𝑛 = 0
𝑙 𝑙
𝑙
𝑠𝑖𝑛0. 𝑠𝑖𝑛0𝑑𝑥 = 0𝑑𝑥 = 𝐶 −𝑙 = 0
−𝑙 −𝑙
12
If 𝑚 = 𝑛
𝑙
𝑚𝜋𝑥
𝑐𝑜𝑠 2 𝑑𝑥
−𝑙 𝑙
𝑙
1 2𝑚𝜋𝑥
= [1 + 𝑐𝑜𝑠 ]𝑑𝑥
−𝑙 2 𝑙
𝑙
1 2𝑚𝜋𝑥 𝑙
= 𝑥 + 𝑠𝑖𝑛 .
2 𝑙 2𝑚𝜋 −𝑙

𝑙 𝑙
= 𝑙 − −𝑙 + (sin 2𝑚𝜋 − sin(−2𝑚𝜋))
2 2𝑚𝜋
𝑙 𝑙
= 2𝑙 + (0 + 0)
2 2𝑚𝜋
=𝑙
If 𝑚 = 𝑛 = 0
𝑙 𝑙
𝑙
𝑐𝑜𝑠0. 𝑐𝑜𝑠0𝑑𝑥 = 1𝑑𝑥 = 𝑥 −𝑙 = 𝑙 − −𝑙 = 2𝑙
−𝑙 −𝑙
13
 𝑙 𝑚𝜋𝑥 𝑛𝜋𝑥
5. −𝑙
𝑠𝑖𝑛
𝑙
𝑐𝑜𝑠
𝑙
𝑑𝑥

If 𝑚 ≠ 𝑛
𝑙
1 𝜋𝑥 𝜋𝑥
= [sin 𝑚 + 𝑛 + sin 𝑚 − 𝑛 ]𝑑𝑥
−𝑙 2 𝑙 𝑙
𝑙
1 𝑚 + 𝑛 𝜋𝑥 𝑙 𝜋𝑥 𝑙
= −𝑐𝑜𝑠 . + −cos 𝑚 − 𝑛 .
2 𝑙 𝑚+𝑛 𝜋 𝑙 (𝑚 − 𝑛)𝜋 −𝑙
𝑙
−𝑙 1 𝜋𝑥 1 𝜋𝑥
= cos 𝑚 + 𝑛 + cos(𝑚 − 𝑛) )
2𝜋 𝑚 + 𝑛 𝑙 𝑚−𝑛 𝑙 −𝑙

−𝑙 1 1
= (cos 𝑚 + 𝑛 𝜋 − cos 𝑚 + 𝑛)(−𝜋 ) + (cos 𝑚 − 𝑛 𝜋 − cos 𝑚 − 𝑛)(−𝜋 )
2𝜋 𝑚 + 𝑛 𝑚−𝑛
−𝑙 1 1
= (1 − 1) + (1 − 1)
2𝜋 𝑚 + 𝑛 𝑚−𝑛
=0

14
If 𝑚 = 𝑛
𝑙
𝑚𝜋𝑥 𝑛𝜋𝑥
𝑠𝑖𝑛 𝑐𝑜𝑠 𝑑𝑥
−𝑙 𝑙 𝑙
𝑙
1 2𝑚𝜋𝑥
= 𝑠𝑖𝑛 𝑑𝑥
−𝑙 2 𝑙
𝑙
1 2𝑚𝜋𝑥 𝑙
= 𝑐𝑜𝑠 .
2 𝑙 2𝑚𝜋 −𝑙

𝑙
= (cos 2𝑚𝜋 − cos(−2𝑚𝜋))
4𝑚𝜋
𝑙
= 1−1
4𝑚𝜋

=0

15
7. 𝐼 = 𝑒 𝑎𝑥 𝑐𝑜𝑠𝑏𝑥 𝑑𝑥
𝑠𝑖𝑛𝑏𝑥
𝑢=𝑒 𝑎𝑥 𝑑𝑢 = 𝑎𝑒 𝑎𝑥 𝑑𝑥 ; 𝑑𝑣 = 𝑐𝑜𝑠𝑠𝑏𝑥 𝑣=
𝑏
𝑠𝑖𝑛𝑏𝑥 𝑠𝑖𝑛𝑏𝑥 𝑎𝑥
𝐼= 𝑒 𝑎𝑥 . − 𝑎𝑒 𝑑𝑥
𝑏 𝑏
𝑠𝑖𝑛𝑏𝑥 𝑎
= 𝑒 𝑎𝑥 . − 𝑒 𝑎𝑥 𝑠𝑖𝑛𝑏𝑥𝑑𝑥
𝑏 𝑏
𝑠𝑖𝑛𝑏𝑥 𝑎
= 𝑒 𝑎𝑥 . − 𝐽
𝑏 𝑏
𝐽= 𝑒 𝑎𝑥 𝑠𝑖𝑛𝑏𝑥𝑑𝑥
−𝑐𝑜𝑠𝑏𝑥
𝑢1 = 𝑒 𝑎𝑥 𝑑𝑢1 = 𝑒 𝑎𝑥 𝑑𝑥 ; 𝑑𝑣1 = 𝑠𝑖𝑛𝑏𝑥 𝑣1 =
𝑏
−𝑐𝑜𝑠𝑏𝑥 −𝑐𝑜𝑠𝑏𝑥 𝑎𝑥
𝐽= 𝑒 𝑎𝑥 . − 𝑎𝑒 𝑑𝑥
𝑏 𝑏
𝑎𝑥
−𝑐𝑜𝑠𝑏𝑥 𝑎
=𝑒 . + 𝑒 𝑎𝑥 𝑐𝑜𝑠𝑏𝑥𝑑𝑥
𝑏 𝑏
𝑎𝑥
−𝑐𝑜𝑠𝑏𝑥 𝑎
=𝑒 . + 𝐼
𝑠𝑜, 𝑏 𝑏
𝑠𝑖𝑛𝑏𝑥 𝑎 𝑎𝑥 −𝑐𝑜𝑠𝑏𝑥 𝑎 𝑒 𝑎𝑥
𝐼 = 𝑒 𝑎𝑥 . − (𝑒 . + 𝐼) 𝐼= 2 (𝑎𝑐𝑜𝑠𝑏𝑥 + 𝑏𝑠𝑖𝑛𝑏𝑥)
𝑏 𝑏 𝑏 𝑏 𝑎 + 𝑏2
16
8. 𝐼 = 𝑒 𝑎𝑥 𝑠𝑖𝑛𝑏𝑥 𝑑𝑥
−𝑐𝑜𝑠𝑏𝑥
𝑢=𝑒 𝑎𝑥 𝑑𝑢 = 𝑎𝑒 𝑎𝑥 𝑑𝑥 ; 𝑑𝑣 = 𝑠𝑖𝑛𝑏𝑥 𝑣=
𝑏
−𝑐𝑜𝑠𝑏𝑥 −𝑐𝑜𝑠𝑏𝑥 𝑎𝑥
𝐼= 𝑒 𝑎𝑥 . − 𝑎𝑒 𝑑𝑥
𝑏 𝑏
−𝑐𝑜𝑠𝑏𝑥 𝑎
= 𝑒 𝑎𝑥 . + 𝑒 𝑎𝑥 𝑐𝑜𝑠𝑏𝑥𝑑𝑥
𝑏 𝑏
−𝑐𝑜𝑠𝑏𝑥 𝑎
= 𝑒 𝑎𝑥 . + 𝐽
𝑏 𝑏
𝐽= 𝑒 𝑎𝑥 𝑐𝑜𝑠𝑏𝑥𝑑𝑥
𝑠𝑖𝑛𝑏𝑥
𝑢1 = 𝑒 𝑎𝑥 𝑑𝑢1 = 𝑒 𝑎𝑥 𝑑𝑥 ; 𝑑𝑣1 = 𝑐𝑜𝑠𝑏𝑥 𝑣1 =
𝑏
𝑠𝑖𝑛𝑏𝑥 𝑠𝑖𝑛𝑏𝑥 𝑎𝑥
𝐽= 𝑒 𝑎𝑥 . − 𝑎𝑒 𝑑𝑥
𝑏 𝑏
𝑎𝑥
𝑠𝑖𝑛𝑏𝑥 𝑎
=𝑒 . − 𝑒 𝑎𝑥 𝑠𝑖𝑛𝑏𝑥𝑑𝑥
𝑏 𝑏
𝑎𝑥
𝑠𝑖𝑛𝑏𝑥 𝑎
=𝑒 . − 𝐼
𝑠𝑜, 𝑏 𝑏
−𝑐𝑜𝑠𝑏𝑥 𝑎 𝑎𝑥 𝑠𝑖𝑛𝑏𝑥 𝑎 𝑒 𝑎𝑥
𝐼 = 𝑒 𝑎𝑥 . + (𝑒 . − 𝐼) 𝐼= 2 (𝑎𝑠𝑖𝑛𝑏𝑥 − 𝑏𝑐𝑜𝑠𝑏𝑥)
𝑏 𝑏 𝑏 𝑏 𝑎 + 𝑏2
17
 Remarks
Let 𝑢 𝑎𝑛𝑑 𝑣 be two functions of variable 𝑥 then,
𝑢𝑣 𝑑𝑥 = 𝑢𝑣1 − 𝑢1 𝑣2 + 𝑢2 𝑣3 − 𝑢3 𝑣4 + ⋯

Here,
𝑣1 = 𝑣𝑑𝑥 ; 𝑣𝑖 = 𝑣𝑖−1 𝑑𝑥
𝑑𝑢 𝑑𝑖 𝑢
𝑢1 = ; 𝑢𝑖 =
𝑑𝑥 𝑑𝑥 𝑖

Try
𝑥 2 𝑠𝑖𝑛𝑥 𝑑𝑥 = ?

2 2
𝑑(𝑥 2 ) 𝑑2 (𝑥 2 )
𝑥 𝑠𝑖𝑛𝑥 𝑑𝑥 = 𝑥 𝑠𝑖𝑛𝑥𝑑𝑥 − . 𝑠𝑖𝑛𝑥𝑑𝑥 𝑑𝑥 + 𝑠𝑖𝑛𝑥𝑑𝑥 𝑑𝑥 𝑑𝑥 + ⋯
𝑑𝑥 𝑑𝑥 2

= 𝑥 2 −𝑐𝑜𝑠𝑥 − 2𝑥 −𝑐𝑜𝑠𝑥 𝑑𝑥 + 2 (−𝑐𝑜𝑠𝑥)𝑑𝑥 𝑑𝑥

= 𝑥 2 −𝑐𝑜𝑠𝑥 − 2𝑥 −𝑠𝑖𝑛𝑥 + 2 −𝑠𝑖𝑛𝑥 𝑑𝑥

= −𝑥 2 𝑐𝑜𝑠𝑥 + 2𝑥𝑠𝑖𝑛𝑥 + 2𝑐𝑜𝑠𝑥
18