Chapter 10

Fourier Series

Contents
10.1 Orthogonal Functions 3

10.2 Fourier Series 4

10.2.1 Fourier series of functions with period 2π . . . . . . . . . . . . . . . . . . . . . . . . . 4
10.2.2 Fourier series of arbitrary period P = 2L . . . . . . . . . . . . . . . . . . . . . . . . . 11
10.2.3 Fourier series of odd and even functions . . . . . . . . . . . . . . . . . . . . . . . . 15

10.3 Half Range Expansion 18

10.4 Fourier Integrals 20

10.5 Exercise 23

Fourier series is a way to represent a wave like function as the sum of simple sine waves. More
formally, it decomposes any periodic function or periodic signal into the sum of a possibly infinite set of
simple oscillating functions namely sines and cosines.

10.1 Orthogonal Functions

Definition 10.1.1 Functions y1 , y2 , y3 , · · · defined on some interval are called orthogonal on a ≤ x ≤ b
with respect to the weight function ρ(x) > 0 if ab ρ(x)ym (x)yn (x) dx = 0 for m 6= n. ab ρ(x)y2n (x) dx
R R

is called the square of the norm of yn (x) and written as kyn (x)k2 .
1. If kyn (x)k2 = 1, then the set is an orthonormal set.
2. If ρ(x) = 1, then the set is orthogonal on a ≤ x ≤ b.

Remark 10.1.1 • cos nx cos mx = 21 [cos(n + m)x + cos(n − m)x]

• sin nx sin mx = 12 [cos(n − m)x − cos(n + m)x]
• sin nx cos mx = 12 [sin(n + m)x + sin(n − m)x]
Rπ
 Example 10.1 Consider −π cos 3x cos 2x dx, such that

1 π
Z π Z
cos 3x cos 2x dx = (cos 5x + cos x) dx
−π 2 −π
1 1
= sin 5x |π−π + sin x |π−π
10 2
= 0

Therefore cos 3x and cos 2x are orthogonal on −π ≤ x ≤ π. 

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
2 Fourier Series
π
R
 Example 10.2 • −π cos nx cos mx dx = 0 (m 6= n).
Therefore
Rπ
cos nx and cos mx are orthogonal on −π ≤ x ≤ π, where m 6= n.
• −π sin nx sin mx dx = 0 (m 6= n)
Therefore
Rπ
sin nx and sin mx are orthogonal on −π ≤ x ≤ π, where m 6= n.
• −π sin nx cos mx dx = 0 (m 6= n) or (m = n)
Therefore sin nx and cos mx are orthogonal on −π ≤ x ≤ π.


 Example 10.3 In general, the set of functions 1, cos(nx), where n = 1, 2, 3, · · · or 1, sin(nx), where
n = 1, 2, 3, · · · or 1, cos(nx), sin(nx), where n = 1, 2, 3, · · · in the interval a ≤ x ≤ a + 2π with weight
functions ρ(x) = 1 for any real constant a are orthogonal functions. 

10.2 Fourier Series

Definition 10.2.1 Let y0 , y1 , y2 , · · · be an orthogonal set with respect to the weight function ρ(x) on
an interval a ≤ x ≤ b. Let f (x) be a given function that can be represented in terms of yn (x) by a
convergent series,
∞
f (x) = ∑ am ym (x) = c0 y0 + c1 y1 + c2 y2 + · · ·
m=0

This is called an orthogonal expansion or the generalized Fourier series, where c0 , c1 , c2 , · · · is Fourier
coefficients of f (x) respect to y0 , y1 , y2 , · · · .

10.2.1 Fourier series of functions with period 2π

Definition 10.2.2 A function f is said to be a periodic function with period p if p is the least positive
numbers such that f (x + p) = f (x) for all x in the domain of f .

 Example 10.4 f (x) = sin x and g(x) = cos x are the familiar periodic functions with period p = 2π but

x, x2 , x3 , ex , cosh x and ln x are examples of non−periodic function. 

Proposition 10.2.1 The constant function h(x) = c is a periodic function, since h(x + p) = c = h(x)
for all p ∈ (0, ∞).

Proposition 10.2.2 If f (x) are periodic of p, it also has the period of 2p because

f (x + 2p) = f (x + p + p)
= f ((x + p) + p)
= f (x + p)

It follows that f (x + np) = f (x) for all x in the domain of f and all integers n.

Proposition 10.2.3 If f and g are periodic functions with period p, then h(x) = a f (x) + bg(x), for
some constants a and b, is a periodic function with period p. Since f (x + p) = f (x) and g(x + p) = g(x)

h(x + p) = a f (x + p) + bg(x + b)
= a f (x) + bg(x)
= h(x)

Therefore, h(x) is a periodic function with period p.

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
10.2 Fourier Series 3

Definition 10.2.3 A functional series of the form

a0
g(x) = + a1 cos x + b1 sin x + a2 cos 2x + b2 sin 2x + · · ·
2
+an cos nx + bn sin nx + · · ·
∞
a0
= + ∑ an cos nx + bn sin nx
2 n=1

is called a trigonometric series: where a0 , an , bn (n = 1, 2, 3, · · · ) are real constants, are called the
coefficients of the trigonometric series.
∞
a0
If the series converges, say to a function f (x); that is, 2 + ∑ (an cos nx + bn sin nx) = f (x). Then f is a
n=1
period function with period p = 2π, by the above proposition. Thus f (x + 2π) = f (x) for all x in the
∞
a0
domain of f , where f (x) = 2 + ∑ (an cos nx + bn sin nx).
n=1

Definition 10.2.4 The Fourier series for the periodic function f (x) in an interval a < x < a + 2π is
given by
∞
a0
f (x) = + ∑ (an cos nx + bn sin nx)
2 n=1
and the coefficients a0 , an , bn (n = 1, 2, 3, · · · ) are called the Fourier coefficients.

Basic Rules
Basic rules of integrationR involving sine and cosine functions.
1. aa+2π cos nx dx = aa+2π sin nx dx = 0, n = 1, 2, 3, · · ·
R

2. aa+2π sin nx sin mx dx = 0 (n 6= m)

R

3. aa+2π cos nx cos mx dx = 0 (n 6= m)

R

4. aa+2π cos mx sin nx dx = 0 (n 6= m or n = m)

R

5. aa+2π cos2 nx dx = aa+2π sin2 nx dx = π, n 6= 0

R R

Z a+2π Z a+2π
1 + cos 2nx
cos2 nx dx = dx
a a 2
1 1
= (x + sin 2nx)|aa+2π
2 2n
1
= (a + 2π − a)
2
= π

6. sin(2n + 1) π2 = cos nπ = (−1)n , where n = 1, 2, 3, · · · and cos(2n + 1) π2 = sin nπ = 0, where

n = 1, 2, 3, · · ·

Theorem 10.2.4 Euler’s Formula

∞
a0
The Fourier coefficients in f (x) = 2 + ∑ (an cos nx + bn sin nx) are given by
n=1
1 R a+2π
• a0 = π Ra f (x) dx
1 a+2π
• an = π Ra f (x) cos nx dx, where n = 1, 2, 3, · · · and
1 a+2π
• bn = π a f (x) sin nx dx, where n = 1, 2, 3, · · ·

Corollary 10.2.5 1. If a = 0, the interval becomes 0 < x < 2π, and Euler’s formula are given by
1 R 2π
• a0 = π R0 f (x) dx
1 2π
• an = π 0 f (x) cos nx dx and

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
4 Fourier Series

• bn = π1 02π f (x) sin nx dx

R

2. If a = −π, the interval becomes −π < x < π, and the Euler’s Formula are given by
• a0 = π1 −π
Rπ
f (x) dx
• an = π1 −π
Rπ
f (x) cos nx dx and
1 Rπ
• bn = π −π f (x) sin nx dx

When f (x) has finite number of discontinuities in any one period, for instance if c is a point of discontinuity
in an interval (a, a + 2π), f (x) is defined by

φ (x) for a < x < c
f (x) =
ψ(x) for c < x < a + 2π
then
1 Rc 1 R a+2π
• a0 = π Ra φ (x) dx + π c ψ(x) dx
1 c 1 R a+2π
• an = π Ra φ (x) cos nx dx + π R c ψ(x) cos nx dx and
1 c 1 a+2π
• bn = π a φ (x) sin nx dx + π c ψ(x) sin nx dx
Example 10.5 Find the Fourier coefficients of the periodic function f (x). The formula is f (x) = ( π−x 2

2 ) ,
∞
π2 cos nx
where 0 ≤ x ≤ 2π and f (x + 2π) = f (x), and show that f (x) = 12 +∑ 2
.
n=1 n
∞
a0
Solution: The Fourier serves for f in (0, 2π) is f (x) = 2 + ∑ an cos nx + bn sin nx, where
n=1

1 2π
Z
a0 = f (x) dx
π 0
1 2π π − x 2
Z
= ( ) dx
π 0 2
Z 2π
1
= (π 2 x − 2πx + x2 ) dx
4π 0
1 2 1
= (π x − πx2 + x3 ) |2π 0
4π 3
π 2
=
6

1 2π
Z
an = f (x) cos nx dx
π 0
1 2π π − x 2
Z
= ( ) cos nx dx
π 0 2
Z 2π
1
= (π − x)2 cos nx dx
4π 0
1 (π − x)2 − cos nx sin nx
= ( sin nx + 2(π − x)( ) + 2( 3 )) |2π
0
4π n n2 n
1 2π 2π
= ( + 3)
4π n2 n
1
= and
n2

1 2π
Z
bn = f (x) sin nx dx
π 0
1 2π π − x 2
Z
= ( ) sin nx dx
π 0 2
1 (π − x)2 − cos nx − sin nx cos nx
= ( ( 2
) 2
+ 2(π − x)( 2
) + 2( 3 )) |2π
0
4π n n n n
= 0

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
10.2 Fourier Series 5
∞
π2 cos nx
Therefore, f (x) = 12 +∑ 2
. 

n=1 n
 Example 10.6 Find the Fourier coefficients of the periodic function f (x) in the figure. The formula is

−k for − π < x < 0
f (x) = and f (x + 2π) = f (x)
k for 0 < x < π

and show that 1 − 13 + 15 − 17 + · · · = π4 .

Figure 10.1:

∞
a0
Solution: The Fourier serves for f in (0, 2π) is f (x) = 2 + ∑ an cos nx + bn sin nx, where
n=1

1 2π
Z
a0 = f (x) dx
π 0
Z 0
1 1 π
Z
= f (x)dx + f (x) dx
π −π π 0
1 0 1 π
Z Z
= −kdx + k dx
π −π π 0
1 0 1 π
Z Z
= − kdx + k dx
π −π π 0
= 0

1 0 1 π
Z Z
an = f (x) cos nxdx + f (x) cos nx dx
π π π 0
Z 0
1 1
Z π
= −k cos nxdx + k cos nx dx
π π π 0
1 0 1 π
Z Z
= − k cos nxdx + k cos nx dx
π π π 0
= 0 and

1 0 1 π
Z Z
bn = f (x) sin nxdx + f (x) sin nx dx
π −π π 0
Z 0
1 1
Z π
= −k sin nxdx + k sin nx dx
π −π π 0
1 0 1 π
Z Z
= − k sin nxdx + k sin nx dx
π −π π 0
2k
= [1 − cos nθ ]
nπ
 4k
nπ for odd n
=
0 for even n
4k 4k 4k
that is, b1 = , b2 = 0, b3 = , b4 = 0, b5 = ···
π 3π 5π

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
6 Fourier Series
4k 1
Therefore, f (x) = π (sin x + 3 sin 3x + 15 sin 5x + · · · ).
Here s1 = 4k
π sin x

4k
Figure 10.2: s1 = π sin x

4k 1
s2 = π (sin x + 3 sin 3x)

4k 1
Figure 10.3: s2 = π (sin x + 3 sin 3x)

4k 1
s3 = π (sin x + 3 sin 3x + 15 sin 5x)

4k 1
Figure 10.4: s3 = π (sin x + 3 sin 3x + 15 sin 5x)

Let x = π2 , then

π 4k π 1 π 1 π
f( ) = (sin( ) + sin 3( ) + sin 5( ) + · · · )
2 π 2 3 2 5 2
4k 1 1 1 π
k = (1 − + − + · · · ) since f (x) = k at x =
π 3 5 7 2
π 1 1 1
⇒ = 1− + − +···
4 3 5 7


Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
10.2 Fourier Series 7

The value of f (x) at a single point does not affect the integral, hence we can leave f (x) undefined at the
end points and at the point of discontinuity if there is.

Dirichlet’s Conditions

Suppose that:
1. f (x) is defined and single–valued except possibly at a finite number of points in (a, a + 2π).
2. f (x) is periodic outside (a, a + 2π) with period 2π.
3. f (x) and f 0 (x) are sectionally(piecewise) continuous in (a, a + 2π).
∞
a0
Then, the series 2 + ∑ (an cos nx + bn sin nx) with coefficients
n=1
1 R a+2π
• a0 = π Ra f (x) dx
1 a+2π
• an = π Ra f (x) cos nx dx and
1 a+2π
• bn = π a f (x) sin nx dx
converges to
• f (x) if x is a point of continuity
• f (c) if c is a point of discontinuity, where

lim f (x) + lim f (x)

x→c− x→c+
= f (c)
2
∞
a0
= + ∑ (an cos nc + bn sin nc)
2 n=1

 Example 10.7 Find the Fourier Series expansion for


−π for − π < x < 0
f (x) = and f (x + 2π) = f (x)
x for 0 < x < π

∞
1 1 1 π2
1 1
and show that ∑ (2n + 1)2 = 12 + 32 + 52 + 72 + · · · = 8
.
n=0
Solution: The Fourier series expansion in the interval (−π, π) is

∞
a0
f (x) = + ∑ (an cos nx + bn sin nx)
2 n=1

Where

1 π
Z
a0 = f (x) dx
π −π
Z 0
1
Z π
= [ f (x)dx + f (x) dx]
π −π 0
1 0
Z Z π
= [ −πdx + x dx]
π −π 0
1 1
= [−πx|0−π + x2 |π0 ]
π 2
1 1 2
= [− π ]
π 2
1
= − π
2

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
8 Fourier Series

1 π
Z
an = f (x) cos nx dx
π −π
Z 0
1
Z π
= [ f (x) cos nxdx + f (x) cos nx dx]
π −π 0
1 0
Z Z π
= [ −π cos nxdx + x cos nx dx]
π −π 0
1 π x 1
= [− sin nx|0−π + ( sin nx + 2 cos nx)|π0 ]
π n n n
1 π π 1 1
= [− sin nπ + ( sin nπ + 2 cos nπ − 2 )]
π n n n n
1
= [cos nπ − 1]
πn2
1
= [(−1)n − 1]
πn2

1 π
Z
bn = f (x) sin nx dx
π −π
1 0
Z Z π
= [ f (x) sin nxdx + f (x) sin nx dx]
π −π 0
1 0
Z Z π
= [ −π sin nxdx + x sin nx dx]
π −π 0
1 π x 1
= [ cos nx|0−π + (− cos nx + 2 sin nx)|π0 ]
π n n n
1 π π
= [ (1 − cos nπ) − cos nπ]
π n n
1 π
= [ (1 − cos nπ − cos nπ)]
π n
1
= [1 − 2 cos nπ]
n
1
= [1 − 2(−1)n ]
n

So

∞
a0
f (x) = + ∑ (an cos nx + bn sin nx)
2 n=1
∞
π 1 1
= − + ∑ ( 2 [(−1)n − 1] cos nx + [1 − 2(−1)n ] sin nx)
4 n=1 πn n
π 2 2 2
= − + (− cos x − cos 3x − cos 5x − · · · ) +
4 π 9π 25π
1 1 3
(3 sin x − sin 2x + sin 3x − sin 4x + sin 5x − · · · )
2 4 5

At x = 0, by dirichelet’s condition

lim f (x) = lim −π

x→0− x→0−
= −π
lim f (x) = lim x
x→0+ x→0−
= 0 and
π 1
f (0) = − + ( 2 [(−1)n − 1])
4 πn

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
10.2 Fourier Series 9

Therefore,

lim f (x) + lim f (x)

x→c− x→c+
= f (c)
2
∞
a0
= + ∑ (an cos nc + bn sin nc)
2 n=1
−π + 0 π 1
= − + ( 2 [(−1)n − 1])
2 4 πn
−π π 1
⇒ + = [(−1)n − 1]
2 4 πn2
−π 1
⇒ = [(−1)n − 1]
4 πn2
−π 2 1
⇒ = [(−1)n − 1]
4 n2
−π 2 2 2 2 2
⇒ = −2 − − 2 − 2 − 2 − · · ·
4 3 5 7 9
−π 2 1 1 1 1
⇒ = −2(1 + + 2 + 2 + 2 + · · · )
4 3 5 7 9
π2 1 1 1 1
⇒ = 1+ + 2 + 2 + 2 +···
8 3 5 7 9
π2 ∞
1
⇒ = ∑ (2n + 1)2
8 n=0

∞
π2 1
Therefore, 8 = ∑ (2n + 1)2 . 

n=0

10.2.2 Fourier series of arbitrary period P = 2L

The Fourier series Expansion of f (x) in the interval (a, a + 2L) is given by
∞
a0 nπ nπ
f (x) = + ∑ (an cos( )x + bn sin( )x)
2 n=1 L L

are given by
• a0 = L1 aa+2L f (x) dx
R

• an = L1 aa+2L f (x) cos( nπ

R
L )x dx and
• bn = L1 aa+2L f (x) sin( nπ
R
L )x dx

Corollary 10.2.6 1. Putting a = 0 in these formula, we get the corresponding Fourier Coefficients
for the interval (0, 2L) are
• a0 = L1 02L f (x) dx
R

• an = L1 02L f (x) cos( nπ

R
L )x dx and
• bn = L1 02L f (x) sin( nπ
R
L )x dx
2. Putting a = −LRL
in the above formula, we get the results
• a0 = L1 −L f (x) dx
RL
• an = L1 −L f (x) cos( nπ
L )x dx and
1 RL
• bn = L −L f (x) sin( nπ L )x dx
for the Fourier coefficients in the interval (−L, L).

 Example 10.8 Find the Fourier series of the periodic function f (x) of period 2, where

−1 for − 1 < x < 0
f (x) = and f (x + 2L) = f (x)
2x for 0 < x < 1

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
10 Fourier Series
2
and show that 112 + 312 + 512 + 712 + · · · = π8 and 112 − 312 + 512 − 712 + · · · = π4
Solution: The Fourier series for f (x) in (−1, 1) in the period P = 2L = 2; that is, L = 1 is given by

∞
a0
f (x) = + ∑ (an cos(nπ)x + bn sin(nπ)x
2 n=1

where

Z 1
1
a0 = f (x) dx
1 −1
Z 0 Z 1
= f (x)dx + f (x) dx
−1 0
Z 0 Z 1
= −1dx + 2x dx
−1 0
Z 0 Z 1
= − dx + 2x dx
−1 0
= −1|0−1 + x2 |10
= −1 + 1
= 0

Z 1
1
an = f (x) cos(nπ)x dx
1 −1
Z 0 Z 1
= f (x) cos(nπ)xdx + f (x) cos(nπ)x dx
−1 0
Z 0 Z 1
= −1 cos(nπ)xdx + 2x cos(nπ)x dx
−1 0
Z 0 Z 1
= − cos(nπ)xdx + 2x cos(nπ)x dx
−1 0
1 x 1
= − sin(nπ)x|0−1 + 2( sin(nπ)x + cos(nπ)x)|10
nπ nπ (nπ)2
1 1
= 2[ 2
cos nπ − ]
(nπ) (nπ)2
2
= [cos nπ − 1]
(nπ)2
2
= [(−1)n − 1] and
(nπ)2

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
10.2 Fourier Series 11
Z 1
1
bn = f (x) sin(nπ)x dx
1 −1
Z 0 Z 1
= f (x) sin(nπ)xdx + f (x) sin(nπ)x dx
−1 0
Z 0 Z 1
= −1 sin(nπ)xdx + 2x sin(nπ)x dx
−1 0
Z 0 Z 1
= − sin(nπ)xdx + 2x sin(nπ)x dx
−1 0
1 x 1
= cos(nπ)x|0−1 + 2(− cos(nπ)x + sin(nπ)x)|10
nπ nπ (nπ)2
1 1
= [1 − (−1)n ] + 2[− (−1)n ]
nπ nπ
1 1 2
= − (−1)n − (−1)n
nπ nπ nπ
1
= [1 − (−1)n − 2(−1)n ]
nπ
1
= [1 − 3(−1)n ]
nπ

∞
2 1
Therefore, f (x) = ∑ ( (nπ)2 [(−1)n − 1] cos(nπ)x + nπ [1 − 3(−1)n ] sin(nπ)x).
n=1
At x = 0, by dirichelet’s condition

lim f (x) = lim −1

x→0− x→0−
= −1
lim f (x) = lim 2x
x→0+ x→0−
= 0 and
2 ∞ 1
f (0) = ∑ n2 [(−1)n − 1]
π 2 n=1

Therefore,

lim f (x) + lim f (x)

x→0− x→0+
= f (0)
2
∞
a0
= + ∑ (an cos n0 + bn sin n0)
2 n=1
−1 + 0 2 ∞ 1
2
= ∑ n2 [(−1)n − 1]
π 2 n=1
−1 2 ∞ 1
⇒
2
= ∑ n2 [(−1)n − 1]
π 2 n=1
−π 2 2 2 2 2
⇒ = −2 − 2
− 2 − 2 − 2 −···
4 3 5 7 9
−π 2 1 1 1 1
⇒ = −2(1 + 2 + 2 + 2 + 2 + · · · )
4 3 5 7 9
π2 1 1 1 1
⇒ = 1+ 2 + 2 + 2 + 2 +···
8 3 5 7 9
π2 ∞
1
⇒ = ∑ 2
8 n=0 (2n + 1)

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
12 Fourier Series

At x = 21 , since f (x) is continuous at x = 1

2

∞
2 1
f (x) = ∑ ( nπ 2 [(−1)n − 1] cos(nπ)x + nπ [1 − 3(−1)n ] sin(nπ)x)
n=1
∞
1 2 1 1 1
f( ) =
2 ∑ ( nπ 2 [(−1)n − 1] cos(nπ) 2 + nπ [1 − 3(−1)n ] sin(nπ) 2 )
n=1
1 4 4 4 1
1 = 2
[4 − + − + · · · ] since f ( ) = 1
π 3 5 7 2
π 1 1 1
= 1− + − +···
4 3 5 7

 Example 10.9 Find the Fourier series of the function f (x) having p = 4 is given by


 0 for − 2 < x < −1
f (x) = k for − 1 < x < 1 and f (x + 2L) = f (x)
0 for 1 < x < 2


Solution: The Fourier series f (x) in (−2, 2) in the period P = 2L = 4; that is, L = 2 is given by

∞
a0 nπ nπ
f (x) = + ∑ (an cos( )x + bn sin( )x)
2 n=1 2 2

where

1 2
Z
a0 = f (x) dx
2 −2
1 −1
Z Z 1 Z 2
= [ f (x)dx + f (x)dx + f (x) dx]
2 −2 −1 1
Z −1 Z 0 Z 2
1
= [ 0dx + kdx + 0 dx]
2 −2 −1 0
= k

1 2 nπ
Z
an = f (x) cos( )x dx
2 −2 2
1 −1
Z 1 Z 2
nπ nπ nπ
Z
= [ f (x) cos( )x dx + f (x) cos( )x dx + f (x) cos( )x dx]
2 −2 2 −1 2 1 2
Z −1 Z 1 Z 2
1 nπ nπ nπ
= [ 0 cos( )x dx + k cos( )x dx + 0 cos( )x dx]
2 −2 2 −1 2 1 2
1 1 nπ
Z
= [ k cos( )x dx]
2 −1 2
k 1 nπ
Z
= cos( )x dx
2 −1 2
k nπ 1
= sin( )x|−1
nπ 2
2k nπ
= sin( )
nπ 2
2k

 nπ if n = 1, 5, 9, · · ·
= 0 if n is even and
 2k
− nπ if n = 3, 7, 11, · · ·

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
10.2 Fourier Series 13

1 2 nπ
Z
bn = f (x) sin( )x dx
2 −2 2
1 −1
Z 1 Z 2
nπ nπ nπ
Z
= [ f (x) sin( )x dx + f (x) sin( )x dx + f (x) sin( )x dx]
2 −2 2 −1 2 1 2
Z −1 Z 1 Z 2
1 nπ nπ nπ
= [ 0 sin( )x dx + k sin( )x dx + 0 sin( )x dx]
2 −2 2 −1 2 1 2
Z 1
1 nπ
= k sin( )x dx
2 −1 2
−k nπ 1
= cos( )x|−1
nπ 2
= 0

Therefore,
∞
a0 nπ nπ
f (x) = + ∑ (an cos( )x + bn sin( )x)
2 n=1 2 2
∞
k 2k nπ nπ nπ
= + ∑ ( sin( ) cos( )x + 0 sin( )x)
2 n=1 nπ 2 2 2
∞
k 2k nπ nπ
= + ∑ ( sin( ) cos( )x)
2 n=1 nπ 2 2
k 2k π 1 3π 1 5π 1 7π
= + [cos( )x − cos( )x + cos( )x − cos( )x + · · · ]
2 π 2 3 2 5 2 7 2


10.2.3 Fourier series of odd and even functions

Definition 10.2.5 A function f (x) is said to be odd if and only if f (−x) = − f (x) and a function is
said to be even if and only if f (−x) = f (x).

 Example 10.10 The functions sin(nx) and tan(nx) are odd functions. Graph of odd function is symmet-
ric about the origin. 

 Example 10.11 The functions cos(nx), x2 , csc(nx) are even functions. Graphs of even functions are
symmetric about y-axis. 

Proposition 10.2.7 If f (x) is a periodic function with period p = 2L, then

Z L  RL
2 0 f (x)dx if f is even
f (x)dx =
−L 0 if f is odd

Recall that a periodic function f (x) defined in (−L, L) can be represented by the Fourier series
∞
a0 nπ nπ
f (x) = + ∑ (an cos x + bn sin x)
2 n=1 L L

where
RL
• a0 = L1 −L f (x) dx
1 RL
• an = L −L f (x) cos( nπ L )x dx and
RL
• bn = L1 −L f (x) sin( nπ
L )x dx
When f (x) is even function
L
• a0 = L1 −L f (x) dx = L2 0L f (x) dx
R R
L 2 RL
• an = L1 −L f (x) cos( nπ nπ nπ
R
L )x dx = L 0 f (x) cos( L )x dx since f (x) cos( L )x is even and
1 L
R nπ nπ
• bn = L −L f (x)( L )x dx = 0 since f (x) sin( L )x is odd.

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
14 Fourier Series

Figure 10.5: Even function

Therefore,
Z L ∞ Z L
1 2 nπ nπ
f (x) = f (x) dx + ∑ [ f (x) cos( )x dx] cos( )x
L 0 n=1 L 0 L L
is fourier cosine series.
When f (x) is odd function

Figure 10.6: Odd function

L
• a0 = L1 −L
R
f (x) dx = 0
1 RL
• an = L −L f (x) cos( nπ nπ
L )x dx = 0 Rsince f (x) cos( L )x is odd and
L
• bn = L1 −L f (x) sin( nπ )x dx = L2 0L f (x) sin( nπ nπ
R
Z LL L )xdx since f (x) sin( L )x is even.
∞
2 nπ nπ
Therefore, f (x) = ∑ [ f (x) sin( )xdx] sin( )x is Fourier sine series.
n=1 L 0 L L

Theorem 10.2.8 Sum and Scalar multiple

• The Fourier coefficients of sums f1 + f2 are the sum of the corresponding Fourier coefficients of
f1 and f2 .
• The Fourier coefficients of c · f are c times the corresponding Fourier coefficients of f .

 Example 10.12 Find a series representation of f (x) = x2 in the interval (−`, `) and f (x + 2L) = f (x).
2 2
Show that π12 = 112 − 212 + 312 − 412 + · · · and π6 = 112 + 212 + 312 + 412 + · · · .
Solution: The Fourier series representation in the interval (−`, `); that is, in the period P = 2l is given by
∞
a0 nπ nπ
f (x) = + ∑ (an cos( )x + bn sin( )x)
2 n=1 ` `

Where R`
• a0 = 1` −` f (x) dx
1 R`
• an = ` −` f (x) cos( nπ ` )x dx and
R`
• bn = 1` −` f (x) sin( nπ
` )x dx

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
10.2 Fourier Series 15

Here our function f (x) = x2 is even function because of this bn = 0. Now

1 `
Z
a0 = f (x) dx
` −`
Z `
2
= f (x) dx
` 0
2 ` 2
Z
= x dx
` 0
21 3 `
= x |
`3 0
2 2
= `
3

1 ` nπ
Z
an = f (x) cos( )x dx
` −` `
Z `
2 nπ
= x2 cos( )x dx
` 0 `
2 2 ` nπ `2 nπ 2`3 nπ
= [x sin( )x + 2x 2 2 cos( )x − 3 3 sin( )x]|`0
` nπ ` n π ` n π `
2 2`3
= [ ((−1)n )]
` n2 π 2
4`2 (−1)n
= [ ]
π 2 n2

∞
4`2 (−1)n nπ
Therefore, f (x) = 13 `2 + ∑ 2
[ 2 ] cos( )x.
n=1 π n `
At x = 0, we have

`2 ∞
4`2 (−1)n nπ
f (0) = + ∑ 2 [ 2 ] cos( )(0)
3 n=1 π n `
1 2 ∞ 4`2 (−1)n nπ
0 = ` + ∑ 2 [ 2 ] cos( )(0) since f (0) = 02
3 n=1 π n `
1 ∞
4`2 (−1)n
⇒ − `2 = ∑ 2 [ n2 ]
3 n=1 π
π2 ∞
(−1)n+1
⇒ = ∑
12 n=1 n2
π2 1 1 1 1 1
⇒ = 1− + − + − +···
12 22 32 42 52 62

At x = `, we have

lim f (x) = lim x2

x→`− x→`−
2
= `
lim f (x) = lim x2
x→`+ x→`−
2
= ` and
1 2 ∞ 4`2 (−1)n nπ
f (`) = l + ∑ 2 [ 2 ] cos( )`
3 n=1 π n `

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
16 Fourier Series

Therefore,
lim f (x) + lim f (x)
x→`− x→`+
= f (`)
2
1 2 ∞ 4`2 (−1)n nπ
= ` + ∑ 2 [ 2 ] cos( )`
3 n=1 π n `
`2 + `2 1 2 ∞ 4`2 (−1)n
= ` + ∑ 2 [ 2 ](−1)n
2 3 n=1 π n
1 2 ∞ 4`2 (−1)n
`2 = ` + ∑ 2 [ 2 ](−1)n
3 n=1 π n
2 2 4`2 ∞ 1
` = ∑ n2
3 π 2 n=1
π2 ∞
1
= ∑ n2
6 n=1


10.3 Half Range Expansion

In many problems of physics and engineering there is a practical need to apply a Fourier series to non–
periodic function f (x) on the interval 0 < x < T . Because of physical or mathematical consideration,
it may be possible to extend f (x) over the interval −T < x < T , making it periodic of period P = 2T .
The following figures illustrate the odd and even extensions of f (x) which have Fourier sine and Fourier
cosine series, respectively.
1. The figure of the f (x)

Figure 10.7: f (x)

2. The odd extensions of the function f (x)

Figure 10.8: Odd extensions of the function f (x)

3. The even extension of the function f (x)

Example 10.13 Express f (x) = x as a half range sine series in 0 < x < 2 and cosine series in 0 < x < 2.
Solution: First we have to find for sine series. Let us extend the function f (x) in the interval 0 < x < 2,

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
10.3 Half Range Expansion 17

Figure 10.9: Even extension of the function f (x)

so that the new function is symmetric about the origin and, therefore, represent an odd function in
−2 < x < 2.
Hence, the Fourier series for f (x) over the full period (−2, 2) will contain only sine series terms given by
∞
nπ
f (x) = ∑ bn sin( )x
n=1 2

Where
2 2 nπ
Z
bn = f (x) sin( )x dx
2 0 2
2 2 nπ
Z
= x sin( )x dx
2 0 2
2 nπ 2 4 nπ
= x cos( )x|0 + 2 2 sin( )x|20
nπ 2 n π 2
−4(−1)n
= +0
nπ
4(−1)n+1
=
nπ
Therefore,

4(−1)n+1
∞
nπ
f (x) = ∑ sin( )x
n=1 nπ 2
4 π 1 π 1 3π
= [sin( )x − sin( )x + sin( )x − · · · ]
π 2 2 1 3 2
Next, let us extend the function f (x) in the interval 0 < x < 2 for cosine series, so that the new function is
symmetric about the y-axis and, therefore, represent an even function in −2 < x < 2.
Hence, the Fourier series for f (x) over the full period (−2, 2) will contain only cosine series terms given
by
∞
a0 nπ
f (x) = + ∑ an cos( )x
2 n=1 2
Where
2 2
Z
a0 = f (x) dx
2 0
Z 2
2
= x dx
2 0
= 2
2 2 nπ
Z
an = f (x) cos( )x dx
2 0 2
Z 2
2 nπ
= x cos( )x dx
2 0 2
2 nπ 2 4 nπ
= x sin( )x|0 − 2 2 cos( )x|20
nπ 2 n π 2
−4((−1)n − 1)
=
n2 π 2

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
18 Fourier Series

Therefore,

∞
a0 nπ
f (x) = + ∑ an cos( )x
2 n=1 2
4 2 π 2 3π 2 5π
= 1+ [− cos( )x − 2 cos( )x − 2 cos( )x − · · · ]
π 2 12 2 3 2 5 2
8 1 π 1 3π 1 5π
= 1 − 2 [ 2 cos( )x + 2 cos( )x + 2 cos( )x + · · · ]
π 1 2 3 2 5 2

10.4 Fourier Integrals

Fourier series are powerful tools for problems involving functions that are periodic or are of interests on a
finite interval only, but many problems involve functions that are non periodic and are of interest on the
whole x-axis. So, we can extend the method of Fourier series to such functions. This idea will lead to
Fourier integrals.
Consider a function f (x) which satisfies the Dirichlet’s conditions in the interval (−L, L) so that

∞
a0 nπx nπx
f (x) = + ∑ (an cos + bn sin )
2 n=1 L L

Where
L
• a0 = L1 −L
R
f (t) dt
1 L
• an = L −L f (t) cos( nπ
R
RL L )t dt and
• bn = L1 −L f (t) sin( nπ
L )t dt

Substituting the values of a0 , an and bn in the Fourier series expansion, we get the form

Z L ∞ Z L
1 1 nπ nπ
fL (x) = f (t) dt + ∑ ( f (t) cos( )t dt cos( )x
2L −L n=1 L −L L L
1 L nπ nπ
Z
+ f (t) sin( )t dt sin( )x)
L −L L L
1 L ∞
1 L nπ
Z Z
= f (t) dt + ∑ ( f (t) cos( )(t − x) dt)
2L −L n=1 L −L L

nπ π 1 4λ
Let λn = L , then 4λ = λ − n + 1 − λn = L ⇒ L = π .
So,

f (x) = lim fL f (x)

L→∞
Z L ∞ Z L
1 1 nπ
= lim f (t) dt + ∑ ( f (t) cos( )(t − x) dt)
L→∞ 2L −L L −L L
n=1

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
10.4 Fourier Integrals 19
∞
| f (t)| dt converges; that is, f (x) is absolutely integrable in (−∞, ∞), then 1st term on the right side
R
If −∞
1 R∞ R∞
of the above equation approaches to 0 as L → ∞, since 2L −∞ f (t) dt| ≤ −∞ | f (t)| dt. Hence
∞ Z L
1 nπ
f (x) = ∑ ( f (t) cos( )(t − x) dt)
n=1 L −L L
∞
4L
Z ∞
= lim ∑ f (t) cos λn (t − x) dt)
4L→0 n=1 π −∞
∞
1
Z ∞
= lim ∑ 4L f (t) cos λn (t − x) dt)
π 4L→0 n=1 −∞
1 ∞ ∞
Z Z
= ( f (t) cos λ (t − x) dt)dλ
π 0 −∞
1
Z ∞ Z ∞
= ( f (t) cos(λt − λ x) dt)dλ
π 0 −∞
1 ∞ ∞
Z Z
= ( f (t)(cos λt cos λ x + sin λt sin λ x) dt) dλ
π 0 −∞
1
Z ∞ Z ∞
= ( ( f (t) cos λt cos λ x + f (t) sin λt sin λ x) dt) dλ
π 0 −∞
1 1 ∞
Z ∞ Z ∞ Z
= ( ( f (t) cos λt cos λ x + f (t) sin λt sin λ x) dt) dλ
0 π −∞ π −∞
1
Z ∞ Z ∞
= ( ( f (t) cos λt dt) cos λ x dx +
π
0
| −∞ {z }
A(λ )
1 ∞
Z ∞ Z
( f (t) sin λt dt) sin λ x dλ
0 π −∞
| {z }
B(λ )
Z ∞ Z ∞
= A(λ ) cos λ x + B(λ ) sin λ x dλ
Z0 ∞ 0

= (A(λ ) cos λ x + B(λ ) sin λ x) dλ

0

Therefore, the Fourier integral of f (x) is represented by

Z ∞
f (x) = (A(λ ) cos λ x + B(λ ) sin λ x) dλ
0
1 R∞ 1 R∞
Where A(λ ) = π −∞ f (x) cos λ x dx and B(λ ) = π −∞ f (x) sin λ x dx.

Remark 10.4.1 If function f is continuous at x, then

Z ∞
f (x) = (A(λ ) cos λ x + B(λ ) sin λ x) dλ
0

If function is not continuous at c, then

lim f (x) + lim f (c)
x→c+ x→c−
= f (c)
2 Z ∞
= (A(λ ) cos λ c + B(λ ) sin λ c) dλ
0

Fourier Sine and Cosine Integral

• If f (x) is an odd function, then f (t) cos λt is also an odd function while f (t) sin λt is even. Then
A(λ ) on the right side of the above equation vanishes and
Z ∞
f (x) = B(λ ) sin λ x dλ
0

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
20 Fourier Series

Which is known as the Fourier Sine integrals.

• If f (x) is an even function, then B(λ ) = 0 the above integrals takes the form of
Z ∞
f (x) = A(λ ) cos λ x dλ
0

Which is known as the Fourier Cosine integrals.

 Example 10.14 Express 
1 if |x| < 1
f (x) =
0 if |x| > 1
R ∞ sin λ cos λ x
as a Fourier sine integral and hence evaluate 0 λ dλ .
Solution: From the Fourier integral
1 ∞
Z
A(λ ) = f (x) cos λ x dx
π −∞
Z −1 Z 1
1
Z ∞
= [ 0 cos λ x dx + 1 cos λ x dx + 0 cos λ x dx]
π −∞ −1 1
1 1
Z
= cos λ x dx
π −1
sin λ x 1
= |
πλ −1
2 sin λ
= and
πλ
1
Z ∞
B(λ ) = f (x) sin λ x dx
π −∞
1 −1
Z Z 1 Z ∞
= [ 0 sin λ x dx + 1 sin λ x dx + 0 sin λ x dx]
π −∞ −1 1
1 1
Z
= sin λ x dx
π −1
= 0

Therefore, the Fourier integral is f (x) = 0∞ sin λ λcos λ x dλ .

R

• Case I: If |x| = 1; that is, at x = 1 and x = −1, then

lim f (x) + lim f (x)

x→1+ x→1−
= f (1)
2
2 ∞ sin λ cos λ x
Z
= dλ
π 0 λ
1 2 ∞ sin λ cos λ x
Z
⇒ = dλ
2 π 0 λ
sin λ cos λ x
Z ∞
π
⇒ = dλ
4 0 λ
• Case II: If |x| > 1; that is, for x > 1 and x < −1, then

sin λ cos λ x
Z ∞
dλ = 0
0 λ
• Case III: If |x| < 1; that is, for −1 < x < 1, then

2 sin λ cos λ x
Z ∞
1 = dλ
π λ 0
sin λ cos λ x
Z ∞
π
⇒ = dλ
2 0 λ


Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
10.5 Exercise 21

10.5 Exercise
RL kπx RL kπx
1. Prove −L sin L dx = −L cos L dx = 0 if k = 1, 2, 3, ·
··.
RL mπx nπx R L mπx nπx 0 m 6= n
2. Prove −L sin L sin L dx = −L cos L cos L dx = L m=n
.
RL
3. Prove −L sin mπx nπx
L cos L dx = 0, where m and n can assume any of the values 1, 2, 3, · · · .
4. For the periodic function f (x) with period=10 is given by

0 for − 5 < x < 0
f (x) = and f (x + 2L) = f (x)
3 for 0 < x < 5
(a) Find the Fourier coefficients.
(b) Write the corresponding Fourier series.
(c) How should f (x) be defined at x = −5, x = 0 and x = 5 in order that the Fourier series will
converge to f (x) for −5 ≤ x ≤ 5?
Ans.
(a) a0 = 3, an = 0 and bn = 3(1−cosnπ
nπ)
∞
3(1 − cos nπ) nπx
(b) 32 + ∑ sin
n=1  nπ 5
3

 2 for x =5
 0 for − 5 < x < 0


3
(c) f (x) = 2 for x = 0 and f (x + 2(5)) = f (x)
3 for 0 <x<5




 3
2 for x=5
5. Expand f (x) = x2 , 0 < x < 2π and f (x + 2π) = f (x) in a Fourier series if the period is 2π and
show that
1 1 1 π2
2
+ 2 + 2 +··· =
1 2 3 6
∞
4π 2 4 4π
Ans. f (x) = 3 +∑ 2
cos nx − sin nx
n=1 n n
6. Classify eachof the following functions according as they are even, odd or neither even nor odd
−2 for − 3 < x < 0
(a) f (x) = and f (x + 2(3)) = f (x)
 2 for 0 < x < 3
cos x for 0 < x < π
(b) f (x) = and f (x + 2π) = f (x)
0 for π < x < 2π
(c) f (x) = x(10 − x), 0 < x < 10 and f (x + 2(5)) = f (x)
Ans. Odd, Neither even nor odd, Even
7. Prove that for −π ≤ x ≤ π
π2 cos x cos 2x cos 3x
x2 = − 4( 2 − 2 + 2 − · · · )
3 1 2 3
8. Prove that for −π ≤ x ≤ π
sin x sin 2x sin 3x
x(π − x)(π + x) = 12( − 3 + 3 −···)
13 2 3
9. Show that for −π ≤ x ≤ π,
1 2 sin 2x 3 sin 3x 4 sin 4x
x cos x = − sin x + 2( − + −···)
2 1·3 2·4 3·5
10. Show that for −π ≤ x ≤ π,
1 cos 2x cos 3x cos 4x
x sin x = 1 − cos x − 2( − + −···)
2 1·3 2·4 3·5

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]
22 Fourier Series

11. Find a Fourier series coefficients x − x2 from x = −π to x = π and hence show that

π2 1 1 1 1
= − + − +···
12 12 22 32 42
2 n n
Ans. a0 = − 2π3 , an = − 4(−1)
n2
, bn = − 2(−1)
n

12. Find a Fourier series coefficients x + x2 from x = −π to x = π and hence show that

π2 1 1 1 1
= 2 + 2 + 2 + 2 +···
12 1 2 3 4
2π 2 4(−1)n n
Ans. a0 = 3 , an = n2
, bn = − 2(−1)
n
2
13. Find a Fourier series representation of f (x) = x + x4 from x = −π to x = π.
2 n n
π2 (−1) cos nx
Ans. x + x4 = 12 + ∑∞
n=1 ( n2
− 2(−1)n sin nx )
14. Find a Fourier series representation of f (x) = |x| from x = −π to x = π.

Ans. |x| = π
2 − π4 [cos x + cos323x + cos525x + · · · ]
15. Expand f (x) = sin x, 0 < x < π in a Fourier series.
2
Ans. f (x) = π − 4( cos 2x
22 −1
+ cos 4x
42 −1
+ cos 6x
62 −1
+···)

16. Prove that for 0 ≤ x ≤ π

π2 cos 2x cos 4x cos 6x

x(π − x) = −( 2 + 2 + 2 +···)
6 1 2 3
17. Prove that for 0 ≤ x ≤ π
8 sin x sin 3x sin 5x
x(π − x) = ( + 3 + 3 +···)
π 13 3 5
18. Show that for 0 ≤ x ≤ π
sin x sin 2x sin 3x
x = 2( − + −···)
1 2 3

2x for 0 < x < 1
19. Find the Fourier half range cosine series of the function f (x) =
2(2 − x) for 1 < x < 2

Ans. f (x) = 1 − π82 cos 2πx 8 cos 6πx

2 − 9π 2 + · · ·

20. Find the periodic function


2 for − 2 < x < 0
f (x) = and f (x + 4) = f (x)
x for 0 < x < 2
∞
1 π2 (−1)n π
and show that ∑ (2n + 1)2 = and ∑∞
n=0 2n+1 = 4
n=0 8

Any one can get this soft-copy from Google site Exodus4Wisdom Natnael
c Nigussie
[email protected]
[email protected]