Fourier Analysis: Part II

Hakan Dogan

Department of Mechanical Engineering

Hacettepe University

MMU604(MMU703) Analytical Methods in Engineering (II)

April 21, 2025

MMU604/MMU703 April 21, 2025 1 / 21

 Orthogonal Functions

Fourier Series (Recap)

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

where
1
Z π
a0 = f (x)dx
2π −π
1
Z π
an = f (x) cos nxdx
π −π
1
Z π
bn = f (x) sin nxdx
π −π

MMU604/MMU703 April 21, 2025 2 / 21

 Orthogonal Functions

Fourier Series with any period of p=2L (Recap)

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

MMU604/MMU703 April 21, 2025 3 / 21

 Orthogonal Functions

Sturm-Liouville Problem (SLP)

′
p(x)y ′ + [q(x) + λ r (x)]y = 0


on some interval a ≡ b, satisfying conditions of the form

1 k1 y + k2 y ′ = 0 at x = a
2 l1 y + l2 y′ =0 at x = b.
λ : parameter
k1 , k2 , l1 , l;2 : real constants
* one of the conditions must be satisfied by non-zero y .

The above equation and conditions are called Sturm-Liouville Equation

and Sturm-Liouville Problem.

MMU604/MMU703 April 21, 2025 4 / 21

 Orthogonal Functions

Orthogonal Functions

Orthogonal functions on a ≦ x ≦ b with respect to the weight function

r (x) > 0:
Z b
(ym , yn ) = r (x)ym (x)yn (x)dx = 0 (m ̸= n)
a

The norm ∥ym ∥ of ym is defined by

s
p Z b
∥ym ∥ = (ym , ym ) = 2 (x)dx
r (x)ym
a

The functions y1 , y2 , · · · are called orthonormal on a ≦ x ≦ b if they are

orthogonal on this interval and all have norm 1.

MMU604/MMU703 April 21, 2025 5 / 21

 Orthogonal Functions

Orthogonal Series

Orthogonality will give us coefficient formulas for the desired generalised

Fourier Series. Let y0 , y1 ... be orthogonal wrt to a weight function r (x).
Then, a function:
∞
f (x) = ∑ am ym (x) = a0 y0 (x) + a1 y1 (x) + ...
m=0

This is called an orthogonal expansion (series). The Fourier constants:

Z b
(f , ym ) 1
am = = r (x)f (x)ym (x)dx (n = 0, 1, · · · )
∥ym ∥2 ∥ym ∥2 a

See Fourier-Legendre and Fourier-Bessel series!

MMU604/MMU703 April 21, 2025 6 / 21

 Fourier Integral

Fourier Integral

Can a similar approach to Fourier Series be applied for non-periodic

functions?
• We know how to deal with periodic functions with a period of 2L
• See what happens to FS if we let L → ∞

MMU604/MMU703 April 21, 2025 7 / 21

 Fourier Integral

Fourier Integral

Can a similar approach to Fourier Series be applied for non-periodic

functions?
• We know how to deal with periodic functions with a period of 2L
• See what happens to FS if we let L → ∞

MMU604/MMU703 April 21, 2025 7 / 21

 Fourier Integral

Fourier Integral

Any periodic function fL (x) of period 2L (with FS):

∞
nπ
fL (x) = a0 + ∑ (an cos ωn x + bn sin ωn x) , ωn =
n=1 L

where (with dummy variable v )

Z L Z L Z L
1 1 1
a0 = fL (v )dv , an = fL (v ) cos ωn vdv , bn = fL (v ) sin ωn vdv
2L −L L −L L −L

Let’s set
(n + 1)π nπ π
∆ω = ωn+1 − ωn = − = .
L L L
and 1/L = ∆ω/π.

MMU604/MMU703 April 21, 2025 8 / 21

 Fourier Integral

Fourier Integral

Z L ∞  L
1 1
Z
fL (x) = fL (v )dv + ∑ (cos ∗omegan x) ∆ω fL (v ) cos ωn vdv
2L −L π n=1 −L
Z L 
+ (sin ωn x) ∆ω fL (v ) sin ωn vdv .
−L
1/L → 0; therefore, the first term becomes 0. The sum term becomes
integral: R 
f (x) = π1 0∞ cos ωx −∞
R∞ R∞ 
f (v ) cos ωvdv + sin ωx −∞ f (v ) sin ωvdv dω
The representation by a Fourier integral:
Z ∞
f (x) = [A(ω) cos ωx + B(ω) sin ωx]dω
0
where

1 1
Z ∞ Z ∞
A(ω) = f (v ) cos ωvdv , B(ω) = f (v ) sin ωvdv .
π −∞ π −∞

MMU604/MMU703 April 21, 2025 9 / 21

 Fourier Integral

Fourier Integral (Example)

2 cos ωx sin ω
Z ∞
Soln: f (x) = dω
π 0 ω

MMU604/MMU703 April 21, 2025 10 / 21

 Fourier Integral

Fourier Sine and Cosine Integrals

The function f has a Fourier representation and is even:

Fourier cosine integral representation of f (x) on [0, ∞):
Z ∞
f (x) = A (ω) cos ωxdω
0

where
2
Z ∞
A(ω) = f (t) cos ωtdt
π 0
is odd: Fourier sine integral representation of f (x) on [0, ∞):
Z ∞
f (x) = B (ω) sin ωxdω (1)
0

where
2
Z ∞
B(ω) = f (t) sin ωtdt (2)
π 0

MMU604/MMU703 April 21, 2025 11 / 21

 Fourier Integral

Fourier Sine and Cosine Integrals (example)

Derive the Fourier cosine and sine integral of f (x) = e −kx , where x > 0
and k > 0.

MMU604/MMU703 April 21, 2025 12 / 21

 Fourier Integral

Fourier Cosine and Sine Transforms

An integral transform is a transformation in the form of an integral that

produces from given functions new functions depending on a different
variable. (e.g, The Laplace transform)

Fourier Cosine Transforms

It concerns even functions f (x):

2
Z ∞ Z ∞
f (x) = A(ω) cos ωxdω, where A(w ) = f (v ) cos ωvdv
0 π 0
Writing v = x in the formula for A(ω):
r Z
2 ∞
fˆc (ω) = f (x) cos ωxdx
π 0
This is a transformation (Fourier cosine transform), which gives from
f (x) a new function fˆc (ω).
MMU604/MMU703 April 21, 2025 13 / 21
 Fourier Integral

Fourier Cosine and Sine Transforms

2
Z ∞ Z ∞
f (x) = A(ω) cos ωxdω, where A(ω) = f (v ) cos ωvdv
0 π 0

r Z
2 ∞ˆ
f (x) = fc (ω) cos ∗omegaxdω
π 0

This formula gives us back f (x) from fˆc (ω), and we therefore call f (x) the
inverse Fourier cosine transform of fˆc (ω).

The process of obtaining the transform fˆc from a given f is also called the
Fourier cosine transform or the Fourier cosine transform method.

MMU604/MMU703 April 21, 2025 14 / 21

 Fourier Integral

Fourier Cosine and Sine Transforms

A similar approach can be followed for the Fourier sine transform and
the inverse Fourier sine transform:
r Z
2 ∞
fˆs (ω) = f (x) sin ωxdx
π 0

r Z
2 ∞ˆ
f (x) = fc (ω) sin ωxdω
π 0

MMU604/MMU703 April 21, 2025 15 / 21

 Fourier Transform

Fourier Transform

• It is an integral transform which can be thought as similar to the

Laplace transform.
• It is obtained from the Fourier integral in complex form.

MMU604/MMU703 April 21, 2025 16 / 21

 Fourier Transform

Fourier Transform

• It is an integral transform which can be thought as similar to the

Laplace transform.
• It is obtained from the Fourier integral in complex form.
The Fourier integral:
Z ∞
f (x) = (A (ω) cos ωx + B (ω) sin ωx) dω
0
where
1
Z ∞
A(ω) = f (v ) cos ωvdv
π −∞
1
Z ∞
B(ω) = f (v ) sin ωvdv
π −∞

MMU604/MMU703 April 21, 2025 16 / 21

 Fourier Transform

Fourier Transform

The resulting expression:

Z ∞ Z ∞ 
1 −iωv
f (x) = f (v )e dv e iωx dω
2π −∞ −∞
or
Z ∞ 
1 1
Z ∞
−iωv
f (x) = √ √ f (v )e dv e iωx dω
2π −∞ 2π −∞
The Fourier Transform:
1
Z ∞
F {f (x)} = F (ω) = √ f (x)e −iωx dx
2π −∞

The Inverse Fourier Transform:

1
Z ∞
F −1 {F (ω)} = f (x) = √ F (ω)e iωx dω
2π −∞

MMU604/MMU703 April 21, 2025 17 / 21

 Fourier Transform

Important properties

Linearity:
F {af (t) + bg (t)} = aF {f (t)} + bF {g (t)}
Shifting property:
F {f (t − t0 )} = F (ω)e −iωt0
Frequency shifting property:
Z ∞ Z ∞
F {e iω0 t f (t)} = e −iω0 t f (t)e −iωt dt = f (t)e −i(ω−ω0 )t dt
−∞ −∞

F {e iω0 t f (t)} = F (ω − ω0 )

MMU604/MMU703 April 21, 2025 18 / 21

 Fourier Transform

Fourier Transform

The Fourier Transform of derivatives:

F {f ′ (t)} = iωF {f (t)}

F {f n (t)} = (iω)n F {f (t)}

Derivative wrt ω:

dF (ω)
= −iF {tf (t)}
dω
d n F (ω)
= (−i)n F {t n f (t)}
dω n

MMU604/MMU703 April 21, 2025 19 / 21

 Fourier Transform

Fourier Transform

The Fourier Transform of Dirac-delta function:

F {δ (t − a)} = e −iωa

F {δ (t)} = 1
Convolution
Z ∞ Z ∞
(f ∗ g )(t) = f (τ)g (t − τ)dτ = f (t − τ)g (τ)dτ
−∞ −∞

F {(f ∗ g )} = F {f }F {g } = F (ω)G (ω)

MMU604/MMU703 April 21, 2025 20 / 21

 Fourier Transform

May the Fourier be with you

MMU604/MMU703 April 21, 2025 21 / 21