Scribd
Upload
11 views

Orthogonality of Eigenfunctions

The document discusses orthogonality of eigenfunctions. It defines what it means for two functions to be orthogonal over an interval, and proves that eigenfunctions corresponding to distinct eigenvalues must be orthogonal over the same interval under the same boundary conditions. It also shows some examples of orthogonal trigonometric functions like sin(x) and sin(2x).

Uploaded by

אייל אברהם
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
11 views

Orthogonality of Eigenfunctions

The document discusses orthogonality of eigenfunctions. It defines what it means for two functions to be orthogonal over an interval, and proves that eigenfunctions corresponding to distinct eigenvalues must be orthogonal over the same interval under the same boundary conditions. It also shows some examples of orthogonal trigonometric functions like sin(x) and sin(2x).

Uploaded by

אייל אברהם
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

3.

8 (SUPPLEMENT) — ORTHOGONALITY OF EIGENFUNCTIONS

We now develop some properties of eigenfunctions, to be used in Chapter 9 for Fourier


Series and Partial Differential Equations.

1. Definition of Orthogonality
Rb
We say functions f (x) and g(x) are orthogonal on a < x < b if a f (x)g(x) dx = 0 .
[Motivation: Let’s approximate the integral with a Riemann sum, as follows. Take a
large integer N , put h = (b − a)/N and partition the interval a < x < b by defining
x1 = a + h, x2 = a + 2h, . . . , xN = a + N h = b. Then
Z b
f (x)g(x) dx ≈ f (x1 )g(x1 )h + · · · + f (xN )g(xN )h
a
= (uN · vN )h
where uN = (f (x1 ), . . . , f (xN )) and vN = (g(x1 ), . . . , g(xN )) are vectors containing the values
of f and g. The vectors uN and vN are said to be orthogonal (or perpendicular) if their dot
product equals zero (uN · vN = 0), and so when we let N → ∞ in the above formula it makes
Rb
sense to say the functions f and g are orthogonal when the integral a f (x)g(x) dx equals
zero.]
Rπ π
Example. sin x and cos x are orthogonal on −π < x < π, since −π sin x cos x dx = 21 sin2 x −π = 0.

2. Integration Lemma
Suppose functions Xn (x) and Xm (x) satisfy the differential equations
Xn00 + λn Xn = 0, a < x < b,
00
Xm + λm Xm = 0, a < x < b,
for some numbers λn , λm . Then
Z b
0
(λn − λm ) Xn (x)Xm (x) dx = [Xn (x)Xm (x) − Xn0 (x)Xm (x)]ba .
a

Proof.
Z b
LHS = [(λn Xn )Xm − Xn (λm Xm )] dx by taking the λ’s inside the integral
a
Z b
= [−Xn00 Xm + Xn Xm
00
] dx since λn Xn = −Xn00 and λm Xm = −Xm
00
a
Z b
0
= [Xn Xm − Xn0 Xm ]0 dx as you can check by differentiating!
a
= RHS by the Fundamental Theorem of Calculus.

1
3. Boundary Conditions
Consider a function X(x) for a < x < b. Four boundary condition (“BC”) types are:
Dirichlet BC X(a) = X(b) = 0
Neumann BC X 0 (a) = X 0 (b) = 0
Mixed1 BC X(a) = X 0 (b) = 0
Mixed2 BC X 0 (a) = X(b) = 0
Periodic BC X(a) = X(b), X 0 (a) = X 0 (b)
(The two varieties of Mixed BC are very similar, differing only as to which endpoint has
X = 0 and which has X 0 = 0.)

4. Orthogonality of Eigenfunctions Theorem:

Eigenfunctions corresponding to distinct eigenvalues must be orthogonal.

Precise statement: suppose Xn00 + λn Xn = 0 and Xm00


+ λm Xm = 0 on a < x < b, and that Xn
and Xm both satisfy the same type of BC. If λn 6= λm then Xn and Xm are orthogonal:
Z b
Xn (x)Xm (x) dx = 0.
a

Proof. By the Integration Lemma, we have


Z b
Xn (x)Xm (x) dx
a
1 0
= [Xn (x)Xm (x) − Xn0 (x)Xm (x)]ba
λn − λm
= 0 under Dirichlet BCs, because Xn (a) = Xn (b) = 0 and Xm (a) = Xm (b) = 0
= 0 under Neumann BCs, because Xn0 (a) = Xn0 (b) = 0 and Xm
0 0
(a) = Xm (b) = 0
= 0 under Mixed BCs, for similar reasons.
0 0
For periodic BCs, we use that Xn (b) = Xn (a) and Xm (b) = Xm (a) and so on, to see
0
[Xn (x)Xm (x) − Xn0 (x)Xm (x)]ba = [Xn (b)Xm
0
(b) − Xn0 (b)Xm (b)]
0
− [Xn (a)Xm (a) − Xn0 (a)Xm (a)] = 0.

5. Example
Show that sin x and sin 2x areRorthogonal for 0 < x < π.
π
Solution. We could just show 0 sin(x) sin(2x) dx = 0 by using trigonometric identities to
evaluate the integral. But it is easier to notice that both X1 (x) = sin(x), with λ1 = 12 , and
X2 (x) = sin(2x), with λ2 = 22 , are eigenfunctions for
X 00 + λX = 0, 0 < x < π,
X(0) = X(π) = 0 (Dirichlet BC).
Rπ
Since λ1 6= λ2 , the Orthogonality Theorem implies 0
X1 X2 dx = 0, so that sin(x) and
sin(2x) are orthogonal for 0 < x < π.

6. Orthogonality of sines and cosines for −π < x < π


The following formulas will be essential for our study of Fourier series, in Chapter 9.
Let n, m ≥ 1 be integers. Then:
Z π
(1) cos(nx) cos(mx) dx = 0 if n 6= m
−π
(2) =π if n = m,
Z π
(3) sin(nx) sin(mx) dx = 0 if n 6= m
−π
(4) =π if n = m,
Z π
(5) cos(nx) sin(mx) dx = 0 if n 6= m
−π
(6) = 0 if n = m.
Proof. For formula (1), we let Xn (x) = cos(nx) and Xm (x) = cos(mx), so that Xn and Xm
are eigenfunctions of X 00 + λX = 0 on −π < x < π, with eigenvalues λn = n2 and λm = m2
respectively. Also, Xn and Xm satisfy periodic boundary conditions (since they are both
2π-periodic). If n 6= m then λn 6= λm , and so the Orthogonality Theorem implies Xn and
Xm are orthogonal, which is equation (1).
Equations (3) and (5) are proved similarly. [Exercise.]
To get equation (2), we substitute n = m and evaluate the integral explicitly as
Z π Z π
2 1
cos (mx) dx = (1 + cos(2mx)) dx = π.
−π −π 2
Equation (4) proceeds similarly. [Exercise.] And finally, equation (6) is true since we can
substitute n = m and evaluate the integral explicitly as
Z π π
1
cos(mx) sin(mx) dx = sin2 (mx) = 0.
−π 2m −π
Remark. Here we have used the Orthogonality Theorem to evaluate integrals (1), (3) and
(5). They can also be evaluated using trigonometric identities (or the integration formulas
inside the back cover of the text). But “orthogonality of eigenfunctions” is the best way to
think about these results, and we will use orthogonality in Chapter 9 for other applications
as well.


You might also like