The 2D wave equation Separation of variables Superposition Examples

The two dimensional wave equation

Ryan C. Daileda

Trinity University

Partial Differential Equations

March 1, 2012

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Physical motivation
Consider a thin elastic membrane stretched tightly over a
rectangular frame. Suppose the dimensions of the frame are a × b
and that we keep the edges of the membrane fixed to the frame.

Perturbing the membrane from equilibrium results in some

sort of vibration of the surface.
Our goal is to mathematically model the vibrations of the
membrane surface.
Daileda The 2D wave equation
The 2D wave equation Separation of variables Superposition Examples

We let
deflection of membrane from equilibrium at
u(x, y , t) =
position (x, y ) and time t.

For a fixed t, the surface z = u(x, y , t) gives the shape of the

membrane at time t.

Under ideal assumptions (e.g. uniform membrane density, uniform

tension, no resistance to motion, small deflection, etc.) one can
show that u satisfies the two dimensional wave equation

utt = c 2 ∇2 u = c 2 (uxx + uyy ) (1)

for 0 < x < a, 0 < y < b.

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Remarks:
For the derivation of the wave equation from Newton’s second
law, see exercise 3.2.8.
As in the one dimensional situation, the constant c has the
units of velocity. It is given by
τ
c2 = ,
ρ
where τ is the tension per unit length, and ρ is mass density.
The operator
∂2 ∂2
∇2 = 2
+ 2
∂x ∂y
is called the Laplacian. It will appear in many of our
subsequent investigations.

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

The fact that we are keeping the edges of the membrane fixed is
expressed by the boundary conditions

u(0, y , t) = u(a, y , t) = 0, 0 ≤ y ≤ b, t ≥ 0,
u(x, 0, t) = u(x, b, t) = 0, 0 ≤ x ≤ a, t ≥ 0. (2)

We must also specify how the membrane is initially deformed and

set into motion. This is done via the initial conditions

u(x, y , 0) = f (x, y ), (x, y ) ∈ R,

ut (x, y , 0) = g (x, y ), (x, y ) ∈ R, (3)

where R = [0, a] × [0, b].

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Solving the 2D wave equation

Goal: Write down a solution to the wave equation (1) subject to

the boundary conditions (2) and initial conditions (3).

We will follow the (hopefully!) familiar process of

using separation of variables to produce simple solutions to
(1) and (2),
and then the principle of superposition to build up a
solution that satisfies (3) as well.

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Separation of variables
We seek nontrivial solutions of the form

u(x, y , t) = X (x)Y (y )T (t).

Plugging this into the wave equation (1) we get

XYT ′′ = c 2 X ′′ YT + XY ′′ T .

If we divide both sides by c 2 XYT this becomes

T ′′ X ′′ Y ′′
= + .
c 2T X Y
Because the two sides are functions of different independent
variables, they must be constant:
T ′′ X ′′ Y ′′
= A = + .
c 2T X Y
Daileda The 2D wave equation
The 2D wave equation Separation of variables Superposition Examples

The first equality becomes

T ′′ − c 2 AT = 0.

The second can be rewritten as

X ′′ Y ′′
=− + A.
X Y
Once again, the two sides involve unrelated variables, so both are
constant:
X ′′ Y ′′
=B =− + A.
X Y
If we let C = A − B these equations can be rewritten as

X ′′ − BX = 0,
Y ′′ − CY = 0.

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

The first boundary condition is

0 = u(0, y , t) = X (0)Y (y )T (t), 0 ≤ y ≤ b, t ≥ 0.

Since we want nontrivial solutions only, we can cancel Y and T ,

yielding
X (0) = 0.
When we perform similar computations with the other three
boundary conditions we also get

X (a) = 0,
Y (0) = Y (b) = 0.

There are no boundary conditions on T .

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Fortunately, we have already solved the two boundary value

problems for X and Y . The nontrivial solutions are
mπ
Xm (x) = sin µm x, µm = , m = 1, 2, 3, . . .
a
nπ
Yn (y ) = sin νn y , νn = , n = 1, 2, 3, . . .
b
with separation constants B = −µ2m and C = −νn2 .

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Recall that T must satisfy

T ′′ − c 2 AT = 0

with A = B + C = − µ2m + νn2 < 0. It follows that for any choice

of m and n the general solution for T is

∗
Tmn (t) = Bmn cos λmn t + Bmn sin λmn t,

where r
m2 n2
q
λmn = c µ2m + νn2+ 2. = cπ
a2 b
These are the characteristic frequencies of the membrane.

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Assembling our results, we find that for any pair m, n ≥ 1 we have

the normal mode

umn (x, y , t) = Xm (x)Yn (y )Tmn (t)

∗
= sin µm x sin νn y (Bmn cos λmn t + Bmn sin λmn t)

where
mπ nπ
q
µm = , νn = , λmn = c µ2m + νn2 .
a b

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Remarks:
Note that the normal modes:
oscillate spatially with frequency µm in the x-direction,

oscillate spatially with frequency νn in the y -direction,

oscillate in time with frequency λmn .

While µm and νn are simply multiples of π/a and π/b,

respectively,
λmn is not a multiple of any basic frequency.

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Superposition

According to the principle of superposition, because the normal

modes umn satisfy the homogeneous conditions (1) and (2), we
may add them to obtain the general solution

∞ X
X ∞
∗
u(x, y , t) = sin µm x sin νn y (Bmn cos λmn t + Bmn sin λmn t).
n=1 m=1

We must use a double series since the indices m and n vary

independently throughout the set

N = {1, 2, 3, . . .}.

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Initial conditions

Finally, we must determine the values of the coefficients Bmn and

∗ that are required so that our solution also satisfies the initial
Bmn
conditions (3). The first of these is
∞
∞ X
X mπ nπ
f (x, y ) = u(x, y , 0) = Bmn sin x sin y
a b
n=1 m=1

and the second is

∞ X
∞
X
∗ mπ nπ
g (x, y ) = ut (x, y , 0) = λmn Bmn sin x sin y.
a b
n=1 m=1

These are examples of double Fourier series.

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Orthogonality (again!)

To compute the coefficients in a double Fourier series we can

appeal to the following result.
Theorem
The functions
mπ nπ
Zmn (x, y ) = sin x sin y , m, n ∈ N
a b
are pairwise orthogonal relative to the inner product
Z aZ b
hf , g i = f (x, y )g (x, y ) dy dx.
0 0

This is easily verified using the orthogonality of the functions

sin(nπx/a) on the interval [0, a].

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Using the usual argument, it follows that assuming we can write

∞
∞ X
X mπ nπ
f (x, y ) = Bmn sin x sin y,
a b
n=1 m=1

then
Z aZ b
mπ nπ
x sinf (x, y ) sin
y dy dx
hf , Zmn i 0 0 a b
Bmn = = Z aZ b
hZmn , Zmn i mπ nπ
sin2 x sin2 y dy dx
0 0 a b

Z aZ b
4 mπ nπ
= f (x, y ) sin x sin y dy dx (4)
ab 0 0 a b

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Representability
The question of whether or not a given function is equal to a
double Fourier series is partially answered by the following result.
Theorem
If f (x, y ) is a C 2 function on the rectangle [0, a] × [0, b], then
∞ X
∞
X mπ nπ
f (x, y ) = Bmn sin x sin y,
a b
n=1 m=1

where Bmn is given by (4).

To say that f (x, y ) is a C 2 function means that f as well as

its first and second order partial derivatives are all continuous.
While not as general as the Fourier representation theorem,
this result is sufficient for our applications.
Daileda The 2D wave equation
The 2D wave equation Separation of variables Superposition Examples

Conclusion

Theorem
Suppose that f (x, y ) and g (x, y ) are C 2 functions on the rectangle
[0, a] × [0, b]. The solution to the wave equation (1) with
boundary conditions (2) and initial conditions (3) is given by
∞ X
X ∞
∗
u(x, y , t) = sin µm x sin νn y (Bmn cos λmn t + Bmn sin λmn t)
n=1 m=1

where
mπ nπ
q
µm = , νn = , λmn = c µ2m + νn2 ,
a b

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Theorem (continued)
∗ are given by
and the coefficients Bmn and Bmn
Z aZ b
4 mπ nπ
Bmn = f (x, y ) sin x sin y dy dx
ab 0 0 a b

and
Z aZ b
∗ 4 mπ nπ
Bmn = g (x, y ) sin x sin y dy dx.
abλmn 0 0 a b

We have not actually verified that this solution is unique, i.e.

that this is the only solution to the wave equation with the
given boundary and initial conditions.
Uniqueness can be proven using an argument involving
conservation of energy in the vibrating membrane.

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Example 1

Example
A 2 × 3 rectangular membrane has c = 6. If we deform it to have
shape given by
f (x, y ) = xy (2 − x)(3 − y ),
keep its edges fixed, and release it at t = 0, find an expression that
gives the shape of the membrane for t > 0.

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

∗ . Since
We must compute the coefficients Bmn and Bmn
g (x, y ) = 0 we immediately have
∗
Bmn = 0.

We also have
Z 2Z 3
4 mπ nπ
Bmn = xy (2 − x)(3 − y ) sin x sin y dy dx
2·3 0 0 2 3
2 2
Z Z 3
mπ nπ
= x(2 − x) sin x dx y (3 − y ) sin y dy
3 0 2 0 3
2 16(1 + (−1)m+1 ) 54(1 + (−1)n+1 )
  
=
3 π 3 m3 π 3 n3
576 (1 + (−1)m+1 )(1 + (−1)n+1 )
= 6 .
π m3 n3

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

The coefficients λmn are given by

r
m2 n2 p
λmn = 6π + = π 9m2 + 4n2 .
4 9
Assembling all of these pieces yields

∞ ∞ 
576 X X (1 + (−1)m+1 )(1 + (−1)n+1 ) mπ
u(x, y , t) = 6 sin x
π m3 n3 2
n=1 m=1
nπ p 
× sin y cos π 9m2 + 4n2 t .
3

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Example 2

Example
Suppose in the previous example we also impose an initial velocity
given by
g (x, y ) = sin 2πx.
Find an expression that gives the shape of the membrane for t > 0.

Because Bmn depends only on the initial shape, which hasn’t

changed, we don’t need to recompute them.
∗ and add the appropriate terms to
We only need to find Bmn
the previous solution.

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Using the values of λmn computed above, we have

Z 2Z 3
∗ 2 mπ nπ
Bmn = √ sin 2πx sin x sin y dy dx
2
3π 9m + 4n 0 02 2 3
Z 2 Z 3
2 mπ nπ
= √ sin2πx sin x dx sin y dy .
2
3π 9m + 4n 0 2 2 0 3

The first integral is zero unless m = 4, in which case it evaluates

to 1. Evaluating the second integral, we have

∗ 1 3(1 + (−1)n+1 ) 1 + (−1)n+1

B4n = √ = √
3π 36 + n2 nπ π 2 n 36 + n2
∗ = 0 for m 6= 4.
and Bmn

Daileda The 2D wave equation

The 2D wave equation Separation of variables Superposition Examples

Let u1 (x, y , t) denote the solution obtained in the previous

example. If we let

∞ X
∞
X
∗ mπ nπ
u2 (x, y , t) = Bmn sin x sin y sin λmn t
2 3
m=1 n=1
∞
sin 2πx X 1 + (−1)n+1 nπ p
= 2
√ sin y sin 2π 36 + n2 t
π n 36 + n2 3
n=1

then the solution to the present problem is

u(x, y , t) = u1 (x, y , t) + u2 (x, y , t).

Daileda The 2D wave equation