7.2
Heat Flow in a Rectangular Domain
In this section we solve the heat equation in two spatial variables inside a rectangle L by H.
The equation is
u
t
= k(u
xx
+ u
yy
),
0 < x < L,
0 < y < H,
(7.2.1)
u(0, y, t) = 0,
(7.2.2)
u(L, y, t) = 0,
(7.2.3)
u(x, 0, t) = 0,
(7.2.4)
u(x, H, t) = 0,
(7.2.5)
u(x, y, 0) = f (x, y).
(7.2.6)
Notice that the term in parentheses in (7.2.1) is
∇
2
u. Note also that we took Dirichlet bound-
ary conditions (i.e. specified temperature on the boundary). We can write this condition
as
u(x, y, t) = 0.
on the boundary
(7.2.7)
Other possible boundary conditions are left to the reader.
The method of separation of variables will proceed as follows :
1.
Let
u(x, y, t) = T (t)φ(x, y)
(7.2.8)
2.
Substitute in (7.2.1) and separate the variables
˙
T φ = kT
∇
2
φ
˙
T
kT
=
∇
2
φ
φ
=
−λ
3.
Write the ODEs
˙
T (t) + kλT (t) = 0
(7.2.9)
∇
2
φ + λφ = 0
(7.2.10)
4.
Use the homogeneous boundary condition (7.2.7) to get the boundary condition associ-
ated with (7.2.10)
φ(x, y) = 0.
on the boundary
(7.2.11)
The only question left is how to get the solution of (7.2.10) - (7.2.11). This can be done in
a similar fashion to solving Laplace’s equation.
Let
φ(x, y) = X(x)Y (y),
(7.2.12)
then (7.2.10) - (7.2.11) yield 2 ODEs
X
+ µX = 0,
(7.2.13)
X(0) = X(L) = 0,
(7.2.14)
148
Y
+ (λ
− µ)Y = 0,
(7.2.15)
Y (0) = Y (H) = 0.
(7.2.16)
The boundary conditions (7.2.14) and (7.2.16) result from (7.2.2) - (7.2.5). Equation (7.2.13)
has a solution
X
n
= sin
nπ
L
x,
n = 1, 2, . . .
(7.2.17)
µ
n
=
nπ
L
2
,
n = 1, 2, . . .
(7.2.18)
as we have seen in Chapter 2. For each n, equation (7.2.15) is solved the same way
Y
mn
= sin
mπ
H
y,
m = 1, 2, . . . , n = 1, 2, . . .
(7.2.19)
λ
mn
− µ
n
=
mπ
H
2
,
m = 1, 2, . . . , n = 1, 2, . . .
(7.2.20)
Therefore by (7.2.12) and (7.2.17)-(7.2.20),
φ
mn
(x, y) = sin
nπ
L
x sin
mπ
H
y,
(7.2.21)
λ
mn
=
nπ
L
2
+
mπ
H
2
,
(7.2.22)
n = 1, 2, . . . , m = 1, 2, . . .
Using (7.2.8) and the principle of superposition, we can write the solution of (7.2.1) as
u(x, y, t) =
∞
n=1
∞
m=1
A
mn
e
−kλ
mn
t
sin
nπ
L
x sin
mπ
H
y,
(7.2.23)
where λ
mn
is given by (7.2.22).
To find the coefficients A
mn
, we use the initial condition (7.2.6), that is for t = 0 in (7.2.23)
we get :
f (x, y) =
∞
n=1
∞
m=1
A
mn
sin
nπ
L
x sin
mπ
H
y,
(7.2.24)
A
mn
are the generalized Fourier coefficients (double Fourier series in this case). We can
compute A
mn
by
A
mn
=
(
L
0
(
H
0
f (x, y) sin
nπ
L
x sin
mπ
H
ydydx
(
L
0
(
H
0
sin
2 nπ
L
x sin
2 mπ
H
ydydx
.
(7.2.25)
(See next section.)
Remarks :
i. Equation (7.2.10) is called Helmholtz equation.
ii. A more general form of the equation is
∇ · (p(x, y)∇φ(x, y)) + q(x, y)φ(x, y) + λσ(x, y)φ(x, y) = 0
(7.2.26)
iii. A more general boundary condition is
β
1
(x, y)φ(x, y) + β
2
(x, y)
∇φ · n = 0
on the boundary
(7.2.27)
where
n is a unit normal vector pointing outward. The special case β
2
≡ 0 yields (7.2.11).
149
Problems
1. Solve the heat equation
u
t
(x, y, t) = k (u
xx
(x, y, t) + u
yy
(x, y, t)) ,
on the rectangle 0 < x < L, 0 < y < H subject to the initial condition
u(x, y, 0) = f (x, y),
and the boundary conditions
a.
u(0, y, t) = u
x
(L, y, t) = 0,
u(x, 0, t) = u(x, H, t) = 0.
b.
u
x
(0, y, t) = u(L, y, t) = 0,
u
y
(x, 0, t) = u
y
(x, H, t) = 0.
c.
u(0, y, t) = u(L, y, t) = 0,
u(x, 0, t) = u
y
(x, H, t) = 0.
2. Solve the heat equation on a rectangular box
0 < x < L, 0 < y < H, 0 < z < W,
u
t
(x, y, z, t) = k(u
xx
+ u
yy
+ u
zz
),
subject to the boundary conditions
u(0, y, z, t) = u(L, y, z, t) = 0,
u(x, 0, z, t) = u(x, H, z, t) = 0,
u(x, y, 0, t) = u(x, y, W, t) = 0,
and the initial condition
u(x, y, z, 0) = f (x, y, z).
150
7.3
Vibrations of a rectangular Membrane
The method of separation of variables in this case will lead to the same Helmholtz equation.
The only difference is in the T equation. the problem to solve is as follows :
u
tt
= c
2
(u
xx
+ u
yy
),
0 < x < L, 0 < y < H,
(7.3.1)
u(0, y, t) = 0,
(7.3.2)
u(L, y, t) = 0,
(7.3.3)
u(x, 0, t) = 0,
(7.3.4)
u
y
(x, H, t) = 0,
(7.3.5)
u(x, y, 0) = f (x, y),
(7.3.6)
u
t
(x, y, 0) = g(x, y).
(7.3.7)
Clearly there are two initial conditions, (7.3.6)-(7.3.7), since the PDE is second order in time.
We have decided to use a Neumann boundary condition at the top y = H, to show how the
solution of Helmholtz equation is affected.
The steps to follow are : (the reader is advised to compare these equations to (7.2.8)-(7.2.25))
u(x, y, t) = T (t)φ(x, y),
(7.3.8)
¨
T
c
2
T
=
∇
2
φ
φ
=
−λ
¨
T + λc
2
T = 0,
(7.3.9)
∇
2
φ + λφ = 0,
(7.3.10)
β
1
φ(x, y) + β
2
φ
y
(x, y) = 0,
(7.3.11)
where either β
1
or β
2
is zero depending on which side of the rectangle we are on.
φ(x, y) = X(x)Y (y),
(7.3.12)
X
+ µX = 0,
(7.3.13)
X(0) = X(L) = 0,
(7.3.14)
Y
+ (λ
− µ)Y = 0,
(7.3.15)
Y (0) = Y
(H) = 0,
(7.3.16)
X
n
= sin
nπ
L
x,
n = 1, 2, . . .
(7.3.17)
µ
n
=
nπ
L
2
,
n = 1, 2, . . .
(7.3.18)
Y
mn
= sin
(m
−
1
2
)π
H
y,
m = 1, 2, . . .
n = 1, 2, . . .
(7.3.19)
151
λ
mn
=
(m
−
1
2
)π
H
2
+
nπ
L
2
,
m = 1, 2, . . .
n = 1, 2, . . .
(7.3.20)
Note the similarity of (7.3.1)-(7.3.20) to the corresponding equations of section 4.2.
The solution
u(x, y, t) =
∞
m=1
∞
n=1
A
mn
cos
+
λ
mn
ct + B
mn
sin
+
λ
mn
ct
sin
nπ
L
x sin
(m
−
1
2
)π
H
y.
(7.3.21)
Since the T equation is of second order, we end up with two sets of parameters A
mn
and
B
mn
. These can be found by using the two initial conditions (7.3.6)-(7.3.7).
f (x, y) =
∞
n=1
∞
m=1
A
mn
sin
nπ
L
x sin
(m
−
1
2
)π
H
y,
(7.3.22)
g(x, y) =
∞
n=1
∞
m=1
c
+
λ
mn
B
mn
sin
nπ
L
x sin
(m
−
1
2
)π
H
y.
(7.3.23)
To get (7.3.23) we need to evaluate u
t
from (7.3.21) and then substitute t = 0. The coeffi-
cients are then
A
mn
=
(
L
0
(
H
0
f (x, y) sin
nπ
L
x sin
(m−
1
2
)π
H
ydydx
(
L
0
(
H
0
sin
2 nπ
L
x sin
2 (m−
1
2
)π
H
ydydx
,
(7.3.24)
c
+
λ
mn
B
mn
=
(
L
0
(
H
0
g(x, y) sin
nπ
L
x sin
(m−
1
2
)π
H
ydydx
(
L
0
(
H
0
sin
2 nπ
L
x sin
2 (m−
1
2
)π
H
ydydx
,
(7.3.25)
152
Problems
1. Solve the wave equation
u
tt
(x, y, t) = c
2
(u
xx
(x, y, t) + u
yy
(x, y, t)) ,
on the rectangle 0 < x < L, 0 < y < H subject to the initial conditions
u(x, y, 0) = f (x, y),
u
t
(x, y, 0) = g(x, y),
and the boundary conditions
a.
u(0, y, t) = u
x
(L, y, t) = 0,
u(x, 0, t) = u(x, H, t) = 0.
b.
u(0, y, t) = u(L, y, t) = 0,
u(x, 0, t) = u(x, H, t) = 0.
c.
u
x
(0, y, t) = u(L, y, t) = 0,
u
y
(x, 0, t) = u
y
(x, H, t) = 0.
2. Solve the wave equation on a rectangular box
0 < x < L, 0 < y < H, 0 < z < W,
u
tt
(x, y, z, t) = c
2
(u
xx
+ u
yy
+ u
zz
),
subject to the boundary conditions
u(0, y, z, t) = u(L, y, z, t) = 0,
u(x, 0, z, t) = u(x, H, z, t) = 0,
u(x, y, 0, t) = u(x, y, W, t) = 0,
and the initial conditions
u(x, y, z, 0) = f (x, y, z),
u
t
(x, y, z, 0) = g(x, y, z).
3. Solve the wave equation on an isosceles right-angle triangle with side of length a
u
tt
(x, y, t) = c
2
(u
xx
+ u
yy
),
153
subject to the boundary conditions
u(x, 0, t) = u(0, y, t) = 0,
u(x, y, t) = 0,
on the line
x + y = a
and the initial conditions
u(x, y, 0) = f (x, y),
u
t
(x, y, 0) = g(x, y).
154
7.4
Helmholtz Equation
As we have seen in this chapter, the method of separation of variables in two independent
variables leads to Helmholtz equation,
∇
2
φ + λφ = 0
subject to the boundary conditions
β
1
φ(x, y) + β
2
φ
x
(x, y) + β
3
φ
y
(x, y) = 0.
Here we state a result generalizing Sturm-Liouville’s from Chapter 6 of Neta.
Theorem:
1. All the eigenvalues are real.
2. There exists an infinite number of eigenvalues. There is a smallest one but no largest.
3. Corresponding to each eigenvalue, there may be many eigenfunctions.
4. The eigenfunctions φ
i
(x, y) form a complete set, i.e. any function f (x, y) can be
represented by
i
a
i
φ
i
(x, y)
(7.4.1)
where the coefficients a
i
are given by,
a
i
=
( (
φ
i
f (x, y)dxdy
( (
φ
2
i
dxdy
(7.4.2)
5. Eigenfunctions belonging to different eigenvalues are orthogonal.
6. An eigenvalue λ can be related to the eigenfunction φ(x, y) by Rayleigh quotient:
λ =
( (
(
∇φ)
2
dxdy
−
,
φ
∇φ · nds
( (
φ
2
dxdy
(7.4.3)
where
,
symbolizes integration on the boundary. For example, the following Helmholtz
problem (see 4.2.10-11)
∇
2
φ + λφ = 0,
0
≤ x ≤ L, 0 ≤ y ≤ H,
(7.4.4)
φ = 0,
on the boundary,
(7.4.5)
was solved and we found
λ
mn
=
nπ
L
2
+
mπ
H
2
,
n = 1, 2, . . . ,
m = 1, 2, . . .
(7.4.6)
φ
mn
(x, y) = sin
nπ
L
x sin
mπ
H
y,
n = 1, 2, . . . ,
m = 1, 2, . . .
(7.4.7)
Clearly all the eigenvalues are real. The smallest one is λ
11
=
π
L
2
+
π
H
2
, λ
mn
→ ∞
as n and m
→ ∞. There may be multiple eigenfunctions in some cases. For example, if
155
L = 2H then λ
41
= λ
22
but the eigenfunctions φ
41
and φ
22
are different. The coefficients of
expansion are
a
mn
=
(
L
0
(
H
0
f (x, y)φ
mn
dxdy
(
L
0
(
H
0
φ
2
mn
dxdy
(7.4.8)
as given by (7.2.25).
156
Problems
1. Solve
∇
2
φ + λφ = 0
[0, 1]
× [0, 1/4]
subject to
φ(0, y) = 0
φ
x
(1, y) = 0
φ(x, 0) = 0
φ
y
(x, 1/4) = 0.
Show that the results of the theorem are true.
2. Solve Helmholtz equation on an isosceles right-angle triangle with side of length a
u
xx
+ u
yy
+ λu = 0,
subject to the boundary conditions
u(x, 0, t) = u(0, y, t) = 0,
u(x, y, t) = 0,
on the line
x + y = a.
157
7.5
Vibrating Circular Membrane
In this section, we discuss the solution of the wave equation inside a circle. As we have
seen in sections 4.2 and 4.3, there is a similarity between the solution of the heat and wave
equations. Thus we will leave the solution of the heat equation to the exercises.
The problem is:
u
tt
(r, θ, t) = c
2
∇
2
u,
0
≤ r ≤ a, 0 ≤ θ ≤ 2π, t > 0
(7.5.1)
subject to the boundary condition
u(a, θ, t) = 0,
(clamped membrane)
(7.5.2)
and the initial conditions
u(r, θ, 0) = α(r, θ),
(7.5.3)
u
t
(r, θ, 0) = β(r, θ).
(7.5.4)
The method of separation of variables leads to the same set of differential equations
¨
T (t) + λc
2
T = 0,
(7.5.5)
∇
2
φ + λφ = 0,
(7.5.6)
φ(a, θ) = 0,
(7.5.7)
Note that in polar coordinates
∇
2
φ =
1
r
∂
∂r
r
∂φ
∂r
+
1
r
2
∂
2
φ
∂θ
2
(7.5.8)
Separating the variables in the Helmholtz equation (7.5.6) we have
φ(r, θ) = R(r)Θ(θ),
(7.5.9)
Θ
+ µΘ = 0
(7.5.10)
d
dr
r
dR
dr
+
λr
−
µ
r
R = 0.
(7.5.11)
The boundary equation (7.5.7) yields
R(a) = 0.
(7.5.12)
What are the other boundary conditions? Check the solution of Laplace’s equation inside a
circle!
Θ(0) = Θ(2π),
(periodicity)
(7.5.13)
Θ
(0) = Θ
(2π),
(7.5.14)
158
|R(0)| < ∞
(boundedness)
(7.5.15)
The equation for Θ(θ) can be solved (see Chapter 2)
µ
m
= m
2
m = 0, 1, 2, . . .
(7.5.16)
Θ
m
=
sin mθ
cos mθ m = 0, 1, 2, . . .
(7.5.17)
In the rest of this section, we discuss the solution of (7.5.11) subject to (7.5.12), (7.5.15).
After substituting the eigenvalues µ
m
from (7.5.16), we have
d
dr
r
dR
m
dr
+
λr
−
m
2
r
R
m
= 0
(7.5.18)
|R
m
(0)
| < ∞
(7.5.19)
R
m
(a) = 0.
(7.5.20)
Using Rayleight quotient for this singular Sturm-Liouville problem, we can show that λ > 0,
thus we can make the transformation
ρ =
√
λr
(7.5.21)
which will yield Bessel’s equation
ρ
2
d
2
R(ρ)
dρ
2
+ ρ
dR(ρ)
dρ
+
ρ
2
− m
2
R(ρ) = 0
(7.5.22)
Consulting a textbook on the solution of Ordinary Differential Equations, we find:
R
m
(ρ) = C
1m
J
m
(ρ) + C
2m
Y
m
(ρ)
(7.5.23)
where J
m
, Y
m
are Bessel functions of the first, second kind of order m respectively. Since we
are interested in a solution satisftying (7.5.15), we should note that near ρ = 0
J
m
(ρ)
∼
1
m = 0
1
2
m
m!
ρ
m
m > 0
(7.5.24)
Y
m
(ρ)
∼
2
π
ln ρ
m = 0
−
2
m
(m−1)!
π
1
ρ
m
m > 0.
(7.5.25)
Thus C
2m
= 0 is necessary to achieve boundedness. Thus
R
m
(r) = C
1m
J
m
(
√
λr).
(7.5.26)
In figure 49 we have plotted the Bessel functions J
0
through J
5
. Note that all J
n
start at
0 except J
0
and all the functions cross the axis infinitely many times. In figure 50 we have
plotted the Bessel functions (also called Neumann functions) Y
0
through Y
5
. Note that the
vertical axis is through x = 3 and so it is not so clear that Y
n
tend to
−∞ as x → 0.
159
J_0
J_1
J_2
J_3
J_4
J_5
Legend
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
2
4
6
8
10
x
Figure 49: Bessel functions J
n
, n = 0, . . . , 5
To satisfy the boundary condition (7.5.20) we get an equation for the eigenvalues λ
J
m
(
√
λa) = 0.
(7.5.27)
There are infinitely many solutions of (7.5.27) for any m. We denote these solutions by
ξ
mn
=
+
λ
mn
a
m = 0, 1, 2, . . .
n = 1, 2, . . .
(7.5.28)
Thus
λ
mn
=
ξ
mn
a
2
,
(7.5.29)
R
mn
(r) = J
m
ξ
mn
a
r
.
(7.5.30)
We leave it as an exercise to show that the general solution to (7.5.1) - (7.5.2) is given by
u(r, θ, t) =
∞
m=0
∞
n=1
J
m
ξ
mn
a
r
{a
mn
cos mθ + b
mn
sin mθ
}
A
mn
cos c
ξ
mn
a
t + B
mn
sin c
ξ
mn
a
t
(7.5.31)
We will find the coefficients by using the initial conditions (7.5.3)-(7.5.4)
α(r, θ) =
∞
m=0
∞
n=1
J
m
ξ
mn
a
r
A
mn
{a
mn
cos mθ + b
mn
sin mθ
}
(7.5.32)
160
Y_0
Y_1
Y_2
Y_3
Y_4
Y_5
Legend
–1.5
–1
–0.5
0
4
5
6
7
8
9
10
x
Figure 50: Bessel functions Y
n
, n = 0, . . . , 5
β(r, θ) =
∞
m=0
∞
n=1
J
m
ξ
mn
a
r
c
ξ
mn
a
B
mn
{a
mn
cos mθ + b
mn
sin mθ
} .
(7.5.33)
A
mn
a
mn
=
(
2π
0
(
a
0
α(r, θ)J
m
ξ
mn
a
r
cos mθrdrdθ
(
2π
0
(
a
0
J
2
m
ξ
mn
a
r
cos
2
mθrdrdθ
,
(7.5.34)
c
ξ
mn
a
B
mn
a
mn
=
(
2π
0
(
a
0
β(r, θ)J
m
ξ
mn
a
r
cos mθrdrdθ
(
2π
0
(
a
0
J
2
m
ξ
mn
a
r
cos
2
mθrdrdθ
.
(7.5.35)
Replacing cos mθ by sin mθ we get A
mn
b
mn
and c
ξ
mn
a
B
mn
b
mn
.
Remarks
1. Note the weight r in the integration. It comes from having λ multiplied by r in
(7.5.18).
2. We are computing the four required combinations A
mn
a
mn
, A
mn
b
mn
, B
mn
a
mn
, and
B
mn
b
mn
. We do not need to find A
mn
or B
mn
and so on.
Example:
Solve the circularly symmetric case
u
tt
(r, t) =
c
2
r
∂
∂r
r
∂u
∂r
,
(7.5.36)
u(a, t) = 0,
(7.5.37)
161
u(r, 0) = α(r),
(7.5.38)
u
t
(r, 0) = β(r).
(7.5.39)
The reader can easily show that the separation of variables give
¨
T + λc
2
T = 0,
(7.5.40)
d
dr
r
dR
dr
+ λrR = 0,
(7.5.41)
R(a) = 0,
(7.5.42)
|R(0)| < ∞.
(7.5.43)
Since there is no dependence on θ , the r equation will have no µ, or which is the same
m = 0. Thus
R
0
(r) = J
0
(
+
λ
n
r)
(7.5.44)
where the eigenvalues λ
n
are computed from
J
0
(
+
λ
n
a) = 0.
(7.5.45)
The general solution is
u(r, t) =
∞
n=1
a
n
J
0
(
+
λ
n
r) cos c
+
λ
n
t + b
n
J
0
(
+
λ
n
r) sin c
+
λ
n
t.
(7.5.46)
The coefficients a
n
, b
n
are given by
a
n
=
(
a
0
J
0
(
√
λ
n
r)α(r)rdr
(
a
0
J
2
0
(
√
λ
n
r)rdr
,
(7.5.47)
b
n
=
(
a
0
J
0
(
√
λ
n
r)β(r)rdr
c
√
λ
n
(
a
0
J
2
0
(
√
λ
n
r)rdr
.
(7.5.48)
162
Do'stlaringiz bilan baham: |