0% found this document useful (0 votes)
82 views3 pages

Solution 4

1. This document summarizes solutions to partial differential equations (PDEs) in rectangular coordinates. 2. Eigenvalues and eigenfunctions are derived for boundary value problems involving the Laplacian operator and other PDEs. 3. Separation of variables is used to find general solutions to initial-boundary value problems as sums of products of spatial and temporal eigenfunctions.

Uploaded by

Chen Shyan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
82 views3 pages

Solution 4

1. This document summarizes solutions to partial differential equations (PDEs) in rectangular coordinates. 2. Eigenvalues and eigenfunctions are derived for boundary value problems involving the Laplacian operator and other PDEs. 3. Separation of variables is used to find general solutions to initial-boundary value problems as sums of products of spatial and temporal eigenfunctions.

Uploaded by

Chen Shyan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

UECM1713 Tutorial : PDEs In Rectangular Coordinates (Solution)

1. Show : u(x, y) = a ln(x2 + y 2 ) + b satisfies uxx + uyy = 0


Solve u = 0 = a ln 1 + b, u = 5 = a ln 9 + b
a = 5/(ln 9), b = 0
2. (a) y 00 + λy = 0, y 0 (0) = y(π) = 0

m2 + λ = 0 ⇒ m = ± −λ
Case 1. λ = −k 2 < 0
m = ±k
y = c1 ekx + c2 e−kx = 0, ∵ y 0 (0) = y(π) = 0 Case 2. λ = 0
m=0
y = c1 + c2 x = 0, ∵ y 0 (0) = y(π) = 0
Case 3 : λ = k 2 > 0
m = ±ki
y 0 = −c1 k sin(kx) + c2 k cos(kx) = c2 = 0, ∵ y 0 (0) = 0, k 6= 0
y = c1 cos(kx) = c1 cos(kπ), ∵ y(0) = 0
For non-0 solns, choose c1 6= 0 so that
k = 2n+12
, n = 0, 1, 2, . . .
 2
∴ λ = 2n+1 2
and yn = cos 2n+1
2
x, n = 0, 1, 2, . . .
(b) y 00 + 2y 0 + λy = 0, y(0) = y(π) =
√0
m2 + 2m + λ = 0 ⇒ m = −1 ± 1 − λ
Case 1 : 1 − λ = −k 2 < 0, k > 0
m = −1 ± ki
y = e−x [c1 cos(kx) + c2 sin(kx)]
y(0) = 0 ⇒ c1 = 0 ⇒ y = c2 e−x sin(kx)
y(π) = 0 ⇒ c2 e−π sin(kπ) = 0
c2 6= 0 ⇒ sin(kπ) = 0 ⇒ k = n = 1, 2, 3, . . .
λ = λn = 1 + k 2 = 1 + n2 , yn = e−x sin(nx),
Case 2 : 1 − λ = 0 or λ = 1
m = −1
y = c1 e−x + c2 xe−x =
y(0) = 0 ⇒ c1 = 0 ⇒ y = c2 xe−x
y(π) = 0 ⇒ c2 πe−π = 0 ⇒ c2 = 0
y=0
Case 3 : 1 − λ = k 2 > 0, k > 0
m = −1 ± k
y = c1 e(−1+k)x + c2 e(−1−k)x
y(0) = 0 ⇒ c1 + c2 = 0 ⇒ c2 = −c1 ⇒ y = c1 e−x ekx − e−kx = 2c1 e−x sinh(kx)
 

y(π) = 0 ⇒ 2c1 e−π sinh(kπ) = 0 ⇒ c2 = 0 since 2e−π sinh(kπ) 6= 0


y=0
(c) x2 y 00 + xy 0 + λy = 0, y 0 (1) = y(eπ ) = 0

Euler equation : m = −λ
Consider 3 cases : λ = k 2> 0, λ = 0, λ= −k 2 < 0; k > 0 (DIY)
 2n − 1 2 2n − 1
λ= , yn = cos ln x , n = 1, 2, . . .
2 2

1
3. u(x, t) = X(x)T (t)
XT 00 = 4X 00 T
X 00 + kX = 0, X(0) = X(π) = 0
T 00 + 4kT = 0, T (0) = 0
k = n2
Xn = sin(nx)
Tn = sin(2nt)
u= ∞
P
n=1 An sin(nx) sin(2nt)
ut (x, 0) = ∞
P
n=1 2nAn sin(nx) = sin 2x
1
u = sin(2x) sin(4t)
4
4. DIY
5. u(x, y) = X(x)Y (y)
X 00 Y 00
X 00 Y + XY 00 = 0 ⇒ =− = λ = k2
X Y
Note: No negative eigenvalues. You can check that the case λ < 0 gives only u = 0.
X 00 − k 2 X = 0, X(0) = 0
Y 00 + k 2 Y = 0, Y 0 (0) = Y 0 (1) = 0
Yn = cos nπy, n = 0, 1, 2, . . .
X0 = x; Xn = sinh nπx, n ≥ 1
X ∞
u(x, y) = A0 x + An sinh nπx cos nπy
n=1

X
u(1, y) = 1 − y = A0 + An sinh nπ cos nπy
n=1
Z 1
1
A0 = (1 − y) dy =
0 2
Z 1
2 [1 − (−1)n ] 4
An sinh nπ = 2 (1 − y) cos nπy dy = = , n = 1, 3, . . . .
0 n2 π 2 n2 π 2

1 4 X sinh(2n − 1)πx cos(2n − 1)πy
u(x, y) = x + 2
2 π n=1 (2n − 1)2 sinh(2n − 1)π

6. (a) Substituting u(x, t) = X(x)T (t) into (i),(ii), (iii), (v) we obtain

X 00 − kX = 0, X(0) = 0, X(1) = 0
T 0 − kT = 0

where k is a constant.
(b) The first DE has eigenfunctions

Xn = sin(nπ) corresponding to k = −n2 π 2 .

With these values of k, the second DE becomes

T 0 + n2 π 2 T = 0, T 0 (0) = 0.

Solving this IVP, we obtain Tn = exp[−n2 π 2 t].


Thus un (x, t) = Xn (x)Tn (t) = sin nπxexp[−n2 π 2 t] satisfies all the conditions except (iv).

2

X ∞
X
(c) Let u(x, t) = An un (x, t) = An sin nπxexp[−n2 π 2 t].
n=1 n=1

X 4
(4) ⇒ u(x, 0) = An sin nπx = 1 ⇒ An =
, n = 1, 3, 5, . . . ; An = 0, n = 2, 4, 6, . . .
n=1

Note : An =the coefficients of the half range sine series of f (x) = 1, 0 < x < 1

X
f ∗ (x) = bn sin nx
n=1
( 4
Z 1
2 h cos nπx i1 2 , n = 1, 3, 5, . . .
bn = sin nπx dx = 2 − = [1 − (−1)n ] = nπ
1 0 nπ 0 nπ 0 , n = 2, 4, 6, . . .
∗ 4 h 1 1 i
Hence, the half range sine series is f (x) = sin x + sin 3x + sin 5x + · · · .
π 3 5
4 −π2 t
 1 −9π2 t 1 −25π2 t 
The solution is u(x, t) = e sin πx + e sin 3πx + e sin 5πx + · · ·
π 3 5
∞ −(2n−1)2 π 2 t
4 X e sin(2n − 1)πx
or u(x, t) =
π n=1 2n − 1
7. Bar with insulated ends
(a) Case 1. k = λ2 > 0 : No nontrivial solutions.
Case 2. k = 0 : X0 = 1

Case 3. k = −λ2 > 0 : λ = λn = nπ L
, X = Xn = cos x, n = 1, 2, . . .
L
Combining Case 2 and Case 3, we conclude that the eigenvalues are
 nπ 2
k=−
L
and that

Xn = cos x
L
are the corresponding eigenfunctions for n = 0, 1, 2, . . . .
(b) X 00 − kX = 0, X 0 (0) = X 0 (π) = 0
L=π
k = −n2 , Xn = cos nx, n = 0, 1, 2, . . . .
T 0 − 4kT = 0
T 0 − 4kT = 0
2
Tn = e4kt = e−4n t , n = 0, 1, 2, . . .

2
X
u(x, t) = a0 + an e−4n t cos nx
n=1

X 1 1
u(x, 0) = a0 + an cos nx = cos2 x = + cos(2x)
n=1
2 2
a0 = 12 , a2 = 21 and other an = 0
1 1
u(x, t) = + e−16t cos(2x)
2 2
 
h 2 1 hπ
8. ψ(x) = − x + + x,
2 π 2
Solve by separation of variables: wxx + wyy = 0, w(0, y) = w(π, y) = 0, w(x, 0) = −ψ(x)

2 (−1)n
   
X
−ny 1 hπ h 2 h n
u(x, y) = An e sin nx + + x − x , An = + 3 [(−1) − 1]
n=1
π 2 2 π n n

You might also like