Steady State- 2D-Heat Equation
Kalpana Mahalingam
March 2020
In this section, we study the solutions of 2D heat equation(steady
state-2D heat equation) that do not vary with time, also called as
2D-Laplacian.
2 ∂ 2u ∂ 2u
∇ u= 2+ 2 (1)
∂x ∂y
A heat problem then consists of this PDE to be considered in some re-
gion R of the xy-plane and a given boundary condition on the boundary
curve C of R. This is a boundary value problem (BVP). If u is pre-
scribed on C (“Dirichlet boundary condition”) then the BVP is called
First BVP or Dirichlet Problem.
Dirichlet Problem on a rectangle :
We consider a Dirichlet problem for Laplace’s equation (1) in a
rectangle R, assuming that the temperature equals a given function
f (x) on the upper side and 0 on the other three sides of the rectangle
and 0 ≤ x ≤ a, 0 ≤ y ≤ b.
Separation of Variables :
1
K.Mahalingam 2D-Laplacian
We use separation of variables technique to solve the given problem.
Let,
u(x, y) = F (x)G(y)
Then, the equation (1) becomes,
00 00
00 00 F (x) G (y)
F (x)G(y) + F (x)G (y) = 0, =⇒ =− =k
F (x) G(y)
where k is a constant.
We now obtain two ODE’s
00 00
F (x) − kF (x) = 0, G (y) + kG(y) = 0
Solving, for F , we get,
√ √
kx − kx
F (x) = Ae + Be
Substituting the boundary conditions F (0) = F (a) = 0, we get,
√ √
F (0) = A + B = 0, F (a) = A(e ka − e− ka
)=0
nπ 2
=⇒ k =
a
and hence, nπx
F (x) = Fn (x) = sin
a
Now solving for G similarly, we get
nπy nπy
Gn (y) = An e a + Bn e− a
Now applying the boundary condition u = 0 on the lower side of rect-
angle, we get Gn (0) = 0 which implies,
Gn (0) = An + Bn = 0, =⇒ An = −Bn
K.Mahalingam 2D-Laplacian
Now applying the condition u(x, b) = f (x) implies that G(b) = f (x)
and substituting in Gn , we get
nπb nπb nπy
Gn (b) = An (e a − An e− a ) = 2An sinh
a
Hence, the general solution of the 2D-Laplacian is
nπx nπy
∗
u(x, y) = Fn (x)Gn (y) = An sin sinh
a a
Since, the given equation is linear, a linear combination of solutions is
also a solution and hence,
X nπx nπy
∗
u(x, y) = An sin sinh
n≥0
a a
We now apply the last boundary condition u(x, b) = f (x) implies
G(b) = f (x). Then,
X nπx nπb
u(x, b) = f (x) = A∗n sin sinh
n≥0
a a
and hence, by Fourier series,
nπb 2 Z a nπx
∗
An sinh = f (x) sin dx
a a 0 a
Hence, X nπx nπy
u(x, y) = A∗n sin sinh
n≥0
a a
with Z a
2 nπx
A∗n = f (x) sin dx
a sinh nπb 0 a
a