ENG2005 Workshop W11-2
ENG2005 Workshop W11-2
Workshop Week 11
∂2u ∂2u
+ 2 =0
∂x2 ∂y
In general, this equation governs physical processes where equilibrium has been reached.
Laplace’s equation is also known as the potential equation and also as the steady state heat equation.
It also serves as the prototype for a wider class of elliptic equations. Unlike the heat and wave
equations, there are no initial conditions and the boundary conditions completely specify
the solution.
Both variables are spatial variables and each variable occurs in a 2-nd order derivative and so we’ll
need two boundary conditions for each variable.
1
1 General Solutions
Laplace’s Equation arises as a steady state problem for the Heat or Wave Equations that do not vary with
∂2u
time so that ∂u
∂t = 0 = ∂t2 .
2D:
∂2u ∂2u
∆u = + 2 =0
∂x2 ∂y
3D:
∂2u ∂2u ∂2u
∆u = + 2 + 2 = 0.
∂x2 ∂y ∂z
Determine whether each of the following functions satisfy the Laplace equation.
2
2 Identifying Boundary Conditions
(a) If u(x, y) represents the temperature across the rectangular plate, each side of the rectangular region
should have a standard boundary condition: either a fixed temperature distribution or is insulated.
In the figure above, the four boundary conditions are shown in a simplified form.Write down the correct
mathematical relation for all boundary conditions shown in the figure.
Example: u(0, y) = ....
(b) The 4 BCs are defined below. Place them in the correct place of the plate:
u(0, y) = f1 (y)
u(a, y) = f2 (y)
u(x, 0) = g1 (x)
u(x, b) = g2 (x)
(c) Sketch the plate, label all important elements of the graph and add the boundary conditions.
∂u
=0
∂x (0,y)
∂u
=0
∂x (W,y)
u(x, 0) = sin(x)
u(x, l) = x
3
(d) Suppose we wish to find the steady-state temperature u(x, y) in a rectangular plate whose vertical
edges x = 0 and x = a are insulated, as shown in the figure:
Complete the statements: When no heat escapes from the lateral faces of the left plate, we solve the
following boundary-value problem:
∂2u ∂2u
+ 2 = 0.
∂x2 ∂y
≤x≤ , ≤y≤
with boundary-values:
(e) A solution u(r, θ) of the Laplace equation on a disk of radius 3 is shown in the image
Identify which one of the following functions will describe the boundary condition: on r = 3 the
condition is u(3, θ) = f (θ) for 0 ≤ θ < 2π
(iv) f (θ) = 3
4
3 Solving the Laplace equation
A thin rectangular plate coincides with the region defined by 0 < x < 1, 0 < y < 1. The left end is held
at temperature 0 and the right end is held at temperature 0 for 0 < y < 1. The top of the plate is held at
temperature sin(πx) and the bottom of the plate is also held at temperature 0. Calculate the temperature
distribution inside the plate.1
(a) Solutions of Laplace’s equation represent equilibrium or steady states in a system. T/F?
(b) In this section we will concentrate on the two-dimensional Laplace equation; we consider a thin (two-
dimensional) plate. How many boundary conditions does this equation need?
(c) We will find the solution of Laplace’s equation on the square plate
∂2u ∂2u
+ 2 = 0 for 0 ≤ x ≤ 1 and 0 ≤ y ≤ 1
∂x2 ∂y
with the boundary conditions:
u(0, y) = 0 for y ∈ (0, 1)
u(1, y) = 0 for y ∈ (0, 1)
u(x, 0) = 0 for x ∈ (0, 1)
u(x, 1) = sin(πx) for x ∈ (0, 1)
(d) Let’s seek a solution of the Laplace equation in the form u(x, y) = X(x)Y (y). This method is widely
used for finding solutions to PDEs and it is known as the “separation of variables”. The PDE after
substitution becomes:
d2 X d2 Y
+ =0
dx2 dy 2
We then introduce the separation constant λ to obtain two ordinary differential equations of simple-
harmonic type:
d2 X
=λ
dx2
d2 Y
= −λ
dy 2
(e) Write the boundary conditions from part (c) in terms of the functions X(x) and Y (y).
Which function, X(x) or Y (y), has homogeneous boundary conditions?
1
Using the detailed work from the example ”Two dimensional Laplace equation - boundary conditions” let solve this problem
- this is Example 6.4.1
5
(f) Solve the boundary value problem for X(x) by considering the three signs for λ.
(g) For each non-trivial solution Xn (x) found above, solve for Yn (y) making use of the one remaining
homogeneous boundary condition. Combine them to form the general solution.
(h) Use the remaining boundary condition to find the temperature distribution on the plate.
(i) Use MATLAB to plot the steady-state temperature on the plate as a surface. Describe the solution
below. Where is the maximum temperature? Where is the minimum temperature?
6
4 Another Laplace equation
A thin rectangular plate coincides with the region defined by 0 < x < 4, 0 < y < 1. The left end is held at
temperature 0 and the right end is held at temperature 0. The top of the plate is held at temperature zero
and the bottom of the plate is held at temperature sin(2πx).
(a) Calculate the temperature distribution inside the plate following the model.2
2
Note that this BC is slightly different that the example from Section 3
7
(b) Create a contour plot of the solution u(x, y) in MATLAB, showing. Describe the solution below.
Where on the plate is the temperature equal to zero?
8
5 Laplace’s Equation on Circular Domains
(a) The Laplace equation in polar coordinates is
∂2u 1 ∂u 1 ∂2u
+ + =0
∂r2 r ∂θ2
∂
Note that on occasions this can be rewritten as
!
1 ∂ ∂u 1 ∂2u
+ =0
r ∂r ∂r ∂θ2
The Laplace equation in polar coordinates arises when seeking steady state solutions over geometries
such as discs, half discs and annuli. This is because the method of separation of variables relies on the
boundaries being described by constant functions. In this section we will concentrate on finding the
general solution of Laplace’s equation in polar coordinates over a disc of radius a units:
0≤r≤ ,0 ≤ θ <
(b) Consider a thin disc that is insulated on either face. This means we will have only one side (compared
to the four sides of a rectangular region) which will require a boundary condition, since the condition
is on r = a then the condition will depend upon θ at most, that is,
Assume u represents a physical quantity such as temperature. Physical intuition suggests that at any
given point (r, θ) on the disc we would expect only one value of u, that is, u is not multi-valued. To
guarantee this condition we have
Furthermore, we expect the value of u at each and every point (r, θ) on the disc will be a finite value,
that is,
u(r, θ) is for all 0 ≤ r ≤ a, 0 ≤ θ < 2π
You may be concerned that there are only three conditions when you would have expected four. For
the moment, let us see what happens if we attempt to find the general solution with only these three
conditions.
Consider separation of variables of Laplace’s equation in polar coordinates u(r, θ) = R(r)Θ(θ). The
two ordinary differential equations that result from using the separation of variables (see 6.5.3) are
discussed below.
9
The constant n must be an integer because
(e) Write down the general solution of Laplace’s equation in circular polar coordinates.
Hint: Using both cases n = 0 and n2 put everything in a single expression:
∞
X Dn
u(r, θ) = + Cn
+ An
+ Bn sin(nθ)
n=1
10
6 Bonus: Laplace’s Equation on an annulus
This seems like a long question but it is not. It was an exam question in the past.
(a) Consider Laplace’s equation on an annulus (a domain bounded by two concentric circles):
(i) Suppose u and v each satisfy (L) on the annulus. Then w = c1 u + c2 v also satisfies (L) on the
annulus for any constants c1 and c2 .
(ii) Suppose u satisfies (L) on the annulus. Then w = c1 u + r2 (cos2 θ − sin2 θ) also satisfies (L) on
the annulus for any constant c1 and c2 .
(iii) Suppose u and v each satisfy (L) on the annulus. Then w = uv also satisfies (L) on the annulus.
(iv) Suppose that u satisfies (L) on the annulus, and also the boundary values u(1, θ) = 0 and
u(2, θ) = 0. Then u = 0 everywhere in the annulus.
Answer
∂u
u(1, θ) = 0 and (2, θ) = 1. (B)
∂r
Identify all the correct statements (more than one is possible):
(i) Suppose u and v each satisfy (L) on the annulus and the boundary conditions (B). Then w =
c1 u + c2 v also satisfies (L) on the annulus and boundary conditions (B) for any constants c1 and
c2 .
(ii) u(r, θ) = 2 log r is a solution of (L) with boundary conditions (B).
(iii) u(r, θ) = r − 1 is a solution of (L) with boundary conditions (B).
(iv) u(r, θ) = ∞ n
P
n=1 r An cos(nθ) + Bn sin(nθ) is a solution of (L) with boundary conditions (B), for
some choice of constants An and Bn .
Answer
11