ENGI 5432 Lecture Notes 4 - PDEs Page 4.

01
4. Partial Differential Equations

Partial differential equations (PDEs) are equations involving functions of more than one
variable and their partial derivatives with respect to those variables.
Most (but not all) physical models in engineering that result in partial differential
equations are of at most second order and are often linear. (Some problems such as
elastic stresses and bending moments of a beam can be of fourth order). In this course
we shall have time to look at only a very small subset of second order linear partial
differential equations.
Sections in this Chapter:
4.1 Major Classifications of Common PDEs
4.2 The Wave Equation dAlembert Solution
4.3 The Wave Equation Vibrating Finite String
4.4 The Maximum-Minimum Principle
4.5 The Heat Equation
ENGI 5432 4.1 - Classification of PDEs Page 4.02
4.1 Major Classifications of Common PDEs
A general second order linear partial differential equation in two Cartesian variables can
be written as
( ) ( ) ( )
2 2 2
2 2
, , , , , , ,
u u u u u
A x y B x y C x y f x y u
x x y y x y
_
+ +


,
Three main types arise, based on the value of D = B
2
4AC (a discriminant):
Hyperbolic, wherever (x, y) is such that D > 0;
Parabolic, wherever (x, y) is such that D = 0;
Elliptic, wherever (x, y) is such that D < 0.
Among the most important partial differential equations in engineering are:
The wave equation:
2
2 2
2
u
c u
t

or its one-dimensional special case
2 2
2
2 2
u u
c
t x

[which is hyperbolic everywhere]
(where u is the displacement and c is the speed of the wave);
The heat (or diffusion) equation:
2
u
K u K u
t

v v
g

a one-dimensional special case of which is
2
2
u K u
t x

[which is parabolic everywhere]
(where u is the temperature, is the specific heat of the medium, is the density and K is
the thermal conductivity);
The potential (or Laplaces) equation:
2
0 u
a special case of which is
2 2
2 2
0
u u
x y

+

[which is elliptic everywhere]
The complete solution of a PDE requires additional information, in the form of initial
conditions (values of the dependent variable and its first partial derivatives at t = 0),
boundary conditions (values of the dependent variable on the boundary of the domain) or
some combination of these conditions.
ENGI 5432 4.2 Wave Equation - dAlembert Solution Page 4.03
4.2 dAlembert Solution
Example 4.2.1
Show that
( )
( ) ( )
,
2
f x ct f x ct
y x t
+ +

is a solution to the wave equation

2 2
2 2 2
1
0
y y
x c t



with initial conditions y(x, 0) = f (x) and
( )
0
, 0
t
y x t
t

for any twice differentiable function f (x).

ENGI 5432 4.2 Wave Equation - dAlembert Solution Page 4.04
Example 4.2.1 (continued)
ENGI 5432 4.2 Wave Equation - dAlembert Solution Page 4.05
A more general dAlembert solution to the wave equation for an infinitely long string is
( )
( ) ( )
( )
1
,
2 2
x ct
x ct
f x ct f x ct
y x t g u du
c
+

+ +
+

This satisfies the wave equation

2 2
2
2 2
for and 0
y y
c x t
t x

< < >

and
Initial configuration of string: y(x, 0) = f (x) for x
and
Initial speed of string:
( )
( )
, 0
for
x
y
g x x
t

for any twice differentiable functions f (x) and g(x).
Physically, this represents two identical waves, moving with speed c in opposite
directions along the string.
Proof that ( ) ( )
1
,
2
x ct
x ct
y x t g u du
c
+

satisfies both initial conditions:

( ) ( ) ( ) ( )
1 1
, , 0 0
2 2
x ct x
x ct x
y x t g u du y x g u du
c c
+

Using a Leibnitz differentiation of the integral:
ENGI 5432 4.2 Wave Equation - dAlembert Solution Page 4.06
Example 4.2.2
An elastic string of infinite length is displaced into the form y = cos x/2 on [1, 1]
only (and y = 0 elsewhere) and is released from rest. Find the displacement y(x, t) at all
locations on the string x and at all subsequent times (t > 0).
See the web page "www.engr.mun.ca/~ggeorge/5432/demos/ex422.html" for an
animation of this solution.
ENGI 5432 4.2 Wave Equation - dAlembert Solution Page 4.07
Example 4.2.2 (continued)
Some snapshots of the solution are shown here:
ENGI 5432 4.2 Wave Equation - dAlembert Solution Page 4.08
A more general case of a dAlembert solution arises for the homogeneous PDE with
constant coefficients
2 2 2
2 2
0
u u u
A B C
x x y y

+ +

The characteristic (or auxiliary) equation for this PDE is
2
0 A B C + +
This leads to the complementary function (which is also the general solution for this
homogeneous PDE)
( ) ( ) ( ) 1 1 2 2
, , u x y f y x f y x + + +
where
1 2
and
2 2
B D B D
A A

+

and D = B
2
4AC
and f
1
, f
2
are arbitrary twice-differentiable functions of their arguments.

1
and
2
are the roots (or eigenvalues) of the characteristic equation.
In the event of equal roots, the solution changes to
( ) ( ) ( ) ( )
1 2
, , u x y f y x h x y f y x + + +
where h(x, y) is any non-trivial linear function of x and/or y (except y + x).
The wave equation is a special case with y = t, A = 1, B = 0, C = 1/c
2
and = 1/c.
ENGI 5432 4.2 Wave Equation - dAlembert Solution Page 4.09
Example 4.2.3
2 2 2
2 2
3 2 0
u u u
x x y y

+

u(x, 0) = x
2

u
y
(x, 0) = 0
(a) Classify the partial differential equation.
(b) Find the value of u at (x, y) = (0, 1).
(a) Compare this PDE to the standard form
2 2 2
2 2
0
u u u
A B C
x x y y

+ +

(b)
ENGI 5432 4.2 Wave Equation - dAlembert Solution Page 4.10
Example 4.2.3 (continued)
ENGI 5432 4.2 Wave Equation - dAlembert Solution Page 4.11
Example 4.2.4
Find the complete solution to
2 2 2
2 2
6 5 14 ,
u u u
x x y y

+

u(x, 0) = 2x + 1 ,
u
y
(x, 0) = 4 6x .
This PDE is non-homogeneous.
For the particular solution, we require a function such that the combination of second
partial derivatives resolves to the constant 14. It is reasonable to try a quadratic function
of x and y as our particular solution.
Try u = ax
2
+ bxy + cy
2

ENGI 5432 4.2 Wave Equation - dAlembert Solution Page 4.12
Example 4.2.4 (continued)
ENGI 5432 4.2 Wave Equation - dAlembert Solution Page 4.13
Example 4.2.5
Find the complete solution to
2 2 2
2 2
2 0 ,
u u u
x x y y

+ +

u = 0 on x = 0 ,
u = x
2
on y = 1 .
ENGI 5432 4.2 Wave Equation - dAlembert Solution Page 4.14
Two-dimensional Laplace Equation
2 2
2 2
0
u u
x y

+


A function f (x, y) is harmonic if and only if
2
f = 0 everywhere inside a domain .
Example 4.2.6
Is u = e
x
sin y harmonic on 1
2
?
ENGI 5432 4.3 Wave Equation - Finite String Page 4.15
4.3 The Wave Equation Vibrating Finite String

The wave equation is
2
2 2
2
u
c u
t

If u(x, t) is the vertical displacement of a point at location x on a vibrating string at time t,

then the governing PDE is
2 2
2
2 2
u u
c
t x

If u(x, y, t) is the vertical displacement of a point at location (x, y) on a vibrating
membrane at time t, then the governing PDE is
2 2 2
2
2 2 2
u u u
c
t x y
_
+


,
or, in plane polar coordinates (r, ), (appropriate for a circular drum),
2 2 2
2
2 2 2 2
1 1 u u u u
c
t r r r r
_
+ +


,
Example 4.3.1
An elastic string of length L is fixed at both ends (x = 0 and x = L). The string is
displaced into the form y = f (x) and is released from rest. Find the displacement y(x, t)
at all locations on the string (0 < x < L) and at all subsequent times (t > 0).
The boundary value problem for the displacement function y(x, t) is:
2 2
2
2 2
for 0 and 0
y y
c x L t
t x

< < >

Both ends fixed for all time:
Initial configuration of string:
String released from rest:
ENGI 5432 4.3 Wave Equation - Finite String Page 4.16
Example 4.3.1 (continued)
Separation of Variables (or Fourier Method)
Attempt a solution of the form y(x, t) = X(x) T(t)
Substitute y(x, t) = X(x) T(t) into the PDE:
( ) ( ) ( ) ( ) ( ) ( )
2 2 2 2
2 2
2 2 2 2
d T d X
X x T t c X x T t X c T
t x dt dx



2 2
2 2 2
1 1 1 d T d X
c T dt X dx

ENGI 5432 4.3 Wave Equation - Finite String Page 4.17
Example 4.3.1 (continued)
Therefore our complete solution is
( ) ( )
0 1
2
, sin sin cos
L
n
n u n x n ct
y x t f u du
L L L L

_
_ _ _



, , ,
,

ENGI 5432 4.3 Wave Equation - Finite String Page 4.18
Example 4.3.1 (continued)
This solution is valid for any initial displacement function f (x) that is continuous with a
piece-wise continuous derivative on [0, L] with f (0) = f (L) = 0.
If the initial displacement is itself sinusoidal ( ) sin for some
n x
f x a n
L
_ _


, ,

,
then the complete solution is a single term from the infinite series,
( ) , sin cos
n x n ct
y x t a
L L
_ _

, ,
.
Suppose that the initial configuration is triangular:
( ) ( )
( )
( )
1
2
1
2
0
, 0
x x L
y x f x
L x L x L

'
<

Then the Fourier sine coefficients are

( )
0
2
sin
L
n
n u
C f u du
L L
_

ENGI 5432 4.3 Wave Equation - Finite String Page 4.19

Example 4.3.1 (continued)
See the web page "www.engr.mun.ca/~ggeorge/5432/demos/ex431.html" for an
animation of this solution.
ENGI 5432 4.3 Wave Equation - Finite String Page 4.20
Example 4.3.1 (continued)
Some snapshots of the solution are shown here:
These graphs were generated from the Fourier
series, truncated after the fifth non-zero term.
ENGI 5432 4.3 Wave Equation - Finite String Page 4.21
Example 4.3.2
An elastic string of length L is fixed at both ends (x = 0 and x = L). The string is initially
in its equilibrium state [y(x, 0) = 0 for all x] and is released with the initial velocity
( )
( )
,0 x
y
g x
t

. Find the displacement y(x, t) at all locations on the string (0 < x < L)
and at all subsequent times (t > 0).
The boundary value problem for the displacement function y(x, t) is:
2 2
2
2 2
for 0 and 0
y y
c x L t
t x

< < >

Both ends fixed for all time: y(0, t) = y(L, t) = 0 for t > 0
Initial configuration of string: y(x, 0) = 0 for 0 < x < L
String released with initial velocity:
( )
( )
, 0
for 0
x
y
g x x L
t

As before, attempt a solution by the method of the separation of variables.

Substitute y(x, t) = X(x) T(t) into the PDE:
( ) ( ) ( ) ( ) ( ) ( )
2 2 2 2
2 2
2 2 2 2
d T d X
X x T t c X x T t X c T
t x dt dx



Again, each side must be a negative constant.
2 2
2
2 2 2
1 1 d T d X
c T dt X dx

We now have the pair of ODEs
2 2
2 2 2
2 2
0 and 0
d X d T
X c T
dx dt
+ +
The general solutions are
( ) ( ) ( ) ( ) ( ) ( ) cos sin and cos sin X x A x B x T t C ct D ct + +
respectively, where A, B, C and D are arbitrary constants.
ENGI 5432/5435 4.3 Wave Equation - Finite String Page 4.22
Example 4.3.2 (continued)
Consider the boundary conditions:
( ) ( ) ( ) 0, 0 0 0 y t X T t t
For a non-trivial solution, this requires ( ) 0 0 0 X A
.
( ) ( ) ( ) ( ) , 0 0 0 y L t X L T t t X L
( ) ( ) sin 0 ,
n
n
B L n
L

We now have a solution only for a discrete set of eigenvalues
n

, with corresponding
eigenfunctions
( ) ( ) sin , 1, 2, 3,
n
n x
X x n
L
_


,
K
and
( ) ( ) ( ) ( ) ( ) , sin , 1, 2, 3,
n n n n
n x
y x t X x T t T t n
L
_


,
K
So far, the solution has been identical to Example 4.3.1.
Consider the initial condition y(x, 0) = 0 :
( ) ( ) ( ) ( ) , 0 0 0 0 0 0 y x X x T x T
The initial value problem for T(t) is now
( )
2 2
0 , 0 0 , where
n
T c T T
L

+
the solution to which is
( ) ( ) sin ,
n n
n ct
T t C n
L
_


,

Our eigenfunctions for y are now
( ) ( ) ( ) ( ) , sin sin ,
n n n n
n x n ct
y x t X x T t C n
L L
_ _


, ,

ENGI 5432/5435 4.3 Wave Equation - Finite String Page 4.23
Example 4.3.2 (continued)
Differentiate term by term and impose the initial velocity condition:
( )
( )
, 0
1
sin
x
n
n
y n c n x
C g x
t L L

_ _

, ,

which is just the Fourier sine series expansion for the function g(x).
The coefficients of the expansion are
( )
0
2
sin
L
n
n c n u
C g u du
L L L
_

which leads to the complete solution

( ) ( )
0 1
2 1
, sin sin sin
L
n
n u n x n ct
y x t g u du
c n L L L

_
_ _ _



, , ,
,

This solution is valid for any initial velocity function g(x) that is continuous with a
piece-wise continuous derivative on [0, L] with g(0) = g(L) = 0.
The solutions for Examples 4.3.1 and 4.3.2 may be superposed.
Let y
1
(x, t) be the solution for initial displacement f (x) and zero initial velocity.
Let y
2
(x, t) be the solution for zero initial displacement and initial velocity g(x).
Then y(x, t) = y
1
(x, t) + y
2
(x, t) satisfies the wave equation
(the sum of any two solutions of a linear homogeneous PDE is also a solution),
and satisfies the boundary conditions y(0, t) = y(L, t) = 0 .
y(x, 0) = y
1
(x, 0) + y
2
(x, 0) = f (x) + 0,
which satisfies the condition for initial displacement f (x).
y
t
(x, 0) = y
1t
(x, 0) + y
2t
(x, 0) = 0 + g(x),
which satisfies the condition for initial velocity g(x).
Therefore the sum of the two solutions is the complete solution for initial displacement
f (x) and initial velocity g(x):
( ) ( )
( )
0
0
1
1
2
, sin sin cos
2 1
sin sin sin
L
n
L
n
n u n x n ct
y x t f u du
L L L L
n u n x n ct
g u du
c n L L L

_
_ _ _



, , ,
,
_
_ _ _
+


, , ,
,

ENGI 5432/5435 4.4 - Max-Min Principle Page 4.24

4.4 The Maximum-Minimum Principle
Let be some finite domain on which a function u(x, y) and its second derivatives are
defined. Let

be the union of the domain with its boundary.

Let m and M be the minimum and maximum values respectively of u on the boundary
of the domain.
If
2
u 0 in , then u is subharmonic and
( ) ( ) or in u M u M < r r r
v v v
If
2
u 0 in , then u is superharmonic and
( ) ( ) or in u m u m > r r r
v v v
If
2
u = 0 in , then u is harmonic (both subharmonic and superharmonic) and
u is either constant on

or m < u < M everywhere on .

Example 4.4.1

2
u = 0 in : x
2
+ y
2
< 1 and u(x, y) = 1 on C : x
2
+ y
2
= 1.
Find u(x, y) on .
ENGI 5432/5435 4.4 - Max-Min Principle Page 4.25
Example 4.4.2

2
u = 0 in the square domain : 2 < x < +2, 2 < y < +2 .
On the boundary C, on the left and right edges (x = 2), u(x, y) = 4 y
2
,
while on the top and bottom edges (y = 2), u(x, y) = x
2
4.
Find bounds on the value of u(x, y) inside the domain .
ENGI 5432/5435 4.5 - Heat Equation Page 4.26
4.5 The Heat Equation
For a material of constant density

, constant specific heat and constant thermal

conductivity K, the partial differential equation governing the temperature u at any
location (x, y, z) and any time t is
2
, where
u K
k u k
t

Example 4.5.1
Heat is conducted along a thin homogeneous bar extending from x = 0 to x = L. There is
no heat loss from the sides of the bar. The two ends of the bar are maintained at
temperatures T
1
(at x = 0) and T
2
(at x = L). The initial temperature throughout the bar at
the cross-section x is f (x).
Find the temperature at any point in the bar at any subsequent time.
The partial differential equation governing the temperature u(x, t) in the bar is
2
2
u u
k
t x

together with the boundary conditions
u(0, t) = T
1
and u(L, t) = T
2

and the initial condition
u(x, 0) = f (x)
[Note that if an end of the bar is insulated, instead of being maintained at a constant
temperature, then the boundary condition changes to
( ) 0, 0
u
t
t

or
( ) , 0
u
L t
t

.]
Attempt a solution by the method of separation of variables.
u(x, t) = X(x) T(t)
T X
X T k X T k c
T X


Again, when a function of t only equals a function of x only, both functions must equal
the same absolute constant. Unfortunately, the two boundary conditions cannot both be
satisfied unless T
1
= T
2
= 0. Therefore we need to treat this more general case as a
perturbation of the simpler (T
1
= T
2
= 0) case.
ENGI 5432/5435 4.5 - Heat Equation Page 4.27
Example 4.5.1 (continued)
Let u(x, t) = v(x, t) + g(x)
Substitute this into the PDE:
( ) ( ) ( ) ( ) ( ) ( ) ( )
2 2
2 2
, ,
v v
v x t g x k v x t g x k g x
t x t x
_
+ + +


,
This is the standard heat PDE for v if we choose g such that g"(x) = 0.
g(x) must therefore be a linear function of x.
We want the perturbation function g(x) to be such that
u(0, t) = T
1
, u(L, t) = T
2

and
v(0, t) = v(L, t) = 0
Therefore g(x) must be the linear function for which g(0) = T
1
and g(L) = T
2
.
It follows that
( )
2 1
1
T T
g x x T
L
_
+

,
and we now have the simpler problem
2
2
v v
k
t x

together with the boundary conditions
v(0, t) = v(L, t) = 0
and the initial condition
v(x, 0) = f (x) g(x)
Now try separation of variables on v(x, t) :
v(x, t) = X(x) T(t)
1 T X
X T k X T c
k T X


But v(0, t) = v(L, t) = 0 X(0) = X(L) = 0
This requires c to be a negative constant, say
2
.
The solution is very similar to that for the wave equation on a finite string with fixed ends
(section 4.3). The eigenvalues are
n
L

and the corresponding eigenfunctions are

any non-zero constant multiples of
( ) sin
n
n x
X x
L
_

,
ENGI 5432/5435 4.5 - Heat Equation Page 4.28
Example 4.5.1 (continued)
The ODE for T(t) becomes
2
0
n
T k T
L
_
+

,
whose general solution is
( )
2 2 2
/ n kt L
n n
T t c e

Therefore
( ) ( ) ( )
2 2
2
, sin exp
n n n n
n x n kt
v x t X x T t c
L L
_
_


,
,
If the initial temperature distribution f (x) g(x) is a simple multiple of
sin
n x
L
_

,
for
some integer n, then the solution for v is just ( )
2 2
2
, sin exp
n
n x n kt
v x t c
L L
_
_


,
,
.
Otherwise, we must attempt a superposition of solutions.
( )
2 2
2
1
, sin exp
n
n
n x n kt
v x t c
L L

_
_


,
,

such that ( ) ( ) ( )
1
, 0 sin
n
n
n x
v x c f x g x
L

_


,
.
The Fourier sine series coefficients are ( ) ( ) ( )
0
2
sin
n
L
n z
c f z g z dz
L L
_


,

so that the complete solution for v(x, t) is

( ) ( )
2 2
2 1
1 2
0 1
2
, sin sin exp
L
n
T T n z n x n kt
v x t f z z T dz
L L L L L

_
_ _ _ _



, , ,
,
,

and the complete solution for u(x, t) is

( ) ( )
2 1
1
, ,
T T
u x t v x t x T
L
_
+ +

,
Note how this solution can be partitioned into a transient part v(x, t) (which decays to
zero as t increases) and a steady-state part g(x) which is the limiting value that the
temperature distribution approaches.
ENGI 5432/5435 4.5 - Heat Equation Page 4.29
Example 4.5.1 (continued)
As a specific example, let k = 9, T
1
= 100, T
2
= 200, L = 2 and
f (x) = 145x
2
240x + 100 , (for which f (0) = 100, f (2) = 200 and f (x) > 0 x).
Then ( )
200 100
100 50 100
2
g x x x

+ +
The Fourier sine series coefficients are
The complete solution is
( )
( )
2 2
3 3
1
1 1 2320 9
, 50 100 sin exp
2 4
n
n
n x n t
u x t x
n

_
_
_
+


,
,
,

Some snapshots of the temperature distribution (from the tenth partial sum) from the
Maple file at "www.engr.mun.ca/~ggeorge/5432/demos/ex451.mws" are shown
on the next page.
ENGI 5432/5435 4.5 - Heat Equation Page 4.30
Example 4.5.1 (continued)
ENGI 5432/5435 4.5 - Heat Equation Page 4.31
The steady state distribution is nearly attained in much less than a second!
END OF CHAPTER 4
END OF ENGI. 5432!