Chapter 11.
Partial Differential Equations
11.1
Differential Equations
Mathematical models of physical phenomena often
involve differential equations with one or more independent variable. For instance, a model of heat
conduction in a region (e.g. an ocean, the atmosphere, or a system of pipes) will normally involve
three-space variables for coordinates of points in the
region and one time variable, as well as other physical
information.
11.1.1
Ordinary differential equation
An ordinary differential equation (o.d.e.) is an
equation that involves an unknown function y(x) of
MA1505 Chapter 11. Partial Differential Equations
exactly one independent variable x and derivatives
of y. We also call y the dependent variable.
11.1.2
Example
(i) y 0 xy = 0
where y 0 =
dy
dx
is the derivative of y w.r.t. x.
This is an o.d.e. that involves the function y(x) with
one independent variable x.
(ii) y 00 3y 0 + 2y = 0
0
where y =
dy
dx
00
and y =
d2 y
dx2
Again, in this o.d.e., x is the only independent variable of the function y.
11.1.3
MA1505 Chapter 11. Partial Differential Equations
Partial differential equation
A partial differential equation (p.d.e.) is an
equation containing an unknown function u(x, y, . . .)
of two or more independent variables x, y, . . . and
its partial derivatives with respect to these variables.
We also call u the dependent variable.
11.1.4
Example
(i) uxy 2x + y = 0
This is a p.d.e. that involves the function u(x, y)
with two independent variables x and y.
(ii) wxy + x(wz )2 = yz
This is a p.d.e. that involves the function w(x, y, z)
with three independent variables x, y and z.
3
11.1.5
MA1505 Chapter 11. Partial Differential Equations
Solutions of Differential Equations
A solution of a differential equation is any function
which satisfies the equation identically.
There are usually one or more family of solutions for
a differential equation. We call such a family of solutions a general solution of the differential equation.
A specific function from the general solution is called
a particular solution of the differential equation.
11.1.6
Example
Let us substitute y = ex
2 /2
11.1.2 (i):
in the o.d.e of example
MA1505 Chapter 11. Partial Differential Equations
By differentiating y = ex
x2 /2
y =e
d
dx
2 /2
using chain rule, we have
2
x
2
On the other hand, xy = xex
So y = ex
2 /2
x2 /2
= xe
2 /2
satisfies the o.d.e. and hence is a par-
ticular solution of the o.d.e.
In fact, we can easily check that y = kex
2 /2
satisfies
the o.d.e. for any constant k.
So a general solution of the o.d.e is y = kex
11.1.7
2 /2
Example
The function
1 2
u(x, y) = x y xy + F (x) + G(y)
2
2
(1)
MA1505 Chapter 11. Partial Differential Equations
is a general solution of the p.d.e. in example 11.1.4
(i). Here F and G can be any (arbitrary) single
variable functions.
Indeed, by taking partial derivatives of (1):
1
ux = 2xy y 2 + F 0(x) and
2
uxy = 2x y,
we see that the function (1) satisfies the p.d.e.
If we set F (x) = 3 sin x and G(y) = 4y 5 6, we get
the particular solution
1
u(x, y) = x2y xy 2 + 3 sin x + 4y 5 6.
2
Suppose we require the p.d.e. to also satisfy the conditions
u(x, 0) = x3 and u(0, y) = sin(3y).
6
MA1505 Chapter 11. Partial Differential Equations
Then using (1), we have
x3 = u(x, 0) = F (x) + G(0)
and
sin(3y) = u(0, y) = F (0) + G(y).
In this case, we can simply take F (x) = x3 and
G(y) = sin(3y) and get the particular solution
1 2
u(x, y) = x y xy + x3 + sin(3y)
2
2
which satisfy the additional conditions.
11.1.8
Exercise
Consider the o.d.e
ay 00 + by 0 + cy = 0 ()
(where a, b, c are some real numbers).
7
MA1505 Chapter 11. Partial Differential Equations
Let k1 and k2 be the solutions of the quadratic equation at2 + bt + c = 0.
Show that the o.d.e. () has the following general
solutions:
(i) If k1 and k2 are real and distinct, then
y = Aek1x + Bek2x
is a general solution of ().
(ii) If k1 and k2 are real and equal, then
y = Aek1x + Bxek1x
is a general solution of ().
(iii) If k1 = + i and k2 = i are complex, then
y = Aex cos x + Bex sin x
8
MA1505 Chapter 11. Partial Differential Equations
is a general solution of ().
Here A and B are arbitrary constants.
11.1.9
Example
In general, the totality of solutions of a p.d.e. is very
large.
The Laplace equation uxx + uyy = 0 has the following solutions
u(x, y) = x2 y 2,
u(x, y) = ex cos y,
u(x, y) = ln (x2 + y 2),
etc
which are entirely different from each other.
10
11.1.10
MA1505 Chapter 11. Partial Differential Equations
Order of Differential Equations
The order of the p.d.e. is the order of the highest
derivative present.
Example 11.1.2 (i) is an o.d.e. of order 1 while (ii) is
an o.d.e. of order 2.
Example 11.1.4 (i) is a p.d.e. of order 2 and (ii) is
also a p.d.e. of order 2.
11.1.11
Linearity and Homogeneity
An order 1 linear o.d.e. has the form
Ay 0 + By = Z
and an order 2 linear o.d.e. has the form
Ay 00 + By 0 + Cy = Z
10
11
MA1505 Chapter 11. Partial Differential Equations
where A, B, C, Z are constants or functions of x but
not functions of y.
An order 1 linear p.d.e. has the form
Aux + Buy + Cu = Z
and an order 2 linear p.d.e. has the form
Auxx + Buxy + Cuyy + Dux + Euy + F u = Z
where A, B, C, D, E, F, Z are constants or functions
of x and y but not functions of u.
An order 1 or 2 linear o.d.e. or p.d.e. is said to be
homogeneous if the Z term in the above form is 0.
Generalizations to differential equation of higher order and those with more than two independent variables can be easily made.
11
12
11.1.12
MA1505 Chapter 11. Partial Differential Equations
Example
p.d.e.
order linear homogeneous
4uxx ut = 0
2
yes
yes
x2Ryyy = y 3Rxx
3
yes
yes
tutx + 2ux = x2
2
yes
no
4uxx uut = 0
2
no
n.a.
(ux)2 + (uy )2 = 2
1
no
n.a.
11.1.13
Superposition Principle
If u1 and u2 are any solutions of a linear homogeneous
differential equation, then
u = c1u1 + c2u2,
where c1 and c2 are any constants, is also a solution
of that equation.
12
13
MA1505 Chapter 11. Partial Differential Equations
11.1.14
Example
Referring to the particular solutions of Laplace equation uxx + uyy = 0 in Example 11.1.9, by superposition principle,
u(x, y) = 3(x2 y 2) 7ex cos y + 10 ln (x2 + y 2)
is again a solution of the Laplace equation.
11.2
Solving Differential Equations
In this section, we will demonstrate some techniques
to solve certain types of simple o.d.e. and p.d.e.
11.2.1
Separable O.D.E.
An order 1 o.d.e is separable if it has the form
A(y)y 0 + B(x) = 0.
13
14
MA1505 Chapter 11. Partial Differential Equations
We can separate the variables by rewriting the equation as
A(y)dy + B(x)dx = 0.
We can then solve the o.d.e. by integration:
Z
Z
A(y)dy +
B(x)dx = c
where c is an arbitrary constant.
11.2.2
Example
The o.d.e. y 0 xy = 0 in example 11.1.2 (i) can be
rewritten as
1 0
y x = 0.
y
14
15
MA1505 Chapter 11. Partial Differential Equations
So it is a separable o.d.e. and can be solved as follows:
Z
x2
xdx = 0 ln |y|
=c
2
x2
2
ln |y| = c +
|y| = ecex /2
2
1
dy
y
y = Kex
2 /2
where K = ec is an arbitrary constant.
So we get the general solution as in example 11.1.6.
11.2.3
Reducing P.D.E. to O.D.E.
To solve a P.D.E., it is very common to first reduce
it to an O.D.E. before solving it. Here we illustrate
two situations where we can use this approach.
15
16
11.2.4
MA1505 Chapter 11. Partial Differential Equations
Example
(Absence of one partial derivative)
uxx u = 0
(2)
Since there is no yderivative in p.d.e. (2), we treat
y as constant and regard the p.d.e. as an o.d.e. in x:
u00(x) u(x) = 0
(3)
To solve this o.d.e., note that it is of the form
au00 + bu0 + cu = 0
with a = 1, b = 0, c = 1.
The quadratic equation t2 1 = 0 has solutions t =
1. So by exercise 11.1.8, so a general solution of
16
17
MA1505 Chapter 11. Partial Differential Equations
the o.d.e. (3) has the form
u(x) = Aex + Bex,
where A and B are constant w.r.t. x. Thus the constants A and B may in fact be functions of y. Hence,
a general solution of the p.d.e. (2) is
u(x, y) = A(y)ex + B(y)ex.
11.2.5
Example
(Common inner derivative)
uxy = ux
(4)
Set ux = p so that the p.d.e. (4) may be viewed as
an o.d.e.
py = p.
17
(5)
18
MA1505 Chapter 11. Partial Differential Equations
This is a separable o.d.e. So we can solve it as in
11.2.1:
Z
1
dp +
p
Z
ydy = 0 ln |p| + y = c
ln |p| = c y |p| = ecey
p = Key
(6)
which is a general solution of (5). Here K is a constant w.r.t. y, so it may be a function of x.
Substituting p by ux in (6) and integrate w.r.t. x,
we obtain a general solution of the p.d.e. (4):
Z
u(x, y) =
K(x)ey dx = f (x)ey + g(y),
Z
where f (x) =
K(x)dx and g(y) is an arbitrary
function in y.
18
19
11.2.6
MA1505 Chapter 11. Partial Differential Equations
Separation of Variables for P.D.E.
This method can be used to solve p.d.e. involving
two independent variables, say x and y, that can be
separated from each other in the p.d.e. There are
similarities between this method and the technique
of separating variables for o.d.e. in (11.2.1). We first
make an observation:
Suppose u(x, y) = X(x)Y (y).
Then
(i) ux(x, y) = X 0(x)Y (y)
(ii) uy (x, y) = X(x)Y 0(y)
(iii) uxx(x, y) = X 00(x)Y (y)
19
20
MA1505 Chapter 11. Partial Differential Equations
(iv) uyy (x, y) = X(x)Y 00(y)
(v) uxy (x, y) = X 0(x)Y 0(y)
Notice that each derivative of u remains separated
as a product of a function of x and a function of y.
We exploit this feature as follows:
11.2.7
Illustration of Separation of Variables
Consider a p.d.e. of the form
ux = f (x)g(y)uy .
If a solution of the form u(x, y) = X(x)Y (y) exists,
then we obtain
X 0(x)Y (y) = f (x)g(y)X(x)Y 0(y)
i.e.,
1
X 0(x)
Y 0(y)
= g(y)
.
f (x) X(x)
Y (y)
20
21
MA1505 Chapter 11. Partial Differential Equations
LHS is a function of x only while RHS is a function
of y only. We conclude that
LHS = RHS = some constant k.
Thus, we obtain two o.d.e.
X 0(x)
1
= k X 0(x) = kf (x)X(x)
f (x) X(x)
k
Y 0(y)
0
= k Y (y) =
Y (y)
g(y)
Y (y)
g(y)
(7)
(8)
Note that (7) is an o.d.e. with independent variable
x and dependent variable X while (8) is an o.d.e.
with independent variable y and dependent variable
Y.
By solving (7) and (8) respectively for X(x) and
Y (y), we obtain the solution u(x, y) = X(x)Y (y).
21
22
11.2.8
MA1505 Chapter 11. Partial Differential Equations
Example
Solve ux + xuy = 0.
Solution: If a solution u(x, y) = X(x)Y (y) exists,
then we obtain
X 0(x)Y (y) + xX(x)Y 0(y) = 0
i.e.,
1 X 0(x)
Y 0(y)
=
x X(x)
Y (y)
(9)
This gives two o.d.e.s :
LHS of (9) = k gives X 0 = kxX.
This o.d.e. has general solution
X(x) = Ae
kx2 /2
Similarly, RHS of (9) = k gives Y 0 = kY .
22
(a)
23
MA1505 Chapter 11. Partial Differential Equations
This o.d.e. has general solution
Y (y) = Beky
(b)
Multiplying (a) and (b), we obtain a general solution
of the p.d.e.
u(x, y) = X(x)Y (y) = Cek(x
23
2 /2 y)