Partial Differential Equations: Definition
Partial Differential Equations: Definition
Mahmoud
Definition:
A partial differential equation (PDE) is any equation involving a function of more than one
independent variable and at least one partial derivative of that function.
Only the simplest physical system can be modeled by ordinary differential equations (ODEs),
whereas many physical problems from fluid mechanics, chemistry, electromagnetism, heat
transfer, rigid body dynamics and elasticity are modelled by PDE's in particular second order.
In general, PDEs are much more difficult to solve analytically than ODEs. They may sometimes
be solved using a Bäcklund transformation, characteristics, Green's function, integral
transform, Lax pair, Laplace transform, separation of variables, or when all else fails (which it
frequently does), numerical methods such as finite differences, finite elements are used.
Independent variables typically include one or more space dimensions and possible time
dimension as well.
Partial derivatives with respect to independent variables are denoted by subscripts, for
example:
• ut=∂u/∂t
• uxy=∂u/∂x∂y
• uxx=∂2u/∂x2
Examples of PDEs:
u 2u
1. c2 2 One-dimensional heat conduction equation
t x
u 2 u 2u
2
2. c 2 2 Two-dimensional heat conduction equation
t x y
2u 2u
3. 0 Two-dimensional Laplace equation
x 2 y 2
2u 2u 2u
4. 0 Three-dimensional Laplace equation
x 2 y 2 z 2
2u 2 u
2
5. a One-dimensional wave equation
t 2 x 2
2u 2 u 2u
2
6. a
x 2 y 2 Two-dimensional wave equation
t 2
1|Page
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
2u 2u
7. f ( x, y ) Two-dimensional Poison equation
x 2 y 2
u 2u
8. cv 2 One-dimensional consolidation equation
t z
1. Order: The order of a PDE is the order of the highest partial derivative in the equation.
Examples:
u 2u u u u 3u
a. 2nd order b. 1st order c. u 3 sin x 3rd order
t x 2 t x t x
2. Degree: the degree of a PDE is the highest exponent of the highest derivative in the
equation.
Examples:
2
u 2u u 2u u
a. st
1 degree b. 2 2nd degree
t x 2 t x y
Examples:
u 2u u 2u 1 u 1 2u
a. 2 variables x and t b. 3 variables t, r and θ
t x 2 t r 2 r r r 2 2
4. Linearity: a PDE is linear if it is of the first degree in the unknown function u (i.e.,
dependent variable) and its partial derivatives. Otherwise it is nonlinear.
Examples:
2u t u
2
2u u
a. e sin t linear b. u 0 non-linear
t 2 x 2 x 2 t
2u 2u u u
c. y 0 linear d. x y u2 0 non-linear
x 2 y 2 x y
2|Page
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
5. Homogeneity: a PDE is called homogenous if after writing the terms in order, the right
hand side is zero (i.e., f(x,y)=0).
Examples:
2u 2u 2u 2u
a. f ( x, y ) non-homogeneous b. u 0 homogeneous
x 2 y 2 t 2 x 2
6. Type of coefficients: If the coefficients in front of each term involving the dependent
variable and its derivatives are independent of the variables (dependent or independent),
then that PDE is one with constant coefficients.
Examples:
2u 2 u
2
2u 2u
1. x 0 variable coefficient b. C 0 constant coefficient
x 2 y 2 x 2 t 2
7. Canonical forms for 2nd order linear PDEs: the standard form for 2nd order linear PDE is
written as follows:
3|Page
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
Examples:
2u u c2 is constant
c 2
x 2 t
Is parabolic (with y t) B2-4AC = 0 – 4(c2) (0) = 0
Note: In this course and for simplicity we will consider only linear and second order PDEs
associated with only two independent variables either:
• Two space variables, denoted by x and y, or
• One space variable denoted by x and one time variable denoted by t
In order to solve PDEs boundary and initials conditions should be given. Time-dependent PDEs
(e.g. wave equation) involve both initial values and boundary values. Whilst steady-state PDEs
involve only boundary values.
4|Page
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
The solution for a second order linear PDE is obtained through the following steps:
1. By the “method of separating variables” or product method, setting u(x,t)= X(x).T(t), then
we obtain from PDE two ODEs, one for X(x) and the other one for T(t).
2. Determine solutions of these ODEs that satisfy the boundary conditions.
3. Finally, using Fourier series, we compose the solutions found in step 2 to obtain a solution
of PDE satisfying both boundary and initial conditions.
The wave equation models vibration in steel plates, and oscillations along elastic strings of
homogeneous materials.
2u 2 u
2
T
a , a2
t 2 x 2
Where T is tension and ρ is density of the string material. The string is placed along x- direction,
stretched to length L, and fixed at the ends x=0 and x=L. The string is then distorted and at some
instant, say time t=0, it is released to allow is to vibrate.
The problem is to determine the vibration of the string, that is, to find deflection u(x,t) at any
point x and at any time t>0.
Solution:
Step 1
By the "method of separating variables" or product method, setting u(x,t) = X(x). T(t), we obtain
from PDE two ODEs, one for X and other for T.
Where:
X = Function of x and,
T= Function of t
5|Page
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
u(x,t) = X.T
Differentiating the above equation with respect to x and t and substituting it in PDE form
u 2 u
2 2
t 2 a x 2 .
u2
2u
X T '' , and X '' T
t 2 x 2
2u 2 u
2
and a
t 2 x 2
X T '' a 2 X '' T Re arrange it ,
T '' X ''
now we separate PDE
a 2T X
Both sides of the above equation must be equal to constant denoted by k
T ''
k T '' a 2 k T 0
a 2T
and ,
X ''
k X '' k X 0
X
Now we have three possible values for k:
Step 2
Satisfying the boundary conditions to determine the solution.
CASE 1 k = λ2
T '' a 2 2 T 0
r2 a2 2 0 r a
T T (t ) C1e at C2 e at
6|Page
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
and ,
X '' 2 X 0
r2 2 0 r
X X ( x ) C3 e x C 4 e x
u( x, t ) X T (C1ect C2e ct )(C3e x C4e x )
u(0,t) = 0
u(L,t) = 0
There are no solutions for CASE 1, try to solve CASE 2 and CASE 3
CASE 2 k = -λ2
T '' a 2 2 T 0
r 2 a 2 2 0 r a i
T T (t ) A cos a t B sin a t
and ,
X '' 2 X 0
r2 2 0 r i
X X ( x ) A1 cos x B1 sin x
u( x, t ) ( A cos a t B sin a t )( A1 cos x B1 sin x )
7|Page
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
u(0,t) = 0
0 ( A cos a t B sin a t )( A1 0)
A1 0
u( x, t ) ( A cos a t B sin a t )( B1 sin x )
( A cos a t B sin a t )(sin x )
u (L,t) = 0
Step 3
u(x,0) = f(x)
n n
f ( x ) A 1 sin x B 0 sin x
n 1 L L
n
f ( x ) A sin x
n 1 L
2L n
A Bn f (x )sin xdx
L0 L
u
g ( x)
t
x ,o
8|Page
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
an n
g ( x) 0 B 1 sin x
n 1 L L
an 2 L n
B. g ( x ).sin xdx
L L0 L
2 L n
B g ( x ).sin xdx
an 0 L
Example
2u 2u
4 2, 0 x ,t 0
t 2 x
u(0, t ) u( , t ) 0, t0
u( x,0) x 2 ( x ), 0 x
u
( x,0) 0, 0 x
t
a2 4 a2 L
f ( x ) x 2 ( x ) g ( x) 0
2L n
A Bn f ( x )sin xdx
L0 L
2
n
A x 2 ( x ).sin xdx
0
2
n n
( x .sin x .sin x )dx
2 3
0
x 2 x 2
2
x3
cos nx 2 sin nx 3 cos nx cos nx
2 n n n n
2
3x 6x 6
sin nx 3 cos nx 4 sin nx 0
n 2 n n
9|Page
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
g (x ) 0 B 0
cn n cn n
u (x , t ) A cos t .sin x B sin t .sin x
n 1 L L L L
2n n 2n n
u (x , t ) A cos t .sin x 0.sin t .sin x
n 1
4
u (x ,t ) (2cos n 1).cos 2nt .sin nx 0.sin 2nt .sin nx
n 1 n3
4
3 (2cos n 1).cos 2nt .sin nx
n 1 n
4
3 (2(1) n 1).cos 2nt .sin nx
n 1 n
or
4 12
u (x ,t ) 3
cos 2 nt .sin nx 3
cos 2nt .sin nx
n 1,3,5,.. n n 2,4,6,.. n
10 | P a g e
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
The one-dimensional heat conduction equation models the flow of temperature in a body in
space and it written as follows.
u 2u
c2 2
t x
u(0, t ) 0 , u( L, t ) 0 for t 0 (boundary conditions )
u( x,0) f ( x ) 0 x L (initial conditions )
L
x=0 x=L
u(0,t) u(L,t)
=0 =0
The constant c2 is called thermal diffusivity and it is equal to c2=K/ρσ, K is thermal conductivity,
σ is specific heat and ρ is the density of material used.
The heat equation is called diffusion equation when used to modelling chemical diffusion
processes of one substance or a gas into in another.
Solution
Step 1
Let
u( x, t ) X T
Where :
X function of x
T function of t
u 2u
X T ' , 2 X '' T Sub. in heat equation
t x
T' X ''
X T ' c 2 X '' T 2 k
c T X
11 | P a g e
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
T' X ''
2
c T
2
X
T'
2 T ' c 2 2T 0
c T
2
r c 2 2 0 , r c 2 2
T C1e c t
2 2
and ,
X ''
2 X '' 2 X 0
X
r2 2 0 , r i
X A cos x B sin x
u( x, t ) ( A cos x B sin x ) C1e c t ( A cos x B sin x ) e c t
2 2 2 2
Step 2
u(0, t ) 0
0 ( A cos 0 B sin 0) e c t
2 2
0 Ae c t A0
2 2
Therefore,
u( x, t ) Be c t sin x
2 2
u( L, t ) 0
0 B e c t sin L
2 2
sin L 0 L n
n
L
2 n 2 c2t
n n c
n
2 2 2
x Bn e L sin
t
u( x, t ) Be L2
sin
2
x
L n 1,2,3,.. L
12 | P a g e
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
Step 3
Substitute initial conditions to find full solution for the PDE.
u( x,0) f ( x )
2n2c2
n
n
B 1 sin
.0
f ( x) Bn e L2
sin x f ( x) n
x
n 1,2,3,.. L n 1,2,3,.. L
2L n
Bn f ( x ) sin xdx
L0 L
Example:
u 2u
, 0 x , t 0,
t x 2
u (0, t ) u ( , t ) 0 , t 0,
f (x ) u (x ,0) x 2 , 0 x ,
Solution:
c2 1 , L , f ( x) x 2
2L n
Bn f ( x ) sin xdx
L0 L
2 2 n 2 2
x sin x x sin nxdx
0 0
0
3
n3 n 3
n
3 ( 2 n 2 2 )cos n 2
2
, n 1,2,3,4,.........
n
Bn 3 ( 2 n 2 2 )( 1) n 2
2
n
13 | P a g e
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
2( n 2 2 4)
for odd n (1,3,5,.....) Bn
n3
2
for even n (2,4,6,.....) Bn
n
2n2c2
n 5 n
n
2 2
u( x, t ) Bn e x Bn e sin
t t
sin
2
L2
x
n 1 L n 1
2 1
Bn e sin nx 3 [( 2 n 2 2 )( 1) n 2] e 5 n t sin nx
5 n t 2 2
n 1 n1 n
or
2
( n 2 2 4) 5 n t
1 5 n t
u ( x, t ) e sin nx
2 2
e sin nx 2
n1,3,5,.. n3 n 2,4,6,.. n
2u 2u
0 (uxx u yy 2u 0)
x 2 y 2
Note that the equation has no dependence on time, just on the spatial variables x and y. This
means that Laplace’s Equation describes steady state situations such as:
The solution aims at computing the distribution u(x,y) over a plane area subjected to a certain
boundary conditions.
14 | P a g e
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
y
u(x, b)=f(x)
u(0, y)=0 b
u(a,y)=0
a x
u(x,0)=0
u(0,y) = 0 for 0 ≤ y ≤ b
u(x,0) = 0 for 0 ≤ x ≤ a
u(a,y) = 0 for 0 ≤ y ≤ b
u(x,b) = f(x) for 0 ≤ x ≤ a
Solution:
Step 1
2u 2u
0
x 2 y 2
let u( x, y ) X ( x ) Y ( y )
X '' Y X Y '' 0
X '' Y '' X '' Y ''
and 0 k
X Y X Y
15 | P a g e
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
X '' Y ''
2
X Y
therefore
X ''
2 X '' 2 X 0 r 2 2 0
X
r i
X ( x ) A cos x B sin x
and ,
Y ( y ) c1e y c2 e y
so,
u( x, y ) ( A cos x B sin x )( c1e y c2e y )
Step 2:
Satisfy boundary conditions
u(0,y) = 0
16 | P a g e
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
Step 3
Substitute other boundary conditions to determine the values of λ and B1
u(a,y) = 0
0 B1 sin a sinh y
we have solution in one case only
sin a 0 a n
n
, n 1,2,3,4,........
a
and
n n
u( x, y ) B1 sin x sinh y
a a
n n
B1 sin x sinh y , Fourier sin e series
n 1 a a
n b 2 a n
B1 sinh f ( x )sin xdx
a a0 a
2 a
n
B1 f ( x )sin xdx
n b 0 a
a sinh
a
17 | P a g e
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
Example
Assume f(x) = 1 , a=4 and b= 4
2 4
n 1 2 n
B1 1. sin xdx cos 4
n 4 sinh n n 4
0
4 sinh 0 4
4
( 1) n 1
2 2
cos n 1
n sinh n n sinh n
4
for odd n B1
n sinh n
for even n B1 0
n n
u( x, y ) B1 sin x sinh y
n 1 a a
n n
( 1) n 1 sin
2
x sinh y
n 1 n sinh n 4 4
4 1 (2n 1) (2n 1)
sin x sinh y
n1 (2n 1)sinh(2n 1) 4 4
18 | P a g e
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
This equation describes the settlement of a saturated clay layer and dissipation of the excess pore
water pressure with time.
u 2u
cv 2
t z
Where u is the excess pore water pressure in the soil, t is the time of load applied, z is the depth
of clay layer, and cv is the coefficient of consolidation.
Example
In an open clay layer of 2d thickness (i.e., z=2d), find the final distribution of excess pore water
pressure u(z, t) given:
Solution
Step 1
u 2u
cv 2
t z
let u( z, t ) Z ( z ) T (t )
Z T ' cv Z '' T
1 T ' Z ''
and k
cv T Z
19 | P a g e
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
1 T ' Z ''
2
cv T Z
therefore
Z ''
2 Z '' 2 Z 0 r 2 2 0
Z
r i
Z ( z ) A cos z B sin z
and ,
1 T'
2
cv T
T (t ) c1e c t
2
v
so,
u( z, t ) ( A cos z B sin z )e c t
2
v
Step 2:
Satisfy boundary conditions
u(0,t) = 0
0 A cos0 A0
u( z, t ) B sin z e c t
2
v
u( z, t ) B sin z e c t
2
v
n
2 d n
2d
n
n
2
c t
u( z, t ) B sin z e 2d n 1,2,3,...........
v
2d
n
n
2
c t
u( z, t ) bn sin z e 2d
v
n 1 2d
20 | P a g e
Engineering Analysis\Partial Differential Equations Dr. Herman A. Mahmoud
Step 3
Substitute initial conditions to determine full solution.
u(z,0) = u0
n
u( z,0) bn sin z 1 u0
n 1 2d
2 2d n
bn u0 sin zdz
2d 0 2d
u 2 d n 2d
bn 0 cos z
d n 2d 0
2u0 2u
bn cos n cos0 0 ( 1) n 1
n n
4u
for n 1,3,5,7............bn 0
n
for n 2,4,6,8............bn 0
21 | P a g e