Sie sind auf Seite 1von 15

Method of Separation of

Variables
Method of Separation of Variables
UCF Example 1 (1)
For three dimensional potential problem defined in 0 x a, 0 y b
0 z h, V 0 for x=0, a; y=0, b; z=0 planes. For z=h plane, V ( x, y)
(given). Find potential distribution V(x,y,z).
z
Solution:
h
The Laplaces equation 2V 0
in Cartesian coordinate can be written as: b
2V 2V 2V a y
0 (1)
x 2
y 2
z 2

together with boundary conditions: x


V (0, y , z ) 0
V (a , y, z ) 0

V ( x ,0 , z ) 0 (2)

V (x,b, z) 0
V ( x , y ,0 ) 0

V ( x , y , h ) ( x , y ) (given)
UCF
Example 1 (2)
For the method of separation of variables, we can let
V ( x, y, z ) X ( x)Y ( y) Z ( z ) (3)
When this is inserted into Equation (1), there results the equation
d 2X d 2Y d 2Z
YZ X Z XY 0
dx 2 dy 2 dz 2
If the above is divided by X ( x )Y ( y ) Z ( z ),
we can obtain:
1 d 2 X 1 d 2Y 1 d 2 Z
2
2
2
0 (4)
X dx Y dy Z dz
Let each of the above three terms be separated constant:
1 d 2X
2

X dx 2
1 d 2Y
2
2 (5) where 2 2 2 (6)
Y dy
1 d 2Z
2
2
Z dz
UCF
Example 1 (3)
The general solutions for the above three equations in (5) are
X ( x) A sin(x) B cos(x)

Y ( y ) C sin(y ) D cos(y ) (7)
Z ( z ) G sinh(z ) H cosh(z )

Note: There are different ways to write down the general solution. Equation
1 d 2X
2

X dx 2
has other mathematical forms of general solutions. For example,
X ( x) Ae jx Be jx

or:
X (x) A sin[(x a)] B cos[(x a)]
But the form in (6) gives the simplest form of solution.
UCF
Example 1 (4)

Substituting (3) into the first five boundary conditions in (2) yields
X (0) 0

Y ( 0 ) 0 (8)
Z (0) 0

and
X (a ) 0
(9)
Y (b ) 0
When (8) is inserted into (7), we find
B 0

D 0 (10)
H 0

Thus
X ( x) A sin(x)
(11)
Y ( y ) C sin(y)
Z ( z ) G sinh(z )

UCF
Example 1 (5)

Inserting (11) into (9) leads to


sin( a ) 0
(12)
sin( b ) 0
which gives
m
, m 1, 2,
a
(13)
n , n 1, 2,
b

From (3), (11) and (13), we can have



m n
V ( x, y, z) Fmn sin( x) sin( y) sinh( mnz) (14)
m1 n1 a b
where from (6)
m n
2 2

mn
a b
UCF
Example 1 (6)
From (14) and the last equation in boundary conditions (2), we have

m n

m1 n1
Fmn sin(
a
x) sin(
b
y) sinh( mnh) ( x, y) (15)

m n
Multiplying (15) by sin x sin y , integrating for x from 0 to a
a b
and for y from 0 to b, and using the following orthogonality conditions:
a m m 0 m m
0 sin( a x) sin( a x)dx a m m (16)
2
n n 0 n n
b (17)
0 sin( b y) sin( b y)dy b n n
we have 2
4 1 b a m n
Fmn
ab sinh( mn h ) 0 0 ( x, y ) sin(
a
x ) sin(
b
y ) dxdy
4 1 b a m n
or Fmn
ab sinh( mn h ) 0 0
( x , y ) sin(
a
x ) sin(
b
y ) dxdy (18)
UCF
Example 2

For the above example, if we assume the boundary condition at z=0 is V ( x, y)


and at z=h is V 0, the general solution for Z(z) should better be defined as:

Z ( z ) G sinh( ( z h )) H cosh( ( z h ))

By this way, from Z (h) 0, we simplify have


Z ( z ) G sinh( ( z h ))
z
h

b
a y
x
Details are similar to Example 1.
UCF
Example 3
2D problem: y 2V 2V
b V2 0
x 2 y 2

V3 V4
V1 = (Superposition)
a x
y y
b 0 b V2

0 0 + 0 0
+
V1 a x 0 a x

y y
b 0 b 0

V3 0 + 0 V4
0 a x 0 a x
Method of Separation of Variables
UCF
in Cylindrical (Polar) Coordinate Example 4 (1)

(2.69)

From Matthew Sadiku, Numerical Techniques


in Electromagnetics, 2nd Edition, pp. 40-42.
UCF
Example 4 (2)
Example 4 (3)
UCF
UCF
Example 4 (4)
UCF
Example 4 (5)

n
UCF
Example 4 (6)

Das könnte Ihnen auch gefallen