0% found this document useful (0 votes)
68 views11 pages

Workedexamplesc3 PDF

The document presents worked examples on various mathematical problems related to steady-state temperature distribution in a cylinder, diffusion of a solute past a solid sphere, heat source near an insulated wall, electrostatics using Poisson's equation, and the solution of Laplace's equation in a 3D half-space. Each example includes the governing equations, boundary conditions, and the final solutions derived through appropriate mathematical methods. The document emphasizes the application of Fourier series, Legendre polynomials, and integral solutions in solving these physical problems.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
68 views11 pages

Workedexamplesc3 PDF

The document presents worked examples on various mathematical problems related to steady-state temperature distribution in a cylinder, diffusion of a solute past a solid sphere, heat source near an insulated wall, electrostatics using Poisson's equation, and the solution of Laplace's equation in a 3D half-space. Each example includes the governing equations, boundary conditions, and the final solutions derived through appropriate mathematical methods. The document emphasizes the application of Fourier series, Legendre polynomials, and integral solutions in solving these physical problems.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Worked Example 7: Steady-State Temperature Distribution

in a Cylinder
T=+T0

T=T0

An infinitely long cylinder of radius a is heated on its


boundary as shown. The steady-state temperature T (r, )
(note no dependence on z) satisfies
2 T = 0
subject to

+T
0
T (a, ) =
T0

in r < a

0 < ,
< 2.

The general solution for plane polar coordinates applies; we


choose to use it in its second form as given in the lecture
notes. We require that the temperature be finite at r = 0 for
a physically realistic solution: so C0 = 0, and also, for all
negative n, An = Bn = 0 (since they are the coefficients of

rn cos
n).
sin

Finally, T must be periodic in (i.e., not multi-valued), so


B0 = 0. Hence
T (r, ) = A0 +

rn (An cos n + Bn sin n).

n=1

On r = a this gives
T (a, ) = A0 +

(An an cos n + Bn an sin n).

n=1

This is a standard Fourier series, so we may calculate the


Fourier coefficients using the standard formulae:
Z 2
1
T (a, ) d = 0
A0 =
(by anti-symmetry of T (a, ))
2 0
Z 2
1
T (a, ) cos n d = 0
An an =
0
Z 2
1
Bn an =
T (a, ) sin n d
0
Z
Z
1 2
1
T0 sin n d
T0 sin n d
=
0

4T /n n odd,
0
=
0
n even.

Hence the final solution for all r and is


4T0 X rn
T =
sin n.
n odd nan

Worked Example 8: Diffusion of a Solute past a Solid Sphere


F
a
axis of
symmetry

Consider fluid at rest surrounding a fixed solid sphere of


radius a at the origin. The fluid contains a solute which
diffuses through the fluid, and we are interested in the
steady state. At large distances from the sphere (where the
sphere has negligible effect) we assume that there is a
constant flux of solute parallel to the z-axis of magnitude F
(possibly due, for example, to an externally imposed
concentration gradient).
The flux is k where is the concentration. There can
be no flux across r = a, so
er . = 0 on r = a, or
equivalently
(a, ) = 0 for all .
r
Far from the sphere, we must have Fk
ez , i.e.,
Fk z; so we require that as r ,
Fk r cos = Fk rP1 (cos ).
We use the general axisymmetric solution, and must choose
the arbitrary constants to ensure the correct behaviour as
r .

This can only occur if A1 = Fk and An = 0 for all n 2.


Thus

X
F
Bn rn1 Pn (cos ).
= A0 rP1 (cos ) +
k
n=0
On r = a, we must have

X

F
(n + 1)Bn an2 Pn (cos ) = 0
= P1 (cos )
r r=a
k
n=0

for all . Using the orthogonality of Legendre polynomials


(multiply by Pm (cos ), substitute = cos , and integrate
from = 1 to 1), or by inspection, we find that B0 = 0,
B1 = F a3 /2k and Bn = 0 for all n 2. So the solution is


a3
F
r + 2 cos ,
= A0
k
2r

and A0 remains an arbitrary constant (it measures, in some


sense, the average of the concentrations far up and
downstream).
Note that the boundary conditions involved only P1 (cos )
and no other Pn ; and so does the solution. This is usual:
boundary conditions can often be expressed in terms of just
a few Pn , and only those terms need be retained from the
general solution. For this purpose it is useful to know the
following:
1 = P0 (cos )
cos = P1 (cos )
cos2 =

1
2
P2 (cos ) + P0 (cos )
3
3

Worked Example 9: Heat Source near an Insulated Wall

flux from x0

flux from x1
+Q

11
00
00
11

x1

11
00
00
11

x0

Hold a heat source of strength Q at x0 = (x0 , y0 , z0 ) near an


insulated plane wall, i.e., one through which no heat can
pass, at x = 0. We must then have no component of heat
flux through the wall; i.e., n . (kT ) = 0 on the wall.
Therefore we must solve
Q
2 T = (x x0 )
k
subject to

in x > 0

T
=0
on x = 0.
n
This is a problem with Neumann (rather than Dirichlet)
boundary conditions.

We use the method of images. Introduce an image source of


strength +Q at x1 = (x0, y0 , z0 ). (Note that for Dirichlet
boundary conditions we would have used Q for the
strength of the image.) Because T is radial from each
source, the total T (from the two sources combined) must
have zero component perpendicular to the wall. Hence we
have T /n = 0 as required. Therefore (by uniqueness) the
solution is


1
1
Q
+
.
T =
4k |x x0 | |x x1 |

Worked Example 10: Electrostatics Using the Integral Solution of Poissons Equation
Consider a wire of length 2L carrying a charge density per
unit length, lying along the z-axis from z = L to +L.
What is the electric potential ?
The charge distribution is (x) = (x)(y) for L z L
(and zero for |z| > L). We shall use the integral solution of
Poissons equation in the whole of space to obtain the
potential at a point (x0 , y0 , z0 ). We need Greens function,
which is simply the fundamental solution here.
(x0 )

(x)
40 |xx0 |

RRR
R3

=
=
=

RL

dV

L 40 |(0,0,z)x0 |

40

40

RL

dz

dz

x20 +y02 +(zz0 )2

0
sinh1 zz
2

x0 +y02

L




1 Lz0
1 L+z0
.
sinh
+
sinh
= 4
2
2
2
2
0
x0 +y0

x0 +y0

This is true for arbitrary locations x0 , so replacing x0 by x


we obtain


L
+
z

z
.
+ sinh1 p
(x, y, z) =
sinh1 p
2
2
2
2
40
x +y
x +y

In particular, the potential at a point in the (x, y)-plane is


given by
p


1
2
2
(x, y, 0) =
sinh L/ x + y .
20
Note, for completeness, that for very large L, i.e., in the
limit as L , it is possible to check (using sinh1 x ln x
as x ) that
p

ln x2 + y 2 + constant,

20
which verifies an earlier result we obtained for the
two-dimensional field around an infinitely long wire.

Worked Example 11: Solution of Laplaces Equation


in a 3D Half-Space
We wish to solve 2 = 0 in the half-space x > 0 of R3 , with
= f (y, z) on the boundary x = 0.

S
1
0
0
1
0
1

V
1
0
0
1
0
1

x0

x1

x=0

We use the integral solution of Poissons equation (with


0) in the half-space, with S being the plane x = 0
(strictly speaking, together with the hemisphere at ):
ZZZ
ZZ
G
dS
(x0 ) =
(x)G(x; x0 ) dV +
f (x)
n
V
S
Z Z

f (y, z) G(x; x0 ) dy dz
=
x

(because

= x
on S).

To calculate this we need to evaluate







G

1
1

=

+

x x=0
x
4|x x0 | 4|x x1 | x=0

1
1
=
p
4 x
(x x0 )2 + (y y0 )2 + (z z0 )2


1

+p
(x + x0 )2 + (y y0 )2 + (z z0 )2 x=0

1
x x0
=
4 {(x x0 )2 + (y y0 )2 + (z z0 )2 }3/2


x + x0


2
2
2
3/2
{(x + x0 ) + (y y0 ) + (z z0 ) }
x=0
x0
.
2{x20 + (y y0 )2 + (z z0 )2 }3/2

Therefore
x0
(x0 ) =
2

f (y, z)
dy dz
{x20 + (y y0 )2 + (z z0 )2 }3/2

or alternatively (swapping x and x0 ),


Z Z
x
f (y0 , z0 )
(x, y, z) =
dy0 dz0 .
2 {x2 + (y y0 )2 + (z z0 )2 }3/2

This is the solution for:


(i) Steady-state temperature distribution with a wall heated
to a specified temperature distribution;
(ii) Steady-state concentration of solute with a wall kept at
given concentration;
(iii) Electrostatic potential with a conducting wall held at
given potential.

You might also like