Conduction Heat Transfer

Solved Sample Problems

The temperature T is maintained at 0 at three edges of a square plate and at 100 at the
fourth edge. The plate is 100 cm in side. For steady state conditions
a) Find T (x, y) in the plate.
b) Calculate T (50, 50).

Answer:
a) Assume Tp = X(x)Y (y). Substituting into Laplaces equation yields

1 d2 X 1 d2 Y
2
= 2
= k2
X dx Y dy
with general solutions

X(x) = a sin kx + b cos kx

Y (y) = eky + eky

Using the three homogeneous boundary conditions reduces the above to

n
Xn (x) = an sin( x)
100
and
n
Yn (y) = n (eky eky ) = n sinh( y)
100
Now, substituting Tp = X(x)Y (y) and superimposing for all n yields

X n n
T (x, y) = An sin( x) sinh( y)
n=1 100 100

1
 Finally, introducing the initial condition T (x, 100) = 100 gives

X X X
n n n
100 = An sin( x) sinh(n) = [An sinh(n)] sin( x) = cn sin( x)
n=1 100 n=1 100 n=1 100

which is the Fourier sine series representation of 100 with the Fourier coefficients
Z 100
2 n 100
cn = An sinh(n) = 100 sin( x)dx = 2 [1 cos(n)]
100 0 100 n
I.e.
(
0 n even
An = 400
n sinh n
n odd

With the above the desired solution can now be rearranged to

X n
n sinh( 100 y)
T (x, y) = cn sin( x)
n=1 100 sinh(n)

with
(
0 n even
cn = 400
n
n odd

b) To find T (50, 50) use the first four terms in the above series to get
3
400 sinh( 100 50) 400 3 sinh( 100 50)
T (x, y) = sin( 50) +0+ sin( 50) +0=
100 sinh() 3 100 sinh(3)
= 25.37 0.3812 = 24.98

2
The temperatures at the ends x = 0 and x = L = 100 of a 100 cm long rod with insulated
sides are held at temperatures of 0 and 100 , respectively until reaching steady state. Then,
the temperatures at the ends are interchanged. Find T (x, t).

Answer: The solution to the problem of T = T1 at x = 0 and T = T2 at x = 100, is

xX n n2 2
T (x, t) = T1 + (T2 T1 ) an sin( x)e l2 t
l n=1 l

2
where
Z L
2 n 2
an = f (x) sin( x)dx + (T2 cos n T1 )
L 0 L n
This can be applied directly noticing that here f (x) = 100x/L is the initial condition (ob-
tained from the previous steady state) and that T1 = 100 and T2 = 0. The Fourier coefficients
are then

(
1 1002 200 200 400/n even
an = [ (1)n ] = [(1)n+1 1] =
50 n n n 0 odd

The result is
X 400 n n2 2
T (x, t) = 100 x sin( x)e 1002 t =
n1,even n 100
X 400 2m (2m)2 2
= 100 x x)e 1002 t =
sin(
m=1 2m 100

200 X 1 m m2 2
= 100 x sin( x)e 2500 t
m=1 m 50

3
A large, 0.1 meter thick steel slab emerges from a rolling mill with a uniform temperature
of 1000 degrees Celsius and is left to cool by convection into an environment at zero degrees
Celsius. For this steel k = 50W/mK, = 7, 900kg/m3 and Cp = 470J/kgK. Assume the
heat transfer coefficient is h = 250W/m2 K.
Evaluate the first three terms in the series solution of this problem obtained by separation
of variables when the temperature at the center of the slab is 500 degrees Celsius.

Answer:
Assume heat is convected aways from both sides of the slab equally. This allows us do
the analysis on just one half of the slab (i.e. L = 0.05). The mathematical formulation of
the problem is to find T (x, t) such that

2 T (x, t) 1 T (x, t)
2
=
x t
where = k/Cp = 50/(7, 900 470) = 1.34 105 .

3
 The boundary conditions consist of the symmetry condition at x = 0
T T
k = =0
x x
and the convective loss into an environment at T = 0 at x = L = 0.05,
T T
k + hT = 50 + 250T = 0
x x
The initial condition is

T (x, 0) = 1000

for all x when t = 0.

Assume the solution is of the form T (x, t) = X(x)(t) and substitute in the heat equation
to obtain
1 d2 X 1 d
2
=
X dx dt
The left hand side is only a function of x. The right hand side is only a function of t. Since
x and t are independent of each other the equality is only possible if both terms are equal
to a constant.
Hot objects exposed to cold environments cool down and cold objects exposed to warm
environments heat up. This requires the constant above to be a negative real number. To
make sure of this we call it 2 .
Therefore, we obtain the following two equations, first
1 d
= 2
dt
for which a general solution is readily obtained by integration as (check!)
2 t
(t) = Ce

where C is a constant of integration and

1 d2 X
2
= 2
X dx
for which a general solution is obtained by inspection as (check!)

X(x) = A cos(x) + B sin(x)

However, at x = 0 we require (check!)

T (X) dX dX
= = = = 0 = A sin(x) + B cos(x)
x x dx dx

4
so that, necessarily B = 0 and X(x) = A cos(x) (check!).
Furthermore, at x = L, (check!)

dX h h
+ X = 0 = sin(L) + cos(L)
dx k k
For this to be true, the values of must satisfy the equation (check!)

hL
(n L) tan(n L) =
k
This equation as infinitely many roots n for n = 1, 2, .... The roots are the eigenvalues of
this problem. The roots can readily be obtained either graphically or numerically. In our
problem, hL/k = 250(0.05)/50 = 0.25 and the above equation becomes

zn tan(zn ) = 0.25

with zn = 0.05 n . The first four roots are then (check!) z1 0.48, z2 3.22 and
z3 6.325 so that 1 z1 /L = 0.48/0.05 = 9.6, 2 64.4 and 3 126.5. The associated
eigenfunctions of the problem are then

Xn (x) = X(n , x) = An cos(n x)

The particular solution obtained by separation of variables and satisfying the heat equa-
tion as well as the boundary conditions is then
2
Tn (x, t) = X(n , x)(n , t) = cn cos(x)en t

where cn = An C. A more general solution is obtained by linear combination of the above

solutions, i.e.

X
X 2
T (x, t) = Tn (x, t) = cn cos(n x)en t
n=1 n=1

All is left to do is introduce the initial condition, i.e. at t = 0,

X
Ti = 1000 = cn cos(n x)
n=1

But this is just the Fourier cosine series representation of the number 1000. Now multiply
both sides of the above by cos(m x) and integrate from 0 to L
Z L Z L
Ti cos(m x)dx = cn cos(n x) cos(m x)dx
0 0

5
But since the cos functions constitute an orthogonal set, the right hand side becomes
Z
L X Z
X L
cn cos(n x) cos(m x)dx = cn cos(n x) cos(m x)dx =
0 n=1 n=1 0
Z L 1
= cn cos(n x)2 dx = (n L + sin(n L) cos(n L))
0 2n

Therefore (check!)

2Ti sin(n L)
cn =
n L + sin(n L) cos(n L)

Finally, the desired solution is then

X sin(n L) cos(n x) 2
T (x, t) = 2Ti en t
n=1 n L + sin(n L) cos(n L)

Using just the first term of the series, introducing numerical values and focusing on the
temperature at x = 0 yields (check!)

sin(0.48) 3
T (x, t) 2000 exp(1.34 105 (9.6)2 t) = 1038e1.2310 t
0.48 + sin(0.48) cos(0.48)

The following table gives numerical values obtained from the above equation for a few values
of time (check!)

time (s) Temperature (C)

600 496
1200 237
1800 113

4
Find the function T (x, t) in 0 x < satisfying

2T 1 T
=
x2 t

6
and subject to
T (0, t) = f (t)

T (x , t) = 0
and
T (x, 0) = 0
using the Laplace Transform method.

Answer:
Taking Laplace transforms one gets
d2 T (x, s) s
T (x, s) = 0
dx2
T (0, s) = f(s)

T (x , s) = 0
The solution of this problem is

T (x, s) = f(s)ex s/ = f(s)g(x, s) = L[f (t) g(x, t)]
Inversion then produces
Z t
T (x, t) = f (t) g(x, t) = f ( )g(x, t )d
0

Inversion of g(x, s) to get g(x, t) finally gives

Z
x t f ( ) x2
T (x, t) = exp[ ]d
4 =0 (t )3/2 4(t )
If f (t) = T0 = constant, the solution is
T0 q
T (x, s) = exp(x s/)
s
and inverting the transform yields
x
T (x, t) = T0 erf c( )
2 t
where the complementary error function erf c(z) is given by
2 Z
erf c(z) = exp(z 02 )dz 0
z

7
5
Find a bounded solution u(r, t) of the following problem using the Laplace transform method.
2 u 1 u u
2
+ =
r r r t
subject to
u(r, 0) = 0
and
u(a, t) = u0
where t > 0, 0 < r < a with a = constant.

Answer:
Taking Laplace transforms of the terms in the given PDE yields
d2 u(r, s) 1 du(r, s)
+ = su(r, s)
dr2 r dr
This is a Bessel-type equation with general solution given by

u(r, s) = AJ0 (ir s) + BY0 (ir s)
Since we look for a bounded solution, necessarily B = 0. Now transformation of the boundary
condition at r = a yields
u0
u(a, s) =
s
and combining with the above gives
u0 1
A=
s J0 (ia s)
so that the solution for u(r, s) becomes

J0 (ir s)
u(r, s) = u0
sJ0 (ia s)
Finally, the inverse transformation gives
2 2
X en t/a J0 (n r/a)
u(r, t) = u0 [1 2 ]
n=1 n J1 (n )
where n are the roots of
J0 (n ) = 0