Principles of Heat and Mass Transfer

F. Incropera, D. Dewitt, T. Bergman, A.

Chapter 5: Lavine, 7th Edition

TRANSIENT CONDUCTION

[ME421] Heat Transfer Dr. Alia H. Marafie

5. Transient Conduction

 Recall from Chapter 4 that:

 The heat conduction equation under two-dimensional, steady-
state conditions may be solved using:
 Analytical methods: which may be obtained for idealized
conditions.
 Numerical methods: to accurately predict temperatures and
heat rates within the medium and its boundaries for more
complex problems and problems involving internal heat
generation.

[ME421] Heat Transfer 2 Dr. Alia H. Marafie

1
 5. Transient Conduction

 In our treatment of conduction we have gradually

considered more complicated conditions:
 Chapter 2: One-dimensional, steady-state conduction
w/out heat generation.
 Chapter 3: One-dimensional, steady-state conduction
w/ and w/out heat generation effects.
 Chapter 4: Two-dimensional, steady-state conduction
w/ and w/out heat generation effects.

 However, we have not yet considered situation of which

conditions change with time.

[ME421] Heat Transfer 3 Dr. Alia H. Marafie

5. Transient Conduction

 In this chapter we will learn that

 Many heat transfer problems are time dependent. Such that
unsteady, or transient, problems typically arise when the
operating conditions of a system are changed.

 2T  2T  2T q 1 T Transient problems,
    transient term  0
x 2 y 2 z 2 k  t
 Changes in operating conditions in a system cause temperature
variation with time, as well as location within a solid, until a new
steady-state (thermal equilibrium) is obtained.

[ME421] Heat Transfer 4 Dr. Alia H. Marafie

2
 5. Transient Conduction

 Objectives
 1st objective: to develop procedures for determining the time
dependence of the temperature distribution within a solid
during a transient process for different geometries, as well as
for determining heat transfer between the solid and its
surroundings assuming:
 Negligible temperature gradients within the solid  Lumped analysis.
 Temperature gradients is NOT negligible within the solid (for plane walls,
long cylinders, spheres and semi-infinite wall)  (a) Exact solution, (b)
approximate solution, and (c) Heisler charts (graphical
representation)
 2nd objective: to use numerical methods to predict the time
dependence of temperatures within the solid, as well as heat
rates at its boundaries for more complex conditions.

[ME421] Heat Transfer 5 Dr. Alia H. Marafie

5. Transient Conduction

 Transient conduction can be induced by changes in:

 surface convection conditions (h, T∞),
 surface radiation conditions (hr, Tsur),
 a surface temperature or heat flux, and/or
 internal energy generation.

 Solution Techniques:
 The Lumped Capacitance Method (approximate)
 Exact Solutions
 Graphical Representation- Heisler Charts (approximate).
 The Finite-Difference Method (approximate).

[ME421] Heat Transfer 6 Dr. Alia H. Marafie

3
 5. Transient Conduction
5.1 The Lumped Capacitance Method

 5.1 The Lumped Capacitance

Method
5.2 Validity of the Lumped Capacitance Method
5.3 General Lumped Capacitance Analysis
5.4 Spatial Effects
5.5 The Plane Wall with Convection
5.6 Radial Systems with Convection
5.7 The Semi-Infinite Solid
5.10 Finite-Difference Methods Numerical

[ME421] Heat Transfer 7 Dr. Alia H. Marafie

5. Transient Conduction
5.1 The Lumped Capacitance Method

 Transient conduction problem is one for which a solid

experience a sudden change in its thermal environment.
 The lumped capacitance method is based on the assumption
of a spatially uniform temperature distribution throughout the
transient process  the temperature gradients within the
solid are negligible.

Figure 5.1 Cooling of a hot metal forging.

[ME421] Heat Transfer 8 Dr. Alia H. Marafie

4
 5. Transient Conduction
5.1 The Lumped Capacitance Method
 As an example, consider a hot metal that is initially at a uniform
temperature, Ti, and at t=0 is quenched by immersion in a cool
liquid, of lower temperature T∞.
 The temperature of the solid will decrease for time t > 0, due to
convection heat transfer at the solid-liquid interface, until it
reaches T∞.

(a) (b) y

T ( x,0)  Ti

x
Figure 5.1 (a) Cooling of a hot metal forging. (b) Metal initial temperature.

[ME421] Heat Transfer 9 Dr. Alia H. Marafie

5. Transient Conduction
5.1 The Lumped Capacitance Method
 If the thermal conductivity of the solid is very high, resistance
to conduction within the solid will be small compared to
resistance to heat transfer between solid and surroundings.
 Temperature gradients within the solid will be negligible, i.e. the
temperature of the solid is spatially uniform at any
instant.
 The transient temperature response is determined by formulating
an overall energy balance on the solid:

dT
 E out  E st (5.1) hAs (T  T )   Vc (5.2)
dt
The rate of heat loss The rate of change of
at the surface the internal energy

[ME421] Heat Transfer 10 Dr. Alia H. Marafie

5
 5. Transient Conduction
5.1 The Lumped Capacitance Method
 Defining a temperature difference for the temperature ( ) and for
the initial temperature (i ), as
  T  T (5.3) i  Ti  T (5.4)
d hAs
   0 (5.e1)


(eq. 5.3 into eq. 5.2)

dt  Vc
 Solving the resulting first order  Vc i
ODE, we find the time (t) takes t ln (5.5)
the solid to reach a temperature, T, hAs 

The temperature, T,   hA  

 T  T
reached by the solid   exp    s  t  (5.6)
at some time, t, can be i Ti  T    Vc  
calculated using,
[ME421] Heat Transfer 11 Dr. Alia H. Marafie

5. Transient Conduction
5.1 The Lumped Capacitance Method

 Vc i
t ln (5.5) This equation
hAs  indicate that the
difference between the
solid and fluid temperature,
 , must decay exponentially
to zero at t approaches
infinity.

Figure 5.2 Transient temperature response

of lumped capacitance solids for different
thermal time constants  t

[ME421] Heat Transfer 12 Dr. Alia H. Marafie

6
 5. Transient Conduction
5.1 The Lumped Capacitance Method
 Defining the thermal time constant (t ) as
Rt: Resistance to convection
1
i    Vc   Rt Ct (5.7)
heat transfer (1/hAs).
Ct: Lumped thermal
hAs
capacitance of the solid (Vc).
 Any increase in Rt or Ct will cause a solid
to respond more slowly to changes in its
thermal environment and will increase the
time required to reach thermal equilibrium
( =0).
 The total energy transfer, Q, occurring
up to some time t is:
  t  
Q   q dt  hAS   dt    Vc i 1  exp   
t t

0 0
   t 
Figure 5.2 Transient temperature
(5.8a) response of lumped capacitance
 Such that Q  Est (5.8b) solids for different thermal time
constants  t

[ME421] Heat Transfer 13 Dr. Alia H. Marafie

5. Transient Conduction
5.2 Validity of the Lumped Capacitance Method

 5.1 The Lumped Capacitance Method

 5.2 Validity of the Lumped

Capacitance Method
5.3 General Lumped Capacitance Analysis
5.4 Spatial Effects
5.5 The Plane Wall with Convection
5.6 Radial Systems with Convection
5.7 The Semi-Infinite Solid
5.10 Finite-Difference Methods

[ME421] Heat Transfer 14 Dr. Alia H. Marafie

7
 5. Transient Conduction
5.2 Validity of the Lumped Capacitance Method

 The lumped capacitance method is the simplest and most

convenient method that can be used to solve transient heating
and cooling problems.
 Under what conditions the lumped capacitance method
could be used with reasonable accuracy?

T < Ts,2 < Ts,1

Under steady-state conductions, the
surface energy balance reduces to
kA
L
Ts,1  Ts,2   hATs,2  T 
Figure 5.3 Effect of Biot number on steady-state
temperature distribution in a plane wall with
surface convection.

[ME421] Heat Transfer 15 Dr. Alia H. Marafie

5. Transient Conduction
5.2 Validity of the Lumped Capacitance Method

 Under what conditions the lumped capacitance method

could be used with reasonable accuracy?
 We need a suitable criterion to determine validity of the
method. This may be done by considering steady-state
condition and then relating the relative magnitudes of
temperature drop in the solid (due to conduction) to the
temperature difference between surface and fluid (due to
convection)  Biot number (Bi).

T solid (due to conduction) T s ,1 T s ,2 (L / kA ) Rcond hL

     Bi
T solid / liquid (due to convection) T s ,2 T  (1 / hA ) Rconv k

(5.9)

[ME421] Heat Transfer 16 Dr. Alia H. Marafie

8
 5. Transient Conduction
5.2 Validity of the Lumped Capacitance Method

Tsolid( due to conduction) ( L / kA) Rcond hL

    Bi (5.9)
Tsolid / liquid( due to convection) (1 / hA) Rconv k
 The Biot number, Bi, is a dimensionless parameter that
plays a fundamental role in conduction problems that involve
surface convection effects and very significant to transient
conduction problems.
 It is the ratio of internal (conduction) thermal resistance
(Rcond) of a solid body to its surface (convection) thermal
resistance (Rconv).
Therefore, a small Biot number represents small
resistance to heat conduction, and thus small
temperature gradients within the body.
Lumped Capacitance Assumption

[ME421] Heat Transfer 17 Dr. Alia H. Marafie

5. Transient Conduction
5.2 Validity of the Lumped Capacitance Method

 For what value of Biot number the assumption of lumped

capacitance would be valid?
 We will consider three different cases:
1st Case: Bi << 1 2nd Case: Bi = 1 3rd Case: Bi >> 1

Figure 5.4 Transient temperature distributions for different Biot numbers in a

plane wall symmetrically cooled by convection.

[ME421] Heat Transfer 18 Dr. Alia H. Marafie

9
 5. Transient Conduction
5.2 Validity of the Lumped Capacitance Method

 1st Case: Bi << 1 ( Bi  hL k  Rcond / Rconv )

 The resistance to conduction within the
solid is much less than the resistance to
convection across the fluid boundary layer.
 The temperature gradient in the solid is
small and T(x,t) ≈ T(t).
 All the temperature difference is between
the solid and the fluid.
 For this condition, it is reasonable to
assume a uniform temperature
distribution within a solid at any time Figure 5.4a Transient
during a transient process  Lumped temperature distributions
for Biot number <<1.
capacitance assumption is VALID for Bi
<< 1.

[ME421] Heat Transfer 19 Dr. Alia H. Marafie

5. Transient Conduction
5.2 Validity of the Lumped Capacitance Method

 2nd and 3rd Cases: Bi = 1, Bi

>>1
 Temperature gradients within the
solid are significant.
 The temperature varies with
location and time, T ≈ T(x,t).
 The temperate difference across
the solid is much larger than that
between the surface and the fluid.
 For these conditions, the
assumption of a uniform Figure 5.4b,c Transient temperature
temperature distribution within a distributions for Biot number =1 and
solid at any time is not valid  Biot number <<1.
Lumped capacitance assumption is
NOT valid.

[ME421] Heat Transfer 20 Dr. Alia H. Marafie

10
 5. Transient Conduction
5.2 Validity of the Lumped Capacitance Method

 For transient heating or cooling problems, the lumped

capacitance method is valid when
hLc Lc = ½ L for plane wall
Bi   0.1 (5.10) Lc = r0/2 for long cylinder
k Lc =r0/3 for sphere
Lc: Characteristic length is defined as the the ratio of the solid’s volume to
surface area, Lc ≈V/As. (case in Figure 5.4)
 Define a “dimensionless time”, as the Fourier number which
is relates the ratio of the rate of heat conduction to the rate of
internal energy storage in a solid.  Very important
: thermal diffusivity
 Fourier number: Fo   t L2c (5.12)  = k / c
 Hence,  T  T
Eqs. 5.10-11 into Eq. 5.6    exp  Bi  Fo  (5.13)
i Ti  T  Valid only for  Bi  0.1

[ME421] Heat Transfer 21 Dr. Alia H. Marafie

5. Transient Conduction
5.2 Validity of the Lumped Capacitance Method

 When the lumped capacitance analysis is NOT valid, we must

solve the partial differential equations analytically or numerically.
 Exact and approximate solutions may be used.
 Tabulated values of coefficients used in the solutions of these
equations are available.
 Transient temperature distributions for commonly encountered
problems involving semi-infinite solids can be found in the
literature.

Resolve Example 5.1 in the textbook

[ME421] Heat Transfer 22 Dr. Alia H. Marafie

11
 5. Transient Conduction
5.2 Validity of the Lumped Capacitance Method

 Problem 5-16: A plane wall of a furnace is

fabricated from plain carbon steel (k= 60 W/m·K, 
= 7850 kg/m3, c = 430 J/kg·K) and is of thickness L
= 10 mm. To protect it from the corrosive effects of
the furnace combustion gases, one surface of the
wall is coated with a thin ceramic film that, for a
unit surface area, has a thermal resistance of R”t,f =
0.01 m2 ·K/W. The opposite surface is well
insulated from the surroundings.
At furnace start-up the wall is at an initial temperature of Ti = 300K, and
combustion gases at T = 1300 K enter the furnace, providing a convection
coefficient of h = 25 W/m·K at the ceramic film. Assuming the film to have
negligible thermal capacitance, how long will it take for the inner
surface of the steel to achieve a temperature of Ts,i = 1200 K?
What is the temperature Ts,o of the exposed surface of the
ceramic film at this time

[ME421] Heat Transfer 23 Dr. Alia H. Marafie

5. Transient Conduction
5.3 General Lumped Capacitance Analysis

 5.1 The Lumped Capacitance Method

 5.2 Validity of the Lumped Capacitance Method
 5.3 General Lumped Capacitance
Analysis
5.3.1. Radiation Only
5.3.2. Negligible Radiation
5.4 Spatial Effects
5.5 The Plane Wall with Convection
5.6 Radial Systems with Convection
5.7 The Semi-Infinite Solid
5.10 Finite-Difference Methods

[ME421] Heat Transfer 24 Dr. Alia H. Marafie

12
 5. Transient Conduction
5.3 General Lumped Capacitance Analysis

 Transient conduction in solid can be induced by:

 Convection heat transfer to/or from
adjoining fluid, (h, T∞),
 Radiation exchange between the
solid and the surroundings
(hr, Tsur),
 Applying heat flux on the surface
(for example, attaching a film or
sheet electrical heater to the
surface) and/or
 Internal energy generation within Figure 5.5 Control surface for general
the solid (for example, passing an lumped capacitance analysis.
electrical current through the solid).

[ME421] Heat Transfer 25 Dr. Alia H. Marafie

5. Transient Conduction
5.3 General Lumped Capacitance Analysis

 Consider a general case, which

includes convection (q”conv), radiation
(q”rad), imposed on surface As(c,r),
and/or an applied heat flux (q”s),
imposed on surface As(h), as well as
internal energy generation:
dEst dT
Ein  Eout  Eg  Est    Vc
dt dt
(5.13)
 Assuming energy outflow due to
convection and radiation and with Figure 5.5 Control surface for general
lumped capacitance analysis.
inflow due to an applied heat flux
dT
qs As ,h  (qconv
  qrad
 ) As ( c ,r )  Eg   Vc (5.14)
dt
[ME421] Heat Transfer 26 Dr. Alia H. Marafie

13
 5. Transient Conduction
5.3 General Lumped Capacitance Analysis

 Or express the convection and radiation heat transfer in term of

the temperatures,
dT
qs As ,h  [h(T  T )   (T 4  Tsur
4
)] As ( c ,r )  Eg   Vc
dt
(5.15)
Nonlinear, first-order, nonhomogeneous, ordinary differential
equation that CANNOT be integrated to obtain an exact solution.

Exact solutions may be obtained for simplified versions of the

equation

1st Case: 2nd Case:

Radiation only Convection, heat generation
& heat flux.
[ME421] Heat Transfer 27 Dr. Alia H. Marafie

5. Transient Conduction
5.3 General Lumped Capacitance Analysis
5.3.1. Radiation Only
 1st Case: Radiation only (no imposed heat flux or heat
generation and convection is either nonexistent (a vacuum) or
negligible relative to radiation, Equation (5.15) reduces to
dT
 Vc   As ,r (T 4  Tsur
4
) (5.16)
dt
 Separation of variables,
 A s ,r t T dT
Evaluating the
integration,  Vc o
dt  T 4
iT
sur  T
4
(5.17)

 Vc  Tsur  T Tsur  Ti   T   Ti   
t  1n  1n  2  tan1
   tan1
  
4 As ,r Tsur
3
 Tsur  T Tsur  Ti   Tsur   Tsur   
 Result necessitates implicit evaluation of T(t).
 Using trial-and-error techniques to evaluate T implicitly (5.18)
[ME421] Heat Transfer 28 Dr. Alia H. Marafie

14
 5. Transient Conduction
5.3 General Lumped Capacitance Analysis
5.3.2. Negligible Radiation
 2nd Case: Convection, heat generation and heat
flux (radiation is negligible)
dT
qs As ,h  h(T  T ) As ,c  Eg   Vc (5.14)
dt
 Introducing a temperature difference  ≡ T–T where
d /dt = dT/dt, equation (5.14) reduces to a linear, first order,
nonhomogenous differential equation of the form
d a
hAs ,c
 a  b  0 where
dt  Vc
(5.20)  qA  Eg 
b   s s ,h 
  Vc 
[ME421] Heat Transfer 29 Dr. Alia H. Marafie

5. Transient Conduction
5.3 General Lumped Capacitance Analysis
5.3.2. Negligible Radiation
 To eliminate the nonhomogeneity by introducing the
transformation
      b / a (5.21)
 Then non-homogeneous differential equation is transformed into a
homogeneous equation of the form:
d 
 a  (5.22)
dt
Separation of variables and integrating from
t=0 to any t and rearranging,

T  T b/a
 exp  at   1  exp  at   (5.25)
Ti  T Ti  T 
Resolve Example 5.2-3 in the textbook

[ME421] Heat Transfer 30 Dr. Alia H. Marafie

15
 5. Transient Conduction
5.4 Spatial Effects
 5.1 The Lumped Capacitance Method
 5.2 Validity of the Lumped Capacitance Method
 5.3 General Lumped Capacitance Analysis

 5.4 Spatial Effects

5.5 The Plane Wall with Convection
5.6 Radial Systems with Convection
5.7 The Semi-Infinite Solid
5.10 Finite-Difference Methods

[ME421] Heat Transfer 31 Dr. Alia H. Marafie

5. Transient Conduction
5.4 Spatial Effects
 In many situations the lumped capacitance method is
inappropriate, i.e. the temperature gradients within
the medium are no longer negligible  an alternative
approach must be used.
 In such case, consideration must be given to spatial
(eg. x), as well as temporal (t), variations in
temperature during the transient process  T = T(x,t)

[ME421] Heat Transfer 32 Dr. Alia H. Marafie

16
 5. Transient Conduction
5.4 Spatial Effects
 For a 1-D plane wall with symmetrical convection conditions,
without internal heat generation and constant thermal
conductivity, the heat equation and initial/boundary conditions are
expressed as 2  T 1 T
Heat equation:  (5.29)
x2  t
Initial condition: T  x,0   T i (5.30)

Boundary conditions:
a) Symmetry condition for T
the midplane of the wall: 0 (5.31)
x x 0
b) Surface condition for t > 0 :
Figure 5.6a One-dimensional
T
 h T  L, t   T  
plane wall with an initial uniform
k (5.32) temperature subjected to
x xL sudden convection conditions.

[ME421] Heat Transfer 33 Dr. Alia H. Marafie

5. Transient Conduction
5.4 Spatial Effects
 From Eqs. 5.29-32, the temperature depends on seven different
physical parameters,
T  T  x, t , T i , T  , k ,  , h  (5.33)
 How may the functional dependence be simplified?
 The functional dependence of the temperature on these
parameters may be simplified greatly by non-
dimensionalizing the governing equations and
initial/boundary conditions  recast some of the dependent
variables (T, x, t) in a non-dimensional form.

[ME421] Heat Transfer 34 Dr. Alia H. Marafie

17
 5. Transient Conduction
5.4 Spatial Effects
 Non-dimensionalization of heat equation and
initial/boundary conditions:
 Dimensionless temperature,  *:
(the ratio of the temperature difference, ,  T T (5.34)
to the maximum possible temperature * 
difference, I ).
 i T i  T  (0 ≤ * ≤1)

 Dimensionless coordinate, x*:

(the coordinate is non-dimensionalized by x (5.35)
the characteristic length L, where L is the x* 
half-thickness of the plane wall. L (0 ≤x* ≤1)

 Dimensionless time, t*:

t
(the time is non-dimensionalized using the t*  Fo (5.36)
thermal diffusivity, , and L2 ). L2
[ME421] Heat Transfer 35 Dr. Alia H. Marafie

5. Transient Conduction
5.4 Spatial Effects
 Non-dimensional form of the heat equation and
initial/boundary conditions:
2 *  *  * (5.37)
Heat equation:  
x*2 t* Fo
Initial condition:  *  x * ,0   1 (5.38)
Boundary conditions:
a) Symmetry condition for  *
the midplane of the wall: x *
 0 (5.39)
x * 0
 *
b) Surface condition for t > 0:   Bi  * 1, t *  (5.40)
x * x * 1  Bi  hL k solid 
 The functional dependence may now be expressed as
 *  f  x * , Fo, Bi  (5.41)
Only three independent variables
[ME421] Heat Transfer 36 Dr. Alia H. Marafie

18
 5. Transient Conduction
5.4 Spatial Effects
 The resulting equation can be solved analytically for a number of
simple geometries. We will discuss some of them in the next
sections:
 The Plane Wall with Convection (section 5.5).

 Radial Systems with Convection (section 5.6).

 The Semi-Infinite Solid (section 5.7).

 Typically, the solution for the dimensionless temperature

distribution, Equation 5.34 is in the form of infinite series.
However, except for very small values of Fourier number, this series
may be approximated by a single term.

[ME421] Heat Transfer 37 Dr. Alia H. Marafie

5. Transient Conduction
5.5 The Plane Waal with Convection
 5.1 The Lumped Capacitance Method
 5.2 Validity of the Lumped Capacitance Method
 5.3 General Lumped Capacitance Analysis
 5.4 Spatial Effects
 5.5 The Plane Wall with Convection
5.5.1 Exact Solution
5.5.2 Approximate Solution
5.5.3 Total Energy Transfer
5.5.4 Additional Considerations
5S.1 Graphical Representation
5.6 Radial Systems with Convection
5.7 The Semi-Infinite Solid
5.10 Finite-Difference Methods

[ME421] Heat Transfer 38 Dr. Alia H. Marafie

19
 5. Transient Conduction
5.5 The Plane Wall with Convection
5.5.1 Exact Solution
 Exact Solution:
 If the thickness is small relative to the width and height of the
wall, it is reasonable to assume that conduction occurs exclusively
in the x-direction.
 Similar convection conditions for the
surfaces at x = ± L ( or, x* = ± 1)
 The temperature distribution at any
instant must be symmetrical about
the mid-plane x = 0 (or, x* =0)

Figure 5.6a One-dimensional plane wall with an

initial uniform temperature subjected to sudden
convection conditions.

[ME421] Heat Transfer 39 Dr. Alia H. Marafie

5. Transient Conduction
5.5 The Plane Wall with Convection
5.5.1 Exact Solution
 An exact solution to this 
problem is of the form
 *   C n exp   n2 Fo  cos  n x * 
n 1
\

(5.42a)
where the Cn is 4sin  n
Cn 
2 n  sin  2 n 
(5.32b)

 The values of n (eigenvalues) are

positive roots of the transcendental  n tan  n  Bi (5.42c)
equation

 See Appendix B.3 for first four roots (eigenvalues 1, 2, …4)
Note:  is pronounced ‘zeta’
 is pronounced ‘xi’

[ME421] Heat Transfer 40 Dr. Alia H. Marafie

20
 5. Transient Conduction
5.5 The Plane Wall with Convection
5.5.2 Approximate Solution
 Approximate Solution:
 For values of Fo > 0.2, the infinite series solution, Eq. 5.39, can
be approximated by the first term of the series.
 An approximate solution will have the form

 *  C 1 exp   12 Fo  cos  1x *  (5.43a)

or  *   o* cos  1x *  (5.43b)
where * represents the mid-plane (x* =0) temperature

T o  T    C exp  2 Fo
 o*   1 
T i  T   1 (5.44)

The coefficients C1 and 1 are given in Table 5.1

[ME421] Heat Transfer 41 Dr. Alia H. Marafie

5. Transient Conduction
5.5 The Plane Wall with Convection
5.5.2 Approximate Solution

Important

[ME421] Heat Transfer 42 Dr. Alia H. Marafie

21
 5. Transient Conduction
5.5 The Plane Wall with Convection
5.5.3 Total Energy Transfer
 The change in thermal energy storage (Est) with time can be
calculated by
Energy Balance: E st  Q
 The total energy transferred (Q) from the wall over the time
interval (t) is given as

 sin  1 *  (5.49)
Q  Q o 1   
 1 o Refer to the text book for the
detailed derivations

 where Qo, is the initial internal energy , of the wall relative to the
fluid temperate. Also known as maximum amount of energy
transfer that could occur if the process were continued to t = .

Q o   cV T i  T   (5.47)
[ME421] Heat Transfer 43 Dr. Alia H. Marafie

5. Transient Conduction
5.5 The Plane Wall with Convection
5.5.4 Additional Considerations

 Can the foregoing results be used for a plane wall that is

well insulated on one side and convectively heated or
cooled on the other?
 Can the foregoing results be used if an isothermal
condition (Ts ≠Ti) is instantaneously imposed on both
surfaces of a plane wall or on one surface of a wall
whose other surface is well insulated?

Resolve Example 5.4 in the textbook

[ME421] Heat Transfer 44 Dr. Alia H. Marafie

22
 5. Transient Conduction
5.5 The Plane Wall with Convection
Summary of the Basic Solution Procedure
1) Calculating the dimensionless numbers at specific time and position:
 Given the properties (h, ksolid, L) Calculate the Biot number ( Bi  hLc k )
 Given the time (t)  calculate Forier number ( Fo   t L )
2

 Given the position (x)  calculate the dimensionless position ( x *  x L)

2) Depending on the value of the Biot number and the Fourier number,
determine the solution method:
(a) Bi < 0.1  use the lumped capacitance method.
(b) Bi > 0.1  use the analytical solution.
(b1) Bi > 0.1, Fo > 0.2  use the first term approximation.
(b2) If not, use more than one term of the series to approximate the
solution (will not be consider at this level).

3) Calculate the temperature of the mid-plane (*o) at a given time (Fo):

(b) Fo > 0.2  use Bi to find C1 and 1 from Table 5.1.
 calculate the variation of mid-plane
temperature with time   o*  C 1 exp   12 Fo  (5.53c)

[ME421] Heat Transfer 45 Dr. Alia H. Marafie

5. Transient Conduction
5.5 The Plane Wall with Convection
Summary of the Basic Solution Procedure
4) Calculate the temperature (*) at a given time (Fo) and position (x*),
other than the mid-plane:
(b) Fo > 0.2  use the values of 1 and x*, in addition to the previously
calculated mid-plane temperature (*o) to calculate the
temperature at any location (*).
 *   o* cos  1x *  (5.43b)
5) Calculate the total energy transferred (Q) from the wall over the
time interval (t):
(b) Fo > 0.2  calculate the maximum energy transfer.
Q o   cV T i  T   (5.47)
 use the values of 1 and *o , in addition to the previously
calculated Qo to calculate Q.
 sin  1 * 
Q  Q o 1    (5.49)
 1 o

[ME421] Heat Transfer 46 Dr. Alia H. Marafie

23
 5. Transient Conduction
5.5 The Plane Wall with Convection
5S.1 Graphical Representation
 Alternative Graphical Representation (one-term approximation):
 In cases where Fourier number > 0.2, we can use graphical
methods to illustrate the functional dependency of the
transient temperature distribution (of plane wall of thickness
2L) on the Biot and Fourier numbers.
 This graphical representation is known as the Heisler Charts:
 Figure 5S.1: May be used to obtain the mid-plane temperature of the
wall at, T(0,t) = To(t) at any time during the transient process (or *o) .
 Figure 5S.2: May be used to determine the surface temperature
(x* ≡ ±1) at some time t (or *) .
 Figure 5S.3: May be used to determine the energy transfer from a
plane wall over a time interval t (or Q).

[ME421] Heat Transfer 47 Dr. Alia H. Marafie

5. Transient Conduction
5.5 The Plane Wall with Convection
5S.1 Graphical Representation
 The charts are used easily to find all quantities of interest
as follows:
1) Calculate the Biot and Fourier numbers and all the parameters as
in the previous procedure.
2) Figure 5S.1: Read the value of θ*o that correspond to the
calculated Bi and Fo numbers. Use equation 5.41 to calculate To(t)
 mid-plane temperature (To) 

T o  T    C exp  2 Fo
 o*   1 
T i  T   1
(5.44)
Figure 5S.1 Mid-plane
temperature as a function of time
[ME421] Heat Transfer for a plane wall
Dr.ofAlia
thickness 2L.
H. Marafie
48

24
 5. Transient Conduction
5.5 The Plane Wall with Convection
5S.1 Graphical Representation
3) Figure 5S.2: Determine the temperature variation in a location
other than the mid-plane temperature.
3.a) To determine the temperature at specific position:
 calculate the values of dimensionless
position (x*= x/L) and Bi -1
(Bi -1 = k/hL).
 knowing Bi -1 and x*  read the
corresponding dimensionless
temperature (θ/θo).
 knowing To (step2) and (θ/θo)
 calculate the temperature at
this position (T).

Figure 5S.2 Temperature

[ME421] Heat Transfer 49 distribution in Dr. Aliawall
a plane H. Marafie
2L.

5. Transient Conduction
5.5 The Plane Wall with Convection
5S.1 Graphical Representation
3.b) To determine the time required for the surface to reach a
prescribed temperature (T):
 calculate the values of dimensionless position (x*= x/L) and Bi -1
(Bi -1 = k/hL).
 knowing Bi -1 and x*  read the
corresponding dimensionless
temperature (θ/θo).
 knowing (T) and (θ/θo)
 calculate To.
 knowing To  calculate θ*o.
 using Figure 5S.1 and knowing
θ*o and Bi -1  find Fo and then
calculate t.

Figure 5S.2 Temperature

[ME421] Heat Transfer 50 distribution in Dr. Aliawall
a plane H. Marafie
2L.

25
 5. Transient Conduction
5.5 The Plane Wall with Convection
5S.1 Graphical Representation
4) Figure 5S.3: To determine the change in thermal energy stored
in the solid with respect to time, Est, which is equivalent to the
energy transfer , Q, from a plane wall over a time interval t (see
Equation 5.42).
 knowing the values of Bi and Bi2Fo  read Q/Qo.

Plane Wall

Figure 5S.3 Internal energy change as a function for a plane wall of thickness 2L.

[ME421] Heat Transfer 51 Dr. Alia H. Marafie

5. Transient Conduction
5.6 Radial Systems with Convection
 5.1 The Lumped Capacitance Method
 5.2 Validity of the Lumped Capacitance Method
 5.3 General Lumped Capacitance Analysis
 5.4 Spatial Effects
 5.5 The Plane Wall with Convection

 5.6 Radial Systems with

Convection
5.7 The Semi-Infinite Solid
5.10 Finite-Difference Methods

[ME421] Heat Transfer 52 Dr. Alia H. Marafie

26
 5. Transient Conduction
5.6 Radial Systems with Convection
 For an infinite cylinder (long rods) or spheres of radius ro , which is
at an initial uniform temperature and then heated or cooled by
convection, two solution type may be obtained:
 Exact Solution.
 Approximate Solution.
Dimensionless temperature:
T  T
* 
Ti  T
 For radial systems, Bi and Fo
numbers are defined using ro: Figure 5.6b One-dimensional infinite cylinder
or sphere with an initial uniform temperature
Bi  hr o / k , Fo   t / r o2 subjected to sudden convection conditions.

[ME421] Heat Transfer 53 Dr. Alia H. Marafie

5. Transient Conduction
5.6 Radial Systems with Convection
5.6.1 Exact Solution (Fo < 0.2)
 Infinite Cylinder or long rod:

 *   Cn exp(  n2 Fo) J 0 ( n r*) (5.50a)
n 1

2 J1 ( n ) Bi  hr o / k
Cn 
 n J 0 ( n )  J12 ( n )
2 (5.50b)
Fo   t / r o2
The quantities Jo and J1 are Bessel functions of the first kind
and their values obtained in Appendix B.4.
 x2k
J 0 ( x)   (1) k , (5.50c)
k 0 22 k (k !)2
x  x 2k
J1 ( x)   (1) 2 k
k
(5.50d)
2 k 0 2 k!(k  1)!
[ME421] Heat Transfer 54 Dr. Alia H. Marafie

27
 5. Transient Conduction
5.6 Radial Systems with Convection
5.6.1 Exact Solution (Fo < 0.2)
 Sphere:
 1
 *   Cn exp(  n2 Fo) sin( n r*) (5.51a)
n 1  nr *

4[sin( n )   n cos( n )] Bi  hr o / k
Cn  (5.51b)
2 n  sin(2 n ) Fo   t / r o2

[ME421] Heat Transfer 55 Dr. Alia H. Marafie

5. Transient Conduction
5.6 Radial Systems with Convection
5.6.1 Exact Solution (Fo < 0.2)
 Infinite Cylinder or long rod (One-term Approximation):

 *  0* J 0 ( 1r*) (5.52b)

0*  [(T0  T ) / Ti  T )] Bi  hr o / k
 C1 exp( Fo)
1
2
(5.52c) Fo   t / r o2
C1 and ζ1 can be found in Table 5.1, page 274.
 Graphical Representation:
 Graphical representation of the one-term approximations
for infinite cylinder are presented in Figures 5S4-5S5.

[ME421] Heat Transfer 56 Dr. Alia H. Marafie

28
 5. Transient Conduction
5.6 Radial Systems with Convection
5.6.2 Approximate Solutions (Fo > 0.2)

Cylinder

Figure 5S.4 Centerline temperature as a function

of time for an infinite cylinder of radius ro.

[ME421] Heat Transfer 57 Dr. Alia H. Marafie

5. Transient Conduction
5.6 Radial Systems with Convection
5.6.2 Approximate Solutions (Fo > 0.2)

Cylinder

Figure 5S.5 Temperature distribution in an infinite cylinder of radius ro.

[ME421] Heat Transfer 58 Dr. Alia H. Marafie

29
 5. Transient Conduction
5.6 Radial Systems with Convection
5.6.2 Approximate Solutions (Fo > 0.2)
 Sphere (One-term approximation):
1
 *   0* sin( 1r*) (5.53b)
 1r *
Bi  hr o / k
0*  [(T0  T ) /(Ti  T )] Fo   t / r o2
(5.53c)
 C1 exp( Fo) 1
2

C1 and ζ1 can be found in Table 5.1, page 274.

 Graphical Representation:
 Graphical representation of the one-term
approximations for sphere are presented in Figures
5S6-5S7.
[ME421] Heat Transfer 59 Dr. Alia H. Marafie

5. Transient Conduction
5.6 Radial Systems with Convection
5.6.2 Approximate Solutions (Fo > 0.2)

Sphere

Figure 5S.6 Centerline temperature as a function

of time for a sphere of radius ro.

[ME421] Heat Transfer 60 Dr. Alia H. Marafie

30
 5. Transient Conduction
5.6 Radial Systems with Convection
5.6.2 Approximate Solutions (Fo > 0.2)

Sphere

Figure 5S.7 Temperature distribution in a sphere of radius ro.

[ME421] Heat Transfer 61 Dr. Alia H. Marafie

5. Transient Conduction
5.6 Radial Systems with Convection
5.6.3 Total Energy Transfer (Fo > 0.2)
 The total energy transferred (Q) from an infinite cylinder or
sphere over the time interval t = t is given as

Q 2 *
 Infinite Cylinder  1  0 J1 ( 1 ) (5.54)
Q0 1
Q 3 *
 Sphere  1  30 [sin( 1 )   1 cos( 1 )]
Q0 1 (5.55)

 ζ1 can be found in Table 5.1, page 301.

 o* can be determined from Equation 5.52c or 5.53c using
the coefficients of Table 5.1 for the appropriate system.
 The graphical representation of Equations 5.54 and 5.55 is
represented in Figures 5S.8 and 5S.9 respectively.

[ME421] Heat Transfer 62 Dr. Alia H. Marafie

31
 5. Transient Conduction
5.6 Radial Systems with Convection
5.6.3 Total Energy Transfer (Fo > 0.2)

Cylinder

Figure 5S.8 Internal energy change as a function of time for an infinite cylinder of radius ro.

[ME421] Heat Transfer 63 Dr. Alia H. Marafie

5. Transient Conduction
5.6 Radial Systems with Convection
5.6.3 Total Energy Transfer (Fo > 0.2)

Sphere

Figure 5S.9 Internal energy change as a function of time for a sphere of radius ro.

[ME421] Heat Transfer 64 Dr. Alia H. Marafie

32
 5. Transient Conduction
Summary of the Transient Solutions of Plane
and Radial Systems with Convection

Geometry Exact Solution Approximate Solution Total Heat Heisler Chart

Equations Equations Transfer (graphical
(infinite series) (one-term approximation) representation)
(θ*) (θ*) (Q/Q0) (θ*0, θ*, Q/Q0)

Plane wall Eq.(5.42) Eq. (5.43a) Eq. (5.49) Charts 5S1 5S3

Long rod Eq. (5.50a) Eq. (5.52b) Eq. (5.54) Charts 5S4, 5S5, 5S8

Sphere Eq. (5.51a) Eq. (5.53b) Eq. (5.55) Charts 5S6, 5S7, 5S9

Note: The constants C, and ζ values are found from Table 5.1 as functions
of the Biot number
Resolve Example 5.5 in the textbook

[ME421] Heat Transfer 65 Dr. Alia H. Marafie

5. Transient Conduction
5.6 Radial Systems with Convection
 Problem 5.67: As part of heat treatment process, cylindrical, 304
stainless steel rods of 125-mm diameter are cooled from an initial
temperature of 500C by suspending them in an oil bath at 30C.
(a) If a convection coefficient of 400 W/m2·K is maintained by circulation of
the oil, how long does it takes for the centerline of a rod to reach a
temperature of 50C, at which point it is withdrawn from the
bath?
(b) If 10 rods of length L = 1 m are processed per hour, what is the
nominal rate at which energy must be exerted from the bath (the
cooling load)?
(c) What is the surface temperature of the rod when it is removed from the
bath?
 If the first term approximation is valid, solve the problem using the
approximate solution and the graphical representation. Compare the
two solutions.

[ME421] Heat Transfer 66 Dr. Alia H. Marafie

33
 5. Transient Conduction
5.7 The Semi-Infinite Solid
 5.1 The Lumped Capacitance Method
 5.2 Validity of the Lumped Capacitance Method
 5.3 General Lumped Capacitance Analysis
 5.4 Spatial Effects
 5.5 The Plane Wall with Convection
 5.6 Radial Systems with Convection

 5.7 The Semi-Infinite Solid

5.10 Finite-Difference Methods

[ME421] Heat Transfer 67 Dr. Alia H. Marafie

5. Transient Conduction
5.7 The Semi-Infinite Solid
 A solid that is initially of uniform temperature Ti and is assumed
to extend to infinity from a surface at which a sudden change of
thermal conditions is imposed at this surface.
Energy equation:  2T 1 T
 (5.29)
 PDE with T = T (x,t) x 2  t
Interior Boundary Condition
T ( x  , t )  Ti (5.56)
 The heat equation is cast in terms of a similarity
variable, η, defined as η ≡ x/(4αt)1/2. After a
few simple chain rule application, the heat Figure 5.7a Semi-infinite
equation looks like d 2T dT
solid exposed to a sudden

 2
change in wall temperature.
(5.57)
d 2 d
 ODE with T = T ()
[ME421] Heat Transfer 68 Dr. Alia H. Marafie

34
 5. Transient Conduction
5.7 The Semi-Infinite Solid
Semi-infinite solid

Case 1: Change in Case 2: Uniform Heat Case 3: Convection

Surface Temperature (Ts). Flux (qs=qo) Heat Transfer (h, T∞)

Figure 5.7 Transient temperature distributions in a semi-infinite solid for three surface conditions:
[ME421]constant
Heat Transfer surface heat flux, and surfaceDr.
surface temperature, constant69 Alia H. Marafie
convection.

5. Transient Conduction
5.7 The Semi-Infinite Solid
 1st Case: Change in Surface Temperature (Ts)
T  0, t   T   0   T s (5.58)
T  x   or t  0   T      T i
(5.59)
Temperature distribution:

T  x, t   T s  x 
 erf   (5.60)
Ti T s  2 t  
x
2 t
erf : Gaussian error function tabulated in Appendix B
Surface Heat Flux:

k T s  T i 
qs  (5.61)
 t
[ME421] Heat Transfer 70 Dr. Alia H. Marafie

35
 5. Transient Conduction
5.7 The Semi-Infinite Solid
 2nd Case: Uniform Heat Flux (qs=qo)
k T x x 0
 q s"
T  x   or t  0   T      T i
(5.59)
Temperature distribution:
2qo  t /   x2 
T  x, t   T i  exp    
x
k  4 t  2 t
q x  x 
 o erfc  
k  2 t  (5.62)
erfc  : Complementary error function
erfc  ≡ 1- erf 

[ME421] Heat Transfer 71 Dr. Alia H. Marafie

5. Transient Conduction
5.7 The Semi-Infinite Solid
 3rd Case: Convection Heat Transfer (h, T∞)
k T x x 0
 h T   T (0, t )
T  x   or t  0   T      T i
(5.59)
Temperature distribution:
T  x, t   T i  x  x
 erfc   
T Ti  2 t  2 t

  hx h 2 t     x h  t 
 exp     erfc   
  k k 2     2 t
 k   (5.63)
Resolve Example 5.7 in the textbook

[ME421] Heat Transfer 72 Dr. Alia H. Marafie

36
 5. Transient Conduction
5.10 Finite-Difference Methods
 5.1 The Lumped Capacitance Method
 5.2 Validity of the Lumped Capacitance Method
 5.3 General Lumped Capacitance Analysis
 5.4 Spatial Effects
 5.5 The Plane Wall with Convection
 5.6 Radial Systems with Convection
 5.7 The Semi-Infinite Solid
 5.10 Finite-Difference Methods

[ME421] Heat Transfer 73 Dr. Alia H. Marafie

5. Transient Conduction
5.10 Finite-Difference Methods
 Analytical solutions to transient problems are restricted to simple
geometries and boundary conditions.
 However, in many cases the geometry and/or boundary conditions
preclude the use of analytical techniques  must use a
numerical method (finite-difference or finite-element
methods).
 Finite-difference method for steady-state conditions was
introduced in Section 4.4 and can be extended to transient
problems.

Heat equation: 1 T  2T  2T
2-D, transient problem   (5.75)
with no heat generation:  t x 2 y 2

[ME421] Heat Transfer 74 Dr. Alia H. Marafie

37
 5. Transient Conduction
5.10 Finite-Difference Methods
 The numerical method of a transient problem provides an
approximate method for determining temperatures at discrete
(nodal) points of a physical system and at discrete times
during the transient process.
 An explicit and implicit forms of the finite-difference solutions
of transient conduction problems will be consider in this section.

[ME421] Heat Transfer 75 Dr. Alia H. Marafie

5. Transient Conduction
5.10 Finite-Difference Methods
5.10.1 Discretization of the H.E.: The Explicit
Method
 Procedure (Explicit Method): 1 T  2T  2T
  (5.75)
1) Represent the physical system by a  t x 2 y 2
nodal network, with an m, n notation
used to designate the x and y locations of
discrete points in the network.

2) Use the central-difference

approximations to express the spatial
derivatives  2T/x2 and  2T/y2 (see
Section 4.4) at time p.

 2T T( mp 1),n  T( mp 1),n  2Tmp,n  2T Tmp,( n1)  Tmp,( n1)  2Tmp,n

 
x 2 m ,n
(x)2 y 2 m ,n
(y )2
(5.77a) (5.77b)
[ME421] Heat Transfer 76 Dr. Alia H. Marafie

38
 5. Transient Conduction
5.10 Finite-Difference Methods
5.10.1 Discretization of the H.E.: The Explicit
Method
3) Discretize the problem in time by designating a time
increment ∆t, and expressing the time as
t  p t (5.76)
where p assumes integer values, (p = 0, 1, 2,…).
 By this we can find the temperature at the specified
nodal point and at some specified time.
4) Use the forward-difference approximation* (or Forward Euler
scheme) to express the time derivative  T/ t at each node.

T Tmp,n1  Tmp,n
 (5.77c)
t m,n t (p) is the previous time
(p+1) is the new time
* In the explicit method of solution, these temperatures are evaluated at the previous (p)
time. Hence the finite-difference approximation equation is considered to be a forward-
difference approximation to the time derivative.

[ME421] Heat Transfer 77 Dr. Alia H. Marafie

5. Transient Conduction
5.10 Finite-Difference Methods
5.10.1 Discretization of the H.E.: The Explicit
Method
5) Use the heat equation to obtain a finite-difference equation for
each node of unknown temperature.

 Exact differential equation  1 T  2T  2T (5.75)

 
 t x 2 y 2
(5.77c)
T Tmp,n1  Tmp,n  2T T( mp 1),n  T( mp 1),n  2Tmp,n (5.77a)
 
t m,n t x 2 m ,n
(x)2 (5.77b)

 2T Tmp,( n1)  Tmp,( n1)  2Tmp,n


 Approximate algebraic equation y 2 (y )2
(x = y) m ,n

 
Tmp,n1  Fo Tmp1,n  Tmp1,n  Tmp,n 1  Tmp,n 1  1  4Fo Tmp,n (5.79)
Fo  Heat
[ME421]  2
t Transfer
x  finite-difference form
78 of Fourier number Dr. Alia H. Marafie

39
 5. Transient Conduction
5.10 Finite-Difference Methods
5.10.1 Discretization of the H.E.: The Explicit
Method

 
Tmp,n1  Fo Tmp1,n  Tmp1,n  Tmp,n 1  Tmp,n 1  1  4Fo Tmp,n (5.79)

Unknown nodal
temperature for the Known adjoining nodes temperature at known nodal temperature
new time (p+1) previous time (p) at previous (p)

Equation 5.76 is an explicit equation because unknown nodal

temperature for the new time (p+1) are determined exclusively
by known nodal temperature at the previous time (p) 
the calculation of the unknown temperatures is straightforward.

6) Solve the resulting set of equations for the nodal temperatures at

t = ∆t, 2∆t, 3∆t, …, until steady-state is reached (see Stability and
Convergence in the next slides).
[ME421] Heat Transfer 79 Dr. Alia H. Marafie

5. Transient Conduction
5.10 Finite-Difference Methods
5.10.1 Discretization of the H.E.: The Explicit
Method
 Solution Accuracy:
How is solution accuracy affected by the choice of ∆x
and ∆t?
 In explicit schemes, such as the forward-Euler, the solution
accuracy is greatly affected by the choice of Δx and Δy.
 The accuracy of the finite-difference solution may be
improved by decreasing the values of Δx and Δt:
 decreasing Δx  increases the number of interior nodal points.
 decreasing Δt  increases the time interval required to solve the
problem.
 Hence the computational time increases with decreasing of
Δx and Δt.

[ME421] Heat Transfer 80 Dr. Alia H. Marafie

40
 5. Transient Conduction
5.10 Finite-Difference Methods
5.10.1 Discretization of the H.E.: The Explicit
Method
 Stability and Convergence:
What factors influence the choice of ∆t?
 We are NOT totally free to choose the size of the time-step (∆t)
used in marching the solution forward. It is, instead,
determined by stability requirements.
 In transient problems, the solution for the nodal temperature
should continuously approach final (steady-state) values with
increasing time  Convergence.
 Convergence is achieved when the updated values of
unknown do not differ from previous values, to within a
specified convergence tolerance ()

Tmp,n1  Tmp,n  

[ME421] Heat Transfer 81 Dr. Alia H. Marafie

5. Transient Conduction
5.10 Finite-Difference Methods
5.10.1 Discretization of the H.E.: The Explicit
Method
 In explicit method, the solution may oscillate. This oscillation
may become unstable, causing the solution to divert from the
actual steady-state conditions.
 To prevent solution diversion, ∆t must be maintained below a
certain limit, which depend on ∆x and other parameters. This
dependence is termed a stability criterion.
 Stability criterion is determined by requiring that the
coefficient associated with the node of interest at the previous
time is greater than or equal to zero.
Finite-difference equation
for 2-D internal node Tmp,n1  ..............................  ATmp,n , A  0
Hence, for the two-dimensional interior node:

1  4Fo   0  Fo 
1
or t 
 x  2
(5.83)
4 4
• See Table 5.3 for finite-difference
equations for common nodal regions. Resolve Example 5.11 in the textbook

[ME421] Heat Transfer 82 Dr. Alia H. Marafie

41
 5. Transient Conduction
5.10 Finite-Difference Methods
5.10.1 Discretization of the H.E.: The Explicit
Method
 Implicit Method:
 In the implicit method all other terms in the energy balance
are evaluated at the new time corresponding to (p+1). the
equation is then termed a backward-difference approximation,
or backward-Euler.
What is the advantage of using the implicit method
instead of using the explicit method in solving a
transient problem?
 Although the explicit method offers computational convenience,
it suffers from limitations on the selection of t. Stability of the
explicit method requires the use of extremely small values of t,
and a very large number of time intervals for convergence.
While the implicit method may reduce the amount of
computation time and it is unconditionally stable (will be
explain later in this sub-section).

[ME421] Heat Transfer 83 Dr. Alia H. Marafie

5. Transient Conduction
5.10 Finite-Difference Methods
5.10.1 Discretization of the H.E.: The Explicit
Method
 Implicit Method:
 Following similar procedure as the explicit
method, but with evaluating the temperature
at the new time (p+1), instead of the previous
time (p) for the special derivatives.

 Considering the backward-difference approximation to the time

derivative, the implicit form of the finite-difference equation for
the interior node of a two-dimensional system with ∆x=∆y is then*

1  4Fo Tmp,n1  Fo Tmp1,1n  Tmp1,1n  Tmp,n11  Tmp,n11   Tmp,n (5.95)

* For the complete derivations, see the text book Section 5.10.2.

[ME421] Heat Transfer 84 Dr. Alia H. Marafie

42
 5. Transient Conduction
5.10 Finite-Difference Methods
5.10.1 Discretization of the H.E.: The Explicit
Method

1  4Fo Tmp,n1  Fo Tmp1,1n  Tmp1,1n  Tmp,n11  Tmp,n11   Tmp,n (5.95)

Unknown nodal known nodal

Unknown adjoining nodes temperature
temperature for the temperature at
at new time (p+1)
new time (p+1) previous (p)
 From the above equation:
 The new temperature of the (m, n) node depends on the new
temperatures of its adjoining nodes, which are in general
unknown.
 To determine the unknown nodal temperatures at t + ∆t, the
corresponding nodal equations must be solved simultaneously.
 System of N finite-difference equations for N unknown nodal
temperatures may be solved by matrix inversion or Gauss-Seidel
iteration as discussed in Section 4.5.

[ME421] Heat Transfer 85 Dr. Alia H. Marafie

5. Transient Conduction
5.10 Finite-Difference Methods
5.10.1 Discretization of the H.E.: The Explicit
Method
 Transient temperature distribution is determined by marching
the solution in time. Beginning with the known initial conditions
and then solving the nodal equations at each time t = ∆t, 2 ∆t, …
, simultaneously, until the desired final time was reached.

p t T1 T2 T3 …………… TN
Known
0 0 T1,i T2,i T3,i …………… TN,i
1 ∆t -- -- -- …………… --
2 2∆t -- -- -- …………… --
3 3∆t -- -- -- …………… --
. .
. .
. .
Steady-state -- -- -- -- …………… --

[ME421] Heat Transfer 86 Dr. Alia H. Marafie

43
 5. Transient Conduction
5.10 Finite-Difference Methods
5.10.1 Discretization of the H.E.: The Explicit
Method
 The solution in the implicit method is unconditionally stable
 That is, the solution remains stable for all space and time
intervals, in which case there are NO restriction on ∆x and ∆t.
 Since larger values of ∆t may therefore be used with an implicit
method, computation time may often be reduced, with little loss
of accuracy.

• See Table 5.3 for finite-difference equations for common

nodal regions using the implicit method.

Resolve the textbook examples

[ME421] Heat Transfer 87 Dr. Alia H. Marafie

5. Transient Conduction
Summary of Finite-Difference Equations for Common
Nodal Regions: Explicit and Implicit Methods

[ME421] Heat Transfer 88 Dr. Alia H. Marafie

44
 End of Chapter 5:
Transient Conduction

After carefully reading Chapter 5 in the

textbook, make sure to read the summary at
the end of chapter and resolve all the
chapter’s examples.

45