CHE-614

Computational Fluid Dynamics

Finite Volume Method

2-D Steady State Diffusion

Dr. Muhammad Nadeem

[email protected]
Room 15, . : 3774
 Finite Volume Method:
2-D Steady State Diffusion
𝜕 𝜕𝜙 𝜕 𝜕𝜙 N
Γ + Γ + 𝑆𝜙 = 0
𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑦
n A
n δyPN
In addition to the east (E) and W Aw P Ae E
e Δy
west (W) neighbours, a general w
grid node P now also has north A δySP
s s
(N) and south (S) neighbours
S
δxWP δxPE
𝜕 𝜕𝜙 𝜕 𝜕𝜙 Δx
න Γ 𝑑𝑉 + න Γ 𝑑𝑉 + න 𝑆𝜙 𝑑𝑉 = 0
𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑦
∆𝑉 ∆𝑉 ∆𝑉
𝜕𝜙 𝜕𝜙
න𝒏 ⋅ Γ 𝑑𝐴 + න 𝒏 ⋅ Γ 𝑑𝐴 + න 𝑆𝜙 𝑑𝑉 = 0
𝜕𝑥 𝜕𝑦
∆𝐴 ∆𝐴 ∆𝑉

𝜕𝜙 𝜕𝜙
෍ 𝑛𝑖 ⋅ Γ 𝐴 𝑖 + ෍ 𝑛𝑖 ⋅ Γ ҧ
𝐴𝑖 + 𝑆∆𝑉 =0
𝜕𝑥 𝑖
𝜕𝑦 𝑖
91
 Finite Volume Method:
2-D Steady State Diffusion
𝜕 𝜕𝜙 𝜕 𝜕𝜙 N
Γ + Γ + 𝑆𝜙 = 0
𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑦
n A
n δyPN
In addition to the east (E) and W Aw P Ae E
e Δy
west (W) neighbours, a general w
grid node P now also has north A δySP
s s
(N) and south (S) neighbours
S
δxWP δxPE
𝜕 𝜕𝜙 𝜕 𝜕𝜙 Δx
න Γ 𝑑𝑉 + න Γ 𝑑𝑉 + න 𝑆𝜙 𝑑𝑉 = 0
𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑦
∆𝑉 ∆𝑉 ∆𝑉
𝑛 𝑒 𝑛 𝑒
𝜕 𝜕𝜙 𝜕 𝜕𝜙
නන Γ 𝑑𝑥. 𝑑𝑦 + න න Γ 𝑑𝑥. 𝑑𝑦 + න 𝑆𝜙 𝑑𝑉 = 0
𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑦
𝑠 𝑤 𝑠 𝑤 ∆𝑉
𝑒 𝑛 𝑛 𝑒
𝜕𝜙 𝜕𝜙
Γ න 𝑑𝑦 + Γ ҧ
න 𝑑𝑥 + 𝑆∆𝑉 =0
𝜕𝑥 𝑤
𝜕𝑦 𝑠 𝑤
𝑠 92
 Finite Volume Method:
2-D Steady State Diffusion
𝜕𝜙 𝜕𝜙 𝜕𝜙 𝜕𝜙
Γ𝑒 𝐴𝑒 − Γ𝑤 𝐴𝑤 + Γ𝑛 𝐴𝑛 − Γ𝑠 𝐴𝑠 ҧ
+ 𝑆∆𝑉 =0
𝜕𝑥 𝑒
𝜕𝑥 𝑤
𝜕𝑦 𝑛
𝜕𝑦 𝑠
𝐴𝑒 = 𝐴𝑤 = ∆𝑦 and 𝐴𝑛 = 𝐴𝑥 = ∆𝑥
𝜕𝜙 𝜙𝑃 − 𝜙𝑊
Flux across the west face = Γ𝑤 𝐴𝑤 ቤ = Γ𝑤 𝐴𝑤
𝜕𝑥 𝑤 𝛿𝑥𝑊𝑃
𝜕𝜙 𝜙𝐸 − 𝜙𝑃
Flux across the east face = Γ𝑒 𝐴𝑒 ቤ = Γ𝑒 𝐴𝑒
𝜕𝑥 𝛿𝑥𝑃𝐸 An δyPN
𝑒
𝜕𝜙 𝜙 − 𝜙𝑆 Aw Ae
Flux across the south face = Γ𝑠 𝐴𝑠 ቤ = Γ𝑠 𝐴𝑠 𝑃
𝜕𝑦 𝑠 𝛿𝑦𝑆𝑃
𝜕𝜙 𝜙𝑁 − 𝜙𝑃
As δySP
Flux across the north face= Γ𝑛 𝐴𝑛 ቤ = Γ𝑛 𝐴𝑛
𝜕𝑦 𝑛 𝛿𝑦𝑃𝑁
By substituting the expressions of fluxes into equation δxWP δxPE
𝜙𝐸 − 𝜙𝑃 𝜙𝑃 − 𝜙𝑊
Γ𝑒 𝐴𝑒 − Γ𝑤 𝐴𝑤 +
𝛿𝑥𝑃𝐸 𝛿𝑥𝑊𝑃
𝜙𝑁 − 𝜙𝑃 𝜙𝑃 − 𝜙𝑆
Γ𝑛 𝐴𝑛 − Γ𝑠 𝐴𝑠 + 𝑆𝑢 + 𝑆𝑝 𝜙𝑃 ∆𝑉 = 0
𝛿𝑦𝑃𝑁 𝛿𝑦𝑆𝑃 93
 Finite Volume Method:
2-D Steady State Diffusion
This equation can be rearranged as
Γ𝑤 𝐴𝑤 Γ𝑒 𝐴𝑒 Γ𝑠 𝐴𝑠 Γ𝑛 𝐴𝑛
+ + + − 𝑆𝑝 ∆𝑉 𝜙𝑃
𝛿𝑥𝑊𝑃 𝛿𝑥𝑃𝐸 𝛿𝑦𝑆𝑃 𝛿𝑦𝑃𝑁
Γ𝑤 𝐴𝑤 Γ𝑒 𝐴𝑒 Γ𝑠 𝐴𝑠 Γ𝑛 𝐴𝑛
= 𝜙 + 𝜙 + 𝜙 + 𝜙 + 𝑆𝑢 ∆𝑉
𝛿𝑥𝑊𝑃 𝑊 𝛿𝑥𝑃𝐸 𝐸 𝛿𝑦𝑆𝑃 𝑆 𝛿𝑦𝑃𝑁 𝑁

𝑎𝑃 𝜙𝑃 = 𝑎𝑊 𝜙𝑊 + 𝑎𝐸 𝜙𝐸 + 𝑎𝑆 𝜙𝑆 + 𝑎𝑁 𝜙𝑁 + 𝑆𝑢 ∆𝑉

with 𝑎𝑃 = 𝑎𝑊 + 𝑎𝐸 + 𝑎𝑆 + 𝑎𝑁 − 𝑆𝑝 ∆𝑉

Γ𝑤 𝐴𝑤 Γ𝑒 𝐴𝑒 Γ𝑠 𝐴𝑠 Γ𝑛 𝐴𝑛
𝑎𝑊 = 𝑎𝐸 = 𝑎𝑆 = 𝑎𝑁 =
𝛿𝑥𝑊𝑃 𝛿𝑥𝑃𝐸 𝛿𝑦𝑆𝑃 𝛿𝑦𝑃𝑁

94
 Von Neumann Boundary Conditions
In this case, a flux (or normal gradient to the face) for the quantity 𝜙 is
specified at the boundary N
𝜕𝜙 𝜕𝜙
Γ𝑒 𝐴𝑒 − Γ𝑤 𝐴𝑤 + n
𝜕𝑥 𝜕𝑥 A δxb
𝑒 𝑤 P E Δy
𝑞𝑏
w e
𝜕𝜙 𝜕𝜙 s
Γ𝑛 𝐴𝑛 − Γ𝑠 𝐴𝑠 ҧ
+ 𝑆∆𝑉 =0
𝜕𝑦 𝑛
𝜕𝑦 𝑠 Δx

S
𝜙𝐸 − 𝜙𝑃 𝜙𝑁 − 𝜙𝑃 𝜙𝑃 − 𝜙𝑆
Γ𝑒 𝐴𝑒 − 𝑞𝑏 𝐴𝑤 + Γ𝑛 𝐴𝑛 − Γ𝑠 𝐴𝑠 + 𝑆𝑢 + 𝑆𝑝 𝜙𝑃 ∆𝑉 = 0
𝛿𝑥 𝛿𝑦 𝛿𝑦

𝑎𝑃 𝜙𝑃 = 𝑎𝑊 𝜙𝑊 + 𝑎𝐸 𝜙𝐸 + 𝑎𝑆 𝜙𝑆 + 𝑎𝑁 𝜙𝑁 + 𝑏

with 𝑎𝑃 = 𝑎𝐸 + 𝑎𝑆 + 𝑎𝑁 − 𝑆𝑝 ∆𝑉

Γ𝑒 𝐴𝑒 Γ𝑠 𝐴𝑠 Γ𝑛 𝐴𝑛
𝑎𝑊 = 0 𝑎𝐸 = 𝑎𝑆 = 𝑎𝑁 = 𝑏 = 𝑞𝑏 𝐴𝑤 + 𝑆𝑢 ∆𝑉
𝛿𝑥𝑃𝐸 𝛿𝑦𝑆𝑃 𝛿𝑦𝑃𝑁 95
 Von Neumann Boundary Conditions
If both 𝑞𝑏 and 𝑆𝑢 are zero, then 𝜙𝑃 will be bounded by its neighbors.
Otherwise, 𝜙𝑃 can exceed (or fall below) the neighbor values of 𝜙 ,
which is admissible.

Once 𝜙𝑃 is computed, the boundary value, 𝜙𝑏 may be computed using

𝑞𝑏 + Γ𝑏 Τ𝛿𝑥𝑏 𝜙𝑃
𝜙𝑏 =
Γ𝑏 Τ𝛿𝑥𝑏

Finally, the Von Neumann condition can be considered as a natural

boundary condition for the Finite Volume method since, for the case
where the specified flux is zero, nothing needs to be done in terms of
discretization for the face, while a specified value of zero (Dirichlet
condition) will still require the discretization to be carried out.

96
 Example: 2-D Steady State Diffusion
A two-dimensional plate of
thickness 1 cm is shown.
The thermal conductivity of the
plate material is k = 1000 W/m.K.
The west boundary receives a steady
heat flux of 500 kW/m2 and the
south and east boundaries are
insulated.

The north boundary is maintained

at a temperature of 100°C,

Use a uniform grid with ∆x = ∆y = 0.1 m to calculate the

steady state temperature distribution at nodes 1, 2, 3, 4, . . . etc.
97
 Example: 2-D Steady State Diffusion
Governing Equation
𝜕 𝜕𝑇 𝜕 𝜕𝑇
𝑘 + 𝑘 =0
𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑦
This can be written in discretised form as
𝑎𝑃 𝑇𝑃 = 𝑎𝑊 𝑇𝑊 + 𝑎𝐸 𝑇𝐸 + 𝑎𝑆 𝑇𝑆 + 𝑎𝑁 𝑇𝑁
where 𝑎 = 𝑘 𝐴 , 𝑎 = 𝑘 𝐴 𝑎 = 𝑘 𝐴 , 𝑎 = 𝑘 𝐴
𝑊
𝛿𝑥 𝑤 𝐸 𝛿𝑥 𝑒 𝑆 𝛿𝑦 𝑠 𝑁 𝛿𝑦 𝑛
𝑎𝑃 = 𝑎𝑊 + 𝑎𝐸 + 𝑎𝑆 + 𝑎𝑁
1000
In this case, 𝑎𝑊 = 𝑎𝐸 = 𝑎𝑆 = 𝑎𝑁 = × 0.1 × 0.01 = 10
0.1
At interior points 6 and 7 𝑎𝑃 = 𝑎𝑊 + 𝑎𝐸 + 𝑎𝑆 + 𝑎𝑁 = 40
40𝑇6 = 10𝑇2 + 10𝑇10 + 10𝑇5 + 10𝑇7
40𝑇7 = 10𝑇3 + 10𝑇11 + 10𝑇6 + 10𝑇8
98
 Example: 2-D Steady State Diffusion
At a boundary node the discretised equation takes the form
𝑎𝑃 𝑇𝑃 = 𝑎𝑊 𝑇𝑊 + 𝑎𝐸 𝑇𝐸 + 𝑎𝑆 𝑇𝑆 + 𝑎𝑁 𝑇𝑁 + 𝑆𝑢 ∆𝑉
with 𝑎𝑃 = 𝑎𝑊 + 𝑎𝐸 + 𝑎𝑆 + 𝑎𝑁 − 𝑆𝑝 ∆𝑉
At node 1
𝑎𝑊 = 0 𝑎𝑆 = 0
Let 𝑏𝑊 be the contribution to the source term from
the west and 𝑏𝑆 be form south
𝑏𝑊 = 𝑞𝑤 ⋅ 𝐴𝑤 = 500 × 103 × 0.1 × 0.01 = 500
𝑏𝑆 = 0
Total source 𝑆𝑢 ∆𝑉 = 𝑏𝑊 + 𝑏𝑆 = 500 𝑆𝑝 ∆𝑉 = 0
𝑎𝑃 = 0 + 10 + 0 + 10 − 0 = 20
The discretised equation at node 1 is
20𝑇1 = 10𝑇2 + 10𝑇5 + 500
99
 Example: 2-D Steady State Diffusion
At node 4
West is a constant flux boundary
𝑎𝑊 = 0 𝑏𝑊 = 𝑞𝑤 ⋅ 𝐴𝑤
= 500 × 103 × 0.1 × 0.01 = 500
North is a constant temperature boundary
2𝑘
𝑎𝑁 = 0 𝑏𝑁 = 𝐴 × 100 = 2000
𝛿𝑦 𝑛
2𝑘
𝑆𝑝𝑁 ∆𝑉 = − 𝐴𝑛 = −20
𝛿𝑦
Total source 𝑆𝑢 ∆𝑉 = 𝑏𝑊 + 𝑏𝑁 = 500 + 2000 = 2500
𝑎𝑃 = 𝑎𝑊 + 𝑎𝐸 + 𝑎𝑆 + 𝑎𝑁 − 𝑆𝑝 ∆𝑉
= 0 + 10 + 10 + 0 − −20 = 40
The discretised equation at node 4 is
40𝑇4 = 10𝑇3 + 10𝑇8 + 2500
100
 Example: 2-D Steady State Diffusion
The coefficients and the source term of the discretisation equation for
all points

101
 Example: 2-D Steady State Diffusion

 20 −10 0 0 −10 0 0 0 0 0 0 0  T1   500 

 −10 30 −10 0 0 −10 0 0 0 0 0 0  T2   500 
     
 0 −10 30 −10 0 0 −10 0 0 0 0 0  T3   500 
     
 0 0 −10 40 0 0 0 −10 0 0 0 0  T4   2500 
 −10 0 0 0 30 −10 0 0 −10 0 0 0  T5   0
     
 0 −10 0 0 −10 40 −10 0 0 −10 0 0  T6  
=
0
 0 0 −10 0 0 −10 40 −10 0 0 −10 
0 T7   0
     
 0 0 0 −10 0 0 −10 50 0 0 0 −10  T8   2000 
 0 0 0 0 −10 0 0 0 20 −10 0 0  T9   0
     
 0 0 0 0 0 −10 0 0 −10 30 −10 0  T10   0
 0 0 0 0 0 0 −10 0 0 −10 30 −10  T11   0
     
 0 0 0 0 0 0 0 −10 0 0 −10 40  T12   2000 

102
 Example: 2-D Steady State Diffusion
Application of the TDMA to two-dimensional problems
𝑎𝑃 𝜙𝑃 = 𝑎𝑊 𝜙𝑊 + 𝑎𝐸 𝜙𝐸 + 𝑎𝑆 𝜙𝑆 + 𝑎𝑁 𝜙𝑁 + 𝑏

−𝑎𝑆 𝜙𝑆 + 𝑎𝑃 𝜙𝑃 − 𝑎𝑁 𝜙𝑁 = 𝑎𝑊 𝜙𝑊 + 𝑎𝐸 𝜙𝐸 + 𝑏

103
 Example: 2-D Steady State Diffusion
Application of the TDMA to two-dimensional problems
All west coefficients are zero at points 1, 2, 3 and 4
East values (points 5, 6, 7 and 8) are required for the evaluation of C
They are unknown at this stage and are assumed to be zero as an initial guess

The values of 𝛼𝑗, βj, 𝐷𝑗 and 𝐶𝑗 can be calculated using equations

−𝑎𝑆 𝜙𝑆 + 𝑎𝑃 𝜙𝑃 − 𝑎𝑁 𝜙𝑁 = 𝑎𝑊 𝜙𝑊 + 𝑎𝐸 𝜙𝐸 + 𝑏
−𝛽𝑗 𝜙𝑗−1 + 𝐷𝑗 𝜙𝑗 − 𝛼𝑗 𝜙𝑗+1 = 𝐶𝑗

104
 Example: 2-D Steady State Diffusion
For line 1: 𝜙𝑗 = 𝐴𝑗 𝜙𝑛+1 + 𝐶𝑗′

105
 Example: 2-D Steady State Diffusion
For line 2: For line 3:

Values at the end of first iteration

The converged solution after 37 iterations

227.7
106
 Example: 2-D Steady State Diffusion

Demonstration of MATLAB Code for Example 7-2 using TDMA

line-by-line

107
 Mixed Boundary Condition
The mixed boundary condition refers to the situation where information
at the boundary is given via a convection transfer coefficient (ℎ∞ ) and
a surrounding value for 𝜙∞ as
ℎ∞ 𝐴 𝜙∞ − 𝜙𝑏
N
For the boundary control volume
Δx
𝜙𝑏 − 𝜙𝑃 ℎ∞
𝑘𝐴 = ℎ∞ 𝐴 𝜙∞ − 𝜙𝑏 P 𝜙𝑏
𝛿𝑥𝑏 W
Δy
𝜙∞
δxb
ℎ∞ 𝜙∞ + 𝑘Τ𝛿𝑥𝑏 𝜙𝑃 ℎ∞ 𝛿𝑥𝑏 Τ𝑘 𝜙∞ + 𝜙𝑃
𝜙𝑏 = =
ℎ∞ + 𝑘Τ𝛿𝑥𝑏 1 + ℎ∞ 𝛿𝑥𝑏 Τ𝑘 S
ℎ∞ 𝜙∞ + 𝑘Τ𝛿𝑥𝑏 𝜙𝑃
ℎ∞ 𝐴 𝜙∞ − 𝜙𝑏 ℎ∞ 𝐴 𝜙∞ −
ℎ∞ + 𝑘Τ𝛿𝑥𝑏
ℎ∞ 𝜙∞ + 𝑘Τ𝛿𝑥𝑏 𝜙∞ − ℎ∞ 𝜙∞ − 𝑘Τ𝛿𝑥𝑏 𝜙𝑃
= ℎ∞ 𝐴
ℎ∞ + 𝑘Τ𝛿𝑥𝑏
𝜙𝑏 − 𝜙𝑃 ℎ∞ 𝑘Τ𝛿𝑥𝑏
𝑘𝐴 = 𝐴 𝜙∞ − 𝜙𝑃 = 𝑅𝑒𝑞 𝐴 𝜙∞ − 𝜙𝑃
𝛿𝑥𝑏 ℎ∞ + 𝑘Τ𝛿𝑥𝑏
108
 Mixed Boundary Condition
The discretization equation for the boundary control volume P.
𝑎𝑃 𝜙𝑃 = 𝑎𝑊 𝜙𝑊 + 𝑎𝑁 𝜙𝑁 + 𝑎𝑆 𝜙𝑆 + 𝑏
𝑘 𝑘 𝑘
𝑎𝐸 = 0 𝑎𝑊 = 𝐴 𝑎𝑁 = 𝐴 𝑎𝑆 = 𝐴
𝛿𝑥 𝛿𝑦 𝛿𝑦
N
𝑎𝑃 = 𝑎𝐸 + 𝑎𝑊 + 𝑎𝑁 + 𝑎𝑆 − 𝑆𝑝 ∆𝑥∆𝑦 Δx
ℎ∞
𝑏 = 𝑅𝑒𝑞 ∆𝑦𝜙∞ + 𝑆𝑢 ∆𝑥∆𝑦 W P 𝜙𝑏 𝜙∞
ℎ∞ 𝑘Τ𝛿𝑥𝑏 ℎ∞ Δy δxb
𝑅𝑒𝑞 = =
Τ
ℎ∞ + 𝑘 𝛿𝑥𝑏 1 + ℎ∞ 𝛿𝑥𝑏 Τ𝑘
S
𝑎𝑃 > ∑𝑎𝑛𝑏 for the boundary cell 𝑃 if 𝑆 = 0.
Thus, mixed boundaries are like Dirichlet boundaries
in that the boundary condition helps ensure that the
Scarborough criterion is met
The cell value 𝜙𝑃 is bounded by its neighbor values and the external value, 𝜙∞
The boundary value, 𝜙𝑏 , may be computed once a solution has been obtained.
It is bounded by 𝜙𝑃 and 𝜙∞ as shown on previous slide
109
 Finite Volume Method:
3-D Steady State Diffusion
𝜕 𝜕𝜙 𝜕 𝜕𝜙 𝜕 𝜕𝜙
Γ + Γ + Γ + 𝑆𝜙 = 0
𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑦 𝜕𝑧 𝜕𝑧
𝜕𝜙 𝜕𝜙
Γ𝑒 𝐴𝑒 − Γ𝑤 𝐴𝑤
𝜕𝑥 𝑒
𝜕𝑥 𝑤

𝜕𝜙 𝜕𝜙
+ Γ𝑛 𝐴𝑛 − Γ𝑠 𝐴𝑠
𝜕𝑦 𝑛
𝜕𝑦 𝑠

𝜕𝜙 𝜕𝜙
+ Γ𝑡 𝐴𝑡 − Γ𝑏 𝐴𝑏 ҧ
+ 𝑆∆𝑉 =0
𝜕𝑧 𝑡
𝜕𝑧 𝑏

𝜙𝐸 − 𝜙𝑃 𝜙𝑃 − 𝜙𝑊 𝜙𝑁 − 𝜙𝑃 𝜙𝑃 − 𝜙𝑆
Γ𝑒 𝐴𝑒 − Γ𝑤 𝐴𝑤 + Γ𝑛 𝐴𝑛 − Γ𝑠 𝐴𝑠
𝛿𝑥𝑃𝐸 𝛿𝑥𝑊𝑃 𝛿𝑦𝑃𝑁 𝛿𝑦𝑆𝑃
𝜙 𝑇 − 𝜙𝑃 𝜙𝑃 − 𝜙𝐵
+ Γ𝑡 𝐴𝑡 − Γ𝑏 𝐴𝑏 + 𝑆𝑢 + 𝑆𝑝 𝜙𝑃 ∆𝑉 = 0
𝛿𝑧𝑃𝑇 𝛿𝑧𝐵𝑃
𝑎𝑃 𝜙𝑃 = 𝑎𝑊 𝜙𝑊 + 𝑎𝐸 𝜙𝐸 + 𝑎𝑆 𝜙𝑆 + 𝑎𝑁 𝜙𝑁 + 𝑎𝐵 𝜙𝐵 + 𝑎 𝑇 𝜙 𝑇 + 𝑏
110
 Finite Volume Method:
3-D Steady State Diffusion

111
 Finite Volume Method:
Steady State Diffusion
𝑎𝑃 𝜙𝑃 = ∑𝑎𝑛𝑏 𝜙𝑛𝑏 + 𝑆𝑢
𝑎𝑊 , 𝑎𝐸 in one dimension

𝑎𝑛𝑏 𝑎𝑊 , 𝑎𝐸 , 𝑎𝑆 , 𝑎𝑁 in two dimension

𝑎𝑊 , 𝑎𝐸 , 𝑎𝑆 , 𝑎𝑁 , 𝑎𝐵 , 𝑎 𝑇 in three dimension
𝑎𝑃 = ∑𝑎𝑛𝑏 − 𝑆𝑝

112
 Assignment
• You are provided with a Python code for the solution of
Example 7-2 by finite volume method using central difference
discretization scheme by incorporating the functions of
 Guass-Seidal iterative scheme and
 TDMA line-by-line technique as discussed in class.
• Replace one of the insulted boundaries with mixed boundary
condition with h = 200 W/(m2.°C), and T∞= 20°C and derive
the discretized algebraic equations and values of coefficients.
• Modify the code accordingly to compute the temperatures for
Δx = Δy = 0.1 and Δx = Δy = 0.05
• Also compute the boundary temperature at boundary with
mixed boundary condition.

114
 Assignment
Temperature 100 °C
Consider a non-uniform grid as shown in figure.
The location of control volume faces are given as
follows:

Insulated
0.000

Heat Flux 500 kW/m2

0.000 0.031
0.033 0.075
0.4 m
0.084 0.133
x= 0.150 y = 0.200
0.216 0.267
0.267 0.325
0.300 0.369
0.400 Insulated
0.3 m

Given the boundary conditions and physical data is same as of

Example 7.2, derive the discretized algebraic equations and values of
the coefficients.
Modify the given code accordingly to compute the temperature
distribution within the plate
115
h∞ = 200 W/m2.°C
T∞ = 20°C

T∞ = 20°C h∞ = 200 W/m2.°C

116