equation), which can be written as,

   
∂ ∂T ∂ ∂T
0= k + k +S (1)
∂x ∂x ∂y ∂y

Discretize this equation using FVM (Eq. 4.57 in Versteeg & Malalasekara [1]) and solve using Gauss-Seidel
iteration. MATLAB is recommended for programming.
ME501M: Computational Fluid Dynamics
The computational domain and boundary conditions are as shown in figure 1 and table 1, respectively.
Discuss the results both from physical Lab andSession
numerical pointProblemsof view. Present the results as contour plots
(wherever applicable). The presentation and the report must include the following:
1. Use different meshes to solve the problem (i.e. 10 × 10, 20 × 20, 40 × 40, etc) and show mesh
independence. Stretch and refine the mesh in regions where you expect large gradients and discuss the
solution outcome.
Problem #1: 2-D steady state heat conduction with zero source term
2. Plot residual error (R) vs number of iterations (N ). Increase or decrease the error tolerance () and
discuss the outcome.
Two sessions: 18/2/25 and 25/2/25
The diffusion equation for temperature T (i.e., heat conduction equation), can be written as,
3. What happens if you change the boundary conditions? e.g. change Neumann to Dirichlet or Dirichlet
   
to Neumann on one side. Discuss the results. 0=
∂
k
∂T
+
∂
k
∂T
+ S, (1)
4. In order to illustrate the heat flow, plot ∂x the∂xheat ∂y flux ∂y
vector as a vector plot and discuss.
where x and y are Cartesian coordinates, k is the thermal conductivity, and S is the source term. Assume
the magnitude of k = 1 and S = 0 units respectively. The equation then reduces to the form that was
asked in quiz-1. Consider a rectangular computational
∂T domain∂Tas shown in Figure 1, wherein the length
scale in x-axis can be taken as L = −kthe length
q̇x8t=and , q̇y = −kin y-axis can be taken as H = 3t, here t is a
scale (2)
multiplier (you can select as per your choice, ∂x ∂y with ∆x = 2 and ∆y = 1 units implies
however t = 1 along
the lengths that were asked in quiz-1). Take boundary conditions for temperature (◦ C) as:

Figure 1: Computational Domain

Figure 1: Computational domain.
Boundary 1 Boundary 2 Boundary 3 Boundary 4
T =0 T =0 T = 100 T =0
T1 T2 T3
15 −5y/H
Table + 15cos(2πy/H)
1: Boundary Conditions 10

1. Discretize Equation 1 using FDM and solve using Jacobi/G-S iteration. You can use any language
and H = 0.5; dT /dx = 0 at boundary 4. Constant source term (per area), S = −1.5.
Table 1: L = 1 programming.
Coefficient of conductivity, k = 5(1 + 100x/L)
2. Use different meshes to solve the problem and discuss the solution outcome.
3. Introduce the relaxation parameter and discuss it’s effect on the number of iterations.
4. Change the boundary condition on the left wall to adiabatic, while temperature at top, bottom,
1
and right wall can be taken as 200, 200, and 500 units respectively.

1
Problem #2: 1-D transient heat conduction with zero source term
Two Sessions: 4/3/2025 and 11/3/2025
A wall 1 unit thick and infinite in other directions (see Figure 2) has an initial uniform temperature
(Ti = 100 units). The surface temperatures (Ts ) at two sides are suddenly increased and maintained at
300 units. The wall is composed of nickel steel with a diffusivity of α = 0.1 units. We are interested in
computing distribution within the wall as function of time (say for total time of 0.1 units).

Figure 2: Schematic of physical problem

The governing equation to be solved is the unsteady one-space dimensional heat conduction equation,
which in Cartesian coordinates is,
∂T ∂2T
=α 2. (2)
∂t ∂x
1. Use the following techniques: FTCS, and step sizes to be used are: {∆x = 0.05, ∆t = 0.01} and
{∆x = 0.05, ∆t = 0.05}.

2. Now change the left wall (x = 0) boundary condition to adiabatic and take surface temperature at
the right wall (x = L) as Ts = 500 units. Use FTCS explicit method to solve the problem. take
step size as {∆x = 0.05, ∆t = 0.01}
3. Equation 2 does have the an analytical solution subjected to the imposed initial and boundary
conditions as,
∞
1 − (−1)m
 
X 2
T = Ts + 2(Ti − Ts ) e−(mπ/L) αt sin(mπx/L) (3)
m=1
mπ
This solution can be used to validate the numerical solutions. Use Equation 3 to compare your
numerical solution that you obtained in part 1.
4. To investigate the effect of step size on the accuracy of the solution and required computation time
of an implicit technique, use the (a) BTCS, and (b) Crank-Nicolson method with the following
step sizes:
∆t = 0.01, ∆t = 0.025, ∆t = 0.05, ∆t = 0.1.

2
Problem #3: 1-D diffusion without source term: FVM
Session: 18/3/25
1. Consider the problem of source-free (q = 0) heat conduction in an insulated rod whose ends
are maintained at constant temperatures of 100°C and 500°C respectively. The one-dimensional
problem sketched in Figure 3 is governed by
 
d dT
k +q =0 (4)
dx dx
Calculate the steady state temperature distribution in the rod. Thermal conductivity k = 1000
W/m.K, cross-sectional area A = 0.01 m2 .

Figure 3: Schematic

2. Now we discuss a problem that includes sources other than those arising from boundary conditions.
A large plate of thickness L = 2 cm with constant thermal conductivity k = 0.5 W/m.K and uniform
heat generation q = 1000 kW/m3 . The faces at ends A and B of the plate are at temperatures
of 100°C and 200°C respectively. Assuming that the dimensions in the y- and z-directions are so
large that temperature gradients are significant in the x- direction only, calculate the steady state
temperature distribution. Compare the numerical result with the analytical solution (Equation 5).
 
TB − TA q
T = + (L − x) x + TA (5)
L 2k

Problem #4: 1-D convection-diffusion: FVM

Session: 25/3/25
A property ϕ is transported by means of convection and diffusion through the one-dimensional domain
sketched in Figure 4. The governing equation is Equation 6; the boundary conditions are ϕ0 = 1 at x = 0
and ϕL = 0 at x = L. Using five equally spaced cells and the central differencing scheme for convection
and diffusion, calculate the distribution of ϕ as a function of x for
(i) Case 1: u = 0.1 m/s,
(ii) Case 2: u = 2.5 m/s,
and compare the results with the analytical solution
ϕ − ϕ0 exp(ρux/τ ) − 1
= (6)
ϕL − ϕ0 exp(ρuL/τ ) − 1

(iii) Case 3: recalculate the solution for u = 2.5 m/s with 20 grid nodes and compare the results with
the analytical solution. The following data apply: length L = 1.0 m, ρ = 1.0 kg/m3, τ = 0.1 kg/m.s.

Figure 4: Schematic of problem

3
Problem #5: 2-D transient heat conduction without source term
To be submitted by 30/04/2025
A long, rectangular bar has dimensions of L = 0.3 m by W = 0.4 m. The bar is initially heated to a
temperature of To = 0.0 ◦ C. Subsequently, its surfaces are subjected to the constant temperatures of
T1 = 40.0 ◦ C (bottom wall), T2 = 0.0 ◦ C (left wall), T3 = 10.0 ◦ C (top wall) and T4 = 0.0 ◦ C (right
wall). It is required to compute the transient solution where the governing equation is
 2
∂2T

∂T ∂ T
=α + (7)
∂t ∂x2 ∂y 2

The bar is composed of copper with a thermal conductivity of 380 W/(m◦ C) and a thermal diffusivity
of 11.234 × 10− 5 m2 /s, both assumed constant for this problem. The computational grid is specified by
31×41. Use the FTCS explicit scheme with time steps of 0.2 s and 1.0 s to compute the transient solution.

1. Print the solutions at interval of 0.05 m in the x-direction and all y-locations at t = 10 s, t = 40 s,
and steady state.
2. Compare the steady-state solution obtained in (1) to the analytical solution T = Ta + Tb , where,
∞   
X 1 − cos(mπ) sinh(mπ(W − y)/L)
Ta = T1 ∗ 2 sin(mπx/L) (8)
m=1
mπ sinh(mπW/L)

∞   
X 1 − cos(mπ) sinh(mπy/L)
Tb = T3 ∗ 2 sin(mπx/L) (9)
m=1
mπ sinh(mπW/L)

3. Plot the transient solution for the following locations: (0.1, 0.05), (0.15, 0.10) and (0.1, 0.3).
4. Print and plot the heat transfer distributions along the sides y = 0.0, y = W , and x = 0.0 for the
steady state solution.