Appendix E

CFD Assignments and

Guideline for CFD Project
A sample of three assignments and a guideline for a CFD project clarifying the
aim and objectives are described in this appendix. For students who do not pos-
sess their own project topics, they are most welcomed to select one from the
attached project topics (CFD projects A–C) exemplified herein.

E.1 ASSIGNMENT 1
E.1.1 Background and Aim
The backward-facing step is commonly used as a benchmark for validating
numerous flow characteristics, including flow recirculation and reattachment,
and testing of numerical models and methods. This problem has numerous
applications in industry, such as for HVAC and combustion chamber.
The aims of this problem are as follows:
1. To learn the process of creating and exporting a mesh by using any available
mesh generation software packages. For this assignment, the mesh genera-
tor in ANSYS Workbench is employed.
2. To learn how to set suitable boundary conditions and numerical models
using any available CFD software packages. ANSYS-FLUENT is used to
solve the flow problem in this assignment.
3. To explore the postprocessing facilities of the CFD code to analyse the
numerical results.
4. To formulate concise professional reports.

E.1.2 Problem Description

The student is required to compute laminar flows through a backward-facing
step as detailed below. The coordinates given for the geometry are normalized
against the characteristic length scale. For the case of backward-facing step, the
characteristic length scale is the step height (in this case, a length of unity is

435
436 Appendix E

assumed). The normalized fluid properties at the velocity inlet are given as
follows:
Inlet velocity: ux ¼ 1 and uy ¼ 0
Fluid properties: Density, ρ ¼ 1 and dynamic viscosity, μ ¼ 1/Re, where Re is
the Reynolds number
The outlet boundary is defined as an outflow condition, whilst the no-slip con-
dition is invoked for the rest of the computational walls. Turbulence is ignored,
and no heat transfer exists within the system. It is noted that the dimensions of
the computational domain may need to be altered to ensure that the flow is
sufficiently developed at the outlet boundary (Fig. E.1).

E.1.3 Instructions
1. Initiate ANSYS Workbench to create a mesh for the backward-facing step.
Assign appropriate boundary conditions to the computational domain.
Structured mesh is preferred, but the user may alternatively generate the
geometry with an unstructured mesh. Ensure that proper mesh quality is
achieved. Provide explanations for areas that require further mesh refine-
ment. Export the two-dimensional mesh to ANSYS-FLUENT, Version 17.
2. Using ANSYS-FLUENT, solve the simulation to obtain the velocity and
pressure contours and the velocity vectors for Re ¼ 100. Ensure that the flow
is fully developed (ensuring no flow reversal) or close-to-developed flow at
the outlet boundary. Discuss any observed flow characteristics using the
physical parameters.
3. Repeat steps 2 and 3 for other meshes of varying densities. Use the same flow
settings for Re ¼ 100 to determine the sensitivity of the mesh to the reattach-
ment point of the recirculation zone. Plot a graph relating the mesh size
against the reattachment location, highlighting the most economical mesh
for numerical computations (hint, when grid independence is achieved, the
reattachment point will not vary with increasing mesh density).
4. Using the mesh determined from step 3, perform simulations for other Reyn-
olds numbers of 50, 150, and 200. Compile your results and create a graph
illustrating the relationship between the Reynolds numbers and reattachment
points. Explain the phenomenon and provide your own conclusions.

(−6, 2) (10, 2)

Inlet
Outlet
(0, 0)
(−6, 0)

(0, −1) (10, −1)

FIG. E.1 A schematic illustration of two-dimensional geometry of a backward-facing step.
 Appendix E 437

E.2 ASSIGNMENT 2
E.2.1 Background and Aim
One common CFD application is the study of flows over external structures. In
the automotive industry, it is important to determine the aerodynamic effects of
the spacing between adjacent motor vehicles. The Ahmed model is often used in
experiments as a representation of the motor vehicle due to its simple geometry
and the ease of varying a number of important parameters.
The aims of simulating these models are as follows:
1. To create a CFD simulation of a single Ahmed model and extract
meaningful data
2. To obtain CFD simulations and study the effects of spacing between two
Ahmed models
3. To gain an understanding on the model requirements for turbulent flow and
the importance of the distribution of y+ values
4. To better understand boundary-layer flows
5. To learn how to distinguish and access results with available published
experimental results

E.2.2 Problem Description

E.2.2.1 Single Car Configuration
The geometry of a typical Ahmed configuration is shown below. Students are
required to develop a model simulation of a single two-dimensional Ahmed
configuration model (Fig. E.2).
The coordinates for the Ahmed model are given as follows:

x y
0.94400 0.00000
0.00000 0.00000
1.04400 0.10000
0.00000 0.17700
1.04400 0.18800
0.94400 0.28800
0.19226 0.28800

FIG. E.2 A two-dimensional geometry of an Ahmed model.

438 Appendix E

Note that the characteristic length (taken as the length of the vehicle) of the
Ahmed configuration is not equivalent to unity. You will need to take into con-
sideration when normalizing the Reynolds number and the spacing between
adjacent vehicles.
Simulations are to be performed in air. The outer domain for the single-
model case should be constructed according to the following Cartesian coordi-
nates below, which should allow any wakes and vortices to be properly resolved
within the computational domain:

x y
9.39600 0.05000
9.39600 4.12600
19.83600 0.05000
19.83600 4.12600

At the inlet, the velocity should be set according to the Reynolds number
(with respect to the car length) of 2.3  106. The flow turbulence intensity is
assumed to be 1.8%.

Instructions
1. Students are required to generate another mesh for a second vehicle trailing
the one created above (hint, the domain may need to be purposefully
extended to accommodate the additional vehicle).
2. Appropriate meshing should be employed, preferably similar to above. The
same meshing methods and boundary conditions as used in previous section
should be used herein.
3. Vary the distance between the trailing and leading vehicles. Formulate at
least three additional cases. Discuss the flow characteristics and compare
the drag and lift coefficients for both the lead and the rear car models
(Figs E.3 and E.4).
4. Compare the drag and lift coefficients against experimental results (Figs E.3
and E.4).
5. Discuss the flow physics obtained from the predicted drag and lift
coefficients.

E.2.2.2 Drafting Configuration

Instructions
1. Generate a mesh of sufficient quality (keeping in mind the numerical con-
siderations of aspect ratios and grid skewness). Ensure mesh independence
is reached (hint, use two or three different types of mesh densities).
2. Students must apply the attained knowledge gained from tutorials and the
previous assignment to determine suitable boundary conditions (note that
 Appendix E 439

0.5

0.4
Drag coefficient
0.3

0.2
Single model
0.1 Leading model
Trailing model
0
0 1 2 3 4
Spacing (x/L)
FIG. E.3 Drag coefficient for single Ahmed model and two Ahmed models at different vehicle
spacing. (From Watkins, S., Vino, G., 2004. On vehicle spacing and its effect on drag and lift.
In: Proceedings of the 5th International Colloquium on Bluff Body Aerodynamics and Applications
(BBAA5), Ottawa, Canada.)

0.6

0.5
Lift coefficient

0.4

0.3
Single model
0.2
Leading model
0.1 Trailing model

0
0 1 2 3 4
Spacing (x/L)
FIG. E.4 Lift coefficient for single Ahmed model and two Ahmed models at different vehicle spac-
ing. (From Watkins, S., Vino, G., 2004. On vehicle spacing and its effect on drag and lift. In: Pro-
ceedings of the 5th International Colloquium on Bluff Body Aerodynamics and Applications
(BBAA5), Ottawa, Canada.)

simulations can be performed by allowing the floor to travel in the same

speed as the air).
3. Discuss the importance of y+ values in turbulent flows. Students must
ensure that acceptable values are achieved. If otherwise, provide an
explanation on this.

E.3 ASSIGNMENT 3
E.3.1 Background and Aim
CFD has the ability to model fluid flows coupled with heat transfer. Basic
understanding of the thermal and hydrodynamic behaviour of fluid within a