FINAL INDEPENDENT PROJECT

simulating flow over a body to calculate the drag coefficient of a car body

Introduction

As the technology is developed, more software engineers have produced much simulation software in order
to reduce the time taken for the research and development process and can save up more money. In addition, time is very crucial things in the most manufacturing company because they need to compete with others company to
produce faster and quality product thus Computational Fluid Dynamic (CFD) software is very important in the automobile industry. There are two types of CFD simulation software which are open source type and commercial type. Most researchers are preferred to use a commercial type software such as SolidWorks and ANSYS to study the aerodynamic behaviours of a vehicle but the problem is which one is the best software to study the aerodynamic behaviours is still debateable. The purpose of this project is to compare the results of aerodynamic behaviour between two different CFD software which are SolidWorks 2018 and ANSYS. CFD is a
branch of fluid mechanics that uses numerical methods with the help of computer to solve and analyse problems
involving fluid flows [1]. CFD simulation technology helps engineers to understand the physical phenomena taking place around the design and provides an environment to optimize

THEORY

The drag coefficient is a common measure in automotive design as it pertains to aerodynamics. Drag is a force that acts parallel and in the same direction as the airflow. The drag coefficient of an automobile impacts the way the automobile passes through the surrounding air lower the drag coefficient lower the amount of fuel consumed by the car.
The average modern automobile achieves a drag coefficient of between 0.25 and 0.3. SUVs, with their typically boxy shapes, typically achieve a Cd=0.35–0.45. The drag coefficient of a vehicle is affected by the shape of body of the vehicle.
The reduction of drag in road vehicles has led to increases in the top speed of the vehicle and the vehicle's fuel efficiency, as well as many other performance characteristics, such as handling and acceleration

here is a detailed view of some production cars with their drag coefficients

the drag force can be expressed as:

F_d = c_d 1/2 ρ v² A (1)

where

F_d = drag force (N)

c_d = drag coefficient

ρ = density of fluid (1.2 kg/m³ for air at NTP)

v = flow velocity (m/s)

A = characteristic frontal area of the body (m^2)

PROJECT DETAILS :

in this project flow over a shell eco marathon protype catagory was studies in 2 sofwares ie ansys and solidworks to calculate the drag coefficient and determmine the error percentage in the calculated results

CAD BODY USED :

the following body was designed by me in solidworks for the shell eco marathon asia 2020 and shell eco marathon 2019

METHORD OF AERODYNAMIC DESIGN:

To design a highly efficient vehicle to move towards an energy-efficient economy, empower people, and spread awareness and consequently innovate towards a better future.

The design was straightforward and given a teardrop shape since a teardrop has the lowest drag
coefficient. In our previous vehicles, there was no wheel covering which was a disadvantage to our fuel
efficiency. With the ameliorate design, we gave wheel coverings to provide a superior aerodynamic
shape to our vehicle. The frontal area of the vehicle was designed to be as low as possible.

Simulation Set-Up:

A SYMMETRY MODEL IS preffered for drag calculations since it computes results more accurately and quickly with more number of elments but a non symmetric approch was choosen for this research

In order to stimulate the air flow around the vehicle, a fluid volume needs to be created. This was done by creating
a fluid enclosure around the model and subtracting the model. This fluid enclosure will act as the wind tunnel for the
simulation. This fluid enclosure is created only for ANSYS FLUENT. skewness of less than 0.86 was achieved

an infation layer was added with a mesh refinement near the body of the vehicle the domain sizes were kept same for both solidworks and ansys fluent size of elements used is 10mm an 15 inflation layers were used to capture the boundry effect near the body

Inlet 40 m/s
Outlet 0 Pa
Wall No slip wall
Fluid Air at 25°C
Analysis type Steady state solver
Turbulence model k-epsilon model

the following mesh type was used in solidworks mesh was kept as refined as possible this was decided by a mesh independency test to achieve this mesh structure and once no change in results was obtained on refining the mesh in solidworks )

mesh indipendency condition 1:converged at 97 iteration mesh according to solidworks with no changes net force force 91 N Cd=0.1767

mesh indipendency condition 2:converged at 120 iteration mesh according to solidworks with no changes net force force 79 N Cd=0.1543

mesh indipendency condition 3:converged at 132 iteration mesh according to solidworks with no changes net force force 72N Cd=0.137

mesh indipendency condition 4:converged at 147 iteration mesh according to solidworks with no changes net force force 71.6603 N Cd=0.137

mesh independency reached

Computational Domain

Size

X min	-3.576 m
X max	5.102 m
Y min	-0.318 m
Y max	1.751 m
Z min	-1.958 m
Z max	1.989 m

Ambient Conditions

Thermodynamic parameters	Static Pressure: 101325.00 Pa Temperature: 293.20 K
Velocity parameters	Velocity vector Velocity in X direction: 40.000 m/s Velocity in Y direction: 0 m/s Velocity in Z direction: 0 m/s
Turbulence parameters

Material Settings

Fluids

Air

Goals

Global Goals

GG Normal Force (X) 1

Type	Global Goal
Goal type	Normal Force (X)
Coordinate system	Global coordinate system
Use in convergence	On

.1 Pre-processing
In this pre-processing, the meshing is one of the significant different the software. The meshing element detail
for the software is shown in the Table 3. The number of elements for both software is different because the
SolidWorks does not have a wall to act as a wind tunnel, it only uses computational domain compare to ANSYS which use solid box with a specific dimension to act as wind tunnel and also a computational domain for the simulation, thus the same accurate and exact size for the computational cannot be done. Besides, ANSYS 18.2 managed to produce high number of elements and nodes because it can easily use the manual mesh setup for the simulation compared to the SolidWorks manual mesh setup, therfore i used the automatic setup with the highest level of the initial mesh to match with the ANSYS mesh setup.

RESULTS:

ansys

pressure contour in ansys

the final cd in ansys was converged at 0.155 after 246 iteration

SOLIDWORKS:

velocity contour solidworks

pressure contour analysis in solidworks

pathhline anlysis for smooth flow of air partices used during modelling analysis

value of force exerted on surface of body is 71.6603 N (mesh independent result used )

MESH USED

Nx=180

NY=120

Nz=140

calculation of drag coefficient using formula (1)

Cd=0.137

- Coefficient of drag and lift for SolidWorks and ANSYS.

SOFTWARE USED	DRAG COEFFICIENT
SolidWorks	0.137
ANSYS FLUENT	0.155

Percentage difference 11.61 percent

CONCLUSION

A Computational Fluid Dynamic (CFD) analysis was carried out by using SolidWorks and ANSYS CFX on a
PROTOTYPE SHELL ECO MARATHON VEHICLE DESIGNED BY ME to analyse the aerodynamics behaviour for that car. The results of the simulation such as drag coefficient for both software were calculated, the percentage
different between the results for drag coefficient is 11.61%.Based on the high percentage different between SolidWorks and ANSYS CFX drag coefficient value, it can be concluded that between SolidWorks and ANSYS FLUENT <FLUENT is the relevant choice in CFD simulation to study the aerodynamic behaviours of the car. The software
used can provide good results if the details are set specifically. However, based on this study, ANSYS fluent
provides a better result to study the aerodynamic behaviours due to its various parameter settings and its flexibility
compared to SolidWorks 2018.

SUGGESTIONS

• use of a bigger computational domain
• use of symmetry plane analysis
• residuals in ansys fell till 1e-3 further meshing can improve the solution to give even more accurate results.