Académique Documents
Professionnel Documents
Culture Documents
P ROJECT
23-1
Title:
GROUP
NUMBER :
Student Names
and IDs:
Supervisors
names:
Clint Steele
Date
of
Submission:
31/10/14
10,500
DECLARATION
We declare that ( the first four boxes must be completed for the assignment to be accepted):
This assignment does not contain any material that has previously been submitted for assessment at this or any
other university.
This is an original piece of work and no part has been completed by any other student than those signed below.
We
have
read
and
understood
the
avoiding
plagiarism
guidelines
at
We have retained a copy of this assignment in the event of it becoming lost or damaged.
(optional) We agree to a copy of the assignment being retained as an exemplar for future students (subject to
identifying details being removed).
Student signatures:
Date: 31/10/14
Declaration
This report contains no material which has been accepted for the award of any other
degree or diploma in any university or tertiary establishment, and to the best of my
knowledge and belief, contains no material previously published or written by another
person except where due reference has been made.
Where a report has been professionally edited, the student must include separate
acknowledgement that such
editing has been and the editing has addressed only the style and/or grammar of the
report and not its substantive content."
ABSTRACT
The idea of downforce inducing devices for race cars have been present almost since the
inception of the race car, they have been developed due to significant research since then.
Recent developments such as the double diffuser; have been left unstudied in recent literature
however the system has been used and proven successful in Formula 1. Although it has been
success in Formula 1 it is unknown if this device would be viable for FSAE because of the speed
differences in both formulas.
The aim of this study is to examine unchartered downforce inducing devices or areas of
downforce creation left un-studied for use in a FSAE platform. The double diffuser was chosen
as it was an area that was left un-researched by previous literature; and also the high potential
for success in FSAE given the achievement of such promising results in Formula 1.
A scientific approach has been utilised in this study to unearth gaps in literature and then based
off these previous gaps in literature experiments are then conducted to gain numerical data (in
the form of CFD) on the effectiveness of the double diffuser model in comparison to a regular
diffuser model (as the last method is currently employed by most FSAE teams).
This study has found that there was a significant increase in performance for the double
diffuser over the regular diffuser. The double diffuser method appears to present as a viable
method for increasing performance for FSAE vehicles or similar. As such further research should
be focused on the optimisation of the double diffuser method for use in FSAE.
ACKNOWLEDGEMENTS
The Author wishes to acknowledge Team Swinburne Formula-SAE for support with the
resources required to fully comprehend and conduct this research project, providing the
catalyst for the described system. The author wishes to acknowledge the assistance of Dr Clint
Steele for his supervision of this research task and as advisor for the 2014 Formula-SAE race car
Project. The following research would not have been possible without the support of Swinburne
University providing the resources for the resulting research.
ii
CONTENTS
Abstract .......................................................................................................................................................... i
Acknowledgements....................................................................................................................................... ii
Table of Figures ............................................................................................................................................ iv
1.0 Introduction and problem statement ..................................................................................................... 7
2.0 Literature Review .................................................................................................................................... 8
2.1 Race car & aerodynamics fundamentals ............................................................................................ 8
2.10 Why downforce............................................................................................................................. 8
2.11 Downforce................................................................................................................................... 11
2.12 Bernoullis Principle .................................................................................................................... 13
2.13 The under-tray ............................................................................................................................ 14
2.14 Ground effect .............................................................................................................................. 15
2.15 Lift induced drag ......................................................................................................................... 18
2.16 The diffuser ................................................................................................................................. 19
2.17 Centre of pressure ...................................................................................................................... 20
2.2 prior art ............................................................................................................................................. 21
2.20 Gurney flaps ................................................................................................................................ 24
2.21 Aero foils ..................................................................................................................................... 25
2.3 recent developments ........................................................................................................................ 27
2.30 DRS - drag reduction system ....................................................................................................... 27
2.31 Vortex generators ....................................................................................................................... 28
2.33 Double deck diffusers ................................................................................................................. 30
2.4 Simulation ......................................................................................................................................... 32
3.0 Research question & objectives ............................................................................................................ 35
4.0 Methodology......................................................................................................................................... 35
4.1 Design modeller set up ..................................................................................................................... 38
4.2 Meshing............................................................................................................................................. 39
4.3 running simulation ............................................................................................................................ 43
5.0 Results and discussion .......................................................................................................................... 45
6.0 Conclusion ............................................................................................................................................. 57
7.0 Appendix ............................................................................................................................................... 58
Down-force generation for a Formula-SAE vehicle
iii
TABLE OF FIGURES
Figure 1: Tyre adheasion and relationship to aerodynamic down force (Katz, 2006) .................................. 8
Figure 2: Aerodynamic force balance (McBeth, 2006) ................................................................................. 9
Figure 3: Frontal Area (McBeth, 2006) ....................................................................................................... 10
Figure 4: different shapes have different drag coefficients (McBeth, 2006).............................................. 10
Figure 5: modern saloon car profile (McBeth, 2006) ................................................................................. 11
Figure 6: Wing profile (McBeth, 2006)........................................................................................................ 11
Figure 7: front WING OF 1993 indy CAR (J.KATZ, 1995) ............................................................................. 12
Figure 8: rear Wing of 1986 zakspeed F-1 car (j.katz, 1995)....................................................................... 12
Figure 9: Typical CL and Cd of various vehicle shapes (Katz, 1995) ............................................................ 12
Figure 10: Venturi effect illustration, (Roberts, 2014) ................................................................................ 13
Figure 11: Typical race car under-tray (Seljak, 2008).................................................................................. 14
Figure 12: Typical underbody tunnels, shown generating Venturi Vortices, (Li & Vallejo Paz, 2000)........ 15
Figure 13: Effect of a single element Wing in ground effect (Katz, 2006) .................................................. 15
Figure 14: NORMAL VELOCITY PROFILE OF BOUNDARY LAYER, (Katz, 1995) ............................................. 16
Figure 15: Velocity profile of boundary layer with moving top plane shows couette flow , (Katz, 1995) . 16
Figure 16: Vs with laminar and turbulent flow), (Katz, 1995) ........................................................ 17
Figure 17: 2D DIFFUSER GEOMETRY, (Peddie & Gonzalez, 2009) .............................................................. 19
Figure 18: Maximum cornering speed with reference to lift coefficient (Katz, 1995) ............................... 20
Figure 19: Pressure distribution of under-tray showing pressure distribution along the x-axis, (Katz,
1995) ........................................................................................................................................................... 20
Figure 20: SKIRT USED TO SEAL THE UNDERTRAY TO THE ROAD, (KATZ, 1995) ......................................... 21
Figure 21: Some different ways of generating down-force (Van Valkenburgh, 2000) ............................... 23
Figure 22: Gunnery flap effect on flow (Haney, 2000) ............................................................................... 24
Figure 23: Effect of adding a 1.7% chordlong Gurney flap on the lift and drag coefficient of a rectangular
wing (AR = 8, NLF 0414 airfoil). AR, aspect ratio. (Data from (Myose, et al., 1996) ................................... 24
Figure 24: Multiple element rear wing (Katz, 2006) ................................................................................... 25
Down-force generation for a Formula-SAE vehicle
iv
Figure 25: Rear wing height effect on aerodynamic performance (Katz, 2006) ......................................... 25
Figure 26: Aerodynamic wing element with and with out end plates (Katz, 2006) ................................... 26
Figure 27: DRS system used on the UTA FSAE car (Merkel & Woods, 2014).............................................. 27
Figure 28: Vortex generator (KOIKE, et al., 2004) ....................................................................................... 28
Figure 29: Various shapes of vortex generators (KOIKE, et al., 2004) ........................................................ 28
Figure 30: Vortex generator effect (Katz, 2006) ......................................................................................... 29
Figure 31: Year 2000 indicar under-tray vortex generators (Katz, 2006) ................................................... 29
Figure 32: Toyota Triple deck diffuser, (Formula1.com, 2009)................................................................... 30
Figure 33: Double deck diffusser entry and exit points (F1Tecnichal.net, 2011) ....................................... 30
Figure 34: Note the large secondary diffuser outlets on this Renult R30 F1 car (F1Technical.net) ........... 31
Figure 35: Various methods for replicating real world conditions for a race car (Katz, 1995) ................... 32
Figure 36: EFFECT OF INCOMING FLOW VELOCITY PROFILE WITH AND WITH OUT A MOVING FLOOR
(HUMINIC & HUMINIC, 2009) ..................................................................................................................... 33
Figure 37: small control volume ................................................................................................................. 33
Figure 38: Body geometry ........................................................................................................................... 38
Figure 39: Ahmed body (Ahmed, 1984) ...................................................................................................... 38
Figure 40: Enclose details table, and symmetry plane selected, result shown in Figure 42 ...................... 39
Figure 41: Symmetry used to decrease calculation time ............................................................................ 39
Figure 42: Default Mesh.............................................................................................................................. 40
Figure 43: Modified sizing ........................................................................................................................... 40
Figure 44: Face sizing selection and resultant mesh .................................................................................. 41
Figure 45: Wake & road body of influence ................................................................................................. 41
Figure 46: Final mesh .................................................................................................................................. 42
Figure 47: Double diffuser meshing ............................................................................................................ 43
Figure 48: Velocity contours basic bulff body ............................................................................................. 45
Figure 49: Velocity streamlines ................................................................................................................... 46
Figure 50: pressure plot .............................................................................................................................. 46
Figure 51: Pressure coefficient over body about symmetry axis................................................................ 47
Figure 52: Pressure contours 10 degree double diffuser ........................................................................... 48
Figure 53: Velocity contours 10 Degree Double diffuer ............................................................................. 48
Figure 54: Velocity contours 7 Degree double diffuser .............................................................................. 49
Figure 55: Pressure contours 7 Degree double diffuser ............................................................................. 49
Figure 56: Velocity CONTOURS 5 DEGREE DOUBLE DIFFUSER ................................................................... 50
Figure 57: PRESSURE CONTOURS 5 DEGREE DOUBLE DIFFUSER ................................................................ 50
Figure 58: Velocity contours 2 Degree Double diffuser .............................................................................. 51
Figure 59: Pressure CONTOURS 2 DEGREE DOUBLE DIFFUSER .................................................................. 51
Figure 60: Lift force Vs Vehicle velocity ...................................................................................................... 52
Figure 61: CL Vs Vehicle velocity ................................................................................................................. 55
Figure 62: CD Vs Vehicle velocity ................................................................................................................ 55
Figure 63: L/D Ration Vs Double diffuser angle .......................................................................................... 56
vi
McBeth (2006) explains aerodynamic force balance as the balance of forces in the horizontal
plane where the tractive force is the force of the engine pushing the vehicle forward and the
drag force is the force produced in the opposite direction by the car pushing the air out of its
way. This is better explained by using Newtons second law of physics (Equation 2) which says
that an object will only accelerate if there is a net force unbalance. Equation 1 shows how
friction and drag forces are subtracted from tractive forces resulting in forward acceleration by
way of Equation 2 below. The vertical forces shown in the diagram represents the weight of the
car pushing down on the ground through the tyres, the upward force is the reaction force of
the tyres on the road pushing up. These forces cancel each other out resulting in a zero net
force in the vertical direction.
The illustration in Figure 2 of force balance in the horizontal plane explains this more clearly.
Equation 1 -
Equation 2 -
The lower the drag force, the larger the net force forward will be and the faster the vehicle will
travel, which is why it is advantageous to reduce drag.
Referring back to Equation 2 shows that a reduction in weight of the aero package will lead to
less force required to accelerate the car and hence a higher top speed.
Therefore weight needs to be taken into consideration when designing an aerodynamic
package for a race car. As the FSAE competition is for electric vehicles, drag and weight will
also have an effect on battery life because unnecessary drag and weight will induce extra force
required to power the vehicle.
The Calculation of drag force is represented by Equation 3 (Aird, 1997) and it also shows why
reducing frontal area is of importance and displays the effect CD has on the drag force which
will be explained further.
Equation 3 -
= 2
Where
=
= (2 )
= ()
Frontal area is the cross sectional area that the object must move through, Figure 3 represents
this as the frontal view of the car.
10
2.11 DOWNFORCE
Down-force is the negative lift created by the car as it travels through the air. A car is generally
shaped like a wing seen in Figure 5, and produces positive lift meaning that the car is lifted
lightly from the road. In a race car this needs to be minimised or if possible designed to create
negative lift, a race car will often be designed to be as low as possible to the road surface in
order to reduce this wing like profile and in turn reduce positive lift.
Coefficient of drag and coefficient of lift are important quantities to the race car, again
is a dimensionless number used to represent the lift generated by the object. These numbers
translate into actual drag and lift values in metric SI using Equation 4 & Equation 5 (Katz, 1995).
Equation 4 -
= 1
2
Equation 5 -
= 1
2
11
On a race car this concept is used to create down-force by using inverted wings on the fronts or
rear of the car as seen in Figure 8 & Figure 7.
12
Equation 6 -
1
2
2 + + =
Where
=
=
9.81
= ( 2 )
=
=
13
The under-tray on modern Formula 1 cars or champ cars still utilise the ground effect principle.
The modern under-tray usually encompasses 3 components. These components include the
inlet which is an opening at the front of the car between the ground, this allows air to enter the
under body. The next component is the restriction component where the air flow area is
reduced from the inlet and is sped up, and hence a reduced static pressure is achieved. The
final component is the diffuser which returns the air back to atmospheric pressure and ideally
with very slight losses. Something that must be noted is that there is a limit to the amount of
flow that can be passed through an under-tray this is a critical point where the low pressure will
reverse and create a dramatic high pressure (Campbell, 1973) due to the high force in the inlet.
14
An additional feature of the modern under-tray is the underbody tunnels. Underbody tunnels
shown in Figure 12, is a section in the under-tray where vortices are produced to furthermore
reduce the pressure under the car and hence increase down-force (Li & Vallejo Paz, 2000).
FIGURE 12: TYPICAL UNDERBODY TUNNELS, SHOWN GENERATING VENTURI VORTICES, (LI & VALLEJO PAZ, 2000)
These vortices are basically air spinning at high speed (Tremayne, 2004) that according to
Bernoullis principle decrease pressure and in turn raise the total down-force of the car. More
information on this vortex system is available in Appendix 1.
When an object is present in the area of a moving ground plane flow asymmetry occurs
Down-force generation for a Formula-SAE vehicle
15
between the underbody which accelerates flow due to the ground limitation (Zerihan & Zhang,
2000), this profile can be seen in Figure 14 and this is different for that of a regular velocity
profile Figure 15.
In Figure 14 the top plate (or underbody) is moving to the right and the lower is the road. The
particles close to the surfaces tend to stick to the surface and you get a non-slip condition
which is why the velocity on the ground is zero whist the velocity on the plate is equal to the
velocity of the car. As the two profiles differ it makes simulation difficult in wind tunnels, this is
where a rolling road or other measures are required to get accurate results (Tremayne, 2004).
This phenomenon causes fluid shear and hence force at high speeds as represented by
Equation 7 (Force being drag or Skin friction( )) (Katz, 1995).
Equation 7 -
Where
=
=
=
=
Boundary layer plays a big role in race cars; the thicker the boundary layer the more viscous
sheer and hence more drag. When at high speeds however the boundary layer is reduced,
Down-force generation for a Formula-SAE vehicle
16
allowing more air flow the results are increased down-force and less drag as compared to how
the above model predicts (Katz, 1995), Laminar and turbulent flow also affect the boundary
layer, as such the boundary layer is thicker with turbulent flow (Katz, 1995), see Figure 16.
17
Equation 8 -
= 2
2
Where
=
=
=
=
=
Equation 9 -
Where
= ( 0.85 0.95 )
=
Equation 8 (Clancy, 1975) displays that in order to gain some lift there must be some amount of
drag penalty for any lift inducing aerodynamic body. Equation 9 shows that the can only be
zero if the is zero. This confirms that there is no such thing as free down-force as some may
consider when it comes to race cars. Particularly the under-tray or wings are affected by this;
they all induce some form of drag but some more than others as the value determines the
amount of drag that will be induced. Table 1 indicates the typical drag associated with various
sections of the car.
TABLE 1: TYPICAL DRAG ASSOSIATED WITH VARIOUS SECTIONS OF THE CAR, (KATZ, 1995)
Skin friction
Cooling drag
Internal flow
Parasitic drag
Lift induced drag
0.04-0.05
0.00-0.06
0.00-0.05
0.00-0.45
0.00-0.60
18
Diffuser pumping is a quality of the diffuser where the diffuser acts like a pump to produce
more flow through the under-tray; this scavenging effect creates a low pressure area under the
car and decreases boundary layer through increased velocity (Peddie & Gonzalez, 2009).
Another performance parameter of the diffuser is the pressure recovery coefficient as seen
below (Peddie & Gonzalez, 2009).This quantity is the recovery of static pressure of flow through
a diffuser (Blevins, 1984).
EQUATION 10 -
Where:
1 =
2 =
=
=
Down-force generation for a Formula-SAE vehicle
19
The centre of pressure of an under-tray usually lies somewhere in the centre of the car (Refer
to Figure 19), this means there should be no problem getting a centre of pressure about 40-60%
rearward bias as was a design consideration for the TS14.
FIGURE 19: PRESSURE DISTRIBUTION OF UNDER-TRAY SHOWING PRESSURE DISTRIBUTION ALONG THE X-AXIS, (KATZ, 1995)
20
FIGURE 20: SKIRT USED TO SEAL THE UNDERTRAY TO THE ROAD, (KATZ, 1995)
Vacuums have been used in the past for example the Chaparral 2J CAN AM series car of 1970
and the F1 Brabham BT46 of 1978. The Chaparral 2J used 2 snow mobile engines to create a
vacuum under the car to suck it to the ground; this meant constant down-force irrespective of
the cars speed, which meant that it was able to take low speed corners much faster than the
conventional competing race car.
It also used flexible Lexan skirts to seal the underbody for even more down force. Another
advantage to the fan car is that the fan blows air into the separated flow area behind the car
thus reducing drag. 8 year later this concept was used for the 1978 Formula one world
championship with the Brabham BT46 but it used the driving engine to power the fan.
It worked so well that it won the first round of the 1978 season (Swedish GP) in the hands of
Niki Lauder by approximately 30 seconds. The instant success raised strong protest among
other teams and it was immediately banned (Katz, 1995).
These technologies worked very well and they were mostly banned because of their resounding
effectiveness so while they would be a viable option they cannot be used as it conflicts with the
FSAE rules (SAE International, 2014)
Down-force generation for a Formula-SAE vehicle
21
This concept would not work on an electric FSAE car for 2 other reasons; the first being that the
rules prohibit this kind of system Refer below.
-T9.4 Ground Effect Devices
No power device may be used to move or remove air from under the vehicle except fans
designed exclusively for cooling. Power ground effects are prohibited. (2014 Formula SAE Rules,
SAE International)
The second reason is that it would require too much power to spin the fan and create the
vacuum, which would be detrimental to an electric car, thus another option must be found.
Another system for generating down-force is using the open-tailed box design which is sealed
and relies on the natural negative wake pressure from the rear (Van Valkenburgh, 2000).
22
FIGURE 21: SOME DIFFERENT WAYS OF GENERATING DOWN-FORCE (VAN VALKENBURGH, 2000)
23
24
Katz (2006) shows the effect that rear wing height has on coefficient of lift in Figure 25.
There have been studies conducted on front wing elements as well and the effect of end plates
has shown increases in negative lift due to reduced pressure loss, which can be seen in Figure
26.
25
FIGURE 26: AERODYNAMIC WING ELEMENT WITH AND WITH OUT END PLATES (KATZ, 2006)
In an experiment publicised by Katz (2006) it was shown that front wings have a large effect on
the under-tray and that only a few degrees of attack angle can be the difference between
diverting flow from the under-tray, it also showed that wing tip vortices from the front wing can
influence the rear wing.
In conclusion it can be seen that there has been allot of research on the wing and gunnery flaps
over the years and as such any further research should be focused elsewhere.
26
FIGURE 27: DRS SYSTEM USED ON THE UTA FSAE CAR (MERKEL & WOODS, 2014)
Merkel & Woods (2014) Suggest that their system provides the following difference when the
system is on and off:
Closed 4.27
1.56
Open
1.78
0.30
This is quite a dramatic difference and allows the added benefit of increased braking forces by
using the system in a braking situation.
27
28
Figure 30 shows how the reduction in ground clearance can increase the effect of the vortex
generator when integrated into an under-tray which shows quite a significant improvement.
There is a point where too little ground clearance will reduce this effect, probably due to a
break down in the vortices. These vortices increase the suction under the car and actually
increase it further by the two vortices entangling and become closer to the vehicles surface
(Katz, 2006).
FIGURE 31: YEAR 2000 INDICAR UNDER-TRAY VORTEX GENERATORS (KATZ, 2006)
Vortex generators can be seen on the circa 2000 Indy car in Figure 31. This effect is also
implemented into the design of underbody tunnels as seen in section 2.3 where the vortices are
trapped in the tunnel.
Down-force generation for a Formula-SAE vehicle
29
DIFFUSSER
ENTRY
AND
EXIT
POINTS
30
The double diffuser can be seen in many different designs, some with quite large openings
(F1Technical.net) and some with smaller openings, there has even been a triple deck diffuser
(formula1.com, 2009).
FIGURE 34: NOTE THE LARGE SECONDARY DIFFUSER OUTLETS ON THIS RENULT R30 F1 CAR (F1TECHNICAL.NET)
31
2.4 SIMULATION
Simulating designs are a complex process; Katz (1995) outlines three ways of conducting flow
analysis tests which are:
Wind tunnel
CFD (Computational Fluid dynamics), and finally
Real world tests
Running tests on a real car on a race
track induces some complexities when
you start trying to measure lift and
drag etc. It becomes very difficult
because of vibration from suspension
and varying ambient conditions; but
ultimately reduced lap times are the
goal so a real world test is the
ultimate vindication. However you
must also understand where change
needs to me made which is where
measurement comes into view. When
developing a vehicle the car obviously
doesnt exist this makes it impossible
to do real road tests, so that
eliminates this method for any
research.
most commonly regarded as the best method is the moving ground plane, where the car
runs on a belt at the same velocity as the incoming air flow; the velocity profile can be seen
in Figure 36 for both conditions, the other advantage to the moving ground plane is that it
takes into account the effect that rotating wheels may have on the simulation (Refer to
Appendix 2) .
FIGURE 36: EFFECT OF INCOMING FLOW VELOCITY PROFILE WITH AND WITH OUT A MOVING FLOOR (HUMINIC & HUMINIC, 2009)
As the facilities and time are not accessible for this research, CFD is the chosen method for any
experiments that this research may require.
With CFD it is easy to make any changes to design and it will cost allot less than any wind tunnel
experiments, and the time taken to make changes or run a test will be much faster usually.
CFD has the advantage of it being very effective at displaying a comprehensive depiction of the
flow field as well which could give more information on the flow patterns.
CFD relies on the input of some factors such as boundary conditions like flow velocity,
turbulence, temperature ect.
Most CFD packages calculate flow problems by using the body
geometry and discretely calculating, the body being tested is
converted into small elements like tetrahedrals (Swinburne
University of Technology, 2014), so the domain is split up into
grids where many small calculations can be conducted. Three
methods for discretely calculating flow problems are used
including: FEM (Finite element method), FDM (Finite
Difference method), FVM (Finite volume method) where FVM
is the most widely used.
Some terminology often associated with CFD are the
governing equations of fluid mechanics that are used in CFD,
like the Navier stokes equations which are used to calculate
velocity, pressure and time(in transient cases) in 3
dimensions. The following equations are used to solve for u, v,
w, p in x, y, z & t.
33
Conservation of mass:
Conservation of Momentum:
Together these 4 equations can solve the 4 unknowns and are known as the Navier stokes
equations.
Turbulence is also an aspect that needs to be considered in CFD and three common approaches
to modeling turbulent flow are
DNS (Direct numerical simulation) which involves setting a grid and time steps fine
enough to resolve the features of the turbulent flow. DNS is unrealistic for most flows
because the required computational time is too large (Crowe, 2010).
Large eddy simulation (LES) involves direct simulation of the large-scale eddies in the
turbulence and approximate simulation of the smaller eddies (Crowe, 2010).
The k-epsilon model (k-e model) models turbulence by introducing two extra equations.
As compared to DNS and LES, the k-e model is computationally efficient (Crowe, 2010).
34
4.0 METHODOLOGY
The methodology that this research will employ is the scientific method which will include
quantitative collection of data. Sanford (1899) shows that there are 4 elements of the scientific
research method:
1. In collecting carefully authenticated acts as the basis of all generalization;
2. In looking for some common causal relation between these facts, which relation is
stated in the form of a general proposition, or a so-called law of nature;
3. In deducing by the methods of both formal logic and mathematical conclusions
concerning other phenomena which have not yet been observed;
4. In experimenting to see if these conclusions are correct
Here the first is the literature review, which has looked at tried means and unearthed
voids in the literature that necessitate inquiry.
In the 2nd the gap in the literature reveals a hypothesis in this case the lack of research
on the Double diffuser.
In the 3rd a method for testing such a hypothesis is constructed and
In the 4th an experiment is conduced to prove or otherwise the hypothesis
35
After examining the various techniques in which downforce can be produced and their
advantages/disadvantages; it is clear that there are gaps in the literature regarding the Double
diffuser as the only recorded literature is from sports news or Formula one web sites. There are
no academic studies of this system and there are no studies of the system in a FSAE application.
Aerodynamics is a very well-studied area and has been studied for some time but as a lot of this
research comes from the aviation industry or from Formula 1, there seems to be very little
documented on the more recent and proven designs; this is most likely because these areas are
kept quiet by the companies that develop them as to gain a completive advantage on the track
or in the air. One such design is the Double deck diffuser, very little is documented on this
system, how it works or why it is used but what is known (at least in a Formula one application)
that it works, so it is the aim of the author to evaluate and better understand the applicability
of the double diffuser for increasing performance in an FSAE vehicle.
To make an educated decision on whether a double diffuser is viable for a FSAE car, we must
first conduct an experiment on the system to compare it to a regular diffuser system.
This will be done by conducting a CFD study; the tests will be conducted through Ansys Fluent
simulation software. The decision has been made to choose CFD over a real wind tunnel test as
the author does not have time nor resources to build a model and test facilities are not
accessible. There is access to a miniature test facility at Swinburne University but a test model
of relatively similar size to the FSAE car is needed to replicate a real world situation and a rolling
road is also required if boundary layer effects are to be accurately replicated. As such CFD will
suit as the most practical method to examine the Double diffuser in a FSAE application.
A simplified model of a vehicle body will be used to investigate this system which will
encompass geometry similar to that of a real FSAE car (both in rough geometry and under-tray
area). The reason for this is that computational power readily available to the author is limited
and in order to simulate conditions accurately a detailed mesh is required which would be
difficult with the hardware available. Another reason for choosing a simplified model is that
there could be an infinite amount of designs of FSAE car so testing on one particular type would
not mean that it would work on another design (or work as well), so by removing other factors
that may influence the result a generalisation can be made for all vehicles of similar geometry
and application.
By utilising a simple model, speed of iteration will be increased and accuracy will be increased
(McAlpine, 2004) by allowing greater mesh density, flow caused by external objects such as
wheels and wings will be removed which could influence wake flow from the diffuser which will
increase turbulence in that area and require longer iteration time. The increase or decrease in
performance can then be stated to be increased or decreased by an amount found by this study
on a real FSAE car taking into consideration that there could be other influences on a real world
application such as the intricacies of the FSAE car geometry such as wheels/wings etc.
CFD results can be almost just as accurate as wind tunnel testing with correct set up and will
likely be more useful because of the visualisation (McAlpine, 2004).
Down-force generation for a Formula-SAE vehicle
36
4 different angle diffusers will be tested 10, 7, 5 and 2. The model will use an under-tray ride
height of 35mm and diffuser angle of 10 on all tests; this should be somewhere close to the
optimum as it is very close to the findings of Amal Seba Thomas T, 2014 (2014). The model
dimensions used on all tests (apart from Double diffuser angle) can be viewed in Appendix 4
and Appendix 5.
After conducting these experiments the results will then be analysed and then a
recommendation made. It is the opinion of the author that there will need to be significant
increases in downforce of at least 2% or 3% for further research to be worthwhile; a judgment
will be made based on the increase in downforce and the extra weight that will be needed to
facilitate the system and any other issues that arise. This will only be a recommendation and
will ultimately be upon future FSAE designers to interpret these results, expand upon them and
make a decision although this study should show roughly the type of benefits/deficits this
device should give for a FSAE race car.
It must be noted that these results will not be verified with wind tunnel testing or any other
method because of time and resource constraints.
Based on what is known about the double diffuser where in Formula one a double diffuser is
worth about a 5% increase in downforce, it is the opinion of the author that there is possibility
for a significant increase in downforce for the current or other FSAE cars using a regular undertray. If an increase in drag is an offset of this device then a mechanism to close the secondary
diffuser could be used.
37
After the model was imported into Fluent a computational domain needed to be constructed.
The following section covers the meshing.
38
4.2 MESHING
The domain needs to be 3 car lengths in front of
the body, 3 to the sides and 5 from the rear of
the body as suggested by Lanfrit (2005). To
simulate the road the enclosure is 35mm from
the bottom of the car.
To simplify and speed up calculation, a symmetry
wall in the longitudinal direction was used to
reduce the number of cells and hence iterations
as suggested by Robert Carrese (2014).
FIGURE 40: ENCLOSE DETAILS TABLE, AND SYMMETRY PLANE SELECTED,
RESULT SHOWN IN FIGURE 41
This default mesh (Figure 42) is very rough and the elements are quite large, in order to
increase accuracy the mesh needs to be refined. This is done by altering the default selections
for meshing to that shown in Figure 43.
Some sizing functions need to be set to make sure that boundary layer effects are predicted
effectively, this is done by right clicking on the mesh icon and selecting all the surfaces of the
body.
39
40
Lanfrit (2005) for flow near wall boundaries, the first aspect ratio is set at 5 which again was
recommended by Lanfrit (2005) for vehicle flow testing. The mesh sizing used in Fluent can be
viewed in Figure 43.
The surfaces of the body need to be selected as to include these faces in the program
controlled inflation, these faces will be set according to Appendix 3; the resulting mesh can be
seen in Figure 44, note the Prismatic Layers (Lanfrit, 2005) (how the mesh follows the body
surface) .
FIGURE 44: FACE SIZING SELECTION AND RESULTANT MESH
As a means to increase accuracy and predict boundary layer flow of the body and particularly
the under tray and wake, some extra bodies will be generated to use as bodies of influence
Figure 45. The image to the left shows a body of influence that will be used to set the boundary
layer of the under-tray and the wake of the vehicle to a size similar to the surface mesh; the
image to the right shows another body of influence that will be set to a slightly lesser value, the
body of influence is not mandatory although it is suggested by Lanfrit (2005) . The final mesh
can be seen in Figure 46 and a near completed mesh of the double diffuser in Figure 47.
41
42
43
Viscous
model
K-epsilon
model
Near wall
treatment
wall
function
110
Moving
wall with
No slip
condition
Coupled
Initialisation
Hybird
44
Figure 48 shows the Velocity contours of the basic bluff body; the high velocity can be seen
under the body, as would be expected due to the venturi effect that takes place here; the
maximum velocity is at the beginning of the body as flow increases in velocity to fill the void
created by the diffuser. The minimum velocity can be seen at the stagnation point at the front
of the body and at the wake of the car where vortex shedding occurs; this can be illustrated
better by a vector diagram in Figure 49.
45
Figure 50 shows the pressure plot over the bodys surface with maximum pressure at the front
(stagnation point) of the body and lower pressure on the under-tray. The pressure over the
body about the symmetry plane can be seen in Figure 51 where the stagnation point is
represented by a pressure coefficient of 1 at position -2m; traveling left the pressure is
represented over the top surface at a pressure coefficient of 0 and then as the line goes under
the diffuser the pressure becomes a minimum at the diffuser exit point where is returns to the
stagnation point.
46
47
48
Note in comparison of the 10 Double diffuser (Figure 53) velocity graph to the 7Double
diffuser (Figure 54) velocity graph that the separation zone is larger in the 10 model, this
results in more drag and less downforce for the 10 Double diffuser.
49
50
51
1000
0
0
50
100
150
200
250
300
350
400
-1000
-3000
-4000
-5000
52
Figure 60 shows the lift and drag of the all the models (simple bluff body, 10 Double diffuser,
the 7 Double diffuser, 5 Double diffuser, 2 Double diffuser.) against vehicle velocity. The lift
force increases with increasing vehicle velocity as does drag force, albeit much less than the lift.
The graph shows that for a vehicle of similar geometry to a FSAE car; there is an increase in
downforce with a double diffuser, where the maximum increase was with a 5 Double diffuser.
Table 3 shows the results in numerical form, here it can be seen that the actual increase in
downforce (Lift) for a 10 degree Double diffuser is 1.27-4.28N over the range that a FSAE car
would be running at(0-100km/h), this also shows for a F1 application about 18N are achievable
at 275km/h. This also shows a dip in performance past 275km/h, this may indicate that more
flow doesnt mean more performance as the downforce seems to decrease or level out after
275km/h in this test.
As the Double diffuser angle is decreased to 7 more performance I gained, with an increase of
4.1-17.1N over the FSAE car velocity range; A deficit of roughly 1.5-5.7N (approximately 9.5%)
in drag over the same velocity range.
At 5 maximum performance is achieved, the actual increase in downforce is a whopping 25
odd percent over the simple model for all ranges of velocity, this translates to a 21.74N increase
in lift at 50 km/h and a 92.92N increase at 100km/h; although there is an increase in drag of
about 5% over the total velocity range.
At 5 Double diffuser angle the performance starts to recede at 2 Double diffuser angle.
53
TABLE 3: RESULTS
Simple model
Velocity(kmh)
Drag Force
Lift Force
CD
CL
Lift/Drag
0
0
0
-
10
50
14.95
-67.99
0.398
-0.376
1.27
1.87
1.15
7.72
4.547
100
250
275
300
350
59.71
371.65
450.34
537.94
730.42
-277.15 -1767.18 -2137.60 -2529.93 -3454.55
0.398
0.396
0.396
0.398
0.397
-0.383
-0.391
-0.391
-0.388
-0.390
4.28
29.15
27.88
18.85
28.08
1.54
1.65
1.30
0.75
0.81
4.40
29.53
35.62
45.56
61.37
7.37
7.95
7.91
8.47
8.40 Ave Lift/Drag
4.642
4.755
4.747
4.703
4.730
4.687
7
50
15.30
-70.82
0.407
-0.391
4.10
5.79
1.50
9.80
4.629
100
250
275
300
350
60.88
378.10
457.82
544.32
740.50
-289.98 -1876.88 -2274.00 -2712.42 -3706.54
0.405
0.403
0.403
0.403
0.402
-0.401
-0.415
-0.415
-0.416
-0.418
17.10
138.85
164.28
201.35
280.07
5.90
7.40
7.22
7.42
7.56
5.57
35.98
43.10
51.94
71.45
9.15
9.52
9.41
9.54
9.65 Ave Lift/Drag
4.763
4.964
4.967
4.983
5.005
4.885
5
50
14.64
-88.46
0.390
-0.489
21.74
24.58
0.84
5.74
6.042
100
250
275
300
350
58.32
358.90
435.45
515.62
700.78
-365.80 -2301.52 -2811.43 -3322.40 -4542.66
0.388
0.382
0.383
0.381
0.381
-0.505
-0.509
-0.514
-0.510
-0.512
92.92
563.49
701.71
811.33 1116.19
25.40
24.48
24.96
24.42
24.57
3.01
16.78
20.73
23.24
31.73
5.16
4.68
4.76
4.51
4.53 Ave Lift/Drag
6.272
6.413
6.456
6.444
6.482
6.352
2
50
13.90
-86.4
0.370
-0.478
88.27
97.84
12.63
90.87
6.216
100
250
275
300
350
54.94
339.12
411.32
487.16
661.70
-356.18 -2290.20 -2821.43 -3311.74 -4522.34
0.366
0.361
0.362
0.360
0.360
-0.492
-0.506
-0.515
-0.508
-0.510
357.72 2291.85 2822.73 3312.49 4523.15
99.57
99.93
99.95
99.98
99.98
50.66
309.97
383.44
468.31
633.62
92.21
91.41
93.22
96.13
95.76 Ave Lift/Drag
6.483
6.753
6.859
6.798
6.834
6.657
0.3178
1.5314
54
CL Vs Vehicle velocity
0.000
0
100
200
300
400
-0.100
Simple model
-0.200
-0.400
-0.500
-0.600
Figure 61 shows the coefficient of lift ( ) for each model tested; it shows how even with the 10
Double diffuser there is an increase in downforce. Figure 62 shows how drag is less for the 5 Double
diffuser than the 7 or 10 Double diffusers, this is because the 7 & 10 Double diffusers are inducing
flow separation, a comparison of the pressure contours of the 10 & the 5 Double diffuser ducts shows
a large separation region on the lower surface of the Double diffuser duct.
CD Vs Vehicle velocity
0.410
0.400
0.390
Simple model
Double diffuser 10 Deg
CD 0.380
0.370
0.360
0.350
0
100
200
300
400
55
L/D
2.000
1.000
0.000
0
10
12
Figure 63 shows the Lift to Drag ratio change over the range of Double diffuser angles, it shows
that 5 gives the optimum Double diffuser angle in relation to L/D ratio and maximum lift
however performance could be increase further if a DRS system was employed to shut off the
Double diffuser inlets at high speed, this would save on about 3N at 100km/h although the
mechanical complexity involved in such a system may outweigh the minor improvement in top
speed. This graph also shows the point where the lessening of the Double diffuser angle results
in negative performance where the minimum is 2 Double diffuser angle.
ERRORS
Some error could be present in this research such as, the geometry of the body is such that the
secondary diffuser ducts are pointing upward which could influence the results. The flow
simulation may not be 100% accurate; Wind tunnel testing would be required to verify this.
56
6.0 CONCLUSION
The purpose of this study was to examine downforce inducing devices for use in a FSAE platform, this
was done by investigating previous mechanisms and unearthing gaps in literature that have been
neglected or un-researched. This study has shown that the use of the double diffuser in a FSAE platform
has the promise to significantly increase performance over the regular diffuser system.
Data for the optimum angle was also found; although this may vary for different geometry vehicles. The
drag and weight penalties for using such a system would be very little; as it would not be a structural
component. Thin carbon fibre could be utilised reducing the weight penalty to a minimum.
These findings will allow FSAE vehicles to produce more downforce and hence higher cornering speeds.
This research will also allow others to conduct more in depth research on the optimum design for the
Double diffuser to achieve a greater understanding for use in FSAE.
Wind tunnel testing of an actual FSAE vehicle would further vindicate this research as would actual on
track FSAE vehicle testing and integration of the system in a FSAE car. This system has proved to have
promise to increase FSAE performance as such it is the view of the author that more research should be
invested in to gain a better understanding of the Double diffuser in relation to placement, exit direction,
inlet direction, Double diffuser area and flow properties. More research could yield even higher
performance than this research has extracted.
57
7.0 APPENDIX
APPENDIX 7.1
Vortex lift used in aircraft wings is where vortices are produced on the convex section of the wing in an
effort to reduce pressure and hence lift; when the vortex is spinning in the correct direction it can also
reattach flow and reduce commencement of wing stall (utexas, 2014). A graphical representation can be
seen below.
One study investigates the flow characteristics of vortices in under body applications for vehicle with a
diffuser, they find that the vortices increase downforce but break down at low ride heights ( Mahon , et
al.).
General vortex behaviour can be characterized by the following equation (Steenbergen, 2004)
=
Where
=
= ,
=
58
APPENDIX 7.2
APPENDIX 2: WHEEL AERODYNAMICS
Wheel aerodynamics is an area that is mostly overlooked when aerodynamic design is studied,
this is because the primary function of the wheel is not aerodynamic and they are not intended
to increase the aerodynamic efficiency of the car (Zhang, et al., 2006). The wheel is essential for
the car and they usually have a fixed shape with poor aerodynamic attributes, and as such are
not a good area to spend time and money. The wheel is also very difficult to study
experimentally as it requires the wheel to be spinning in ground effect ie. road moving at air
speed , this being said they are very important to the open wheel racecar and contribute 40% of
total drag, they also create lift and effect other components of the car such as diffuser and
wings (Zhang, et al., 2006). There has been much study of the wheel over the years Fackrell and
Harveys method for pressure measurement over the surface of the wheel has stood mostly
unchanged for 30 years; they used pressure sensors inside the wheel in order to measure this
without interfering with the flow of the wheel.
59
APPENDIX 7.3
60
APPENDIX 7.4
61
APPENDIX 7.5
62
APPENDIX 7.6
Work Schedule
Writing report
introduction
Literature Review
Methadology
Experimental proceedure
Discussion
Conclusion
References
Proofreading
6/11/2014
30/10/2014
23/10/2014
16/10/2014
9/10/2014
2/10/2014
25/09/2014
18/09/2014
11/09/2014
4/09/2014
28/08/2014
21/08/2014
14/08/2014
7/08/2014
31/07/2014
24/07/2014
17/07/2014
10/07/2014
3/07/2014
26/06/2014
19/06/2014
12/06/2014
5/06/2014
29/05/2014
63
Down-force generation for a Formula-SAE vehicle
22/05/2014
15/05/2014
8/05/2014
1/05/2014
24/04/2014
17/04/2014
10/04/2014
3/04/2014
27/03/2014
20/03/2014
Research
8.0 REFERENCES
Mahon , S., Zhang, X. & Gage, C., n.d. The Evolution of Edge Vortices Underneath a Diffuser Equipped
Bluff Body, Southampton: s.n.
Ahmed, G. R., 1984. Some Salient Features of the Time-Averaged Ground Vehicle Wake, SAE-Paper
840300, s.l.: SAE.
Aird, F., 1997. Aerodynamics for racing and performance cars. 1 ed. 375 Hudson Street New York: HP
Books.
Amal Seba Thomas T, V. T. M. &., 2014. EFFECT OF DIFFUSER HEIGHT AND ANGLE ON PERFORMANCE OF
FSAE CAR. nternational Journal of Advanced Engineering and Global Technology, p. 1008.
Anderson, G., 2012. Red Bulls secrets revealed. [Online]
Available at: http://www.bbc.com/sport/0/formula1/20118852
[Accessed 20 May 2014].
Ansys, Inc., 2012. Inlet Boundary Conditions. s.l.:Ansys, Inc..
Benson, T., 2010. Lift to Drag ratio. [Online]
Available at: https://www.grc.nasa.gov/www/k-12/airplane/ldrat.html
[Accessed 14 May 2014].
Blevins, R. D., 1984. Applied Fluid Dynamics Handbook. GA Technologies, Inc.; California, University; San
Diego State University, San Diego, CA: Van Nostrand Reinhold Co..
Campbell, C., 1973. Design of racing sports cars. 1 ed. Massachusetts: Robert Bently, INC..
Clancy, L., 1975. Aerodynamics. London: Pitman Publishing Limited, London.
Clayton T, C., n.d. Engineering Fluid Mechanics. 10 ed. s.l.:Wiley.
Crowe, C., 2010. Engineering Fluid Mechanics. 9 ed. s.l.:John Wiley & Sons.
Evans, P., 2009. Formula One Double Deck Diffuser explained. [Online]
Available at: http://www.gizmag.com/formula-one-double-deck-diffuser/11260/
[Accessed 19 May 2014].
F1Technical.net, n.d. F1Technical.net. [Online]
Available at: http://i.kinja-img.com/gawker-media/image/upload/s--4wfiY_jH-/c_fit,fl_progressive,q_80,w_636/18ky5iuwmxsg2jpg.jpg
[Accessed 21 10 2014].
64
65
Li, Y. & Vallejo Paz, H., 2000. Downforce without wings vehicle aerodynamics. Department of Thermo
and Fluid Dynamics.
McAlpine, J., 2004. Computational Fluid Dynamics or Wind Tunnel Modeling. [Online]
Available at: http://www.envirometrics.com/abstracts/index.html
[Accessed 20 10 2014].
McBeth, S., 2006. Competition Car Aerodynamics. 1 ed. 861 Lawrence Drive, Newbury Park California
USA: Haynes North America Inc..
Merkel, J. P. & Woods, D. B., 2014. Four-Quadrant, Fully Active Aero. Race Car Engineering Magazine, 1.
Michael , W. & Isaac, S., n.d. Handbook In Research and Evaluation. s.l.:s.n.
Molina, M. J. T., 2013. The Global Scientific Method. 2 ed. s.l.:Publidisa.
Myose, R., Papadakis, M. & Heron, I., 1996. The effect of Gurney flaps on three-dimensional wings with
and without taper. SAE Technical Paper 965514.
Peddie, K. & Gonzalez, L., 2009. CFD Study on The Diffuser of A Formula 3 Racecar. Orbit: The University
of Sydney Undergraduate Research Journal, pp. 19, 20, 30, 31, 32..
Peddie, K. & Gonzalez, L., 2009. CFD Study on The Diffuser of A Formula 3 Racecar. The University of
Sydney Undergraduate Research Journal, pp. 19, 20, 30, 31, 32.
Race Car Engineering Magazine , 2012. Formula 1 2012 Round 9: British Grand Prix. [Online]
Available at: http://www.racecar-engineering.com/articles/f1/formula-1-2012-round-9-british-grandprix/
[Accessed 9 10 2014].
Race Car Engineering, 2009. Race Car Engineering Magazine. [Online]
Available at: http://www.racecar-engineering.com/technology-explained/diffusers-engineering-basicsaerodynamics/
[Accessed 10 11 2014].
Robert Carrese, L. A. S. W. (. U., 2014. Performing automotive aerodynamic simulations using ANSYS
CFD, s.l.: s.n.
SAE International, 2014. 2014 Formula SAE Rules. s.l.:s.n.
Sanford, F., 1899. The scientific method and its limitations. s.l.:Stanford University Press.
SAS IP, Inc, 2011. Num Cells Across Gap. [Online]
Available at: https://www.sharcnet.ca/Software/Fluent14/help/wb_msh/msh_Num_Cells_Gap_SF.html
[Accessed 25 10 2014].
Seljak, G., 2008. Race Car Aerodynamics, s.l.: s.n.
Down-force generation for a Formula-SAE vehicle
66
Smith, C., 1978. Tune to win. 1 ed. Fallbrook, CA: Aero Publishers.
Society of Automotive Engineers, 2014. 2014 Formula SAE Rules. s.l.:s.n.
Steenbergen, C. v., 2004. Vortices in favorable pressure gradients, Assessment of Vortex behavior in
Formula 1 Underbody conditions. p. 15.
Swinburne University of Technology, 2014. Introduction to CFD. Melbourne: Swinburne University of
Technology.
Tremayne, D., 2004. Science of Formula 1 Design. 1 ed. Sparkford, Yeovil, Somerset: Hynes Publishing.
utexas, 2014. Aerodynamics: Increasing Lift Coefficient. [Online]
Available at: http://courses.ae.utexas.edu/ase463q/design_pages/fall02/AWE_web/aero.htm
[Accessed 21 10 2014].
Van Valkenburgh, P., 2000. Race Car Engineering & Mechanics. 1 ed. Seal Beach, Ca: HP Books.
Zerihan, J. & Zhang, X., 2000. Aerodynamics of a single wing in ground effect. Journal of Aircraft,
(University of Southhampton, Southhampton England), 6(37), p. 1064.
Zhang, X., Toet, W. & Zerihan, J., 2006. Ground Effect Aerodynamics of Race Cars. Applied mechanics
review, p. 59.
67