MECE E4430 Automotive Dynamics

Professor Josh Browne

Fall 2018

FSAE Aerodynamics Project

Design Report

Aaron Thompson

Columbia University
Department of Mechanical Engineering

December 2018
Contents
1 Introduction 3

2 Theory 3
2.1 Bernoulli’s Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2 Lift and Downforce . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.1 Coefficient of Lift . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.3 Drag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.3.1 Top Speed Effects . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.4 Airfoils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.5 Angle of Attack and Stall . . . . . . . . . . . . . . . . . . . . . . . . 6
2.6 Multielement Airfoils . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.7 Ground Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.8 Grip and Tires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

3 Design 10
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2 Body . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2.1 Sidepods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.3 Wings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.4 Undertray . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

4 Optimization 12
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.2 2D Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.2.1 Front Wing Optimization . . . . . . . . . . . . . . . . . . . . 13
4.2.2 Rear Wing Optimization . . . . . . . . . . . . . . . . . . . . . 14
4.3 3D Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.4 Lap Time Simulator . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4.4.1 Skidpad . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.5 Points Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

5 Manufacturing 17
5.1 Molds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5.2 Mounting to Chassis . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5.2.1 Rear Wing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5.2.2 Front Wing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5.3 Assembling Parts of Assemblies . . . . . . . . . . . . . . . . . . . . . 17

6 Conclusion 17

7 Appendix 19
7.1 2D Simulation Mesh Parameters . . . . . . . . . . . . . . . . . . . . . 19
7.2 Front Wing 2D Simulations . . . . . . . . . . . . . . . . . . . . . . . 19
7.3 Rear Wing 2D Simulations . . . . . . . . . . . . . . . . . . . . . . . . 24

1
8 References 29

2
1 Introduction
In Formula SAE (FSAE), aerodynamic elements fall in three basic categories: front
wing, rear wing, and undertray. The rule-set restricts the location, size and design
of many components. This aerokit is designed to generate maximum downforce.
After designing these components, testing is imported into ANSYS for Computational
Fluid Dynamics (CFD) in 2 dimensions for individual components, and then into 3
dimensions for overall car effects.
Optimization does not stop at lift and drag forces. These values are then im-
ported into a lap time simulator, which converts the downforce gains into perfor-
mance improvements at competition. From the event performances, a points analysis
is performed to complete the justification of implementing aerodynamics.
Finally, once the aerokit is fully designed, optimized, and justified, preparation for
manufacturing begins. For this endeavour, the team hopes to work with Fibreworks
Composites to have the system professionally built. This necessitates understanding
exactly how each component will be built, assembled, and mounted.

2 Theory
2.1 Bernoulli’s Principle
The Bernoulli equation is an important part of fluid dynamics. Its equation is below:
1
P1 + ρv12 + ρgh1 = constant (1)
2
In this equation:
P = pressure
ρ = fluid density
v = fluid velocity
g = acceleration due to gravity
h = height
(Bernoulli Equation, n.d.)

Use of this specific equation to determine specific pressure, velocity, or height val-
ues for a flow is restricted by the following assumptions:
1. Velocity across the cross-section of the flow is constant.
2. There is no viscous drag in the fluid.
3. There are no energy losses in the flow.
4. If flow travels around a bend, the centrifugal force is not negligible.

Practically speaking, none of these assumptions are reasonable for this project. What
the equation does, however, is develop the fact that flow velocity and pressure
are inversely related. Thus, when generating negative lift, creating pressure dif-
ferentials using flow velocity is important.

3
2.2 Lift and Downforce
Lift is a force applied to a body travelling through a fluid, in the vertical direction
perpendicular to the object’s direction of travel. In the case of an airplane or car,
this force would act along the same axis as the weight.
Lift is generated due to the changing of the momentum of the fluid flowing around
the object. “The flow is turned in one direction, and the lift is generated in the
opposite direction, according to Newton’s Third Law of action and reaction” (Lift
Force ,n.d.). In the case of an airfoil, both sides of the object do the turning. The
equation of lift is given below:
1
L = cL ρAv 2 (2)
2
In this equation:
L = Lift force
cL = Coefficient of lift
ρ = Density of the fluid
A = Area of the object passing through the fluid
v = Velocity of the object passing through the fluid
(Lift Equation, n.d.)

2.2.1 Coefficient of Lift

The coefficient of lift is a nondimensional quantity that influences the lift force applied
to an object in a flow field. Generally determined experimentally, it simplifies all of
the complexities of an object in motion, including surface roughness, object shape,
and flow properties.
This value serves as a great reference point to compare the aerodynamic perfor-
mance of two elements, without having to control for other factors like area or flow
velocity.

2.3 Drag
Drag is a force applied to an body travelling through a fluid, opposing the direction
of travel of that object. For example, a car travelling through air has a drag force
opposing the acceleration of the car. This is a force that is detrimental to car perfor-
mance, reducing both top speed and acceleration of a racecar. The equation of drag
is given below:
1
D = cD ρAv 2 (3)
2
In this equation:
D = Lift force
cD = Coefficient of drag
ρ = Density of the fluid
A = Area of the object passing through the fluid

4
 v = Velocity of the object passing through the fluid
(Drag Equation, n.d.)

2.3.1 Top Speed Effects

Drag has its most prevalent effects on the top speed of the car. For reference, in
the FSAE competition, the cars almost never reach their top speed, and most have
speeds limited by gearing as opposed to the engine being drag limited. Nonetheless,
it is helpful to consider how the increase of drag affects the top speed of a car.
Knowing the peak power of an engine, we can then set up a free body diagram of
a car accelerating. Looking, the only force opposing the acceleration of the car is the
drag force.

Figure 1: Freebody diagram of a driving car. (Forces and Diagrams, n.d.)

The first fact to know is that:

P = Fv (4)
This equation states that power, P , is equal to force, F , times velocity, v. Knowing
the drag force equation, we can then substitute this into the equation:
1
P = cD ρAv 3 (5)
2
With this equation, one can determine the terminal velocity of the car, assuming
it is drag limited. An extension of this is understanding that the drag increases
substantially compared to velocity, thus accelerating becomes increasingly challenging
as speeds increase.
In the case of this car, knowing that the peak power is approximately 60 horse-
power (44742 watts), the frontal area is approximately 1.205 m2 , the CD is 1.029
kg
(listed later in the Optimization section), and air density is 1.225 3 , we can calcu-
m
late top speed.

5
 1
∗ 1.029 ∗ 1.225 ∗ v 3 = 44742 (6)
2
This then simplifies to find:
Vmax = 39.91m/s = 87.04mph

2.4 Airfoils
An airfoil is a 2 dimensional cross section of a wing. This cross section is intended to
generate lift when travelling through a flow field. Airplane wings’ cross sections are
specially designed to maximize lift within certain angles of attack and flow parameters.
When considering racecar design, one would like to implement these wings, but
inverted. Logically, implementing inverted wings generates downforce instead of lift,
which has been established as being desirable for racecars. An example of a common
airfoil is included below:

Figure 2: Diagram of an airfoil. (Airfoil, 2018)

2.5 Angle of Attack and Stall

The angle of attack, α, is the angle between free stream and chord line. Lift and drag
are functions of the angle of attack. Based on the angle of attack, different pressures
will exist on either side of the airfoil. The illustration below demonstrates the effects
on an airfoil:

6
 Figure 3: An airfoil travelling through a flow field. (Angle of Attack, n.d.)

When an airfoil is increased past its critical angle of attack, it reaches a point called
stall. At this point, the coefficient of lift drastically decreases and consequently the
lift force drastically falls as well. This is caused by substantial flow separation. Many
accidents involving airplanes come down to an over-aggressive angle of attack. The
figure below demonstrates the relationship between lift, drag, and angle of attack in
a moving airfoil:

Figure 4: An airfoil undergoing stall. (Inclination Angle Effects, n.d.)

When designing the front and rear wings, it is critical to maximize the part’s lift
without passing the stall point. This is checked further in the Optimization section
of the paper, where real cases of this are found for the front and rear wings.

2.6 Multielement Airfoils

Multielement airfoils are the next frontier of aerodynamic performance for racecars.
In their simplest form, these are multiple airfoils arranged about one another such that
the overall angle of attack of these components can be maximized. Using multielement
airfoils, the angle of attack of some elements can be increased as much as 5 times
that of other airfoils, without running into stall effects, which allows the generation
of greater downforce.

7
 This is done using slot gaps. The first element has an angle of attack of approx-
imately 0. Essentially, each element’s angle of attack follows the curvature of the
previous element. Thus, the air flow from the tip of the previous foil energizes the
flow beneath the following airfoil, which prevents the propagation of flow separation.
An example of this in action on the front wing is shown below:

Figure 5: A multielement airfoil in action.

As one can see, the multiple elements allow airflow to come from the top of the
previous element to stave off flow separation.

2.7 Ground Effect

The ground effect is the increased lift generated by wings when close to the ground.
As a rule of thumb, the closer a wing is to the ground, the greater lift and drag it
generates. The chart below demonstrates:

8
Figure 6: The change in aerodynamic lift and drag with respect to height. (Katz, J.,
2006)

2.8 Grip and Tires

All grip performance eventually passes through the tires. The most basic trend that
has been applied in this test is that with additional normal forces through the tires,
there is a proportional increase (within the linear range) in the amount of lateral force
that the tires can accept before slipping. The chart below demonstrates a trend line
generated from tests:

Figure 7: The relationship between Fz and Fy, as well as slip angle. (Metz, D. L.,
Milliken, W. F., & Milliken, D. L., 1998)

9
 By implementing aerodynamics, this project intends to maximize the normal force
through the tires, giving greater cornering loads and consequently greater cornering
speeds.

3 Design
3.1 Introduction
Designing the aerodynamics is an integrated process, including designing the body
alongside aerodynamic elements. The undertray attaches directly to the body, and
the front has to interface with the body to be bolted to the chassis. Because of these
things, designing all of the components together is critical for success. Solidworks
was used to model each part.
The most important thing to understand, however, is the fact that this project
does not aim to be the most intense or perfect design possible. Instead, it is focused
on building a system that provably improves performance of the car, and can be built
within the confines of time restrictions for the team. If the initial design goals of
65lbs of downforce and improved performance on track were met, then the design
was considered reasonable.

3.2 Body
The design of the body was focused on a few key factors. Namely, the body was
expected to be as close to the chassis as possible, was expected to clear the front
springs and dampers, and finally was expected include sidepods.
To complete this, a chassis assembly with critical parts was put in place, and sur-
face modeling was done to clear all components. The body was designed to minimize
frontal area where possible, while maintaining rules compliance. The rules in place
included:
T3.23.1 Sharp edges on the forward facing bodywork or other protruding com-
ponents are prohibited.
T3.23.2 All forward facing edges on the bodywork that could impact people, e.g.
the nose, must have forward facing radii of at least 38 mm (1.5 inches). This minimum
radius must extend to at least forty-five degrees (45) relative to the forward direction,
along the top, sides and bottom of all affected edges
(FSAE Rules, 2017-18)

3.2.1 Sidepods
The sidepods are intended to shuttle air to the radiator to improve its efficiency. The
downside of this is that it is aerodynamically inefficient. Thus, it was sized to be just
large enough to fit the worst-case-scenario radiator and no larger.

10
3.3 Wings
The airfoil chosen for this project is the Eppler 423. It is known for its very high
lift. To develop this choice, multiple options for wings were found by researching
examples of high lift airfoils. After developing a list of potential options, these were
implemented into a basic 2D simulation. Looking at the results, and then comparing
them to previous years’ tests, the E423 appeared to be the optimal choice. It showed
the least gains in drag compared to increases in angle of attack, and was the same
airfoil used in previous years’ designs.
An equally important choice was to use three element airfoils in front and rear.
This choice was initially made based on researching other teams’ designs and looking
for consistent trends: Three elements seemed to be a common theme. Had the stated
downforce goals not been reached using the three element setup, then changes in
number of elements would be implemented. More elements have increasing gains, but
diminishing returns.
The design was intended to fit within a bounding box, as prescribed by the rules
below:
T9.3.1 In plan view, any part of any aerodynamic device, wing, undertray or
splitter must not be:
a. Further rearward than 250 mm (9.8 inches) rearward of the rear of the rear
tires
b. Further forward than a vertical plane through the rearmost portion of the front
face of the driver head restraint support, excluding any padding, set (if adjustable)
in its fully rearward position (excluding undertrays).
c. Wider than the inside of the rear tires, measured at the height of the hub
centerline.
T9.3.2 In side elevation, no part of the rear wing or aerodynamic device (including
end-plates) must be higher than 1.2 meters above the ground when measured without
a driver in the vehicle
T9.5.1 All forward facing wing edges including wings, end plates, Gurney flaps,
wicker bills and undertrays that could contact a pedestrian must have a minimum
radius of 5 mm (0.2 inches) for all horizontal edges and 3mm (0.12 inches) for vertical
edges (end plates). These radius requirements must be achieved with permanently
affixed components and with specific design intent to meet this radius requirement.
(FSAE Rules, 2017-18)
Looking forward, the airfoil choice represents the greatest area for development.
Experimenting further with different airfoils, and more or fewer elements could prove
fruitful.

3.4 Undertray
The development of the undertray was also focused on remaining rules compliant,
and pushing as far as the bounding boxes of the rules allowed. Analyzing research
from other teams, optimal diffuser angle was between 15 and 25 degrees. The design
focused on making a part that would fit within the regulations, fit the car, and avoid

11
any interference with other parts in the rear end assembly. An important change for
this season is that the width extended to the width of the outside of the rear wheels.
The diffuser angle was a “guess and optimize” process. As a whole, the initial
angle of 25 degrees was found to be effective and capable in reaching the original
downforce goals. The overall part was designed considering the rule below:
T9.4.1 Between the centerlines of the front and rear wheel axles, an aerodynamic
device (e.g. undertray) may extend outboard in plan view to a line drawn connecting
the outer surfaces of the front and rear tires at the height of the wheel centers T9.4.2
Except as permitted under T9.3.1, any aerodynamic devices, or other bodywork,
located between the transverse vertical planes positioned at the front and rear axle
centerlines must not exceed a height of 500 mm (19.7 inches) above the ground when
measured without a driver in the vehicle. (Bodywork within vertical fore and aft
planes set at 400 mm (15.75 inches) outboard from the centerline on each side of the
vehicle is excluded from this requirement).
(FSAE Rules, 2017-18)

Figure 8: The design of the car in Solidworks.

4 Optimization
4.1 Introduction
After designing the components in Solidworks in individual assemblies, it is necessary
to optimize the components, including chord length, slot gap width, end plate size

12
and shape, and diffuser angle.
ANSYS Fluent was used to determine overall car effects of each component. Using
3D simulations is highly computationally intensive, and prone to random meshing
errors that slowed progress. To compensate for this, 2D CFD was performed on the
front and rear wing profiles to nearly optimize them. These 2D simulations are quick
to setup, run, and review. The wings were then imported into the 3D assembly in
a near-optimized form. Finally, once full 3D simulations were performed on the full
car, the lift, drag and weight numbers were imported into a lap time simulator which
facilitated a points analysis to justify the implementation of the aerodynamics system.

4.2 2D Simulation
To perform a 2D CFD simulation, the front and rear wing profile cross sections were
imported into ANSYS Fluent. A dense mesh was generated surrounding the airfoils,
which then generated a flow field around the wings.
In the Appendix, images of each iteration on the front and rear wing are included,
as well as the chosen mesh parameters for quality performance. 2D simulations will
serve to be a critical tool for testing and trying different types of airfoils for future
designs.

4.2.1 Front Wing Optimization

Front Wing
Trials Cd Cl L/D Notes
1 0.103 -0.911 -8.845 Free Air
2 0.101 -0.916 -9.069 Free Air
3 0.038 -1.102 -29.000 Free Air
4 0.0336 -1.14 -33.929 Free Air
5 0.0832 -1.25 -15.024 Moving Ground, same orientation as T4

Trial 1: The first and second airfoil suffered from flow separation. The flow coming
from top of the first airfoil to the second appears to be very choked, and the second
foil appears to be too far back compared to the trailing edge of the first airfoil. To
remedy this, the first airfoil’s AoA (angle of attack) was reduced to 0, the second
airfoil was raised up compared to the trailing edge of the first airfoil, and the second
airfoil was moved further forward.
Trial 2: The first airfoil’s flow separation was fixed. There was more flow attach-
ment to the bottom surface of the second airfoil, but it still separated about a third
of the way of the chord. The third airfoil still has attached flow. To remedy these
issues, the second airfoil’s AoA was reduced, down to approximately 22 degrees. The
third airfoil was held in place as a control.
Trial 3: The first and second airfoil’s flows remained fully attached along the full
flow surface. The L/D reflects this improvement. The third airfoil’s flow separates
about halfway along the bottom surface. To remedy this, the third airfoil’s AoA was

13
reduced to approximately 38 degrees, and lowered compared to the trailing edge of
the second airfoil.
Trial 4: The flow remained fully attached for each of the airfoils. There is a bit
of trailing edge separation, but this was considered acceptable as gurney flaps will be
implemented in the future as an underclassman project. The improvement in L/D
reflected these improvements.
Trial 5: The test remained the same as trial 4. The only change was implementing
a moving ground at 40MPH at the bottom. This showed a few important changes.
The actual value of −CL increased, which is in line with the understanding of the
ground effect. There was slight flow separation, and an increase in drag. This agrees
with our expectation, since ground effect also increases drag. This was not an issue,
and still allowed us to reach downforce goals.

4.2.2 Rear Wing Optimization

Rear Wing
Trials Cd Cl L/D Notes
1 1.21 -4.23 -3.496 Flow separation on 2nd airfoil
2 7.82 -40.44 -5.171 Flow separation on 3rd airfoil
3 0.5 -6.75 -13.5 No notable flow separation

Trial 1: The first airfoil AoA was at 0 degrees, and showed no flow separation. The
second airfoil, however, had flow separation about 60% of the way along the bottom
edge. The third airfoil had fully attached flow. To solve these issues, the slot gap was
increased from first to second airfoil. The third airfoil was held in place.
Trial 2: The flow remained fully attached for all three airfoils. The pressure
differential, however, for the third airfoil certainly could be larger. To solve this
issue, the third airfoil was moved forward, towards the leading edge of the first and
second airfoil. Additionally, the AoA was increased to follow the trailing edge of the
second airfoil closer. One important thing to note is how large the CL and CD are
with respect to expected reality. They were considered to be off by some scaling
factor, as the L/D remained reasonable. This was a meshing error.
Trial 3: The final test showed fully attached flow for all three airfoils. There was
a substantial pressure differential for all three airfoils, and showed a large L/D. From
this point, both the front and rear wings were imported into the full car ANSYS
simulation.

4.3 3D Simulation
Once the front and rear wings were brought to be nearly optimized, they were imple-
mented into the full-car analysis in ANSYS Fluent. Working with Charles Johnson,
whose accompanying paper has more information regarding the details of the 3D sim-
ulations, the car’s aerodynamic performance was tested. This paper will not delve

14
deeply into the specifics and settings for performing 3D simulations. Instead, this
paper will discuss the decisions and goals of these simulations.
For performing simulations on the full car, mesh settings were determined for
optimal results. This was tuned using a “dummy car” and a sphere. Both of these
items had known aerodynamic performance results, including coefficients of lift and
drag. Once the simulation was able to reproduce these results within 5%, the simu-
lation was usable for the full car. Iterations were focused on building up from just
the car body up to the full aerokit. This would allow the visualization of how new
components affected the performance of the car. Thus, the simulation followed the
process of: body alone, body and wheels, body, wheels and undertray, and body, wheels,
undertray and wings.
Returning to the initial goals, the simulations looked to return drag and lift values
that would enable better performance in competition. Considering our goals, once
65 lbs of downforce was achieved, lap time simulations would begin. By the final
simulation, the car was comfortably past this benchmark.
3D Simulations
Trial Date Title Cd Cl L/D Area (mˆ2) F D (lb) F L (lb) F D adj (lb) F L adj (lb) Configuraton Change
1 11/13/18 KMR 19’ Body Diffuser V1 1.018 -1.343 -1.319 0.373 16.860 -59.642 8.430 -29.821 Body
2 11/14/18 KMR 19’ Wing Kit V1 (Front/Rear) 1.071 -1.956 -1.826 1.205 57.341 -86.874 28.670 -43.437 Rear Wing High Attitude
3 11/15/18 KMR 19’ Wing Kit V2 1.029 -1.934 -1.879 1.205 55.086 -85.887 27.543 -42.943 Rear Wing Plates

In the chart, there are adjusted lift and drag values. This is considered to be
a worst case scenario, in which the real life case diverges from the simulation with
approximately 50% performance loss. This is just a reference point that we hope to
keep in mind as a minimum baseline.

4.4 Lap Time Simulator

Once the 3D simulations reached initial goals, a lap time simulator was used to
jusify gains. To perform the lap time simulation, a pre-made solution was used
(Development, n.d.). By calculating the weight gain of the front and rear wings and
diffuser to be approximately 40 lbs, and then performing a weighted average of center
of gravity to find an increase of 1.4 inches, these values were implemented into the
simulator along with the L/D and downforce values.
After including all of these assumptions and values into the simulator, it returned
projected performances for the 2011 FSAE West competition. A table of event per-
formances is included below, comparing the car with no aerodynamics to the car with
full aerodynamics had both competed in that competition:
Lap Time Simulation
2011 Event Original Result (s) Adjusted Original Result (s) Original Position New Result (s) Adjusted New Result (s) New Position
Skidpad 5.149 5.659 28 4.833 5.312 12
Acceleration 4.75 5.221 39 5.1 5.606 42
Endurance 1528.438 1679.957 19 1476.351 1622.706 15
Autocross 77.4 85.073 38 74.806 82.222 33

There is an adjusted result to incorporate real-world inefficiencies and reduction in

performance compared to ideal cases. It measured up to be a 10% gain in time. This

15
value was not randomly determined; rather it was based on the car’s expected per-
formance in skidpad. Using the skidpad analysis below, a time was calculated. This
steady state value was greater than the lap time simulator’s projected performance
by approximately 10%, thus this was used to adjust all track performance values.

4.4.1 Skidpad
The steady-state skidpad simulation was done in accordance with that taught by
Josh Browne in his course Automotive Dynamics. The general process is first taking
a known skidpad time. Then, knowing the radius, determining a velocity and lateral
v2
acceleration (knowing Ay = ). Knowing the weight of the car, its weight distri-
R
bution, and the FLLTD, static and dynamic cornering weights could be calculated.
Then, incorporating additional Fz from aerodynamics, and using known tire data,
the additional cornering Fy could be calculated. This could then be converted into a
lateral acceleration, which converts to a lap time.

4.5 Points Analysis

Referencing the projected event performance, the track position could be converted
to points at the competition. Given that the simulator was designed for the 2011
FSAE West competition, these results (FSAE West Results, 2011) were found on the
FSAE website. The chart below demonstrates the difference:

Points Performance
2011 Event Original Position Original Points New Position New Points
Skidpad 28 16.8 12 30.13
Acceleration 39 17.83 42 11
Endurance 19 134 15 174
Autocross 38 7.5 33 22.67
Total Points 176.13 Total Points 237.8

The addition of aerodynamics, based on the 2011 FSAE West competition, would
net over 61 points. A full 60 points could be the biggest factor in bringing the team
into the top 20 at FSAE Michigan. Thus, the performance of the aerodynamics system
was considered points positive, and efforts were turned fully to manufacturing.
An interesting fact found through these lap time analyses is that drag is not
important to consider with this design. Even relatively low L/D values (within reason
for an acceptably designed aerokit) still generated positive points gains. This actually
makes sense considering that in FSAE competition, the car is almost never power
limited; it is almost exclusively grip limited in most events. Focusing on downforce
and less on L/D will be invaluable for future designs of an aerokit.

16
5 Manufacturing
5.1 Molds
As desired, Emile from Fibreworks has committed to designing and manufacturing
the molds for this project. Thus, it is professionally made and intended to improve
the quality of build for the car.

5.2 Mounting to Chassis

5.2.1 Rear Wing
The rear wing is mounted using 2-force members with adjustable rod ends. The
brackets are inset into the carbon fiber layup. There are 5 members to fully constrain
the motion of the rear wing. It must not flex under loading, and thus the rods used
are as robust as possible within reason.

5.2.2 Front Wing

Instead of the rods, the front wing is designed to be mounted using custom machined
aluminum brackets that bolt directly into the front wing, and are slotted on the
other end to bolt to the front bulkhead of the chassis. The slotting allows micro-
adjustability. Further, it remains as rigid as possible.

5.3 Assembling Parts of Assemblies

The hollow wing profiles must be mounted to the end plates, as the end plates are
designed to mount to the chassis (in the case of the rear wing). Thus, there are
machined aluminum end caps that are inset in the carbon fiber. These end caps have
tapped holes that allow the wings to be bolted to the end plate.

6 Conclusion
In this project, the aerodynamics system for the Knickerbocker Motorsports Formula
SAE team was designed, optimized, and planned for manufacturing. Originally set
about to achieve 65lbs of downforce at 40MPH, the system instead achieved as much
as 85lbs of downforce in CFD simulations in ANSYS, with an L/D of approximately
-1.92.
The system was designed in Solidworks, and optimized in both 2 and 3 dimen-
sions in ANSYS. After CFD simulations, it was validated using lap time simulations
to generate a points analysis, which justified the system in terms of higher scoring at
competition. The system is expected to generate an additional 61 points at competi-
tion compared to the same car without aerodynamics.
Going forward, the system has room for improvement, but the initial goals to
design, optimize, and plan the manufacturing of the system were met in their entirety.

17
Figure 9: The design of the car in Solidworks, from the front.

18
7 Appendix
7.1 2D Simulation Mesh Parameters
The basic mesh remained the same as the default settings, except for a few key
changes:
Max Face Size = 0.009
Smoothing = High
Max Inflation Layers = 6
Add inflation feature
Select the airfoil faces, and add 6 layers of inflation
Simulation box dimensions:
Extend 2 meters in front of wings, set as velocity inlet
Extend 2 meters above and below wings, set as symmetry (or a ground)
Extend 13 meters past the end of the wings, set as pressure outlet at ambient

7.2 Front Wing 2D Simulations

Figure 10: Countour plot of front wing test Trial 1.

19
Figure 11: Streamline plot of front wing test Trial 1.

Figure 12: Countour plot of front wing test Trial 2.

20
Figure 13: Streamline plot of front wing test Trial 2.

Figure 14: Countour plot of front wing test Trial 3.

21
Figure 15: Streamline plot of front wing test Trial 3.

Figure 16: Countour plot of front wing test Trial 4.

22
Figure 17: Streamline plot of front wing test Trial 4.

Figure 18: Countour plot of front wing test Trial 5.

23
 Figure 19: Streamline plot of front wing test Trial 5.

7.3 Rear Wing 2D Simulations

Figure 20: Countour plot of rear wing test Trial 1.

24
Figure 21: Streamline plot of rear wing test Trial 1.

Figure 22: Countour plot of rear wing test Trial 2.

25
Figure 23: Streamline plot of rear wing test Trial 2.

26
Figure 24: Countour plot of rear wing test Trial 3.

27
Figure 25: Streamline plot of rear wing test Trial 3.

28
8 References
Airfoil. (2018, December 04). Retrieved from
https://en.wikipedia.org/wiki/Airfoil/media/File:Wingprofilenomenclature.svg

Angle of Attack. (n.d.). Retrieved from

http://www.aviationchief.com/angle-of-attack.html

Bernoulli Equation. (n.d.). Retrieved from

http://hyperphysics.phy-astr.gsu.edu/hbase/pber.html

Development. (n.d.). Formula SAE Simulation. Retrieved from

http://www.fsaesim.com/simulation.html

Drag Equation. (n.d.). Retrieved from

https://www.grc.nasa.gov/www/k-12/airplane/drageq.html

Forces and Diagrams. (n.d.). Retrieved from

https://www.pathwayz.org/Tree/Plain/FORCES - NAMES AND DIAGRAMS

FSAE Rules 2017-18. (n.d.). Retrieved from Formula SAE.

FSAE West Results 2011. (2011). Retrieved from

https://www.sae.org/binaries/content/assets/cm/content/attend/2018/
student-events/fsae/results/fsaew2011results.pdf.

Inclination Angle Effects. (n.d.). Retrieved from

https://www.grc.nasa.gov/www/k-12/airplane/incline.html

Katz, J. (2006). Race car aerodynamics: Designing for speed. Cambridge, MA:
Bentley.

Lift Equation. (n.d.). Retrieved from

https://www.grc.nasa.gov/www/k-12/airplane/lifteq.html

Lift Force. (n.d.). Retrieved from

https://www.grc.nasa.gov/WWW/K-12/airplane/lift1.html

Metz, D. L., Milliken, W. F., Milliken, D. L. (1998). Race car vehicle dynamics.
Warrendale, PA: Society of Automotive Engineers Internat.

29