energies
Article
Joint Research on Aerodynamic Characteristics and Handling
Stability of Racing Car under Different Body Attitudes
Zhe Zhang, Qiang Wang, Shida Song, Chengchun Zhang, Luquan Ren and Yingchao Zhang *






Citation: Zhang, Z.; Wang, Q.; Song,
S.; Zhang, C.; Ren, L.; Zhang, Y. Joint
Research on Aerodynamic
Characteristics and Handling
Stability of Racing Car under
Different Body Attitudes. Energies
2022, 15, 393. https://doi.org/
10.3390/en15010393
Academic Editor: Calin Iclodean
Received: 23 November 2021
Accepted: 29 December 2021
Published: 5 January 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Energies 2022, 15, 393. https://doi.org/10.3390/en15010393
State Key Laboratory of Automotive Simulation and Control, Jilin University, Changchun 130015, China;
zhangzhejlu@jlu.edu.cn (Z.Z.); qiangwang21@mails.jlu.edu.cn (Q.W.); caesar_star_1996@163.com (S.S.);
jluzcc@jlu.edu.cn (C.Z.); lqren@jlu.edu.cn (L.R.)
* Correspondence: yingchao@jlu.edu.cn; Tel.: +86-13504439217
Abstract: With the rapid development of FSAE, the speed of racing cars has increased year by year.
As the main research content of racing cars, aerodynamics has received extensive attention from
foreign teams. For racing cars, the aerodynamic force on the aerodynamic device ultimately acts on
the tires through the transmission of the body and the suspension. When the wheel is subjected to
the vertical load generated by the aerodynamic device, the ultimate adhesion capacity of the wheel is
improved. Under changing conditions, racing wheels can withstand greater lateral and tangential
forces. Therefore, the effects of aerodynamics have a more significant impact on handling stability. The
FSAE racing car of Jilin University was taken as the research object, and this paper combines the wind
tunnel test, the numerical simulation and the dynamics simulation of the racing system. The closed-
loop design process of the aerodynamics of the FSAE racing car was established, and the joint study
of aerodynamic characteristics and handling stability of racing car under different body attitudes was
realized. Meanwhile, the FSAE car was made the modification of aerodynamic parameter on the basis
of handling stability. The results show that, after the modification of the aerodynamic parameters, the
critical speed of the car when cornering is increased, the maneuverability of the car is improved, the
horoscope test time is reduced by 0.525 s, the downforce of the car is increased by 11.39%, the drag is
reduced by 2.85% and the lift-to-drag ratio is increased by 14.70%. Moreover, the pitching moment is
reduced by 82.34%, and the aerodynamic characteristics and aerodynamic efficiency of the racing car
are obviously improved. On the basis of not changing the shape of the body and the aerodynamic
kit, the car is put forward to shorten the running time of the car and improve the comprehensive
performance of the car, so as to improve the performance of the car in the race.
Keywords: aerodynamic characteristics; handling stability; wind tunnel test; closed-loop design;
crosswind; pitching motion; body attitude
1. Introduction
With the popularity of Formula Student China [1] in China, the FSAE racing car pays
more and more attention to aerodynamics. Using aerodynamic means to optimize the
external flow field of racing car, we can make the vehicle obtain good power performance
and handling stability, and then improve the overall performance of the vehicle. At present,
the analysis of the aerodynamic characteristics of the racing car is still in the steady-state
basic condition, and a few of the body attitude analysis and aerodynamic characteristics are
under the action of unsteady wind analysis, but the influence of aerodynamic characteristics
on the dynamics and handling stability of the racing car is ignored.
Most of the research studies on racing cars are to optimize their aerodynamics: In the
1970s, the lotus [2] was first to put forward the theory of ground effect and the application
technology of ground effect on the car. With the imitation and improvement of other teams,
the application of this technology is becoming increasingly mature. Since then, the research
on the aerodynamics of racing cars has been gradually refined, and each aerodynamic
https://www.mdpi.com/journal/energies
Energies 2022, 15, 393 2 of 29

component has been optimized and the influence of individual aerodynamic components
on the aerodynamic characteristics of the vehicle has been studied. Al Muharrami et al. [3]
conducted wind tunnel tests with wheels and ground moving on closed-wheel GT2 cars
and open-wheel F1 cars. The test results showed that the downforce of the GT2 car is
mainly provided by diffuser and rear wing, and the drag coefficient increases by more
than 3% when the wind speed increases, while the lift coefficient decreases by more than
2%. The lift coefficient and drag coefficient of F1 car vary greatly under different tracks,
and the angle of attack of the front and rear wings greatly affects the aerodynamic force
of the car. Sneh [4] conducted a numerical study on a FSAE racing car with rear engine.
The K -ε turbulence model was selected to reduce drag on the racing car under the premise
of ensuring aerodynamic downforce and improve the running speed of the car, but the
influence of aerodynamic torque on the racing process was not taken into account. Craig [5]
used porous pressure probes to conduct wind tunnel tests to explore the potential airflow
interaction between the driver’s helmet and tail and used data in two-dimensional CFD
calculations to accurately predict possible downforce. Weingart [6] took FSAE racing cars
of the 14th season as the research object and measured aerodynamic parameters through
low-cost coasting tests and compared the measured data with simulation results. The
results show that the difference between simulation results and test data is about 5%,
and the downforce data deviation is up to 18%. He [7] studied the crosswind stability
of racing cars and adopted active front-wheel steering to actively control the crosswind
stability of racing cars. The simulation results show that, under crosswind conditions, the
crosswind angular velocity and lateral displacement of the car can effectively follow the
ideal model, the side-slip angle of the centroid is optimized, the crosswind stability of the
car is improved and the driving track is ensured. In order to reduce the error of simulation
calculation, Lu [8] established a more accurate CAD model and used a computational fluid
dynamics method to simulate the fluid of a racing car. The influence of front and rear wings
on the vehicle flow field was explored, and several suggestions for optimizing aerodynamic
characteristics were put forward through the quantitative study of pressure distribution.
The running time of the car was reduced, and the race performance was improved by
optimizing aerodynamic characteristics. Zhang [9] studied the aerodynamic characteristics
of F1 racing car models at different pitching angles; the aerodynamic performance of the
car was comprehensively analyzed from aerodynamic characteristics analysis to flow field
analysis, and the results showed that, with the increase of pitch angle, the drag and the
lift decrease, the downforce of a car is mainly provided by the front and rear wings, and
the lift is provided by the front, bottom and rear wings and gradually decreases as the
pitch angle increases; the drag is more sensitive to the change of pitch angle. Wsik [10]
reduces the impact of air flow on the wheel arch clearance by changing the shape of the
front bumper and adding air intake channels and diversion plates, thus reducing the air
drag of the vehicle.
There are also a few studies on the influence of system dynamics on FSAE racing
performance: Ma et al. [11] established the aerodynamic model of FSAE racing car and
the multi-body dynamics model of “double wish-bone suspension” and “rack and gear
steering” after the simulation calculation of the aerodynamic parameters of the whole
vehicle. After the track model is established, the optimal driving trajectory is optimized ac-
cording to the minimum curvature of each bend and the weight distribution of the shortest
bend. Through the simulation of endurance race, the influence of the aerodynamic effect
on the control stability of the vehicle was analyzed, and the maximum lateral acceleration,
roll angle and other performances of the vehicle were evaluated. According to the speed
characteristic and the vertical distribution of lateral acceleration of the car to evaluate
performance limits of the car, the comparison of aerodynamic device and aerodynamic
device provided under pressure size vehicle performance index, but he did not put forward
a specific adjustment improvement program to increase the car’s performance. Based on
automobile dynamics, Li et al. [12], from Jilin University, established a car model and the
corresponding aerodynamic model through CarSim, simulated the crosswind condition of
Energies 2022, 15, x FOR PEER REVIEW 3 of 29

Energies 2022, 15, 393 3 of 29

corresponding aerodynamic model through CarSim, simulated the crosswind condition
of high-speed cars, analyzed and studied the influence of the driver model on the stability
of cars and showed the importance of studying crosswind condition of cars. Carbonne
high-speed
[13] predicted cars, the analyzed andofstudied
sensitivity vehicles thetoinfluence
transientofcrosswind
the driverthrough
model on the stability
numerical simu-of
cars and
lation, showed the
evaluated the importance
full-dynamicofcoupling
studyingperformance
crosswind condition
between of cars. Carbonne
aerodynamics and[13]ve-
predicted
hicle dynamicsthe sensitivity
simulation of vehicles to transient
and obtained crosswind
the degree through numerical
of complexity required to simulation,
simulate
evaluated the full-dynamic coupling performance between
vehicles passing crosswind. Jia et al. [14], from the Liaoning University of Technology,aerodynamics and vehicle dy-
namics
establishedsimulation
the vehicle and obtained
dynamicsthe degree
model of of
FSAEcomplexity
racing car required
through to simulate
CarSim and vehicles
con-
passing crosswind. Jia et al. [14], from the Liaoning University
ducted the transient response test simulation and snaking test simulation of steering- of Technology, established
the
wheelvehicle
angledynamics
step input model
in the of software,
FSAE racing car through
according to theCarSim and conducted
requirements the tran-
of the race. The
sient response test simulation and snaking test simulation
simulation results showed that the racing car had good stability characteristics. Huang Tof steering-wheel angle step
input in the
et al. [15] software,and
simulated according
analyzed to the
the pitch
requirements of the race.
motion model of a car Thebody
simulation
throughresults
large
showed
eddy simulation, verified the accuracy of the numerical simulation, discussedsimulated
that the racing car had good stability characteristics. Huang T et al. [15] the influ-
and
enceanalyzed
law of the thepitch
pitchmotion
motionon model of a car body
aerodynamic through
lift in large eddy
quasi-static simulation,
simulation verified
and transient
the accuracy of the numerical simulation, discussed
simulation and analyzed the mechanism of the change of transient aerodynamic the influence law of the pitch motion lift
on aerodynamic lift in quasi-static
caused by the pitch motion of the car body. simulation and transient simulation and analyzed the
mechanism of the change of transient aerodynamic lift caused by the pitch motion of the
The research of domestic and foreign scholars on racing cars mainly focuses on aer-
car body.
odynamics. They improve the running speed of racing cars by improving aerodynamic
The research of domestic and foreign scholars on racing cars mainly focuses on aero-
kits, and some scholars reduce aerodynamic drag to reduce the running time of racing
dynamics. They improve the running speed of racing cars by improving aerodynamic
cars by improving the appearance of car bodies, or study the working conditions of racing
kits, and some scholars reduce aerodynamic drag to reduce the running time of racing
cars from the perspective of dynamics. However, the interaction between aerodynamic
cars by improving the appearance of car bodies, or study the working conditions of racing
characteristics and handling characteristics is seldom considered in the simulation design
cars from the perspective of dynamics. However, the interaction between aerodynamic
stage, but this interaction happens all the time in the running process of the car, so the
characteristics and handling characteristics is seldom considered in the simulation design
joint research on aerodynamic characteristics and dynamics system of the car becomes
stage, but this interaction happens all the time in the running process of the car, so the
particularly important, as it has very important practical significance in guiding the car to
joint research on aerodynamic characteristics and dynamics system of the car becomes
adjust and improve the performance of the race.
particularly important, as it has very important practical significance in guiding the car to
adjustThisandstudy
improve tookthetheperformance
FSAE racingofcar theofrace.
Jilin University as the research object. It com-
binedThisa wind-tunnel test, aerodynamic numerical
study took the FSAE racing car of Jilin University simulationasand thedynamics simulation
research object. It com-of
racing car system; established a closed-loop design process
bined a wind-tunnel test, aerodynamic numerical simulation and dynamics simulation for aerodynamic characteris-
ticsracing
of of thecarFSAE system;racingestablished
car; conducted joint research
a closed-loop design onprocess
aerodynamic characteristics
for aerodynamic and
charac-
handlingofstability
teristics the FSAE of FSAE
racingracing car under joint
car; conducted different bodyon
research attitudes;
aerodynamicand adjusted the aer-
characteristics
odynamic parameters of FSAE racing car based on handling
and handling stability of FSAE racing car under different body attitudes; and adjusted stability. On the basis of the
not
changing the shape of the body and the aerodynamic kit,
aerodynamic parameters of FSAE racing car based on handling stability. On the basis ofthe adjustment plan of the aer-
odynamic
not changing parameters
the shapeofofthe thecar were
body putthe
and forward to shorten
aerodynamic kit, the
the running
adjustment timeplan
of the car
of the
and improve the comprehensive performance of the car, so as
aerodynamic parameters of the car were put forward to shorten the running time of the car to improve the performance
of the
and car in the
improve therace.
comprehensive performance of the car, so as to improve the performance
of the car in the race.
2. Simulation Strategy Determination
2. Simulation
2.1. Car Model Strategy Determination
2.1. Car Model
The car model studied in this paper is a simplified vehicle model of FSAE racing car
in theThe19thcarseason
modelof studied in this paper
Jilin University, and is its
a simplified vehicle model
design conforms of FSAE racing
to the relevant rules ofcar in
For-
the 19th season of Jilin University, and its design conforms
mula Student China. In order to ensure the quality of the grid and save computing re- to the relevant rules of Formula
Student
sources,China. In order to
the simplified ensure
model was thesimplified
quality of by theremoving
grid and save computing resources,
the suspension, as shownthe in
simplified
Figure 1. model was simplified by removing the suspension, as shown in Figure 1.

Figure 1. Simplified racing model.

Energies 2022, 15, x FOR PEER REVIEW 4 of 29

Energies 2022, 15, 393 4 of 29

Figure 1. Simplified racing model.

2.2. Numerical Simulation

2.2. Numerical Simulation
Using a triangular mesh model for the car surface mesh division, we set the vehicle
Using a triangular mesh model for the car surface mesh division, we set the vehicle
body and the surface of the tire grid size to 0.016 m; the curvature for the vehicle surface
body and the surface of the tire grid size to 0.016 m; the curvature for the vehicle surface
grid refinement
grid refinementwas
wasset
setto
to 36.0 points,and
36.0 points, andthe
theminimum
minimum sizesize
waswas set0.0005
set to to 0.0005 m. Through
m. Through
thethe
grid properly after repair, a better quality of a grid model was obtained.
grid properly after repair, a better quality of a grid model was obtained. The overall The overall
surface
surfacegrid
gridnumber
number ofofthis
thisracing
racing
carcar model
model is million
is 1.79 1.79 million cells,
cells, and and the
the details of details
the body,of the
body, the leading
the leading edge ofedge of the
the front wingfront
and wing
the rearand
wingthearerear
keptwing are vehicle
well. The kept well. The
surface vehicle
grid
and local detail surface grid of the racing model are shown in Figure 2.
surface grid and local detail surface grid of the racing model are shown in Figure 2.

(a)

(b)
Figure 2. Surface
Figure 2. Surfacemesh
meshand
andlocal
local detail surfacemesh
detail surface meshofof simplified
simplified racing
racing model:
model: (a) vehicle
(a) vehicle surface
surface
grid and
grid (b)(b)
and front
frontwing
winglocal
local surface mesh.
surface mesh.

Generally,the
Generally, theblockage
blockage ratio
ratioshould
shouldbebeless than
less than5%5%
[16–19]. In combination
[16–19]. In combinationwith with
the the
size of the simulation model in this paper, in order to improve the calculation
size of the simulation model in this paper, in order to improve the calculation accuracy accuracy and
eliminate the influence of the boundary, the computational domain is 13 times as long as
and eliminate the influence of the boundary, the computational domain is 13 times as long
the vehicle and 11 times the width of the vehicle, as high as 5 times of the vehicle. The front
as the vehicle and 11 times the width of the vehicle, as high as 5 times of the vehicle. The
end of the model is at least 4 times the model length from the entrance of the computing
front
domain,of
end the
the backmodel is at
end is atleast
least84times
times the
the model
model length
length from from the and
the exit entrance of the com-
the blocking
puting domain,
ratio is 1.2%. the back end is at least 8 times the model length from the exit and the
blocking ratio
After the is surface
1.2%. mesh was divided, the volume mesh was divided. The type of
volume
After mesh is cut volume
the surface mesh was mesh, which ensures
divided, the accuracy
the volume mesh was of simulation
divided. Thecalculation
type of vol-
umeand coordinates
mesh the feasibility
is cut volume mesh,ofwhich
simulation scheme.
ensures the Encryption
accuracy ofdomains of different
simulation sizes and
calculation
were set. The following encryption strategy was adopted: a total of three
coordinates the feasibility of simulation scheme. Encryption domains of different sizes layered body-grid
encryption domains were set up to encrypt the model and computing domain layer by
were set. The following encryption strategy was adopted: a total of three layered body-
layer, as shown in Figure 3.
grid encryption domains were set up to encrypt the model and computing domain layer
The grid size of the three encryption domains was set in a certain proportion based
by on
layer, as shown
the basic in Figure
size, and different3. grid numbers were obtained by changing the basic size for
grid independence verification. In this simulation, the cells number of the volume grid
is 16 million to meet the requirements of grid independence, and the final generated grid
number is about 16.15 million. The encrypted computing domain is shown in Figure 4.
Energies 2022,15,
Energies2022, 15,393
x FOR PEER REVIEW 55 of
of 29
29

Figure 3. Volume mesh encryption domain.

The grid size of the three encryption domains was set in a certain proportion based
on the basic size, and different grid numbers were obtained by changing the basic size for
grid independence verification. In this simulation, the cells number of the volume grid is
16 million to meet the requirements of grid independence, and the final generated grid
Figure
number
Figure 3.3.Volume
Volume
is mesh
aboutmesh
16.15encryption domain.
million. The
encryption encrypted computing domain is shown in Figure 4.
domain.

The grid size of the three encryption domains was set in a certain proportion based
on the basic size, and different grid numbers were obtained by changing the basic size for
grid independence verification. In this simulation, the cells number of the volume grid is
16 million to meet the requirements of grid independence, and the final generated grid
number is about 16.15 million. The encrypted computing domain is shown in Figure 4.

Figure4.4.Encrypted
Figure Encryptedcomputing
computingdomain.
domain.

Theheight
The heightratio
ratioof ofeach
eachlayer
layerof ofboundary
boundarylayer
layergridgridshould
shouldbe beless
lessthan
than1.4,1.4,and
andthethe
heightratio
height ratioof ofoutermost
outermostboundary
boundarylayerlayergrid
gridtotothe
thefirst
firstlayer
layergrid
gridshould
shouldalso alsomeet
meetthisthis
requirement
requirement[20]. [20].Based
Basedon onthe
theabove
abovefactors,
factors,the
theboundary
boundarylayer layergrid
gridexpansion
expansionratio ratiowas
was
selected
selectedasas1.2, 1.2,and
andthethetotal
totalboundary
boundary layer thickness
layer thickness is 0.008
is 0.008m. m.
TheThesimulation
simulation calculation
calcula-
isFigure
carried
tion out forout
is 4.carried
Encryptedmodels with domain.
computing
for models different boundary
with different layers. When
boundary layers.the
Whenboundary layer number
the boundary layer
exceeds
number11 layers, 11
exceeds thelayers,
lift coefficient tends to betends
the lift coefficient stable,toand the mesh
be stable, andnumber
the mesh of the
numbermodel of
body The height
increases ratio
with theof each layer
increase of of
the boundary
boundary layer
layer grid should
number.
the model body increases with the increase of the boundary layer number. Considering be less
Considering than 1.4,
the and
problem the
height
of
the ratio ofof
computing
problem outermost
resources,
computing theboundary
SST K -ωlayer
resources, thegrid
SSTto
turbulence the
-ω first
Kmodel layer
with
turbulence 11grid
layers
modelshould
ofwithalso
11 meet
boundary this
layer
layers of
requirement
grid is finally
boundary [20]. Based
adopted.
layer grid on theadopted.
is finally above factors, the boundary layer grid expansion ratio was
selected
Through
Throughas 1.2, and the total
numerical
numerical boundary
simulation
simulation layer thickness
calculation,
calculation, the is 0.008and
theaccuracy
accuracy m. reliability
and The simulation
reliability of calcula-
ofnumerical
numerical
simulation
tion is are
carried verified,
out for and
models more
withaccurate simulation
different boundary strategies
layers.
simulation are verified, and more accurate simulation strategies are determined for are
When determined
the boundaryfor subse-
layer
sub-
quent
number numerical
exceeds simulation
11 layers, of
the various
lift working
coefficient conditions
tends to be under
stable,
sequent numerical simulation of various working conditions under different body atti- different
and the body
mesh attitudes
number of
and
the different
model crosswind
body increases effects.
with The
the overall
increasedetermined
of the simulation
boundary layer
tudes and different crosswind effects. The overall determined simulation strategies are strategies
number. are shown
Considering in
Table
shown 1. in Table
the problem of 1.
computing resources, the SST K -ω turbulence model with 11 layers of
boundary layer grid is finally adopted.
Table 1. Final simulation strategy.
TableThrough numerical
1. Final simulation simulation calculation, the accuracy and reliability of numerical
strategy.
simulation are verified, Type Typemore accurate simulation strategies
and Related are determined for sub-
Parameter
Related Parameter
sequent numerical simulation of various working conditions under different body atti-
Turbulence
Turbulencemodel model SST k-ω model
tudes and different crosswind effects. The overall determined SST k-ω model
simulation strategies are
shown in Table 1. Inlet Inlet velocity
velocity v = 15 vm/s
= 15 m/s
Outlet pressure
Outlet p = 0 Pa
pressure p = 0 Pa
Boundary condition Air density ρ = 1.18415 kg/m3
Table 1. Final simulation
Boundary strategy.
condition AirSliding
density ρ = 1.18415 kg/m3
surface
Type Sliding
Ground
Related slip surface
Parameter
Type
Turbulence model RelatedSSTGround
Parameter
k-ω modelslip
Inlet number
Layer velocity11v = 15 m/s
Boundary layer Thickness 8 mm p = 0 Pa
Outlet pressure
Expansion ratio 1.2
Boundary condition Air density ρ = 1.18415 kg/m3
Grid base size 0.512 m
Sliding surface
Calculation domain 11 length × 12 Ground
width × 5slip
height
 Type Related Parameter
Layer number 11
Boundary layer Thickness 8 mm
Expansion ratio 1.2
Energies 2022, 15, 393 Grid base size 0.512 m 6 of 29
Calculation domain 11 length × 12 width × 5 height

3.3. Wind
Wind Tunnel
TunnelTest
Testand
andAerodynamic
AerodynamicCharacteristics Analysis
Characteristics Analysis
3.1. Wind Tunnel Test
3.1. Wind Tunnel Test
This test
This testwas
wascarried
carriedout
outinin
the Automotive
the Wind
Automotive Tunnel
Wind Laboratory
Tunnel of Jilin
Laboratory ofUniver-
Jilin University.
sity. The FSAE racing car was placed in the open test section, and the overall layout
The FSAE racing car was placed in the open test section, and the overall layout scheme of
scheme of four-wheel support was adopted [21,22]. The racing car model is shown in Fig-
four-wheel support was adopted [21,22]. The racing car model is shown in Figure 5.
ure 5.

Figure 5. FSAE racing car of Jilin University.

Figure 5. FSAE racing car of Jilin University.
First, the
First, theaerodynamic
aerodynamic six-component forceforce
six-component of theofFSAE racingracing
the FSAE car wascar
measured
was measured
under linear conditions. The wind velocity was set to increase from 15 to 30 m/s, with an
under linear conditions. The wind velocity was set 3to increase from 15 to 30 m/s, with
increment of 5 m/s, and the air density was 1.18415 kg/m . The test results are shown in
Energies 2022, 15, x FOR PEER REVIEW an increment of 5 m/s, and the air density was 1.18415 kg/m3 . The test results 7 of 29are
Figure 6. With the increase of wind velocity, aerodynamic downforce and aerodynamic
shown in Figure 6. With the increase of wind velocity, aerodynamic downforce
drag of the car continued to increase, and the lift-drag ratio of the vehicle basically re- and
aerodynamic drag
mained unchanged. of the car continued to increase, and the lift-drag ratio of the vehicle
basically remained unchanged.

Figure
Figure 6. Lifting
6. Lifting drag
drag characteristic
characteristic curve
curve at 15–30
at 15–30 m/sm/s wind
wind velocity
velocity in straight-line
in straight-line condition.
condition.

Then the crosswind angle of side-wind dynamic characteristics was tested from −15°
to +15°. The yaw model method by changing the attitude of the car body to achieve the
crosswind encountered in the driving process was adopted in the test, and the crosswind
angle of the test model changed by 5° each time. The diagram of the yaw model method
Energies 2022, 15, 393 7 of 29
Figure 6. Lifting drag characteristic curve at 15–30 m/s wind velocity in straight-line condition.

Then the crosswind angle of side-wind dynamic characteristics was tested from −15°
Then the crosswind angle of side-wind dynamic characteristics was tested from −15◦
to +15°. The yaw model method by changing the attitude of the car body to achieve the
to +15◦ . The yaw model method by changing the attitude of the car body to achieve the
crosswindencountered
crosswind encounteredininthe thedriving
drivingprocess
processwas wasadopted
adoptedininthethetest,
test,and
andthe
thecrosswind
crosswind
angle of the test model changed by 5°
◦ each time. The diagram
angle of the test model changed by 5 each time. The diagram of the yaw model method of the yaw model method
in the wind tunnel test is shown in Figure 7. During the driving
in the wind tunnel test is shown in Figure 7. During the driving process of the car, process of the car,the
the
angle β between the incoming flow direction and the car body
angle β between the incoming flow direction and the car body is the crosswind angle is the crosswind angle of
car,
of car,and
and it it
is isalso thethe
also yawyaw angle
anglein in
yawyaw action.
action.TheTheangle β isβdivided
angle is dividedintointo
positive and
positive
and negative angles, which defines the clockwise yaw of car in the top view, and theβ
negative angles, which defines the clockwise yaw of car in the top view, and the angle
is theβpositive
angle crosswind
is the positive angle. Now,
crosswind angle.by the yaw
Now, by the model
yaw method,
model method,in the car body
in the carcoordi-
body
coordinates for reference, the vector triangle method is adopted to decompose velocity,
nates for reference, the vector triangle method is adopted to decompose to flow to flow
v, according
velocity, to the relative
v, according motion,motion,
to the relative along the X direction
along of the car
the X direction of body velocity
the car compo-
body velocity
nent—vA equivalent
component—v to the car speed, vW along the Y direction of the speed car body com-
A equivalent to the car speed, vW along the Y direction of the speed car body
ponents can can
components be equivalent
be equivalent to crosswind
to crosswindvelocity and the
velocity andvelocity of twoofequivalent
the velocity veloc-
two equivalent
ities is the product of incoming velocity v and β angle trig function.
velocities is the product of incoming velocity v and β angle trig function. The formula The formula is ex-
is
pressed as follows:
expressed as follows:
v A = v·sinβ (1)
𝑣 𝑣 ∙ 𝑠𝑖𝑛𝛽 (1)
vw = v·cosβ (2)
𝑣 𝑣 ∙ 𝑐𝑜𝑠𝛽 (2)

Figure7.7.Yaw
Figure Yawmodel
modelmethod
methodfor
forwind
windtunnel
tunneltest.
test.

The aerodynamic parameters of the numerical simulation and wind-tunnel test under
the condition that the wind velocity is 15 m/s, the crosswind angle is 0◦ and the body
attitude is unchanged are shown in Table 2. The simulation data of the yaw model method
and the aerodynamic characteristics data of the yaw model method in the wind tunnel test
are shown in Table 3 for when the synthetic wind is 15 m/s and the crosswind angle is 10◦ .

Table 2. Comparison of lift coefficient and drag coefficient between numerical simulation and wind
tunnel test.

Numerical Simulation Wind Tunnel Test Relative Error

Lift coefficient −2.089 −1.968 6.14%
Drag coefficient 1.254 1.195 4.94%

Table 3. Aerodynamic characteristics of FSAE racing car under the action of 10◦ crosswind angle.

Lift/N Drag/N Lateral Force/N

Wind-tunnel-test data −280.389 182.319 −87.743
Yaw-model-method simulation −292.389 194.784 −85.709
Energies 2022, 15, 393 8 of 29

According to the data in Table 2, under the same simulation and test conditions, the
relative error of the lift coefficient is 6.14%, and the relative error of the drag coefficient
is 4.94%. There is little difference between the aerodynamic parameters of the simulation
and the test. According to the data in Table 3, the crosswind data simulated by the yaw
model method are slightly larger than the test values, but the errors are kept within 5%. In
addition, the lateral force of the racing car in crosswind condition conforms to the pretest
value. In the wind-tunnel test, due to the influence of equipment, external interference
and other factors, as well as the simulation accuracy of software, there are certain errors
between the simulation and test, but the relative errors are within the allowable range of
engineering practice, so the numerical simulation results have high accuracy.

3.2. Analysis of Aerodynamic Characteristics

Based on the actual racing conditions, the FSAE racing car’s aerodynamic simulation
analysis was conducted under multiple working conditions by using STAR-CCM+ software.
The aerodynamic characteristics of FSAE racing car were studied from the aspects of
aerodynamic force, aerodynamic torque, vehicle flow field structure, etc., and we analyzed
the reasons for the aerodynamic characteristics’ differences in each working condition.
The influence of each aerodynamic component on the aerodynamic characteristics of the
vehicle was explored. The aerodynamic parameters of the FSAE racing car in each working
condition were collected to provide data support for the dynamics simulation.
This paper mainly studies some typical aerodynamic conditions of the figure-eight sur-
round test and high-speed obstacle-avoidance test in FSAE. The influence of aerodynamic
characteristics on the handling stability was deeply evaluated by using CarSim software.
The aerodynamic six-component force data were obtained under various operating condi-
tions in an input system module.

3.2.1. Analysis of Pitch Working Condition

According to the actual situation, the pitch-angle data of the car are −2.0◦ ~1.0◦ , and
the body pitching angle under acceleration is defined as positive. In the simulation analysis,
the wind velocity was set as 15 m/s, the air density was 1.18415 kg/m3 and the angle
increment of body attitude change was 0.5◦ .
When analyzing the pitch condition of the car, the range of pitch angle of the car was
determined to be from −2.0◦ to 1.0◦ , the simulation wind velocity was set to 15 m/s and
the increment of body pitch angle was 0.5◦ . A total of 14 conditions were simulated and
analyzed. Figure 8 shows the trend of lift coefficient, drag coefficient and pitching moment
variation curves in various working conditions.
By analyzing the change characteristic curves of lift and drag in various pitching
conditions, it can be found that the change of drag is small in the pitching condition: −2.0◦
that the body leans forward, 2.0◦ in acceleration condition means the minimum drag, −1.0◦
that the body leans forward, 1.0◦ in acceleration condition means the maximum downforce
and 0.5◦ that the body leans backward 0.5◦ in braking condition means the minimum
downforce. It is calculated that the lift–drag ratio of the vehicle is the maximum at −1.0◦ ,
and the aerodynamic efficiency of the vehicle is the highest in various pitching conditions.
The lift–drag ratio is the minimum at 0.5◦ , and the aerodynamic efficiency of the vehicle
is the lowest in various pitching conditions. When the pitch angle is −1.0◦ , the pitching
moment coefficient is the smallest, indicating that the pitching moment is the smallest,
the front and rear axles of the car are least affected by aerodynamic force distribution,
and the car is likely to run more stably. However, the pitching moment coefficient is the
largest when the pitching angle is 1.0◦ , indicating that the pitching moment is the largest
in this condition, and the front and rear axles of the racing car are most affected by the
aerodynamic force distribution, and the stability of the racing car may be worse.
 analyzed. Figure 8 shows the trend of lift coefficient, drag coefficient and
variation curves in various working conditions.
Energies 2022, 15, 393 9 of 29

(a) (b)

(c)
Figure 8. Pitch working condition of the car drag coefficient, lift coefficient and pitching moment
Figure 8. Pitch working condition of the car drag coefficient, lift coefficient and
curve: (a) drag coefficient, (b) lift coefficient and (c) pitching moment.
curve: (a) drag coefficient, (b) lift coefficient and (c) pitching moment.
Table 4 compares the lift-drag characteristics of each component in the two working
conditions and the basic working conditions under 15 m/s wind velocity. By analyzing the
By analyzing
drag data of each thecomponent
changein characteristic curvesin of
the three pitching conditions the lift
table and
above,drag
it can bein vari
found that compared with the pitch 0◦ in the basic condition, the downforce of the front
ditions, itwing
canandbe tail found
wing in thethat
pitch −the
1.0◦ change
condition is of drag
greater than is
thatsmall in condition,
in the basic the pitching
that the body
and theleans
downforce forward,
of the front2.0°
wingin andacceleration
body increases morecondition
than that ofmeans the min
the tail wing.
This is because the working condition of pitch −1.0◦ body attitude forward, front wing and
that the body
body wing leans forward,
structure 1.0°clearance
from the ground in acceleration
is reduced, when condition means
the car speed racing car the m
force andforward,
0.5° thatcar front
thewing
body and leans
the wingbackward
near the optimum 0.5°aerodynamic
in braking kit from regional, mea
condition
venturi jet effect strengthen, airflow velocity increases under the wing surface, negative
downforce. It iszone
pressure calculated thatupthe
also increase, blade andlift–drag
down surfaceratio of the
of differential vehicle
pressure is the m
increases,
the downforce on the front and flanks also increases. When the car is in pitch of −1.0◦ , the
and the aerodynamic efficiency of the vehicle is the highest in various pit
car body leans forward, and the equivalent angle of attack of the main wing and flap of the
The lift–drag
rear wingratio is the
increases; minimum
meanwhile, the area ofat
the0.5°, and
rear wing thebyaerodynamic
blocked the driver’s head and efficien
other components decreases, and the aerodynamic efficiency of the single part of the rear
is the lowest in various
wing increases. However,pitching conditions.
the up-flow angle When
of the front wing the
and the side pitch
increasesangle
as the is −
moment coefficient is the smallest, indicating that the pitching moment i
front and rear axles of the car are least affected by aerodynamic force dis
car is likely to run more stably. However, the pitching moment coeffic
Energies 2022, 15, 393 10 of 29

front wing and the side approach the optimal departure area. Therefore, the downforce of
the tail increases less in pitch of −1.0◦ .
The pitching condition of −1.0◦ and the pitching condition of 0.5◦ are compared and
analyzed. The comparison of pressure cloud images of the two conditions is shown in
Figure 9. By comparing the vehicle and local pressure cloud images, it can be seen that
there is a pitch angle 1.5◦ difference in body attitude between the two working conditions.
Due to the change of front wing position, the positive pressure area of front wheel increases
and front wheel drag increases in pitch 0.5◦ working condition. In the pitch condition of
−1.0◦ , the lower surfaces of the front wing and the side wing are closer to the optimal
separation area, and the area of the negative pressure area increases as the flow velocity
passing under the wing increases, and the downforce increases as the pressure difference
between the upper and lower parts increases. In the pitch condition of −1.0◦ , the shield
area of the tail decreases and the negative pressure area under the main wing increases.
However, the negative pressure area of the tail flap decreases, due to the influence of the
up-flow flowing through the front aerodynamic components. The downforce of the single
tail part increases by combining the two different influences.

Table 4. Lift data and drag data of racing components in pitching conditions of −1.0◦ , 0◦ and 0.5◦ .

Component Pitch −1.0◦ Pitch 0◦ Pitch 0.5◦ Pitch −1.0◦ Pitch 0◦ Pitch 0.5◦
Name (Lift/N) (Lift/N) (Lift/N) (Drag/N) (Drag/N) (Drag/N)
Body −110.806 −97.763 −95.096 76.603 81.689 81.517
Front wheel 10.198 10.660 13.797 9.023 10.651 12.136
Front wing −114.849 −94.890 −93.482 24.117 21.990 22.133
Rear wheel 10.527 12.746 10.834 11.428 12.068 10.953
Tail −148.934 −145.772 −140.723 66.254 64.890 65.166
Total −353.866 −315.027 −304.670 187.425 191.299 191.904

3.2.2. Analysis of Crosswind Condition

It is defined that the direction of environmental wind inflow is positive on the left
side of the driving direction of the car. By default, the FSAE racing model is symmetric
on the left and right, so the absolute values of the aerodynamic six components of the
car are the same under the condition of the same absolute value of the crosswind angle.
In order to save computing resources, this paper only simulates the working condition
with positive crosswind angle, and the working-condition data with negative crosswind
angle are obtained by the positive working condition. Now, the crosswind conditions
corresponding to each coefficient are divided. The crosswind angle variation conditions
of body attitude stability are shown in Table 5. In simulation analysis, the wind velocity
was set as 15 m/s, the air density was 1.18415 kg/m3 , the crosswind angle varied from
0◦ to 180◦ and the angle increment was 10◦ . The downflow velocity, vehicle velocity and
crosswind velocity of some crosswind declination angles in the crosswind simulation are
shown in Table 6.
According to the data analysis, when the absolute value of the angle β is the same, the
lift and drag of the car are basically the same, and the aerodynamic characteristics of the
car show a symmetric trend under the action of the positive and negative side wind. To
simplify the calculation process, the aerodynamic six-component force data of the racing car
are the same when the absolute value of the angle β is the same in the crosswind condition,
and the crosswind simulation needs to be simulated at 0◦ ~180◦ with increments of 10◦ .
Figure 10 shows the variation curves of lift coefficient, drag coefficient and lateral force
coefficient of the racing car at a 0◦ ~180◦ crosswind angle.
 difference between the upper and lower parts increases. In the pitch condition of −1.0°,
the shield area of the tail decreases and the negative pressure area under the main wing
increases. However, the negative pressure area of the tail flap decreases, due to the influ-
ence
Energies 2022, of the up-flow flowing through the front aerodynamic components. The downforce
15, 393 11 of 29
of the single tail part increases by combining the two different influences.

PEER REVIEW 11 of 29

(a)

(b)

(c)
of −1.0 ◦ (left)
Figure 9. ComparisonFigureof pressure cloudofimages
9. Comparison pressureof racing
cloud cars
images in pitching
of racing condition
cars in pitching of −1.0°
condition (left)
and 0.5° (right): (a) theand 0.5◦ view
main (right):of(a) thevehicle,
the main view
(b)ofthe
the vehicle
vehicle, (b)
andthe(c)
vehicle and (c)wing
the front the front
andwing and the
the side
side wing.
wing.

3.2.2. Analysis of Crosswind Condition

It is defined that the direction of environmental wind inflow is positive on the left
side of the driving direction of the car. By default, the FSAE racing model is symmetric on
 Wind velocity V (m/s) 15 15 15 15 15
Vehicle velocity VA (m/s） 14.772 14.095 0 −14.095 −15
Crosswind velocity VW (m/s） ±2.605 ±5.130 15 ±5.130 0
Energies 2022, 15, 393 12 of 29

According to the data analysis, when the absolute value of the angle β is the sam
Table 5. Variation conditions of crosswind angle of vehicle attitude stability.
the lift and drag of the car are basically the same, and the aerodynamic characteristic
Vehicle Attitude Level of Stability
the car show a symmetric trend under ◦the action of the positive and negative side wi
Crosswind angle/ 0, ±10, ±20, ±90, ±160, 180
To simplify the calculation process, the aerodynamic six-component force data of the r
ing car are the same when
Table 6. Data the absolute
of downflow value
velocity, of theand
vehicle velocity angle β is
crosswind the atsame
velocity in angle
crosswind the incrossw
crosswind simulation part.
condition, and the crosswind simulation needs to be simulated at 0°~180° with increme
±10of◦ 0◦ ±0◦ drag ◦ 180◦
of 10°. Figure 10 showsCrosswind angle β
the variation curves lift ±coefficient, ±160
coefficient and late
Wind velocity V (m/s) 15 15 15 15 15
force coefficient of the racing car at a 0°~180° crosswind angle.
Vehicle velocity VA (m/s) 14.772 14.095 0 −14.095 −15
Crosswind velocity VW (m/s) ±2.605 ±5.130 15 ±5.130 0

(a) (b)

(c)
Variation curves of lift coefficient, drag coefficientand
and lateral force coefficient ◦ ◦
at 0 ~180 at 0°~1
Figure 10. Variation Figure
curves 10. of lift coefficient, drag coefficient lateral force coefficient
crosswind angle: (a) lift coefficient, (b) drag coefficient and (c) lateral force coefficient.
crosswind angle: (a) lift coefficient, (b) drag coefficient and (c) lateral force coefficient.
Figure 10 shows that, with the increase of crosswind angle β, the car can gradually
reduce aerodynamic downforce. When the angle β approaches 50◦ , the lift force of the
Figure 10 shows
car isthat, with
0; at this time, the
due toincrease of effect,
the crosswind crosswind anglekitβ,ofthe
the aerodynamic car
the car no can
longergradua
◦
reduce aerodynamic downforce.
provides When the
additional downforce. Whenangle
angle ββisapproaches 50°,
more than 50 , the the
lift of the lift force of the
car increases
gradually, the ultimate adhesion capacity of the tire decreases gradually, and the ability of
is 0; at this time, due to the crosswind effect, the aerodynamic kit of the car no lon
provides additional downforce. When angle β is more than 50°, the lift of the car increa
gradually, the ultimate adhesion capacity of the tire decreases gradually, and the abi
of the wheel to bear the lateral and tangential reaction forces on the ground decreases;
Energies 2022, 15, 393 13 of 29

the wheel to bear the lateral and tangential reaction forces on the ground decreases; the
car’s lift peaks when the angle β is around 130◦ , and then the car’s lift decreases with the
increase of the crosswind angle β.
When the simulated wind velocity remains unchanged, the drag increases with the
increase of the orthographic area, so the drag increases with increase of the crosswind
angle β. When the angle β approaches 90◦ , the drag reaches the peak, and then it decreases
due to the decrease of the orthographic area.
For lateral force, when the angle β is near 40◦ and 130◦ , the side force of the car reaches
the peak, and the lateral force, for which the angle β is within the range of 0◦ ~90◦ , is greater
than lateral force that the angle β is in the range of 90◦ ~180◦ .
It can be seen from the three sets of curves that, when the crosswind angle β is
between 0◦ and 40◦ , the aerodynamic characteristics of the car are significantly reduced,
but additional downforce can still be generated. The aerodynamic suite can improve
the handling stability of the car. When the crosswire angle β is greater than 40◦ , the
aerodynamic suite of the car fails. Moreover, 0◦ ~40◦ can be regarded as the effective
working range of racing aerodynamic kit under crosswind. Three representative crosswind
deflection angles were selected in this study to analyze and study the vehicle outflow
field structure.

4. Analysis of Vehicle Handling Stability under Multiple Operating Conditions

Based on the aerodynamic characteristics of the car, the vehicle outflow field simulation
of FSAE car was carried out, and the aerodynamic data of the car under different body
attitudes were obtained through the simulation. In order to analyze the influence of the
aerodynamic characteristics on the vehicle handling stability, the aerodynamic parameters
obtained from the numerical simulation were input into the dynamics model based on the
characteristics of the racing car assembly. In the dynamic events of FSAE, the aerodynamic
performance of the car has a crucial impact on the figure-eight loop test, high-speed obstacle
avoidance test and endurance test, among which the high-speed obstacle avoidance test
and endurance test can be simplified into multiple snaking driving and extreme steering.
Therefore, based on the FSAE event and the characteristics of the car, this study carried out
the car stability research, snaking test simulation and figure-eight surround test simulation
under the crosswind effect to analyze the influence of the car aerodynamic characteristics
on the vehicle handling stability.

4.1. Dynamic Simulation Model

The dynamics model of the racing car includes the body assembly, steering system
and other systems. The specific parameters were set as shown in Table 7.

4.2. Snaking Test Simulation

The snaking test, also known as S-row round pile test, is an important index of vehicle
stability. According to the national standard [23], the test speed should be half of the
benchmark speed specified for small vehicles, rounded to an integer multiple of 10, and
the speed should be driven in a stable and straight line until it passes the test area. After
the test is completed, the test speed should be gradually increased, and the test should be
repeated for a total of 10 times (excluding the times of knocking down the marker). The
maximum speed shall not exceed 80 km/h. By comparing the real car data of FSAE racing
car, we see that the initial speed of this test is 65 km/h, and the distance of road model L is
30 m. There are 3 tests, with each speed increment of 5.0 km/h, and the final speed of the
test is 75 km/h, which meets the requirements of national standard test. The brake control
adopts open-loop brake to control, and the brake pressure is 0. In order to simplify the
test, closed-loop automatic shift control was adopted in this paper, and 0.15 s pre-sighting
steering was adopted in steering control.
Energies 2022, 15, x FOR PEER REVIEW 14 of 29

Energies 2022, 15, 393 14 of 29

Unsprung mass kg 22.000

Rear
Table 7. Parameters of dynamic simulation
Stiffness of model.
suspension
spring N/mm 45.000
over-
N/(mm/
Component/System hang Shock absorber damping
Parameter Name Unit 2.400
Numerical
s)
Total weight kg 260.000
Height of the center ofOuter
mass diameter mm inch 18.300
280.000
Moment of inertia about theTread
X axis width 2
kg.m inch 304.7637.500
Tyre
Moment of inertia about the Y axis kg.m 2 109.800
Body assembly Applicable rim width 2 inch 7.000–8.000
Moment of inertia about the Z axis kg.m 122.875
Weight
Longitudinal position of center of kg 4.540
mm 866.250
Aerodynamics
mass from front axis reference point XYZ mm (−866.250,0,46.100)
Center of mass to the longitudinal
Orthographic area mm A/m² 0 1.145
Aerodynamics
plane in a transverse position
Wheelbase ◦
L/mm 1575.000
Kingpin inclination angle 7.000
Steering system Kingpin rake angle Air density D ◦ kg/m³ 25.0001.184
Steering force ratio 1/3.800
4.2. Snaking Test Simulation
Unsprung mass kg 18
The snaking test,Front-wheel camberas S-row round pile test, is an important
also known −0.500
index of vehicle
Front overhang Front-wheel beam Angle −0.700
stability. According
Stiffnessto the national
of suspension standard [23], the test speed should
spring 30.000be half of the
Suspension system benchmark speed specified for small vehicles, rounded to an integer multiple of 10, and
Shock absorber damping N/(mm/s) 3.600
Unsprung mass kg 22.000
the speed should be driven in a stable and straight line until it passes the test area. After
Rear overhang Stiffness of suspension spring N/mm increased, and45.000
the test is completed, the test speed should be gradually the test should be
Shock absorber damping N/(mm/s) 2.400
repeated for a total of 10 times (excluding the times of knocking down the marker). The
Outer diameter inch 18.300
maximum speed shall not exceed 80 km/h. By comparing the real car data of FSAE racing
Tread width inch 7.500
Tyre car, we see that the initial speed of this test is 65 km/h,
Applicable rim width inch and the distance of road model L
7.000–8.000
is 30 m. There are 3 tests, with each speed increment
Weight kg of 5.0 km/h, and 4.540
the final speed of
the test is 75 km/h, which
Aerodynamics meetspoint
reference theXYZ
requirementsmm of national standard test. The brake
(−866.250,0,46.100)
control adopts open-loop brake
Orthographic to
area control, and the A/m2 pressure is 0. In1.145
brake order to simplify
Aerodynamics
Wheelbase
the test, closed-loop automatic shift control was adopted L/mm 1575.000
in this paper, and 0.15 s pre-
Air density D kg/m3 1.184
sighting steering was adopted in steering control.
According to the test requirements of snaking test in GB/T 6233-2014 [23], the road
According to the test requirements of snaking test in GB/T 6233-2014 [23], the road
modeling
modelingsection
section isismodeled
modeled according
according toto the
the pile
pile arrangement diagram. The
arrangement diagram. Theground
ground fric-
tion system is 1.0, and the pile arrangement diagram is shown in
friction system is 1.0, and the pile arrangement diagram is shown in Figure 11.Figure 11.

Figure 11. Pile-arrangement diagram of snake test.

Figure 11. Pile-arrangement diagram of snake test.
Considering the actual driving conditions, the simulation conditions can be divided
into two types: one is the snaking test simulation under no crosswind action, and the
Considering the actual driving conditions, the simulation conditions can be divided
other is the snaking test simulation under crosswind action. The influence of aerodynamic
into two types: one is the snaking test simulation under no crosswind action, and the other
characteristics on vehicle handling stability under crosswind is judged by comparing the
istwo
theworking
snakingconditions.
test simulation under crosswind action. The influence of aerodynamic char-
acteristics on vehicle handling stability under crosswind is judged by comparing the two
4.2.1. Evaluation
working Criteria for Snaking Test
conditions.
The simulation was evaluated according to the serpentine test evaluation method
4.2.1.
in theEvaluation Criteria
automobile for and
handling Snaking Testindex limitation and evaluation method of
stability
QC/T
The480–1999 automobile
simulation industry index
was evaluated of theto
according People’s Republic test
the serpentine of China [24]. The
evaluation test
method in
the automobile handling and stability index limitation and evaluation method of QC/T
480–1999 automobile industry index of the People’s Republic of China [24]. The test was
 evaluated and scored according to the average crosswind offset velocity p
the average steering-wheel-angle peak value, θ, at the reference speed.
The scoring formula of the average crosswind offset velocity peak
lows:
Energies 2022, 15, 393 15 of 29

𝑁 60 𝑟 𝑟
was evaluated and scored according to the average crosswind offset velocity peak value, r,
where Naverage
and the r is the evaluation score
steering-wheel-angle peak value
value, θ,of thereference
at the peak speed.
value of the average
velocity; r is the
The scoring experimental
formula of the averagevalue,
crosswind(°)
offset r100 is peak
/s; velocity the value,
upper r, islimit, which is
as follows:

is the lower limit, which is 25.0 (°) 40 /s.

Nr = 60 + (r60 − r ) (3)
The scoring formula of the raverage 60 − r100 steering-wheel-angle peak, θ, is
where Nr is the evaluation score value of the peak value of the average crosswind offset
40
velocity; r is the experimental value, (◦ )/s; r100 is the upper limit, which is 10.0 (◦ )/s; and
𝑁 60 𝜃 𝜃
◦
r60 is the lower limit, which is 25.0 ( )/s. 𝜃 𝜃
The scoring formula of the average steering-wheel-angle peak, θ, is as follows:
where Nθ is the evaluation score value 40
of the average steering-wheel-an
experimental value, (°); θθ100 is the N = 60 + ( θ − θ) (4)
θ60 upper
− θ100 60limit, which is 60.0 (°); and θ60 i
which
whereisNθ180.0 (°).
is the evaluation score value of the average steering-wheel-angle peak; θ is the
The comprehensive
experimental evaluation
value, (◦ ); θ 100 is the scoreis of
upper limit, which 60.0the carθ 60
(◦ ); and is isthe calculated
the lower limit, valu
◦
which is 180.0 ( ).
crosswind velocity peak value, r, and the average steering-wheel-angle
The comprehensive evaluation score of the car is the calculated value of the average
thecrosswind
reference speed.
velocity peak value, r, and the average steering-wheel-angle peak value, θ, at
the The
reference speed.
scoring formula for comprehensive evaluation of racing snake
The scoring formula for comprehensive evaluation of racing snake test is as follows:

2Nr + Nθ𝑁
2𝑁 𝑁
Ns = (5)
3 3
4.2.2. Analysis of Results without Crosswind Action
According toof
4.2.2. Analysis theResults
national standard
withoutevaluation method, the
Crosswind simulation of the snaking
Action
test needs to output the lateral acceleration curve, the peak curve of average crosswind
According
angle to the
velocity and the peak national standard
curve of average evaluation
steering-wheel method,
angle. The simulationthe simulati
curves
of 65, 70 and 75 km/h without crosswind are shown in Figures 12 and 13.
test needs to output the lateral acceleration curve, the peak curve of av
According to the requirements of national standard, the corresponding data of each
angle velocity
parameter and the peak
were intercepted into thecurve of average
simulation curve and steering-wheel angle.
data, and the simulation per- The s
of 65, 70 and 75 km/h without crosswind are shown in Figures 128.and 13
formance index data of 65, 70 and 75 km/h without crosswinds are shown in Table
According to the national standard, Nr , Nθ and Ns of snaking test are calculated. The
performance index score of each working condition is shown in Table 9.

Figure 12. Simulation curve of lateral acceleration without crosswind effect.

Figure 12. Simulation curve of lateral acceleration without crosswind effect.

Energies 2022, 15, x FOR PEER REVIEW 16 of 29

Energies 2022, 15, 393 16 of 29

(a) (b)
Figure
Figure 13.
13. Peak
Peak curve without crosswind
curve without crosswindeffect:
effect:(a)
(a)average
averagecrosswind
crosswindangle
anglevelocity
velocity and
and (b)(b) aver-
average
age steering-wheel angle.
steering-wheel angle.

TableAccording to the
8. Performance requirements
indexes of snakingof national
test at 65, 70standard, thewithout
and 75 km/h corresponding
crosswind.data of each
parameter were intercepted into the simulation curve and data, and the simulation per-
formance index dataAverage
of 65, 70 andLateral
75 km/h without Peak of Average are shown
crosswinds Peak in
of Table
Average
8. Ac-
Average Peak
Transit Time/s
cording to the national Crosswind Angle Steering-Wheel
standard, Nr, −N2 θ) and Ns of snaking test are calculated. The perfor-
Velocity/(km/h) Acceleration/(m.s
Velocity/(deg.s−1 ) Angle/deg.
mance index score of each working condition is shown in Table 9.
64.805 11.7 5.834 18.497 39.087
69.772 Table 8.
10.875Performance indexes of snaking
6.804 test at 65, 70 and 75 km/h
20.020 without crosswind.
40.833
74.734 10.150 7.845 Average Peak 21.538
Peak of Average Peak 42.572
of Average
Average Velocity Transit
Lateral Acceler- Crosswind Angle Steering-Wheel
/(km/h)
Table 9. Performance Timeof
index scores /s65, 70 and 75 km/h
ation/(m.s−2) serpentine
Velocity tests without
/(deg.s−1) crosswind.
Angle /deg.
64.805
Working Condition 11.7 Nr 5.834 Nθ 18.497 39.087
N s
69.772 10.875 6.804 20.020 40.833
65 km/h 77.341 100 (106.971) 84.894
74.734 10.150 7.845 21.538 42.572
70 km/h 73.280 100 (106.389) 82.187
Table 9. 75 km/h
Performance index scores69.891
of 65, 70 and 75 km/h100 (105.809)tests without crosswind.
serpentine 79.927

According to the data in Tables 8𝑵and

Working Condition 𝒓 9, the lateral𝑵acceleration
𝜽 𝑵𝒔 increases
of the car
65 km/h
and the passing time decreases with77.341
the increase of 100 (106.971)
the vehicle speed in the 84.894
three working
conditions. 70Thekm/h
average steering-wheel 73.280
angle peak, θ,100 of(106.389) 82.187
cars at different speeds is less than
the θ range75 specified
km/h in national standard, 69.891 leading to 100the(105.809)
scoring index of the79.927
peak angle θ
greater than 100. Because the steering transmission ratio of the FSAE cars is smaller than
that of civil cars,tothe
According thesteering-wheel-angle
data in Tables 8 andrange
9, theoflateral
the cars is small. of the car increases
acceleration
and the passing time decreases with the increase of the vehicle speed in the three working
4.2.3. Analysis of Simulation Results of Crosswind Action
conditions. The average steering-wheel angle peak, θ, of cars at different speeds is less
than This
the θpaper
rangestudies thein
specified snaking test
national simulation
standard, undertocrosswind
leading the scoringaction.
indexForof the
the open-
peak
loop snaking
angle θ greater test, the 100.
than steering system
Because theadopts
steeringa snaking test with
transmission direct
ratio input
of the FSAEa speed
cars of
is
65 km/h without crosswind. The variation curve of the steering-wheel
smaller than that of civil cars, the steering-wheel-angle range of the cars is small. angle is shown in
Figure 14.
4.2.3.In order toofrealize
Analysis the simulation
Simulation Results ofstudy of handling
Crosswind Actionand stability of the FSAE racing
car with and without crosswind, the car speed was set at 65 km/h and the crosswind angle
This paper studies the snaking test simulation under crosswind action. For the open-
was set at 90◦ . The car was subjected to the slope step crosswind with crosswind speed
loop snaking test, the steering system adopts a snaking test with direct input a speed of
of 12 m/s and ramp time of 0.1 s when driving for 0.2 s, and the acting time was until the
65 km/h without crosswind. The variation curve of the steering-wheel angle is shown in
end of the simulation. Figure 15 shows the comparison curves of aerodynamic forces and
Figure 14.
aerodynamic torque of racing cars with and without crosswind effects over time.
Energies 2022, 15, x FOR PEER REVIEW 17

Energies 2022, 15, 393 17 of 29

Figure 14. Input curve of steering-wheel angle in open-loop test.

In order to realize the simulation study of handling and stability of the FSAE r
car with and without crosswind, the car speed was set at 65 km/h and the crosswind
was set at 90°. The car was subjected to the slope step crosswind with crosswind sp
12 m/s and ramp time of 0.1 s when driving for 0.2 s, and the acting time was until th
of the simulation. Figure 15 shows the comparison curves of aerodynamic forces an
odynamic torque of racing cars with and without crosswind effects over time.
Figure 14. Input curve of steering-wheel angle in open-loop test.
Figure 14. Input curve of steering-wheel angle in open-loop test.

In order to realize the simulation study of handling and stability of the FSAE r
car with and without crosswind, the car speed was set at 65 km/h and the crosswind a
was set at 90°. The car was subjected to the slope step crosswind with crosswind spe
12 m/s and ramp time of 0.1 s when driving for 0.2 s, and the acting time was until th
of the simulation. Figure 15 shows the comparison curves of aerodynamic forces and
odynamic torque of racing cars with and without crosswind effects over time.

nergies 2022, 15, x FOR PEER REVIEW 18

Figure 15. Comparison curves of aerodynamic force and aerodynamic torque with and without
Figure 15. Comparison
crosswind curves
over time: C—A of aerodynamic force
is the crosswind-aerodynamic, and isaerodynamic
NC—A torque
the no crosswind with and wit
aerodynamic,
crosswind over time:
C is the crosswind andC—A is no
NC is the thecrosswind.
crosswind-aerodynamic, NC—A is the no crosswind aer
namic, C is the crosswind and NC is the no crosswind.

Figure 15 shows that, in the absence of crosswind, the aerodynamic side force
and the values of aerodynamic drag and aerodynamic lift are consistent with those ca
 Energies 2022, 15, 393 18 of 29

Figure 15 shows that, in the absence of crosswind, the aerodynamic side force is 0, and
the values of aerodynamic drag and aerodynamic lift are consistent with those calculated
by the fluid software; the aerodynamic lift and aerodynamic drag still have an impact
on the handling stability of the car. In crosswind condition, the aerodynamic force of the
car changes greatly. The aerodynamic lateral force changes the most, the aerodynamic
lift is the second and the aerodynamic drag is less affected. The racing curve always
fluctuates on both sides of the curve without crosswind. Influenced by the car’s serpentine
driving, the crosswind angle changes constantly and presents serpentine fluctuation, and
the aerodynamic curve also presents serpentine fluctuation.
Figure 15 shows the aerodynamic troque variation curve with and without crosswind
Energies 2022, 15, x FOR PEER REVIEWeffect over time. When there is no crosswind effect, the handling stability is only affected by 19 of
the pitching moment, the roll moment and yaw moment of the car are both 0, the maximum
pitching moment of the car is −64.883 N·m and the car presents an understeer state. When
the car is affected by crosswind, the roll moment and yaw moment both change. Because
the As
car can
is in be
theseen from
snaking test,Figure
the yaw 18, the maximum
moment changes the difference
most, and between
the minimum thevalue
crosswind
is an
velocity and the crest of the snaking track with and without
−55.739 N·m. In the simulation, the aerodynamic sideslip angle is between 30 and 40 , crosswind ◦ is achieved.
◦ T
peak value
the roll of the
moment at average
this angle crosswind angle
is small and the velocity,
minimum −8.807
Nris, is
value N·m. Whenindex
an important the rollto estim
moment reaches the peak value, the pitching moment and yaw
the snaking track. Now, the performance index of the snaking track under the action moment are the smallest,
and the pitching moment changes little, and the numerical fluctuation is on both sides of
slope step crosswind at 12 m/s and slope time of 0.1 s is evaluated. As the steering-wh
the pitching moment curve in the non-crosswind condition.
angle is too small,
Through Figurethe16,evaluation
it can be seenscore of theacceleration
that lateral average peak change steering-wheel
with and without angle, Nθ
100 in both is
crosswinds working conditions.
small. Lateral Theisfinal-score
acceleration data ofbythe
mainly determined the two working
longitudinal conditions
velocity
shown in Table
and driving radius 10.under
It canthebe seen
action of from Table 10
the crosswind car that the comprehensive
longitudinal velocity change performance
being
smaller while
reduced underthethe lateral
actiondisplacement increases;
of crosswind, andhowever, the steering-wheel
the crosswind angle influence
has a great did not on
change, and the car
stability characteristics. driving radius basically remained unchanged, Therefore, the maximum
lateral acceleration difference between the two conditions occurs at the wave peak of the
serpentine driving route, and the maximum lateral acceleration difference is 0.2 m/s2 .

Figure 16. Comparison curve and difference curve of lateral acceleration with and without cross-
Figure 16. Comparison curve and difference curve of lateral acceleration with and without cro
wind effect.
wind effect.
It can be seen from Figure 17 that the curve of roll speed change and its difference
presents a sinusoidal trend: the roll speed curve of the racing car under the effect of cross-
wind fluctuates significantly when the crosswind is introduced. The maximum difference
of the roll speed in the two working conditions also occurs in this period, and the max-
imum difference is −0.283 deg/s. When the attitude of the car tends to be stable, the
rear-roll-speed curve basically coincides with that without crosswind.
 Figure 16. Comparison curve and difference curve of lateral acceleration with and without cros
Energies 2022, 15, 393 wind effect. 19 of 29

Figure 17. Comparison curves and difference curves of roll angle velocity with and without cross-
Figure
wind 17. Comparison curves and difference curves of roll angle velocity with and without cros
effect.
wind effect.
As can be seen from Figure 18, the maximum difference between the crosswind angle
velocity and the crest of the snaking track with and without crosswind is achieved. The
peak value of the average crosswind angle velocity, Nr , is an important index to estimate
the snaking track. Now, the performance index of the snaking track under the action of
slope step crosswind at 12 m/s and slope time of 0.1 s is evaluated. As the steering-wheel
angle is too small, the evaluation score of the average peak steering-wheel angle, Nθ , is
Energies 2022, 15, x FOR PEER REVIEW 100 in both working conditions. The final-score data of the two working conditions are 20 of
shown in Table 10. It can be seen from Table 10 that the comprehensive performance is
reduced under the action of crosswind, and the crosswind has a great influence on the
stability characteristics.

Figure 18. Comparison curves and difference curves of crosswind offset velocity with and without
Figure 18. Comparison curves and difference curves of crosswind offset velocity with and witho
crosswind action.
crosswind action.
Table 10. Performance index score of snaking test with and without crosswind effect.
Table 10. Performance index score of snaking test with and without crosswind effect.
Working Condition Nr Nθ Ns
Working
No crosswind Condition
65 km/h 77.341 𝑵𝒓 100 𝑵𝜽 84.894 𝑵𝒔
No crosswind
Crosswind 65 km/h65 km/h 74.876 77.341 100 100 83.251 84.894
Crosswind 65 km/h 74.876 100 83.251
As communicated by the above analysis, when the crosswind direction changes, the
movement track of the car has a tendency to go back, but the horizontal offset continues to
As communicated
increase. by the
The car’s roll angle above
velocity analysis,
and when
crosswind anglethe crosswind
velocity direction
in the same side of changes,
the th
movement track of the car has a tendency to go back, but the horizontal offset continu
to increase. The car’s roll angle velocity and crosswind angle velocity in the same side
the monsoon period reached the maximum value. If the driver has improper operatio
the car will leave the track, and the car’s stability will be affected severely.
 Energies 2022, 15, 393 20 of 29

monsoon period reached the maximum value. If the driver has improper operation, the car
will leave the track, and the car’s stability will be affected severely.

5. Modification of Aerodynamic Characteristics

The aerodynamic and dynamics system of the FSAE racing car were joint simulated
to explore the influence of different aerodynamic conditions on the handling stability
of the racing car, and the performance indexes of the racing car under each working
condition were evaluated and analyzed. The car model of the study was defined as base
car, and the problem that aerodynamic parameters of the car fluctuated greatly when the
body attitudes changed in the joint simulation of base car was summarized and analyzed.
The aerodynamic parameters of the vehicle were adjusted built on the pitch angle of
FSAE car based on the handling stability, and the dynamics simulation verification of
the adjusted car was conducted to explore the influence of aerodynamic modification on
the vehicle performance. After the modification plan was determined, the aerodynamic
parameters after modification were analyzed and evaluated to explore the influence of
each aerodynamic component on the aerodynamic coefficient. Finally, the aerodynamic
parameters were allocated to each aerodynamic component of the racing car to obtain the
downforce interval of the aerodynamic component of the adjusting plan, which provides
guidance and suggestions for the design of the aerodynamic component of the racing car.

5.1. Aerodynamic Force Modification

Based on the aerodynamic force of the base car to analyze and adjust, the aerodynamic
torque parameters remain unchanged. The pitch angle of the FSAE car varies from −2.0◦
to 1.0◦ in the process of movement. Figure 19 shows the curves of the lift coefficient, drag
Energies 2022, 15, x FOR PEER REVIEW
coefficient and lift-to-drag ratio of the car. In order to more intuitively display the variation
trend of aerodynamic coefficient, the drag coefficient in Figure 19 is the negative of the
simulation data.

Figure 19. Curves of lift coefficient, drag coefficient and lift–drag ratio.
Figure 19. Curves of lift coefficient, drag coefficient and lift–drag ratio.
Through Figure 19, with the decrease of pitch angle, the aerodynamic drag increases
gradually, and theFigure
Through aerodynamic downforce
19, with coefficient of
the decrease increases
pitchfirst and then
angle, decreases.
the aerodynamic
d
The lift–drag ratio curve reaches the extreme value of −1.888 when the pitch angle is
gradually, and the aerodynamic downforce coefficient increases first and th
The lift–drag ratio curve reaches the extreme value of −1.888 when the pitch
The lift–drag ratio is defined as the ratio of lift coefficient to drag coefficien
culation formula is shown in Equation (6):
 Through Figure 19, with the decrease of pitch angle, the aerodynamic drag
gradually, and the aerodynamic downforce coefficient increases first and then d
The lift–drag ratio curve reaches the extreme value of −1.888 when the pitch angl
Energies 2022, 15, 393 The lift–drag ratio is defined as the ratio of lift coefficient to drag coefficient,
21 of 29 an
culation formula is shown in Equation (6):

−1.0◦ . The lift–drag ratio is defined as the ratio of 𝑆lift coefficient to drag coefficient, and its
calculation formula is shown in Equation (6):
According to the above formula, the lift-to-drag ratio can reflect the aero
L Cl
efficiency of the car. Under a certain S=
D
=downforce,
Cd the larger the lift-to-drag (6) ra
higher the aerodynamic efficiency of the vehicle is, and this is also the ultima
According to the above formula, the lift-to-drag ratio can reflect the aerodynamic
aerodynamic design of FSAE racing car. Therefore, in the pitch angle range, th
efficiency of the car. Under a certain downforce, the larger the lift-to-drag ratio is, the
namic
highercharacteristics and aerodynamic
the aerodynamic efficiency efficiency
of the vehicle is, and this is are
also the best when
the ultimate goal ofthe pitch is
aero-
Afterdesign
dynamic the optimal pitchcar.
of FSAE racing angle of theinaerodynamic
Therefore, efficiency
the pitch angle range, of the car in the
the aerodynamic
characteristics and aerodynamic efficiency are the best when the pitch is − 1.0 ◦.
range is obtained, the aerodynamic parameters are adjusted. After the modific
After the optimal pitch angle of the aerodynamic efficiency of the car in the pitching
working condition
range is obtained, the of the car is parameters
aerodynamic the base car are that the After
adjusted. pitchthe
angle is −1.0; this
modification, the plan i
asworking
modification
conditionplay
of the 1,
carand
is thethe
baseaerodynamic
car that the pitchparameters
angle is −1.0;ofthisthe
plan car after modific
is defined
shown in Figure
as modification play20.
1, and the aerodynamic parameters of the car after modification are
shown in Figure 20.

Figure 20. Curves of lift coefficient, drag coefficient and lift-to-drag ratio after adjusting aerodynamic
parameters in pitching condition.

According to Figure 20, after adjusting the aerodynamic parameters in pitch condition,
the aerodynamic efficiency of the car is the best in the basic condition, indicating that
the aerodynamic characteristics of the car remain in the best state most of the time in the
driving process.
The influence of the adjusted aerodynamic parameters on the handling stability is
verified by dynamics simulation. The speed-change interval of the car in the figure-eight
surround project is small; under this condition, the test is the ultimate cornering speed of
the car and the anti-roll ability of the car. When the aerodynamic downforce generated by
the aerodynamic device acts on the tire, the ground lateral reaction force and tangential
reaction force of the car wheel are increased, the ultimate lateral acceleration of the car is
increased and the time of the figure-eight surround is shortened. Therefore, aerodynamic
characteristics have a crucial impact on the figure-eight surround. The figure-eight surround
layout is shown in Figure 21.
 the aerodynamic device acts on the tire, the ground lateral reaction force an
reaction force of the car wheel are increased, the ultimate lateral acceleration
increased and the time of the figure-eight surround is shortened. Therefore, a
Energies 2022, 15, 393
characteristics have a crucial impact on the figure-eight surround. The figu
22 of 29
round layout is shown in Figure 21.

Figure 21.Figure-eight
Figure 21. Figure-eightsurround test layout.
surround test layout.
In this simulation, the radius of the figure-eight surround track is 10.625 m, and the
roadIn thiscoefficient
friction simulation, the
is set as 1.0.radius of the
The test stops whenfigure-eight
the car passessurround trackInis 10.62
the track smoothly.
order to simplify the test, closed-loop automatic shift control
road friction coefficient is set as 1.0. The test stops when the car passes the trais adopted in this simulation.
In steering control, 0.15 s pre-sighting steering is adopted. The longitudinal speed of the
Incar
order
is the to simplify
same, 48.9 km/h. the test, closed-loop automatic shift control is adopted in
tion. In steering
Figures 22 andcontrol, 0.15 s pre-sighting
23 are the comparison steeringsurround
curves of figure-eight is adopted.
simulationThetrack
longitud
route, aerodynamic force
the car is the same, 48.9 km/h. and steering-wheel angle before and after adjusting aerodynamic
parameters. In Figure 22, the racing contrast curve can be seen; when two simulations of
Figuresvelocity
longitudinal 22 and are 2348.9are
km/h,thethecomparison
basic conditioncurves of the carofin figure-eight
the 7.975 s, the car surround
track
pulls route, aerodynamic
off the track and the simulationforceends;
andafter steering-wheel angle before
adjusting the aerodynamic and in
parameters after ad
pitch condition, the figure-eight surround simulation can
dynamic parameters. In Figure 22, the racing contrast curve can be seen; whe be successfully completed under
the same dynamic parameters and longitudinal velocity. According to the aerodynamic
lations of longitudinal
force comparison curve of the velocity are 48.9
car in Figure km/h,
23, neither of the
the twobasic condition
simulated ofthe
cars left the car in
the car
track pulls
in the firstoff
7.0 s.the track and
Compared thebase
with the simulation
condition, the ends; after adjusting
aerodynamic parameter plan the aero
of the car under the pitch condition significantly improved
rameters in pitch condition, the figure-eight surround simulation can be succe the downforce of the car, with
the maximum value of −303.364 N. Compared with the base condition, the downforce of
pleted
the car under
increasedthe sameThe
by 8.56%. dynamic
drag of theparameters
adjusted plan and is lower longitudinal velocity. Acco
than the base condition,
aerodynamic force
so the lift-drag ratio of comparison
the adjusted plan curve
is much ofhigher
the car than intheFigure 23, neither
base condition, and theof the tw
aerodynamic efficiency and aerodynamic characteristics of the
cars left the track in the first 7.0 s. Compared with the base condition, the a adjusted plan are better than
the base condition. It can be seen from the steering-wheel-angle curve that the steering-
parameter
wheel-angle plan
curvesof thetwo
of the carsimulations
under the pitchoverlapped
basically condition significantly
before improved th
6.0 s, thus indicating
ofthat
thethecar, with theparameter
aerodynamic maximum value of
modification −303.364
plan N. Compared
has little influence withWhen
on the driver. the base c
downforce of the car increased by 8.56%. The drag of the adjusted will
the car is actually running and the driver has similar skills, the performance of the car plan is lo
be greatly improved.
base According
condition, so analysis
to the the lift-drag ratio of
and simulation, whenthetheadjusted
aerodynamic plan is much
parameters higher th
are ad-
condition, and the
justed, the ultimate aerodynamic
cornering efficiency
speed of the car is improvedand when
aerodynamic
the dynamics characteristics
parameters of
of the car are the same, the car of figure-eight surround project can drive at a faster speed
on the track and the time is reduced.
 plan are better than the base condition. It can be seen from the steering-wheel-angle curve
that the steering-wheel-angle curves of the two simulations basically overlapped before
6.0 s, thus indicating that the aerodynamic parameter modification plan has little influence
on the driver. When the car is actually running and the driver has similar skills, the per-
Energies 2022, 15, 393 formance of the car will be greatly improved. 23 of 29

Figure 22. Comparison curve of simulation trajectory of figure-eight surround with aerodynam
parameters modification.
Figure 22. Comparison curve of simulation trajectory of figure-eight surround with aerodynamic
parameters modification.
Figure 22. Comparison curve of simulation trajectory of figure-eight surround with aerodynamic
parameters modification.

Figure 23. Comparison curve of aerodynamic force and steering–wheel angle in figure-eight sur-
Figure
round 23. Comparison curve of aerodynamic force and steering–wheel angle in figure-eight su
simulation.
Figure 23. Comparison curve of aerodynamic force and steering–wheel angle in figure-eight sur-
round simulation.
round simulation.
5.2. Aerodynamic Torque Modification
Based on the
According modification
toanalysis
the analysis ofand
aerodynamic
simulation, parameters,
when thethe aerodynamic parameters
aerodynamic torque pa- are a
Energies 2022, 15, x FOR PEER REVIEW
According to the and simulation, when the aerodynamic parameters
rameters of the car are further adjusted. As can be seen from Figure 15, when there are
24
ad- is no
of 29
justed,
justed, the the ultimate
ultimate cornering
cornering speed speed
of the of
car isthe car
improved is improved
when the when
dynamics the dynamics
parameters paramete
crosswind action, the yaw moment and roll moment of the dynamic simulation car are
of of
thethe
car car
both 0, are
aresothe
onlythethesame,
same, the carthe
pitching car of figure-eight
ofmoment
figure-eight surround
of the surround
project in
car is adjusted project
canthis
drive at can drive
a faster
section. Figure atisathe
speed
24 faster spe
ononthe track and
the track and
comparison the time
curvethe is reduced.
time
of the is reduced.
pitching moment coefficient modification.

5.2. Aerodynamic Torque Modification

5.2. Aerodynamic Torque Modification
Based on the modification of aerodynamic parameters, the aerodynamic torque pa-
rametersBased
of the on
car the modification
are further adjusted.ofAs
aerodynamic
can be seen fromparameters, the aerodynamic
Figure 15, when there is no torque p
rameters
crosswind of the
action, thecar
yawaremoment
furtherand
adjusted. As can
roll moment be dynamic
of the seen from Figure 15,
simulation car when
are there is
crosswind
both 0, so only action, the moment
the pitching yaw moment and
of the car roll moment
is adjusted in this of the dynamic
section. Figure 24 simulation
is the car a
comparison
both 0, socurve
onlyofthe
the pitching
pitching moment
moment coefficient modification.
of the car is adjusted in this section. Figure 24 is t
comparison curve of the pitching moment coefficient modification.

Figure 24. Comparison curve of pitching moment coefficient modification.

Figure 24. Comparison curve of pitching moment coefficient modification.

It can be seen from Figure 24 that two pitching moment coefficient adjustment plans
with different understeering characteristics are proposed based on base conditions, and
the understeering characteristics of Modification Plan 2 are slightly greater than that of
 Figure 24. Comparison curve of pitching moment coefficient modification.

Energies 2022, 15, 393 It can be seen from Figure 24 that two pitching moment coefficient adjustment 24 of 29 pla
with different understeering characteristics are proposed based on base conditions, an
the understeering characteristics of Modification Plan 2 are slightly greater than that
Modification
It can bePlan 3. The
seen from coefficient
Figure 24 that twofluctuation of the coefficient
pitching moment two plansadjustment
is less thanplans that of th
base
withcondition. The basic body
different understeering attitude of
characteristics arethe car is the
proposed basedbest
on attitude of the aerodynam
base conditions, and
the understeering characteristics of Modification Plan 2 are slightly
characteristics of the car. The coefficient fluctuation of different pitching angles greater than that of is sma
Modification Plan 3. The coefficient fluctuation of the two plans is less than that of the
the comprehensive adaptability of the car is stronger and the sensitivity is lower.
base condition. The basic body attitude of the car is the best attitude of the aerodynamic
The influence
characteristics of theofcar.
theThe
adjusted aerodynamic
coefficient fluctuation oftorque
differentparameters on is
pitching angles the handling
small, the stab
itycomprehensive
of the car wasadaptability
verified. In of this simulation,
the car is stronger the
and longitudinal
the sensitivity isspeed
lower.of the car with diffe
ent modification
The influenceplans
of theisadjusted
the same, which istorque
aerodynamic 49 km/h, and the
parameters onspeed increases
the handling by 0.1 km/
stability
of the25
Figure carshows
was verified. In this simulation,
the comparison curvethe longitudinal
between speed of the
the pitching car with
moment ofdifferent
the car adjuste
bymodification
the pitching plans is the same,
moment whichand
coefficient is 49the
km/h, and the speedangle.
steering-wheel increases by 0.1 km/h.
Figure 25 shows the comparison curve between the pitching moment of the car adjusted by
the pitching moment coefficient and the steering-wheel angle.

Figure 25. Comparison curve of pitching moment and steering-wheel angle.

Figure 25. Comparison curve of pitching moment and steering-wheel angle.
It can be seen from the curve of the steering-wheel angle that, when the car runs at
a longitudinal speed
It can be seen of 49
from km/h,
the curve theofcar
theinsteering-wheel
base condition rolls over
angle at 7.975
that, when s, and the runs at
the car
car in Modification Plan 2 rolls off the track at 8.775 s. The simulation of Modification
longitudinal speed of 49 km/h, the car in base condition rolls over at 7.975 s, and the c
in Plan 3 is successfully completed. Through the simulation, it can be seen that the limit
Modification Plan 2 rolls off the track at 8.775 s. The simulation of Modification Plan
cornering speed of the car in the base condition of the figure-eight surround simulation
is is
successfully
less than 49 completed.
km/h. The car Through the simulation,
in Modification it can
Plan 2 still be off
drives seenthethat the
track butlimit
runscornerin
speed
0.8 s more than that in the base condition, while the simulation of Modification Plan 3 is
of the car in the base condition of the figure-eight surround simulation is less tha
49successfully
km/h. Thecompleted.
car in Modification
It can be seenPlan
from2the
still drives
curve off themoment
of pitching track but runs 0.8that
comparison s more tha
the pitching moment of Modification Plan 3 is the smallest, but it is still negative, which
indicates that, the greater the degree of understeer in the figure-eight surround condition,
the worse the passing performance of the car, but the oversteer will still have a poor impact
on the car, and the driver’s correction of the car will become worse.
The comprehensive analysis of the curves shows that the ultimate cornering speed
of the car is further improved under the condition that the dynamics parameters and
aerodynamic parameters of the car are the same after the modification of aerodynamic
torque parameters, but the increase is less than that of the car after the modification of
aerodynamic parameters.
In order to further explore the influence of modification of aerodynamic torque param-
eters on the handling stability of the car, the fixed closed-loop snaking test simulation was
carried out for Modification Plan 2 and Modification Plan 3, and the longitudinal speed of
the car is 65 km/h. Figure 26 shows the simulation steering-wheel angle and difference
curve of snaking test adjusting pitching moment parameters.
 In order to further explore the influence of modification of aerodynamic torque p
rameters on the handling stability of the car, the fixed closed-loop snaking test simulatio
was carried out for Modification Plan 2 and Modification Plan 3, and the longitudin
Energies 2022, 15, 393
speed of the car is 65 km/h. Figure 26 shows the simulation steering-wheel 25angle
of 29
and di
ference curve of snaking test adjusting pitching moment parameters.

Figure 26. Comparison curve of steering-wheel angle and difference of snaking test.
Figure 26. Comparison curve of steering-wheel angle and difference of snaking test.
It can be seen from the curve for the steering-wheel angle that the peak values of
It can be seen
steering-wheel anglefrom the curvePlan
of Modification for 2the
andsteering-wheel
Modification Planangle thatsmaller
3 are both the peak
thanvalues
the base condition. According to the difference curve, the peak value of the
steering-wheel angle of Modification Plan 2 and Modification Plan 3 are both smaller tha steering-wheel
theangle
baseofcondition.
Modification Plan 3 is the
According to smallest, and thecurve,
the difference maximum the difference
peak value between
of theModifi-
steering-whe
cation Plan 3 and base condition is 0.5756◦ , which indicates that other parameters of the car
angle of Modification Plan 3 is the smallest, and the maximum difference
remain unchanged in the serpentine closed-loop test, the steering-wheel angle is reduced
between Mod
fication
and thePlan 3 and
driver basemore
operates condition is 0.5756°,
conveniently which
after the indicates
modification of that other parameters
aerodynamic torque. of th
car remain
Whenunchanged
the dynamicsin the serpentine
parameters of the carclosed-loop
are the same, test, the steering-wheel
the ultimate cornering speedangle
is is r
duced and slightly,
increased the driver operates more
the steering-wheel conveniently
angle of the snakingafter the
test is modification
reduced of aerodynam
and the handling
performance of the car is improved when other parameters remain unchanged.
torque.
5.3.When the dynamics
Final Modification Plan parameters of the car are the same, the ultimate cornering spee
is increased slightly, the steering-wheel angle of the snaking test is reduced and the han
The aerodynamic parameter plan under the final modification pitching condition was
dling performance
analyzed by using of the carsimulation,
dynamic is improved
andwhen
the carother
beforeparameters
modificationremain unchanged.
was compared.
Dynamics simulation is the figure-eight surround condition of FSAE race, the longitudinal
5.3.speed
FinalofModification Plan
the car in base condition is 48 km/h and the modification plan is 49 km/h.
Figures 27 and 28 show the comparison curves of aerodynamic force, steering-wheel
The aerodynamic parameter plan under the final modification pitching conditio
angle, roll angle velocity and lateral acceleration before and after aerodynamic parameters
was analyzed
under bypitch
different using dynamic conditions.
modification simulation, Asand theseen
can be carfrom
before modification
Figure 27, when thewas com
pared. Dynamics simulation is the figure-eight surround condition
driving speed of the car is reduced by 1 km/h, the aerodynamic downforce in base of FSAE race, the lon
condition
gitudinal
is far lessspeed of of
than that the
thecar in base
adjusted car, condition is 48 km/h
and the maximum andofthe
difference modification
downforce is 29.05%plan is 4
of
km/h. the pressure in base condition. As the wind speed decreases, the drag of the car in
base condition is slightly less than that of the adjusted car. The aerodynamic
Figures 27 and 28 show the comparison curves of aerodynamic force, steering-whe force of the
car fluctuates significantly in 0–2.0 s and 6.0–8.0 s. The aerodynamic force fluctuation of
angle, roll angle velocity and lateral acceleration before and after aerodynamic paramete
the car in 0–2.0 s is caused by the unstable speed control of the car in the initial stage. It
under
can bedifferent
seen from pitch modification conditions.
the steering-wheel-angle Asthe
curve that can be seen from
steering-wheel Figure
angle of the27,
carwhen th
around 6.0 s is close to 400◦ in a short time, and the car enters the right circle from the
left circle; the aerodynamic force fluctuates slightly, due to the change of body attitude.
According to the analysis of the roll-angle-velocity curve in Figure 28, the car body’s roll
angle fluctuates significantly in the three stages of the car that include turning into the
left lap, entering the right lap from the left lap and turning back out of the right lap. The
adjusted car’s transient response is relatively rapid. The roll-angle-speed curve shows that
the car tires with the ground appeared near 8.0 s brief separation after the modification.
Since the longitudinal velocity is greater than that in base condition, when the car starts to
turn, the lateral acceleration of the car after adjusting the aerodynamic parameters in pitch
condition is greater than that in base condition.
 to
theturn, the lateral
car tires acceleration
with the of the carnear
ground appeared after
8.0adjusting the aerodynamic
s brief separation after theparameters
modificatio
pitch condition is greater than that in base condition.
Since the longitudinal velocity is greater than that in base condition, when the car sta
to turn, the lateral acceleration of the car after adjusting the aerodynamic parameters
Energies 2022, 15, 393
pitch condition is greater than that in base condition. 26 of 29

Figure 27. Comparison curves of aerodynamic force and steering-wheel angle of figure-eight su
round simulation with different aerodynamic parameters.
Figure 27. Comparison curves of aerodynamic force and steering-wheel angle of figure-eight sur-
Figure
round 27. Comparison
simulation curvesaerodynamic
with different of aerodynamic force and steering-wheel angle of figure-eight s
parameters.
round simulation with different aerodynamic parameters.

Figure 28. Comparison curves of roll angle velocity and lateral acceleration of figure-eight surround
Figure 28. Comparison curves of roll angle velocity and lateral acceleration of figure-eight surroun
simulation with different aerodynamic parameters.
simulation with different aerodynamic parameters.
FigureTable 11 shows the
28. Comparison comparison
curves data velocity
of roll angle of performance parameters
and lateral of the
acceleration offigure-eight
figure-eight surrou
Table
surround 11 shows
simulation the
beforecomparison
and after data
adjusting
simulation with different aerodynamic parameters. of performance
aerodynamic parameters
parameters in of the
pitching figure-eig
condi-
tion. According to the comprehensive analysis of the curve and Table 11, after
surround simulation before and after adjusting aerodynamic parameters in pitching co adjusting the
aerodynamic parameters under pitching condition, the speed of the car’s figure-eight sur-
dition. According
Table 11 shows to the
thecomprehensive
comparison data analysis of the curve
of performance and Table of
parameters 11,the
after adjustin
figure-eig
round is increased by 1 km/h, and the downforce of the race car is increased by −53.693 N
surround simulation before and after adjusting aerodynamic parameters
around 6.0 s; the difference is about 21.50% of the base race car’s downforce. The lateral in pitching co
dition. According
acceleration of the to
carthe comprehensive
increases by 0.25 g at analysis
4.0 s, and of
thethe curve lateral
increased and Table 11, after
acceleration is adjusti
about 15.11% of the base car’s lateral acceleration. Finally, after modifying the aerodynamic
parameter scheme under pitching condition, the simulation time of the racing car is 11.3 s
and the base condition is 11.825 s. After adjusting the aerodynamic parameters under
pitching condition, the running time of the racing car is shortened by 0.525 s, about 4.40%
of the total time. After adjusting the aerodynamic parameters under pitching condition,
under the same dynamic parameters of racing car, the overall performance of the car is
greatly improved.
In this section, after the analysis and study of the racing car with the aerodynamic
parameters adjusted under the pitching condition, it can be known that the aerodynamic
downforce of the whole vehicle increases by 11.39%, the drag decreases by 2.85%, the
lift-to-drag ratio increases by 14.70% and the pitching moment decreases by 82.34%.
Energies 2022, 15, 393 27 of 29

Table 11. Comparison of performance parameters of figure-eight surround simulation before and
after adjusting aerodynamic parameters under pitching condition.

Lateral
Aerodynamic Lift /N Simulation Time/s
Acceleration/g
Base car −249.782 1.654 11.825
Modification plan −303.475 1.904 11.3
Parameter difference −53.693 0.25 −0.525
Differential ratio 21.50% 15.11% −4.40%

6. Conclusions
This study took the FSAE racing car of Jilin University as the research object. It
combined the wind-tunnel test for the racing car, aerodynamic numerical simulation of
the simplified model of the racing car and dynamics simulation of the racing car system;
established the aerodynamic closed-loop design process of the FSAE racing car; and realized
the joint research on aerodynamic characteristics and handling stability of the racing car in
different body attitudes. The conclusions are summarized as follows:
(1) As the pitch conditions of body attitude change, the vehicle aerodynamic character-
istics undergo great changes: pitch racing drag under the condition of less change,
least drag when the body forward 2.0◦ , maximum drag when the body back is 1.0◦ ,
body leaning forward negative lift is the largest, 1.0◦ in the car body back 0.5◦ in the
least stressed nowadays and body leaning forward 1.0◦ vehicle lift-to-drag ratio is the
largest. The aerodynamic efficiency was the highest, and the lift-drag ratio was the
lowest when the body reclined 0.5◦ . By comparing the pressure cloud diagram, it can
be seen that the aerodynamic characteristics of the vehicle change due to the change
of the ground clearance of the front wing and tail wing after the body attitude change
and the interaction between the aerodynamic components, thus indicating that the
aerodynamic components have a great impact on the aerodynamic characteristics of
the vehicle.
(2) Through the simulation of crosswind condition, when the crosswind angle β is be-
tween 0◦ and 40◦ , the aerodynamic characteristics of the car are significantly reduced,
but additional downforce can still be generated. The aerodynamic suite can improve
the handling stability of the car. When the crosswire angle β is greater than 40◦ , the
aerodynamic suite of the car fails. A crosswind angle β with 0◦ ~40◦ can be regarded
as the effective working range of racing aerodynamic kit under crosswind. Three
representative crosswind deflection angles were selected in this paper to analyze and
study the vehicle outflow field structure.
(3) The aerodynamic parameters obtained from the numerical simulation were input into
the dynamics model. Under the action of crosswind snaking test simulation, the results
show that the aerodynamic lift and lateral force with the crosswind input generate
step change, aerodynamic drag is relatively small, the influence of the aerodynamic
yawing moment has the biggest change and the change of the car’s aerodynamic
pitching moment is small. When the aerodynamic yaw moment reaches the peak
value, the aerodynamic pitching moment and aerodynamic yaw moment are at the
minimum. Under the crosswind effect, the car deviates from the predetermined track,
and the lateral offset far exceeds the safety limit. The aerodynamic lateral force and
yaw moment under the action of crosswind have a great influence on the handling
stability of the car. When the driver encounters strong crosswind during driving, the
car will leave the track if the operation is improper, and the stability of the car will be
seriously affected.
(4) After completing the modification of aerodynamic force, the results show that the
critical cornering speed of the FSAE car is improved after the aerodynamic force
adjustment. After completing the aerodynamic torque modification, the ultimate
Energies 2022, 15, 393 28 of 29

cornering speed is increased slightly under the condition of the same dynamics
parameters of the car, the steering-wheel angle of the snaking test is reduced and the
handling performance of the car is improved. In the final aerodynamic modification
plan under the final modification pitching condition, the aerodynamic downforce of
the whole car is increased by 11.39%, aerodynamic drag is reduced by 2.85%, lift-to-
drag ratio is increased by 14.70% and pitching moment is reduced by 82.34%. The
aerodynamic characteristics and aerodynamic efficiency of the car are significantly
improved, and the aerodynamic performance of the car is greatly improved. The
running time of the figure-eight surround test car is shortened by 0.525 s, about
4.40% of the total time, thus proving that aerodynamic modification can improve
the handling stability of the car, and the overall performance of the car can also be
greatly improved.

Author Contributions: Conceptualization, Y.Z. and Z.Z.; methodology, Z.Z. and S.S.; software, Q.W.
and S.S.; validation, Z.Z.; formal analysis, Z.Z.; investigation, Y.Z.; resources, Z.Z.; data curation,
S.S.; writing—original draft preparation, Q.W.; writing—review and editing, Q.W.; visualization,
Q.W. and Z.Z.; supervision, Y.Z.; project administration, Y.Z.; funding acquisition, L.R. and C.Z. All
authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by National Key Research and Development Program of China
(Grant No. 2018YFA0703300) and National Natural Science Foundation of China through (Grant
No.11772140 and 11702109).
Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Qiao, J. Formula SAE of China. Automot. Eng. 2012, 2, 146.
2. Katz, J. Aerodynamics of race cars. Annu. Rev. Fluid Mech. 2006, 38, 27–63. [CrossRef]
3. Al Muharrami, M.S.; Almousa, S.M.; Almutairi, A.F.M. Study of Open and Closed Wheels Aerodynamics of Racing Cars in Wind
Tunnels with Movable Ground at Different Car Speeds. In Proceedings of the 18th Australasian Fluid Mechanics Conference
Launceston, Launceston, Australia, 3–7 December 2012.
4. Hetawal, S.; Gophane, M.; Ajay, B.K.; Mukkamala, Y. Aerodynamic Study of Formula SAE Car. Procedia Eng. 2014, 97, 1198–1207.
[CrossRef]
5. Craig, C.; Passmore, M.A. Methodology for the Design of an Aerodynamic Package for a Formula SAE Vehicle. SAE Int. 2014, 7,
575–585. [CrossRef]
6. Weingart, R. On-Track Testing as a Validation Method of Computational Fluid Dynamic Simulations of a Formula SAE Vehicle.
Master’s Thesis, The Graduate Faculty of the University of Kansas, Lawrence, KS, USA, 2015.
7. He, H. Research on Crosswind Stability and Control Strategy of FSAE Racing Car; Harbin Institute of Technology: Harbin, China, 2017.
8. Lu, X. Research on the Flow Field around a Formula SAE Car. SAE Tech. Pap. 2015, 26, 208.
9. Zhang, Y.; Yang, C.; Wang, Q.; Zhan, D.; Zhang, Z. Aerodynamics of Open Wheel Racing Car in Pitching Position. SAE Tech. Pap.
2018, 1, 729.
10. Wasik,
˛ M.; Skarka, W. Aerodynamic Features Optimization of Front Wheels Surroundings for Energy Efficient Car. Adv.
Transdiscipl. Eng. 2016, 4, 483–492.
11. De Vita, A.; di Angelo, L.; Andreassi, L.; Romagnuolo, S. CFD-Aided Design of an Airbox for Race Cars. SAE Tech. Pap. 2002, 1, 2167.
12. Li, J.; Zhang, Z.; Zhang, Y.; Jiang, Z. Research on Effects of Crosswind on Handling and Stability of Vehicle with High Speed.
SAE-China 2010, 172, 1380–1383.
13. Carbonne, L.; Winkler, N.; Efraimsson, G. Use of Full Coupling of Aerodynamics and Vehicle Dynamics for Numerical Simulation
of the Crosswind Stability of Ground Vehicles. SAE Int. J. Commer. Veh. 2016, 9, 359–370. [CrossRef]
14. Jia, L.J.; Han, Z.H.; Li, G. Simulation Study of FSC Racing Car Modeling and Handling Stability Based on CarSim. J. Liaoning
Univ. Technol. (Nat. Sci. Ed.) 2015, 35, 172–176.
15. Huang, T.M.; Gu, Z.-Q.; Chen, Z.; Wen, Q.; Luo, Z.-X. Influence of Vehicle Body in Pitching Motion on Aerodynamic Lift Force.
China J. Highw. Transp. 2017, 30, 150–158.
16. Guerrero, A.; Castilla, R. Aerodynamic Study of the Wake Effects on a Formula 1 Car. Energies 2020, 13, 5183. [CrossRef]
17. An, Y. Research on the Aerodynamic Drag Reduction of the Passenger Car based on Rear Accessories; Jilin University: Changchun, China, 2012.
18. Cheng, S.Y.; Tsubokura, M.; Nakashima, T.; Okada, Y.; Nouzawa, T. Numerical quantification of aerodynamic damping on
pitching of vehicle-inspired bluff body. J. Fluids Struct. 2012, 30, 188–204. [CrossRef]
Energies 2022, 15, 393 29 of 29

19. Martins, D.; Correia, J.; Silva, A. The Influence of Front Wing Pressure Distribution on Wheel Wake Aerodynamic of a F1 Car.
Energies 2021, 14, 4421. [CrossRef]
20. Zhu, Z.-W. Feasibility Investigation on Prediction of Vortex Shedding of Flat Box Girders Based on 2D RANS Model. China J.
Highw. Transp. 2015, 28, 24–33.
21. Cao, H. Optimization of Aerodynamic Drag Reduction for Flat-head Commercial Vehicles considering the Influence of Crosswind; Jilin
University: Changchun, China, 2020.
22. Jessing, C.; Stoll, D.; Kuthada, T.; Wiedemann, J. New horizons of vehicle aerodynamics. Proc. Inst. Mech. Eng. Part D J. Automob.
Eng. 2017, 231, 1190–1202. [CrossRef]
23. GB/T 6323-2014; Automotive industry standard. Controllability and Stability Test Procedure for Automobile Snaking Test. AQSIQ:
Beijing, China, 2014.
24. QC/T 480-1999; Automotive industry standard. Criterion Thresholds and Evaluation of Controllability and Stability for Automo-
biles. State Machinery Industry Bureau: Beijing, China, 1999.