See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/352407708

Design, Optimization, and Analysis of the Chassis of a Racecar

Conference Paper in SAE Technical Papers · June 2021

DOI: 10.4271/2021-01-5058

CITATIONS READS
0 339

3 authors, including:

Greg Wheatley
James Cook University
74 PUBLICATIONS 469 CITATIONS

SEE PROFILE

All content following this page was uploaded by Greg Wheatley on 23 January 2023.

The user has requested enhancement of the downloaded file.

 Downloaded from SAE International by Samuel Popoola, Friday, July 16, 2021

2021-01-5058 Published 11 Jun 2021

Design, Optimization, and Analysis of the Chassis

of a Racecar
Greg Wheatley James Cook University

Sam Popoola University of Lagos

Jayke Collins James Cook University

Citation: Wheatley, G., Popoola, S., and Collins, J., “Design, Optimization, and Analysis of the Chassis of a Racecar,” SAE Technical
Paper 2021-01-5058, XXXX, doi:10.4271/2021-01-5058.

Abstract
increasing the torsional stiffness, and decreasing the weight.

T
his article includes the design procedure and the Finite Computer software ANSYS and SolidWorks were used to analyze
Element Analysis (FEA) that was undertaken to design a the major performance indicators of the chassis design. The
chassis to be used in a Formula Society of Automotive torque of the car was increased by approximately 30%, torsional
Engineers (FSAE) racecar competition. After conducting an audit stiffness was increased by 10%, the chassis was shortened by
on the previous chassis design, major design goals were created 12.8%, and the weight was reduced by 3.5%. The results from the
for optimization and improvement, some of which included analysis suggest that there are many improvements in the new
removing the rear box section to allow a 52-tooth sprocket, chassis design, and it will be highly beneficial to the new FSAE car.

Keywords
Finite element analysis, FSAE racecar, Torsional stiff-
ness, ANSYS

I. Introduction
where F is the force applied at the mounts, L is the length from

T
his work looks at the investigation of major perfor- center to location of force, ∆y𝑙𝑒𝑓𝑡 is the displacement on the
mance indicators of a chassis design [1] for a Formula left side, ∆y𝑟𝑖𝑔ℎ𝑡 is the displacement on the right side, and 𝜑 is
Society of Automotive Engineers (FSAE) racecar the angular deflection
competition, which are to investigate the torsional stiffness For a given height of the center of gravity, the longer the
within a chassis, how the length of a car affects handling and wheelbase the smaller the load transfer, as shown in the longi-
how the weight [9] of a chassis affects the acceleration of a car. tudinal weight transfer formula in the analysis methods of
The chassis is the backbone of a vehicle, of which the this article. A greater load transfer allows a car to accelerate
torsional stiffness is one of the most important characteristics at a much higher rate [2].
of a chassis design [5], affecting the handling, safety, and ride The reduction in vehicle mass without the sacrifice of
comfort of the vehicle [2]. The challenge of any chassis design performance or safety is, as with virtually every automotive
[6] is to increase torsional stiffness without majorly increasing field, an ever-present demand within the FSAE competition
the weight of the chassis, or overengineering the design [7]. [4]. Weight reduction inherently provides improved accelera-
Computer software using Finite Element Analysis (FEA) tion and handling for any given FSAE vehicle, along with a
is used to analyze the torsional stiffness of a chassis by fixing variety of other benefits. As such, weight reduction is often a
the rear mounting points and applying a torque to the front major point of focus for any team aiming to improve their
mounting points. The torque is applied using equal and chassis design [8]. Therefore, for any given force, a lighter
opposite forces to each front mount. The torsional stiffness is vehicle will accelerate faster than a heavier counterpart will.
calculated by using the displacement of each mount [3], as This condition applies to all transient cases, including braking
shown in Equation 1. and cornering, resulting in an all-around improvement in
vehicle mobility.
T F×L Materials are a vital element of every chassis. They have
KT = = Eq. (1) a large influence on vehicle weight, torsional stiffness, strength,
ϕ  ∆y left + ∆y right 
tan −1   and durability. The major criteria for choosing the best
 2L 
 Downloaded from SAE International by Samuel Popoola, Friday, July 16, 2021

2 DESIGN, OPTIMIZATION, AND ANALYSIS OF THE CHASSIS OF A RACECAR

material for a chassis include material weight, strength, cost, FIGURE 2 Design with box section removed (top view).
and workability.

II. Design
The design [10] process adopted throughout the development
of the chassis was inherently an iterative process with several
redesigns of particular regions of the chassis required as the
needs of other parts of the racecar design changed; Figure 1
shows the previous design. In order to assist in developing an
initial design direction for the chassis, core issues were identi-

© SAE International.
fied, which were established as the most important issues that
must be addressed with the new chassis design.
In the previous design, the FSAE vehicle used the engine
from a 2007 Honda CBR 600RR motorcycle. A larger sprocket
was needed to adjust the drivetrain ratio of this adapted
motorcycle engine/transmission to suit an FSAE vehicle, but
the previous frame design did not permit this. The primary FIGURE 3 Original design with box section (top view).
function of the rear box section was to support the differential
and all associated parts; however, following the inclusion of
an engine-mounted differential, the removal of this rear box
section was a feasible modification, as done in Figures 2 and
3 showing the original design and the design with the box
section removed. In removing this section, ample space was
provided for the new rear sprocket, and a core issue of the
chassis design was addressed.
In removing the rear box section of the chassis, several
rear suspension appropriate points were also removed. As
such, consideration had to be made as to the relocation of
these points. After consideration of the Rear Suspension

© SAE International.
design, it was established that the A-arm appropriate points
could be relocated at points further toward the chassis front
with little alteration to the existing geometry required. As
shown in Figures 4-7, the rearmost appropriate points were
shifted to the preexisting nodes at the rear of the chassis.
Additional nodes were then created to accommodate the
FIGURE 4 Original design without node (side view).
frontmost appropriate points with appropriate member trian-
gulation included to maintain the chassis strength. Finally,
the rocker and spring-damper appropriate points were relo-
cated to the curved rear brace of the chassis due to its ideal
location for these points.
With the front suspension running a direct-action
suspension system, an extra node was needed for the

FIGURE 1 Chassis design.

© SAE International.

suspension to attach. An additional node was created to

© SAE International.

accommodate the need for the extra appropriate point.

Additional frame members were added for appropriate
member triangulation to maintain strength within
the chassis.
 Downloaded from SAE International by Samuel Popoola, Friday, July 16, 2021

DESIGN, OPTIMIZATION, AND ANALYSIS OF THE CHASSIS OF A RACECAR 3

FIGURE 5 Design with extra nodes (side view). assigned to each line corresponding to the member diameter
and thickness. A comparison between the chassis model,
before and after conversion to beam elements, is provided in
Figures 8 and 9.
The simplistic nature of beam elements inherently
resulted in a minimal solving time and modelled the broad
chassis behavior accurately. However, several issues were
encountered, the joint detail between each member was lost
in this conversion; furthermore, significant interference can
be observed between separate members. An additional issue
observed with the use of beam elements was that of the over-
approximation of stresses within each member. Given that
several members of the chassis experienced bending
throughout simulations, regions of maximum (tensile) stress
© SAE International.

and minimum (compressive) stress were present at opposite

ends of the member cross-section, as illustrated in Figure 24.
The entirety of each cross-section, however, was modelled by
only one element resulting in the maximum and minimum
stresses within a member existing at a single point. This
resulted in a significant overapproximation of the range of
FIGURE 6 Original design without the fifth node
stress, which any point of the chassis could be expected to
(side view).
experience. Finally, the “Fatigue Tool” module of ANSYS was
unavailable with the use of beam elements resulting in the
alternative, and potentially less detailed, methods having to
be used to analyze the fatigue strength.
Having identified several issues with the use of beam
elements, it was concluded that an alternative method to
model the chassis was required to conduct an effective analysis.
© SAE International.

FIGURE 8 Initial SolidWorks chassis section.

FIGURE 7 Design with the fifth node (side view).

© SAE International.
© SAE International.

FIGURE 9 ANSYS beam element mesh.

III. Discussion and

Analysis
A key issue encountered in the analysis process was estab-
lishing an appropriate method in which to effectively model
the complex geometry of the chassis. This issue was initially
resolved with the use of beam elements. In order to develop
© SAE International.

the corresponding model, the chassis geometry was reduced

to a combination of lines and nodes representing members
and joints, respectively. The geometry file was then imported
into ANSYS where pre-developed cross-sectional profiles were
 Downloaded from SAE International by Samuel Popoola, Friday, July 16, 2021

4 DESIGN, OPTIMIZATION, AND ANALYSIS OF THE CHASSIS OF A RACECAR

An appropriate alternative was found to be shell elements, solution was then developed whereby the engine was substi-
whereby each member was reduced to an outer shell within tuted with a solid block of material. With the appropriate
SolidWorks. This geometry was again reimported into ANSYS calculations below, a suitable Young’s Modulus, E, and density,
with an appropriate thickness being assigned to each member. ρ, were then calculated such that the block possessed similar
The resultant meshing, again with a comparison to the original deformational and mass properties to that of the engine.
geometry, is provided in Figures 10 and 11.
In modelling with shell elements, a far more accurate Engine deformation along the length,
representation of the chassis was developed. The joint detail
L E = Block deformation along the length,L B
was preserved, minimal interference between members was
observed, and mesh resolution across each member cross-
section was improved. It was also established that the “Fatigue ∆L E = ∆L B
Tool” module was compatible with shell elements. As such,
modelling the chassis with shell elements removed each of the ∴ε E LE = ε B LB
key issues associated with using beam elements, and was
therefore the method used in the analysis of the final ∴σ E E E LE = σ B E B LB
chassis design.
An additional factor of the consideration required for ∴ FE AE E E LE = FB AB E B LB
chassis analysis was that of engine modelling. Bolted directly
to the chassis, the engine strongly influences the behavior of ∴1WE H E E E LE = 1WB H B E B LB
the chassis under loading and, as such, could not be neglected
in the analysis process. However, due to the highly complex ∴ E B = LBWB H B × E EWE H E LE
geometry of the engine, it was deemed unsuitable to directly
Using SolidWorks, the average dimensions of the engine
import the engine into ANSYS for analysis. An initial
were approximated as follows:
approach to modelling the engine within the chassis was to
set each engine mount (the engine mounts used are 9 mm
thick steel tubes attached to the chassis) as a “fixed support.” Length, L ~ 321.3 mm; Width,
This, however, was found to result in highly misrepresentative W ~ 363.3 mm; Height, H ~ 367.5 mm.
stresses within the surrounding members. An alternative
It was also approximated that the Young’s Modulus of the
FIGURE 10 Initial SolidWorks chassis section.
engine would lie between those of gray cast iron (E ≈ 130 GPa)
and aluminum alloy (E ≈ 70.3 GPa). Furthermore, the inherent
presence of voids within the engine was expected to result in
a Modulus lower than that of a solid piece of material. A “void
factor” of 0.85 was therefore assigned to account for this. The
Young’s Modulus of the engine was therefore approximated
as follows:

E E ≈ (130 + 70.32 ) / 0.85 = 85.139 GPa

From this, Young’s Modulus required for equivalent

deformation along the length of the components was estab-
© SAE International.

lished. This process was then repeated for the remaining two
directions, width and height, and a mean average was taken
of the required Modulus for each block along each direction.
The final Young’s Modulus approximated for each block
was, therefore,
FIGURE 11 ANSYS shell element mesh.
= =
E B ,Chassis 119 .07 GPa; E B ,Final Chassis 105.55 GPa
Similarly, having conservatively approximated the engine
mass as 75 kg and measuring the volume of each block, an
appropriate block density was established:

ρB ,Chassis = m Engine /VB ,Original Chassis = 75 kg/0.01645 m 3

= 4559 kg/m 3
© SAE International.

ρB ,Chassis = m Engine /VB ,Final Chassis = 75 kg/0.022851 m 3

= 3282 kg/m 3
 Downloaded from SAE International by Samuel Popoola, Friday, July 16, 2021

DESIGN, OPTIMIZATION, AND ANALYSIS OF THE CHASSIS OF A RACECAR 5

FIGURE 12 Comparison of the original engine (left), stiffness could not be made as the specific method in which they
original chassis block (center), and final chassis block (right). were established was not stated. However, provided that a consis-
tent method was applied for all analysis, all internal findings
could be directly compared to establish the difference in
torsional stiffness between the original and final chassis design.
For both the original and final chassis design, the torsional
stiffness was evaluated through the following method. Firstly,
the rearmost upright members of the chassis were fixed in
© SAE International.
space (indicated by the purple members in Figure 13). These
Having approximated the suitable Young’s Moduli and were chosen in order to place as much of the chassis under
densities, materials with these properties were then developed torsional loading for a thorough measurement of the entire
within ANSYS and assigned to the respective blocks used chassis’ stiffness. They were also chosen as they were identified
throughout the analysis. as similar members between the original and new chassis
Given that the final chassis designed included three in location and orientation. This, therefore, minimized the
chassis mounts compared to the two of the original design, difference in model setup between each chassis, allowing for
separate blocks had to be developed for the original and final more reliable comparisons to be made between the resultant
designs. A comparison of these blocks to the original engine findings. After fixing the rearmost upright members, a couple
is provided in Figure 12. was applied at the frontmost appropriate points of the upper
From this, the influence of the engine on the chassis’ A-arms of the front suspension (indicated by the purple
torsional stiffness and behavior under fatigue loading could members in Figure 13). Deformation probes were also placed
be incorporated into the subsequent analysis. at these points to measure the vertical deformation of the
As previously specified for the design, 4130 Chromoly chassis. Knowing the magnitude of the applied forces, 𝐹, the
steel tubing was assigned to each cylindrical member of the distance between the two nodes of the couple, 𝐿, and the
chassis. It was also established that, following the consider- vertical deformation of these nodes, Δ𝑦𝑙𝑒𝑓𝑡 and Δ𝑦𝑟𝑖𝑔ℎ𝑡, the
ation of the Front and Rear Suspension design, grade HA250 torsional stiffness was calculated for both the original and
steel plating (Sy = 250 MPa, Sut = 350 MPa) was to be used for final chassis using the previously introduced Equation 1.
each suspension’s appropriate point. It was therefore assumed Table 1 provides a summary of the key data used in calcu-
that this material was also applicable for the gusset plates and lating torsional stiffness.
engine mount tabs. While this specific material was not In Table 1, a clear increase in torsional stiffness can
present within ANSYS, AISI 1012 steel was found to have be observed between the original and final chassis design; this
similar yield and ultimate tensile strengths (Sy = 233 MPa, can quantitatively be described as a 10.1% increase.
Sut = 333 MPa) and was, therefore, the material used in model- One approach to developing load cases for the vehicle is
ling all non-cylindrical features of the chassis. Having inte- that of software such as the Royal Melbourne Institute of
grated all the required aspects of chassis modelling into both Technology (RMIT) Track Simulator. It was established that
the original and final chassis designs, the analysis could then the RMIT Track Simulator was a suitable tool for the analysis
be conducted. of the system; analysis of the chassis was conducted using data
As discussed, the torsional stiffness is a key performance from the simulator. Following the use of additional software
aspect of the chassis with a higher torsional stiffness providing packages including MATLAB, Microsoft Excel, and
improved vehicle performance. Following the initial investiga- Engineering Equation Solver, the uniaxial forces present
tion, it was found that slight changes in the model setup provided within each suspension member for a standard FSAE track
vastly different measurements for torsional stiffness. was established. Knowing the geometry of each suspension
Furthermore, a standardized process to analytically determine member, these uniaxial forces were then converted to lateral,
torsional stiffness could not be found. It was therefore concluded longitudinal, and vertical forces acting on each appropriate
that any comparison with the competing designs’ torsional point of the chassis for each load case. These load cases were

FIGURE 13 Model setup for torsional stiffness calculation for the original (left) and final (right) chassis.
© SAE International.
 Downloaded from SAE International by Samuel Popoola, Friday, July 16, 2021

6 DESIGN, OPTIMIZATION, AND ANALYSIS OF THE CHASSIS OF A RACECAR

TABLE 1 Summary of torsional stiffness calculations. TABLE 2 Fatigue strength factor calculation.

Chassis design Endurance limit Assigned

Key figures Original Final modifying factors value Justification
Applied force, 250 500 250 500 Load factor, Ca 0.9 Axial forces expected to
F [N] be dominant
Left node 8.74 17.5 8.76 17.5 Size factor, Cb 1.0 Axial forces expected to
deformation, be dominant
Δyleft [×10−5 m] Surface factor, Cc 0.5 Finish assumed to
Right node 8.50 17.0 6.90 13.8 be equivalent to “cold
deformation, drawn” surface, Sut ≈
Δyright [×10−5 m] 0.895 GPa
Node distance, L 0.53475 Temperature factor, Cd 1.0 All members assumed to
[m] consistently operate

© SAE International.
under 350°C
Torsional 3,618,609 3,618,692 3,983,949 3,983,924
stiffness, KT Reliability factor, Ce 0.814 Reliability of R = 0.99
[N·m/deg] assigned
Average 3,618,651 3,983,937 Stress concentration 0.9 Welds assumed to remove
factor, Cf the majority of stress
concentration

© SAE International.
classified as the critical acceleration, deceleration, left turn, Miscellaneous factor, Cg 1.0 No additional factors
and right turn. assumed
Having established the lateral, longitudinal, and vertical Fatigue strength 0.330
forces exerted on each chassis node for each load case, fatigue factor, Kf
analysis could then be completed. However, given the limita-
tions of the ANSYS “Fatigue Tool” module, the complete
fatigue history for each chassis node could not be imported FIGURE 14 Left-turn load case maximum stress.
and analyzed. In order to bypass this issue, an alternative
method of fatigue analysis was completed whereby only one
load case was analyzed at a time. A zero-based loading of
the associated forces was applied and the “Goodman” Mean
Stress Theory was selected. An additional component that
then had to be established prior to analysis was that of the
Fatigue Strength Factor, 𝐾𝑓. However, given that the final
chassis consisted of a combination of geometric features,
namely, cylinders and flat plates, a single Fatigue Strength
Factor could not be determined. As a result, two separate
fatigue analyses were conducted: one with all features
included with a Fatigue Strength Factor of 1 and another
with all non-cylindrical features suppressed and a Fatigue
Strength Factor of 0.330. This factor was calculated as the
product of all individual Endurance Limit Modifying Factors
© SAE International.

as summarized in Table 2.
Having established all relevant figures required for
fatigue analysis, consideration was then made as to the
support settings of the model. It was concluded that simple
supports at both the frontmost and rearmost, lower appro-
priate points for suspension were the most effective approach, be identified as the weakest region of the chassis, being partic-
providing sufficient support to the model while minimizing ularly vulnerable during both left and right turns. This is an
their effect on chassis behavior under loading. The resultant understandable finding given that, during cornering, a slight
analysis findings are presented in Figures 14-25 and summa- twisting of the engine may be expected. This would then trans-
rized in Table 3. late to torsional forces acting on the tab mounts, which, being
In Table 3, the most critical load case in terms of flat plates, is a mode of applied force that is resisted poorly.
maximum stress was found to be the “left turn” load case. After discounting all non-cylindrical features for the
Here a maximum stress of 221.58 MPa was observed to have second fatigue analysis, it was concluded that the members of
occurred on the left engine mount tab, which, having a yield the chassis are sufficiently designed for their intended lifetime.
strength of approximately 250 MPa, is close to the yield stress. All four load cases yielded fatigue lives above the widely
Considering that for the full member and plate fatigue accepted “infinite life” threshold of 1 × 106 cycles, and while
analyses, no Fatigue Strength Factor was assigned, it is highly one cycle only correlates to one zero-based application of the
likely that the left engine mount tab possesses very little respective forces, these results are still expected to imply
fatigue life. As such, these mounting tabs can clearly sufficient design.
 Downloaded from SAE International by Samuel Popoola, Friday, July 16, 2021

DESIGN, OPTIMIZATION, AND ANALYSIS OF THE CHASSIS OF A RACECAR 7

FIGURE 15 Left-turn load case maximum (zoomed). FIGURE 17 Right-turn load case maximum stress.

© SAE International.
© SAE International.

FIGURE 18 Right-turn load case maximum stress.

FIGURE 16 Left-turn load case—Minimum member

fatigue life.
© SAE International.
© SAE International.

FIGURE 19 Right-turn load case—Minimum member

fatigue life.
It can be concluded from the above results that both the
chassis members and the rear suspension gusset plates are
sufficiently designed for their intended lifetime. Both engine
mount tabs, however, are likely to fail after minimal usage.

IV. Conclusion
The new design considered has been improved in the following
areas: firstly, the overall chassis torsional rigidity was increased
© SAE International.

by 10.1%, and also the rear section of the chassis was rede-
signed to accommodate the larger rear drive sprocket, which
results in a 30% increase in torque. The chassis weight was
reduced by 3.5% and the chassis length by 12.8%. A central
 Downloaded from SAE International by Samuel Popoola, Friday, July 16, 2021

8 DESIGN, OPTIMIZATION, AND ANALYSIS OF THE CHASSIS OF A RACECAR

FIGURE 20 Deceleration load case maximum stress. FIGURE 22 Deceleration load case—Minimum member
fatigue life.

© SAE International.
FIGURE 23 Acceleration load case—Maximum stress.
© SAE International.

© SAE International.
FIGURE 21 Deceleration load case maximum
stress (zoomed).

FIGURE 24 Acceleration load case—Maximum

stress (zoomed).
© SAE International.

© SAE International.
 Downloaded from SAE International by Samuel Popoola, Friday, July 16, 2021

DESIGN, OPTIMIZATION, AND ANALYSIS OF THE CHASSIS OF A RACECAR 9

FIGURE 25 Acceleration load case—Minimum member mounting point was added for both the right and left rear
fatigue life. suspensions, with some of the suspension forces cancelling
each other, resulting in less forces applied to the chassis. An
additional node was created in the front section for a direct
suspension method of suspension.

References
1. Salter, J.I., “Design, Analysis and Manufacture of 2011 REV
Formula SAE Vehicle Chassis,” Edu.au, accessed October 16,
2020, http://robotics.ee.uwa.edu.au/theses/2011-REV-SAE-
© SAE International.

Chassis-Salter.pdf.
2. Tebby, S., “Methods to Determine Torsion Stiffness in an
Automotive Chassis,” Comput. Aided Des. Appl. 8, no. PACE
(2011): 67-75.
3. Foale, T., Motorcycle Handling and Chassis Design: The Art
and Science (Tony Foale Designs, USA, 2006).
4. SEAS, “Centre of Gravity,” Formula1-dictionary.net,
accessed October 16, 2020, http://www.formula1-dictionary.
TABLE 3 Summary of analysis results. net/centre_of_gravity.html.
Minimum 5. Milliken, W.F. and Milliken, D.L., Race Car Vehicle
Maximum Maximum stress member Dynamics (Warrendale, PA: SAE International, 1995)
Load case stress [MPa] location fatigue life 6. Anonymous, “Stiffness Comparison of Welded Parts—Part 1,”
Left turn 221.58 Left engine mount 37.455 × 106 Com.au, accessed October 16, 2020, http://ezymechanic.blogspot.
tab com.au/2015/07/stiffness-comparison-of-welded-parts.html.
Right turn 195.52 Right engine mount 39.979 × 106 7. Idea-rs.com, accessed October 16, 2020, http://www.idea-rs.
tab com/release-2016s/version-5-2/.
Deceleration 205.41 Frontmost 4.831 × 106
8. “Student Events,” Sae.org, accessed October 16, 2020, http://
appropriate point
of the lower-right students.sae.org/cds/formulaseries/about.htm.
© SAE International.

rear A-arm 9. Furnace, L.V., “Weight Reduction Techniques Applied to

Acceleration 118.22 Frontmost 9420.8 × 106 Formula SAE Vehicle Design: An Investigation in Topology
appropriate point Optimization,” 2006, https://citeseerx.ist.psu.edu/viewdoc/
of the lower-right download?doi=10.1.1.172.4165&rep=rep1&type=pdf.
rear A-arm
10. Rijeka, “Formula1-Dictionary,” accessed November 2018,
http://www.formula1-dictionary.net/centre_of_gravity.html.

© 2021 SAE International. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of SAE International.

Positions and opinions advanced in this work are those of the author(s) and not necessarily those of SAE International. Responsibility for the content of the work lies
solely with the author(s).

ISSN 0148-7191

View publication stats