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

net/publication/332448488

Optimised ultrafast lightweight design and finite element modelling of a

CFRP monocoque electric car chassis

Article  in  International Journal of Electric and Hybrid Vehicles · April 2019

DOI: 10.1504/IJEHV.2019.101300

CITATION READS

1 2,050

3 authors:

Evangelos Ch. Tsirogiannis Georgios Eleftherios Stavroulakis

University of Patras Technical University of Crete
9 PUBLICATIONS   43 CITATIONS    414 PUBLICATIONS   3,555 CITATIONS   

SEE PROFILE SEE PROFILE

Sofoklis Makridis
University of Patras
87 PUBLICATIONS   522 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Electric Car Chassis Design Development and Manufacturing View project

Vulnerability of Cultural Heritage View project

All content following this page was uploaded by Evangelos Ch. Tsirogiannis on 30 July 2019.

The user has requested enhancement of the downloaded file.

Int. J. Electric and Hybrid Vehicles, Vol. 11, No. 3, 2019 255

Optimised ultrafast lightweight design and finite

element modelling of a CFRP monocoque electric car
chassis

Evangelos Ch. Tsirogiannis*

School of Mechanical Engineering,
National Technical University of Athens,
Athens, GR15780, Greece
Email: etsirogian@central.ntua.gr
*Corresponding author

Georgios E. Stavroulakis
School of Production Engineering and Management,
Technical University of Crete,
Chania, GR73100, Greece
Email: gestavroulakis@isc.tuc.gr

Sofoklis S. Makridis
Department of Environmental and Natural Resources Management,
University of Patras,
Agrinio, GR30100, Greece
Email: smakridis@upatras.gr

Abstract: A breakthrough in the process method has been developed by

combined considerations in the demands of “Shell Eco Marathon” using finite
element modelling (FEM). Ultrafast calculations have revealed novel chassis
development. The most critical factors in designing the new chassis are the
reduction of the weight, the improvement of strength and stiffness, the
reduction of material and the manufacturing cost. The simple structural
surfaces (SSS) method was used for an electric car so as the type of loading
conditions which are applied to the chassis structure to be predicted.
Afterwards, a new design approach for a lightweight carbon-fibre reinforced
polymer (CFRP) monocoque chassis is proposed which conforms to structural,
ergonomic, safety and aesthetic requirements. The SSS method in conjunction
with the creation of the chassis load calculator (CLC) model and both the
specialised and integrated methodology of the processing procedure through
FEM, have attained the overcoming of the time consuming conceptual design
process.

Keywords: parametric design; lightweight; composites; CFRP; carbon-fibre

reinforced polymer; monocoque; CAD; vehicle dynamics; FEM; finite element
modelling; electric car; shell eco marathon; chassis design; stress scenario;
ROM; rule of mixture.

Copyright © 2019 Inderscience Enterprises Ltd.

256 E.C. Tsirogiannis et al.

Reference to this paper should be made as follows: Tsirogiannis, E.C.,

Stavroulakis, G.E. and Makridis, S.S. (2019) ‘Optimised ultrafast lightweight
design and finite element modelling of a CFRP monocoque electric car
chassis’, Int. J. Electric and Hybrid Vehicles, Vol. 11, No. 3, pp.255–287.

Biographical notes: Evangelos Ch. Tsirogiannis is a chartered engineer with

specialisation in Mechanical Design and Manufacturing and with proven
Managerial and Communicational skills. He is currently Research and
Development Engineer in Defence/Automotive Industry. He has great work and
research experience, being involved in numerous Industrial and Research
projects. He has obtained an MEng in Production Engineering and
Management, a postgraduate diploma ‘International Welding Engineer’ and a
MSc in Manufacturing and Production Systems as well as more than
25 honours and achievements. Furthermore, he is starting his PhD in Electric
Cars.

Georgios E. Stavroulakis is Professor at the School of Production Engineering

and Management, Technical University of Crete, Greece, Privatdozent at the
Technical University of Braunschweig, Germany and Honorary Professor
at Jordan University of Science and Technology, Irbid, Jordan. He is civil
engineer with diploma, 1985, and PhD 1991, both with distinctions, from the
Aristotle University of Thessaloniki, Greece, both with distinctions, and a
Habilitation from TU Braunschweig, 2000. His postdoctoral works have been
supported by the Alexander von Humboldt Foundation and by the Marie Curie
Project of European Union. He has served as president of the Greek
Association of Computational Mechanics.

Sofoklis S. Makridis is an Assistant Professor (tenured) at the Department of

Environmental and Natural Resources Management, University of Patras &
senior visiting fellow at Loughborough University. He is a physicist while
he received the MEng in 2002 and the PhD in 2004 in fields of Materials
Science and Engineering (NanoMagnetic Materials). His Postdoc research has
been accomplished on Materials for Hydrogen Storage by having a fellowship
from the State Scholarships Foundation while he holds an MBA in spin-offs.
The last decade gives lectures more than ten hours per week by covering
physics, materials science and technology, hydrogen technologies, energy and
innovation.

1 Introduction

The main goal of this publication, which continues our previous works (Tsirogiannis
et al., 2017, 2018, 2019; Tsirogiannis, 2015, 2017), is to demonstrate a strategy plan
concerning the designing process and guidelines, the materials, the worst case stress
scenario, the loads and the finite element modelling (FEM) process, in order to maximise
the car’s structural efficiency in terms of high strength and performance in low mass
(Tsirogiannis and Vosniakos; 2019; Budynas-Nisbett, 2006). By applying the new
strategy plan, the chassis of an electric car could have less weight and become more
durable. Historically, the studied prototype of this electric car was employing an
aluminium space frame and has already won four trophies in six years, in the Shell Eco
Marathon, being placed among the best cars in this European competition.
 Optimised ultrafast lightweight design and finite element modelling 257

According to the strategy plan, a CFRP monocoque chassis design is proposed that
offers great design freedom and it is lighter, stiffer, stronger, easier to manufacture and
more spacious than the previous one. To achieve a high quality design, the design
specifications were compromised with the team’s targets, the ergonomic and safety issues
were evaluated, the structural possibilities and limitations regarding the available
materials were taken into account, the structural engineering constraints regarding a
lightweight, stiff, strong and easy to manufacture design were investigated, the loads that
act on the axles were analysed and calculated and finally the FEM method was applied to
the new chassis. For the three-dimensional design, the ProEngineer Wildfire 5 software
was used (Anon, 2008b). Also, for the FEM, the ANSA pre-processor of the BETA CAE
Systems (BETA CAE Systems S.A., 2014) and the Ansys solver (LSTC, 2014) were
employed.
The common lightweight urban vehicles, either for Shell Eco Marathon racing
competition (Mat and Ghani, 2012) or for urban use (Nayak et al., 2012) often employ
space frame structures in order to meet strength, low manufacturing cost and aesthetics
demands. Nevertheless, a monocoque chassis with an innovative geometry approach that
comply with these demands is suggested. This chassis became lighter (only 5.38 kg) than
the previous chassis (10.85 kg), which was space frame and was the lightest chassis of
the competition. At the concept stage of the design process of the car chassis, the simple
structural surfaces (SSS) method was useful for determining that there is continuity for
load paths and hence for determining the integrity of the structure (Brown et al., 2002).
Also, in this work (Tebby et al., 2011) the structural feasibility of a preliminary sedan
design concept was confirmed using the SSS method. In literature (Lee et al., 2012) a
smart way on how to sketch using parameters and function based dimensions instead of
using standardised dimensions was presented, so as the design of the structure to be
easily updated and developed by changing one or more of its parameters values. In that
work the designing was not referred for chassis structures, however, that way of
designing is used for our application and also the pre-final and the final design stage are
depicted. In bibliography (Duleba and Greskovic, 2013; Harris, 1999) a method for the
calculation of the mechanical properties of composites was demonstrated. Specifically, in
the work (Duleba and Greskovic, 2013), they used composite material for the arm of their
car and aluminium for their chassis. However, in our project, we recommend the usage of
composite material for the chassis and steel for the arm, in order for the whole structure
to meet higher strength and performance. These works (Riley and George, 2002; Broad
and Gilbert, 2009; Stigliano et al., 2010) have proved that their chassis meets the strength
requirements, under standard deformation modes, without calculating the exact loads
which create these deformation modes. On the other side, an attempt was made by
another study (Reyes Ruiz et al., 2012), where the loads which were applied on the
suspension of a commercial passenger bus were calculated in different loading
conditions. Nevertheless, the chassis was not examined. Apart from that, the loading
scenarios were examined individually, without combining them in order to analyse the
structure under a combined stress condition resulting from the combination of various
loading cases. Hence, in our work, the calculation of the loads according to vehicle
dynamics under a combined condition, named as the worst case stress scenario, is
proposed as a more accurate method and as a method that gives the designers the
knowledge of the amplitude and the direction of the applied loads. This helps them to
model and optimise the structure more efficiently. The manner in which the vehicle
structure is analysed could yield dramatically varying results on identical structures
258 E.C. Tsirogiannis et al.

(Moeller et al., 2004). These works (Prasad et al., 2013; Stigliano et al., 2010; Tebby
et al., 2011; Moeller et al., 2004; Broad and Gilbert, 2009) proposed the car chassis to be
constrained at the rear axles with all DOF’s. Moreover, at the same time the loads were
applied especially to the front axles of the chassis.
In the current work, a novel approach is followed and it has been observed that
there was no platform that automatically calculates, by importing just some basic
characteristics of the vehicle, the magnitude and the direction of the loads acting on each
vehicle,. Thus, the chassis load calculator (CLC) model is implemented so as to calculate
easily and quickly the applied loads of the chassis. This model helps the designer to
redesign and develop the chassis, in the conceptual design phase, without the need to
carry out the calculations from the beginning. Therefore, the novelty of this work is the
overcoming of the time consuming conceptual design process which is attained not only
with the creation of the CLC model but also with the use of the SSS method for an
electric car. Via SSS method, the types of loading conditions which are applied to the
chassis structure are predicted and then the chassis is designed taking under consideration
these loading conditions. Furthermore, a specialised and integrated methodology of the
FEM method is developed, in which all the steps are pointed out thoroughly. Contrary to
other previous works which modelled their chassis using standard deformation modes
and without combined stress conditions, in the present study, the chassis is modelled after
calculating the exact vehicle’s dynamic loads under a combined stress condition as a
more accurate method and as a method that helps the designers to investigate the paths
and the amplitude of the loads. Finally, another great outcome of this work is the
achievement of the lightest chassis in the Shell Eco Marathon competition that combines
ergonomics, safety, aesthetic and strength demands.

2 Vehicle and chassis specifications

The previous vehicle was a one seat, four wheel car for urban environments, particularly
designed for the racing competition Shell Eco Marathon. The chassis was employing an
aluminium space frame constructed by (30x30) mm 6082-T6 aluminium hollow sections
with 1.5 mm wall thickness, welded together accurately and a 1 mm thickness carbon-
fibre body shell which was not used as a structural part of the chassis. The energy system
of the vehicle consisted of a brushless electric motor with 4 Nm max motor torque and
4000 RPM max motor rotational frequency, powered from a 1.2 kW H2 fuel cell. The fuel
cell used a hydrogen 200 bar bottle as a fuel tank so as to power the vehicle. The
powertrain consisted of one-stage geared transmission with ratio 1:10 placed between the
electric motor and the wheel in order to provide the needed torque and rpm. The drag
coefficient (Cd) was 0.21 and the maximum vehicle speed was 37 km/h. The dimensions
of the vehicle were (2.5x1.25x1) m (LxWxH) with 1.295 m wheelbase and 0.910 m track
width. The dimensions of the chassis were 1.740x0.730x0.740 m (LxWxH). The total
vehicle weight with the driver on seat (73 kg driver’s weight) was 155.1 kg and excluding
the driver was 82.1 kg. The space frame chassis weight was 10.85 kg. The achieved
autonomy was around 63 km/kWh and the CO2 emissions were 0 gr. The vehicle and the
chassis with the aforementioned specifications are shown in Figure 1 during the race in
the Shell Eco Marathon competition.
 Optimised ultrafast lightweight design and finite element modelling 259

Figure 1 Shell Eco Marathon racing competition: (a) previous vehicle and (b) previous space
frame chassis (see online version for colours)

In contrast with the aforementioned specifications, the new chassis achieved a mass
reduction greater than 30% in a low manufacturing cost upgrading in parallel the
stiffness, the ergonomics, the safety and the aesthetics. The other specifications remain
the same.

3 Chassis type selection

The design principle of the space frame is based on the construction of a structure where
the tubes are mainly loaded in their strongest direction, in tension and compression loads,
and not in bending loads. Hence, the tubes of the previous chassis were small, light and
thin walled. As a result, a structure with low stiffness was obtained. Moreover, the space
frame of the previous chassis had not enough surfaces to install the mechanical and
electrical parts because the chassis construction consisted of so many voids. Accordingly,
the creation of extra housings on the chassis in order to fit in the mechanical and
electrical parts was maximising the chassis weight. Furthermore, the manufacturing was
complex because of the many tubes which needed to be cut and welded together
accurately. Therefore, the production of the structure entailed not only high risk but also
many working hours leading to a higher manufacturing cost.
Consequently, the monocoque chassis has been proposed as the new chassis structure.
Thus, an entire panel provides the strength instead of tubes. Forming complicated shapes
‘bracing members’ were placed in more directions with the design freedom the
monocoque chassis type permitted. As a result, a stiffer, lighter and more spacious
structure has been constructed. Taking cost and the efficiency of production into account
it has been remained the most preferred option. The manufacturing and the assembly was
260 E.C. Tsirogiannis et al.

easier and more accurate. Additionally, in case of future mass production its repeatability
will be high and its manufacturing will be easy for robotised factories. Finally, the
production speed will be faster (Brown et al., 2002).

4 Simple structural surfaces conceptual design method

Due to the fact that time in this kind of competitions is limited, one cannot rely on the
feedback one receives after a car is designed so as to redesign it and produce a better one.
Taking this factor into account, the computer aided engineering (CAE) in conjunction
with the SSS method was used. Once the loads were defined (bending load, shear load,
tension load or compression load), then one can move ahead with designing the
indispensable characteristics of structure to constituent. One need not know the accuracy
of the loads in the conceptual design stage (Tebby et al., 2011; Brown et al., 2002).

4.1 Illustration of load paths in open car with the SSS method
The proposed chassis design has no upper sideframe and the roof does not carry loads.
Consequently, it ‘behaves’ similarly with the open car type. Figure 2 demonstrates the
SSS model of an open car in a bending load case scenario (Brown et al., 2002).

Figure 2 Simple structural surfaces model of open car (Brown et al., 2002)

A vertical load PZ (PZ/2 on each side) is applied to the cross-member of the structure.
The simple structure surface edge loads that are produced from the vertical load are
presented in Figure 3.
There are equal and opposite forces so as the equilibrium for each component to be
achieved as shown in Table 1 (Brown et al., 2002).

Table 1 The edge forces in the end structures (Brown et al., 2002)

Forces Front Rear

Upper and lower boom force Px1 = RFa1 / h1 Px2 = RRa2 / h2
Bulkhead side reaction P1 = RF P2 = RR
Parcel shelf to pillar force P3 = Px1 = RFa1 / h1 P4 = Px2 = RRa2 / h2
Floor edge force (same on both sides) P5 = Px2 – Px1
 Optimised ultrafast lightweight design and finite element modelling 261

Figure 3 Edge loads in bending load case on open car (Brown et al., 2002)

Figure 4 shows the torsion load case scenario for an open car. Equal and opposite couples
RFTSF, RRTSR are applied at the front and the rear suspension, respectively.

Figure 4 Edge loads in torsion load case on open car (Brown et al., 2002)

There are equal and opposite forces so that the equilibrium for each component is
achieved as shown in Table 2 (Brown et al., 2002).

Table 2 The equilibrium equations of torsion load case (Brown et al., 2002)

Forces Front Rear

Suspension force RF = T / SF RR = T / SR
Upper and lower boom force Px1 = RFa1 / h1 Px2 = RRa2 / h2
Bulkhead side reaction Q1 = RFSF / B = T / B Q2 = RRSR / B = T / B
Parcel shelf to pillar force Q3 = Px1SF / B Q4 = Px2SR / B
Floor (in-plane) Px1SF + Px2SR – Q5B = 0
262 E.C. Tsirogiannis et al.

where B is the car’s width in Table 2.

Nevertheless, equal and opposite loads FZ which can be acted only from the floor, are
needed, for the reason that all the edge forces on the sidewall create moments in the same
direction.

5 Ergonomics and safety

The driver can be aided in his or her performance by ensuring that all controls can be
easily reached, he or she has a comfortable seating position and that visibility over the
front of the chassis is sufficient (Carello and Messana, 2015). The variables for a good
seating position are the vertical and horizontal position of the steering wheel, the
horizontal position and angle of the seat with respect to the horizontal, the horizontal and
vertical position of the pedal assembly, the height and horizontal position of the
dashboard and front roll hoop. Besides being comfortable, the driver must be safe at all
times (Anon, 2008a). This mainly involves that many rules were followed in order to
design a safe car. Some major regulations of Shell Eco Marathon are the existence of a
roll bar that withstands 700 N (applied in all directions) and extends 5 cm around driver’s
helmet, a bulkhead that secures the driver, a wide and long enough chassis design to
protect the driver’s body and dimensional demands for the chassis to allow for quick
driver egress in case of accidents or fire (Anon, 2014b). These regulations are often with
respect to a so called 95th percentile male (Van Kerkhoven, 2008) as shown in Figure 5.

Figure 5 95th percentile male driver template on our new chassis (see online version for colours)

6 Results and discussion

6.1 New chassis’ parametric design stages

The Pro Engineer Wildfire 5.0 was employed for the parametric design of the new
monocoque chassis, a CAD software package of the Parametric Technology Corporation
(PTC). The boundary dimensions of the new chassis design are (1740x730x740) mm.
 Optimised ultrafast lightweight design and finite element modelling 263

Working with parameters and relations in Pro Engineer, the new chassis design was
generated parametrically (Anon, 2008b). It could be updated by changing one or more
values of its parameters. To do so, variables and algorithms were used to generate a
hierarchy of mathematical and geometric relations (Lee et al., 2012). Changing the
parameter values, some optimal steps were created for the final design stage of the
new chassis. One of the intermediate steps of the parametric design process is shown in
Figure 6 and the final design stage is presented in Figure 7.

6.2 Design of the parts fitted on the chassis

The front spindle base, the front spindle, the rear semi-axle kit and the roll bar are
illustrated in Figure 8.
Τhe body shell is illustrated in Figure 9.

Figure 6 Pre-final design stage: (a) view 1; (b) view 2; (c) view 3; (d) view 4; (e) view 5
and (f) view 6 (see online version for colours)

6.3 Mechanical properties of the new CFRP chassis

Τhe carbon-fibre reinforced epoxy composite has been selected as the construction
material for the principal characteristics strength-to-weight ratios, stiffness-to-weight
ratios, durability, fatigue, thermal expansion, energy dampening, corrosion resistance and
design and production flexibility (Hull and Clyne, n.d.; Eurenius et al., 2013).
Eight unidirectional plies were stacked with a sequence that is described by formulae
[0/30/0/–15/15/0/–30/0].
264 E.C. Tsirogiannis et al.

Figure 7 Final design stage: (a) view 1; (b) view 2; (c) view 3; (d) view 4; (e) view 5
and (f) view 6 (see online version for colours)

Figure 8 Parts fitted on the chassis: (a) front spindle base; (b) front spindle; (c) rear semi-axle kit
and (d) roll bar (see online version for colours)
 Optimised ultrafast lightweight design and finite element modelling 265

Figure 9 Body shell adopted by Athanasios Tzanakis: (a) and (b) different views
(Tzanakis, 2012) (see online version for colours)

The volume fraction of fibre/resin, can be calculated (Harapin et al., 2008):

V f = (W f / d f ) / ⎡⎣(W f / d f ) + (Wm / d m ) ⎤⎦ (1)

Vm = (1 − V f ) (2)

where df is the density of the fibre, dm is the density of the resin, Wf is the weight of the
fibre, and Wm is the weight of the resin.
The density of the composite can be given by:

d c = 1 / ⎡⎣(V f / d f ) + (Vm / d m ) ⎤⎦ (3)

Assuming that the structure is a simple beam with length L, consisting of fibre and
resin that deform together and the deformation is time independent, a method of
estimating the stiffness of a unidirectional composite is performed (rule of mixtures)
(Hadigheh et al., 2014).

Ec = E f V f + Em (1 − V f ) (4)

where Ef is the elastic modulus of the fibre, and Em is the elastic modulus of the epoxy.
Assuming an anisotropic thin composite lamina with the fibres aligned in the x1
direction, transverse to the x2 direction and vertically to the x3 direction, Young’s
modulus E, shear modulus G and Poisson ratios v, in all three axes, are required for its
characterisation (Duleba and Greskovic, 2013).

Ex = E f V f + Em (1 − V f ) (5)

(
E y = E f Em / EmV f + E f (1 − V f )) (6)
266 E.C. Tsirogiannis et al.

Ez = E y (7)

ν xy = ν f V f + ν m (1 − V f ) (8)

ν yz = ν xy Et / Ec (9)

ν xz = ν xy (10)

(
Gxy = Gm G f / GmV f + G f (1 − V f )) (11)

(
Gyz = E y / 2 (1 −ν yz ) ) (12)

Gxz = Gxy (13)

Respectively, the longitudinal tensile strength, the transverse tensile strength and the
compression strength on the composite are listed below.

σ x = σ f V f + σ m (1 − V f ) (14)

(
σ y = σ m 1 − 4V f / π ) (15)

σ compx = Gm / (1 − V f ) (16)

where σf is the fibres stress levels, and σm is the resin stress levels.
For the multi-ply laminates, the tensile modulus, the shear modulus and
the Poisson ratio of a random continuous-fibre composite can be calculated by (Harris,
1999):
E = ( 3 / 8 ) E1 + ( 5 / 8 ) E2 (17)

G = (1 / 8 ) E1 + (1 / 4 ) E2 (18)

ν = ( E − 2G ) / 2G (19)

where E1 is the longitudinal modulus, and E2 is the transverse modulus for a

unidirectional lamina.
The Krenchel model was utilised for the approximation of the strengths
of multi-ply laminates. The efficiency factor, nθ, was used in a mixture-rule calculation
(Harris, 1999):
nθ = ∑an cos 4 θ (20)

σ c = nθ σ fuV f + σ m (1 − V f ) (21)

Tables 3 and 4 provide the properties of unidirectional CFRP and multi-ply

laminates.
 Optimised ultrafast lightweight design and finite element modelling 267

Table 3 Properties of unidirectional CFRP

Value Unit
Elastic modulus Ex 380.100 GPa
Elastic modulus Ey 28.269 GPa
Elastic modulus Ez 28.269 GPa
Poisson ratio νxy 0.336
Poisson ratio νyz 0.025
Poisson ratio νxz 0.336
Shear modulus Gxy 4.213 GPa
Shear modulus Gyz 13.790 GPa
Shear modulus Gxz 4.213 GPa
Tensile strength σx 2539.400 MPa
Tensile strength σy 8.251 MPa
Compressive strength σcompx 4.722 GPa

Table 4 Properties of multi-ply laminates

Properties Value Unit

Elastic modulus E 160,206 MPa
Shear modulus G 54,580 MPa
Tensile strength σc 971.400 MPa
Poisson ratio ν 0.468
Density dc 1.5e-9 tonne/mm3

6.4 Mechanical properties of the parts fitted on the chassis

The front spindle base, the front spindle and the rear semi-axle kit were made of AISI
9000 Series Steel and the roll bar was made of Aluminium 6082-T6 (Anon, n.d.).
Τhe properties of the AISI 9000 Series Steel and the Aluminium 6082-T6 are shown in
Tables 5 and 6, respectively.

Table 5 Properties of AISI 9000 Series Steel

Properties Value Unit

Elastic modulus E 207000 MPa
Shear modulus G 80000 MPa
Tensile strength σ 850 MPa
Poisson ratio ν 0.29
Density 7.8e-9 tonne/mm3
268 E.C. Tsirogiannis et al.

Table 6 Properties of Aluminium 6082-T6

Properties Value Unit

Elastic modulus E 72000 MPa
Shear modulus G 33800 MPa
Tensile strength σ 270 MPa
Poisson ratio ν 0.330
Density 2.7e-9 tonne/mm3

6.5 Mass and centre of gravity definition

The software Pro Engineer gives the result of the new chassis mass as 5.38 kg. Thus, the
total vehicle’s weight (with the driver on seat) is only 149.64 kg. Additionally, Pro
Engineer shows us the centre of gravity of the new chassis. Thereafter, the different
centres of gravity (battery light, steering system, seat plus the driver, chassis, fuel cell,
electric motor) are added up, and form the main centre of gravity. The weights of the
battery light, steering system, seat (plus the driver), chassis, fuel cell and electric motor
are 2 kg, 7 kg, 75 kg, 5.38 kg, 12 kg, and 3 kg, respectively. The wheels’ mass are
omitted because they have no effect on finding the gravity centre (Wagtendonk, 1996;
Seddon, 1990; Prouty, 1990; Hakewill, 2007; Baraff, 2001; Tsirogiannis, 2017). The
electric car’s centre of gravity is given by:
Wt = W1 + W2 + W3 + W4 + W5 + W6 (22)

COG x = ( Lx1W1 + Lx2W2 + Lx3W3 + Lx4W4 + Lx5W5 + Lx6W6 ) / Wt (23)

COG y = ( Ly1W1 + Ly2W2 + Ly3W3 + Ly4W4 + Ly5W5 + Ly6W6 ) / Wt (24)

COG z = ( Lz1W1 + Lz2W2 + Lz3W3 + Lz4W4 + Lz5W5 + Lz6W6 ) / Wt (25)

where ( Lxx , Ly y , Lzz ) represent the different centres of gravity (battery light, steering
system, seat plus the driver, chassis, fuel cell, electric motor) which are added up, and Wi
their weights.
The COG of the new vehicle is shown in Table 7.

Table 7 New vehicle’s centre of gravity

COGx (mm) COGy (mm) COGz (mm)

867.900 391.178 419.625

6.6 Determination of the worst case stress scenario

A commonly used assumption at the early design stage is: “If the structure can resist the
worst possible loading which can be encountered, then it is likely to have sufficient
fatigue strength” (Brown et al., 2002). In general, the different deformation modes are
divided into longitudinal torsion, vertical bending, lateral bending and horizontal
lozenging. The loads which cause these deformation modes are crash, ride, towing,
aerodynamic, cornering, braking and tractive loads. Crash cases are often the most
 Optimised ultrafast lightweight design and finite element modelling 269

difficult and critical to design. Nevertheless, crash loads are outside the scope of this
publication, since the structure is moved out of the elastic regime into deep collapse. The
ride loads are one of the most important criteria by which the engineer can judge the
design and construction ‘quality’ of a car. However, in this case, the track road is not
rough, but quite smooth (Gillespie, 1992). Towing loads cannot be neglected, but the
vehicle will not need to tow another vehicle. Aerodynamic loads also stress the vehicle
structure. Nevertheless, these loads are negligible because the forward velocity is very
small (20–30 km/h). Consequently, static, cornering and braking loads have been taken
into account.
Cornering loads are maximised when the vehicle’s speed is at maximum speed and its
turning radius is minimised. At the Shell Eco Marathon’s track, in Rotterdam, there are
four counterclockwise turns and one clockwise turn with approximately the same angle
(90°). Regarding to the driving strategy the vehicle runs at high speed (30 km/h) on first
turn, which is a counterclockwise turn (Omar et al., 2013). Therefore, the first turn has
been investigated while the chosen racing line depends on the characteristics of the car,
the cornering strategies and the conditions around. In the apex point of the corner, the
maximum speed and stress are reached. Therefore, the apex of the first turn is the point
where there is the cornering worst case stress scenario.
Hence, we considered that the motion of the car follows a parabola and R is the
distance between the focus point and the points of the parabola as shown in Figure 10.

Figure 10 The motion of the car follows a parabola (see online version for colours)

At low speed, the tyres need develop no lateral forces. If the rear wheels have no slip
angle, the centre of turn has to lie in the projection of the rear axle. Likewise, the
perpendicular from each front wheel should pass through the same point (the centre of
turn). Assuming small angles, the steer angles δO, δi and the Ackerman Angle δ are
given by (Gillespie, 1992):
δO ≅ L / ( R + t / 2) (26)

δi ≅ L / ( R − t / 2) (27)

δ = L/R (28)
where L is the wheelbase, and t is the track width.
270 E.C. Tsirogiannis et al.

The distances between the focus point and the points of parabola are equal to the
distance between the points of parabola and the directrix. Therefore, the minimum
distance between the focus point and the points of the parabola, is the minimum distance
between the points of the parabola and the directrix.
KX = R = 10000 mm (29)
At this time the vehicle is located in X position:
δ 0 max = 1295 / (10000 + 455 ) = 0.12 rad = 6.88° (30)

δ i max = 1295 / (10000 − 455 ) = 0.14 = 8.02° (31)

Braking loads cause larger loads than tractive loads (Brown et al., 2002). Thus, a real
situation needs to be considered when the chassis is overloaded during braking.
Supposing that while the vehicle is moving on the track, with its maximum speed, 30 km
per hour, the preceding vehicle suddenly brakes. Therefore, the driver is forced to brake
immediately, to avoid the collision. At this point, it is needed to find a realistic
‘deceleration scenario’ for urban cars, to determine the deceleration value. The
‘autonomous emergency braking’ (AEB) test of Euro NCAP was chosen. Randomly,
the Fiat’s braking control system was selected to see how it behaves in braking tests
(Anon, 2014a). According to this system, at speeds between 20 km/h and 30 km/h, the
brakes apply a maximum deceleration of 6 m/s2. In our case it is also supposed, that the
driver’s reflexes during braking, are as good as Fiat’s braking control system. Thus,
the vehicle will be subjected to brake with a deceleration equal to 6 m/s2 from 30 km/h
to 0 km/h (Schram et al., 2013; Hulshof et al., 2013; Sandner, 2006).
In order to demonstrate the strength of the chassis, it only has to be shown that it
withstands the total load worst case stress scenario that is the combination of cornering
and braking worst case stress scenario. Thus, it is needed to study the scenario where the
vehicle is turning in the 1st corner and while it is positioned in the apex with 30 km/h, it
encounters a stationary preceding vehicle and decelerates immediately (6 m/s2), to avoid
the accident.

6.7 Dynamic axle loads

Presuming that the vehicle sits statically on level ground, the vertical loads can be
calculated (Gillespie, 1992).
W f = Mg ( c / L ) (32)

Wr = Mg ( b / L ) (33)

where M is the vehicle mass, g is the gravity acceleration, b is the distance from the front
axle to the CG, and c is the distance from the rear axle to the CG.
According to the lateral dynamics, the two front wheels can be represented by one
wheel at a steer angle δ, with a cornering force equivalent to both wheels. The same
assumption is made for the rear wheels (Gillespie, 1992):
Fy = Fyf + Fyr = MV 2 / R (34)

where V is the forward velocity.

 Optimised ultrafast lightweight design and finite element modelling 271

During cornering, a dynamic load transfer from the inside to the outside wheels
occurs (the second mechanism for this study is zero, because the chassis has not springs)
(Reyes Ruiz et al., 2012).
Fzo − Fzi = 2 Fy hr / t + 2 Kϕ ϕ / t (35)

where, hr is the roll centre height, Kφ is the roll stiffness of the suspension, and φ is the
roll angle of the body.
The torque generated by the rotor, for each wheel brake, as well as the total braking
force is defined (Walker, 2005; Limberg, 2015; Hamilton and Klang, 2009).
Fbp = Fd { L2 / L1} ⇒
Pmc = Fbp / Amc ⇒
Pcal = Pmc ⇒
Fcal = Pcal Acal ⇒
Fclamp = 2 Fcal ⇒
(36)
Ffriction = Fclamp μbp ⇒
Tr = Ffriction Reff ⇒
Tt = Tw = Tr ⇒
Ftyre = Tt / Rt ⇒
Ftotal = ∑Ftire LF , RF , LR , RR

where Fbp is the force output of the brake pedal, Fd is the force applied to the pedal pad
by the driver, L1 is the distance from the brake pedal arm pivot to the output rod clevis
attachment, L2 is the distance from the brake pedal arm pivot to the brake pedal pad,
Pmc is the hydraulic pressure by the master cylinder, Amc is the effective area of the master
cylinder hydraulic piston, Pcal is the hydraulic pressure to the calliper, Fcal is the linear
mechanical force by the calliper, Acal is the effective area of the calliper hydraulic piston,
Fclamp is the clamp force by the calliper, Ffriction is the frictional force by the brake pads,
μbp is the coefficient of friction between the brake pad and the rotor, Tr is the torque
generated by the rotor, Reff is the effective radius of the rotor, Tt is the torque in the tyre,
Tw is the torque in the wheel, Ftyre is the force in the tyre, and Rt is the effective rolling
radius of the loaded tyre.
During braking, a dynamic load transfer from the rear to the front axle occurs
(Walker, 2005; Gritt, 2002).

WT = ( aν / g ) × ( hcg / L ) × ( M × g ) (37)

where aν is the deceleration, and hcg is the vertical distance from the CG to ground.

6.8 Chassis load calculator (CLC) model

The values of the forces acting on a vehicle structure change depending on the
characteristics of the structure. Therefore, for academic and research purposes, a model
has been created which automatically calculates the magnitude and the direction of the
loads acting on each vehicle, by importing the characteristics of the vehicle. In order to
272 E.C. Tsirogiannis et al.

validate the derived model, data such as track width, wheelbase, centre of gravity, mass,
et cetera were used as inputs. Equations (27)–(32) that were utilised to implement this
model have been derived from the theory of vehicle dynamics. According to the chosen
stress scenario and the imported characteristics of the new chassis, the applied loads were
calculated.
First of all, the vertical loads were identified as shown in Tables 8 and 9.
Then, the lateral loads were calculated according to the cornering worst case stress
scenario as presented in Tables 10 and 11.

Table 8 The data used for the calculation of vertical dynamics

Distance from the front axle to the CG 602.900 mm

Distance from the rear axle to the CG 692.100 mm
Distance from the left side of the chassis to the CG 391.178 mm
Distance from the right side of the chassis to the CG 338.822 mm

Table 9 Vertical dynamics

Static load on the left front wheel 372.264 N

Static load on the right front wheel 429.788 N
Static load on the left rear wheel 324.286 N
Static load on the right rear wheel 374.396 N

Table 10 The data used for the calculation of lateral dynamics

Initial forward velocity 8333 mm/s

Turn radius 10000 mm
Gravity acceleration 9810 mm/s2
Roll centre height 280 mm
Track width 910 mm
Distance of chassis on y axis 730 mm
Track width (–) Distance of chassis on y axis 180 mm
Track width (–) Distance of chassis on y axis (from one side) 90 mm

Table 11 Lateral dynamics

Cornering force 1062.276 N

Load transfer on the right 653.708 N
Load transfer on the left –653.708 N

During cornering the mass distribution changes, as well as the centre of gravity
(Hakewill, 2007; Baraff, 2001; Bender et al., 2014). Assuming that there is no
mass transfer in the z axis, since the car has not shock absorbers as well as the fact that if
there is a mass transfer in the x axis, it will be negligible, then the new COG is shown in
Table 12.
 Optimised ultrafast lightweight design and finite element modelling 273

Table 12 New centre of gravity during cornering

COGx (mm) COGy (mm) COGz (mm)

867.900 704.022 419.625

The mass distribution changes during the ¼ turn as given in Table 13.
The braking loads were calculated according to the braking worst case stress scenario
as presented in Tables 14–20.

Table 13 Mass distribution on left and right wheels

Distribution of mass on the right side 96.441%

Distribution of mass on the left side 3.559%

Table 14 Data used for the calculation of braking loads

Final forward velocity 0 mm/s

Absolute value of velocity change 8333 mm/s
Braking time 10 s
Braking distance 83330 mm
Maximum deceleration 833.300 mm/s2
Wheelbase 1295 mm
Front area of front axle 265 mm
Tyre coefficient of friction 0.0025

Table 15 Data for the brake system dimensions

Distance from the brake pedal arm pivot to the output rod clevis attachment L1
Distance from the brake pedal arm pivot to the brake pedal pad L2
Front Rear
Wheel radius 280 mm 280 mm
Master cylinder diameter 12.7mm 12.7 mm
Distance-pushrod to balance bar pivot 30 mm 40 mm
The effective area of the calliper hydraulic piston found on one half of the 800 mm2 800 mm2
calliper body
Pad coefficient of friction 0.35 0.35
Disc diameter 160 mm 160 mm
Pad depth 3 mm 3 mm
Gap between top of pad and disc 1 mm 1 mm

Resulting from all the above, a 6.19 m/s2 deceleration was achieved as shown in Table
19, whose value is been greater than the maximum value of the Fiat’s deceleration (6
m/s2), which was first set as a goal.
274 E.C. Tsirogiannis et al.

Table 16 Data for the dynamic characteristics of the vehicle

Cg Height 419.625 mm
Wheelbase 1295 mm
Front wheel rolling radius 280 mm
Rear wheel rolling radius 280 mm
Weight on the front axle 81.759 kg
Weight on the rear axle 71.221 kg
Total weight 152.98 kg
Percentage weight on the front axle 0.534%
Percentage weight on the rear axle 0.466%

Table 17 Force applied on the balance bar by the driver

kgf applied to pedal 10 kgf

Force applied to pedal 98.1 N
Pedal ratio 4:1
Force on balance bar 392.4 N

Table 18 Braking force calculation

Front Rear
Balance bar proportion 0.571 0.429
Force on M Cyl piston 224.229 N 168.171 N
Master/cylinder area 126.613 mm2 126.613 mm2
Line pressure generated by the master cylinder 1.771 N/mm2 1.328 N/mm2
2
Line hydraulic pressure transmitted to the calliper 1.771 N/mm 1.328 N/mm2
The one sided linear mechanical force generated by the 1416.785 N 1062.588 N
calliper
Clapping force on disc generated by the calliper 2833.569 N 2125.177 N
The frictional force generated by the brake pads opposing 991.749 N 743.812 N
the rotation of the rotor
Fx = Fsin45 701.273 N 525.955 N
Fy = Fcos45 701.273 N 525.955 N
Disc effective radius 77.5 mm 77.5 mm
Disc torque, the torque generated by the rotor (both pads 1 76860.566 Nmm 57645.425 Nmm
wheel)
The torque found on the tyre = torque wheel = torque by the 76860.566 Nmm 57645.425 Nmm
rotor
The force reacted between the tyre and the ground 274.502 N 205.877 N
(assuming friction exists to support the force
 Optimised ultrafast lightweight design and finite element modelling 275

Table 19 Deceleration and stopping distance

Total force (4 wheels) 960.757 N

Deceleration a 6194.436 mm/s2
Stopping distance 5604.940 mm

Table 20 Load transfer from braking

Front Rear
Weight transfer 31.735 kg –31.735 kg
Axle load under braking 113.493 kg 39.487 kg
Dynamic axle load 1113.371 N 387.363 N
Load transfer from braking on the left side 11.079 N –11.079 N
Load transfer from braking on the right side 300.240 N –300.240 N

In the case of both braking and turning loads, the mass distribution changes, as well as
the centre of gravity of the vehicle (Hakewill, 2007; Baraff, 2001; Bender et al., 2014).
Assuming that there is no mass transfer in the z axis, since the car has not shock
absorbers, as well as the fact that if there is a mass transfer in the y axis, it will be
negligible, then the new COG is shown in Table 21.

Table 21 New centre of gravity during braking and cornering

COGx COGy COGz

599.260 704.022 419.630

This is the centre of mass that the vehicle has, during braking and cornering coexistence.
It is observed that after such a sudden stop in a ¼ turn the mass distribution changes as
presented in Table 22.

Table 22 Mass distribution on front and rear axles

Distribution of mass on front axle 74.188%

Distribution of mass on rear axle 25.812%

With the new centre of mass, the cornering force on each wheel can be found in Table 23.

Table 23 The cornering force on each wheel with the new centre of gravity

Cornering force (front) 788.086 N

Cornering force (rear) 274.190 N
Cornering force on the left front wheel 28.045 N
Cornering force on the right front wheel 760.041 N
Cornering force on the left rear wheel 9.757 N
Cornering force on the right rear wheel 264.433 N
276 E.C. Tsirogiannis et al.

Summarising, the loads that act on each semi-axle of the chassis, in the z axis, are
presented in Table 24.
Every total force applied to each semi-axle, in the upward direction on the z axis, is
calculated by the sum of the above loads as shown in Table 25.
The cornering forces Fy are transferred from the contact patch to the centre of the axle
(Emri and Voloshin, 2016). The equivalent system consists of the cornering forces (Fy)
plus the moments (Mx) that are created from the cornering forces. These moments are the
result of the cornering forces multiplied by the vertical distance, which is z = 280 mm as
presented in Table 26.

Table 24 Combination load case

Static load on the left front wheel 372.264 N

Static load on the right front wheel 429.788 N
Static load on the left rear wheel 324.286 N
Static load on the right rear wheel 374.396 N
Load transfer from cornering on the right front wheel 484.976 N
Load transfer from cornering on the right rear wheel 168.732 N
Load transfer from cornering on the left front wheel –484.976 N
Load transfer from cornering on the left rear wheel –168.732 N
Load transfer from braking on the left front wheel 11.079 N
Load transfer from braking on the right front wheel 300.240 N
Load transfer from braking on the left rear wheel –11.079 N
Load transfer from braking on the right rear wheel –300.240 N

Table 25 Total dynamic loads on each wheel (z axis)

Total dynamic load on the left front wheel –101.633 N

Total dynamic load on the right front wheel 1215.004 N
Total dynamic load on the left rear wheel 144.475 N
Total dynamic load on the right rear wheel 242.888 N

Table 26 Cornering forces Fy and moments Mx from contact patch to the centre of axle

Fy on the left front wheel 28.045 N

Fy on the right front wheel 760.041 N
Fy on the left rear wheel 9.757 N
Fy on the right rear wheel 264.433 N
Mx on the left front axle 7852.662 Nmm
Mx on the right front axle 212811.430 Nmm
Mx on the left rear axle 2732.090 Nmm
Mx on the right rear axle 74041.136 Nmm
 Optimised ultrafast lightweight design and finite element modelling 277

The braking force needs to be analysed in x and z axes (Fx, Fz) as shown in Table 27.
The vertical distance of Fz from the end of the axle is calculated as well as the vertical
distance of Fx from the centre of the axle as presented in Table 28.
Fx, Fz are transferred to the axle. The equivalent system consists of the braking forces
(Fx, Fz) plus the moments (My1, My2) that are created from the braking forces as shown in
Table 29.

Table 27 Braking forces

Fx Fz
Braking force on the left front axle 701.273 N 701.273 N
Braking force on the right front axle 701.273 N 701.273 N
Braking force on the left rear axle 525.954 N 525.954 N
Braking force on the right rear axle 525.954 N 525.954 N

Table 28 Vertical distances

x = (cos45)*0.0075 + 0.0075 12.803 mm

y = (sin45)*0.0075 5.303 mm

Table 29 Braking forces Fx, Fz and moments My1, My2 from disc effective radius to the axle

Braking forces Fx Fz
Left front axle 701.273 N 701.273 N
Right front axle 701.273 N 701.273 N
Left rear axle 525.954 N 525.954 N
Right rear axle 525.954 N 525.954 N

Moments from braking forces My1 My2

Left front axle 3719.060 Nmm 8978.604 Nmm
Right front axle 3719.060 Nmm 8978.604 Nmm
Left rear axle 2789.295 Nmm 6733.953 Nmm
Right rear axle 2789.295 Nmm 6733.953 Nmm

6.9 Pre-processing procedure through finite element modelling (FEM)

The ANSA pre-processor of the BETA CAE Systems was employed for the preparation
of the model. The model was prepared properly in order to be solved, using the ANSYS
solver, a software package of the Livermore Software Technology Corporation (LSTC)
(LSTC, 2014). By importing the CAD file from Pro Engineer to ANSA, the 14 different
parts of the assembly were recognised as shown in Figure 11 (BETA CAE Systems S.A.,
2014).
278 E.C. Tsirogiannis et al.

6.9.1 Design errors

The geometry problems were located on some faces and are appeared in yellow colour as
shown in Figure 12. The two faces were intersecting. That was a design error which
occurred during the designing process. The redesign process was performed and then the
design error was fixed.

6.9.2 Geometry cleanup

Subsequently, a geometry check was performed. Any triple cons errors were observed in
cyan colour as shown in Figure 13(a). Erasing the surplus cons the geometry errors were
fixed. In the following, intersections check was carried out. Intersections were appeared
in red colour as shown in Figure 13(b) and thereafter were fixed with the use of geometry
fixing tools.

Figure 11 Recognition of the assembly parts (see online version for colours)

Figure 12 Face problem (see online version for colours)

6.9.3 Materials and properties definition

The CFRP properties of the monocoque chassis, the AISI 9000 Series Steel properties of
the front spindle base, the front spindle and the rear semi-axle kit and the Aluminium
6082-T6 properties of the roll bar were inserted into the material forms.
 Optimised ultrafast lightweight design and finite element modelling 279

6.9.4 Shell meshing

With the shell mesh command, an initial mesh at the surfaces of the solid was
constructed. In the field of mesh parameters, the maximum and the minimum dimension
of the mesh were assigned as 5 mm and 3 mm respectively, so as to include a great
number of elements inside the model. Moreover, the meshing scenario was created in
every single part separately. Regions which were anticipated to experience high changes
in stress required a higher finite element mesh density than those which were anticipated
to experience little or no stress variation. Additionally, points of interest such as holes,
fillets or corners were meshed properly.

Figure 13 Geometry cleanup: (a) geometry check and (b) intersections check (see online version
for colours)

6.9.5 Volume meshing

By using the already created surface mesh as a base, the volume mesh was derived by the
mesh volume command. The mesh in the highlighted areas was performed with higher
density compared to the rest of the model and the points of interest were meshed properly
as shown in Figure 14. Finally, the finite element model was composed of 343244 shell
elements (trias) and 561398 volume elements (tetras).

6.9.6 Meshing check

A series of checks, negative volume check, penetration-intersections check, duplicate
elements check and undefined check, were conducted in order to ensure the functionality
of the model. The model was error-free and ready for further processing.
280 E.C. Tsirogiannis et al.

Figure 14 Volume mesh: (a) front end structure and (b) rear end structure (see online version
for colours)

6.9.7 Bolts and contact flanges connections

Bolt connections and contact flanges were applied properly as shown in Figure 15.

Figure 15 Connections: (a) front bolt connections; (b) rear bolt connections; (c) rear contact
flanges and (d) front contact flanges (see online version for colours)
 Optimised ultrafast lightweight design and finite element modelling 281

6.9.8 Loadstep manager

In the loadstep manager, one step was determined. The step includes the permanent static
loads (seat plus the driver load, fuel cell load, battery light load, steering system load and
electric motor load), the total dynamic loads on the front axles which act during the
coexistence of braking and cornering and finally the boundary conditions.

6.9.9 Boundary conditions selection

In the current work, the cornering and braking load case was selected to validate the
strength of the chassis. As a vehicle brakes while it is cornering, an instantaneous weight
transfer occurs to the front, outside axle (Broad and Gilbert, 2009). At this very moment,
the loads are applied especially to the front axles of the vehicle and it could be assumed
that the rear axles of the vehicle remained fixed (Tebby et al., 2011; Prasad et al., 2013).
Thus, the rearmost section of the chassis was selected to be constrained with all DOF’s as
shown in Figure 16 (Stigliano et al., 2010; Moeller et al., 2004).

Figure 16 Loading constraints (see online version for colours)

6.9.10 Permanent static loads

The permanent static loads are the pressure which is applied from the total driver’s
weight plus the seat’s weight, the fuel cell force, the battery light force, the steering
system force and the electric motor force. The fuel cell is assembling at the rear side of
the chassis by four bases. The total load of the fuel cell is applied on the chassis through
these four bases.
The pressure which is applied from the driver (73 kg driver’s weight) plus the seat
(2 kg seat’s weight) was calculated, dividing the force applied by both the driver and the
seat by the area that the driver’s seat takes up on the chassis.
Force = Weight × 9.810 = 75 kg × 9.81 = 735.750 N (38)

Area = 199.858 × 96.568 = 19300.095 mm 2 (39)

Pressure = Force / Area = 0.038 MPa (40)

282 E.C. Tsirogiannis et al.

The permanent static pressure of the total driver’s weight plus the seat’s weight
was applied on the elements of the model as shown in Figure 17(a) and the
permanent static forces (fuel cell forces, battery light force, steering system
force and electric motor force) were applied on the nodes of the model as shown
in Figure 17(b). The battery light, steering system, electric motor forces are 19.62 N,
68.67 N, 29.43 N, respectively. The fuel cell force is divided into four forces resulting in
29.43 N per force.

6.9.11 Front axle loads

Thereupon, the longitudinal, the lateral and the vertical nodal forces as well as
the roll and the pitch nodal moments were applied on the front axles as shown in
Figure 18.

Figure 17 Permanent static loads: (a) pressure applied with the driver on seat and (b) fuel cell
forces, battery light force, steering system force and electric motor force (see online
version for colours)

6.9.12 Gravity acceleration

Finally, the gravity acceleration was taken under consideration (9810 mm/s2) on the
z axis as shown in Figure 19.
 Optimised ultrafast lightweight design and finite element modelling 283

Figure 18 Front axle loads (see online version for colours)

Figure 19 Gravity acceleration (see online version for colours)

6.10 FEM results

The linear static analysis was performed by the Ansys solver. The equivalent von Mises
stresses and deformations of the CFRP monocoque chassis design are presented in
Figure 20. The results present a maximum stress value of 917.20 MPa and deformations
up to 21.38 mm, both being considered acceptable. A 8 GB RAM, Intel i7 2.90 GHz
computer was used and a mean duration of 45 min was needed in order to solve the
problem.
Based on mass, the new chassis design is better than the previous one: its mass
is reduced by 50%. Furthermore, it sustains the mechanical loads while respecting
acceptable maximal displacement and maximal von Mises stress.
284 E.C. Tsirogiannis et al.

Figure 20 FEM results: (a) von Mises stresses and (b) deformations (see online version for
colours)

7 Conclusion

The criteria of comparison of the best design are the weight, the energy consumption, the
ergonomics, the safety and the aesthetic acceptance of the chassis where they have
optimised. As it revealed from our work the new chassis is extremely light, only 5.38 kg
and consequently less energy is consumed to move it, contrary to the previous one that
weighed 10.85 kg. This energy decrease is significantly high, taking into account that the
previous one was the lightest chassis of the competition. Furthermore, the ergonomics,
the safety and the aesthetic acceptance of the new chassis are better than the previous
one.
Consequently, the novelty in this research work is not only the achievement of the
lightest chassis design in the Shell Eco Marathon competition that combined ergonomics,
safety, aesthetic and strength demands but also the overcoming of the time consuming
process by both using the SSS method for an electric car and creating the CLC model.
Furthermore, the methodology for the processing procedure through FEM which has been
developed under a combined stress scenario and by using the exact vehicle’s dynamic
loads could be suggested as an accurate ultrafast method.
Our fast calculations compared to previous studies suggest a new chassis
development method. The resistance of the new chassis design under the aforementioned
extreme stress scenario was demonstrated using the ANSA pre-processor and the Ansys
 Optimised ultrafast lightweight design and finite element modelling 285

solver. Based on that, design optimisation steps will be followed, in the future, so as the
chassis structure will be as light as possible, keeping in parallel a great strength and
stiffness performance.

Acknowledgements

The authors would like to express their gratitude to the TUCER (http://www.tucer.tuc.gr)
team of the Technical University of Crete for providing them the opportunity to work
with data of their racing car, within the framework of the first author's diploma thesis.
Furthermore, we would like to thank BETA CAE Systems for providing access to their
software.

References
Anon (2008a) Formula SAE Rules, Available at: http://fsaeonline.com
Anon (2008b) Pro Engineer User Guide, Available at: http://www.ptc.com
Anon (2014a) Euroncap, Available at: http://www.euroncap.com/
Anon (2014b) Shell Eco Marathon Official Rules, Available at: http://www.shell.com
Anon (n.d.) Material Property Data, Available at: http://www.matweb.com
Baraff, D. (2001) Physically Based Modeling-Rigid Body Simulation, Available at:
http://graphics.cs.cmu.edu/courses/15-869-F08/lec/14/notesg.pdf.
Bender, J., Erleben, K. and Trinkle, J. (2014) ‘Interactive simulation of rigid body dynamics in
computer graphics’, Computer Graphics Forum, Vol. 33, No. 1, pp.246–270, Available at:
http://onlinelibrary.wiley.com/doi/10.1111/cgf.12272/full%5Cnhttp://doi.wiley.com/10.1111/c
gf.12272
BETA CAE Systems S.A. (2014) ANSA and μΕΤΑ User Guide.
Broad, M. and Gilbert, T. (2009) Design, Development and Analysis of the NCSHFH.09 Chassis.
Brown, J.C. and John, A. and Robertson, S.T.S. (2002) Motor Vehicle Structures, Butterworth-
Heinemann.
Budynas-Nisbett (2006) Mechanical Engineering, McGraw-Hill, Available at: www.uotechnology.
edu.iq/.../Mechanical Engineering Design.pd...%5Cn
Carello, M. and Messana, A. (2015) ‘IDRA pegasus: a fuel-cell prototype for 3000 km/L’,
Computer-Aided Design and Applications, Vol. 12, Suppl. 1, pp.56–66, Available at:
http://www.tandfonline.com/doi/full/10.1080/16864360.2015.1077076
Duleba, B. and Greskovic, F. (2013) ‘Simulation of loading the polymer/carbon fiber composites
and prediction of safety factors’, International Journal of Engineering and Innovative
Technology (IJEIT), Vol. 2, No. 8, pp.134–140.
Emri, I. and Voloshin, A. (2016) Statics, Springer-Verlag, New York.
Eurenius, C.A., Danielsson, N., Khokar, A., Krane, E., Olofsson, M. and Wass, J. (2013) Analysis
of Composite Chassis, Chalmers University of Technology.
Gillespie, T.D. (1992) Fundamentals of Vehicle Dynamics, Warrendale: Society of Automotive
Engineers, Inc., Available at: http://books.sae.org/book-r-114
Gritt, P.S. (2002) ‘An introduction to brake systems’, SAE Brake Colloquium, DaimlerChrysler,
pp.1–52.
Hadigheh, S., Gravina, R., Setunge, S. and Kim, S. (2014) ‘Bond characterization of adhesively
bonded joints made with the resin infusion (RI) process’, International Journal of Adhesion
and Adhesives, Vol. 57, pp.13–21, Available at: http://dx.doi.org/10.1016/
j.ijadhadh.2014.10.001
286 E.C. Tsirogiannis et al.

Hakewill, J. (2007) Measuring Center-of-Gravity Height on a Formula Car, Available at:

www.jameshakewill.com
Hamilton, E. and Klang, E. (2009) ‘Design of formula SAE brake systems’, Wolfpack Motorsports,
Raleigh.
Harapin, A., Radnic, J. and Cubela, D. (2008) ‘Numerical model for composite structures with
experimental confirmation’, Materialwissenschaft und Werkstofftechnik, Vol. 39, No. 2,
pp.143–156.
Harris, B. (1999) Engineering Composite Materials, The Institute of Materials, London.
Hull, D. and Clyne, T.W. (n.d.) An Introduction to Composite Materials.
Hulshof, W., Knight, I., Edwards, A., Avery, M. and Grover, C. (2013) ‘Autonomous emergency
braking test results’, Proceedings of the 23rd International Technical Conference on the
Enhanced Safety of Vehicles (ESV), UK, pp.1–13, Available at: http://www-
nrd.nhtsa.dot.gov/Pdf/ESV/esv23/23ESV-000168.pdf
Lee, J.H., Gu, N., Jupp, J. and Sherratt, S. (2012) ‘Evaluating creativity in parametric design
processes and products: a pilot study’, Proceedings of Design Computing and Cognition
(DCC), Springer, Available at: http://mason.gmu.edu/~jgero/conferences/dcc12/
DCC12DigitalProceedings/Digital pdf/Lee.pdf
Limberg, J. (2015) Introduction to Foundation Brake Design.
LSTC (2014) Ansys Mechanical User Guide.
Mat, M.H. and Ghani, A.R.A. (2012) ‘Design and analysis of ‘eco’ car chassis’, International
Symposium on Robotics and Intelligent Sensors (IRIS), pp.1756–1760, Available at: http://
dx.doi.org/10.1016/j.proeng.2012.07.379
Moeller, J.M., Thomas, S.R., Maruvada, H., Chandra, S.N. and Zebrowski, M. (2004)
‘An assessment of an FEA body model for design capability’, Sound and Vibration, pp.24–28.
Nayak, O.A., Kalaivanan, S., Manikandan, D., Ramkumar, G., Manoj, T. and Kannan, M.A. (2012)
‘Complete design and finite element analysis of an all terrain vehicle’, International Journal
of Mechanical and Industrial Engineering (IJMIE), Vol. 2231, pp.85–95.
Omar, S.M.H.S., Arshad, N.M., Fakharuzi, M.H.A.M. and Ward, T.A. (2013) ‘Development of an
energy efficient driving strategy for a fuel cell vehicle over a fixed distance and average
velocity’, Proceedings – 2013 IEEE Conference on Systems, Process and Control, ICSPC
2013, Kuala Lumpur, Malaysia, pp.117–120.
Prasad, T.R.K., Solasa, G., Satyadeep, N.S.D. and Babu, G.S. (2013) ‘Static analysis and
optimisation of chassis and suspension of an all-terrain vehicle’, International Journal of
Engineering and Advanced Technology (IJEAT), Vol. 2, No. 5, pp.1–6.
Prouty, R.W. (1990) Helicopter Performance Stability and Control, Krieger Publishing Company,
Florida.
Reyes Ruiz, C.A., Ramírez Díaz, E.I., Ruiz Cervantes, O. and Rafael Schouwenaars, A.O.P. (2012)
‘Modeling of the suspension of a passenger bus by finite element software’, 3rd International
Conference on Engineering Optimization, Rio de Janeiro, Brazil.
Riley, W.B. and George, A.R. (2002) ‘Design, analysis and testing of a formula SAE car chassis’,
Proceedings of the 2002 SAE Motorsports Engineering Conference and Exhibition, p.382.
Sandner, V. (2006) ‘Development of a test target for AEB systems-development process of a
device to test AEB systems for consumer tests’, ADAC, Germany, pp.1–7.
Schram, R., Williams, A. and van Ratingen, M. (2013) ‘Implementation of autonomous emergency
braking (AEB), the next step in euro NCAP’s safety assessment’, The 23rd International
Technical Conference on the Enhanced Safety of Vehicles (ESV), Belgium, pp.1–6.
Seddon, J. (1990) Basic Helicopter Aerodynamics, BSP Professional Books, Oxford.
Stigliano, G., Mundo, D., Donders, S. and Tamarozzi, T. (2010) ‘Advanced vehicle body concept
modeling approach using reduced models of beams and joints’, PROCEEDINGS OF
ISMA2010 INCLUDING USD2010, pp.4179–4190.
 Optimised ultrafast lightweight design and finite element modelling 287

Tebby, S., Esmailzadeh, E. and Barari, A. (2011) ‘Methods to determine torsion stiffness in an
automotive chassis’, Computer-Aided Design and Applications, Vol. 8, No. 1, pp.67–75.
Tsirogiannis, E. and Vosniakos, G-C. (2019) ‘Redesign and topology optimisation of an industrial
robot link for additive manufacturing’, Facta Universitatis, Series: Mechanical Engineering,
published online 9/2/2019, DOI: 10.22190/FUME181219003T.
Tsirogiannis, E.C. (2015) Design of an Efficient and Lightweight Chassis, Suitable for an Electric
Car, Diploma Thesis, https://www.researchgate.net/publication/313475487_Design_
of_an_efficient_and_lightweight_chassis_suitable_for_an_electric_car, Technical University
of Crete. Available at: https://www.researchgate.net/publication/313475487_Design_of_an_
efficient_and_lightweight_chassis_suitable_for_an_electric_car
Tsirogiannis, E.C. (2017) Reverse Engineering, Redesign and Topology Optimization for Additive
Manufacturing of an Industrial Robot Arm Link, National Technical University of Athens.
Tsirogiannis, E.C., Siasos, G.I., Stavroulakis, G.E. and Makridis, S.S. (2018) ‘Lightweight design
and welding manufacturing of a hydrogen fuel cell powered car’s chassis’, Challenges, Vol. 9,
No. 1, pp.25–40.
Tsirogiannis, E.C., Stavroulakis, G.E. and Makridis, S.S. (2017) ‘Design and modelling
methodologies of an efficient and lightweight carbon-fiber reinforced epoxy monocoque
chassis, suitable for an electric car’, Mater. Sci. Eng. Adv. Res., Vol. 2, No. 1, pp.5–12,
Available at: http://verizonaonlinepublishing.com/MSCPDF/MaterialScienceandEngineering
withAdvancedResear ch21.pdf
Tsirogiannis, E.C., Stavroulakis, G.E. and Makridis, S.S. (2019) ‘Electric car chassis for Shell Eco
Marathon competition: design, modelling and finite element analysis’, World Electr. Veh. J.,
Vol. 10, No. 8, pp.8–21.
Tzanakis, A. (2012) Conceptual and Aerodynamic Design of an Urban Vehicle, Diploma Thesis,
Technical University of Crete.
Van Kerkhoven, J.D.G. (2008) Design of a Formula Student Race Car Chassis, Technische
Universiteit Eindhoven.
Wagtendonk, W.J. (1996) Principles of Helicopter Flight, Aviation Supplies & Academics Inc.,
Washington.
Walker, J. (2005) ‘J. of scR motorsports’, The Physics of Braking Systems, https://www.
scribd.com/document/179047012/the-physics-of-braking-systems-pdf

View publication stats