Journal of the Society of Automotive Engineers Malaysia

Volume 3, Issue 4 (Special Issue), pp 57-63, October 2019

e-ISSN 2550-2239 & ISSN 2600-8092

Finite Element Bus Rollover Test Verification

I. Abdul Hamid*1, K. A. Kamarudin2, M. R. Osman1, A. N. S. Zainal Abidin1, Z. H. Zulkipli1,
Z. Mohd Jawi1, M. H. Md Isa1, M. S. Solah1, A. Hamzah1 and A. H. Ariffin1
1
Malaysian Institute of Road Safety Research (MIROS), Lot 125-135 Jalan TKS 1, Taman
Kajang Sentral, 43000 Kajang, Selangor, Malaysia
2
Universiti Tun Hussein Onn Malaysia (UTHM), 86400 Parit Raja, Batu Pahat, Johor,
Malaysia
*
Corresponding author: iskandar@miros.gov.my

ORIGINAL ARTICLE Open Access

Article History: Abstract – This study is conducted to verify the rollover test of the finite
element bus model. Verification is a process of determining that a
Received computational model accurately represents the theoretical mathematical
4 Dec 2018 model and its solutions. In case of bus rollover simulation, the verification
process is by looking into its energy balance, in which all energy
Received in associated with the bus rollover process must be equal and preserved
revised form
before and after the rollover. Any energy imbalance indicates errors in the
25 Aug 2019
rollover process and must be rectified so that the errors remain in
Accepted acceptable tolerance. The main energies involved in calculations are
1 Sep 2019 potential energy, kinetic energy, internal energy, contact energy, rigidwall
energy, damping energy, hourglass energy, and total energy. The finite
Available online element bus model used is CONTRAST bus developed by CM/E Group,
1 Oct 2019 Politecnico di Milano. The rollover test standard used is UN R66, and the
software used to set up and solve the simulation are LS-PrePost and LS-
DYNA. Energy balance and energy ratio from the rollover simulation
shows that the verification procedure is followed, results are within
acceptable values, and the rollover test is verified.

Keywords: finite element, bus model, rollover simulation, verification, energy

balance

Copyright © 2019 Society of Automotive Engineers Malaysia - All rights reserved.

Journal homepage: www.jsaem.saemalaysia.org.my

1.0 INTRODUCTION

Severe bus rollover crashes usually resulting in a high number of casualties. One of the main
reasons for this is the failure of bus structure when it collapses during the rollover crash. The
structure collapses and intrudes into survivable space of the bus, and this causes the occupants
trapped in the bus. The examples of such cases are shown in Figure 1. The bus structural
deformation is worsening if it is affected by ageing effects (Abdul Hamid and Li, 2013).

57
© Journal of the Society of Automotive Engineers Malaysia
www.jsaem.saemalaysia.org.my

(a) (b)

Figure 1: Examples of severe bus rollover crashes in Malaysia (Hamzah et al., 2012)

Because of this problem, standard bus rollover tests were made available around the
world to define and describe the level of bus structural deformation and its intrusion level into
the survivable space, so that its structural strength can be further improved to at least withstand
its minimum survivable space.

The bus rollover test can be conducted in various ways, e.g. full body rollover test on a
new bus, rollover test using bus body sections, body sections quasi-static loading test, or by
quasi-static calculation. For full-body rollover, bus body sections rollover and quasi-static
loading test, it can be done either by real test or visual test (UNECE, 2006). Visual test refers
to computer simulation test using finite element (FE) software. For this test, the finite element
model needs to be verified to ensure its results are accurate and close to the real test results.
The purpose of this paper is to show how the process of FE bus model rollover test is verified,
and how the results of the rollover should be within the acceptable range.

2.0 METHODOLOGY

2.1 Selection and Rollover of FE Bus Model

The finite element bus is modelled from a single deck high-floor, intercity CONTRAST bus
assembled by VEST-BUSSCAR AS Company, which is the distributor for Iveco bus in
Norway and Sweden (Revolvy, 2017). The bus model was developed by Pezzucchi (2016), and
further improved and validated by Cortese and Gazzaniga (2014). LS-Prepost software was
used to apply parameters in the bus model, e.g. material models, assignment of properties,
element formulation, contact definition, initial conditions, and boundary conditions. The solver
used to simulate the bus rollover is LS-DYNA. The rollover of the bus was setup according to
UN R66 regulation (UNECE, 2006).

Figure 2 shows the snapshots of bus rollover simulation. Figure 1(a) shows the bus at its
unstable condition (49 o) at t = 0.0 sec. The bus then slowly falls and hits the ground at t = 1.36
sec. The contours of effective plastic strain (unitless) can be seen at several locations of the
structural points after the impact with the maximum value of 5.077 x103. The bus eventually
at its rest position after t = 2.00 sec.

58
© Journal of the Society of Automotive Engineers Malaysia
www.jsaem.saemalaysia.org.my

Figure 2: Snapshots of the bus rollover simulation in its original condition at

(a) t = 0.0 sec. (b) t = 1.2 sec. (c) t = 1.4 sec. (d) t = 1.6 sec. (e) t = 1.8 sec. (f) t = 2.0 sec

2.2 Verification by Energy Balance

Verification is the process of determining that a computational model accurately represents the
theoretical mathematical model and its solution (Schwer, 2007). In case of bus rollover
simulation, verification can be made by checking the quantity of energy balance, deformations,
kinematics, kinetics and other quantities (Matolcsky and Molnar, 1999). Bojanowski (2009)
introduced the procedure of bus rollover simulation, in which all energy components during
the rollover should be conserved and all non-physical energy should be kept minimum.
Bojanowski (2009) specifies the conservation of energy equation, in which
0 0
𝐸𝐿𝑆 𝑡𝑜𝑡𝑎𝑙 = 𝐸𝑘𝑖𝑛 + 𝐸𝑖𝑛𝑡 + 𝐸𝑠𝑖 + 𝐸𝑟𝑤 + 𝐸𝑑𝑎𝑚𝑝 + 𝐸ℎ𝑔 = 𝐸𝑘𝑖𝑛 + 𝐸𝑖𝑛𝑡 + 𝑊𝑒𝑥𝑡 (1)

where

𝐸𝐿𝑆 𝑡𝑜𝑡𝑎𝑙 = LS-DYNA total energy, 𝐸𝑑𝑎𝑚𝑝 = system damping energy,

𝐸𝑘𝑖𝑛 = kinetic energy, 𝐸ℎ𝑔 = hourglass energy,
𝐸𝑖𝑛𝑡 = internal energy, 0
𝐸𝑘𝑖𝑛 = initial kinetic energy,
𝐸𝑠𝑖 = contact (sliding) energy, 0
𝐸𝑖𝑛𝑡 = initial internal energy,
𝐸𝑟𝑤 = rigidwall (stonewall) energy, 𝑊𝑒𝑥𝑡 = external work from applied loads.

59
© Journal of the Society of Automotive Engineers Malaysia
www.jsaem.saemalaysia.org.my

The energy balance is perfect if the energy ratio is equal to 1 (LSTC, 2015).
𝐸𝑡𝑜𝑡𝑎𝑙
𝑒𝑟𝑎𝑡𝑖𝑜 = 0 (2)
𝐸𝑡𝑜𝑡𝑎𝑙 + 𝑊𝑒𝑥𝑡

where
0
𝐸𝑡𝑜𝑡𝑎𝑙 = initial total energy

The acceptable value of the ratio is 1.00 +/- 0.07 (Pezzucchi, 2016). The discrepancies
in the ratio are due to spurious energy produced in the system and simulation error.

3.0 RESULTS AND DISCUSSION

From GLSTAT ASCII file in LS-Prepost, the energy balance for the rollover simulation was
extracted and is shown in Figure 3. Potential energy, Epot and the real Total Energy, Etotal were
hand calculated and added in the figure in the dotted lines.

Figure 3: Energy balance for the rollover simulation of the CONTRAST bus

The summary of energies expected theoretical results, actual results and their level of
tolerance are shown in Table 1. Note that the unit used for energy in Figure 2 is tonne.mm2.sec.2
which is used in LS-Prepost. This is equivalent to the typical unit of kilojoule (kJ).

Before impact, 𝐸𝑘𝑖𝑛 exponentially increased until it reached its maximum value of 98.2
kJ during impact at t = 1.36 sec. It then suddenly decreased and stabilised at 3.1 kJ. After the
impact, the kinetic energy transforms into several mechanical works like 𝐸𝑖𝑛𝑡 , 𝐸𝑠𝑖 , 𝐸𝑟𝑤 , and
𝐸ℎ𝑔 . These energies emerged but within their acceptable values, while 𝐸𝑑𝑎𝑚𝑝 remains zero
throughout the simulation.

60
© Journal of the Society of Automotive Engineers Malaysia
www.jsaem.saemalaysia.org.my

Potential energy, Epot, is calculated using the typical equation of

EP = Mg∆ℎ (3)

where

M = total mass of the bus

g = gravitational acceleration (constant)

∆h = vertical movement of the bus from the highest and unstable position of CG of the bus to its
final location.

Table 1: Summary of energies obtained during the FE bus rollover simulation

Type of energy Expected Expected % of Actual Actual % of Tolerance

theoretical 𝑬𝑳𝑺 𝒕𝒐𝒕𝒂𝒍 results (kJ) 𝑬𝑳𝑺 𝒕𝒐𝒕𝒂𝒍
results (kJ)
𝐸𝐿𝑆 𝑡𝑜𝑡𝑎𝑙 (at t = 3.0 sec.) - - 131.0 100 -
𝐸𝑘𝑖𝑛 (max. at point of - - 98.2 75.0 acceptable
impact when t = 1.36
sec.)
𝐸𝑘𝑖𝑛 (at t = 3.0 sec.) - - 3.1 2.4 acceptable
𝐸𝑖𝑛𝑡 (at t = 3.0 sec.) - - 84.2 64.3 acceptable
𝐸𝑠𝑖 (at t = 3.0 sec.) - - 3.3 2.5 acceptable
𝐸𝑟𝑤 (at t = 3.0 sec.) 32.75 25 35.5 27.1 acceptable
𝐸𝑑𝑎𝑚𝑝 (throughout 0.0 0.0 0.0 0.0 acceptable
simulation)
𝐸ℎ𝑔 (at t = 3.0 sec) < 6.55 <5 4.9 3.7 acceptable
Epot (at t = 0.0 sec) - - 131.0 100.0 acceptable
Etotal (throughout 131.0 100.0 acceptable
simulation)

At the highest, unstable position of CG, Epot is 131 kJ (at t = 0.0 sec.). The value decreases
until the point of impact, where the value stabilises and reaches its zero level. By adding Epot
and 𝐸𝐿𝑆 𝑡𝑜𝑡𝑎𝑙 , the real Total Energy, Etotal is obtained. Etotal is maintained throughout the
simulation, which is 131 kJ. This shows the total energy is conserved before, during and after
the bus impact. The energy ratio obtained from the simulation is shown in Figure 4. The ratio
is in between 0.95 and 1.06, which is within the acceptable range for further analysis.

61
© Journal of the Society of Automotive Engineers Malaysia
www.jsaem.saemalaysia.org.my

Figure 4: Energy ratio for the FE bus rollover simulation

4.0 CONCLUSION

From the results of energy balance, it can be seen that the energy components throughout the
simulation of CONTRAST bus rollover, together with its energy ratio, are within the acceptable
range. Hence, it is concluded that the bus rollover is verified. It can be used as a baseline for
further analysis.

REFERENCES

Abdul Hamid, I., Li, Q.M. (2013). Ageing Effect on Crashworthiness of Bus Rollover, 9th European LS-
DYNA Conference 2013, Manchester, UK.

Bojanowski, C., (2009). Verification, validation and optimization of finite element model of bus
structure for rollover test. Florida State University.

Cortese, M., & Gazzaniga, M. (2015). Numerical approach to assess critical installation of vehicle
restraint systems (Master's thesis, [M. Cortese & M. Gazzaniga]).

Hamzah, A., Manap, A.R.A., Muntalip, M.H., Solah, M.S., & Voon, W.S. (2012). Heavy Commercial
Passenger Vehicle Service Life in Malaysia (No. MRev 01/2012). Malaysian Institute of Road
Safety Research (MIROS).

LSTC (2015). Total Energy [Homepage of LSTC], Retrieved from https://www.dynasupport.

com/howtos/general/total-energy

Matolcsky, M., Molnar, C., (1999). Bus Rollover Test as a Process and Its Energy Balance. 30th
Meeting of Bus and Coach Experts, Gyor, Hungary.

62
© Journal of the Society of Automotive Engineers Malaysia
www.jsaem.saemalaysia.org.my

Pezzucchi, M., (2016). Norwegian Public Roads Administration, e-mail:

matteo.pezzucchi@vegvesen.no, (Personal Conversation).

Revolvy, (2017). Vest Vusscar, Retrieved from

https://www.revolvy.com/main/index.php?s=Vest%20Busscar&item_type= topic

Schwer, L.E., (2007). Guide for Verification and Validation in Computational Solid Mechanics,
American Society of Mechanical Engineers (ASME) Standards Committee on Verification and
Validation in Computational Solid Mechanics (PTC 60/V&V 10).

UNECE (2006). Regulation 66, Addendum 65, Uniform technical prescriptions concerning the
approval of large passenger vehicles with regard to the strength of their superstructure. United
Nations Economic Commission for Europe (UNECE).

63