Bus Rollover and Impact Review

Review Report
Stability of High-Deck Bus in Rollover and

Contact-Impact with Traffic Barriers
Review Report
Stability of High-Deck Bus in Rollover and Contact-Impact with Traffic Barriers
Disclaimer
The information contained in the report is restricted and/or privileged information and is intended
only for authorised screening and/or confidential presentation of MIROSs discretion. This report
should not be disseminated, modified, copied/plagiarised or action taken in reliance upon it,
unless permitted by MIROS. None of the materials provided in this report may be used,
reproduced or transmitted. In any form or by any means, electronic or mechanical, including
recording or used of any information storage and retrieval system, without written permission
from MIROS. Any conclusion and opinions in the report may be subject to reevaluation in the
event of any forthcoming additional information or investigations. MIROS declares that all the
inquiries which MIROS believes are necessary and appropriate and that nothing significance
which MIROS regards as relevant have, to MIROS knowledge, been withheld from the report.
1
Page
(Source of photos on cover: RMP and Sin Chew Daily)

Review Report
Table of Contents
Pages
9 Acknowledgements
10 Abstract
11 1. Introduction
12 2. State of Current Local Bus Construction
12 2.1 Bus Construction and Regulations
12 2.1.1 Adoption of UNECE Regulations in Malaysia
15 2.1.2 Overview of UNECE Regulation 66
15 2.1.3 Overview of UNECE Regulation 107
17 2.1.4 Malaysias Road Transport Rules
18 3. Physics of Rollover
18 3.1 Stability Issue Determination of Centre of Gravity (CG)
18 3.1.1 Definition of Rollover
19 3.1.2 Factors of Rollover
22 3.1.3 Centre of Gravity Calculation
26 4. Longitudinal Traffic Barriers
28 4.1 Flexible Barrier Systems
28 4.1.1 General Description and Behaviour Under Impact
31 4.1.2 Crash Test and Result
32 4.1.3 Advantages and Disadvantages
33 4.2 Semi-Rigid Barrier Systems
35 4.2.2 Crash Tests and Results
37 4.3 Rigid Barrier Systems
38 4.3.2 Crash Tests and Results
43 5. Analysis of High-Deck Bus vs. Traffic Barrier Collision
2
43 5.1 Impact Severity Analysis

Page
45 5.2 Rollover Analysis Bus-Traffic Barrier Collision

Review Report
45 5.2.1 Rollover on Straight Road

47 5.2.2 Rollover at Curve Road
48 5.2.3 Design of Vehicle
49 6. Conclusion
49 7. Recommendations
49 7.1 Bus Construction
51 7.2 Traffic Safety Barrier Consideration
52 References
3
Page
Review Report
List of Figures
Pages
13 Figure 2-1 Conventional frame for single-deck bus/coach

13 Figure 2-2 Example of single-deck superstructure frame
19 Figure 3-1 Relationship between CG and rollover
20 Figure 3-2 Diagram of bus before pivoting
21 Figure 3-3 Critical point of rollover
22 Figure 3-4 Lateral component of vehicle moving with velocity ( ) and angle ( )
23 Figure 3-5 Longitudinal position of CG
24 Figure 3-6 Transverse position of CG
24 Figure 3-7 Tilting test for determining the height of CG
29 Figure 4-1 Cross section of two typical designs of wire rope barrier system
34 Figure 4-2 Profiles of semi-rigid safety barriers used on road shoulder and median
37 Figure 4-3 Profiles of rigid barriers
41 Figure 4-4 General profile of STEP barrier
43 Figure 5-1 Impact severity plot
45 Figure 5-2 Free Body Diagram (FBD) of bus turning moment
46 Figure 5-3 FBD of bus turning moment for collision with rigid barrier
4
Page
Review Report
List of Plates
Pages
30 Plate 4-1 Flexible wire rope barrier at Kuala Lumpur-Karak Expressway (E8) in
Gombak
30 Plate 4-2 Flexible wire rope barrier on flyover at Middle Ring Road 2 (MRR2) in
Gombak
31 Plate 4-3 Flexible wire rope barrier on median at Second Link Expressway (E3)
near Perling toll plaza, Johor
34 Plate 4-4 Single-mounted W-beam guardrail median at North-South Expressway
(E2) in Serdang
35 Plate 4-5 Double-mounted W-beam guardrail median at Kepong road
35 Plate 4-6 Customized stacked mounted W-beam guardrail at Karak Expressway
(E8)
38 Plate 4-7 Concrete barrier at MRR2 in Selayang
38 Plate 4-8 Concrete barrier with antiglare screen at Karak Expressway (E8)
5
Page
Review Report
List of Tables
Pages
25 Table 4-1 Test Level for Longitudinal Traffic Barrier

28 Table 4-2 Typical Severity Indices for Safety Barriers and Various Design Speeds
32 Table 4-3 Results for Deflections of Crash Tested Wire Rope Safety Barriers
36 Table 4-4 General Summary for Semi-Rigid Barriers Installation
41 Table 4-5 Crash Test Data Summary
43 Table 5-1 Results of Impact Severity Calculation Analysis
44 Table 5-2 Impact Severity Comparison
47 Table 5-3 Minimum Radius of a Curve
6
Page
Review Report
List of Acronyms and Abbreviations Used
4WD Four-Wheel Drive

AA Average Acceleration
ADR Australian Design Rules
ARRB Australian Road Research Board
ASI Acceleration Severity Index
CG Centre of Gravity
CRASE Crash Safety Engineering Unit of MIROS
EN European Norm
FBD Free Body Diagram
FHWA Federal Highway Administration
(A division of the United States Department of Transportation)
IS Impact Severity
LLM Lembaga Lebuhraya Malaysia
MIROS Malaysian Institute of Road Safety Research
MOT Ministry of Transport
MRR2 Middle Ring Road 2
(Federal Route 28 which connects Kuala Lumpur and Selangor)
NCHRP National Cooperative Highway Research Program
(United States)
NHTSA National Highway Traffic Safety Association
(An agency of the Executive Branch of the United States Government
Department of Transportation)
OIV Occupant Impact Velocity
ORA Occupant Ridedown Acceleration
PCB Portable Concrete Barrier
PHD Post-Impact Head Deceleration
PRA Protectable Rollover Accident
PWD Public Works Department
(Jabatan Kerja Raya JKR)
R&R Rest and Recuperate
7
REAM Road Engineering Association of Malaysia

Page
Review Report
RMP Royal Malaysian Police

(Polis Diraja Malaysia PDRM)
RSD Road Safety Department
(Jabatan Keselamatan Jalan Raya JKJR)
RTD Road Transport Department
(Jabatan Pengangkutan Jalan JPJ)
SI Severity Index
SSF Static Stability Factor
THIV Theoretical Head Impact Velocity
TL Test Level
TRAPTER Transport Planning and Traffic Engineering Unit of MIROS
UNECE United Nations Economic Commission for Europe
UMTRI University of Michigan Transportation Research Institute
VSB Vehicle Safety and Biomechanics Research Centre of MIROS
WRSB Wire Rope Safety Barriers
8
Page
Review Report
Acknowledgements
This study would not have been possible without the considerable support and cooperation of
many talented people, who have played a part, directly or indirectly, throughout its completion.
The authors would like to express their sincere gratitude to: (i) Road Transport Department
(RTD) of Malaysia and local coachbuilders for their cooperation in providing input related to the
local bus construction and regulation, and (ii) Malaysian Highway Authority (LLM) for their
assistance on traffic barrier matters. Special thanks are due to the following individuals for all
the help and advice to make completion of the study and its review report possible.
Malaysian Institute of Road Safety Research (MIROS)
Prof. Dr. Wong Shaw Voon Director General

Prof. Dr. Ahmad Farhan Mohd Sadullah Former Director General
Ir. Muhammad Marizwan Abdul Manan Unit Head TRAPTER
Crash Safety Engineering (CRASE) Unit
Publication and Knowledge Management Unit
Authors
Aqbal Hafeez Ariffin

Mohd Khairudin Rahman
Mohd Syazwan Solah
Khairil Anwar Abu Kassim
9
Page
Review Report
Abstract
Despite the positive impact of high-deck buses for the operators by increasing their profits
through reduction of fuel consumption and vehicle operating maintenance as well as ability to
carry more passengers per trip compared to single-deck buses, the increasing popularity of the
high-deck buses in Malaysia has become a major concern in road safety. As a result of a fatal
accident involving a high-deck bus in Behrang (2008) which saw failure of median guardrail to
contain and redirect the errant bus on track, a feasibility study about the suitable traffic safety
barrier to be used on Malaysias expressways specifically for single vehicle collision involving
high-deck buses was initiated. The study was carried with two main objectives: (i) to analyze
stability of high-deck bus in rollover collision with crash barrier using calculation method based
on formulas derived in few related literatures, and (ii) to assess existing literatures for any type
of traffic safety barriers currently available worldwide for recommendation on adopting specific
crash barrier for high-deck buses to be used in Malaysias highways or roads. From the study, it
is concluded that location of CG and speed influence stability of a high-deck bus during rollover
event, especially at curve road, and rigid barrier performs better than semi-rigid barrier in
preventing rollover of high-deck bus during collision with traffic barriers.
10
Page
Review Report
1. Introduction
The use of high-deck buses as passenger vehicles for long-distance travel in Malaysia is
gaining popularity among coach operators. This trend of increasing usage of high-deck
buses by coach operators is most probably due to their capabilities of having more seats
(able to carry more passengers per trip) compared to the single-deck buses. In addition, by
shifting the operation from single to high-deck busses, coach operators can cut the cost of
fuel consumption and reduce vehicle operating maintenance, hence, increasing their profit.
However, there is a negative impact with this shift. If the high-deck bus is designed with full
headroom throughout both decks, there is a possibility for the bus to be top heavy and
become unstable (except with the use of counterweight system), thus unsafe for travelling.
This is a major concern in road safety since it can lead to serious high-deck bus accidents if
the use of this type of commercial vehicle kept on increasing, unless safety
countermeasures are considered and implemented.
The Malaysia Road Safety Department (RSD) formally requested the Malaysian Institute of
Road Safety Research (MIROS) to conduct a feasibility study on the suitable traffic safety
barrier to be used on Malaysias expressways specifically for single vehicle collision
involving high-deck buses. A team from the Vehicle Safety and Biomechanics (VSB)
Research Centre led by the Director took up the challenge to carry out the study which
focused on (i) overview of local bus construction including adaptation of United Nations
Economic Commission for Europe (UNECE) regulations, (ii) types of traffic barrier systems
currently available or used in Malaysias highways or roads, (iii) analysis of high-deck bus
collisions with crash barrier, and (iv) recommendations on adopting specific crash barrier for
high-deck buses.
This study was carried out as a preliminary literature research. All data gathered and used
to meet the objectives of this study was gained from related literatures, publications or
reports since no actual crash test and research regarding high-deck bus collision with crash
barrier was performed, the crash test results for assessment of existing traffic safety barriers
included in this report might not cover all crash tests carried out worldwide. However, only
11
selected crash tests results were used for comparison purpose.

Page
Review Report
2. Current State of Local Bus Construction
2.1 Bus Construction and Regulations
2.1.1 Adoption of UNECE Regulations in Malaysia
Conventional bus/coach structure manufacturing involves labour-intensive arc welding

of tube stock. In order to build the body of a bus, it must consist of five separate major
units: side, floor, roof, front and rear units. All of these units are constructed separately
and they will be joined together later during the final assembly. After the final structure
assembly, the frame is subjected to grit blasting and zinc phosphate coating. Most of
the structure joining processes is done using the arc weld method. Figure 2-1 shows
an example of a conventional frame for single-deck bus/coach.
In all vehicles studied, fractures at the welding connection took place after a given
period of service. This failure mechanism was identified as fatigue and it influenced the
structures integrity, especially when the bus was involved in a serious accident such
as a rollover.
Late in 2007, a bus accident occurred in Bukit Gantang, Perak and killed 23 people
onboard. Since that incident, the Malaysian government reinforced the bus
construction law to protect customers and coach builders. Adaptation of UNECE
Regulation 66 Uniform Technical Prescriptions Concerning the Approval of
Large Passenger Vehicles With Regard To the Strength of Their Superstructure
into Malaysias Road Transport Rule was done and the new regulation was enforced
by the Road Transport Department (RTD). However, a grace period was given to
coach builders to implement the rules so as to avoid sudden burden to the coach
builders.
In Malaysia, there are a large number of coach builders and they are moving towards
12
that to make sure their market can be expanded not only in Malaysia but also to other
countries. As of to date, many of the coach builders had already implemented the
Page
UNECE R66 in their bus or coach body construction practice. The regulation basically
Review Report
focuses on the body frames integrity and durability. The main frame itself must be built
with a continuous transversal frame to sustain it in a rollover impact as shown in
Figure 2-2.
Apart from that, the regulation also highlights the method to verify the bus or coach
structure. However, the method to verify the bus or coach structure is not valid for
high-deck buses.
There is a another UNECE regulation that states the special requirement for double-
deck vehicle, which is the UNECE Regulation 107 Uniform Provisions
Concerning the Approval of Double-deck Large Passenger Vehicle With Regard
To Their General Construction. However, there is no specific requirement stated in
the regulation to verify the integrity of the double-deck super structure.
Non-continuous
transversal frame
Figure 2-1 Conventional frame for single-deck bus/coach

(Source: F. Lan et al., 2004)
13
Page
Review Report
Figure 2-2 Example of single-deck superstructure frame

(Source: E. Larrod et al., 1995)
14
Page
Review Report
2.1.2 Overview of UNECE Regulation 66
The UNECE R66 was initiated to prevent severe damage on buses, thus ensuring the
safety of bus passengers, via mitigation of injuries sustained, in the case of rollover.
By definition, superstructure refers to the components of a bus structure that contribute
to the strength of the vehicle in the event of a rollover. Ultimately, an UNECE R66
approved bus is supposed to be able to withstand impact from rollover accident such
that the residual space is intact during and after the accident. In other words, intrusion
to residual space with luggage should not happen and no parts from the residual
space should be projected outside in the event of a rollover.
There are few equivalent methods for approval test other than rollover test on a
complete vehicle. It is also acceptable to carry out the rollover or quasi-static test on
body sections which are representative of the complete vehicle. Other equivalent tests
are quasi-static calculations based on the results of component tests and computer
simulation via dynamic calculations.
Currently, the existing UNECE R66 regulation relates only to large, single-deck buses.
Double-deck buses are excluded from it. There is no regulation for the strength of
superstructure for double-decker coaches.
High-deck buses are also not well represented in the UNECE R66 regulation. The
existing approval test is not specifically tailored for it even though some modifications
of the test can be done to better suit high-deck buses. The main problem is that the
approval test does not appropriately separate the weak superstructure from the strong
one (UNECE R66, 2006).
2.1.3 Overview of UNECE R107 Regulation
The following is an overview of R107 Uniform Provisions Concerning the

15
Approval of Double-deck Large Passenger Vehicles with Regards to Their

Page
General Construction Annex 9.

Review Report
The regulation applies to every type of single-deck bus, double-deck bus, rigid
or articulated vehicle in category M2 (vehicles used for the carriage of
passengers, comprising more than eight seats in addition to the driver's seat,
and having a maximum mass not exceeding five tonnes) and M3 (vehicles
used for the carriage of passengers, comprising more than eight seats in
addition to the driver's seat, and having a maximum mass exceeding five
tonnes).
The requirement is an additional regulation for double-deck buses. However,
Annex 3 shall apply to double-deck vehicles if there is no regulation that is not
stated in this requirement.
Annex 9 encloses the special requirements for double-deck vehicle. The
abstract requirements are :
Total load of vehicle should be in running order. The load of passengers
shall be placed on each upper deck passenger seat. If the vehicle is
intended to be used with a crew member who is not seated, the centre of
gravity (CG) of the mass of 75kg representing the crew member shall be
placed in the upper deck gangway at a height of 875mm. The baggage
compartments shall not contain any baggage.
Fire extinguishers and first aid equipment shall be provided at two places,
one near the driver and one on the upper deck.
The number of exits (including service and emergency required for
double-deck bus) depends on the number of passengers and crew
onboard. Technical requirements of each door are stated.
The design of intercommunication staircases as an access way (between
lower and upper deck) shall not be endangering of passengers being
projected downwards during heavy braking. This requirement is
considered to be fulfilled if at least one of the following conditions is met:
No part of the staircase is forward descending;
The staircase is equipped with guards or a similar provision;
There is an automatic device in the upper part of the staircase which
prevents the use of the staircase when the vehicle is in motion; this
16
device shall be easily operable in an emergency.

Page
Review Report
Handrails or/and handhold for inter communications staircases shall be

provided at each side including a grasping point available to a person
standing on the lower and upper deck adjacent to the inter
communication staircase.
The requirement for seating position for lower and upper deck states that
each seat shall have a free height of not less than 900mm measured from
the highest point of the uncompressed seat cushion. This free height shall
extend over the vertical projection of the whole area of the seat and the
associated foot space. In the case of the upper deck, this free height may
be reduced to 850 mm.
The maximum emergency doors' steps are 850mm for lower deck and
1500mm for upper deck.
The gangways condition is different from a single-deck vehicle. The
gauging device is designed for upper deck to fulfil the requirement.
2.1.4 Malaysias Road Transport Rules (Compilation of 46 Rules, version 25th February
2008)
Stability test for high-deck vehicle is also stated in Motor Vehicle Rules (Construction
and Use 1959). The mass of the driver and passengers can be replaced by other
elements equivalent to represent the mass of the upper deck compartment. The
vehicle shall be tilted without rocking and without dynamic effects until it reaches
unstable equilibrium and commences its rollover. The maximum angle of tilt is 28
degrees and the vehicle shall fail if it rolls over before reaching the maximum angle.
Other requirements stated in the rules for high-deck vehicle focus on the construction
of the vehicle itself. 17
Page
Review Report
3. Physics of Rollover
3.1 Stability Issue Determination of Centre of Gravity (CG)
3.1.1 Definition of Rollover
It is apparent that UNECE R66 is focusing on the ability of a vehicle to sustain its
residual space after a rollover crash. Rollover crash is by far the worst scenario
possible for a vehicle to undergo because the resulting damage could be very
extreme, as compared to other types of crashes such as head-on and side collisions.
Due to the rollover crashs nature, all bus occupants have a higher possibility to
sustain serious to fatal injury when the vehicle is toppling upside down. At that point of
time, structure of the roof is vulnerable to intrusion and projection.
Rollover causes can be generally divided into two main categories; tripped and
untripped (Deshmukh, 2006). A tripped rollover is caused by an object that a vehicles
tires comes in contact with, abruptly stopping the lateral motion of the tire and sending
it to roll around that object. Examples of tripping objects are curbs, ramps, rocks, and
soil. Untripped rollover usually occurs because of severe steering manoeuvres such as
J-hooks, lane changes, and fast turns.
Matolcsy (2003), in his analysis of rollover cases throughout Europe, suggested that
rollovers can be categorized into different groups based on a few characteristics.
Mostly, the characteristics lie around the number of rotations experienced by a vehicle
during the rollover event. For example, turn on side equals rotation. Turn into a ditch
is between and rotation, and rollover from the road is between to two full
rotations. These three types of rollovers fall into the Protectable Rollover Accidents
(PRA) category, a kind of rollover accident that when the bus occupants are involved,
they have a high probability of survival. Other kind of rollovers are serious rollover
(more than two rotations), and combined rollover, a rollover followed by a fire, fall into
18
a lake, etc.
Page
Review Report
3.1.2 Factors of Rollover
There are a few factors that can influence the probability of the occurrence of a
rollover. These include position of the vehicles CG, vehicle speed, angle of impact
upon collision with barrier, and barrier as tripped factor. The first three factors are
explained in this section while in-depth details about barrier are explained in the next
section.
a) Centre of Gravity
Rollover of a vehicle is directly related to its CG in the sense that a vehicle is
unlikely to rollover if its CG is in the region of gravitational patch, viewed from the
top. The CG is always balanced within the gravitational geometry of some support
structure. For this case, it is initially formed by the tires. A vehicle rolls over
because its CG is no longer balanced within the gravitational geometry formed by
the tires. It is now being contained by another support structure, which is the
offside of the bus.
19
Figure 3-1 Relationship between CG and rollover (Illustrated with gravitational

patch)
Page
Review Report
In Figure 3-1, the vehicle on the left is on level ground. From the top, the CG is
contained within the gravitational support structure (formed by tires). Therefore
rollover will not happen, according to the basic laws of physics.
The vehicle on the right is bus cambering to the driver side. From the top, the CG
has moved outside the gravitational support structure. Therefore, rollover will
occur.
b) Speed
Rollover occurs when a vehicle is moving or sliding sideways until it strikes a solid
object such as a curb. The curb provides a pivot point for the vehicle to rotate. The
figures below show more details on rollover physics.
CG
Figure 3-2 Diagram of bus before pivoting
Figure 3-2 shows a vehicle with height of CG, and track width, is moving
sideways with lateral velocity, . The vehicle is about to trip at the pivot point,
which is represented by the curb.
20
Page
Review Report
CG
Figure 3-3 Critical point of rollover
At this point, the vehicle is at the critical point for rollover. The critical point is
achieved when the CG is at the maximum height. According to Pythagoras
Theorem, at the maximum CG height, equals the hypotenuse of a right angle with
sides and .
r = ( ) (Equation 1)
Energy conservation requires that kinetic energy (lateral) before pivoting equals
potential energy at the critical point. The end result yields a formula of critical
speed, ,
( ) (Equation 2)
where
The value of or widely known as Static Stability Factor (SSF) is adopted by

National Highway Traffic Safety Administration (NHTSA) as the parameter for
rollover tendency.
c) Angle of Impact
For a vehicle to rollover as a result of sliding sideways, there shall be lateral
21
movement of the vehicle. In a real-life rollover crash, vehicle approaches pivot

point (curb, barrier, or solid object) at a certain angle. The magnitude of the angle
Page
can influence whether or not rollover will occur as illustrated in the figure below.
Review Report
Figure 3-4 Lateral component of vehicle moving with velocity ( ) and angle ( )
Using trigonometry theorem, the lateral speed of the bus can be calculated by the
following equation.
(Equation 3)
The theorem proves that a narrow angle of impact will certainly lower the
likelihood of rollover from occurring. This is because the lateral component of the
vehicles speed is less as the angle of impact decreases.
3.1.3 Centre of Gravity Calculation
A rollover accident is directly related to the CG of the vehicle. To determine the

position of the CG of a bus, three parameters need to be defined; the longitudinal
distance ( ) from the centre line of gravity, the transverse distance ( ) from the
vertical longitudinal central plane of the vehicle, and the vertical height ( ) above the
flat horizontal ground level when the tires are inflated. The CG can be calculated with
or without considering the effect of the total occupant mass.
The transverse position ( ) of the vehicles CG needs to be determined first in order to

find the vertical height ( ) of the CG. Furthermore, the bus needs to be tilted
22
longitudinally to find the load cells at the wheels of two axles while tilting.
Page
Review Report
The following formulas to determine the CG of a bus are derived from UNECE R66
regulation.
Figure 3-5 Longitudinal position of CG

(Source: UNECE R66, 2006)
The longitudinal position ( ) of the CG relative to the centre of the contact point of the
front wheels (Figure 3-5) is given by,
( ) ( )
(Equation 4)
where:
P1 = reaction load on the load cell under the left-hand wheel of the first axle,
P2 = reaction load on the load cell under the right-hand wheel of the first axle,
P3 = reaction load on the load cell under the left-hand wheel(s) of the second axle,
P4= reaction load on the load cell under the right-hand wheel(s) of the second axle,
P5 = reaction load on the load cell under the left-hand wheel(s) of the second axle,
P6 = reaction load on the load cell under the right-hand wheel(s) of the second axle,
Ptotal = P1+P2+P3+P4+P5+P6 = Mk unladen kerb mass; or,
= Mt total effective mass,
L1 = wheelbase distance from 1st axle to 2nd axle, and
L2 = wheelbase distance from 2nd axle to 3rd axle, if fitted.
23
Page
Review Report
Figure 3-6 Transverse position of CG

The second parameter, which is the transverse position (t) of the CG relative to its
longitudinal vertical centre plane as shown in Figure 3-6 is given by,
(( ) ( ) ( ) ) (Equation 5)
where:
T1 = track width of 1st axle,
T2 = track width of 2nd axle, and
T3 = track width of 3rd axle.
Bear in mind that the CG is situated to the right of the centre line of the bus if the
value of is negative and it goes the other way around if the value is positive.
24
Figure 3-7 Tilting test for determining the height of CG

Page

Review Report
Finally, the third parameter of the CG, the vertical height shall be determined by tilting
the vehicle longitudinally and using individual load-cells at the wheels of two axles.
Refer Figure 3-7 for more comprehension.
In determining the vertical height ( ), more accurate calculation shall be obtained if

the angle is greater. Initially, the inclination of the tilting test shall be determined by
the equation
( ) (Equation 6)
Where H is the height difference between the footprints of the wheels of the 1st and
2nd axles and is the wheelbase distance between 1st and 2nd axles.
Next, the unladen kerb mass of the bus shall be checked as follows:
(Equation 7)
where:
F1 = reaction load on the load cell under the left hand wheel of the 1st axle,
F2 = reaction load on the load cell under the right hand wheel of the 1st axle,
F3 = reaction load on the load cell under the left hand wheel of the 2nd axle, and
F4 = reaction load on the load cell under the right hand wheel of the 2nd axle.
Then, the angle measured from inclination test and the resultant loads shall be used
in the vertical height calculation of the CG, which is given by:
( )( ) (Equation 8)
where is the height of wheel centre (on first axle) above the load cell top surface.
25
Page
Review Report
4. Longitudinal Traffic Barriers
Currently, there are three distinct longitudinal safety traffic barriers (flexible, semi-rigid and
rigid) being used on the roads and expressways in Malaysia. The following sections
describe a number of typical traffic safety barriers but the list does not contain data of all
available barrier systems.
These existing traffic barriers had been adopted by the Malaysian Public Works Department
(PWD) of Malaysia based on the test results of performance evaluation for road safety
features carried out by the National Cooperative Highway Research Program (NCHRP)
Report 350, published in 1993.
The report provides guidelines on the recommended testing procedure for the performance
evaluation of various highway safety features. There are six Test Levels (TL) recommended
in the guidelines and the summary of descriptions (including the impact severity) are shown
in the table below.
26
Page
Review Report
Table 4-1 Test Level for Longitudinal Traffic Barrier

(Source: Main Roads Western Australia, 2006)
Height of
Test Vehicle Test Impact Impact
Test Centre of
Speed Angle Severity
Level Gravity
Mass (kg) Type (km/h) (degrees) (kJ)
(mm)
820 Small car 50 20 550 9.3
1
2000 *4WD/utility truck 50 25 700 34.5
820 Small car 70 20 550 18.1
2
2000 *4WD/utility truck 70 25 700 67.5
820 Small car 100 20 550 37.0
3
2000 *4WD/utility truck 100 25 700 137.8
820 Small car 100 20 550 37.0
4
2000 Single-unit van truck 80 15 1250 132.3
820 Small car 100 20 550 37.0
5
36000 Van type semi-trailer 80 15 1850 595.4
820 Small car 100 20 550 37.0
6 Tanker type semi-
36000 80 15 2050 595.4
trailer
*Four wheel drive
As shown in Table 4-1 above, the Impact Severity (IS) is used as a basis to compare the
test levels and is calculated based on the principle of kinetic energy (Main Roads Western
Australia, 2006). The formula is given by,
[ ( )] (Equation 9)
where impact severity (kJ), mass of vehicle (kg), vehicle velocity (m/s) and
impact angle (degrees).
Other than the calculated impact severity, assessment of Occupant Impact Velocity (OIV)
and Occupant Ridedown Acceleration (ORA) are also required as the main requirements for
27
the test to assess the occupants injury risk (Groe et al., 2004).
Page
Review Report
In order to measure the expected severity outcome of an impact with an object, the Severity
Index (SI) is used. The index is described by a value between 0 to 10, with the lowest value
(0) representing an expected crash involving no property damage or injury, while the highest
value (10) represents an expected crash with a 100% probability of a fatality (Main Roads
Western Australia, 2006). Table 4-2 shows the expected severity indices for safety barriers
for all vehicles at varying speeds.
Table 4-2 Typical Severity Indices for Safety Barriers and Various Design Speeds
Design Speeds (km/h)

Safety Barriers
70 80 90 100
4 wire rope (flexible) 1.5 2.0 2.5
W-beam (*G4) 2.0 2.5 3.0
Thrie-beam 2.0 2.5 3.0
Type F (concrete) 2.0 2.5 3.5
*Blocked-out Strong Post
4.1 Flexible Barrier Systems
4.1.1 General Description and Behaviour Under Impact (Sources: Australian Road
Research Board (ARRB) Transport Research, n.d.; Main Roads Western Australia,
2006; Road Engineering Association of Malaysia (REAM), 2006)
Wire Rope Safety Barriers (WRSB) or also known as Wire Rope Safety Fences, are
classified as flexible barrier systems which create large deflections when impacted by
vehicles. Because of this unique characteristic, the flexible wire rope barriers cause
the least damage and the smallest injury risk to the vehicles and occupants
respectively, as compared to other existing barrier types. There are two types of
flexible barrier systems, namely vertical array and twisted array wire rope safety
fences. The latter is the commonly used type of flexible barrier in Malaysia.
28
Generally, the barrier comprises three or four tensioned galvanized steel wire ropes
suspended by frangible posts at varying heights between 690mm to 710mm and at an
Page
interval of 2.4 or 3.2 meter each along the barrier. The upper ropes which consist of
Review Report
one or two wire ropes are located in a slot on top of the post at a height of 600mm
10mm from the ground. The other two lower ropes are normally located on rollers or
wire rope, hung at a height of 500mm 10mm from the ground, and interwoven
between each pair of line and deflection posts. Figure 4-2 shows the general
arrangement of the barrier system installation.
Twisted Array Vertical Array
Figure 4-1 Cross section for two typical designs of wire rope barrier system
(Source: ARRB Transport Research, n.d.)
Currently, there are two types of wire rope safety barrier systems used in Malaysia: 1.
Wire rope barrier system with double curve shaped posts, and 2. Wire rope barrier
system with circular posts. The following plates (4-1, 4-2 and 4-3) show the typical
installation of this type of barrier on Malaysian roads.
29
Page
Review Report
Plate 4-1 Flexible wire rope barrier at Kuala Lumpur-Karak Expressway (E8) in Gombak
Plate 4-2 Flexible wire rope barrier on flyover at Middle Ring Road 2 (MRR2) in Gombak
30
Page
Review Report
Plate 4-3 Flexible wire rope barrier on median at Second Link Expressway (E3) near
Perling toll plaza, Johor
4.1.2 Crash Test and Result
Prior to installation, the barrier system should be crash tested according to the United
States (US) NCHRP Report 350 (1993) testing conditions by the Federal Highway
Administration (FHWA) or any procedures recognized internationally. Descriptions of
requirement for the crash test and result (REAM, 2006) are as below.
US Test Level: Complies with TL-3

Deflection: Varies with type of barrier (Typically 1.7 to 3.4m for a
2000kg vehicle at 100km/h impact speed and 25
approach angle)
Offset from obstruction approximately 1.7m and 1.0m
for 2.4m and 1.0m post spacing, respectively
Passenger Injury Risk: OIV of maximum speed of 12m/s (43.2km/h)
ORA of 20g (maximum)
31
Page
Review Report
The following table shows some of the results of the test carried out on a 4 wire rope
safety barrier system.
Table 4-3 Results for Deflections of Crash Tested Wire Rope Safety Barriers
(Source: REAM, 2006)
Anchor Post Impact Impact Measured

Vehicle Speed
Spacing Spacing Angle Severity Deflection
Mass (kg) (km/h)
(m) (m) (deg) (kJ) (m)
1 319.2 1.0 753 113 20 43.4 0.86
2 100.8 3.2 875 104 19 38.3 1.00
3 103.2 2.4 1260 83 30 83.7 1.20
4 319.2 1.0 1480 115 20 88.3 1.00
5 100.8 3.2 1492 111 19 75.6 1.40
6 626.4 2.4 1500 113 20 86.4 1.70
7 106.1 3.2 1505 115 20 90.0 1.73
8 192.0 2.4 2010 102 25 144.1 1.65
From the results shown in the Table 4-3 above, it can be generally concluded that an
increase in the post spacing increases the deflection length of the barrier system.
4.1.3 Advantages and Disadvantages (Sources: Main Roads Western Australia, 2006;
Monash University Accident Research Centre (MUARC), 2006; REAM, 2006;
Federation of European Motorcyclists Associations (FEMA), 2000)
Although the flexible barrier system has relatively large deflections, greater clearance
should be provided for installation within medians. It should not be installed in median
of width less than 2.5m. In addition, the system should not be used on a vertical sag
curve of radius less than 3000m as well as on a horizontal curve of less than 200m. In
32
addition, flexible wire rope safety barrier should not be installed and connected directly
Page
to other barriers of bridge parapets since its deflection cannot be safely guaranteed if
Review Report
vehicles collided in the transition area between the flexible barrier and other barrier
systems.
If the wire rope system is properly installed according to guidelines, it is capable of

redirecting the errant vehicles smoothly when impacted by vehicles. Snagging effect is
minimized during impact since its unique design enables the cables to strip from the
collapsible posts.
Last but not least, the maintenance costs associated with repairing the damaged
barrier are minimal; hence the rope barrier system provides significant cost advantage.
4.2 Semi-Rigid Barrier Systems
4.2.1 General Description and Behaviour Under Impact (Sources: Main Roads Western
Australia, 2006; REAM, 2006)
There are three types of semi-rigid safety barriers commonly used on Malaysian roads
which are the W-beam, Thrie-beam, and Modified Thrie-beam guardrails. The barrier
system deforms significantly but not excessively (greater deflection properties
compared to the rigid system, but less than the flexible system) when impacted by
vehicles and has a moderate deflection of a maximum of 1.2m. It can be categorized
into two groups. The first one, which is the strong beam with weak post, is purposely
designed to break away so that the impact force is distributed by the beam action to a
relatively large number of posts. In contrast, the strong beam with strong post is
purposely designed to only deflect moderately and the impact force is distributed by
beam action to a smaller number of posts.
The system mainly includes a steel beam attached to block out units supported on
posts. The block out units and posts are normally constructed of steel. Although there
are posts constructed of wood and concrete, they are not favoured because of poor
33
impact performance. The profiles for each type of semi-rigid barriers typically installed
Page
on road shoulder are shown below in Figure 4-2.

Review Report
Figure 4-2 Profiles of semi-rigid safety barriers used on road shoulder and median
The following plates illustrate the various types of semi-rigid barriers installed in
Malaysia.
Plate 4-4 Single-mounted W-beam guardrail median at North-South Expressway (E2) in

Serdang
34
Page
Review Report
Plate 4-5 Double-mounted W-beam guardrail median at Kepong road
Plate 4-6 Customized stacked mounted W-beam guardrail at Karak Expressway (E8)
4.2.2 Crash Tests and Results
Similar to flexible barrier system, NCHRP Report 350 is adopted by the Malaysian
35
government for the guidelines of standard crash test procedure to evaluate the safety
performance for semi rigid barrier system to be used in Malaysia. Details of the semi-
Page
rigid barrier installation and test results are as shown in the following table.
Review Report
Table 4-4 General Summary for Semi-Rigid Barriers Installation

(Source: REAM, 2006)
ON MEDIAN ON SHOULDER
MB4S
MB4S
Blocked-
MB4S Blocked-Out
Out W- Blocked- Modified
Blocked-Out Blocked-Out Modified Thrie W-Beam
Beam Out Thrie Thrie
W-Beam (Steel Thrie Beam beam (Steel
(Steel Beam Beam
Post) Beam) with
Beam) with
2.0 Spacing
4.0 Spacing
TEST RESULTS:
Test Level 3 3 4 3 3 3 4
Maximum
0.6m 0.9m 0.5m 1.0m 1.2m 0.6m 0.9m
Deflection
Passenger Injury
No detailed information
Risk
DESCRIPTIONS:
Double W- Double Thrie-
Beam 2 Thrie-beams Single W-beam Single Thrie-beam
beams beams
Post Spacing 2.0m 4.0m 2.0m
C Section
C Section C Section C Section
Post 150 x 76 x 6
150 x 110 x 6 mm 150 x 76 x 6 mm 150 x 110 x 6 mm
mm
2 C Sections 2 C Sections 2 C Sections
1 C Section 1 C Section
Off Set Brackets 150 x 76 x 6 150 x 110 x 6 350 x 110 x 6
150 x 76 x 6 mm 150 x 110 x 6 mm
mm mm mm
Mountings 16mm dia. Steel Bolts
Footing None (Except at points of transition)
Median Width 2.5m min <2.5m <2.5m 2.5m min 2.5m min <2.5m <2.5m
4.2.3 Advantage and Disadvantage (Sources: Main Roads Western Australia, 2006;
REAM, 2006)
The advantage of this system is that it has low initial installation cost. Nevertheless, its
maintenance cost is relatively high and care should be taken to install this system on
locations where the frequency of vehicles running off the road is high.
36
Page
Review Report
4.3 Rigid Barrier Systems
4.3.1 General Description and Behaviour Under Impact
A concrete barrier is classified as rigid safety barrier that does not deflect upon impact.
Generally, it is designed to contain and redirect errant vehicles upon impact. Even
though the barrier system is able to redirect the colliding vehicle stably without any
rolling movement, the severity of impact experienced by the vehicle is higher
compared to semi-rigid or flexible barrier. The rigid barrier can be classified into two
categories which are single slope (e.g. Texas Constant Slope Barrier and Californian
Single Slope Barrier) and multi slopes barrier (e.g. New Jersey Barrier and F-Type
Concrete Barrier). The figure below illustrates the profiles of some of the rigid concrete
barriers currently available.
Vertical Wall
New Jersey Profile F-Type Profile Single Slope Profile
Profile
*Note: All dimensions are in millimetres
Figure 4-3 Profiles of rigid barriers

(Sources: ARRB Transport Research, n.d.; Main Roads Western Australia, 2006)
As shown in the subsequent plates are the typical rigid barrier systems installation in
Malaysia.
37
Page
Review Report
Plate 4-7 Concrete barrier at MRR2 in Selayang
Plate 4-8 Concrete barrier with antiglare screen at Karak Expressway (E8)
4.3.2 Crash Tests and Results
The concrete barriers are rigid and normally, they do not result into permanent
38
deflection (if properly installed according to specification) when impacted by vehicles. If

Page
the height of the barrier is reduced to less than 725mm by pavement overlays,
Review Report
impacting vehicles may roll over the barrier (REAM, 2006). Crash tests for concrete
barriers need to comply with the NCHRP requirements or other internationally
recognized testing procedures (i.e. European Norm EN 1317). The followings describe
results for some of the crash tests carried out on various rigid concrete barriers.
a) New Jersey Profile Barrier (1270mm Portable Concrete Barrier (PCB))

Test performed by Office of Research & Development, Ohio Department of
Transportation (Report date: June 2006)
Barrier Data: Height of 457.2mm more than the standard 32-inch New
Jersey PCB
Slope of 3 degrees steeper than typical New Jersey shape
(from first break point to top of barrier)
Joint connection each section is connected with a single pin
passing through 3 set of loops at each segment end
Test Date: No detailed information
Test 2000kg pickup truck at 100km/h with impact angle of 25
Description: (NCHRP Report 350 Test 3-11)
Deflection: Nil
Passenger OIV of 4.5m/s (longitudinal) and 6.1m/s (lateral)
Injury Risk: ORA of -5.4m/s (longitudinal) and -8.6m/s (lateral)
b) Standard F-Type Precast Concrete Barrier

Test performed by Oregon Department of Transportation (Report date:
December 2001)
Barrier Data: 810mm in height; width of 610mm (base) and 240mm (top)
Barrier section length of 3810mm (each section is held
together with an assembly of pin and steel bar loops)
Each section is anchored with 2 galvanized pins (25mm
39
diameter and 750mm long), 50mm deep into the asphalt

Page
(Refer report for more detailed specification)

Review Report

Test Date: 17 April 2001
Test 2000kg pickup truck at 100km/h with impact angle of 25
Description: (NCHRP Report 350 Test 3-11)
Deflection: 760mm max. barrier deflection (No NCHRP requirement,
Oregon Department of Transportations requirement
914mm)
Passenger OIV of 0.0m/s (longitudinal) and -5.85m/s (lateral)
Injury Risk: ORA of -18.23m/s (longitudinal) and -12.52m/s (lateral)
c) Single Slope Barrier (Texas Constant Slope Type 60G, 9.1 slope)
Test performed by FHWA (Acceptance letter: February 1998)
Barrier Data: Slip-formed and reinforced

1420mm in height; width of 610mm (base) and 150mm (top)
Footings 3050mm (length) x 250mm (depth) footing on both
ends with additional reinforcing steel, anchored in asphalt
while the middle section stands freely on the asphalt
Test Date: 28 November 1995
Test 2000kg pickup truck at 97.7km/h speed with impact angle of
Description: 25.5
Deflection: No information
Passenger OIV of 6.80m/s (longitudinal) and -9.51m/s (lateral)
Injury Risk: ORA of -6.7g (longitudinal) and 2.3g (lateral)
d) Concrete STEP Barriers (from the Netherlands & Germany) (Source: Groe
et al., 2004)
The system is a combination of the New Jersey shape and single slope
barrier. Generally, the base of the barrier is wider than the single slope
barrier; although the height is more (with narrow width) than the New Jersey
40
barrier. The general profile of the STEP barrier is shown in Figure 4-4 below.
Page
Review Report
(Note: 9 gon = 8.1 degree)

Figure 4-4 General profile of STEP barrier
(Source: Groe et al., 2004)
The crash tests were performed according to the European Norm EN 1317 respectively
prEN 1317 and the summary of the results are shown in the following table.
Table 4-5 Crash Test Data Summary
(Note: Theoretical Head Impact Velocity, THIV OIV; Post-impact Head Deceleration, PHD ORA; Acceleration
Severity Index, ASI Max. 50ms Average Acceleration, AA)
(Source: Groe et al., 2004) 41
Page
Review Report
4.3.3 Advantages and Disadvantages (Sources: Main Roads Western Australia, 2006;
MUARC, 2006; REAM, 2006; Groe et al., 2004)
One of the primary advantages of the rigid barrier system is its ability to prevent
vehicles (especially small cars) from the risk of rollover, hence avoiding big damages
due to rollover. The F-shape, the New Jersey profile, the single slope and the STEP
barrier (as only the vehicle tires will touch the barrier is slight collision occurs) are
among the rigid barriers which have this unique characteristic. In comparison
excluding the single slope and the STEP barrier, the F-shape performs better than the
New Jersey profile barrier in reducing the tendency for vehicles to roll. However, not all
rigid barriers perform better for all type of vehicles in reducing the risk of vehicle
rollover. For this case, the vertical concrete wall only performs better for heavy
vehicles except for small cars and pickup trucks with mass of less than 2000kg or
motorcycles. The wall does not have the energy management feature of vehicle lifting.
Another advantage of the rigid barrier system, especially for the vertical concrete wall
and the single slope barrier (California and Texas profile), is that resurfacing would be
possible several times (255mm overlay until height is reduced to 815mm) without
affecting its performance.
On the contrary, the disadvantage of the rigid barrier system is that it will result in
severe collision if impacted by a vehicle with an impact angle of greater or equal to 20
degrees. Hence, it will cause severe injuries to vehicle occupants and extensive
damage to impacting vehicles. Compared to other rigid barriers, the single slope will
result in greater vehicle damages if impacted at shallow impact angles. In addition,
some rigid barriers have poor safety values and permanent delfection of more than
zero (0). Both California and Texas profile slope barriers have poor safety values, even
though the California (9.1 degree slope) has better result than the Texas (10.8 degree
slope). The precast concrete barrier (not suitable to be installed on median) and both
pre-fabricated H2 and slip-form H4b STEP barriers are the type of rigid barriers which
have permanent deflection as mentioned earlier.
42
Page
Review Report
5. Analysis of High-Deck Bus vs. Traffic Barrier Collision
5.1 Impact Severity Analysis
A simple calculation analysis has been carried out using the formula in Equation 9 to
compute the Impact Severity (IS) of high-deck bus collision with safety barrier, with
different impact angles and velocities. An approximate maximum load of 24350kg for
standard double-axles high-deck bus similar to the bus in the Behrang case as
mentioned earlier was used for this analysis. The results are shown in Table 5-1.
Table 5-1 Results of Impact Severity Calculation Analysis

Impact Angles (degree)
Energy (kJ)
5 10 15 20 25 30 35 40 45
10 0.7 2.8 6.3 11.0 16.8 23.5 30.9 38.8 47.0
20 2.9 11.3 25.2 44.0 67.1 94.0 123.7 155.3 187.9
30 6.4 25.5 56.7 98.9 151.0 211.4 278.2 349.4 422.8
40 11.4 45.3 100.7 175.9 268.5 375.9 494.6 621.2 751.7
50 17.8 70.8 157.4 274.8 419.6 587.3 772.8 970.6 1,174.5
Velocity 60 25.7 102.0 226.6 395.7 604.2 845.7 1,112.9 1,397.6 1,691.3
(km/h) 70 35.0 138.8 308.4 538.6 822.4 1,151.1 1,514.7 1,902.3 2,302.1
80 45.7 181.3 402.9 703.5 1,074.1 1,503.4 1,978.4 2,484.7 3,006.8
90 57.8 229.5 509.9 890.3 1,359.4 1,902.8 2,504.0 3,144.7 3,805.5
100 71.4 283.3 629.5 1,099.2 1,678.3 2,349.1 3,091.3 3,882.3 4,698.1
110 86.4 342.8 761.6 1,330.0 2,030.7 2,842.4 3,740.5 4,697.6 5,684.7
120 102.8 408.0 906.4 1,582.8 2,416.7 3,382.7 4,451.5 5,590.6 6,765.3
43
Page
Figure 5-1 Impact severity plot

Review Report
Based on the results above, it can be concluded that as the values of the impact angle
and velocity increases, IS increases. The IS, which is based on the principle of kinetic
energy, increases exponentially and is directly proportional to the square of vehicles
speed. Furthermore, the significant change is more eminent for higher impact angle as
illustrated in the Figure 5-1.
Another analysis was carried out to compare the results of IS from the above calculation
with the test results of the NCHRP 350. The values of IS from the NCHRP 350 were then
used to find the range of calculated IS for high-deck bus collision with safety barriers.
From there, the range of speed was determined as shown in Table 5-2.
Table 5-2 Impact Severity Comparison

NCHRP 350 Test Impact Severity Analysis for High-deck Bus Collision
Requirements
Test Level IS (kJ) Impact Angle Calculated IS (kJ) Range Speed (km/h)
3 137.8 25 67.1 151.0 20 30
4 132.3 15 100.7 157.4 40 50
5 595.4 15 509.9 629.5 90 100
6 595.4 15 509.9 629.5 90 100
From the comparison, it can be assumed that if a high-deck bus impacted the TL-3
barrier (presumed to be equivalent to W-beam guardrail) at an angle of 25, the range of
speed between 20 to 30km/h was needed to achieve the IS value of 137.8kJ obtained
from the NCHRP 350 test. The range of speed between 90 to 100km/h is required by a
high-deck bus to achieve the IS value of 595.4kJ if it was to collide with the TL-5/6
barrier (which is equivalent to concrete barrier) at a 15 angle. Hence, from the
comparison analysis:
The possibility of a high-deck bus to penetrate the TL-3 barrier is high for
travelling speed of more than 30km/h at a given angle of 25.
Similarly, the risk of a high-deck bus to penetrate the TL-5/6 barrier is high for
travelling speed of more than 100km/h at a given angle of 15.
However, it is to be noted that these assumptions were based solely on the calculation
analysis of impact severity from the formula in Equation 5 that does not consider the
44
dimensions of the test vehicle, as well as differences in the height of the CG and bumper.
Page
Review Report
5.2 Rollover Analysis Bus-Traffic Barrier Collision
A rollover analysis was performed to study the effect of increasing the CG of a high-deck
bus in relation to the height of the crash barrier. Different worst-case high-deck bus
rollover scenarios at straight and curve roads were explored for this section.
5.2.1 Rollover on Straight Road
According to Matolcsy (2007), the turning moment (M), as described in Figure 5-2, is a
main factor contributing to the initiation of rollover accidents on straight roads.
Figure 5-2 Free Body Diagram (FBD) of bus turning moment
In order for a rollover to occur, the conditions explained hereafter must be satisfied.
The first condition would be the rotation of the vehicle around the axis with the outside
wheel as the pivot point. This situation is somewhat similar to one of the worst-case
scenario when a bus collided with a failed guardrail causing the bus to rollover. The
bus will tip on its side given that the lateral sliding moment is larger than the turning
moment:
(Equation 10)
and the kinetic energy of the bus is greater than the potential energy resulted from CG
displacement in height to make it unstable:
45
(Equation 11)
Page
Review Report
In the figure and equations,
m is the total mass of the bus,

is drag factor of the road,
is the gravitational constant,
is the extended track width, and
is the height of CG from pivot point.
Based on the analysis, high-deck buses in Malaysia would most likely to rollover if they
were travelling at 90km/h with an impact angle of 15. The speed of 90 and 120km/h
were considered in the analysis since 90km/h is the speed limit for buses on
expressways while the latter is the maximum speed on the bus tachometer. Impact
angles of 15 and 25 considered were based on the minimum and maximum impact
angles recommended in the NCHRP 350 test. In addition, the range of CG height
between 1.05m to 1.40m is used to represent Malaysian buses.
For another worst-case scenario when a bus collided with a concrete barrier, similar
conditions as described by Matolcsy (2007) still apply. However, this time, the pivot
point is shifted from the outside wheel to top of the barrier (900mm high STEP barrier)
as illustrated in Figure 5-3.
Figure 5-3 FBD of bus turning moment for collision with rigid barrier
The analysis shows that high-deck buses will most probably survive a collision with a
rigid barrier without rolling over if the vehicle is travelling at 90km/h at a maximum
46
impact angle of 8. Similarly, at 120km/h the bus will most likely to escape rollover even
Page
at an impact angle of 6 during collision. Additionally, the critical CG height that will not
Review Report
result in rollover was found to be 1.56m. From this analysis, it is proven that a rigid
barrier works better than a semi-rigid barrier in preventing high-deck bus rollover.
5.2.2 Rollover at Curve Road
Table 5-3 was taken from A Guide on Geometric Design of Roads by Malaysian PWD.
It lists the standard minimum radius of a curve to be used for the designated speed
and maximum super elevation rates in Malaysian urban roads. In this table, the
minimum radius of a curve is calculated based on the predetermined designated
speed at the road stretch.
Table 5-3 Minimum Radius of a Curve
Design Speed Minimum Radius (m)

(km/h) e=0.06 e= 0.10
120 710 570
100 465 375
80 280 230
60 150 125
50 100 85
40 60 50
30 35 30
20 15 15
Apart from that, the critical speed of a curve can also be determined based on how
sharp the curve is, how much bank is in the roadway, and the coefficient of friction of
the road surface. A vehicle cornering at a speed exceeding this critical speed will begin
to spin around its centre of mass and leave the roadway. Subsequently, the vehicle will
lose control, which can further lead to a rollover. This can happen to anyone regardless
of a persons driving skill level or years of experience.
A vehicle with higher CG such as a high-deck bus is more vulnerable to rollover, even if
at a speed well below the critical speed of a curve. As a high-deck bus is manoeuvring
a curve, the weight of the vehicle together with its occupant will shift to the front outside
47
tire due to centrifugal force. Rollover will result if the bus is travelling at high speed
Page
because the CG of the bus shifts as well.

Review Report
5.2.3 Design of Vehicle
Current design standards for Malaysia buses require modern engineering techniques
to ensure the integrity of the construction to pursue the new regulation. At the same
time, the Malaysia bus industry stepped forward in bus construction with new the
design of high-deck buses. Most of the designs follow the Australian Design Rules
(ADR). In basic principles, the design of the vehicle shall differ especially in terms of
dimension, weight and interior configuration. Additional factors can also influence the
dynamics of the vehicle.
In order to ensure the safety of high-deck buses during travelling, special safety aspect
should be considered such as a counterweight. The purpose of the counterweight is to
stabilize the vehicle when travelling. There are several methods used as
counterweights in bus constructions such as ballast tanks and steel weights. The
ballast tank is commonly placed at the rear side of the vehicle and it is filled up with
water or other liquids. Adding the ballast tank helps the vehicle to lower its CG. It also
prevents the increment of centrifugal force associated with curvy roads. One of the
setbacks when using the ballast tank is when the water is lesser than original settings.
This can cause increment of the CG and high tendency for rollover to occur. The steel
weight is the method that is usually used when constructing the bus frame. Four
pieces of steel, each weighing 1000kg (total 4000kg) is welded on the structure to
lower the CG. Bus manufacturers should also consider in improving the vehicle
structural integrity. The structure must be rollover-proof to prevent injuries and fatalities
of passengers during accidents especially in an event of a rollover. Hence, combining
improvements in rollover strength, seat and seat anchorage strength should be
implemented in all types of buses in Malaysia.
48
Page
Review Report
6. Conclusion
This preliminary literature study was initiated with two main objectives. The first objective of
the study was to carry out analyses of rollover for high-deck bus collisions with crash
barriers through calculations based on formulas derived from a few related literatures. From
the analyses, the location of CG and speed highly affect the stability of a high-deck bus
during rollover event, especially at curved road. The second objective was to assess
existing literatures for any type of traffic safety barriers currently available worldwide for
recommendation on adopting specific crash barrier for high-deck buses to be used in
Malaysia s highways or roads. Based on the study, it is concluded that rigid barrier performs
better than semi-rigid barrier in preventing rollover of high-deck bus during collision with
traffic barriers.
Finally, since the data obtained from the literature research is not extensive enough to carry
out a thorough study regarding the cost-benefit of barrier selection and installation, data
related to statistics of road accidents involving high-deck bus (from RMP) is required with
the RSDs assistance.
7. Recommendations
7.1 Bus Construction
It is time for related regulations to be implemented in Malaysia towards improving local

bus construction. Based on a study carried out on the rollover of heavy commercial
vehicles by the University of Michigan Transport Research Institute (2000), and other
literature reviews, MIROS recommendations for high-deck bus or coach are as follows:
a) Vehicle construction and testing
Construction of high-deck buses must consider all the safety requirements and
must comply with the standard regulation. Stability of the vehicle is a crucial part
49
in the vehicle construction that can influence a vehicle tendency to rollover. Every
Page
aspect shall be considered such as:

Review Report
(i) Location of CG
The higher the CG, the more unstable the vehicle will be. The CG
calculation should be compulsory and properly done by a certified
engineer.
(ii) Selection of bus chassis
The track width, suspension and tire can influence the vehicle stability in
manoeuvring and cornering. Also, selection of the chassis is crucial to
determine the maximum mass (allowable mass) that a bus can have.
(iii) Dimension and design of the bus frame/body
The dimension of the vehicle shall comply with regulation. Location of the
passenger seat for the high-deck bus must be done properly to ensure
equal weight distribution. Furthermore, the vehicle should be designed
accordingly to suit the road geometric design. The maximum allowable
dimension of vehicle under the current Malaysian legislation is 2.5 meter
wide, 4.57 meter high and 12 meter long.
b) Speed management
High speed can affect the stability and controllability of the high-deck buses
during manoeuvring, cornering and braking. Speeding at corners can impart high
centrifugal force to a vehicle especially to heavier and higher vehicles. This can
cause the driver to lose control of the vehicle and increase the tendency to
rollover. Driving training is a significant method to change the drivers' and fleet
management's attitude, as it is also aimed at improving the efficiency of their
operations and developing driving skills to provide safer services for public
transport customers.
c) Enforcement
MIROS recommends that bus developers should be audited by a government
body to ensure all the regulations are being followed. The inspection should
cover all; from the initial documentation till the finish product. In order to make
sure that the vehicle is safe to be used, it must be tested and certified by a
50
government body. The test procedure is explained in the Malaysia Road

Page
Transport Rules.
Review Report
7.2 Traffic Safety Barrier Consideration
In order to select the use of suitable safety barriers, consideration should be given to the
cost of maintenance as compared to the cost resulted from an accident. The collision
maintenance should play an important role in the selection of barrier systems since the
majority of maintenance costs are normally due to collision repairs. Therefore, it is not
cost-effective to install safety barrier along the road. Once the installation of barriers is
justified, specific barrier type must be selected considering the barriers performance
capability, deflection characteristics including occupant's risk, site conditions,
compatibility, life cycle costs, maintenance, aesthetic and environmental considerations
as well as field experience, as recommended in the REAMs guidelines. Besides, the
selected barriers must structurally be able to contain and redirect the vehicle (bus) as
well as preventing it from undergoing rollover during collision.
Several studies carried out by international road safety institutions clearly revealed that
traffic safety barriers should only be installed to reduce accident severity at known spots
which have a history of related run-off-road accidents. Hence, based on the findings
through literature reviews and analysis, it is recommended for rigid barriers to be
installed in areas which are known to have a history of bus-rollover accidents. In this
case, concrete barriers such as STEP and vertical wall which can withstand crash
energy of TL-4 and up to TL-6 should be considered for installation especially on
hazardous curves where high-risk of bus-rollover crashes are identified. The selections
are based on the unique energy management characteristics of both barriers (STEP and
concrete vertical wall) in preventing vehicle rollover and lifting.
51
Page
Review Report
References
Abdul Rahmat, A.M., S.V., Wong (2008). Crashes Related to Overloading in Commercial Vehicles.
Malaysian Institute of Road Safety Research. Report Submitted to the Minister of Transport Malaysia.
ARRB Transport Research (n.d.). Road Hazard Management Guide. Department of Infrastructure, Energy
and Resources, Tasmanian Government, Australia.
E. Larrod, A. Miravete & F.J. Fernndez (1995). A new concept of a bus structure made of composite
materials made by using continuous transversal transversal frames. Elsevier Science Limited.
Composite Structures, Vol. 32, pp. 345-356.
ECE Regulation No. 66, Agreement (22 Feb 2006). United Nations. E/ECE/TRANS/505, Rev. 1/Add.
65/Rev. 1.
ECE Regulation No. 107, Agreement (8 Oct 2004). United Nations. E/ECE/TRANS/505, Rev.
2/Add.106/Rev. 1.
F. Lan, J. Chen & J. Lin (2004). Comparative analysis for bus side structures and lightweight optimization.
Proceeding. Institution of Mechanical Engineers. Vol. 218, Part 5: Journal of Automobile Engineering.
FEMA (2000). Final report of the Motorcyclists and Crash Barriers Project. Federation of European
Motorcyclists Association, Brussels.
Government of Western Australia (Mar 2006). Main Roads Western Australia Assessment of Roadside
Hazards. Road and Traffic Engineering, Technology and Environment Directorate. File No. 05/9104,
Doc. D06#26105.
Groe, U., Valtonen, J. (Oct 2004). Suitable Concrete Barriers for Finland. Laboratory of Highway
Engineering, Helsinki University of Technology.
Jewell, J., Rowhani, P., Stoughton, R., Crozier, W. (Dec 1997). Vehicular Crash Tests of a Slip-formed,
Single Slope, Concrete Median Barrier. Material Testing and Engineering Services, California
Department of Transportation. Report No. FHWA/CA/ESC-98/02.
Kennedy, J.C. Jr., Plaxico, C.A., Miele, C.R. (Jun 2006). Development of NCHRP Report 350 TL-3 New
Jersey Shape 50-inch Portable Concrete Barrier. Executive Summary Report. Office of Research &
Development, Ohio Department of Transportation.
Larsson, M., Candappa, N.L., Corben, B.F. (Dec 2006). Flexible Barrier Systems along High-Speed
Roads: A Lifesaving Opportunity. Monash University Accident Research Centre.
MacDonald, D.J., Kirk, A.R. (Dec 2001). Precast Concrete Barrier Crash Testing Final Report. Oregon
Department of Transportation, Oregon. Report No. FHWA-OR-RD-01-07.
52
Page
Review Report
Matolcsy, M. (May 2003). Lesson and Conclusions Learned from the Analysis of Bus Rollover
th
Accidents. Informal Document # 4, 84 GRSG 5-9, International Conference of Science and Motor
Vehicles.
Matolcsy, M. (2003). Severity of Bus Rollover Accidents. Paper Number 07-0989, Scientific Society of
Mechanical Engineers, Hungary.
Pankaj S. Deshmukh (Dec 2006). Rollover and Roof Crush Analysis of Low-Floor Mass Transit Bus.
Master Thesis, Wichita State University.
Road Engineering Association of Malaysia (REAM) (Sep 2006). Guidelines on Design and Selection of
Longitudinal Traffic Safety Barrier.
Ross, H.E. Jr., D.L. Sicking, R.A. Zimmer and J.D. Michie (1993). National Cooperative Highway
Research Program Report 350: Recommended Procedure for the Safety Performance Evaluation of
Highway Features. Transportation Research Board. National Academy Press, Washington, D.C.
Viner, J.G., Tamanini, F.J. (1973). Effective Highway Barriers. Elsevier. Accident Analysis & Prevention,
Vol. 5, pp. 203-214.
Winkler, C.B. (2000). Rollover of Heavy Commercial Vehicles. University of Michigan Transportation
Research Institute (UMTRI). Research Review, Vol. 31, No. 4.
53
Page

Bus Rollover and Impact Review

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Bus Rollover and Impact Review

Uploaded by

Copyright:

Available Formats

Review Report

Stability of High-Deck Bus in Rollover and

(Source of photos on cover: RMP and Sin Chew Daily)

43 5.1 Impact Severity Analysis

45 5.2 Rollover Analysis Bus-Traffic Barrier Collision

45 5.2.1 Rollover on Straight Road

13 Figure 2-1 Conventional frame for single-deck bus/coach

25 Table 4-1 Test Level for Longitudinal Traffic Barrier

List of Acronyms and Abbreviations Used

4WD Four-Wheel Drive

REAM Road Engineering Association of Malaysia

RMP Royal Malaysian Police

Malaysian Institute of Road Safety Research (MIROS)

Prof. Dr. Wong Shaw Voon Director General

Aqbal Hafeez Ariffin

selected crash tests results were used for comparison purpose.

2. Current State of Local Bus Construction

2.1 Bus Construction and Regulations

2.1.1 Adoption of UNECE Regulations in Malaysia

Conventional bus/coach structure manufacturing involves labour-intensive arc welding

Figure 2-1 Conventional frame for single-deck bus/coach

Figure 2-2 Example of single-deck superstructure frame

2.1.2 Overview of UNECE Regulation 66

2.1.3 Overview of UNECE R107 Regulation

The following is an overview of R107 Uniform Provisions Concerning the

Approval of Double-deck Large Passenger Vehicles with Regards to Their

General Construction Annex 9.

device shall be easily operable in an emergency.

Handrails or/and handhold for inter communications staircases shall be

3.1 Stability Issue Determination of Centre of Gravity (CG)

3.1.1 Definition of Rollover

3.1.2 Factors of Rollover

Figure 3-1 Relationship between CG and rollover (Illustrated with gravitational

Figure 3-2 Diagram of bus before pivoting

Figure 3-3 Critical point of rollover

The value of or widely known as Static Stability Factor (SSF) is adopted by

movement of the vehicle. In a real-life rollover crash, vehicle approaches pivot

3.1.3 Centre of Gravity Calculation

A rollover accident is directly related to the CG of the vehicle. To determine the

The transverse position ( ) of the vehicles CG needs to be determined first in order to

Figure 3-5 Longitudinal position of CG

Figure 3-6 Transverse position of CG

Figure 3-7 Tilting test for determining the height of CG

(Source: UNECE R66, 2006)

In determining the vertical height ( ), more accurate calculation shall be obtained if

4. Longitudinal Traffic Barriers

Table 4-1 Test Level for Longitudinal Traffic Barrier

Design Speeds (km/h)

4.1 Flexible Barrier Systems

Twisted Array Vertical Array

4.1.2 Crash Test and Result

US Test Level: Complies with TL-3

Anchor Post Impact Impact Measured

If the wire rope system is properly installed according to guidelines, it is capable of

4.2 Semi-Rigid Barrier Systems

on road shoulder are shown below in Figure 4-2.

Plate 4-4 Single-mounted W-beam guardrail median at North-South Expressway (E2) in

Plate 4-5 Double-mounted W-beam guardrail median at Kepong road

4.2.2 Crash Tests and Results

Table 4-4 General Summary for Semi-Rigid Barriers Installation

4.3 Rigid Barrier Systems

4.3.1 General Description and Behaviour Under Impact

Figure 4-3 Profiles of rigid barriers