Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018

SAE TECHNICAL
PAPER SERIES 1999-01-0431

The Contribution of Occupant and Vehicle

Dynamics Simulation to Testing Occupant Safety
in Passenger Cars During Rollover
R. Bardini and M. Hiller
University of Duisburg Germany

Reprinted From: Safety Test Methodology

(SP-1434)

International Congress and Exposition

Detroit, Michigan
March 1-4, 1999

400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760
 Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018

The appearance of this ISSN code at the bottom of this page indicates SAE’s consent that copies of the
paper may be made for personal or internal use of specific clients. This consent is given on the condition,
however, that the copier pay a $7.00 per article copy fee through the Copyright Clearance Center, Inc.
Operations Center, 222 Rosewood Drive, Danvers, MA 01923 for copying beyond that permitted by Sec-
tions 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying such as
copying for general distribution, for advertising or promotional purposes, for creating new collective works,
or for resale.

SAE routinely stocks printed papers for a period of three years following date of publication. Direct your
orders to SAE Customer Sales and Satisfaction Department.

Quantity reprint rates can be obtained from the Customer Sales and Satisfaction Department.

To request permission to reprint a technical paper or permission to use copyrighted SAE publications in
other works, contact the SAE Publications Group.

All SAE papers, standards, and selected

books are abstracted and indexed in the
Global Mobility Database

No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written
permission of the publisher.

ISSN 0148-7191
Copyright 1999 Society of Automotive Engineers, Inc.

Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely
responsible for the content of the paper. A process is available by which discussions will be printed with the paper if it is published in
SAE Transactions. For permission to publish this paper in full or in part, contact the SAE Publications Group.

Persons wishing to submit papers to be considered for presentation or publication through SAE should send the manuscript or a 300
word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.

Printed in USA
 Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018

1999-01-0431

The Contribution of Occupant and Vehicle Dynamics Simulation

to Testing Occupant Safety in Passenger Cars During Rollover

R. Bardini and M. Hiller

University of Duisburg Germany

Copyright © 1999 Society of Automotive Engineers, Inc.

ABSTRACT restraint system should guarantee occupant protection in

frontal, side and rear impact as well as during rollover.
In the field of vehicle rollover, the quantification of occu-
pant safety provided by a passenger car has, up until Fatalities in Single-Vehicle
now, been quite difficult. As existing test procedures Crashes (total: 16.523)
such as FMVSS 208 and FMVSS 216 mainly aim at the
structural safety of vehicles, there is a lack of standard- 51%
ized injury criteria for the occupants. The results pre- Rollover
sented in this paper are based on the experience gained Non-rollover
during the development of a rollover detection system.
For this purpose a large number of rollover simulations 49%
using the modular vehicle modeling program FASIM_C++
have been conducted. This multi-body simulation tool
has been developed at the Department of Mechatronics,
Fatalities in Multi-Vehicle
and provides the possibility to model complex three-
dimensional vehicle models and include complex mecha- Crashes (total: 19.170)
tronic elements like hydraulic systems, actuators, sen-
sors and electronic controllers. Furthermore the 38% 8% Other
occupant behavior during rollover has been studied. For Rear-end
this purpose the commercial occupant simulation toolset 10% Angle
MADYMO is used. Various accident scenarios are mod-
eled using FASIM_C++ and the subsequent occupant Head-on
behavior is analyzed in MADYMO. 45%

By discussing the simulation results of various rollover Figure 1. Occupant fatalities in traffic crashes in USA in
maneuvers, this paper suggests requirements for a stan- 1997[18].
dardized rollover test procedure that focuses more on
occupant safety than the existing regulations. Looking at the traffic crash statistics (Figure 1) it can be
seen that the involvement rate of rollover crashes in all
INTRODUCTION occupant fatalities in 1997 was higher than 22 percent.
Although this rate is quite high, until now there has been
Optimized occupant protection in passenger cars today is only little investigation in occupant protection during roll-
a fundamental requirement. Featuring front bags, side over. As the existing test procedures mainly aim at the
bags, belt pretensioners and load limiters, existing structural safety of vehicles, there has been not much
restraint systems provide a high level of protection. Addi- pressure on the international automotive industry to
tionally so now that knee bags and head protecting side develop a sophisticated protection system that takes care
bags are starting to come onto market. For the activation of the occupant as well in rollover situations.
of these protective devices comprehensive sensor sys-
tems are required which can react with the appropriate ROLLOVER CRASH TEST STANDARDS
deployment of restraint systems, taking into account any
relevant accident parameter. For this reason future sen- Contrary to other crash types a fully developed test pro-
sor concepts must supply information about vehicle sta- cedure for the quantification of occupant safety in rollover
bility, approaching obstacles, vehicle interior conditions, crashes does not exist. Subsequent a brief overview of
accident type and crash severity [6]. Thus the future the existing regulations that are involved with rollover
safety is given here.

1
 Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018

FMVSS 208 AND SAE J2114 – The tilted table rollover • FMVSS 216,
test procedures according to FMVSS208 and SAE J2114 • FMVSS 201 and
are conducted using a platform with 23° tilt angle. The
• ECE R16.
vehicle is thrown laterally from that platform at a speed of
48 km/h. FMVSS216 deals with the structural safety of the vehicle.
Its purpose is to reduce deaths and injuries due to the
crushing of the roof into the passenger compartment in
rollover accidents. To pass the roof crush test the roof
structure must be stiff enough to prevent excessive defor-
mation when applying a force to the roof frame of 1½
times the unloaded vehicle weight with a velocity of less
than 13 millimeters per Second.
23°
v=48 km/h FMVSS 201 specifies requirements to afford protection to
occupants in interior head impact. The standard estab-
Figure 2. Tilted table test configuration. lishes the performance criterion HIC(d) that shall not
exceed 1000 when a special headform hits the interior
The FMVSS 208 rollover test procedure stipulates one lining of the passenger compartment. Contacts with the
Hybrid III Dummy placed on the driver seat. A measure- windows are not considered.
ment of the dummy loads is not applied. No recommen- In Europe there actually exists no regulation that directly
dations for maximum occupant loads are given but it is deals with vehicle rollover. ECE R16 merely is related to
stated that all doors must keep closed and that no this subject by setting forward the requirements for seat
dummy ejection should occur during the roll movement. belts in various accident situations.
The dolly rollover according to SAE J2114 recommended
practice has mainly been established in order to stan- INJURY MECHANISMS DURING ROLLOVER
dardize the tilted table rollover test between different test
facilities. It also suggest the observation of the vehicle To understand the injury mechanisms that occur in roll-
and occupant kinematics, but it does not determine maxi- over crashes it is first necessary to analyze the character-
mum occupant loads istics of typical rollover accidents. After this it is possible
to draw conclusions about resulting occupant kinematics
SAE J857 – and related injury potentials.

curved guide rail Z ROLLOVER CHARACTERISTICS AND KINEMATICS –

Y As shown in Figure 1, rollover accidents primarily occur
X
in single vehicle crashes. According to [11] leaving the
roadway with the involvement of an object struck, such as
a boulder, a ditch or an embankment along-side the road
ramp
0,84 m is the most likely cause for a rollover. Accident situations
v=80 km/h
where a vehicle skids laterally impacting an object, such
as a curb or soft ground , which trips the vehicle into roll-
0,35 m
3,66 m over are also of importance (see [1] and [2]).
In rollover crashes vehicles carry out complex 3-dimen-
Figure 3. Curved rail-ramp test configuration. sional movements in space. A representative example
would be the trajectory of a passenger car in an embank-
Although the SAE J857 recommended practice for roll- ment maneuver, as shown in Figure 4. In that particular
over tests was drawn up quite a few years ago it should computer simulation the vehicle drives at an angle of 12°
be mentioned here because it provides a realistic simula- positive x-direction and starts to roll when the left front
tion of passenger car rollover accidents. Using a curved wheel reaches the embankment. At the end of the
rail-ramp procedure (Figure 3) a screw movement is maneuver the vehicle has rolled 180° around the x-axis
applied to the vehicle at a speed of 80 km/h. To obtain of the vehicle local coordinate system. In this maneuver
occupant loading and movement during the test SAE the vehicle conducts rotational and translational move-
J857 suggests use of anthropomorphic test devices, but ments around and along all three axes. But only the
there is no detailed information given about dummy types translational movements in the direction of the longitudi-
or injury criteria. nal and the vertical vehicle axes, and the rotational move-
ment around the longitudinal vehicle axis, are
CRASH TEST STANDARDS RELATED TO significantly large. Neglecting the small pitch angle that
ROLLOVER – There are some further international regu- occurs in that maneuver the movement can be described
lations that play an important role in rollover safety: as a screw motion along an inclined axis.

2
 Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018

Z [m]
0.5 impacted side. Thus there is a high risk of occupant injury
0 in the very early stage of the rollover.
-0.5
-1 Occupant kinematics will be analyzed more closely later
-1.5 in the Simulation Results section.
-2
0
5
10
X [m] INJURIES – Like is set forward in the FMVSS 208 it is a
Y [m] 45 40 35 30 25 20 15 10 5 0
basic requirement for occupant protection in a rollover
Figure 4. Trajectory of a passenger car passing over a accident that the occupant stays within the passenger
steep embankment. compartment. Due to mandatory seat belt use laws, ejec-
tion has become quite seldom. Thus today the injury
Another typical rollover trajectory is shown in Figure 5. In mechanisms of the restrained occupant are of major
that computer simulation a passenger car is skidding lat- interest [4]. As the belt keeps the torso of the occupant in
erally against a curb and rolls 380° around its longitudinal place (within a certain range), the arms and especially
axis. As the maneuver is nearly free from yaw and pitch the head, face and neck regions sustain a substantial
motions as well as from translation in longitudinal direc- number of injuries [8].
tion it can be described as a pure rolling motion, that is
superimposed on a translational motion in the vertical According to [4] intrusions of the roof up to 15cm do not
plane. cause serious injuries for restrained occupants, but the
accelerations induced by the rolling movement due to the
Z [m]
1.6 impact cause excessive head motions (see Figures 13
1.4
1.2
and 14). These head motions are responsible for com-
1.0 pression, tension and bending loads in the cervical spine.
0.8 Furthermore, with an unpretensioned lap and shoulder
0.6
0.4
belt a significant torso displacement still occurs that
0.2 makes contacts with the roof, the window or the A- or B-
0.0 pillars very likely. How far the maximum occupant loads
-0.2
defined in existing crash regulations are applicable in
-0.4 these load cases still has to be examined, but, especially
X [m]
-0.2
0.0 4 3 for the lateral bending motion of the neck that is relevant
Y [m] 7 6 5
8 in many rollover accidents, there are no limits set forward
Figure 5. Trajectory of a passenger car skidding laterally in any regulation. At least many publications such as
against a curb. [13], [17], [5], and [15] have analyzed the biomechanical
properties of the human neck under various loads and
Regarding these two maneuvers, the vehicle kinematics indicated some of the possible injuries.
in rollover accidents can be divided into two groups:
ROLLOVER SIMULATION
1. screw motion along an inclined axis and
2. pure rolling motion with translation in the vertical For the development of a rollover detection system com-
plane. puter simulation plays an important role [9]. Vehicle
The first group covers rollover accidents that involve dynamics simulation, for example, provides the possibility
impacted objects, such as boulders (working as a ramp), to test out in advance various sensors and algorithms for
ditches or embankments. The second group covers rollover detection. Furthermore occupant simulation can
tripped rollovers that occur in an impact with a curb or be used to establish trigger times for rollover detection.
other obstacles, or respectively in a sideslip through soft In the field of vehicle dynamics, the modular computer
soil. program FASIM_C++ has proved to be extremely helpful.
For occupant simulation the commercial simulation
OCCUPANT KINEMATICS – The typical occupant toolset MADYMO is used. MADYMO provides sophisti-
motions that occur in rollover crashes correspond to the cated and validated multi-body models of crash test dum-
vehicle kinematics. Consequently the two representative mies, that can be placed in any accident scenario [12].
rollover kinematics described in the previous section gen- The following describes the simulation tools used for the
erate two different kinds of occupant kinematics. In the development of the rollover detection system, ROSE, of
initial stage of a screw motion rollover the occupant tends Robert Bosch GmbH [14]. Various accident scenarios are
to move relatively slowly with respect to the passenger modeled with FASIM_C++ and the resulting occupant
compartment. When the vehicle first hits an obstacle, i. e. behavior is analyzed in MADYMO. For this purpose posi-
the ground, the occupant starts to move rapidly towards tion and orientation data versus time, computed by
the side the impact has come from. As the pure roll Fasim_C++ can serve as input data for the MADYMO
motion type is initiated by an impact it causes an immedi- occupant simulations.
ate and rapid displacement of the occupant towards the

3
 Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018

Controllers Sensors

Stabilizer
1=abstraction Z
2
2=mathematical modeling X
McPherson
Suspension
McPherson
Suspension
Wheel Wheel
1 2
Z 1 2

Steering
System

1
Rigid body model of the
McPherson strut suspension
Engine
Chassis Suspension

Hydraulic
Systems Power
Train

Wheel Wheel
Rear Axle
3 4

Environment
(Street, Ramp, Embankment, etc.)

McPherson strut front suspension Modular structure in FASIM_C++

Figure 6. Modular structure of a vehicle model in FASIM_C++

THE SIMULATION TOOL FASIM_ C++ – The multi-body without any recompilation of the program. F ASIM_C++
simulation tool FASIM_C++ has been developed. Its main contains a large library of different vehicle modules like
task is to support the development of vehicle controllers wheel suspensions, tire models, powertrains, engines,
[10]. Controller systems such as antilock braking systems engine suspensions, controllers, sensors, elasticities,
(ABS), vehicle dynamics control (VDC or ESPTM) [19] or rigid and flexible car bodies, several hydraulic braking
rollover detection are currently developed with the help of systems, a driver and an environment model.
FASIM_C++. For that purpose an appropriate vehicle
The similar structure of the modules makes it easy to
model, included in a versatile and flexible simulation envi-
expand the library by adding new modules. For modeling
ronment, is required. This simulation environment has to
new modules, first the physical model of a real mecha-
be able to simulate different vehicle types or models with-
nism is generated (abstraction). The type and complexity
out any recompilation. The multi-body vehicle model of a
of the physical model depends on later simulation tasks,
typical passenger car has to have a modular structure, so
so in this step it is very important to specify exactly the
that single components of the vehicle may be exchanged
requirements. To transform the physical model into a
depending on the simulation task.
mathematical one, appropriate procedures for solving the
The modular structure of a passenger car, as it is used in arising system of differential algebraic equations (DAE)
FASIM_C++ is shown in the right hand portion of Figure 6. has to be found (mathematical modeling). After this the
This modular structure enables the user to switch easily system of equations is transformed into source code in
between various vehicle models. Since each module is C++. For further information about the generation of the
independent and has its own data file it is, for example, equations of motions in FASIM_C++ see [10].
possible to use the sub-module McPherson front suspen-
The mechanical part of the vehicle is modeled as a three
sion in various vehicles, and in a rear as well a front axle
dimensional complex multi-body system consisting of
configuration. Some changes in the data files (dimen-
rigid and flexible bodies and closed kinematic loops.
sions, masses, inertia, etc.) only have to be carried out to
More detailed information about the principles of multi-
effect these changes.
body simulation can be found in [7] and [20]. Figure 7
The modular concept was realized using the program- shows the topological structure of the complete vehicle,
ming language C++. In this way it is possible to decide that has been used for the rollover simulations presented
during runtime, which configuration of a vehicle is used in this paper. The vehicle system is described with

4
 Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018

respect to the inertial system by a fictitious joint with six a rollover situation, it is very important that this maneuver
degrees of freedom, so that all degrees of freedom of a is detected by a rollover protective system. Therefore, a
rigid body can be taken into consideration. full scale rollover test with a middle class car (Figure 7)
has been investigated in detail.
Engine Suspension L18
Z

McPherson Strut Power Train Y b

Street X
L3 µ=0,8
L1 L4 Wheel
µ=0,45

Ramp h
L2 L5
Car Body Embankment 1
Five Point Wheel
Suspension
H
Embankment 2

Figure 8. Example of the surface contour used for

rollover simulation

As shown in Figure 9 a very good correlation between the

L6,..., L 11 L12,..., L17 simulation and the real experiment has been achieved.
Only when the car body hits the ground does the simula-
Joint with one Degree of Freedom Rigid Body tion yield wrong results, as the contact interactions
Independent Joint Coordinate Transmission
between the exterior of the vehicle and the environment
Joint with three Degrees of Freedom Differential
have not been modeled. Since this phase of the rollover
Isolated Degree of Freedom Kinematical Loop
is no longer of importance for rollover detection these
errors have been neglected: When the car body hits the
ground rollover detection must have taken place long
System DOF Bodies Loops ago.
Car Body 6 1 0
The sensor module that has been implemented in the
Front Axle 3 11 5
vehicle during simulation is fed with translational acceler-
Rear Axle 2 12 10 ation and angular velocity data by the chassis module,
Power Train 6 12 3 and gives back the trigger signal for controlling the roll-
Wheel Suspension 2 1 0 over protective devices. The instant of rollover detection
Vehicle 19 37 18 is visualized by a cone which has been added to the ani-
mation and which becomes visible above the hood when
Figure 7. Topology of a vehicle modeled in Fasim_C++ the sensor triggers (see Figure 9). Unfortunately in this
particular case a recording equipment failure means that
For conducting rollover simulations some special this comparison can only be qualitative. However other
enhancements have been made in FASIM_C++. Firstly customer data confirms the quantitative validity of the
the sensor including the ROSE rollover detection algo- model.
rithm was implemented. Thus it was now possible to
analyze the trigger behavior in any simulated maneuver.
To incorporate the possible collision between the vehicle
and different obstacles, e.g between wheel rim and a
curbstone, the original model of the power train was
extended by an impact model. For this purpose a simple
Kelvin-Voigt model, which consists of a parallel array of a
spring and a damper element connected to an impact
mass, is used [10].
Furthermore it was necessary to enhance the modeling
of the environment. For the simulation of embankment
and ramp maneuvers it is now possible to configure sur-
faces like those shown in Figure 8. Figure 9. Comparison of simulation with full scale
rollover
Validation – The application of the modeling techniques
to the analysis of real-world vehicle problems shall be With the validated model described above it is possible to
illustrated by the simulation of an embankment rollover. perform parameter studies in order to optimize the roll-
As leaving the road is statistically the most likely cause of over sensing concept, and to establish trigger times for

5
 Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018

rollover protective devices like belt pretensioners and As defined in the section “Rollover characteristics and
window bags. By combining FASIM_C++ and MADYMO Kinematics”, vehicle kinematics in rollover accidents can
an application and development environment can be be divided into two groups. As a representative simula-
formed, as illustrated in Figure 10. tion for the group one (see above) the simulation results
of an embankment maneuver are discussed as follows.
OCCUPANT SIMULATION USING MADYMO – Whereas, as representative for the second group (see
MADYMO (MAthematical DYnamic MOdel) is a commer- above) a curbstone impact is analyzed.
cial computer package which has mainly been developed
for studying occupant behavior during car crashes [12]. Embankment Rollover – In Figure 12 an animation
MADYMO combines in one simulation program the capa- sequence of the MADYMO occupant simulation of the
bilities offered by multi-body and finite element tech- embankment maneuver is shown (Figure 4). For that
niques. Furthermore, MADYMO offers a set of standard simulation the position and orientation data created by
force models e.g. for belts, airbags and the contact of FASIM_C++ is used to describe the movements of the
bodies with each other or with their surroundings. Con- passenger compartment until the car body hits the
sequently, it also allows assessments to be made on the ground. In order to be able to study rolling when the car
suitability of various restraint systems. One main feature body is in contact with the ground the simulation is con-
of MADYMO is the availability of validated data-base of tinued in MADYMO. In Figure 12 it can bee seen that the
crash-test dummy multi-body models, which are suitable restrained driver (50th percentile male Hybrid III Dummy)
to study injuries sustained by accident victims with a high experiences only small displacements during the first 90°
accuracy. The ''50th percentile male Hybrid III Dummy'' of the rollover. Right when the car hits the ground the left
represents an ''average'' of the USA adult male popula- shoulder and arm come into contact with the drivers door,
tion. It is modeled as a multi-body system and has 57 and the head moves fast diagonally forward and toward
degrees of freedom. The interaction between dummy the side window. The severity of the impact to the head is
and passenger compartment can be represented by geo- proportional to the vehicle height and the embankment
metric primitives such as planes, ellipsoids and cylinders height.
attached to the bodies of the multi-body system (Figure
In the context of these occupant kinematics, similar
11). As in rollover maneuvers there often occurs a slip
effects have been observed in ramp rollover simulations
movement between the dummy body and the seat belt, a
where the vehicle is passing over the ramp with the right
finite element belt model for the seat belt was imple-
hand side wheels. Here the impact severity is propor-
mented. As the roof structure is assumed to be stiff
tional to the vehicle velocity and the ramp height.
enough to prevent collapsing, no intrusions of the roof
are taken into account (compare [4]).
Curbstone Rollover – Due to the plastic deformation that
In order to be able to make estimations about the occu- occurs in the vehicle body and suspension components
pant behavior when the vehicle is rolling on its bodywork during the curbstone impact the multi-body model does
the exterior contour of the vehicle was also included in not give a very good representation of the subsequent
the MADYMO contact modeling. vehicle behavior, as was shown during full scale crash
testing. Consequently actual crash test angular velocity
SIMULATION RESULTS – The following table gives an and translational acceleration data (customer confiden-
overview of all the rollover and close-to-rollover situations tial) is used to drive the occupant simulation in MADYMO.
that have been simulated with FASIM_C++ in order to ver- The occupant kinematics that typically occur in a rollover
ify the ROSE-Algorithm. Especially in rollover situations, accident induced by lateral skidding against an obstacle
the detailed examination of the occupant behavior in are described in Figure 13. Due to the lateral impact the
MADYMO is of importance since rollover detection has to vehicle is abruptly decelerated, whereas the driver is still
take place early enough for an “on time” deployment of traveling at the initial lateral speed, and due to the sus-
the protective devices. pension preload he is slightly inclined towards the door.
The seat belt is not able to hold the driver firmly in the
seat and so he moves quickly towards the left hand side
Without Rollover
door. The impact of the left arm, the shoulder and the
Maneuver Simulation Purpose
serpentine resonance behavior of the sensor
head takes place long before the vehicle has rolled 90°.
skidding against curbstone effect of lateral acceleration pulse The severity of the head impact, as well as the roll angle
low ramp (2 wheels) triggering behavior close to rollover at which the head impact occurs, are directly related to
jump over ramp (4 wheels) behavior during free fall the lateral skid velocity of the vehicle.
circular course effect of transvers acceleration
A belt pretensioner strongly influences the occupant kine-
gentle embankment triggering behavior close to rollover
matics. Comparing the trajectory of the dummy head in
With Rollover
an 180° curbstone rollover at 36km/h without a belt pre-
Maneuver Simulation Purpose
rolling without ground contact trigger time
tensioner (dotted line in Figure 14) with the same maneu-
steep embankment trigger time ver using a pretensioner (solid line in Figure 14) the
skidding against curbstone trigger time following conclusions can be made:
high ramp (2 wheels) trigger time

6
 Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018

P o s itio n a n d O rie n ta tio n D a ta

,ω Z

, ω Y

,ω
X

M AD YM O 3D , aZ
F A S IM _ C + +
,a
Y

aX

S e n s o r S im u la tio n

Ve
r if R O S E A lg o rith m io n
ic a s at
tio a li
n su
Vi
Trigg er D e cisio n
Figure 10. Using FASIM_C++ and MADYMO 3D as calibration tools for ROSE

rollover crashes. As FMVSS 208 and FMVSS 216 focus

on structural safety and the rollover test configuration set
forward in the FMVSS 208 is not a very realistic repre-
sentation of a true crash scenario (compare [1] and [2]) a
new or improved rollover safety test method should be
defined. This test method must fulfill the following
requirements:
• utilize a realistic and reproducible rollover maneuver,
Figure 11. Dummy sitting in the passenger compartment that covers the worst case,
• set forward injury criteria,
On the one Hand the reduction in the vertical direction is • combine structural testing with occupant safety
very helpful for occupant protection as it increases the criteria.
safety space. This is especially important when roof
intrusion occurs, or the vehicle is a cabriolet. However Considering the experiences gained through computer
due to the stronger restraint of the upper torso the forces simulations conducted for the development of the rollover
in the neck are increased, and the head moves much detection system ROSE, it can be concluded that rolling
faster towards the window. That leads to a higher impact over due to a curbstone impact induces the most critical
force when the head hits the window. Although the lateral occupant kinematics. Assuming that no roof collapse
displacement of the upper torso is reduced the lateral occurs a rollover protection system that is able to provide
holding force of the shoulder belt is not high enough to protection in a curbstone rollover should easily be able to
avoid the head contacting the window. Thus in this situa- give protection in other rollover types. To determine the
tion it would be very beneficial to deploy a window bag to quality of a rollover protection system it is especially nec-
cushion the impact. essary to define injury criteria for the head and neck
region. Additionally there must be a suitable anthropo-
• the displacement of the dummy in vertical direction is morphic test device available, that especially provides
reduced and more realistic neck kinematics. As observed already in
• the lateral displacement of the dummy head is [16] and also in the occupant simulation the neck of the
increased. Hybrid III Dummy seemed to be to stiff in the lateral direc-
tion. At the moment occupant simulations are being con-
CONCLUSIONS ducted using a more realistic neck model according to
the results published in [3].
Rollover sensing is a new feature on the market, and it
represents a big enhancement in the restraint systems ACKNOWLEDGMENTS
area. The idea is to include the rollover function into the
actual air bag and seat belt concept so that it will improve A major part of the work presented in this paper has been
the performance of existing restraint systems. A major supported by Robert Bosch GmbH Stuttgart, Germany.
problem however will be the application of such a system
The Authors wish to express appreciation to Derek Ward
to real vehicles. This is because at the moment there
for his contributions to this paper.
exists no safety standard that defines occupant Safety in

7
 Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018

Figure 12. Animation of the embankment rollover

REFERENCES

1. Cooperrider, N. K.; Hammoud, S. A.; Colwell J.: Character-

istics of Soil-Tripped Rollovers. SAE Technical Paper
Series 980022. Warrendale, PA 1998.
2. Cooperrider, N. K.; Thomas, T. M.; Hammoud, S. A.: Test-
ing and Analysis of Vehicle Rollover Behavior. SAE Techni-
cal Paper Series 900366. Warrendale, PA 1990.
3. de Jager, M.: Mathematical Head-Neck Models for Acceler-
ation Impacts. Dissertation: Eindhoven University of Tech-
nology, Netherlands 1996
4. Friedman, D.; Friedman, K.: Roof Crush Versus Occupant
Injury from 1988 to 1992 NASS. SAE Technical Paper
Series 980210. Warrendale, PA 1998.
5. Gadd, C. W.; Culver, C. C.; Nahum, A. M.: A Study of
Response and Tolerances of the Neck. SAE Technical
Paper Series 710856. Warrendale, PA 1971.
6. Grösch, L.; Mattes, B.; Schramm, D.: Smart Restraints. In
Proceedings: Airbag 2000, 2nd International Conference.
Karlsruhe, Germany 1996
7. Haug, E. J.; Computer Aided Kinematics and Dynamics of
Mechanical Systems. Volume I: Basic Methods. Allyn and
Bacon Series in Engineering. Allyn and Bacon, Boston,
Figure 13. Animation of the curbstone impact 1989.
8. Hight, P. V.; Siegel, A. W.; Nahum, A. M.: Injury Mecha-
Displacement in Z-Direction [m] nisms in Rollover Collisions. SAE Technical Paper Series
0,25 without Pretensioner Roof Contour 720966. Warrendale, PA 1972.
with Pretensioner
9. Hiller, M. and Bardini, R.: Vehicle and Occupant Dynamics
0,20 Simulation – Important Tools in the Development of Sensor
Concepts for the Control of Restraint Systems. In Proceed-
0,15
ings: 8th German-Japanese Seminar on "Nonlinear Prob-
lems in Dynamical Systems - Theory and Applications":
0,10
p. 15-39. Kobe, 1998.
0,05 Initial COG of 10. Hiller, M.; Schuster, C.; Adamski, D.: FASIM_C++ - A Versa-
Dummy Head
tile Developing Environment for Vehicle Dynamics Simula-
0,00
tion. Int. Journal of Crashworthiness, Volume 2(1): p. 91-
105. 1997
-0,05
11. Insurance Institute for Highway Safety: 1997 Fatality Facts:
-0,40 -0,35 -0,30 -0,25 -0,20 -0,15 -0,10 -0,05 0,00 0,05 0,10 0,15
Displacement in Y-Direction [m]
Roadside Hazards. Arlington, VA 1998.
12. Lupker, H. A.: MADYMO, a Versatile Tool for Vehicle Safety
Analyses. In Proceedings: 6th International MADYMO
Figure 14. Trajectory of head during curb-stone Rollover
Users’ Meeting: p. 139-145. Amsterdam, 1996

8
 Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018

13. McElhaney, J. H.; Myers, B. S.: Biomechanical Aspects of 20. W. Schiehlen, editor. Multiboby Systems Handbook.
Cervical Trauma. In: Accidential Injury; Nahum, A. M.. Springer Verlag, Berlin, Heidelberg, New York, 1990.
Springer Verlag, New York 1993.
14. Mehler, G.; Mattes, B.; Henne, M.; Lang, H.-P.; Wottreng, CONTACT
W.: Rollover Sensing (ROSE). . In Proceedings: 2nd Inter-
national Conference on Advanced Microsystems for Auto- Prof. Dr.-Ing. habil. Manfred Hiller
motive Applications. Berlin 1998 Gerhard-Mercator-Universität - GH Duisburg
Fachbereich Maschinenbau, Fachgebiet Mechatronik
15. Pintar, F. A.; Yoganandan, N.; Sances A.; Harris G.; et al.:
Lotharstraße 1
Kinematic and Anatomical Analysis of the Human cervical
47057 Duisburg, Germany
Spinal Column Under Axial Loading. SAE Technical Paper
email: [email protected]
Series 892436. Warrendale, PA 1989.
16. Piziali, R.; Hopper, R.; Girvan D.; Merala R.: Injury Causa- Dipl.-Ing. Roberto Bardini
tion in rollover Accidents and the Biofidelity of Hybrid III Gerhard-Mercator-Universität - GH Duisburg
Data in Rollover Tests. SAE Technical Paper Series Fachbereich Maschinenbau, Fachgebiet Mechatronik
980362. Warrendale, PA 1998. Lotharstraße 1
47057 Duisburg, Germany
17. Schneider, L. W.; Foust, D. R.; Bowman B. M.; Snyder R.
email: [email protected]
G.; et al.: Biomechanical Properties of Human Neck in Lat-
eral Flexion. SAE Technical Paper Series 751156. Warren-
DEFINITIONS, ACRONYMS, ABBREVIATIONS
dale, PA 1975.
18. U.S. Department of Transportation NHTSA: Traffic Safety ECE: Economic Commission for Europe
Facts 1997 - Overview. Washington, D.C. 1998. FMVSS: Federal Motor Vehicle Safety Standard
19. van Zanten, A.; Erhardt, R.; Lutz, A.; Neuwald, W.; Bartels, ROSE: Rollover Sensing
H.: Simulation for the Development of the Bosch-VDC. SAE
Technical Paper Series 960486. Warrendale, PA 1996.