schriren

SCHRIREN
REHE NR.96
Vermessung des
50%-Hybrid III Dummy
zur Ermittlung eines
verbesserten Datensatzes
für Crashsimulationen
Vermessung des 50%-Hybrid III Dummy
zur Ermittlung eines verbesserten
Datensatzes für Crashsimulationen
Auftraggeber:
Forschungsgemeinschaft Automobiltechnik e.V. (FAT)
Auftragnehmer:
TNO Road-Vehicles Research Institute
Delft, Niederlande
Verfasser:
Mat Philippens, John Nieboer, Jacques Wismans,
Martin Heinz, Bernd Pletschen, Harald Wester,
Thomas Scharnhorst
Postanschrift:
Postfach 170563 • 6000 Frankfurt/M. 17
Telefon (069) 75 70-1
Drahtanschrift: Autoverband
Telex 411293
Druckerei Henrich GmbH
Schwanheimer Straße 110
6000 Frankfurt am Main 71
Vervielfältigung, auch auszugsweise nur
mit ausdrücklicher Genehmigung der FAT
VORWORT
Seit langem ist die Entwicklung sicherer Autos eine der wichtigsten Aufgaben der Automobilingenieure. Als Maßstab zur Bewertung der passiven Sicherheit der Fahrzeuge gelten die bei Fahrzeugunfällen auf die Insassen ausgeübten Belastungen. Um diese vorab schätzen zu können, werden mit
neuen Fahrzeugtypen aufwendige Kollisionstests durchgeführt, bei denen die physikalischen Belastungen mit Hilfe von Meßpuppen, den sog. Dummies, ermittelt werden. Auf diese Weise wird dargestellt, inwieweit Fahrzeuginsassen selbst bei extremen Unfällen in einem Auto noch geschützt
sind.
Da jedoch Versuche dieser Art sehr langwierig und kostenspielig sind, wird zunehmend unter Anwendung der weiter entwickelten Simulationstechnik versucht, Erprobungsdauer, Versuchsumfang,
Auswertung und Umsetzung der Versuchsergebnisse zu verkürzen. Hierzu bedarf es allerdings
ausgefeilter Simulationsverfahren, mit denen das kinematische Verhalten der Dummies bei CrashVersuchen genau nachgebildet werden kann.
In Verfolgung dieser Zielsetzung haben die Mitglieder des FAT-AK 5 Unterausschuß 'CrashDummy', die im Anhang namentlich genannt sind, beschlossen, dem TNO Road-Vehicles Research
Institut in Delft den Auftrag für eine genaue Vermessung der bei Kollisionstests eingesetzten
Dummies zu erteilen.
Mit der vorliegenden Broschüre wird die Kurzfassung des Abschlußberichts zu diesem Projekt vorgelegt, das die genaue Vermessung des Hybrid III Dummy (50% Mann) und die Übernahme dieses
Datensatzes zur Verbesserung eines Modells zur mathematischen Simulation von Fahrzeugkollisionen zum Ziele hatte. Hierfür mußten teilweise neue Meßverfahren entwickelt und aufwendige
Messungen statischer und dynamischer Art durchgeführt werden, um die geometrischen und physikalischen Eigenschaften eines realen Dummy's im Computermodell abbilden zu können.
Die Verfügbarkeit eines solchen Simulationsmodells ist von erheblicher wirtschaftlicher Bedeutung,
da der vermessene Dummy alternativ zum Hybrid II Dummy nach FMVSS 208 (Federal Motor
Vehicle Standard) für die Zertifizierung von Neufahrzeugen in den USA vorgeschrieben ist.
Um die Verständlichkeit der Broschüre zu erhalten, wurde der Wortlaut der von den einzelnen Bearbeitern niedergeschriebenen Kapitel nicht verändert, d.h. in dieser Broschüre im deutschen bzw.
englischen Original beibehalten.
FORSCHUNGSVEREINIGUNG AUTOMOBILTECHNIK EV (FAT)
Frankfurt am Main, im Februar 1992
INHALTSVERZEICHNIS
Seite
Zusammenfassung
7
1.
Einleitung
9
2.
An Advanced Database of the 50th Percentile Hybrid III-Dummy
11
3.
An Advanced 50th Percentile Hybrid III-Dummy Database : Validation
23
4.
Ausblick
39
5.
Literatur
41
Anhang: Mitglieder des FAT-AK5 Unterausschuß "Crash-Dummy"
Zusammenfassung
Ziel des Forschungsvorhabens ist es, einen neuen, für Frontal-Crash-Tests nach USA-Standard FMVSS 208 zunächst alternativ zum Hybrid II vorgesehenen Hybrid Ill-Dummy (50 %Mann) in Bezug auf seine geometrischen und physikalischen Eigenschaften statisch und
dynamisch
zu vermessen und in ein mathematisches Modell zu überführen. Der erste Teil
der Aufgabe umfaßt auch die Entwicklung einer problemorientierten Meßmethodik, die dann
konkret angewandt wird auf die Dummy-Vermessung. Das 3D-Simulationsmodell des
Dummys besteht aus 20 Segmenten und ist damit verfeinert gegenüber einem früheren Modell mit nur 15 Segmenten. Die zusätzlichen Segmente beschreiben Hände, Schultern und
Brustbein.
Darüber hinaus hat sich der Arbeitskreis die Aufgabe gestellt, den so entstandenen DummyDatensatz des 50 %-Hybrid III zu validieren. Dies geschieht durch Abgleich von Schlittenversuchen (von Volkswagen durchgeführt) und entsprechenden Simulationsrechnungen. Die
jeweiligen Ergebnisse zeigen eine gute bis befriedigende Übereinstimmung. Der Datensatz
konnte durch den Arbeitskreis gezielt weiterverbessert werden, so daß nun ein
mathematisches Hybrid HI-Modell des 50 %-Mann von
prognosefähigen
Einsatz
bei
Frontalcrashversuchen vorliegt.
der
hoher Qualität zum produktiven,
computergestützten
Insassensimulation
von
1. Einleitung
Die Forderung nach Zeit- und Kosteneinsparung im Automobilentwicklungsprozeß besteht
für alle Hersteller gleichermaßen. Neue Technologien in Konstruktion, Versuch und Berechnung finden vor diesem Hintergrund vermehrt frühzeitig Anwendung. Im Hinblick auf in Zukunft immer kürzer werdende Entwicklungsintervalle von Automobilen gewinnen numerische, computergestützte Simulationsverfahren stetig an Bedeutung. Ohne daß ein reales
Prototypfahrzeug existieren muß, wird es somit möglich, bereits in den frühen Phasen der
Automobilentwicklung Aussagen über den Stand z. B. der passiven Sicherheit im Auto zu
machen.
Die computergestützte Insassensimulation beschäftigt sich mit der numerischen Analyse
des dynamischen Verhaltens von Fahrzeuginsassen beim Crash. Die dabei verwendeten Insassenmodelle orientieren sich überwiegend an den vom Gesetzgeber vorgeschriebenen
physikalischen Versuchsdummies.
Die Insassensimulation hat sich als Teilgebiet der Unfallforschung seit Anfang der siebziger
Jahre mit der Verfügbarkeit erster zweidimensionaler Berechnungsprogramme auf der Basis
von Starrkörpermechanismen - sog. Mehrkörpersysteme - entwickelt. Heute sind SD-Programme wie MADYMO-3D und CAL 3D international als Standard anzusehen.
Voraussetzung für die Prognosefähigkeit von Ergebnissen der Insassensimulation sind
hochwertige Datensätze für die mathematischen Modelle der für Crash-Tests relevanten
Dummies. Der FAT-AK5 Unterausschuß "Crash-Dummy" hat sich die Entwicklung bzw. Verbesserung entsprechender mathematischer Modelle zum Ziel gesetzt.
Das erste in diesem Arbeitskreis definierte Forschungsvorhaben, dessen Ergebnisse mit
diesem Bericht zusammengefaßt werden, hatte zum Ziel, einen neuen, für Frontal-CrashTests nach USA-Standard FMVSS 208 vorgesehenen 50 %-Hybrid Ill-Dummy in Bezug auf
seine geometrischen und physikalischen Eigenschaften zu vermessen
und in ein
mathematisches Modell zu überführen. Parallel hierzu definierte der Arbeitskreis Schlittenversuche entsprechend dreier Frontalaufprallgeschwindigkeiten (40, 50, 60 km/h), deren
Durchführung von VW übernommen wurde. Der Forschungsnehmer des Vermessungsprojekts, TNO, lieferte darüber hinaus die versuchsentsprechende Simulation mit dem Programm MADYMO. Der Vergleich zwischen Versuchs- und Simulationsergebnissen, der
durch den UA "Crash-Dummy" durchgeführt wurde, zeigte
9
-2-
die gute Qualität aber auch einige Schwachstellen des Dummydatensatzes auf und führte
dann zu einer Verbesserung des endgültigen Datensatzes seitens des FAT-AK. Es liegt
damit ein validiertes mathematisches Modell des 50 %-Hybrid III Dummy vor.
Alle Daten des Dummy-Vermessungsprojektes sind in Form eines umfangreichen Abschlußberichts dokumentiert und den AK-Mitgliedern übergeben worden. Da der große Umfang des Projektberichtes (1), (2), (3) für eine übersichtliche Darstellung der Ergebnisse
komprimiert werden mußte, hat sich der AK entschieden, an dieser Stelle zwei Vortragsveröffentlichungen (4), (5), (6), (7) seitens des Forschungsnehmers TNO - siehe Kapitel 2 - sowie des FAT-AK5 UA "Crash-Dummy" - siehe Kapitel 3 - wiederzugeben, die in englischer
Sprache und weitgehend identisch auf der MADYMO-User's Conference, Mai 1990, in
Noordwijk/Niederlande
und der SAE-Conference and Exposition, Februar
Detroit/USA gehalten wurden.
10
1991, in
2.
An Advanced Database of the 50th
Percentile Hybrid Ill-Dummy
(M. Philippens, J. J. Nieboer, J. Wismans; TNO Road Vehicles
Research Institute, Delft, Niederlande)
ABSTRACT
This paper presents a measurement program of a sitting 50 t h percentile Hybrid
III Dummy to determine a database for Computer simulations. Geometrical,
inertial, Joint property and surface compliance measurements have been carried
out.
On the basis of these measurements a 20 segment database for the MADYMO 3D
occupant Simulation program is developed. The major advancements of this
database compared to an earlier 15-segment database developed by TNO [1] can
de summarized as follows:
Five additional Segments are incorporated in this database to account for
the hands, the shoulders(clavicles) and the sternum.
The database includes a complete omni-directional description for the
neck as well as the lumbar spine.
A detailed mathematical surface description is available to be used, for
instance for Computer animations.
Segment ellipsoids for contact interactions have been determined in a
more accurate way.
Joint properties are determined with a Special developed static Joint
measuring device.
Segment surface and thoracic stiffness data are based on a large series
of tests with different impactor faces, including tests with a seat belt.
In this paper the equipment used irt the tests and the applied measurement
techniques will be described. Some examples of typical measurement results
will be given.
INTRODUCTION
In the field of automobile crash research, Computer simulations have shown a
strong increase in use, particularly aue to the developments in Computer
hardware and Simulation Software in ehe past years. Specifically Computer
modeis have proven to be beneficial in reducing the development time of a
new vehicle model and in reducing the number of crash tests required. They
allow an efficient means of evaluating the influence of parameter changes and
moreover they can be used to evaluate the Performance of new design coneepts
even before a prototype has been built. An important requirement for an
effective use of Computer modeis is that rcliable well validated databases are
available particularly for the Simulation of the human being in a crash
environment.
The objeetive of this study is the development of a crash dummy database on
the basis of a set of well defined measurements and experiments which speeify
the properties of this dummy.
This study deals with a database for the 50 t h percentile Hybrid III dummy
(Fig. 1). This dummy is generally considered to be one of the most advanced
crash test dummies available at the moment. Use of this dummy for the
evaluation of vehicle safety Performance is Standard practice now within the
11
automotive industry. Earlier efforts to develop a database for this dummy were
carried out at Wright Patterson Air Force Base (VPAFB) in Dayton, Ohio [1,2].
In their study two dummies were measured: a Standing and a sitting one. TNO
used the data resulting from these measurements to formulate in 1987 a
preliminary 15 segments database of the Hybrid III dummy for the MADYMO 3D
occupant Simulation model. This database further referred to as SAE database
was validated using a series of Hybrid III sied tests conducted by Ford Motor
Company. A quite good agreement between simulations and experiments could be
observed [3].
Fig. 1 The 50th Percentile Hybrid III dummy
The present research program seeks to optimize this SAE database on the basis
of a new more extensive measurement program of the Hybrid III dummy. The
measurements are conducted in close co-operation with Wright Patterson Air
Force Base where the inertial properties of the dummy have been determined.
The most important differences with the earlier WPAFB measurements [1,2] are:
12
A division in a larger number of dummy segments to be measured allowing
a more detailed, Computer representation.
A detailed surface discretisation of the dummy segments alloving a more
realistic visualisation of the simulated dummy kinematics.
A more accurate contact ellipsoid selection on the basis of the segment
surface discretisation.
A detailed measurement of surface compliancies including ribcage response
using static and dynamic tests with different compression faces.
A more extensive measurement of static and dynamic Joint properties
particularly with respect to the dummy neck.
The characteristics of the dummy are obtained by various measurements. A
subdivision can be made into geometric, mass distribution, Joint property and
segment surface stiffness measurements. The equipment used in the tests and
the applied measurement techniques will be described here, in conjunction
with examples of typical measurement results. In this study one sitting Hybrid
III dummy was measured. Detailed results of the measurements are presented
in Ref. [4].
SEGMENT AND JOINT SPECIFICATION
The first step in developing a multi-body model of a crash dummy is the
division of the dummy in a number of segments and the specification of the
parts which belong to each segment. The segments are selected by dividing the
dummy into functional components. Each part of the dummy having significant
mass and a flexible connection with other parts is considered as a segment.
Dummy parts which do not show any relative motion are considered to be part
of one segment. The dummy is divided in 17 main segments listed in Table 1.
Table 1 Division of dummy in segments
Main segments:
- Head
- Neck
- Upper Torso
- Abdomen
- Lower Torso
- Upper Arms (2)
-
Lower Arms (2)
Hands (2)
Upper Legs (2)
Lower Legs (2)
Feet (2).
Subsegments upper Torso:
- Thoracic Spine
- Clavicles (left & right)
- Sternum assembly and part of the ribs
The Upper Torso has been devided into a number of subsegments allowing a more
detailed representation of this segment. These subsegments are included in
Table 1. The clavicle subsegments are proposed
to account for
shoulder/clavicle flexibility while a separate ribcage subsegment is important
for simulating inertial effects of the ribcage. Moreover due to this
subdivision, dummy interaction with a belt or airbag restraint system can be
simulated more realistic.
Ref. [4] provides a detailed description of each segment. A hardware listof
the various parts is included in this reference.
13
The following Joint types can be distinguished in the Hybrid III dummy in
agreement vith above definitions:
neck, and lumbar spine:
knee and clavicle:
shoulders, elbow and wrist:
hip and ankle:
flexible rubber structures represented by
Joint centres in the end points,
pin joints,
universal joints,
ball and socket joints
A more detailed description of these joints can be found in reference [4].
A local right-handed coordinate system has been defined for each segment and
subsegment. Georaetrical and mass distribution properties will be expressed
relative to these coordinate Systems. The coordinate axes are selected in such
a way that in general in the sitting position of the dummy all z-axes are
directed upward, all x-axes are forward and all y-axes are directed to the
left. The location of the origin and the coordinate axes is selected on the
basis of well defined landmarks in the dummy segments. Usually the origin of
the local coordinate system has been selected in one of the Joint centres of
the segment while the coordinate axes often coincide with Joint axes. An
illustration of each segment coordinate system is given in Fig. 2.
Fig. 2
14
Hybrid III dummy model with local segment coordinate Systems.
GEOMETRY
Different types of geometrical measurements have been carried out within this
study. The most important ones for the model development are the determination
of the Joint locations within the individual Segments and the determination
of a detailed surface description. These measurements are conducted at a
disassembled dummy.
Determination of the position of the Joint centres and Joint axes directions
has been conducted with a so-called Perceptor. This is a 3-dimensional
measuring device provided with a digitizing arm to measure the x,y,z
coordinates of a point in space [1,4]. Most of the geometrical Joint data have
been determined in an indirect way since the requested Joint data usually are
not directly accessible by the perceptor. In case of a ball and socket Joint
for instance the position of a number of points on the Joint surface can be
measured, from which the Joint centre coordinates can be calculated.
Additional measurements of landmarks specifying segment local co-ordinate
Systems and corresponding co-ordinate transformations were performed to
express the Joint centre positions and Joint axis orientation in the segment
local co-ordinate Systems.
The outside surface of the dummy segments in most Crash Victim Simulation
modeis is represented by means of ellipsoids. These ellipsoids are used for
Visual presentation of the occupant kinematics as well as for the calculation
of the contact interaction between dummy segments and environment (e.g. the
vehicle interior). The location and dimensions of these ellipsoids is usually
estimated on the basis of global segment dimensions like width and length. In
this study for the visualization of the dummy an accurate graphical surface
description of the dummy segments was required in addition to an ellipsoid
description for contact interactions. It was decided to integrate the
generation of both surface descriptions in such a way that the ellipsoid
description is derived from the detailed graphical surface description.
The first step is the measurement of the position of a large number of points
on the segment surfaces. The x, y, z co-ordinates of these points were
digitized with an accuracy of 2-3 mm. In order to obtain a complete
description of the segment surfaces the dummy segments had to be measured in
several positions. For each position also the co-ordinates of a number of
landmarks (at least three), specifying the segment local co-ordinate System,
were digitized. In this way a complete surface description could be obtained
by combining the measurement results in the various positions using coordinate
transformations. The digitized points were further processed using a CAD/CAM
System
resulting
in
a
surface
description
according
to
the
VDA-Flächenschnittstelleformat (VDAFS Version 2.0). Fig. 3 illustrates the
results of this mathematical description.
15
Fig. 3 A VDAFS representation of the HYBRID III dummy
16
The specification of the segment ellipsoids was also performed on the CAD/CAM
system. The Parameters describing the ellipsoids are the ellipsoid axes, the
location of the ellipsoid centre relative to the segment co-ordinate system
and the orientation relative to this co-ordinate system [4]. These parameters
were optimized visually by simultaneous graphical presentation from different
view points, of the ellipsoid and the detailed surface description. In this
procedure also the potential contact function of the ellipsoid was taken into
account. In other words potential contact areas of the dummy were approximated
more accurately than areas of the dummy surface which usually do not have
interactions with other Segments or with the environment. Fig. 2 illustrates
the resulting ellipsoids.
INERTIAL PROPERTIES
The mass and moments of inertia measurements are carried out at Wright
Patterson Air Force Base (WPAFB). The following properties are measured for
each dummy segment:
mass,
location of centre of gravity in local co-ordinate system,
principal moments of inertia at the segment centre of gravity,
orientation of the principal axes co-ordinate system.
A detailed description of the measurement methodology can be found in Refs.
[1,4],
The following accuracies could be obtained, except for the smaller elements
like hands, feet and neck:
- weight: 0.02 N
centre of gravity: 0.004 m
moments of inertia: IX
principal axes orientation: 3 degrees.
The mass distribution measurements also have been conducted for the
subsegments except for the soft, flexible structures (i.e. abdominal insert
and ribcage-sternum). For these flexible Segments only the mass values have
been determined.
JOINT PROPERTIES
The stiffness of the connection between the different Segments is one of the
Parameters having a major effect on the movement and position of the dummy
Segments in a crash environment. In order to specify the Joint properties two
concepts will be introduced here first:
Degrees of freedom of a Joint, defined as the number of independent Joint
motions possible in a Joint.
Range of motion, defined as the total motion possible for a degree of
freedom. In most cases the ränge of motion will be dependent on the
external load applied on the Joint.
Free ränge of motion, defined as the ränge of motion if only a small load
is acting on the Joint to compensate for friction and the effect of
gravity.
In the Hybrid III dummy, according to above definitions, the lumbar spine,
neck and the two clavicle joints do not posses any free ränge of motion.
Static and dynamic Joint properties have been determined in this study with
existing or Special developed methods. In these tests the Joint ränge of
17
motion is determined as function of the externally applied load. For joints
(or Joint motions) were these test methods could not be applied only the free
ränge of motion has been determined. This measurement was done manually using
an inclinometer.
Most of the pin- and ball- and socket joints in the dummy vere measured with
a Special developed apparatus, illustrated in Fig. 4. The device consists of
a horizontal non-moving platform and a rotating unit povered by a hydraulic
device (1). This rotating unit can rotate about a vertical rotation axis
oriented perpendicular to the platform. This axis is oriented vertically in
order to minimize gravity effects.
For a Joint under investigation only the two adjoining Segments are used in
this measurement. One segment is rigidly mounted in such a way that the
rotation axis of the measuring device is aligned with the Joint centre (or
Joint axis) of the Joint.
1
2
3
4
=
=
=
=
hydraulic motor
torque transducer
potentiometer
lever
Fig. 4 Test set-up for the static Joint properties.
The other Joint segment is moved by a lever(4) parallel to the rotation axis
of the measuring device and connected to the rotating unit. The torque applied
by the rotating unit is measured with a torque transducer(2), while the
rotation angle is measured with a rotational potentiometer(3). The guidance
of the rotating unit is such that the friction is very small.
The apparatus is manually controlled and the rotational motion of the rotation
unit stops if the applied torque is close to a specified value (first Joint
stop). Then the motion is reversed until the second Joint stop is detected.
Finally the rotation is reversed until the initial starting position is
reached again. A typical result of this test is shown in Fig. 5 for the
flexion-extension motion in the shoulder.
18
_
600T
7
500-
•-
400-
El * U n i I op
300-
500
100-
0-100-200-300-400
-500
Flfilon
-600
-250 -200 -150 -100
50
-50
Rngl«
(dig )
Fig. 5 Static shoulder characteristics: flexion-extension.
For the neck and spine both static bending and torsion tests are carried out.
The loads in these tests were applied manually. Neck and spine rotations and
displacements as well as external loads have been recorded. Tests have been
conducted in the following directions:
flexion - extension
oblique flexion - extension
lateral flexion
torsion
(forward/backvard)
(45° forward/backward)
(sideward right/left)
Fig. 6 illustrates the spine frontal bending test.
Fig. 6 Set-up of spine bending test.
19
00*12*
A. Rigid half sphere
radius 76 mm.
B. Rigid cylinder
radius 76 mm.
D. Rigid flat Square
300 x 300 mm.
E. Seat belt 14.5 %
strain.
C. Rigid flat disk
radius 76 mm.
Fig. 7 Impactor faces used in the static compliance tests.
The number of loading locations is dependent on the segment to be tested. For
the head, for instance, seven different locations have been selected;
Dependent on the segment one or more impactor faces were used. In total more
than 60 tests have been conducted.
The dynamic force-deflection characteristics are measured with a spring loaded
guided impactor.Tests are conducted with two impactor faces i.e. face A (a
rigid half sphere) and face C (a flat rigid disk) as presented in Fig. 7. The
impactor mass is 6.875 kg (face A) and 9.15 kg (face B ) . Impactor velocities
were varied between 1 and 5.5 m/s. No higher velocities or impactor masses
were selected in order to avoid damage of the tested dummy Segments. The
impactor face is provided with an uni-axial accelerometer. The impactor
displacement is measured with a linear potentiometer.
For most of the segment locations for which dynamic measurements were carried
out, also static test results were available. This allows a direct comparison
between the dynamic and static compliance of a dummy surface. Fig. 8 shows
typical results for tests on the upper arm. In these tests face A was used.
A large difference can be observed here between static and the (high velocity)
dynamic response which could be explained by the dynamical behaviour of the
metal insert in the arm. A similar difference between static and dynamic
response was observed in the other limbs.
20
4.0}.(.
3.0
l.l
2.01.6
1.0
0.6
0.0
, ,['- • • SI.IK
• /'
'V. I -/t
V. 2 m/%
V. 4 ./.
40
SO
Fig. 8 Comparison between static and dynamic force-deflection response of
the upper arm tested with face A.
DISCUSSIONS AND CONCLUSIONS
This paper presents a measurement program of a sitting Hybrid III 50th
percentile dummy to determine a database of this dummy for Computer
simulations. A description of the measuring methodology is given.
Dimensional and inertial properties are measured using similar methodologies
as employed in earlier studies dealing with the subject of databases for
Computer simulations. New in comparison with earlier work. is the detailed
segment surface discretisation. As a result of these measurements a realistic
surface description of the individual segments becomes available which in
particular will be useful for Computer animation of the dummy model. As shown
in the present study this discretisation allows the selection of the contact
ellipsoid parameters to be carried out in a more accurate way. In future new
methods for contact calculations might become available in occupant Simulation
modeis which directly employ the surface descriptions in the contact
algori thms.
Joint resistive properties are measured using static and dynamic test set-ups.
For the static properties of joints with a Single centre of rotation a Special
Joint measuring device has been developed where the external torque is exerted
on the Joint segment by a hydraulic powered unit. The external applied torque
as function of the Joint rotation is automatically recorded in this test
set-up, allowing a fast and accurate registration of the static Joint
properties. The advantage of this method is that a relative large torque can
be applied in a very accurate way.
In our study a pendulum set-up is used to determine dynamic properties of the
neck. In order to determine the dynamic properties of the other joints a
Special pendulum set-up was constructed. This set-up needs further
improvements due to the resonances.
Special test set-ups are used in this study to measure static and dynamic
segment surface compliance properties. These methods allow a continuous
registration of the force-deflection properties. Different compression faces
were used including a set-up with a belt system. In addition to the surface
compliancies also the thorax stiffness at different locations is measured. For
the limbs dynamic tests appear to be less reliable due to the dynamic effect
of metal inserts in the limbs.
21
This study relates to the measurement of one dummy i.e. a 5Oth percentile
Hybrid III sitting dummy. Although the present measurement methodology can be
improved further it is believed that in general the present technologies are
adequate to determine Computer Simulation databases for other dummies like
the present side impact dummies (EUROSID, SID and BIOSID), different sized
frontal crash dummies (5th and 95th percentile Hybrid III) and various sized
child dummies.
REFERENCES
1. I. Kaleps, R.P. White, R. Beecher, J. Whitestone and L.A. Obergefell:
"Measurement of Hybrid III Dummy Properties and Analytical Simulation Data
Base Development", draft report; Harry G. Armstrong Aerospace, Medical
Research Laboratory, Vright Patterson Air Force Base, Ohio, February 1988.
2. I. Kaleps, J. Whitestone: "Hybrid III Geometrical and Inertial
Properties"; SAE 880638, International Congress and Exposition, Detroit,
Michigan.
3. J. Wismans and J.H.A. Hermans: "MADYMO 3D Simulations of Hybrid III Dummy
Sied Tests"; SAE 880645, International Congress and Exposition, Detroit,
1988.
4. M. Philippens, J. Wismans, and J.J. Nieboer: "50th Percentile Hybrid III
Database Development"; TNO report 751860026 prepared for FAT-AK5, UA
"Crash-Dummy", Frankfurt, Germany.
22
3.
An Advanced 50th Percentile Hybrid IIIDummy Database: Validation
(M. Heinz, B. Pletschen, H. Wester, T. Scharnhorst; FAT-AK5, UA
"Crash-Dummy", Frankfurt/Main)
SUMMARY
INTRODUCTION
The Integration and productive use of Computer simulation in the car development process requires
highly sophisticated programs and high quality databases, especially in the very sensitive area of safety. Therefore, measurements with crashdummys
have to be conducted and analyzed in detail.
In 1987, a subcommittee of the German Automobile
Technology Research Association FAT (Forschungs
Vereinigung Automobiltechnik) was founded to perform measurements of the mechanical and geometrical crash dummy properties in order to achieve
validated databases for numerical validation. The
project budget was financed by the FAT. The idea of
Company overlapping database development is
based on considerations such as:
The project reported here, was supervised by the
German Automobile Technology Research Association (Forschungsvereinigung Automobiltechnik)
and carried out by a team recruited from various
German car manufacturers. Based on the outcomes
of the measurements of mechanical and geometrical
crash dummy properties, a database for the dynamic
rigid body Simulation program MADYM0-3D has
been developed.
The Simulation results proposed for this new database are compared with the experimental results
from sied tests conducted by Volkswagen. The dummy used as an occupant for the sied tests is the same
as the dummy used for the measurements of its
physical properties. The present study deals with ths
questions and problems occurring during the validation process of the 50th percentile HYBRID IIIDummy Database. Assumptions, approximations
and changes made in order to optimize the dataset
are documented.
'
The quality of the actual dataset is confirmed by the
correspondence of the test and Simulation results
and also compares favorably with the earlier states
of the dataset evaluation.
- The worldwide conformity in crash dummy Standards
- The saving of time and money
- A fixed and common data and program basis
The working group decided to Start the program
with the 50th percentile HYBRID IH-dummy. This
dummy is considered to be one of the most advanced test dummies for frontal collision situations.
All simulations were carried out with the MADYM0-3D occupant Simulation program, release 4.2.
This program is under continued development and
available for any partner participating in the project.
PROJECT ORGANIZATION
The development of a dummy database for a rigid
body Simulation model shall have two separate
parts. The first section includes the measurements of
the physical properties of a dummy. In the second
one the created dummy dataset is validated. Therefore the activities can be distributed as follows:
23
Topics phase 1:
- Dummy segment and subsegment selection, specification qnd roordinate System definition
- Geometry and suriace descretisation of all segments and the assembled dummy in VDAFS-format [2]
- Segment dimensions and Joint locations
- Masses, moments of inertia and center of gravity
location of all segments
- Static and dynamic Joint properties
- Static and dynamic surface compliance tests at
specified locations of all segments
- Data transformation and database development
Topics phase 2:
- Analyses and discussions of the results of phase 1
- Sied tests at different acceleration levels
- Calculations with the preliminary database
- Comparison between sied test and calculation results
- Analyses of error sources
- Evaluation of a final database
Because of its experience in Crash and safety matters and its responsibility for the future MADYMO
program development, the working group charges
TNO in Delft, Netherlands, to perform the project
phase 1. Phase 1 is reported separately and in detail
[3].
DATASET EVALUATION
The results of phase 1 lead to a completely new
dataset, which is totally different from the already
existing 5Oth percentile dummy data [4,5]. The
dummy of phase 1 consists of 20 segments (Fig. 1),
due to separate shoulder elements, separate hand elements and an additional segment representing the
deformable sternum element of the dummy. By adding these model elements, which are most important
for the belt- and airbag-dummy interaction, an advanced and more realistic Simulation model was
created. [6].
Great attention has to be given to the modelling of
the neck and lumbar spine joints which consist of
continuously deformable rubber materials strengthened by steel elements and equipped with asymmetric drillings.In the Simulation model the deformable
rubber joints have to be replaced by simple pin or
ball and socket joints, representing the characteristics of the real parts as closely as possible. There-
24
fore, based on the various component tests the neck
model was verified and prevalidated by means of a
separate Simulation model.
DATASET VALIDATION
Validation tests - To get a reliable validation basis
for the database quality checks, the test configuration has to be defined in detail. The test results are
compared with the Simulation results of the developed database. Therefore the passenger compartment has to be precisely defined using only a few
well defined parameters. Then follows the validation of the dummy database. However there is no
validation of the passenger compartment or the restraint System included here.
The first proposal for the test conditions was verified by pretests to ensure reliability and reproduction. The following test setup finally was chosen:
- 50th percentile HYBRID III - dummy
- Front passenger position
- Wooden seat according to ECE R 16
- Rectangular impact pulse according to ECE R 16
(Fig. 2)
- Zero degree frontal impact configuration
- Two separate static belt Systems for lap and shoulder belts (Fig. 3)
The test configuration is documented in side view
(Fig. 4) and rear view (Fig. 5). The coordinates of
seat, belts and dummy position were measured and
documented aswell.
The sied tests were conducted at Volkswagen completely independent from the database development.
The dummy used in the tests was exactly the same
as the one measured before at TNO. Thereby any
statistical scatter of the dummy characteristics could
be eliminated. The tests were conducted at three different velocity and acceleration levels, with three
tests at each level. Evaluating the sied test results
(e.g. Fig. 6 to 8), a very small scatter ränge was noticed, which is an indicator for the excellent test reliability.
The following data were made available for TNO's
database development:
- Geometry and coordinates of seat, belts and dummy position
- Sied test acceleration pulses
- Initial belt length
- Force-dcflection characteristics of the bclt material
ANALYSIS OF ERROR SOURCES
With Ulio complete dataset, i.e. dummy and test environment, TNO simulates and predicts the dummy
behaviour without any knowledge of test results.
In the following documentation on!y the x-acceleration computed at the ehest element is documented
representing all the changes and effects of data
modification.
Calculations with the Preliminary Database
Phase 1 - Fig.6 to 8 show the good correspondence
of the sied and Simulation results in the time ränge
up to 80 ms.There is similar correspondence for all
other velocity and acceleration levels. Above 80 ms
the calculated head and ehest accelerations differ
significantly from the test results. The calculations
show a second peak almost as high as the first one,
which cannot be seen in the test results. This second
peak oecurs at all acceleration levels and increases
with the acceleration level. The calculation results
indicate that these second peaks are caused by oscillations established at various element accelerations
(e.g. Fig. 9). Detailed discussions between the
working group members and the TNO projeet team
took place in order to find out the reasons for the
differences; alrcady at this early State of Simulation
the general quality and reliability of the database
was undisputed.
Deformable Chest Element - Chest oscillations induced by vibrations of the separate stemum element
and transferred by the point-restraint elements are
not in question here. As shown in Fig. 10 even a
doubling of the damping coefficients of the point-restraint elements did not influence the results in any
way. Varying the masses of the deformable stemum
elements between 1.1 and 3.1 kg (with constant total
mass) no mass sensitivity depending on different
mass distributions between the chest and the deformable stemum element can be established (Fig.
11). The plastic deformation characteristic, modelled for the point-restraints connecting stemum and
chest, is unrealistic. Therefore, the working group
recommends to remodel these elements with an
elastic characteristic (Fig. 12). The calculated chest
defiection, which was much higher than the test value, could be corrected by increasing the spring stiffnesses in the point-restraint elements. A good
correlation between the calculated and the test chest
deformation values is shown in [7]. Generally no influence of the separate stemum element on the double peak effect could be remarked (compare with
Fig.6 to 8).
Sources of Error in the Preliminary Database
Phase 1 - The Simulation results based on the published dataset [4] did not show any double peak effects. Therefore, the double peak must have been
caused by the model innovations introduced in
phase 1. Additionally the following non-definite parameters infiuencing the model had to be investigated:
- Mass sensitivity and distribution of the deformable
ehest element
- Oscillation induced by ehest defiection
- Influence of the plastic ehest defiection characteristic
- Pelvis rotation and dislocation depending on the
belt attachment points
- Belt forces as a funetion of elastic shoulder and deformable ehest
- Oscillation of the separate arms
- Damping in cardan joints
All potential error sources were discussed and investigated by the working group.
Pelvis Kinematics - The pelvis kinematics (excessive rotation and horizontal translation) could be
improved by vertically repositioning the lap belt attachment points. The attachment points are now
0,02 m lower than in the preliminary dataset. This
improves the pelvis kinematics but has no influence
on chest oscillation (Fig. 13 and 14).
Changes in Joint Properties - As shown by the
analysis of high speed films the predicted horizontal
and vertical translations of the upper torso are much
too high. These differences could be reduced by
stiffer characteristics of the flexion torsion joints of
the spine and upper torso. These modifications induce an improved chest kinematic without infiuencing the double peak (Fig. 15). Therefore, the
working group deeided not to change the Joint stiffness especially since the measurement did not vindicate this.
25
Belt Attachment Points at Chest and Sternum The shoulder belt attachment points raised some
questions ;orr cming the connection of a System element represented by the chest and the deformable
sternum part. In principle the MADYMO belt model allows such modelling, correcting the deflection
of the deformable element by belt slip.
the Joint damping coefficients a very simple model
was evaluated: the arm System was separated and
treated as a one-mass System. For this system the
critical damping coefheient is easy to determine.
The damping coefficients used in the dataset are
lower than critical damping
shoulder
Alternative computations with a direct belt link between the car body and the deformable chest
showed an influence on double peaks (Fig. 16).
Nevertheless, the model was not modified to preserve the requested influence of a non-fixed belt attachment point at the shoulder.
Model of the Dummy Arm - Each dummy arm is
built of four segments:
- Shoulder
- Upper arm
- Lower arm
-Hand
The segments are connected by cardan joints.
Damping coefficients were only defined for the
fixed Joint rotation directions. The kind of disrurbance which would produce oscillations of the arm
System had to be investigated. The vibrations could
only be induced by the shoulder arm or the shoulder
belt connections because no other active forces have
an impact on the arm System. The rotational ränge
of motion of the shoulder segment is very small due
to the stiff characteristic. As long as the force at the
shoulder is high enough to compensate the shoulder
momentum reaction, no influence can be noticed.
With decreasing belt forces (Fig. 17) the shoulder
reaction momentum induces a fast forward movement of the shoulder leading to an oscillation of the
whole arm System (Fig. 18 to 21). Due to the symmetric strueture of the dummy, this effect also causes vibrations in the opposite dummy arm (Fig. 18 to
21) and (at the same time) peaks oeeur in the rotational psi-directions of the elbows (Fig. 22). These
peaks cannot be caused by the oscillation of the arm
System since the rotation directions observed are the
same on the left and the right sides.
Damping Parameters - Damping coefficients were
included in the model only if they could be derived
directly from the measurements. As documented in
the chapters above the introduetion of damping has
a major influence on the oscillations. To determine
26
elbow
phi-rotation: ION s/m
psi-rotation: 50 N s/m
psi-rotation: 1 N s/m
Table 1: Additional damping coefficients of the
shoulder and elbow joints
The values mentioned here were not optimized. The
introduetion of damping in the shoulder joints
(Fig.23 to 24) already improved the chest oscillation. However, with the additional damping in the elbows, the chest Vibration problem was finally solved
(Fig. 25). In this context it should be noted that for
all joints with a small ränge of motion and stiff characteristics the introduetion of damping is indispensable. In other joints no damping was used.
Nevertheless, the working group members agreed
that future model improvements could be necessary,
especially for the Joint damping characteristics.
FINAL COMPARSION OF TEST AND SIMULATION
Dataset manipulations made in order to validate the
model should be minimized due to the reliability of
the measurements and the direct correspondence of
dataset and measurement results. Therefore, only the
z-coordinates of the lap belt attachment points at the
pelvis and the damping coefficients in the shoulder
and elbow were modified. The quality of the actual
dataset is confirmed by the correspondence of the
test and Simulation results and also compares
favourably with the earlier states of the dataset evaluation. Figs. 26 to 32 show the results mentioned before, i.e. the comparison of the 15-element dummy
(Status 1988 before the projeet Starts), the 20-element dummy (preliminary Status phase 1 derived
from measurements and recommended by TNO) and
the 20 -element dummy modified by the working
group (modified aecording to the working group
ideas).
PROJECT EVALUATION
The Company overlapping dataset development has
been conducted with great success. The determined
dataset is vcry convincing because of its prognosis
ability. Undoubtedly, dataset development at each
Company wculd have required morc work, more time
and more money. However, it has to be questioned
whether the same results would have been obtained.
The FAT working group conducting the project will
also be in Charge of the dataset in the future. The simplicity of the validation test configuration will probably make dataset adaptations necessary. These
adaptations will be discussed and harmonized in the
working group. A general question which has to be
discussed is the measurement expenditure. Based on
the experience made in this first project the efforts to
determine the Joint damping characteristics must be
increased.
Further dummy measuring projects are planned, e.g.
the development of equivalent datasets for 5th or
95th percentile HYBRID-EI dummies and the legally defined side impact US-Sid-dummy. For one year,
starting from the project final date, the presented HYBRID-III dummy database will be exclusively available only to FAT members as stipulated between FAT
and TNO. However Non-FAT member companies
also will have the Chance to acquire the data. Inquiries for further information should be addressed to
TNO. TNO will have the right to publish the validated dummy dataset one year after the project final
date.
ACKNOWLEDGEMENTS
The project was initiated and supervised by the FAT
Working Group 5 (biodynamics) Subcommittee
Crash Dummy. The following members belong to
this group:
Dipl.-Ing. H. Baldauf
Dipl.-Ing. D. Braun
Dipl.-Ing. M. Heinz
Dipl.-Ing. K. Klingbeil
Dr. B. Pletschen
Dr. T. Scharnhorst
Dr. R. Schöneburg ,
Dipl.-Ing. A. Velikic
Dipl.-Ing. H. Wester
Dr. P. Wienecke
the project by contributing important information
and suggestions.
REFERENCES
[1] I. Kaleps, R. P. White, R. Beecher, J. Whitestone
and L.A. Obergefell: "Measurement of Hybrid
EI Dummy Properties an Analytical Simulation
Data Base Development", Draft Report; Harry
G. Armstrong Aerospace Medical Research
Laboratory, Wright Patterson Air Force Base,
Ohio, USA, February 1988.
[2] VDA Working Group CAD/CAM: "VDA Surface Interface (VDAFS)", Version 2.0; Frankfurt/M,
D, 1986.
[3] M. Philippens, J. Wismans and J. J. Nieboer:
"50th Percentile HYBRID m Data Base Development ", Contractor: FAT AK 5 Working
Group Crash Dummy; Vol. 1+2, Delft, NL,
1989.
[4] I. Wismans and J. H. A. Hermans: "MADYMO
3D Simulations of Hybrid III Dummy Sied
Tests", SAE 880645.
[5] "MADYMO Data Bases", Version 4.2, The
Hague, Nl, 1988.
[6] M. Heinz, R. Hoefs: "Vehicle Occupant Crash
Simulation Using MADYMO Porsche Airbag
2D", SAE 890755.
[7] M. Heinz: "Simulation von Gurtsystemen in
Sport- und Rennfahrzeugen", VDI-Tagung 'Berechnung im Automobilbau', Würzburg, D,
1990.
BMW AG
Keiper Recaro GmbH& Co.
Porsche AG
Adam Opel AG
Mercedes Benz AG
Volkswagen AG
AUDI AG
Ford Werke AG
Volkswagen AG
FAT
For the discussion of the validation results thanks are
due to Dipl.-Ing. J. Nieboer, TNO, and to Dipl.-Ing.
H.-J. Petit from Mercedes-Benz AG, who assisted
27
Fig. 1: The 50th percentile Hybrid EI Dummy after Phase 1
12.
10.
oo
53 -100.
0
Time
Fig. 2: Example of the sied test acceleration and velocity
28
16
1210-
I .:
QJ
u
O
6-
42-
0
0
2
4
6
B
10
12
rel. alongsiton (>)
Fig. 3: Characteristic of the belt material
13
12
Fig. 4: Test configuration (side view)
Fig. 5: Test configuration (rear view)
29
600.
Fig. 6: Resultant head acceleration, Simulation
versus test results (15 g, 40 km/h)
-
C-J
500.
-
400.
-
300.
-
200.
-
Tests
Simulation
O
1
I
100. -
I
Time
600. -,
c
o
•a
500.
-
400.
-
300.
-
200.
-
100.
-
Tests
Simulation
2
o
<
0
Time
30
Fig. 7: Resultant ehest acceleration, Simulation
versus test results (15 g, 40 km/h)
Fig. 8: Resultant pelvis acceleration, Simulation
versus test results (15 g, 40 km/h)
Tests
Simulation
*
C/3
C
o
1
o
o
<t/3
•>
Time
£•
250.
250.
40 km/h 15 g
Double Damping Coefficiem
50 km/h 20 g
Original Result Phase 1
-250. "
500.
0.
0.05
Time
0.10
Fig. 9: Fonvard ehest acceleration
0.15
0.20
0.20
Fig. 10: Forward ehest acceleration, influence of
damping
31
250.
— - New Lap Bell Aiuchinem Poini
Original Result Phase 1
0.15
0.20
0.20
Fig. 11: Forward ehest acceleration, influence of the
mass of the deformable ehest element
«
pltstic ehest chtncleriiuc
elislic ehest chtnoeristie
lestnsult
0.048'
0.040-
Fig. 14: Forward ehest acceleration as a funetion of
the location of the lap belt attachment point
250.
Changed Joint Söffness
Original Result Phase I
0.0320.0240.016-
-250. -
0.0080.000
1
-0.008
0.
0.05
0.10
0.20
0.15
Time
-500.
0.
1
0.05
0.10
0.15
0.20
Time
Fig. 12: Chest defiection as a funetion of the ehest
modelling
500.
— - New Lap Bell Altüchmenl
Original Resull
I
[Poim
250. -
I
S
u
0.15
0.
0.20
Time
Fig. 13: Resultant pelvis acceleration, influence of
the location of the lap belt attachment point
32
Fig. 15: Forward chest acceleration as a funetion of
the rotational stiffness of the spine and upper torso joints
?
«
250.
— - New Shoulder Belt Aliachment Poini
Original Resull Phase 1
Fig. 16: Forward ehest acceleration as a function of
the location of the shoulder belt attachment
point at the sternum element
Fig. 18: Coordinate System of the left shoulder Joint,
including the rotation directions
6000.
E
>uu. -
A\ *11
/
150. -
1
/ \ •• A
/ \i: l
.9
/ \i\'•A
y
-»»-Ä-
0. -
\
i\
'1 11
i i|
M 1!
I M
1
11 1 . l
i«
v y
150. -
l-
1
Right Shoulder
2
Lefl Shoulder
- i
0.15
0.20
i
i
i
1
0.05
•
i
'
i
|
0.10
'
0.15
0.20
Time
Fig. 17: Shoulder belt force
Fig. 19: Momentum in the left and right shoulder
joints (phi-rotation)
33
E
150.
300.
s
Right Elbow
z
Left Elbow
150. -|
c
S.
B "130. -
I
-150.
-300.
-
0.
I
I
-
—-
Right Shoulder
0.05
Time
0.10
0.15
0.20
Left Shoulder
-300.
-1
0.
1
|
0.05
Time
r
I
0.10
r
0.15
0.20
Fig. 20: Momentum in the left and right shoulder
joints (theta-rotation)
Fig. 22: Momentum in the left and right elbow
joints (psi-rotation)
f
*
250.
Changed Daniping Coefficicnis
Original Resull Phase 1
450.
Righi Shoulder
Left Shoulder
I
300.
-
150.
-H
0.20
0.
Fig. 23: Forward ehest acceleration as a funetion of
damping in the shoulder Joint
0.
0.20
Fig. 21: Momentum in the left and right shoulder
joints (psi-rotation)
34
E
300.
?
Changed Damping Coeflicicnis
250.
-
Original Resull Phase 1
— — Changed Dumping Coefiiciems
Original Resull Phase 1
0.
-15U. 0.
I
0.05
0.10
0.20
0.15
Time
I
0.10
0.05
0.15
0.20
Time
Fig. 25: Forward chest acceleration including damping in shoulders and elbows
Fig. 24: Momentum in the shoulder Joint as a function of damping in the shoulder Joint (phirotation)
600. - i
CM
«
E
c
I
—
o
o
<
Simulation with
naa Validated Dataset
+++ Dataset Phase 1
xxx Dataset SAE 880645
500.
400.
-
300.
-
200.
-
100. -
0.
. ^
0.20
Time
Fig. 26: Resultant head acceleration, test versus Simulation results
35
600. -i
*
oo
500.
400.
c
o
1
QJ
O
Fig. 27: Resultant ehest acceleration, test
versus Simulation results
Simulation with
- •••Validated Dataset
+++ Dataset Phase 1
xxx Dataset SAE 880645
-
300. -i
200.
-
100.
-
<
U
l/D
0.15
0.20
Time
600. -1
500.
Test
Simulation with
- •am Validated Dataset
+++ Dataset Phase 1
xxx Dataset SAE 880645
400.
-
300.
-
200.
-
100.
-
*
oo
•2
8
oö
QJ
Fig. 28: Resultant pelvis acceleration, test
versus Simulation results
0.20
Time
36
Fig. 29: Shoulder belt force (attachment part),
test versus Simulation results
7000. 6000. -
Simulation with
Validated Dataset
+++ Dataset Phase 1
x x x Dataset SAE 880645
nnD
OJ
PQ
o
5o
0.05
0.10
0.15
0.20
Time
Fig. 30: Shoulder belt force (stemum buckle part),
test versus Simulation results
Test
Simulation with
aaa Validated Dataset
+++ Dataset Phase 1
x x x Dataset SAE 880645
0.
0.
Time
37
11200.9600. -
Fig. 31: Lap belt force (attachment part),
test versus Simulation results
Test
Simulation with
citua Validated Dataset
+++ Dataset Phase 1
XXX Dataset SAE 880645
8000. 6400. -
n
4800. 3200. L
a
1600.
-
J
0.
J
0.05
\
i
\
i
r~\
0.10
ii
i
1
i
0.15
i
0.20
Time
11200.
9600.
Test
Simulation with
CDQQ Validated Dataset
+++ Dataset Phase 1
xxx Dataset SAE 880645
Fig. 32: Lap belt force (pelvis buckle part),
test versus Simulation results
8000. 6400. 4800. -
0.15
Time
38
0.20
4. Ausblick
Mit diesem Forschungsvorhaben wurde ein Verfahren zur Vermessung der geometrischen
und physikalischen Eigenschaften von Crashdummies sowie ein validiertes 50 %-Hybrid III
Dummy-Modell zur Anwendung in Computersimulationen erarbeitet. Das Simulationsmodell
wird bereits in den einzelnen Häusern der Arbeitskreismitglieder erfolgreich angewendet. Die
entwickelte Vorgehensweise - modellspezifische Vermessung des Dummys und anschließende Erstellung eines Dummy-Datensatzes durch den Forschungsnehmer TNO sowie Validierung des Modells anhand von Vergleichen zwischen Ergebnissen eigens durchgeführter
Versuche und Simulationen - ist zielfindend und schafft ein Dummy-Modell hoher Qualität für
Frontalaufprallsimulationen.
Die Arbeit des FAT-Arbeitskreises 5 Unterausschuß "Crash-Dummy" wird weitergeführt mit
den derzeitigen Vermessungen und Modellentwicklungen für 2 weitere Frontalaufpralldummies (5 %-Hybrid III Frau, 95 %-Hybrid III Mann) sowie einen Seitenaufpralldummy (US-SID;
FMVSS214). Damit werden dann Computer-Modelle aller wichtigsten Dummies zur Simulation von Dummy-Verhalten und -Belastung beim Fahrzeugcrash verfügbar sein und können
bei der Automobilentwicklung verstärkt unterstützend genützt werden.
39
5. Literatur
(1) Philippens, M.; Wismans, J.; Nieboer, J. J.:
Hybrid III 50th Percentile Database Development Program: Phase 1
Draft Report, TNO, The Nethderlands (Febr. 1989)
(2) Philippens, M.; Wismans, J.; Nieboer, J. J.:
50th Percentile Database Development
Final Report, TNO-Report 751860026, Volume I,
TNO, The Netherlands (December 1989)
(3) Philipens, M.; Wismans, J.; Nieboer, J. J.:
50th Percentile Database Development
Final Report, TNO-Report 751860026, Volume II,
TNO, The Netherlands (December 1989)
(4) Philipens, M.; Nieboer, J. J.; Wismans, J.:
An Advanced Database of the 50th Percentile Hybrid IM
Proceedings of the Second International MADYMO
Users Meeting, p. 9 - 21, May 1990,
Noordwijk, The Netherlands
(5) Wester, H.; Heinz, M.; Pletschen, B.; Scharnhorst, T.:
An Advanced Database of the 50th Percentile Hybrid III: Validation
Proceedings of the Second International MADYMO
Users Meeting, p. 23 - 68, May 1990,
Noordwijk, The Netherlands
(6) Philipens, M.; Nieboer, J. J.; Wismans, J.:
An Advanced Database of the 50th Percentile Hybrid III Dummy
SAE-paper 910813, Detroit, USA (1991)
(7) Heinz, M.; Pletschen, B.; Wester, H.; Scharnhorst, T.:
An Advanced Database 50th Percentile Hybrid III Dummy Database
SAE-paper 910658, Detroit, USA (1991)
41
Anhang
Mitglieder des Unterausschusses "Crash-Dummy"
FAT-AK 5 "Biomechanik"
Dr. T. Scharnhorst
Volkswagen AG
Abt. E/FT
Postfach
3180 Wolfsburg 1
Dr. P. Wienecke
Forschungsvereinigung
Automobiltechnik
Postfach 17 05 63
Dipl.- Ing. Baldauf
BMW AG
ABT. EW-40
Postfach 40 02 40
Dipl.-Ing. A. Velikic
Ford Werke AG
Abt. MC/PK-15
Postfach 60 40 02
6000 Frankfurt 17
8000 München 40
5000 Köln 60
Dipl.-Ing. M. Heinz
Dr.-Ing.h.c.F. Porsche AG
Abt. EFF2
Postfach 11 40
Dipl.-Ing. D. Braun
Keiper Recaro GmbH & Co.
Hertelsbrunner Ring 2
Dipl.-Ing. Klingbeil
Adam Opel AG
VE-Fahrzeugberechnung
Postfach 15 60
6750 Kaiserslautern
7251 Weissach
6090 Rüsselsheim
Dipl.-Ing. H. Wester
Volkswagen AG
E/FT-BE
Postfach
Dr. R. Schöneburg
AUDI AG
Abt. I/ERS
Postfach 2 20
Dr.-Ing. B. Pletschen
Mercedes-Benz AG
Abt. EP/ADBT
Postfach 226
3180 Wolfsburg
8070 Ingolstadt
7032 Sindelfingen
Dr.-Ing. F. Zeidler
Mercedes-Benz AG
Abt. EP/ADUS
Postfach 2 26
7032 Sindelfingen
43
Bisher in der FAT-Sch ritten reihe erschienen:
Nr. 1
Nr.
Nr.
Nr.
Nr.
Nr
Nr.
Nr.
Nr.
Nr
Nr
Nr.
Nr
2
3
4
5
6
7
8
9
10
11
12
13
Nr. 14
Nr
Nr.
Nr.
Nr
Nr.
15
16
17
18
19
Nr 20
Nr. 21
Nr 22
Nr 23
Nr 24
Nr 25
Nr.
Nr.
Nr
Nr.
Nr.
Nr.
26
27
28
29
30
31
Nr.
Nr.
Nr.
Nr.
Nr.
32
33
34
35
36
Nr 37
Nr. 38
Nr. 39
Nr. 40
Nr. 41
Nr 42
Nr 43
Nr 44
Nr. 45
Nr. 46
Nr. 47
Nr. 48
Nr. 49
Nr. 50
Nr. 51
Nr. 52
Nr. 53
Nr. 54
Nr. 55
Nr. 56
Nr. 57
Nr. 58
Nr. 59
Nr. 60
Nr. 61
Nr. 62
Nr. 63
Nr. 64
Nr. 65
Nr. 66
Nr. 67
Nr. 68
Nr. 69
Nr. 70
Nr. 71
Immissionssituation durch den Kraftverkehr in der Bundesrepublik Deutschland
Systematik der vorgeschlagenen Verkehrslenkungssysteme
Literaturstudie über die Beanspruchung der Fahrbahn durch schwere Kraftfahrzeuge
Unfallforschung/Westeuropäische Forschungsprogramme und ihre Ergebnisse/Eine Übersicht
Nutzen/Kosten-Untersuchungen von Verkehrssicherheitsmaßnahmen
Belastbarkeitsgrenze und Verletzungsmechanik des angegurteten Fahrzeuginsassen
Biomechanik des Fußgängerunfalls
Der Mensch als Fahrzeugführer
Güterfernverkehr auf Bundesautobahnen
Recyclfng im Automobilbau - Literaturstudie
Rückführung und Substitution von Kupfer im Kraftfahrzeugbereich
Der Mensch als Fahrzeugführer
Sicherheitsmaßnahmen im Straßenverkehr
Sammlung, Beschreibung und Auswahl für die Anwendung der Nutzen/Kosten-Analyse
Tierexperimentelle und epidemiologische Untersuchungen zur biologischen Wirkung von Abgasen
aus Verbrennungsmotoren (Otto- und Dieselmotoren) - Literaturstudie
Belastbarkeitsgrenzen des angegurteten Fahrzeuginsassen bei der Frontalkollision
Güterfernverkehr auf Bundesautobahnen - Ein Systemmodell, 2. Teil
Ladezustandsanzeiger für Akkumulatoren
Emission, Immission und Wirkung von Kraftfahrzeugabgasen
Sicherheitsmaßnahmen im Straßenverkehr
Ergebnisse einer Nutzen/Kosten-Analyse von ausgewählten Maßnahmen
Aluminiumverwendung im Automobilbau und Recycling
Fahrbahnbeanspruchung und Fahrsicherheit ungelenkter Dreiachsaggregate in engen Kurven
Umskalierung von Verletzungsdaten nach AIS - 80 (Anhang zu Schrift Nr. 15}
Grundlagen und Möglichkeiten der Nutzung sprachlicher Informationssysteme im Kraftfahrzeug
Altteileverwendung im Automobilbau
Energie für den Verkehr - Eine systemanalytische Untersuchung der langfristigen Perspektiven
des Verkehrssektors in der Bundesrepublik Deutschland und dessen Versorgung mit Kraftstoffen
im energiewirtschaftlichen Wettbewerb Wirtschaftlichkeit des Einsatzes von Aluminium im Lkw-Bau
Äußere Sicherheit von Lkws und Anhängern
Dämpfung und Tilgung von Torsionsschwingungen im Triebstrang von Kraftfahrzeugen
Wirkungsgradmessung an Getrieben und Getriebeelementen
Fahrverhalten von Lastzügen und hierbei insbesondere von Anhängern
Entwicklung, Aufbau und Test eines Ladezustandsanzeigegerätes für Bleiakkumulatoren
in Elektrostraßenfahrzeugen
Rollwiderstand und Lenkwilligkeit von Mehrachsanhängern mit Zwillings- und Einzelbereifung
Fußgängerschutz am Pkw - Ergebnisse mathematischer Simulation Verfahren zur Analyse von Unfallursachen - Definitionen, Erfassung und Bewertung von Datenquellen Untersuchungen über kraftstoffsparende Investitionsmaßnahmen im Straßenbau
Belastbarkeitsgrenzen und Verletzungsmechanik der angegurteten Fahrzeuginsassen beim Seitenaufprall.
Phase I: Kinematik und Belastungen im Vergleich Dummy/Leiche
Konstruktive Einflüsse auf das Fahrverhalten von Lastzügen
Studie über Energieeinsparungsgeräte zur Mitführung im Kraftfahrzeug (Bordlader)
Grundlagen und Möglichkeiten der Nutzung sprachlicher Informationssysteme im Kraftfahrzeug
- Hauptstudie Sprachausgaben im Kraftfahrzeug - Ein Handbuch für Anwender Auswertung von Forschungsberichten über:
Die Auswirkung der Nutzfahrzeugkonstruktion auf die Straßenbeanspruchung
Fußgängersicherheit - Ergebnisse eines Symposiums über konstruktive Maßnahmen am Auto Auswirkungen der Nutzfahrzeugkonstruktion auf die Straßenbeanspruchung - Gesamtbericht Sprachliche Informationssysteme und Anwendungsmöglichkeiten im Kraftfahrzeug
- Ergebnisse eines Symposiums Abgasemissions- und Kraftstoffverbrauchsprognosen für den Pkw-Verkehr in der Bundesrepublik
Deutschland im Zeitraum von 1970 bis 2000 auf der Basis verschiedener Grenzwertsituationen
Bewertung von Personenverkehrssystemen - Systemanalytische Untersuchungen von Angebotsund Nachfrageelementen einschließlich ihrer Wechselwirkungen Nutzen/Kosten-Analyse für einen Pkw-Frontunterfahrschutz an Nutzfahrzeugen
Radlastschwankungen und dynamische Seitenkräfte bei zwillingsbereiften Achsen
Studie über die Wirtschaftlichkeit von Verbundwerkstoffen mit Aluminiummatrix im Nutzfahrzeugbau
Rechnerische Simulation des dynamischen Verhaltens von nicht stationär betriebenen Antrieben
und Antriebselementen
Simulationsmodell - Schwingungsprogramm zur Ermittlung der Beanspruchung von Antriebssträngen Verwendung von Kunststoff im Automobil und Wiederverwertungsmöglichkeiten
Entwicklung eines hochgenauen, normfähigen Verfahrens zur Wirkungsgradmessung an Antriebselementen
Erhebung und Auswertung von Straßenverkehrsunfalldaten in der Bundesrepublik Deutschland Ergebnisse eines VDA/FAT-Fachgesprächs
Untersuchungen zur subakuten und chronischen Wirkung von Ottomotorabgasen auf den Säugetierorganismus
Pilotzelle zur Steuerung von Batterien in Fahrzeugen mit Elektro- oder Elektro-Hybrid-Antrieb
Wirkungen von Automobilabgas und seiner Inhaltsstoffe auf Pflanzen - Literaturstudie Rekonstruktionen von fünf realen Seitenkollisions-Unfällen - Ergänzende Auswertung der KOB-Daten Luftqualität in Fahrgasträumen
Belastbarkeitsgrenzen und Verletzungsmechanik des angegurteten Pkw-Insassen beim Seitenaufprall
Phase II: Ansätze für Verletzungsprädiktionen
Erhebung und Analyse von Pkw-Fahrleistungsdaten mit Hilfe eines mobilen Datenerfassungssystems
- Methodische und meßtechnische Ansätze für eine Pilotstudie Technische Erfahrungen und Entwicklungsmöglichkeiten bei Sicherheitsgurten im Fond von Pkw
- Ergebnisse eines Symposiums Untersuchungen über Wirkungen von Automobilabgas auf pflanzliche Bioindikatoren
im Umfeld einer verkehrsreichen Straße in einem Waldschadensgebiet
Sicherheitsorientierte Bewertung von Anzeige- und Bedienungselementen in Kraftfahrzeugen - Grundlagen Quantifizierung der Radlastdynamik bei Einfach-, Doppel- und Dreifachachsen in Abhängigkeit
vom Federungs- und Dämpfungssystem des Fahrzeugs
Seitenverkleidung am Lkw - Technische Analyse
Vorstudie für die Durchführung von Tracermessungen zur Bestimmung von Immissionskonzentrationen
durch Automobilabgase
Untersuchung fahrdynamischer Eigenschaften kurzgekuppelter Lastzüge bei Kursänderungen
Abschlußbericht der Pilotstudie zum Fahrleistungspanel „Autofahren in Deutschland"
Herstellung und Analyse charakteristischer Abgaskondensate von Verbrennungsmotoren
für die Untersuchung ihrer biologischen Wirkung bei nichtinhalativen Tests
Bewertung von Personenverkehrssystemen Teil II: Auswirkungen aus Angebots- und Nachfrageänderungen im Personenverkehr
vergriffen
DM 20,DM 30,vergriffen
DM 60,DM 50,DM 30,ver( jriffen
DM 50,DM 50,DM 50,DM 50,DM
60,-
DM
DM
DM
DM
DM
60,50,-
50,50,30,-
vergriffen
DM 50,DM 50,DM 50,DM 50,DM 50,DM
DM
DM
DM
DM
DM
60,50,60,50,50,50,-
DM
DM
DM
DM
DM
60,60,75,75,-
DM
DM
DM
60,50,30,-
DM
DM
60,25,-
DM
DM
DM
30,30,20,-
DM
30,-
DM
50,-
50-
vergriffen
DM 30,DM 40,DM 50,DM 250,DM 275,vergriffen
DM 160,DM
50,-
DM
DM
DM
DM
DM
75,40,30,35,50,-
DM
95,-
DM
35,-
DM
60,-
95,vergriffen
DM
DM
DM
30,50,-
DM
DM
DM
30,85,85,-
DM
55,-
DM
65,-
Bisher in der FAT-Schriftenreihe erschienen:
Untersuchung über das Emissionsverhalten der Leichtmüllfraktion aus Autoshredderanlagen
beim Verbrennen
Verletzungsfolgekosten nach Straßenverkehrsunfällen
Sicherheitsonentierte Bewertung von Anzeige- und Bedienelementen in Kraftfahrzeugen
- Empirische Ergebnisse Retrospektive Untersuchung über die innere Sicherheit von Lkw-Fahrerhäusern
Aufbau und Labortest eines wartungsarmen, sich selbst überwachenden Batterieaggregates
für Straßenfahrzeuge mit Elektro- und Elektro-Hybrid-Antrieb - Vorbereitende Untersuchungen Belastungsgrenze und Verletzungsmechanik des angegurteten Pkw-Insassen beim 90 -Seitenaufprall Phase III: Vertiefende Analyse der überarbeiteten und zum Teil neu berechneten HeidelbergerSeitenauf prall-Daten
Ermittlung von ertragbaren Schnittkräften für die betriebsfeste Bemessung von Punktschweißverbindungen
im Automobilbau
Verhalten des EUROSID beim 90 -Seitenaufprall im Vergleich zu PMTO sowie US-SID, HYBRID II und APROD
Demontagefreundliche Gestaltung von Automobilien - Teil I
Grundlagenuntersuchung zum Einfluß der Sonneneinstrahlung auf die thermische Behaglichkeit
in Kraftfahrzeugen
Einsatz von Retardern in der Betriebsbremsaniage von Nutzfahrzeugen - Zweiachsiges Fahrzeug Zwei Bände
Belastungen und Verhalten des EUROSID bei unterschiedlichen Prüfverfahren zum Seitenaufprall
Kosten einer kontinuierlichen Pkw-Fahrleistungserhebung
Auswirkungen der Nutzfahrzeugkonstruktion auf die Straßenbeanspruchung
Seitenkräfte an Mehrfachachsen von Sattelanhängern bei Kurvenfahrt und durch Spurrinnen
Verfahren zur Umwandlung polymerer Misch abfalle aus der Autositz-Produktion in Polyole
Methoden zur Vorausberechnung der Faserorientierung beim Pressen von SMC mit geschnittenen Glasfasern
Teil I: Unverrippte Bauelemente
Teil II: Verrippte Bauelemente
Fahrzeugerprobung eines wartungsarmen Batterieaggregates
Grundsatzuntersuchungen zum Festigkeitsverhalten von Durchsetzfügeverbindungen aus Stahl
Fahrverhalten von Lkw mit Zentralachsanhängern
Der Fahrer als adaptiver Regler
Einfluß realer Betriebsverhältnisse auf die Reproduzierbarkeit von Wirkungsgradbestimmungen an nicht
stationär betriebenen Getrieben
Mobilität - Automobil - Energiebedarf
Rationalisierungspotentiale im Straßenverkehr I
Abschlußbericht „Einsatz von Retardern in der Betriebsbremsaniage von zweigliedrigen Lastzügen"
Vermessung des 50 °'o-Hybrid III Dummy zur Ermittlung eines verbesserten Datensatzes für Crashsimulationen
vergriffen
DM 95,vergriffen
DM 90,-