www.ATZonline.

com

WORLDWIDE

03
March 2015 | Volume 117

C OVER STORY

Lightweight Design
between Performance and Costs
SAFE MANAGEMENT SYSTEMS INTERIOR CONCEPT KNOWLEDGE BASE
for Lithium-ion Batteries with Light Centre Console for Valid Driving Simulation
C ONTENTS

DE V E L OP MEN T

VEHICLE ELECTRICAL SYSTEM

16 Realisation of Safe Management
Systems for Lithium-ion Batteries
Marc Nalbach, Christian Wagner [Hella]

C OVER STORY SIMULATION AND TEST

20 Potential of Online Model Identification

Lightweight Design for Vehicle Dynamics Controls of

Load Sensitive Lightweight Vehicles
Florian Kohlhuber, Elias Schneider, Markus Lienkamp

between Performance
[Technische Universität München]

INTERIOR

and Costs
24 Interior Concept with Light
Centre Console
Isabella Schmiedel, Manfred Winklbauer,
Lion Pfeufer [Dräxlmaier]

The ambitious targets for reducing fuel consump- SOFTWARE

28 Reduction of Total Cost of Ownership
tion and emissions are making lightweight design
by Use of Electric Vehicles
increasingly important. Although new lightweight Achim Kampker, Christoph Deutskens,

materials offer better performance, they are often Philip Müller [RWTH Aachen], Thomas C. M. Müller
[Forschungszentrum Jülich]
more expensive. Designers are also spoilt for choice
when faced with up to 30 different grades of steel, MATERIALS
32 Lignin Powder as a Filler for Thermo-
complex composites or polymers which can be
plastic Lightweight Design Components
used to exploit lightweight potentials in bodies, Hendrik Mainka [Volkswagen], Maria Laue,

roofs and B-pillars. Chokri Cherif [TU Dresden], Axel S. Herrmann

[Faserinstitut Bremen]

R E S E A RCH

37 Peer Review

SIMULATION
38 Simulating Reliability with Respect
to Ride Comfort
Katrin Mutter [Daimler and University of Stuttgart],
Axel Röder, Jörg U. Brünis [Daimler],
Bernd Bertsche [University of Stuttgart]

46 Knowledge Base for Valid

Driving Simulation
Ilka Zöller, Nicole Mautes, Wen Ren,
Bettina Abendroth [TU Darmstadt]

4 Economically Viable Lightweight Design

Concept for a Hybrid B-pillar
Thomas Muhr, Johannes Weber [Mubea], Andreas Theobald, RUBRIC S | S ER V ICE
Martin Hillebrecht [Edag]

3 Editorial
36 Imprint, Scientific Advisory Board
10 Trends in Lightweight Construction
of Roofs
Matthias Arleth [Webasto]
COVER FIGURE © Edag
FIGURE ABOVE © ThyssenKrupp

2
 EDITORIAL

Material Concepts
Dear Reader, Recent innovations in components and
materials are evidence of significant pro-
Emission levels of less than 90 g of gress in aspects of energy and weight
CO2 per km are hardly possible without efficiency, safety, functionality, design
electrification, alternative fuels or and comfort. This topic is an ideal com-
lightweight design. The latest Shell plement to the latest issue of our maga-
Passenger Car Scenarios forecast a fleet zine lightweight design (lwd), which, in
of 10.1 million electric vehicles by 2040, the first edition of 2015, turns its atten-
while the number of petrol- or diesel- tion to chassis components, and shows
powered vehicles will fall to 30.7 million. how ZF combines material substitution
McKinsey expects that only around 5 % and lightweight design in a wheel-guid-
of vehicles will have an internal combus- ing transverse spring made of GRP.
tion engine in 2050. The 2011 EU White
Paper on transport policy predicts that Choosing the right material at the right
the use of conventionally fuelled cars in place is an ongoing challenge for devel-
urban transport will halve by 2030. opers and design engineers. It involves
enabling new material concepts to be
There are therefore no unanimous put into production from an economic
forecasts for the number of electric cars perspective, while at the same time also
in the future. Electric mobility is not addressing life cycle issues. The focus
predictable. Customers need to be con- remains on the efficient use of energy
vinced – both emotionally and econo- and therefore the sustainable protection
mically. At the same time, the need for of resources. In modern automotive
mobility is falling, and mobility is still engineering, the entire process chain
strongly determined by emotions. So must remain visible at all times.
what are customers looking for and what
price are they prepared to pay for an Best regards,
energy-optimised vehicle?

In addition to applying suitable measures

to the powertrain, lightweight design is
an effective means of reducing vehicle
weight and emissions. In automotive
engineering, there is huge potential not
only in the vehicle body but also in load- Dr. Alexander Heintzel, Editor in Chief
Wiesbaden, 22 January 2015
bearing or safety-relevant components,
especially if composite materials are
used. The key point, however, is the
cost-effectiveness of lightweight design
concepts, particularly when we need
to consider not only the additional costs
of lightweight construction of up to
5 Euros/kg but also the added cost of an
electrified powertrain, as well as the fact
that the use of different materials will
mean that repairs might become more
complex and therefore more expensive.

These questions were reason enough to

make the subject of lightweight design
the leading topic of this issue of ATZ.
03I2015 Volume 117 3
C OVER STORY LIGHT WEIGHT DESIGN

Economically Viable Lightweight

Design Concept for a Hybrid B-pillar
Edag and Mubea Carbo Tech were searching a best-in-class solution to manufacture hybrid B-pillars
economically. The result is a lightweight design concept that permits a weight reduction of approximately
20 %. The success factors are tailor rolled blanks, an automated fibre reinforced plastics manufacturing
process and load specific design.

AUTHORS

Dr.-Ing. Thomas Muhr Johannes Weber, B. Eng. Dipl.-Ing. (FH) Andreas Theobald Dr.-Ing. Martin Hillebrecht
is General Partner of the Mubea Group is Project Leader at the Mubea Carbo is Team Leader within the is Head of the Competence Center
in Attendorn (Germany). Tech GmbH in Salzburg (Austria). CAE at the Edag Engineering AG Lightweight Design, Materials
in Fulda (Germany). and Technologies at the Edag
Engineering AG in Fulda (Germany).

4
CONCEPT APPROACH degrees of complexity, cooperated with regarded as a critical load case, was to
Edag Engineering AG to develop a hybrid be of significant importance in this pro-
Car manufacturers are currently work- part for high-volume production, taking ject. To serve as a reference, a Mubea
ing under high pressure on the develop- its structural integration in the vehicle TRB B-pillar with standard crash perfor-
ment of new vehicles due to go into pro- body into account in the process. Also mance in side impact tests (IIHS rating
duction between 2015 and 2020. The involved in this cooperation was the good) was installed in the vehicle.
vehicle bodies are to weight considerably Mubea (Muhr und Bender, Attendorn) At the beginning of the concept devel-
less, provide high stiffness and fulfill Body Division for components made of opment, the main focus was on the ana-
ambitious crash load cases. Intelligent flexible tailor rolled blanks (TRB). lyse of tensile and compressive stress
lightweight design should make vehicles The objective of the cooperation distribution in the B-pillar for the load-
up to 100 kg lighter, depending on which was to demonstrate the feasibility and case IIHS. The result of the analyse
segment they belong to [1]. economic viability of a hybrid B-pillar showing the zones in which the use of
As far as weight, stiffness and eigen based on steel and fibre-reinforced FRP would be effective. Due to the load-
frequency are concerned, the B-pillar is plastic (FRP) and examine the technical specific alignment of the fibres, stresses
one of the most challenging vehicle body design. can be transferred most effective. On
components. Side impact requirements account of the deformation, the greatest
continue to be a constant challenge for tensile stress occurs on the B-pillar
CONCEPT ENGINEERING
vehicle manufacturers. In order to inner, and the greatest compressive
AND CAE DESIGN
develop an innovative solution, it was stress on the B-pillar reinforcement,
decided that the key aspect of the project In order to create realistic basic condi- FIGURE 1. Bearing in mind the fact that
would be the B-pillar. tions, the CAE model of a generic basic FRP has only very slight elastic and
Mubea Carbo Tech, a supplier of car- vehicle was defined first. It was decided almost no plastic properties, this result
bon fibre-reinforced structural and visi- that the IIHS side impact crash test, a US helped to identify the zones for the most
ble automotive components of varying consumer protection measure that is promising use of a FRP reinforcement.
03I2015 Volume 117 5
C OVER STORY LIGHT WEIGHT DESIGN

Various layouts of different FRP rein-

forcements with regard to topology, layer Stress in Z
structure and fibre alignment in interac- Zug
Pressure
tion with gradually optimised sheet thick- -500
ness profiles in the steel sheet metal outer -400
shell were subsequently examined and
-300
assessed by means of simulation in
numerous iteration loops. It was also nec- -200

essary to examine how the FRP reinforce- -100

ment would perform in peak stress areas. 0
Simulation indicated that an FRP hol-
100
low profile structually bonded with the B-pillar B-pillar
B-pillar above the hinge areas would give 200 inner reinforcement
the best results. Working on the sheet 300
thickness profile in the sheet metal com-
400
ponents, it was possible to deduct weight
from the upper area and channel the 500
Tension
weight into the lower section of the B-pil-
lar. By applying a very high fibre volume
content and optimum fibre orientation,
this made it ideal for use in the FRP rein-
forcement. A conscious decision was FIGURE 1 Distribution of tensile and compressive stress on the TRB B-pillar in the basic vehicle for
made to position the end trims outside of IIHS side impact load case
the highly deformed areas, to prevent any
failure of the steel plate components as a
result of stress peaks, FIGURE 2.
In the FEM, CAE tests were carried PRODUCTION OF PROTOT YPE The hybrid design first of all calls for a
out using a LS-Dyna solver. Energy FOR COMPONENT TESTING differentiation between the combined
absorption until start of damage was material groups. On the one hand, the
shown in the material card. Breaking Following concept development, detail- production of the sheet metal parts, and
behaviour (delamination in the mate- ing and simulation, the next challenge on the other, the FRP reinforcement in
rial) was not modelled. As a result the was to produce prototypes as close-to- the form of a hollow profile.
statements regarding the time of failure serialproduction as possible. The most The steel sheets were produced by the
tend to be conservative. In reality later important things here were to validate Mubea Tailor Rolled Blank GmbH in
failure and higher energy absorption the manufacturing process, collect new Attendorn. Two standard processes in
are expected. insights and validate the FEM analysis. the automotive industry were used in the

TRB steel B-pillar Hybrid B-pillar

(reference) (concept/prototype)
FRP reinforcement Tailor rolled blank

t = 2.0

t = 2.5
t = 1.1
t = 2.9

t = 2.0
t = 2.5 t = 2.2

t = 1.8
t = 2.0

t = 1.45 t = 1.45

FIGURE 2 Due to the FRP hollow profile adhesive bonded into the TRB B-pillar, the sheet thickness can be reduced to bring about a weight saving of 20 % if the
IIHS load case is fulfilled

6
 adjusted to the specified geometry. This
Preform by braiding TRB steel B-pillar and B-pillar inner
Braiding onto the core Tailor rolled blank, hot stamping
blown core can subsequently be de-
moulded with very little effort. To achieve
a suitable glass transition temperature, a
further tempering process was carried out
after the profile ends had been trimmed
by corundum blasting, before preparing
the part for gluing into the steel compo-
nent, FIGURE 3.
Next to this, the individual components
were adhesive bonded, and in addition
the steel components joined by conven-
tional spot welding. In order to avoid cor-
FRP reinforcement (RTM) Assembly rosion of the steel component the adhe-
Resin transfer moulding Mass 6.2 kg (vs 7.7 kg reference)
sive layer also helps as electrical insula-
tor, hence corrosion protection. For the
concept selected, this would also be the
delivery status, alternatives in the order
in which parts are joined are possible.

VALIDATION

Following production of the hybrid B-pil-

lar, a crash test should be carried out to
confirm the positive CAE results. As it
FIGURE 3 Production process chain from raw material to component; all processes were selected and was not possible to validate the complete
analysed to ensure their suitability for series production vehicle, a drop tower test was developed
instead. For the IIHS load case, this was
required to produce the same deforma-
tion pattern of the hybrid B-pillar as the
1 CAE model of the complete vehicle. In
close coordination with the test institute
conducting the test, the drop tower test
equipment was defined in the virtual
model. This ensured that it would be
2 possible to transfer the CAE settings to
3
the test equipment, FIGURE 4.
The test component moved transla-
tionally in the upper bearing, and rota-
4 tionally in the lower bearing. In order to
be able to roughly simulate the overall
FIGURE 4 Drop tower simulation in CAE: upper mounting point (2) can transfer the translation with a
defined stiffness (4); lower mounting point (3) can only transfer the rotation with a defined stiffness; vehicle performance, bearing stiffness
the impactor goes into free fall towards (1) on the test vehicle was also taken into account in the test
set-up. The impactor was geometrically
coordinated to produce buckling behav-
iour comparable with that in the com-
production of the semi-finished products advantages of this method are that very plete vehicle crash test. Once the virtual
and the final products. The established little fibre is wasted and complex pre- reference model with the conventional
technology for the flexible rolling of raw forming work during production is elimi- steel TRB B-pillar corresponded with the
material and blanks made of this (TRB), nated. To create the hollow prototype, a deformation in the complete vehicle and
using the sheet thicknesses defined in foam core was used as a support element, drop tower test, the final development
the development, and supplemented by around which carbon fibres were braided status of the hybrid B-pillar was imple-
hot forming of the component were used in a robot-controlled process. This process mented and simulated in the drop tower
for the B-pillar outer. The method used enables fibres to be applied in layers and CAE model.
for the B-pillar inner was conventional at defined angles, to create a kind of In pilot tests, the stiffness was coordi-
deep drawing of a blank with a constant sleeve. Low-pressure RTM (resin transfer nated in the drop tower so that it was at
sheet thickness. moulding) was used for the core process. the same level as that of the CAE model.
With a view to a mass production sce- This process involves resin impregnation Various subsequent test runs showed
nario, automated braiding was used for and curing in the RTM tool; a blown core that the test and CAE corresponded very
the creation of the FRP preforms. The is used to ensure that the material is fully closely, FIGURE 5.
03I2015 Volume 117 7
C OVER STORY LIGHT WEIGHT DESIGN

SUMMARY AND CONCLUSION

Drop tower test Simulation

As specialists in the composite sector,

Mubea Carbo Tech is keen to further the
idea of lightweight design: this motivated
them to develop a project centring on
hybrid lightweight construction for high-
volume production. The tests decribed
above have shown that an intelligent
material mix is a good choice, even in
high-stress areas such as the B-pillar,
where extremely heavy demands need to
be met. In the correct design and posi-
tion, FRP reinforcements can support the
function of a metal structure, and help to
FIGURE 5 Hybrid TRB B-pillar in drop tower at maximum deformation; it shows how closely deformation in make tried and tested concepts even
CAE and the drop tower test correspond more efficient and lightweight.
From a point of view of weight, the
hybrid B-pillar is a best-in-class solution
Benchmark which cannot be achieved using conven-
TRB steel B-pillar Hybrid B-pillar
(advanced tional manufacturing methods, and
(reference) (hybrid concept)
state-of-art) which also offers an additional weight
saving of approximately 20 % compared
B-pillar outer B-pillar outer (TRB) B-pillar outer
to the highly optimised TRB B-pillars,
Part list Reinforcement FRP reinforcement
B-pillar inner B-pillar, inner B-pillar inner TABLE 1. Modern, automated manufactur-
ing methods help to keep the high mate-
Tailor rolled blank rial costs involved in FRP volumes to a
Tailor rolled blank,
(option), cold forming
Manufacturing process or hot forming
Tailor rolled blank, hot forming reasonable level, FIGURE 6.
hot forming Braiding/RTM
Tailor tempering The results underlines the technical
adhesive bonding
(option) feasibility of the project, and serves as the
starting point for further developments
Weight 7.7 to 9.5 kg 7.9 kg 6.2 kg in which vehicle and manufacturer-spe-
cific requirements, material values and
developments can be in focus.
IIHS rating It is now the declared intention of
(intrusion distance to Not examined Good (160 mm) Good (127 mm)
centre of seat)
Mubea Carbo Tech to take this construc-
tion method into serial production and
TABLE 1 Comparison of design variants support the vehicle manufacturers in
pre-development activities.

REFERENCES
[1] Hillebrecht, M.; Begert, M.; Reul, W.; Kiel, B.:
Technik für Mischbauweisen ab 2020. In: ATZ 116
(2014), No. 5, pp. 48-51
[2] Carhs Safety Companion 2014, Wissen für
Mass:
die Fahrzeugentwicklung von morgen. Carhs,
20 % reduced Aschaffenburg 2014
[3] Schmitz, S.: Hybrid vehicle structures
IIHS rating: CFRP/Steel Carbon Tailor Rolled Blank B-Pillar.
”Good“ FutureCarBody, Bad Nauheim, 3 June 2014

FIGURE 6 CAE simulation of the IIHS load case

THANKS
The authors thank Dr.-Ing. Sebastian Schmitz and
Florian Strobl from the Mubea Carbo Tech GmbH,
Salzburg (Austria), and Dr.-Ing. Philip Beiter,
The difference between the two con- tional TRB B-pillar. The degree of defor- Benedict Göddeke, M. Sc., and Niklas Klinke, M. Sc.,
cepts meant that there was no reason to mation of the CAE model and of the final from the Mubea Tailor Rolled Blanks GmbH,
expect the same degree of deformation drop tower test with the hybrid B-pillar Attendorn (Germany), for the qualified technical
for the hybrid B-pillar and the conven- prototypes did correspond closely. support within the project.

8
 chassis tech . plus

6th International Munich Chassis Symposium

16 and 17 June 2015 www
.ATZ
live .de

Munich | Germany
C A R -T
O
-X C O
Bene MMU
fits fo N IC A
e com T IO N
safety r rid
and fo
effic
iency rt,
NEW
CH A S S IS
Dem SYST
ands EMS
of to on th
morr e cha
ow ssis
CHAS
S IS T
COMM EC
CAR-TO-X COMMUNICATION Inno
E R C IA H N O L O G Y
vatio
ns an
L VEH
d
FO
IC L E S R
requ
irem

Benefits for ride comfort,

ents

safety and efficiency /// SC

IE NT IF
IC DI
Prof. RE CT
Dr. P OR
Mun eter
ich U E.
Appli nivers Pfeffer
ed S ity of

NEW CHASSIS SYSTEMS

cienc
es
/// ON
E FO
R
AL L
Four
cong
Demands on the chassis in on
e eve
resse
nt
s

of tomorrow chas
sis.te Simult
ch +
a
steer
in Germa neous Inte

chas
g.tec
CHASSIS TECHNOLOGY FOR n and rp
Englis reting

.
h + br
ake.t h

sis te
ec
COMMERCIAL VEHICLES h +
tire-w
heel.t

ch
Innovations and requirements ech
6th In
te
16 an rnational M
© Op
el

d1 un plus
7 June ich Ch
2015 a
| Mun ssis Sympo
ich | G s
//// SCIENTIFIC DIRECTOR
erman ium
Prof. Dr. Peter E. Pfeffer y
Munich University of
Applied Sciences

/// ONE FOR ALL

Four congresses
in one event

/// KINDLY SUPPORTED BY

Phone +49 611 7878-131 PROGRAM AND REGISTRATION

Abraham-Lincoln-Straße 46
65189 Wiesbaden | Germany
Fax +49 611 7878-452
[email protected] www.ATZlive.com
C OVER STORY LIGHT WEIGHT DESIGN

Trends in Lightweight
Construction of Roofs
The bodywork is a major factor in the total weight of a vehicle, making up approximately
35 to 40 %. Its potential for reducing weight an CO2 emissions is great. Working with
innovative technologies ranging from polycarbonate to cellulose, Webasto is developing
trendsetting roof systems that help meet the targets for lightweight car body construction.

10
 AUTHOR

Matthias Arleth
is Deputy Chairman of the Board
of Webasto Roof & Components SE
in Stockdorf (Germany).

MOTIVATION

The lighter the better – is the formula

for success in automotive construction.
Every gram counts in the struggle to
achieve lower fuel consumption amid
the global demand for emission reduc-
tion. The rule of thumb is: 100 kg less
weight lowers the fuel consumption of a
vehicle by about 0.15 l. To achieve these
goals, super lightweight materials have
been developed. They include titanium,
magnesium or carbon fiber composites,
all of which feature a combination of
extremely low weight and high stiffness.
But they also come at a high price. One
that is generally too expensive for auto-
motive mass production.
That notwithstanding, at Webasto
these materials are brought up again and
again. For visions do produce innova-
tions. Based on the lightweight con-
struction study of a panorama roof, a
development team explored what would
be technically feasible – without consid-
eration of costs and mass production
suitability. The result: the current weight
was cut in half. Admittedly only with
individual components that were in part
very expensive and not appropriate for
series production. Yet some of the inno-
vations emerging from this study will
still find their way into car lines. The
value derived from such prototypes is
the opportunity to gain know-how and
accumulate experience in the practical
implementation of visionary concepts.
It all comes down to the distinction
between what is technically feasible and
that which is economically reasonable.
03I2015 Volume 117 11
C OVER STORY LIGHT WEIGHT DESIGN

LIGHTWEIGHT TECHNOLOGIES IN
THE ROOF SECTOR

In its drive to develop forward-looking

solutions that associate lightweight con-
struction with safety and comfort, the
automotive supplier works closely with
car manufacturers around the world.
Polycarbonate (PC) has been conquering
car construction for decades. The pio-
neering lightweight developments in the
roof segment include the use of plastics
and a super lightweight paper honey-
comb material, copied from the beehive.
Paper honeycomb (PHC) in conjunction
with reinforcing fiberglass facilitates
innovative lightweight construction solu-
tions – and that includes the outer shell
of the car body. The thermoplastic is
characterised by good mechanical,
thermal and optical properties. It was FIGURE 1 Structure of a PU composite with paper honeycomb core
most notably these features that paved
the way for its use in the automotive
industry.
Glass also remains an essential material (e.g. fiberglass), organic reinforcing fib- ogy, creating a complex honeycomb
in the automotive industry. Webasto ers (e.g. carbon fibers) or natural fibers structure that can absorb high compres-
developed the safety glass Webasto Glas (hemp, flax, etc.). sion forces. The insensitivity to moisture
Protec for use in the roof area. The weight Depending on the requirements and is ensured by the waterproof outer shell
advantage of the glass-plastic composite intended area of application, the core or the fiberglass/matrix composite.
as compared to conventional composite can consist of particle foam (e.g. polysty- The load impacting the component is
safety glass is around 10 to 15 %. rene) or cellulose (paper). Foam cores absorbed by the fibers. The matrix, e.g.
This paper presents innovative pro- have advantages with regard to the spe- polyurethane (PU), bonds the fibers,
cesses, technologies and materials cific weight and heat-insulation, but have thus enabling load input and output, pre-
geared towards weight reduction. They disadvantages in terms of temperature venting buckling under pressure that is
are: stability and costs. Cores made of paper parallel to the grain. Moreover, the
– PU composites for roof elements honeycombs are manufactured from matrix Protects the fibers from environ-
– Polycarbonate for roof windows and cellulose fibers that are predominantly mental influences.
panels obtained from recycled waste paper. What’s more: PU composite materials
– glass and plastic composite: Webasto The individual paper layers are joined are lightweight, which puts them among
Glas Protec. together by means of ultrasonic technol- the important ingredients for future

PU COMPOSITES: STABLE,
LIGHTWEIGHT, ECONOMICAL

PU composites are generally noted for

their high degree of stiffness with low
weight, good 3-D moldability and acous-
tic as well as heat-insulating properties.
They are constructed in a symmetrical
or asymmetrical sandwich structure and
in exterior components consist of a
structural outer layer (shell), fiber/PU
matrix and a core. In addition, fastening
elements or other comfort-enhancing
features can be integrated.
The stiffness of the material is essen-
tially achieved by the fibers embedded in
the matrix and the sandwich thickness.
Depending on the stiffness requirements, FIGURE 2 Jeep New Renegade: two roof panels that can be quickly removed from the roof system
these are composed of inorganic fibers with just a few simple steps

12
lightweight formulas. Due to their great POLYCARBONATE ROOFS technology that enables this: The infra-
stiffness these materials are well suited red absorbers integrated into the polycar-
for body shell parts of up to 2 m2. Plastics such as polycarbonate are suit- bonate – and that already in the granu-
Thanks to the low tooling costs, espe- able for use as lightweight materials. late material of which the panes are
cially for order volumes with small to The material is transparent, extremely manufactured. The new smart marks the
medium unit numbers, PU composites impact resistant, neither splintering first time use of the product called Mak-
are also economically interesting nor breaking on impact. And yet it is rolon AG2677 in the color and tempera-
materials. very lightweight. With a density of ture-stable, heat absorbing color 771079.
In the new Jeep Renegade, for the first 1.20 g/cm3, polycarbonate components Also taking ecological aspects into
time Webasto introduces a two-piece roof weight up to 50 % less than identical consideration, the infrared absorption
system that can be opened and whose parts made of glass. The extremely high has benefits. If the interior heats up less,
elements can be completely removed and impact resistance is particularly relevant the need for using the air-conditioner for
stored in the trunk of the car. The core for application in the roof area. Protec- cooling decreases. That further reduces
consists of paper honeycomb encased in tion of the passengers in case of an acci- fuel consumption.
a mixture of polyurethane and reinforc- dent or falling rocks is always ensured,
ing fiberglass, FIGURE 1. The exterior sur- even when the vehicle rolls over. In
PC PANELS
faces are covered with a grained, thermo- addition, an appropriate varnish coating
formed and weatherproof film, while the provides scratch resistance as well as In addition to glazing components in a
inside surfaces are flocked to match the resistance to weather-related effects and Class-A-look, Schierling produces light-
interior décor. UV radiation. weight panels made of polycarbonate for
The special challenge posed by this Automobile designers appreciate the various automakers. That includes 14
roof system involved the compensation of freely moldable material that offers them models alone for the Volkswagen brand
the curvature change. The asymmetrical multi-faceted design options that go far universe. These panels border the tilt/
sandwich construction creates a so-called beyond traditional sliding and panorama slide sunroofs in the front and at the
bimetallic effect, which results in differ- roofs. Polycarbonate allows for creating sides, creating a seamless transition to
ent curvatures in varying temperatures. very large roofs and just as visionary the vehicle’s roof. This gives the entire
By means of FEM calculation, this ideas like boldly curving windows. roof surface a high-quality solid glass
dynamic was taken into consideration in The first polycarbonate roof system for appearance. Such panels in 3-D shapes
the sealing design. The result: a tight seal the smart fortwo rolled off the produc- are not feasible in glass, FIGURE 4.
under all conditions, no wind noises and tion line at the Webasto Schierling plant Currently, the proportion of polycar-
a harmonious curvature shape, FIGURE 2 in 2007. The latest generation of the city bonate used in automobile construction
and FIGURE 3. car is also equipped with a panorama is about 15 %. Experts assume that
The weight savings of this PU compos- roof made of polycarbonate. The roof within the next five years that share will
ite development versus steel are about window size is 1.20 m2, it has a thick- increase to about 20 %. For the potential
50 %. Moreover, it is the first use of ness of 5.25 mm and weighs 9.80 kg. But of polycarbonate has by no means been
Paper Honeycomb (PHC) in the exterior the real technical highlight is its capacity fully exploited as yet. For instance, fi xed
shell of a vehicle undertaken in serial to absorb solar energy and thus to keep rear side and rear windows, windshields,
production. Up to now PHC was used the interior of the car from heating up. spoilers with integrated taillights and
primarily as a lightweight material for That means that the infrared radiation even movable (hinged) side windows are
headliners of sunroofs. stays where it belongs – outside. The conceivable. Additional functions can
also be integrated into PC parts, such as
heating wires. Webasto is already work-
ing on developments on these subjects.

WEBASTO GLAS PROTEC

That being said, glass remains an essen-

tial material for the automotive industry –
provided that the high requirements for
a further weight reduction can be met.
Because the customer appreciates light in
his car interior, and the brightness engen-
dered by a transparent roof system cre-
ates a feeling of well being. There is
another reason for the growth of the pro-
portion of glass in the car: Cars currently
have approximately 4.60 m2 glass surface,
while in 1985 it was only 3.50 m2.
But glass is heavy. Another lightweight
FIGURE 3 FEM simulation of the Jeep New Renegade roof elements alternative is called Webasto Glas Protec.
03I2015 Volume 117 13
C OVER STORY LIGHT WEIGHT DESIGN

This safety glazing, a glass-plastic com- secure anchoring of the glass pane in the to a special coating can contribute to
posite, enables the achievement of 10 to roof system is ensured by the patented reduce consumption.
15 % weight savings as compared to perforation of the film. Roof systems As it were, radiation of cold in the area
conventional laminated safety glass. A made of Webasto Glas Protec can be of the head could be fended off. As is
highly tear-resistant PET film (polyethyl- manufactured in lengths of up to 1.60 m, already common in building glazing, a
ene terephthalate) is applied to the inner FIGURE 5. thin metal coating can produce the effect
surface of convex single-pane safety Then again, weight isn’t everything of having the thermal radiation of pas-
glass (ESG). In the event of glass break- when it comes to innovative roof sys- sengers’ own bodies reflected onto the
age, the film holds the fragments tems. Ecological aspects play an ever inner surface of the pane. This lowers
together, thereby Protecting passengers greater role here, especially with a view the heating requirements – an energy
from injuries from flying glass splinters. to e-mobility. Additional comfort fea- topic primarily relevant for electric vehi-
The protective effect of the film allows tures, as for example, with Webasto cles. The developers are also working on
for the usage of very thin (< 3 mm) and Glas Protec a tint in the fi lm or the the use of organic photovoltaics in con-
thus lighter weight glass panes. The absorption of infrared radiation thanks junction with Webasto Glas Protec.
This safety glazing is being installed
in series production since 2005. Current
models include the Range Rover, Mer-
cedes R-Class and Ford Lincoln. The
Jaguar XJ sedan features a large two-
part panorama roof made of Webasto
Glas Protec as standard equipment.

MULTI OPTIONAL ROOF

The Multi Optional Roof (MOR) is a com-

pletely new dimension in lightweight car
body construction. The idea: One vehicle
interface – multiple roof versions. What
this means is that different roof systems
can be fitted onto one vehicle shell ver-
sion. The installation interface is identi-
cal for all roof variants. The roof mod-
ules are delivered on demand, ready to
install and are then built in on the
assembly line. The great advantage: The
automotive manufacturer does not have
FIGURE 4 Tilt/slide sunroof for Golf VII: panel made of polycarbonate in Class-A look to decide early in the construction stage
which roof system should be installed on
later models of a vehicle. This provides
enormous flexibility that allows for rapid
reactions to market trends or for upgrad-
ing facelifts with attractive roof systems.
And it reduces complexity, since the
number of vehicle shell constructions is
brought down, FIGURE 6.
Webasto consolidates all it lightweight
construction technologies in the MOR. It
is also conceivable that the surface of the
hardtop element could be equipped with
solar cells. The solar power thus gener-
ated can be supplied directly to the bat-
tery. This results in a drop in fuel con-
sumption. That is why solar roofs are
listed as an Eco Innovation by the Euro-
pean Union. The CO2 savings potential of
solar roofs is more than respectable: An
Otto engine emits an average of 2.3 g/km
less CO2, a diesel vehicle about 1.6 g/km
less when equipped with a solar roof for
FIGURE 5 Quality check of the PET film for the Webasto Glas ProTec glazing technology charging the battery.
14
FIGURE 6 Multi optional roof: multiple roof versions on one shell construction version

OUTLOOK thing we have succeeded in accom-

plishing with our lightweight panes
Fuel savings and CO2 reduction will made of polycarbonate, Webasto Glas
remain one of the predominant issues Protec and creative ideas in the area of
in the automotive industry for some PU composites. But progress demands
time to come. Therefore, lightweight ongoing further development. That is
construction technologies – and that why the tradition-rich Bavarian com-
also includes the roof area – continue pany continues to invest in these tech-
to gain in significance. As an innova- nologies. Above all, it is the electrifica-
tive supplier, the development and tion of the automobile that poses great
implementation of sustainable solu- new challenges for lightweight
tions is an obligation. That is some- construction.
03I2015 Volume 117 15
DE VELO PMENT V EHICL E EL ECTRICAL SYSTEM

Realisation of Safe
Management Systems for
Lithium-ion Batteries

AUTHORS When developing high-voltage traction batteries for electric vehicles,

the focus is not only on the costs but also on the safety of lithium-
ion cells and the protection of occupants and other people touching
the car against electric shocks. Secure system architecture for the
battery monitoring electronics, customised on every vehicle, is an
Dr.-Ing. Marc Nalbach
is Director of Design & Development
important variable for complying with safety objectives. Hella presents
Energy Management at Hella KGaA
Hueck & Co. in Lippstadt
how the system architecture of a battery management system can
(Germany). be developed sustainably and cost-effectively with the support of
model-based methods and in compliance with the ISO 26262.

Dr.-Ing. Christian Wagner DEVELOPMENT OF LITHIUM-ION Today traction batteries with a cathode
is System Engineer for Battery BATTERY TECHNOLOGY material of lithium-nickel-manganese-
Management Systems at Hella
KGaA Hueck & Co. in Lippstadt cobalt oxide are used in automotive
(Germany). Due to the future worldwide CO2 emis- applications. In particular, the so-called
sion targets, the distribution of electric third-mix NMC cells with a mixing ratio
and hybrid vehicles in the coming years of one to one to one (1-1-1) are a common
will steadily increase. The system costs technology. The influence of the individ-
will be reduced and the evolution of the ual components are: manganese
electric drive systems must proceed increases the intrinsic safety of the
quickly [1]. chemical, cobalt increases the life time
16
and nickel increases the energy density. proper connection and protection of all SECURE SYSTEM ARCHITECTURE
Nevertheless, the 1-1-1 NMC technology high-voltage components. Furthermore,
must be strictly protected against over- a constant monitoring of the insulation In this context, a secure architecture is
charge and deep discharge to avoid a level of the high-voltage system is imple- defined as the hierarchical and func-
thermo chemical accident. With empha- mented by active and passive measure- tional break down of safety goals into
sis on longer ranges the proportion of ment methods in current and future sys- system functions. The structured distri-
nickel has to increase significantly in the tems [3]. In summary, different electronic bution of these functions on system com-
future. Today mixtures in the ratio 6-2-2 components are necessary for monitoring ponents (control units) and the further
are already in development. By 2020, a and conditioning of the battery chemistry refinement down to the hardware com-
ratio of 8-1-1 is forecasted for automotive and the protection against high-voltage. ponent level is part of a reliable system
applications. The last one is already in design in terms of security and robust-
development for consumer electronics. ness. Additionally, all functions which
BATTERY MANAGEMENT SYSTEM
However, the chemical safety of the are non-safety related but necessary for
high-energy compositions will decrease The electronic monitoring devices are operation of a BMS must be incorporated
significantly. Therefore the monitoring of summarised as Battery Management in the consideration area.
the battery cells becomes increasingly System (BMS). As shown in FIGURE 1, the For dealing with the complexity aris-
crucial [2]. In addition to the electrical BMS consists of several system compo- ing from this context, the ISO 26262
management, the thermal conditioning nents. Typical tasks of the BMS are guar- defines a specific process model includ-
of the many cells for the traction battery anteed achievement of safety goals, pro- ing work products and applied methods.
equally emerges as a challenging task. viding accurately measured data for the The aim of the ISO standard is to docu-
For this, the demanding cooling and battery application and realisation of ment the traceability and verifiability of
even heating systems were and are inte- established vehicle functions such as the development with the current state of
grated in the battery, they are electroni- diagnostics or network management. the art methods.
cally controlled to keep the battery One challenge in developing a secure The standard includes recommenda-
chemistry in an optimal working range. architecture for a BMS is the functional tions on requirements management as
Besides, the voltage level of a high volt- separation of safety-critical and non- well as for the description of system
age battery requires additional safety safety-critical parts, for example for sys- architectures [5]. The requirements man-
mechanisms. The focus is on measures to tem functions, vehicle functions or com- agement is therefore the basis of a secure
protect the vehicle passengers as well as munication data as well as interaction system architecture. The ISO 26262
external staff against electric shocks. The between different BMS control units places special emphasis on traceability
high-voltage (HV) monitoring circuit among each other and their interfaces to not only between system requirements
(High Voltage Interlock) ensures the the system context. and customer requirements (including

FIGURE 1 Battery Management System (BMS) architecture for monitoring of the high-voltage accumulator
(BMU = Battery Management Unit, LV = Low Voltage, CMU = Cell Measurement Unit, HVCS = High Voltage Current Sensor,
IMD = Insulation Monitoring Device)

03I2015 Volume 117 17

DE VELO PMENT V EHICL E EL ECTRICAL SYSTEM

the functional safety concept), but also sidered as an overall system. Below, fur- In contrast to the functional architec-
demands a strict link between system ther layers include detailed elements ture, the technical architecture describes
architecture and system requirements. such as associated control units on level a specific solution. At the top level L1,
Considering this, the influence of L2 and hardware components in level the system context is the reference point
requirements on the architecture and L3. This structure enables a systematic of the technical domain.
vice versa is comprehensible. function break down of the overall sys- The distribution of functions and
The BMS safety goals are usually rated tem down to individual system elements. function blocks to the control units are
with ASIL B or ASIL C level. Therefore, a By linking the architectural elements to realised in the next L2 abstraction level.
semi-formal notation for describing the corresponding requirements the tracea- Basically, the technical architecture dis-
design, TABLE 1, is recommended by the bility required by the ISO standard is ful- tinguishes two characteristics of model-
ISO standard. Model-based description filled including FSC and TSC. ling methods. The static modelling
languages like SysML are one character- In detail, a solution-independent describes the interfaces and the distribu-
istic of semi-formal notation. The inher- model of the system is described by the tion of functions (point ④ in FIGURE 2).
ent structuring of the development functional architecture. The first step The dynamic modelling is used to
approach and its associated mastering of in the development process is the crea- describe the behaviour of the system
complexity are advantages of model- tion of use case diagrams (point ① in (point ⑤ in FIGURE 2). The interaction
based methods. Furthermore, the use of FIGURE 2). The protection of passengers with external and internal control units
graphical modelling languages supports against electric shock is a common use is detailed in L2. For the behavioural
the intuitive understanding of the rela- case of the BMS. The subsequent refine- aspects, sequence and state diagrams are
tionships within the system. ment with activity diagrams (point ② in the preferred modelling notation. The
For high voltage batteries, the monitor- FIGURE 2) describes the sum of all func- individual elements of the architecture
ing of the cell chemistry is often classi- tions that must be covered by the BMS. are in turn linked with the requirements
fied as ASIL C due to a high exposure The just mentioned use case can be (point ⑥ in FIGURE 2) to justify the
and low controllability. In contrast, the refined in the following functions: design decision and maintain the
protections of passengers is usually rated – measurement of insulation resistance traceability.
slightly lower with ASIL B. – comparison of the insulation resist- The modelling of the underlying layer
ance against a threshold values (L3) differs only in the degree of abstrac-
– warning of the driver in case of an tion from layer L2. Here the hardware
MODELLING CONCEPT
error. modules of a control unit are the element
Due to the amount of options offered by Thereafter, the functional architecture under study. The functional blocks of the
SysML (over 200 language elements), it is (point ③ in FIGURE 2) contains groups of previously introduced example are allo-
necessary to develop a modelling con- all identified systems. The functions of the cated to one control unit, the BMU (L2).
cept which restricts the variability. It use case that protect against electric shock The insulation measurement is allocated
describes the purpose and the applica- are grouped into a function block insula- to a dedicated controller (L3) and the
tion of specific modelling language ele- tion measurement and a communication communication of the warning message
ments in the development process [6]. block. The preliminary architecture will is operated via CAN bus. The resulting
In FIGURE 2, the general aspects of such not provide a concrete solution, but can be interfaces, the CAN interface and the
a modelling concept are illustrated. The evaluated against the requirements in the voltage measurement interface to calcu-
vertical separation between functional FSC. Thus inadequacies in the design can late the insulation resistance, could be
and technical architecture is important. be discovered and resolved at a very early aligned with the system context (L1).
Both aspects are also represented in the stage. Additionally, each function can be Due to this process model it becomes
ISO 26262 standard as functionally (FSC) ASIL classified into a safety level by the automatically a constant interaction
and technically safety concepts (TSC). linkage of the FSC. This allows an effi- between requirements and architecture.
The horizontal separation is used to cient design of the function blocks and a Usually new requirements arise during
describe the BMS at different abstraction separation of safety-critical and non-safety system modelling. These must be taken
levels. On top level L1, the BMS is con- critical functions. into account and incorporated back into
the existing requirements. This process
is not a negative one – in contrast it
shows systematically specification gaps
ASIL level and closes them. This process is also
Three methods
A B C D known as “zig-zag” pattern [6]. More-
1. Informal notation (description languages with a weakly defined over the model-driven structured
++ ++ + +
syntax, for example diagrams, sketches) approach reduces design errors and
2. Semi-formal notation (languages with formally described promotes cost-effective development.
syntax, but semantics is not completely and unique specifi ed, + ++ ++ ++
for example UML)
CONCLUSION
3. Formal notation (both syntax and semantic are completely
defined, for example Z, VDM, PVS) + + + +
The seamless integration of a battery
TABLE 1 Process model with three methods according to ISO 26262, volume 6, page 18 management system into a vehicle is a
(++ = highly recommended, + = recommended) [4] complex task. The interplay of system
18
 Functional architecture Technical architecture

FSC

L1 -- BMS
Require- System context
ments
Use
cases

L2 -- BMU/CMU/HVCS
Structural Behavioural
Activity model model
diagrams

Functional
architecture Requirements
+ TSC

L3 -- HW components
Derive
Satisfy Structural Behavioural
Refines model model
Realises/details

FIGURE 2 Overview of the modelling concepts for functional and for technical architecture

functions, safety goals and a plurality

of control units require a structured
approach to control complexity as cost
driver. The described procedure model
for the development of secure architec-
tures has been successfully applied and
optimised [7]. The mentioned require-
ments of ISO 26262 are met and the nec-
essary work products to demonstrate
compliance with the safety goals are
made directly within the process. In
addition, the efficiency of this methodol-
ogy results from a change from a docu-
ment-based to a model-based approach.

REFERENCES
[1] Schulte, A.: Market Study E-Mobility: Market
Development, Infrastructure, Costs and Regulations.
In: Elektrik/Elektronik in Hybrid- und Elektro-
fahrzeugen und elektrisches Energiemanagement V
(2014), pp. 201-213, Renningen: Expert-Verlag,
2014
[2] Theis, C.; Hackmann, M.: Die Zelle zu unter-
schätzen, könnte fatale Folgen haben. Interview.
In: ATZelektronik 9 (2014), No. 3, pp.16-18
[3] Nalbach, M.; Mehling, F.: Isolationsfehler in
Bordnetzen erkennen. In: Hanser eMobility 1
(2014), pp. 18-20
[4] ISO 26262, Volume 6
[5] ISO 26262, Volume 4
[6] Weilkins, T.: Systems Engineering with SysML/
UML. Modeling, Analysis, Design. 1st edition, The
MK /OMG Press, February 12, 2008
[7] Wagner, C.; Meyer, J.: Erfahrungen mit dem
modellbasiertem Systementwurf am Beispiel des
automobilen Steuergerätes Battery Management
System (BMS) I. In: Maurer, M.; Schulze, S.-O.:
Tag des Systems Engineering (12.-14.11.2014),
pp. 215-224, München: Hanser, 2015

03I2015 Volume 117 19

DE VELO PMENT SIMUL ATION AND TEST

Potential of Online Model Identification

for Vehicle Dynamics Controls
of Load Sensitive Lightweight Vehicles

AUTHORS

Dipl.-Ing. Florian Kohlhuber Elias Schneider, B. Sc. Prof. Dr.-Ing. Markus Lienkamp
is Research Associate at the is Master Student in the course is Director of the Institute of
Institute of Automotive Technology Vehicle and Motor Technology at Automotive Technology at the
at the Technische Universität the Technische Universität Technische Universität München
München (Germany). München (Germany). (Germany).

Future efficiency-focused vehicle concepts with curb weights around 500 kg

suffer from a high influence of everyday load on vehicle dynamics character-
istics. Currently, researchers at the TU München work on driving-synchronous
estimation methods to determine valid vehicle parameters for each trip to
improve the quality of chassis control systems.
20
PROBLEM OF LOAD SENSITIVIT Y load distribution of 45:55 and a wheel values of the rear axle, FIGURE 2 shows
base of 2100 mm. Different weights for the minimum and maximum value as
In order to achieve ambitious efficiency passengers as well as luggage in the well as the relative deviation based on
improvements for future automobiles, a front and rear trunk were combined to the seven setups.
trend towards design and production of seven load scenarios, each representing One can clearly observe that all rele-
lightweight vehicle concepts has started. a maximum for a characteristic physical vant physical parameters underlie trip
With about 500 kg, these cars have con- property. individual deviations of around 30 %.
siderably lower curb weights than nowa- In this way, the variance of the most Furthermore, it should be pointed out,
days cars, which usually range from 900 important physical vehicle properties that the longitudinal COG position can
up to over 2000 kg. Examples are the can be observed. These are for the over- vary by more than 200 mm which is
Nils (Volkswagen, 430 kg), the Urban all vehicle the total mass, the yaw iner- almost 10 % of the wheel base. In com-
Concept (Audi, 500 kg) or the Twizy tia, the longitudinal COG position and bination with the corresponding total
(Renault, 480 kg). But, on the other the relative front axle load. Basic axle masses, axle load distributions between
hand, new challenges for passive and behaviour as a combination of suspen- 39:61 and 49:51 are possible. Although,
active driving dynamics arise in this sion and tyre (dimension: 115/70 R16) one could state, that these high parame-
new vehicle class. behaviour is represented by three ter deviations can occasionally occur
These challenges result from the parameters, namely the mean axle for conventional cars as well, very high
fact that the reduction of a vehicle’s cornering stiffness, the mean lateral loads of several hundred kilograms
unloaded weight leads to an exponen- relaxation length and the mean pneu- would be necessary. With respect to
tial increase of the influence of load, matic trail. For each of the three overall lightweight vehicles, just little everyday
meaning passengers as well as luggage, vehicle parameters and for the three loads can cause the same effect.
on a vehicle’s physics, FIGURE 1. Regard-
ing conventional nowadays cars, a
typical load with driver and co-driver
(80 kg each) rarely represents more
than 15 % additional weight, whereas
for lightweight vehicles the ratio is
around 1:3. Assuming a maximum
load of 210 kg by adding 50 kg of lug-
gage, the load to curb weight ratio of
lightweight vehicles can easily double
or triple compared to conventional cars.
This, in turn, leads to an increasing
load influence on the overall para-
meters of the loaded vehicle, both on
the inertial values like the centre of
gravity (COG) location and moments of
inertia as well as on chassis parameters
like axle loads or tyre characteristics.
In this way, vehicle dynamics differ FIGURE 1 Load influence with respect to different vehicle classes
considerably between trips depending
on varying loads that are transported.
This problem is called load sensitivity.
Rear Front
Driver Co-driver Total
Setup luggage luggage Annotation
[kg] [kg] mass [kg]
VARIANCE OF VEHICLE AND [kg] [kg]
T YRE PARAMETERS
Minimum load with light
S550 50 0 0 0 550
female driver
To investigate the impact of different
everyday loads on the driving dynamics Standard, most-likely
S580 80 0 0 0 580
setup
of lightweight vehicles and to derive the
Maximum relative front
possible spectrum of values vehicle and S585 50 0 0 35 585
axle load
tyre parameters can attain, the exem-
High yaw inertia with
plary load scenarios in TABLE 1 have been S610 50 0 30 30 610
low load
analysed. For this, the electric vehicle High absolute and
concept of the Technische Universität S695 80 80 0 35 695
relative front axle load
München, which has been presented at
S730 80 80 35 35 730 Maximum of yaw inertia
the International Motor Show 2011, has
been used as a reference. This vehicle Overload with maximum
S740 90 90 60 0 740
of rear axle load
has a curb weight of 500 kg, a yaw
moment of inertia of 500 kgm², an axle TABLE 1 Analysed everyday load scenarios

03I2015 Volume 117 21

DE VELO PMENT SIMUL ATION AND TEST

LOAD INFLUENCE ON PASSIVE

DRIVING DYNAMICS

In order to demonstrate the effect of the

presented parameter deviations on the
passive vehicle behaviour, step steer
manoeuvres have been simulated for all
setups in a multi-body simulation
(Adams/Car). FIGURE 3 presents the lat-
eral accelerations and yaw rates of the
setups S695, S740 and the most likely
“driver-only” scenario S580. As steering
wheel step input an amplitude of 20°
was chosen at a vehicle speed of 120 FIGURE 2 Load dependent variance of basic vehicle parameters
km/h. Regarding the maximum ampli-
tudes, the transient overshoots and the
eigenfrequencies of the measured out-
puts, the vehicle setups’ responses are POTENTIAL OF ONLINE MODEL ward control as well as in the online
very different. For example, setup S695 IDENTIFICATION FOR VEHICLE identification algorithm. On the one
reaches a maximum yaw rate of 10 °/s, DYNAMICS CONTROLS hand, this model should work with a
whereas the “driver-only” setup´s yaw small set of parameters and, on the other
rate maximum is at almost 15 °/s. Con- Model-based vehicle dynamics controls hand, it should be able to represent the
cerning the maximum lateral accelera- normally rely on constant values for the dynamic vehicle behaviour very well
tion, the bandwidth of all setups ranges internal model parameters. It was shown within limits of everyday driving. A
from 0,4 to almost 0,6 g representing an that for lightweight vehicles these good choice for a model matching these
obviously high difference in vehicle reac- parameters differ on a wide range for requirements is based on the single-track
tion, too. each trip. Therefore, the standard values model which is extended for several
A more detailed analysis of the used in the controller are not valid for aspects. To simplify estimation of total
vehicle’s parameter variation and the most of the trips. To solve this problem, mass, longitudinal dynamics are incor-
effect on the passive driving behaviour those parameters have to be determined porated into the model. To account for
can be found in [1]. In summary, it can online for each trip to update the inter- the typical narrow tyre decrease of lat-
be stated that already small differences nal model. The potential of such an eral stiffness already at low wheel lateral
of masses or positions of daily loads online model identification is presented slip angles, a nonlinear forming parame-
have a high effect on the agility and the in the following based on a Torque Vec- ter is added to the linear stiffness factor.
self-steering behaviour as well as the toring (TV) feed-forward control. In addition, the delayed lateral force
threshold between stable and instable In advance to that, a vehicle model has buildup is incorporated through mean
driving behaviour. to be chosen, that is used in the feed-for- lateral relaxation lengths. In total, only

FIGURE 3 Responses of differently loaded vehicles FIGURE 4 Torque Vectoring (TV) feed-forward control and online model identification

22
 FIGURE 5 Single lane change
manoeuvres without and with
standard as well as optimal
feed-forward control

three parameters are used to model each The requested yaw torques as output multi-body simulation. This results are
axle’s suspension/tyre system. Together from the different feed-forward controls based on real customer relevant inner
with the three overall vehicle parameters differ much. The TV torque is applied via and outer city driving cycles with veloci-
mass, yaw inertia and longitudinal COG an active differential on the rear axle of ties ranging from 0 to 120 km/h and lat-
position, the vehicle model in total can the vehicles. The example shows that the eral accelerations up to 0,3 g.
be parameterised by nine parameters. If quality of feed-forward controls of load
those are chosen physically correct, the sensitive cars can be considerably REFERENCES
model is able to represent real vehicle increased if the internal model is [1] Kohlhuber, F., Lienkamp, M.: Load problem of
lightweight electric vehicles and solution by online
behaviour in terms of lateral forces, side updated to the trip individual correct model adaption. 5 th International Munich Chassis
slip angle, lateral acceleration and yaw parameter set. Symposium, 24/25 June 2014, Munich,
rate very well. pp. 281-302
Similar to [2] the feed-forward algo- [2] Obermüller, A.: Modellbasierte Fahrzustands-
APPROACH FOR ONLINE schätzung zur Ansteuerung einer aktiven Hinterach-
rithm uses this model and calculates
PARAMETER ESTIMATION BASED skinematik, München, Technische Universität, Dis-
-without any feed-back loop an additional sertation, Cuvillier-Verlag, Göttingen, 2012
ON CUSTOMER RELEVANT DRIVING
yaw torque based on velocity, desired yaw [3] Kohlhuber, F., Büchner, S., Lienkamp, M.: Trip-
synchronous parameter estimation of vehicle and
rate and axle steering angles, that is An algorithm that is able to estimate this
tire model parameters as virtual sensor for load-sen-
needed to make the vehicle follow the set consisting of nine parameters has sitive lightweight vehicles. In: IEEE International
desired yaw rate trajectory. This is shown been presented in [1] and further devel- Conference on Vehicular Electronics and Safety
in the upper half of FIGURE 4. oped in [3]. Without the need for addi- (ICVES), 16-18 December 2014, Hyderabad/India,
accepted paper
To determine the possible improve- tional measurement hardware other than
[4] Best, M. C.: Identifying tyre models directly from
ment of the feed-forward control by ESC sensors (steering wheel angle, motor vehicle test data using an extended Kalman filter.
using trip-optimal parameters, lane torque, velocity, yaw rate, lateral/longi- In: Vehicle System Dynamics: International Journal
change manoeuvres at a velocity of 100 tudinal acceleration) a recursive filter of Vehicle Mechanics and Mobility 48 (2010), No. 2,
pp. 171-187
km/h have been simulated in multi-body identifies the described nine-parameter [5] Hon, S. et al.: Vehicle inertial parameter identifi-
simulation with all setups. Each setup model within the first minutes of random cation using extended and unscented Kalman fil-
was driven without control and twice customer relevant trips. The lower part ters: In: Proceedings of the 16 th International IEEE
Annual Conference on Intelligent Transportation
with feed-forward control, once with the of FIGURE 4 shows that the optimal
Systems, The Hague, 2013, pp. 1436-1441
controller values set to the standard, parameter set is found comparing the
most-likely S580 case and in a second model outputs to the measurement of the
simulation set to the values that fit the real vehicle. Therefore, an Unscented
corresponding setup. In this way, the Kalman Filter (UKF) is applied which
vehicle response with a non-adaptive uses a priori known suspension kinemat-
standard feed-forward control can be ics and delivers accurate values for the
compared to the behaviour of an opti- three chassis and the six axle parame-
mally parameterised controller. ters. Compared to related work, there is
FIGURE 5 shows the results for the set- no need to know certain values like COG
ups S695 and S740. In both cases the position in advance [3] or to make use of
vehicle with a correct parameterised standardised driving manoeuvres [5],
control (red) is able to follow the desired which would not be applicable for the
yaw rate trajectory (black) very well given case anyway. Regarding the esti-
even without a feed-back control. If the mated parameters, relative accuracies of
TV controller just uses the standard S580 far more than 90 % have just been
set of parameters (blue), results are poor. achieved after 5 to 7 min journey time in
03I2015 Volume 117 23
DE VELO PMENT INTERIOR

Interior Concept with

Light Centre Console

AUTHORS

Dr. Isabella Schmiedel

is Director of Interior Advance
Development for the Dräxlmaier
Group in Vilsbiburg (Germany).

Dipl.-Ing. Manfred Winklbauer

is Department Head of Light and
Product Innovations in Interior
Dräxlmaier Group’s new X² centre console architecture opens Advance Development for the
Dräxlmaier Group in Vilsbiburg
up great potential for lightweight design. Thanks to its integral (Germany).

design, this innovative product weighs 30 % less than typical

centre consoles generally installed in mid-size vehicles today.
Numerous studies confirming series production readiness
have already been completed.
Lion Pfeufer
is Online Editor of the Corporate
Communications department in the
Dräxlmaier Group in Vilsbiburg
(Germany).

24
COMBINATION OF TWO THE STATUS QUO enclose the structural component –
ARCHITECTURES shown here in red, FIGURE 2. The large
Two centre console architectures currently number of panels in different shapes
Companies today face wide-ranging chal- dominate the market: the shell design and need to be manufactured separately as
lenges in the area of sustainability. Car the support element design. With regard well as requiring many assembly steps,
makers and automotive suppliers operate to the former, the outer skin is largely since the panels must be fastened using
within a complex environment with high formed by the shell itself (shown in red), screws, clips or other joining methods.
expectations regarding social responsibil- which is generally manufactured with a
ity that opens up new potential, economic grain texture or painted, FIGURE 1. Rein-
ADVANTAGES OF THE NEW
aspects such as rising fuel prices and forcing interior elements provide addi-
INTEGRAL DESIGN
increasing resource scarcity, and strict tional support for loads. Thus, the disad-
statutory requirements for the reduction vantage is that the strength required of The “integral design” described here
of CO2 emissions. Products with low envi- the centre console has to be provided by refers to the supporting, structure inte-
ronmental impact and reduced weight separate internal structural components. grated into the console sides. This new
make an important contribution to over- These must be manufactured in addition X² centre console, FIGURE 3, combines
coming these challenges. and installed, leading to more processes the two typical centre console architec-
Against this backdrop, the Dräxlmaier and increased investment in tooling. tures, improving their common advan-
Group spent three years developing a Additionally, the injection moulding pro- tages while largely dispensing with
centre console that opens up new poten- duction process allows for only slightly their disadvantages. Of particular note
tials in lightweight design thanks to the contoured, plain console sides, placing is the lightweight design: The integral
combination of two typical architec- limits on styling options such as patterns design and the centrally divided sup-
tures [1]. Thus, the new X² centre con- or ornaments. port element both significantly reduce
sole combines a shell design with a sup- The support element design offers the number of components needed to
port element design, resulting in a greater design freedom, but this is achieve the required strength. For
noticeable reduction in weight. achieved with numerous panels that example, most of the panels are elimi-

FIGURE 1 Exploded view of a centre console with conventional shell design

(the red coloured shell itself is the outer skin)

03I2015 Volume 117 25

DE VELO PMENT INTERIOR

FIGURE 2 Exploded view of a centre console

with conventional support element design
(structural component shown in red)

nated, and the reinforcing interior ele- LIGHT MATERIALS currently found in mid-range vehicles.
ments can be reduced to a single com- This corresponds to weight savings of
ponent, FIGURE 3. Another important aspect is the use of around 1.5 kg. The new laminated design
The closure on the armrest is also lightweight materials in the X² centre and the the centre console’s tool-free
eliminated, which further supports the console. For example, the two lightweight installation help optimise costs.
principle of lightweight design. This is design side support elements are pro-
replaced by an innovative opening con- duced using a Thermoplastic Foam
ST YLING POSSIBILITIES
cept driven by a special spring. The base Moulding (TFM) process. Plastics rein-
configuration, which is the lightest, can forced with natural and glass fibres are At the same time, the X² centre console
be upgraded with an armrest height also used. The storage compartment and permits alterations according to cus-
adjustment feature, if desired. The mod- the panel are made, for instance, of the tomer requirements. For example, side
ularity of the X² centre console comes light, sustainable composite material Nat- padding can be subsequently installed in
into play here: The upgrade only requires ural Fibre Polypropylene (NFPP). Overall, the knee area. The styling of decorative
two easily installed additional parts on the interior concept is roughly 30 % elements and the decorative panel can
the console sides. lighter than a conventional centre console also be varied. Decorative styling of
26
plastic, aluminum or wood is possible.
Special features such as applications or
decorative designs on the armrest cover-
ing as well as premium leather stylings
are conceivable. In other words, this is a
huge leap towards maximum styling
freedom. However, the lightweight
design potential is best achieved in the
entry-level variants.

SERIES REQUIREMENTS SATISFIED

The goal is to ready the X² centre console

for series production. The simulations
required for this goal have already been
completed. These include testing of warp-
ing properties, filling pressure, crash
behaviour and stiffness load cases. For the
latter, the forces acting on multiple loca-
tions of the X² centre console were deter-
mined using the finite element method,
FIGURE 4. The result: All values comply
with the standard. The same is true of the
other simulations mentioned, all of which FIGURE 3 Exploded view of the new X² centre console with an integral design that combines the outer
satisfy all OEM requirements. surface and formative structural elements into a single unit with little parts
The ease of assembly was also tested
for series production compatibility. For
Load case
this the developers focused on clip at position 5
(500 N)
assembly ensuring efficient production
without compromising on crash or sta- Load case
Load case at
bility requirements. An interlocking at position 6
position 4
(200 N)
(800 N) Load case
clamp mount enables screw-free fasten- at position 8
ing in the vehicle. Load case at Load case (300 N)
position 3 at position
(800 N) 7 (200 N)

SUMMARY Load case

Load case Load case at
at position 1 position 2 at position 9
Lightweight design offers huge potential. (800 N) (800 N) (500 N)
This is especially true when applied
inventively as is the case with the inte-
gral design approach of Dräxlmaier
Group’s X² centre console. It is a good Load case at
example of systematic lightweight position 10
(500 N)
design, thanks to a combination of the Load case at
position 11
surface and structure by merging the (500 N)
Load case at
shell and support element design position 12
(500 N)
approaches. In conjunction with topol-
ogy optimisation of the components in Load case at
the interests of constructive lightweight position 13
(200 N)
design and the use of light materials, it is
possible to achieve weight savings of
FIGURE 4 FEA load case simulation – pressure was applied to the X² centre console with a piston
roughly 30 %. In its base configuration, to simulate load cases; the result: the measurement values in N exceed the required values at
the X² centre console is roughly 1.5 kg all 13 positions
lighter than centre consoles currently
installed in mid-size vehicles.
Numerous analytical studies for con-
firming the series production compatibil- REFERENCES
ity of the interior concept have been [1] N. N.: Mittelkonsolen – Aktuell Produkte. In: rior with an Integrated Design.] In: Springer für Pro-
ATZ 116 (2014), No. 11, pp. 30 fessionals, http://www.springerprofessional.de/mit-
completed, and several patents have [2] Hofacker, A.: Leichtbau: Mit integraler Bauweise integraler-bauweise-spart-draexlmaier-gewicht-im-
already been granted. A number of spart Dräxlmaier Gewicht im Innenraum. [Light- innenraum/5302332.html, access date: 25
OEMs have already shown interest. weight Design: Dräxlmaier Saves Weight in the Inte- November 2014

03I2015 Volume 117 27

DE VELO PMENT S OF T WARE

© StreetScooter GmbH
Reduction of Total Cost
of Ownership by Use of
Electric Vehicles

Today in the commercial sector vehicles with internal combustion engines come
overwhelming to use. To convince the fleet operator that electric vehicles can also
be economically attractive, a new software was developed by the Department of
Production Engineering E-Mobility Components of the RWTH Aachen, which can
be determine the optimum combination from conventional and electric vehicles in
vehicle fleets.

AUTHORS

Prof. Dr.-Ing. Achim Kampker Dr.-Ing. Christoph Deutskens Philip Müller, M. Sc. Dr. rer. nat. Thomas C. M. Müller
is Head of the Chair of Production is Chief Engineer of the Chair is Research Assistant at the Centre is Post-Doctoral Researcher at
Engineering of E-Mobility of Production Engineering of for International Automobile IEK-5 Photovoltaics at
Components at the RWTH Aachen E-Mobility Components at the RWTH Management at the RWTH Aachen the Forschungszentrum Jülich
University (Germany). Aachen University (Germany). University (Germany). (Germany).

28
MOTIVATION It is the capacity of the battery that file but also on the battery technology in
often limits the operational range of an use [9]. Finally, scenarios on how resid-
Electromobility is still in its starting electric vehicle. In a corporate fleet with ual values of electric vehicles develop
phase. In 2013, the Federal Motor Trans- calculable routes, however, this does not over time and usage are considered.
port Authority recorded a total of 6051 have to be an elimination criterion. Bat-
initial registrations of electric vehicles in teries that cannot be used in vehicles for
FUNCTIONING OF THE SOFTWARE
Germany. While this makes Germany one common ranges due to decreased capac-
of the leading countries in the field of ity can then be used in vehicles operat- Initially, the software is supplied with
electromobility in Europe, this figure rep- ing on shorter routes. Such an approach the direct and indirect costs of the differ-
resents merely 0.2 % of all newly regis- – combined with a selective use of initial ent types of vehicles in operation. Then,
tered cars [1]. The lack of customer battery capacities (number of battery all the vehicle models that are consid-
acceptance is still a crucial problem. How modules) – leads to a reduction of the ered by the decision maker are selected.
can it be achieved, that by experiencing total cost of ownership of a vehicle fleet Macroeconomic factors can be adjusted
significant growth in quantities electro- despite persistently high battery costs. in order for country-specific conditions
mobility will be established as a main The objective decision criterion has to to be considered. Examples for these fac-
factor in mobility? One promising strategy be the minimisation of total cost of own- tors include revenue tax rates, main refi-
for transforming this niche product into a ership. The Chair of Production Engi- nancing rates as well as developments in
mass product is the use of electric-pow- neering of E-Mobility Components at the prices of liquid fuels and electric
ered vehicles in corporate vehicle fleets. RWTH Aachen University has developed energy in the respective country.
According to the association of brand- a computer algorithm that allows for this DFO works with the entire range of
independent vehicle fleet management minimisation problem to be solved for occurring direct costs such as price of the
companies (VMF) there are approxi- any arbitrary composition of any vehicle vehicle, taxes, residual value of the vehi-
mately 1.6 million corporate vehicle fleets fleet. This dynamic and high-dimen- cle, periodic duties, as well as all relevant
in Germany, consisting of about 4.4 mil- sional algorithm for calculating an indirect costs such as cost of energy,
lion corporate vehicles (passenger vehi- investment plan is specially fitted to capacity of the battery, maintenance
cles and vans) [2]. For companies it is the vehicle pools operating on calculable costs, vehicle taxes, and CO2 taxes. Fur-
cost of the new technology that mainly route profiles. Interdependencies in oper- thermore, the software allows the user to
drives the decision whether to invest in ating times of vehicles as well as numer- choose whether battery replacements
electromobility or not. The new software ous modelling parameters for possible shall be allowed or not and to include
Dynamic Fleet Optimizer (DFO) devel- combinations are considered. Simply vehicles with varying battery configura-
oped at the Chair of Production Engineer- speaking, the software aims at providing tions (different capacities) in the algo-
ing of E-Mobility at RWTH Aachen Uni- cost minimal mobility for a given com- rithm. Based on the extensive dataset
versity shows that electromobility can be mercial route profile. The obtained combined with the mileage and route pro-
an economically attractive choice for investment plan also accounts for the files of the fleet operator, cost-optimal
companies. Hence, implementing electric findings of the latest research on this decisions concerning the implementation
vehicles in the company’s vehicle pool topic. One example is the decreasing of electric vehicles are made by the soft-
can be rationally justified instead of being capacity of a battery over its lifetime ware. The time horizon can be freely
a solely emotional decision. depending not only on the operation pro- selected and it is only limited by comput-

FORESIGHTED RATIONAL
DECISIONS CONCERNING 1 0.2 0.3
ELECTROMOBILIT Y
0.25
When deciding on a drive technology for 0.8
Specific energy consumption

0.15
carbon dioxide [kg/km]

the company’s vehicle fleet, the fleet

Specific emission of

0.2
Specific costs

manager is ultimately responsible for the 0.6

[kWh/km]

[€/km]

incurred costs. Several comparisons of 0.1 0.15

total cost of ownership (TCO) of compet-
0.4
ing drive concepts reveal, that these 0.1
costs are anything but trivial [3-8]. This 0.05
0.2
concept compares the total costs of oper- 0.05
ating a certain kind of vehicle summed
over its entire lifecycle. When applied to 0 0 0
an entire vehicle fleet consisting of dif-
an

ity

ay

an

ity

ay

an

ty

ay
i
w

w
rb

rc

rb

rc

rb

rc
ee

ee

ee
te

te

te
U

ferent types, however, the decision

Fr

Fr

Fr
In

In

In

becomes much more complex: What are

the impacts of adding a number of elec- Conventional vehicle Electric vehicle
tric vehicles to an existing fleet, which
FIGURE 1 The individual route profile has to be considered: while an electric vehicle is particularly
reliably operates a fi xed set of routes in advantageous in urban areas, this advantage becomes less important for intercity or freeway routes;
its current configuration? a conventional vehicle shows an opposing trend

03I2015 Volume 117 29

DE VELO PMENT S OF T WARE Total costs of ownership Carbon dioxide emissions
Use of Use of
electric vehicles electric vehicles

ing power. The derived decisions might

concern sales of current vehicles, pur- 3.980 -5.21 % 3.773 -9.24 %
1816 t 1648 t
chases of new vehicles, as well as battery Mio. € Mio. €
replacements. The dynamic calculation

Special energy
Special energy

contract
under consideration of arbitrary risk fac-

-7,92 %
-2.90 %
contract

+1.28 %
-1.92 %

-7

-1
.9

6.
tors therefore offers a cost-optimal vehicle

4
%

%
combination fulfilling the operator’s route
profile at all times, FIGURE 1. Free param-
eterisation of all variables, scientifically
3.904 -6.15 % 3.664 -17.5 %
founded scenarios, and the possibility to Mio. € Mio. €
1839 t 1518 t
include guarantees for residual battery
capacities and leasing models facilitate
FIGURE 2 Possible savings in total cost of ownership as well as carbon dioxide (left: potential reduction of
very high resolutions of reality when
total cost of ownership by use of electric vehicles – further reductions may be possible if special con-
compared to competing methods of tracts with energy suppliers can be negotiated; right: potential reduction of total cost of ownership by use
optimisation. of electric vehicles with simultaneous implementation of special contracts with energy suppliers)

REDUCTION OF TOTAL COST

OF OWNERSHIP IN FLEETS IS
the Corporate Vehicle Observatory Plat- [3] Al-Alawi, B. M.; Bradley, T. H.: Total cost of
POSSIBLE ownership, payback, and consumer preference
form in 2013 only 18 % of the 3,652 modeling of plug-in hybrid electric vehicles. In:
The purpose of this software is to examine responding enterprises stated that the Applied Energy 103 (2013), pp. 488-506
whether including electric vehicles in cor- future use of electromobility was an [4] Ernst, C.-S.; Hackbarth, A.; Madlener, R.; Lunz,
B.; Sauer, D. U.; Eckstein, L.: Battery sizing for serial
porate vehicle fleets can lead to cost sav- option for them at all [10]. When making plug-in hybrid electric vehicles: A model-based eco-
ings and if so, which investment plan is a decision about electric vehicles in cor- nomic analysis for Germany. RWTH Aachen, FCN
Working Paper 14/2010, Aachen, 2011
optimal. Based on this algorithm’s results, porate vehicle pools, an objective evalua- [5] Neubauer, J.; Brooker, A.; Wood, E.: Sensitivity
portfolio-optimisation in former customer tion of quantifiable factors using the latest of battery electric vehicle economics to drive pat-
terns, vehicle range, and charge strategies. National
projects leads to calculated cost savings of findings from research as well as from Renewable Energy Laboratory, Golden, USA, 2012
5 to 10 %, FIGURE 2. Another effect – practical application is essential. The [6] Plötz, P.; Gnann, T.; Wietschel, M.: Total Owner-
ship Cost Projection for the German Electric Vehicle
besides the reduction of total operating combination of conventional and electric Market with Implications for its Future Power and
cost – is the decreased emission of CO2 vehicles calculated by the DFO software, Electricity Demand. Fraunhofer-Institut für System-
und Innovationsforschung, Karlsruhe, 2012
by up to one-fifth. Depending on the price allows for an optimal use of the specific [7] Redelbach, M.; Propfe, B.; Friedrich, H.:
of electric energy the optimal share of advantages of all possible alternatives. Competitive Cost Analysis of Alternative Powertrain
Technologies. Institut für Fahrzeugkonzepte,
electric vehicles in a fleet varies between When it comes to corporate vehicle fleets, Deutsches Zentrum für Luft- und Raumfahrt e. V.,
30 and 50 %, FIGURE 3. In FIGURE 3 the electromobility is not a yes or no decision. Stuttgart, 2012
[8] Sharma, R. et al.: Conventional, hybrid and elec-
electric vehicle StreetScooter Compact of
tric vehicles for Australian driving conditions – Part
the StreetScooter GmbH and the model REFERENCES 1: Technical and financial analysis. School of Infor-
[1] www.kba.de/cln_031/nn_124832/SharedDocs/ mation Technology and Electrical Engineering, The
Polo of the Volkswagen AG, with a 1.2-l Publikationen/PM/2014/pm_ _02_ _14_ _fahrzeugzu- University of Queensland, Brisbane, Australia, 2012
engine for both cases (petrol engine or die- lassungen_ _12_ _2013_ _pdf: Fahrzeugzulassungen [9] Ecker, M.; Nieto, N.; Käbitz, S.; Schmalstieg, J.;
sel engine), are compared. The results im Dezember 2013. Kraftfahrt-Bundesamt, balance Blanke, H.; Warnecke, A.; Sauer, D.: Calendar and
sheet, 24.3.2014 cycle life study of Li(NiMnCo)O2-based 18650 lith-
show that even in the very short term an [2] www.vmf-fuhrparkmanagement.de/branche. iumion. In: Journal of Power Sources 248 (2014),
addition of electric vehicles of 58 % in the html: Markt Deutschland: Überschaubar und mit pp. 839-851
wachsender Bedeutung. Verband markenunabhäng- [10] N. N.: CVO Fuhrpark-Barometer 2013. Arval
fleet is optimal. Special contracts with iger Fuhrparkmanagementgesellschaften e. V., Deutschland GmbH, Kirchheim, 2013
energy suppliers were assumed to be in 24.3.2014

effect. The conducted calculations show

that in a framework of fleet scenarios elec-
tromobility is indeed an economically via- 100 Petrol-powered vehicle
ble option. It is not necessary to show that 90 Electric vehicle
a given electric vehicle is able to cover any Diesel-powered vehicle
80
route, but only that it is able to cover a spe-
cific part of all routes at significantly lower 70
Number of cars [-]

cost. This fundamental realisation suffices 60

to grasp the full potential of electromobil- 50
ity in commercial fleets.
40

30
SUMMARY
20
Still, not only the vision of the decision
10 FIGURE 3 Composition of
maker is crucial, but electromobility also the vehicle fleet over time
has to be recognised as a viable option in 0 after dynamic optimisation
102
108
114
120
12
18
24
30
36
42
48
54
60
66
72
78
84
90
96
0
6
-6

the first place. In a survey carried out by was applied at time 0

Time [months]
30
 Driver Assistance Systems
From Assistance to Automated Driving

1st International ATZ Conference www

.ATZ
liv e.de

28 and 29 April 2015

Frankfurt am Main | Germany
ROAD
MA
Corre P FOR AUT
c OMAT
the b tly asses ION
asic sing
cond
ition
SIMU s
LAT
New ION & TES
meth TIN
ods a G
nd s
SAFET tand
Y&S ards
Fu n c EC
tiona URITY
put to l safe
the te ty
st
ROADMAP FOR AUTOMATION
Correctly assessing /// K
EYN
OTE
SPEE
Step CHE
the basic conditions Prof.
han W
Dr. Jü
olfsri
S
ed, D
aimle
Karlh rgen r
einz Leoh
Haup old,
Dr. M t, Co VW
atthia ntine
s Kla ntal
SIMULATION & TESTING uda,
Robe

OM
rt Bo

OTO.C
sc h
New methods and standards

CKPH
/ISTO
ENG
Simult

UDIF
a
SAFETY & SECURITY Drive Germa neous Inter

© ZH
pr
Functional safety r Ass n and
Englis eting
From istan h
put to the test
Assist
ance t ce S
1st In
te
28 an rnational AT
o Auto
mated ystem
d 29 A
pril
Z Con
fere
Drivin
g s
2015 nce
| Fran
kfurt a
m Ma
in | Ge
/// KEYNOTE SPEECHES rmany

Stephan Wolfsried, Daimler

Prof. Dr. Jürgen Leohold, VW
Karlheinz Haupt, Continental
Dr. Matthias Klauda, Robert Bosch

/// KINDLY SUPPORTED BY

Phone +49 611 7878-131 PROGRAM AND REGISTRATION

Abraham-Lincoln-Straße 46
65189 Wiesbaden | Germany
Fax +49 611 7878-452
[email protected] www.ATZlive.com
DE VELO PMENT MATERIAL S

Lignin Powder as a Filler for Thermoplastic

Lightweight Design Components

The use of the wood material lignin as a filler instead of ash, talcum or glass fibres
in thermoplastics offers an enormous lightweight potential. Additionally, life cycle
analysis, sustainability and CO2 emissions improve when using this organic nature
product. Volkswagen, the TU Dresden and the Faserinstitut Bremen investigated the
effect of lignin powder as filler in polypropylene (PP) with different weight fractions.
Here, the safe-guarding of the mechanical properties of the new compound is
essential also for series production.

32
INTRODUCTION an established material pre-validation AUTHORS
process as well as a final economic effi-
In recent years, lightweight design has ciency and sustainability analysis.
gained enormous importance in the Furthermore, the process ability of the
automotive industry. Especially, renewa- new products as well as the suitability
ble and sustainable low-density materi- for high volume production of the devel-
als are in the focus for applications. The oped processes are investigated as main
low density of these materials allows issues for successful implementation in Dipl.-Ing. Hendrik Mainka
reducing the structural weight of end future vehicle concepts. is Project Engineer in the Materials
products, which can help to save energy and Manufacturing Processes
Division at the Volkswagen AG in
and thus CO2 emissions.
EXPERIMENTAL SETUP Wolfsburg (Germany).
To reduce the weight and the cost of
AND PROCEDURE
polymers, different fillers are used. Cur-
rently, inorganic materials like glass In this study hardwood lignin from a
fibres or balls, soot, talcum, aluminum bio refinery was used. PP with high
oxides and silicate but also organic crystallinity for automotive injection
materials like wood powder and cellu- moulding with a Melt Volume-flow Rate
lose are applicable as filler [1, 2]. (MFR) of 15 g per 10 min was used for
Maria Laue
The low cost and abundant availability blending with lignin at 230 °C and is Graduand at the Institute of
makes lignin attractive as filler in auto- 2.16 kg load. Textile Machinery and High
motive industry. Lignin is the second The lignin powder was dried for 24 h at Performance Material Technology
(ITM) at the TU Dresden (Germany).
most abundant polymers from biomass in 50 °C before extrusion process. Polymer
nature after cellulose. The three dimen- and powder were compounded using a
sional structure consists of p-hydroxyphe- Leisteritz ZSE 40 DL co-rotating twin-
nyl, guaiacyl and syringyl. These are screw extruder. PP and PP resin (MAPP),
cross-linked differently again [3, 4]. which is grafted with maleic anhydride
Lignin is obtained just as by-product in (MAH) adhesion promoter, were weighted
the paper and biofuel industry and serves and dry-mixed to be added to feeder 1.
Univ.-Prof. Dr.-Ing. habil.
for energy generation. Though in the The lignin powder was added to a sepa-
Dipl.-Wirt.-Ing. Chokri Cherif
recent years, many researches were pub- rate feeder 2. Feeding of the materials was is Director of the Institute of Textile
lished [5-7] about lignin as filler material. performed by K-tron gravimetric feeders. Machinery and High Performance
Lightweight design is an essential part The screw speed was 100 rpm and the Material Technology (ITM) at the
TU Dresden (Germany).
of the overall Volkswagen strategy reduc- temperature at the extruder die varied
ing the CO2 emissions. The use of the between 175 and 190 °C. Afterwards the
wood material lignin as a filler in thermo- extruded material cooled down and was
plastics like polypropylene (PP) offers an shredded to pellets.
enormous lightweight potential. Here, a Producing PP/lignin resin with and
PP/lignin compound filled with up to without MAH adhesion promoter, the
30 % lignin powder offers a 20 % weight same parameters were used. The differ-
reduction compared to traditional filled PP ent material formulations of the seven Prof. Dr.-Ing. Axel S. Herrmann
is Director of the Faserinstitut
compounds assuring the same mechanical charges are shown in TABLE 1. The Bremen e. V. in Bremen
performance. Furthermore, in comparison amount of lignin powder in the com- (Germany).
to unfilled thermoplastics a potential cost posites is given in weight per cents. All
reduction potential of up to 30 % by using results in the figures and tables are given
lignin as filler seems possible. Today, the as mean values.
use of lignin as filler for thermoplastic It is essential for all tensile, three-point
materials in automotive components in bending, and notch impact as well as
series production applications is unknown.
The central subject of this study are
the mechanical properties of PP/lignin
compounds with up to 30 weight %. Charge/sample PP [%] Lignin [%] MAH [%]

They were measured with tensile, three- 1 100 – –

point bending and impact strength test. 2 89 10 1
Key aspects for the investigation of 3 78 20 2
novel lignin based fillers are: the exami- 4 67 30 3
nation and quantification of lignin, the
5 90 10 –
optimisation of the manufacturing pro-
6 80 20 –
cesses, the characterisation and quantifi-
cation of the mechanical properties of the 7 70 30 –

novel lignin filled thermoplastics within TABLE 1 Composition of the seven PP/lignin charges with and without MAH adhesion promoter

03I2015 Volume 117 33

DE VELO PMENT MATERIAL S

FIGURE 1 Tensile moduli and tensile strength of all PP/lignin compounds: tensile moduli (left) and tensile strengths (right)

hardness tests: Before testing, all speci- of MAH. Also, the increasing lignin con- strain by almost 80 %. With 30 % lignin
mens were conditioned for at least 72 h at tent causes high standard deviations. weight blend the elongation at fracture is
a temperature of 23 °C and a relative This indicates an inhomogeneous distri- 3.4 %, which is only 0.67 % of the start-
humidity of 50 %. bution of the lignin in the PP melt. The ing value. 10 % lignin content without
The extruded material was com- tensile and break stress values of all addition of MAH shows a similar
pression moulded into samples for differ- material blends are nearly similar with decreasing of the value.
ent mechanical testing. Compression some exceptions. The MAH adhesion promoter seems
moulding was performed at 185 °C with However, the elongations at break not to influence the homogeneity in the
350 bar pressure using an Arburg All- decrease with increasing lignin content. specimens because the standard devia-
rounder 420 C. It was worked with a Just 10 % lignin causes a deterioration of tions are stable.
dwell pressure at 260 bar for 20 s.
Tensile properties were measured on
a tensile test machine from ZwickRoell.
15 specimens of each blend were tested. Lignin content Tensile modulus Tensile strength Strength at break Elongation at
Matrix
[weight %] [MPa] [MPa] [MPa] break [%]
The measurement of the strain changes
were detected optical. PP 0 1326 24.6 19.6 508.3
All three-point bending tests for
PP and MAH 10 1237 21.7 13.1 98.6
flexural properties were tested on the
Zmart.pro static material testing PP and MAH 20 1648 23.3 13.8 13.8
machine from the ZwickRoell company.
PP and MAH 30 1845 23.3 19.1 3.4
Ten specimens of each blend were tested.
Impact strength of notched samples was PP 10 1302 22.2 13.5 94.5
tested on a ZwickRoell instrument with a PP 20 1592 21.5 9.0 53.0
2-joule pendulum. Ten specimens of each
blend were tested. TABLE 2 Mechanical properties of all PP/lignin compounds
The hardness of all samples was deter-
mined with Shore D in the tests.

MEASUREMENT RESULTS
AND DISCUSSION

To compare and assess the mechanical

properties of all blended materials with
each other different methods like tensile,
three-point bending and impact strength
test were performed.
The results of tensile strength proper-
ties are shown in FIGURE 1. It becomes
obvious that the tensile moduli values
increase with increasing lignin content.
FIGURE 2 Impact
Furthermore there is no significant strength of all PP/lignin
enhancement of strength with addition compounds

34
FIGURE 3 Flexural moduli (left) and flexural strengths (right) of all PP/lignin compounds

In summary, lignin has a positive 2 % MAH is higher than the measured SUMMARY
effect on the tensile moduli. However, strength with 30 % lignin.
the tensile strength values decrease com- Based to flexural strain, you can see In this study Volkswagen, TU Dresden
pared with unfilled PP. It is conceivable also that the materials are brittle with and Faserinstitut Bremen investigated
to use lignin as filler for cost reduction. high lignin content because the values the effect of lignin powder as filler in
TABLE 2 shows the results of the mechani- decrease. The decrease is less for the ten- polypropylene with different weight
cal properties. sile tests. Unfilled PP shows the highest fractions. Furthermore there were strong
The results of the impact strength value with 6.3 % flexural strain. In con- improvements of the mechanical pro-
tests are shown in FIGURE 2. The values trast, the sample with 30 % lignin con- perties if the MAH adhesion promoter
of the impact strengths with and with- tent has just 3.6 % flexural strain. is added.
out MAH continuously decrease with In summary, the addition of lignin has In general the properties of the dif-
increasing lignin content. But without a increase effect on the flexural moduli. ferent PP/lignin compounds are influ-
addition of MAH the deterioration is However the strength values decrease enced by the lignin content. If filler was
considerably lower. It can be concluded with 10 % lignin compared to unfilled included the materials were stiffer. The
that the function of MAH as adhesion PP. With 20 % filler or more the flexural best values were measured with 20 %
promoter have a negative effect on the strength increases. Therefore the state- lignin content and MAH. However, the
impact strength of the compound of PP ments so far support this results and it scope of applications still has to be
and lignin. is conceivable to use lignin as filler for defined for the automotive industry.
The flexural moduli increase with cost reduction. Moreover it has to be investigated if
increasing lignin content, which was The hardness of the different blended lower values without adding MAH are
noted already in the tensile strength polymers was measured with shore D. The acceptable for the final utilisation, so
test, FIGURE 3. The flexural strength values are shown in FIGURE 4. It can be that further cost can be reduced.
varies to a lesser extent whereas notice- seen that there is no significant increase
able that the value with 20 % lignin and of hardness with addition of MAH. REFERENCES
[1] Myer Kutz (Ed.): Applied Plastics Engineering
Handbook: Processing and Materials. London,
Amsterdam, New York: Elsevier, 2011
[2] Baur, E. et al.; Saechtling, H. (Hrsg.):
Saechtling Kunststoff-Taschenbuch. München:
Carl Hanser, 2013
[3] Laurichesse S.; Avérous, L.: Chemical Modification
of Lignins: Towards Biobased Polymers. In: Progress
in Polymer Science 39 (2014), No. 7, pp. 1266-1290
[4] Hatakeyama, H.; Hatakeyama, T.: Biopolymers.
pp. 1-63, Berlin, Heidelberg: Springer, 2010
[5] Morandim-Giannetti, A. et al.: Lignin as Additive
in Polypropylene/Coir Composites: Thermal, Mechan-
ical and Morphological Properties. In: Carbohydrate
Polymers 87 (2012), No. 4, pp. 2563-2568
[6] Sailaja, R. R. N. et al.: Mechanical and Thermal
Properties of Compatibilized Composites of Poly-
FIGURE 4 Results of ethylene and Esterified Lignin. In: Materials &
Shore D hardness Design 31 (2010), No. 9, pp. 4369-4379
determination for PP/ [7] Nitz, H. et al.: Influence of Lignin Type on the
lignin compounds with Mechanical Properties of Lignin Based Compounds.
and without MAH In: Macromolecular Materials and Engineering 286
adhesion promoter (2001), No. 12, pp. 737-743

03I2015 Volume 117 35

IMPRINT

W O R L D W I D E

www.ATZonline.com

Founded 1898 as „Der Motorwagen“ Organ of the VDI-Gesellschaft Fahrzeug- und Verkehrstechnik (FVT)
Organ of the Forschungsvereinigung Automobiltechnik e. V. (FAT) and of the Normenausschuss Automobiltechnik
(NAAutomobil), formerly known as FAKRA, in the DIN Deutsches Institut für Normung e. V.
Organ of the Wissenschaftliche Gesellschaft für Kraftfahrzeug- und Motorentechnik e. V. (WKM)

03I2015 _ Volume 117

Springer Vieweg | Springer Fachmedien Wiesbaden GmbH
P. O. Box 15 46 · 65173 Wiesbaden · Germany | Abraham-Lincoln-Straße 46 · 65189 Wiesbaden · Germany

Amtsgericht Wiesbaden, HRB 9754, USt-IdNr. DE811148419

Managing Directors Armin Gross, Petrus W. J. Hendriks, Joachim Krieger

Managing Director Marketing & Sales Armin Gross | Director Magazines Stefanie Burgmaier | Director Production Olga Chiarcos

EDITORS IN CHARGE ADDRESS

SCIENTIFIC ADVISORY BOARD Dr. Johannes Liebl, Wolfgang Siebenpfeiffer Abraham-Lincoln-Straße 46 · 65189 Wiesbaden
P. O. Box 1546 · 65173 Wiesbaden, Germany
YOUR HOTLINE
Dipl.-Ing. Dietmar Bichler EDITOR IN CHIEF
Bertrandt AG Dr. Alexander Heintzel
[email protected]
TO ATZ
phone +49 611 7878-342 · fax +49 611 7878-462
ADVERTISING
Dipl.-Ing. Kurt Blumenröder [email protected] Editorial Staff
IAV GmbH SALES MANAGEMENT
VICE-EDITOR IN CHIEF
Dipl.-Ing. Michael Reichenbach Volker Hesedenz ☎ +49 611 7878-341
Prof. Dr.-Ing. Lutz Eckstein phone +49 611 7878-269 · fax +49 611 7878-78269 Customer Service
phone +49 611 7878-341 · fax +49 611 7878-462
RWTH Aachen, WKM [email protected] [email protected]
☎ +49 6221 3454-303
Dr.-Ing. Ulrich Eichhorn MANAGING EDITOR KEY ACCOUNT MANAGEMENT Advertising
Rouwen Bastian
VDA/FAT Kirsten Beckmann M. A.
phone +49 611 7878-343 · fax +49 611 7878-462 phone +49 611 7878-399 · fax +49 611 7878-78399
☎ +49 611 7878-269

Dr. rer. nat. Andreas Eilemann [email protected] [email protected]

Mahle Behr GmbH & Co. KG DISPLAY AD MANAGER
EDITORIAL STAFF
Klaus Fröhlich Dipl.-Ing. (FH) Richard Backhaus Susanne Bretschneider HINTS FOR AUTHORS
phone +49 611 5045-982 · fax +49 611 5045-983 phone +49 611 7878-153 · fax +49 611 7878-443 All manuscripts should be sent directly to the editors.
BMW AG
[email protected] [email protected] By submitting photographs and drawings the sender
Prof. Dr.-Ing. Burkhard Göschel Dipl.-Ing. (FH) Andreas Fuchs releases the publishers from claims by third parties.
AD PRICES
Burkhard Goeschel Consultancy phone +49 6146 837-056 · fax +49 6146 837-058 Only works not yet published in Germany or abroad can
Advertising ratecard from October 2014.
[email protected] generally be accepted for publication. The manuscripts
Prof. Dr.-Ing. Peter Gutzmer Dipl.-Ing. Ulrich Knorra must not be offered for publication to other journals
MARKETING | OFFPRINTS
Schaeffler AG phone +49 611 7878-314 · fax +49 611 7878-462 simultaneously. In accepting the manuscript the pub-
[email protected] HEAD OF MARKETING lisher acquires the right to produce royalty-free off-
Prof. Dr.-Ing. Ulrich Hackenberg Roland Schedel MAGAZINES + EVENTS prints. The journal and all articles and figures are pro-
Audi AG phone +49 6128 853-758 · fax +49 6128 853-759 Jens Fischer tected by copyright. Any utilisation beyond the strict
[email protected] limits of the copyright law without permission of the
PRODUCT MANAGEMENT publisher is illegal. This applies particularly to duplica-
Dipl.-Ing. Wolfgang Hatz Stefan Schlott
AUTOMOTIVE MEDIA tions, translations, microfilming and storage and pro-
Dr. Ing. h.c. F. Porsche AG phone +49 8726 967-572
Jens Fischer cessing in electronic systems.
[email protected]
Prof. Dr.-Ing. Pim van der Jagt phone +49 611 7878-340 · fax +49 611 7878-407
Markus Schöttle
[email protected]
Ford-Forschungszentrum Aachen GmbH phone +49 611 7878-257 · fax +49 611 7878-462
[email protected] OFFPRINTS
Dipl.-Ing. Bernd Maierhofer Martina Schraad Martin Leopold
MAN Truck & Bus AG phone +49 611 7878-276 · fax +49 611 7878-462 phone +49 2642 907-596 · fax +49 2642 907-597
[email protected] [email protected]
Joachim Mathes
PERMANENT CONTRIBUTORS
Valeo Schalter und Sensoren GmbH PRODUCTION | LAYOUT
Andreas Burkert, Prof. Dr.-Ing. Stefan Breuer, Harmut
Hammer, Detlef Krehl, Prof. Dr.-Ing. Fred Schäfer Kerstin Brüderlin
Dr.-Ing. Harald Naunheimer
phone +49 611 7878-173 · fax +49 611 7878-78173
ZF Friedrichshafen AG ENGLISH LANGUAGE CONSULTANT [email protected]
Paul Willin
Dipl.-Ing. Jörg Ohlsen
ONLINE | ELECTRONIC MEDIA SUBSCRIPTIONS
Edag GmbH & Co. KGaA
Portal Manager Automotive Springer Customer Service Center GmbH
Prof. Dr. Dipl.-Ing. Peter Pfeffer Christiane Brünglinghaus Haberstraße 7 · 69126 Heidelberg, Germany
phone +49 611 7878-136 · fax +49 611 7878-462 phone +49 6221 3454-303 · fax +49 6221 3454-229
Hochschule München
[email protected] Monday to Friday, 8 a.m. to 6 p.m.
Dr.-Ing. Peter E. Rieth Editorial Staff [email protected]
Continental Teves AG & Co. oHG Katrin Pudenz M. A. www.my-specialized-knowledge.com/automotive
phone +49 6172 301-288 · fax +49 6172 301-299
Wolf-Henning Scheider [email protected] SUBSCRIPTION CONDITIONS
Robert Bosch GmbH The eMagazine appears 11 times a year at an annual
SPECIAL PROJECTS subscription rate 199 € for private persons and 333 €
Prof. Dr.-Ing. Rodolfo Schöneburg Managing Editorial Journalist for companies. Special rate for students on proof of
Markus Bereszewski status in the form of current registration certificate
VDI-FVT
phone +49 611 7878-122 · fax +49 611 7878-462 98 €. Special rate for VDI/ÖVK members on proof of
Dipl.-Ing. (FH) Dipl.-Wirtsch.-Ing. [email protected] status in the form of current member certificate 172 €.
Editorial Staff I Coordination Special rate for studying VDI members on proof of
(FH) Wolfgang Schwenk
Dipl.-Reg.-Wiss. Caroline Behle status in the form of current registration and member
Adam Opel AG phone +49 611 7878-393 · fax +49 611 7878-462 certificate 71 €. Annual subscription rate for combina-
[email protected] tion ATZworldwide (eMagazine) and ATZ (print) 399 €.
Prof. Dr.-Ing. Thomas Weber All prices include VAT at 7%. Every subscription comes
© Springer Vieweg |
Christiane Imhof M. A. Springer Fachmedien Wiesbaden GmbH,
Daimler AG phone +49 611 7878-154 · fax +49 611 7878-462 with access to the online archive. However, access is
Wiesbaden 2015
[email protected] only available for the individual subscription holder.
Dr.-Ing. Christian Wiehen To obtain access for your entire company/library/organi- Springer Vieweg is part of
Wabco GmbH PROJECT MANAGEMENT I ASSISTANCE sation, please contact [email protected] Springer Science+Business Media.
Yeliz Konar or phone +49 611 7878-686. The subscription can be
Prof. Dr. rer. nat. Martin Winterkorn phone +49 611 7878-180 · fax +49 611 7878-462 cancelled in written form at any time with effect from
Volkswagen AG [email protected] the next available issue.
 PEER RE VIE W RESE ARCH

PEER REVIEW

Peer Review Process for Research Articles

in ATZ, MTZ and ATZelektronik
STEERING COMMITTEE
Scientific articles of universities in ATZ
Prof. Dr.-Ing. Lutz Eckstein RWTH Aachen University Institut für
Kraftfahrzeuge Aachen
Automobiltechnische Zeitschrift, MTZ
Motortechnische Zeitschrift and ATZelek-
Prof. Dipl.-Des. Wolfgang Kraus HAW Hamburg Department Fahrzeugtechnik
und Flugzeugbau tronik are subject to a proofing method, the
so-called peer review process. Articles
Prof. Dr.-Ing. Ferit Küçükay Technische Universität Institut für Fahrzeugtechnik
Braunschweig accepted by the editors are reviewed by
Prof. Dr.-Ing. Stefan Pischinger RWTH Aachen University Lehrstuhl für
experts from research and industry before
Verbrennungskraftmaschinen publication. For the reader, the peer review
Prof. Dr.-Ing. Universität Stuttgart Institut für Verbrennungs- process further enhances the quality of the
Hans-Christian Reuss motoren und Kraftfahrwesen magazines‘ content on a national and
Prof. Dr.-Ing. Ulrich Spicher MOT international level. For authors in the insti-
tutes, it provides a scientifically recognised
Prof. Dr.-Ing. Hans Zellbeck Technische Universität Dresden Lehrstuhl für
Verbrennungsmotoren publication platform.
In the Peer Review Process, once the
ADVISORY BOARD editors has received an article, it is
reviewed by two experts from the Advisory
Prof. Dr.-Ing. Klaus Augsburg Dr. Malte Lewerenz Board. If these experts do not reach a
Prof. Dr.-Ing. Michael Bargende Prof. Dr.-Ing. Markus Lienkamp unanimous agreement, a member of the
Prof. Dipl.-Ing. Dr. techn. Christian Beidl Prof. Dr. rer. nat. habil. Ulrich Maas Steering Committee acts as an arbitrator.
Prof. Dr. sc. techn. Konstantinos Boulouchos Prof. Dr.-Ing. Markus Maurer Following the experts’ recommended cor-
Prof. Dr. Dr. h.c. Manfred Broy Prof. Dr.-Ing. Martin Meywerk rections and subsequent editing by the
Prof. Dr.-Ing. Ralph Bruder Ao. Univ.-Prof. Dr. Gregor Mori author, the article is accepted.
Dr. Gerhard Bruner Prof. Dr.-Ing. Klaus D. Müller-Glaser In 2008, the peer review process utilised
Prof. Dr. rer. nat. Heiner Bubb Dr. techn. Reinhard Mundl by ATZ and MTZ was presented by the
Prof. Dr. rer. nat. habil. Olaf Deutschmann Prof. Dr. rer. nat. Peter Neugebauer WKM (Wissenschaftliche Gesellschaft für
Prof. Dr.-Ing. Klaus Dietmayer Prof. Dr. rer. nat. Cornelius Neumann Kraftfahrzeug- und Motorentechnik e. V./
Dr. techn. Arno Eichberger Prof. Dr.-Ing. Nejila Parspour German Professional Association for Auto-
Prof. Dr. techn. Helmut Eichlseder Prof. Dr.-Ing. Peter Pelz motive and Motor Engineering) to the DFG
Prof. Dr. Wilfried Eichlseder Prof. Dr. techn. Ernst Pucher (Deutsche Forschungsgemeinschaft/Ger-
Dr.-Ing. Gerald Eifler Dr. Jochen Rauh man Research Foundation) for official
Prof. Dr.-Ing. Wolfgang Eifler Prof. Dr.-Ing. Konrad Reif recognition. ATZelektronik participates in
Prof. Dr. rer. nat. Frank Gauterin Prof. Dr.-Ing. Stephan Rinderknecht the Peer Review since 2011.
Prof. Dr. techn. Bernhard Geringer Prof. Dr.-Ing. Jörg Roth-Stielow
Prof. Dr.-Ing. Uwe Dieter Grebe Dr.-Ing. Swen Schaub
Dr. mont. Christoph Guster Prof. Dr. sc. nat. Christoph Schierz
Prof. Dr.-Ing. Holger Hanselka Prof. Dr. rer. nat. Christof Schulz
Prof. Dr.-Ing. Horst Harndorf Prof. Dr. rer. nat. Andy Schürr
Prof. Dr. techn. Wolfgang Hirschberg Prof. Dr.-Ing. Ulrich Seiffert
Prof. Dr. techn. Peter Hofmann Prof. Dr.-Ing. Hermann J. Stadtfeld
Prof. Dr. rer. nat. Peter Holstein Prof. Dr.-Ing. Karsten Stahl
Prof. Dr.-Ing. Volker von Holt Prof. Dr. techn. Hermann Steffan
Prof. Dr.-Ing. habil. Werner Hufenbach Prof. Dr.-Ing. Wolfgang Steiger
Prof. Dr.-Ing. Armin Huß Prof. Dr.-Ing. Peter Steinberg
Dr. techn. Heidelinde Jetzinger Dr.-Ing. Peter Stommel
Prof. Dr.-Ing. Roland Kasper Dr.-Ing. Ralph Sundermeier
L FOR SCIENTIFI
O VA CA
Prof. Dr.-Ing. Prof. E. h. mult. Rudolf Kawalla Prof. Dr.-Ing. Wolfgang Thiemann PR RT
AP I
F

CL

Prof. Dr.-Ing. Tran Quoc Khanh Prof. Dr.-Ing. Dr. h.c. Helmut Tschöke
LO

ES
SEA

IN A

Dr. Philip Köhn Prof. Dr.-Ing. Georg Wachtmeister

THE

TZ.

Prof. Dr.-Ing. Ulrich Konigorski Prof. Dr.-Ing. Jochen Wiedemann

PEER REVIEW
|

Prof. Dr. Oliver Kröcher Prof. Dr. rer. nat. Gerhard Wiegleb
REV

Y.

Prof. Dr.-Ing. Peter Krug Prof. Dr. techn. Andreas Wimmer

STR
IEW

DU

Dr. Christian Krüger Prof. Dr. rer. nat. Hermann Winner

ED

IN

Y N
B

EX
D

Prof. Dr. techn. Thomas Lauer Prof. Dr. med. habil. Hartmut Witte PER HA
TS FR RC
Prof. Dr. rer. nat. Uli Lemmer Dr.-Ing. Michael Wittler OM RESEA

03I2015 Volume 117 37

 RESE ARCH SIMUL ATION

AUTHORS
Simulating Reliability with
Respect to Ride Comfort

Key potential areas for optimising the efficiency during vehicle development
Dipl.-Kffr. techn.
Katrin Mutter process are therefore rigidly aligned with safeguarding activities of customer
is Project Coordinator in oriented overall vehicle functions in early phases of development. Based on this,
the Team Testing Manage-
ment of the Daimler AG in the Institute of Machine Components, field of Reliability Engineering, University
Sindelfingen and PhD at
the Institute of Machine of Stuttgart, applied in cooperation with the Mercedes-Benz Cars Development
Components at the as a part of Daimler an advanced reliability concept. With this concept, the
University of Stuttgart in
the fi eld of Reliability Engi- numerical simulation methods used are capable for the first time of evaluating
neering (Germany).
and assessing the long-term quality and reliability of a vehicle’s ride comfort.

Dr.-Ing. Axel Röder

is Manager of the Team
Testing Management in the
Mercedes-Benz Cars
Development Corporate
Unit of the Daimler AG in
Sindelfingen (Germany).

L FOR SCIENTIFI
O VA CA
PR RT
AP I
F

CL

Prof. Dr.-Ing.
LO

ES
SEA

Bernd Bertsche
IN A
THE

TZ.

is Head of the Institute of

Machine Components at PEER REVIEW
|

the University of Stuttgart RECEIVED 2014-05-07

REV

Y.

(Germany). REVIEWED 2014-06-18

STR
IEW

ACCEPTED 2014-12-17
DU
ED

IN

Y N
B

EX
D

PER HA
TS FR RC
OM RESEA

Dipl.-Betriebswirt
Jörg U. Brünis
is Head of Overall Testing
Management in the
Mercedes-Benz Cars
Development Corporate
Unit of Daimler AG
in Stuttgart (Germany).

38
1 MOTIVATION 2 OVERALL VEHICLE FUNCTIONS IN THE
2 OVER ALL VEHICLE FUNCTIONS IN THE DIGITAL PROTOT YPE DIGITAL PROTOT YPE
3 STRUCTURE OF AN END -TO - END SAFEGUARDING PRO CESS
4 SIMUL ATING RELIABILIT Y WITH RESPECT TO RIDE C OMFORT The entire product development process can be divided into two
5 SUMMARY AND OUTLO OK basic phases: one digital and one physical, FIGURE 1. Initial safe-
guarding measures are implemented in the digital phase by mak-
ing calculations and running simulations only. In the physical
development phase, safeguarding occurs on test stands as well as
on the road using available parts, components and trial testing
vehicles. During street testing, globally distributed testing grounds
and public roadways are used. Digital safeguarding is also prac-
tised throughout the physical development phase to ensure plau-
sibility and validate the trial testing results obtained in addition to
gaining important information about how the simulation methods
can and should be developed further.
From a customer perspective, it is not the parts, components
or systems of a vehicle that matter so much as the overall func-
tions offered when it comes to perceiving the vehicle. These func-
tions pertain to thermal development, driving performance, con-
sumption, vehicle dynamics and ride comfort, for example,
whereby the latter two are associated with the category of ride and
handling, FIGURE 2. Overall vehicle functions are the result of the
coordinated interaction of various components and systems as well
as the complex effects they produce when using the vehicle. It is
in this context that Mercedes-Benz leverages the concept of the
digital prototype. FIGURE 2 shows separate overall vehicle functions
[2, 4]. In the digital prototype environment, corresponding overall
vehicle simulation models are introduced to assess and evaluate
a standardised, uniform state of development across all overall
vehicle functions to obtain consistent and clear results for one and
the same vehicle.

3 STRUCTURE OF AN END -TO -END

1 MOTIVATION SAFEGUARDING PROCESS

In the automotive industry, ever shorter development times, In the past, overall vehicle functions were divided, classified and
increasing levels of customisation and options for products, and safeguarded in the digital development phase only. To this end,
the resulting boom in the number of variants available, together standardised, functional safeguarding tasks in the form of load
with rising cost pressure, have led to a reduction in the number scenarios exist to safeguard the digital prototype [1, 2]. By apply-
of trial testing vehicles developed. At the same time, the require- ing these load scenarios, designers can compare the simulation
ments pertaining to the functional quality and reliability of end results for different model series as well as for varying stages of
products are at an all-time high. Key potential areas for optimising development within the confines of the development process for
the efficiency with which vehicle design work is safeguarded are a particular model series. Standardised test descriptions are also
therefore rigidly aligned with the digital, test stand and street in place to safeguard exercises carried out in the test stand and
phases. This approach also forms the basis for targeted front load-
ing to shift safeguarding activities back to early phases of
development.
Functional short-term safeguarding based on digital prototypes Concept and Testing
has already become established in the development processes simulation
practised by manufacturers [1, 2]. Safeguarding reliability in the
long term, however, remains very much a manual activity, as com- Physical development phase
Digital development process
puters have not yet been fully integrated and current component Test stand
Street
and system-related reliability techniques still do not offer a com-
Test vehicles,
prehensive concept for evaluating the durability of overall vehicle Preliminary Mature test
series
Digital prototype vehicles test vehicles vehicles
production
functions such as ride comfort and vehicle dynamics. In this con- (development) (validation)
development
text, reliability as defined in [3] refers to “the probability of a prod-
Competitor vehicles, major assembly carriers Process-related digital safeguarding
uct not failing within a defined time window under specific oper-
Series
ating and ambient conditions”. Target reliability must be adapted Product development process development
process
accordingly where it applies to overall vehicle functions (see sec-
tion 4). FIGURE 1 Digital and physical safeguarding in the product life cycle

03I2015 Volume 117 39

RESE ARCH SIMUL ATION

street environments but are not directly comparable with those of methods used are capable of evaluating and assessing the long-
the digital trial testing phase. This, in turn, makes it more difficult term quality and reliability of a vehicle’s ride comfort. Digital reli-
to hand over the safeguarding results in a binding, confidential ability is assessed by applying the constraints and evaluation cri-
manner as well as track the state of development of the safeguard- teria that originally applied to the physical development phase in
ing exercises throughout the overarching safeguarding process. carrying out trial testing exercises on test stands and on
Redundancies in the content of the safeguarding programme are roadways.
also much more difficult to identify.
In striving to improve the efficiency of the safeguarding process,
4 SIMULATING RELIABILIT Y WITH
Mercedes-Benz has extended the application content within the
RESPECT TO RIDE COMFORT
structure of the overall vehicle functions from the digital prototype
to the physical development phase that includes the test stand The design requirement of shifting the need to make clear, bind-
and street sub processes. In the target objective state, these sub ing statements about the reliability of the overall vehicle to the
processes encompass standardised, clearly defined safeguarding digital phase as a preliminary step to avoid provisioning expensive
scopes that are assigned to the phases in a well-defined manner, trial testing vehicles led to the development of an advanced reli-
whereby the scopes permit, or allow, an ideal mix of the trial test- ability concept for overall vehicle functions [5]. Based on the defi-
ing methods available. The only preliminary requirement in this nition of reliability as explained by Bertsche [3], the reliability of
context is that the methods applied have attained an appropriate overall vehicle functions was clarified as the probability of the per-
level of maturity in the three phases. formance criteria for the overall vehicle function not exceeding the
To achieve the target state described, the first step is to analyse limits of a defined target area, or range, under a given set of envi-
and compare the test descriptions across all trial testing special- ronmental and usage conditions up to a certain vehicle mileage
ist departments, TABLE 1. Doing this makes it possible to determine figure.
which part or aspect of the overall safeguarding scope is currently
being processed and how activities should be divided and assigned 4.1 THE FIVE STEPS OF THE CONCEPT
in future. In a further step, a target-actual comparison is drawn to The concept introduced in this article attempts to characterise
identify redundancies, which must then be alleviated. Inadequa- the stochastic scatter present at a mileage figure of 0 km as well
cies or deficiencies can also be detected to prevent safeguarding as quantify the divergent degradation behaviour of the compo-
scopes from transitioning to other phases to ensure a complete nents that influence the overall vehicle function. Degradation in
and seamless integration of methods in the simulation, for exam- this context refers to the change in characteristic component
ple. Currently, action is required to safeguard the reliability of over- properties across the fictitious vehicle lifespan and serves as a
all vehicle functions by carrying out numerical simulation basis for determining real-world reliability over the entire mileage
exercises. spectrum. Examples of degradation include aging and fatigue of
In view of the fact that long-term quality as it pertains to ride rubber bushes and mounts in the form of hardening or changes
comfort has previously not been safeguarded in the digital phase, in dampening elasticity. To this end, five steps were defined,
the following concept was devised for the overall vehicle function FIGURE 3, and are explained in terms of ride comfort in the follow-
of ride and handling. When this concept is applied, the simulation ing sections.

Passive safety, Passive safety, Rigidity/ Noise, vibration, Durability of Ride and handling
crash occupant/ durability and harshness (NVH) suspension,
pedestrian protection body load spectrums

Active safety Climate control/ Thermal management EMC Longitudinal dynamics Consumption/
thermal comfort electromagnetic energy efficiency
compatibility

Driving safety and safety-

related assistance systems

Aerodynamics Production processes Engine process/ Factory load scenarios

drivetrain

FIGURE 2 Overall vehicle functions

in the digital prototype

40
 Conception and
simulation
Testing

Physical development phase

Digital development phase [D]
Test stand (P) Street (S)

Safeguarding task A1 Safeguarding scope, digital (D) for A1 Scope P for A1 Scope S for A1

Safeguarding task A 2 Safeguarding scope, digital (D) for A 2 Scope P for A 2 Scope S for A 2

... ... ... ...

Safeguarding task A m Safeguarding scope, digital (D) for A m Scope P for A m Scope S for A m

Safeguarding task B1 Safeguarding scope, digital (D) for B1 Scope P for B1 Scope S for B1

Safeguarding task B2 Safeguarding scope, digital (D) for B2 Scope P for B2 Scope S for B2

... ... ... ...

Safeguarding task B n Safeguarding scope, digital (D) for B n Scope P for B n Scope S for B n

Safeguarding task C1 Safeguarding scope, digital (D) for C1 Scope P for C1 Scope S for C1

Safeguarding task C 2 Safeguarding scope, digital (D) for C 2 Scope P for C 2 Scope S for C 2

... ... ... ...

Safeguarding task C o Safeguarding scope, digital (D) for C o Scope P for C o Scope S for C o

... ... ... ...

... ... ... ...

... ... ... ...

... ... ... ...

TABLE 1 End-to-end safeguarding scopes for the digital, test stand and street phases

4.1.1 QUANTIF YING RIDE COMFORT To realise a numerically-based simulation exercise, however, objec-
Vibration or ride comfort encompasses all mechanical and acous- tive characteristic variables must be derived that accurately rep-
tic vibrations that the occupants in a vehicle are subjected to [4]. resent, or characterise, the subjective evaluation. An overview of
The term NVH, which is frequently used in literature and in prac- the different processes and procedures for objectively assessing
tice in this context, distinguishes between noise (audible > 100 perceived vibration is provided in [9].
Hz), vibration (perceived by touch or sight < 25 Hz) and harshness Many automakers evaluate vibrational performance by targeting
(intermediate range that is perceived differently depending on age the driver seat console or seat assembly under maximum vehicle
and constitution, approximate 25 to 100 Hz) in relation to the acceleration to define an objective criterion, as the main area in
range of frequency to which humans are sensitive [6, 7]. which forces are transferred from the vehicle to its occupants is
Ride comfort is evaluated based on different phenomena that the seats they sit on. Maximum acceleration continues to be used
occur on varying road surfaces and in specific driving situations. as an objectivity variable within the scope of the reliability analy-
In the digital prototype, the phenomenon known as stuttering, or sis. The limits for the design criteria can be set in variable fashion,
vibration, is viewed as part of standard procedure [8], whereby depending on the market position of the vehicle and manufacturer,
several definitions for the vibrational effect have been derived from and are specified in relation to the target design state for a vehi-
literature and practical applications. Most of the time, engine cle in the vehicle technical specifications. These limits directly
vibrations are what is being referred to and upset the otherwise
smooth control and stability of the front end of the vehicle as a
result of the engine block starting to resonate with the body at a
coupling frequency [8]. The term can also be expanded in scope,
Step 1: Step 2: Step 3: Step 4: Step 5:
however, to cover the vibrations exhibited in the chassis and sus-
pension, engine and detachable parts. When this definition is Objectively Analyse Prepare data Conduct
perceive transfer Simulate reliability
applied, the vibrational effect encompasses frequencies ranging ride comfort path(s)
and sample
analysis
from approximately 4 to 25 Hz. Ride comfort and the phenomena
associated with it are evaluated primarily in a subjective manner.
When vibration is very noticeable, for example, it is perceived as
an unpleasant hopping of the vehicle as it travels down the road. FIGURE 3 Five steps for simulating ride comfort reliability

03I2015 Volume 117 41

RESE ARCH SIMUL ATION

correlate with the levels of tolerable acceptance on the part of the

Vehicle occupants
vehicle occupants and are routinely validated with customers. In
the context of reliability engineering, exceeding the defined limits Seat
can be viewed as an active occurrence of the phenomenon and Engine/transmission
compromise the overall operative function of the vehicle as a
Bearing Bearing
result.
Engine mount Body

4.1.2 TRANSFER PATH ANALYSIS Bearing Bearing

Within the confines of traditional reliability engineering, the dura-
Springs/dampers Springs/dampers
bility of the overall system hinges on the reliability performance
of each individual component. Those combinations of component Front wheels Rear wheels

failures that can cause the overall system to fail are depicted in Street
a Boolean model [3]. Due to the complex interactions between
the components, however, a comfort phenomenon cannot be rep- FIGURE 4 Sample transfer path of a vehicle model for the vibration phenomenon
resented at the overall vehicle level using the classic Boolean
theory. Constellations can be conceived whereby the requirements
defined for vibrational response are maintained even though the
individual components perform outside the window of tolerance 4.1.3 DATA PREPARATION AND SAMPLING
specified. This is the case when degradation effects mutually The transfer path analysis also points to the minimum scope of
oppose, or counteract, each other. Conducting a transfer path components that must be tested in the simulation model. The
analysis can prove beneficial in this regard, as it allows the pro- response variable, however, still does not exhibit an identical, con-
cess by which the vibration phenomenon starts to be traced and sistent reaction to a change or modification to the individual com-
analysed. In this scenario, a detailed transfer path analysis was ponentry and, thus, to its degradation. This is why a sensitivity
conducted by experts and validated using trial-testing vehicles. analysis is conducted by drawing on a design-of-experiments exer-
A simplified version is illustrated in FIGURE 4. The transfer path cise [10], for example, to investigate the effects of each compo-
maps the chain of events from the point at which road vibrations nent parameter on the characteristic vibration value. For this pur-
occur and are transmitted to the seat. Road bumps place the pose, the component parameters are varied individually within a
wheels of a vehicle in a state of vibration. The wheels then trans- specific window of parameters and the intensity with which they
fer these vibrations via the wheel carriers and steering guide influence the simulation output is analysed. To ensure that the
assemblies to the suspension struts and from there on to the body outlay expended on the simulation is also compatible or compliant
up to the seat. The longitudinal members also act as a transfer with future applications, only those components should be
path that connects with the engine mountings and the engine and degraded in the long-term investigation that are significant with
transmission assembly. These vibrations are likewise routed along respect to the initial parameters or yield a comparably high level
the body to the seat and are perceived there by the vehicle occu- of degradation for less significant relevance. Reducing or restrict-
pant. At this time, not only the dampers and damper head bear- ing focus to critical components also minimises simulation exer-
ings, but also the suspension mountings and seat respond to cise times and reduces the outlay required to accumulate and pre-
minimise vibration levels. pare the output data.

Production scatter
Degradation process flows
WDF
[-]
for random sample

f (F,D,k) [-]

F1 F2 ... Fx Production
status [-]

Degradation scatter t0 t1 t2 ... tn Mileage [-]

WDF
[-] Fi: Production status of componen i, i = 1,…,x
Di: Degradation parameters of component i, i = 1,…,x
k: Coefficient for component characteristic (constant)
f (t, Fi, Di, k): Degradation process flow for component i
D1 D2 ... Dx Degradation
f (t, Fi, Di, k): Degradation process flow for component i
parameters [-]
WDF: Probability density function FIGURE 5 Statistical preparation of
the input data for a component

42
 ried out for the current S-Class. In fact, the vibration phenomenon
did not reach the described vehicle specifications limit carried out
Acceleration times graph during the analysis of the observed mileage. The resulting high

Acceleration [m/s2]
reliability is given by the fact that the simulation was based on
Time [s]
already near-series, optimised components. However, to demon-
Excitation of road
strate the application of the method in the practical example, the
Vibrating system of
surfaces, vehicle and occupants limit has been changed so that failures were recorded. This allows
handling/driving style
Frequency spectrum a theoretical replication of a previous stage of development. The

Acceleration [m/s2]
following remarks and conclusions are based on the reduced value.
The characteristic vibration value determined from the acceleration
Frequency [Hz] frequency spectrum is normalised across a vehicle mileage figure of
300,000 km in FIGURE 7. This value corresponds with the ampli-
tude of the maximum vibrational respectively stuttering accelera-
FIGURE 6 Inputs and outputs of the dynamic simulation tion amax,st. FIGURE 8 also shows the normalised limit value (red line)
for amax,st as defined in the requirement specification. If an occur-
rence of the vibration phenomenon is defined as any point in time
at which the limit value defined in the requirement specification
When vehicle random checks are carried out during the simula- is initially exceeded, the times of non-compliance can be derived
tion, the scattering effects associated with the new parts and those from the intersecting points of the characteristic and limit value
that relate to degradation performance are taken into account by curves. These are also plotted in the illustration.
way of statistical distributions. In real-world trial testing vehicles, Statistical reliability analyses in the context of mechanical engi-
various different combinations of these parameters can be encoun- neering are typically conducted by leveraging the Weibull distribu-
tered. To map this information accordingly in the simulation exer- tion [3]. In the process, the times of non-compliance are used to
cise, the first step taken is to quantify a random initial state or determine the Weibull straight line with its shape parameter b and
condition for each component by referring to the distribution that location parameter T. In so doing, the Weibull straight line depicts
maps the production scatter for the component in question, the characteristic median lines, meaning that 50 % of the non-
FIGURE 5. This can be a specific parameter such as the static rigid- compliance events occur to the left of the straight line, while the
ity of an elastomer bearing or a mathematical function such as the other 50 % are found at the right of this line. The shape param-
characteristic damper curve or dynamic rigidity. The combination eters of the Weibull distribution indicate the slope of the straight
of the values extracted characterises the first vehicle targeted by line and thus correlate with the scatter, or spread, of the non-com-
the random sample check. All additional vehicle samples are gen- pliance incidents. Location parameter T specifies at which mean
erated in the same fashion. The second step involves assigning a or average point in time 63.2 % of the vehicles tested failed. A
characteristic degradation curve to each vehicle. Here, too, a ran- confidence bound is also generally defined for the Weibull straight
dom generator is used for this purpose. To this end, statistical line that indicates the percentage of values that lie within the
distributions must first be defined for the degradation scatter. The bound.
result of the sampling process is a digital random sample that The dual-parameter Weibull straight line (with 90 % confidence
accurately represents the physical vehicles. bound) determined for the maximum stuttering acceleration amax, st
of the S-Class is depicted in FIGURE 8. This line corresponds with
4.1.4 SIMUL ATION a time-censored evaluation, meaning that all incidents of non-
An MBS model is typically used to simulate shaking and vibrational compliance that have not yet taken place after a fictitious vehicle
response [11]. The vehicle samples generated serve as input data lifecycle (here: 300,000 km) has been run through are incorpo-
in this context, FIGURE 6. As carrying out a dynamic simulation that rated as intact units in the non-compliance data analysis. With
spans the entire lifecycle of the vehicle is still unrealistic at pre- respect to the random sample this article focuses on, seven inci-
sent, a limited number of time windows is simulated for the ran- dents of non-compliance were registered over 300,000 km, while
dom sample as permitted by available computing power, and the three vehicles remained intact throughout this time frame due to
incremental progressions are quantified by applying regression their compliant vibrational response.
techniques. Vibrational excitation is characterised in amplitude Shape parameter b = 7.3 of the Weibull straight line points to
trend displays that are derived from plotting a road surface that the type of wear-related failures (compare bathtub curve [5]).
offers typical levels of vibration or an amplitude spectrum of sto- This was anticipated as such for the ride comfort, as it represents
chastically generated signals. The simulation output at any respec- the result of the interactions among and between degrading com-
tive point in time is initially depicted in a trend display for the ponents. The value b = 7.3 also indicates that the non-compli-
acceleration exhibited at the seat which, via the fast Fourier trans- ance incidents are all within close proximity to each other. Loca-
form algorithm, can transition to a frequency spectrum that quan- tion parameter T = 4.5 lies over the fictitious service life of a
tifies the acceleration level with respect to the given frequency vehicle, which is consistent with later failures caused by wear in
[12]. The target characteristic vibration value can then be deter- conjunction with the shape parameter determined. In addition to
mined based on this spectrum. the shape and location parameters, the Weibull curve can be
used to read characteristic service life value B10, which com-
4.1.5 RELIABILIT Y ANALYSIS monly applies to preventive quality assurance. This value is 3.3
The reliability analysis for the vibration phenomenon is depicted and indicates the time by which 10 % of the vehicles (with
in the following section by referring to a simulation exercise car- respect to amax,st) have failed, or exhibited non-compliance. In
03I2015 Volume 117 43
RESE ARCH SIMUL ATION

6
Requirement specifications
amax,St [normalised]

limit value amax,St

5

2
Vehicles 1 to 20
1

0 Mileage/service life t [in 1000 km]

Vehicles 1 to 20

0 1 2 3 4
Mileage/service life t [in 1000 km]

Vehicle 04 06 11 13 16 17 20
Failure time t 3.88 3.54 2.78 3.81 3.73 3.78 3.3
FIGURE 7 Failure data for vibration phenomenon

FIGURE 8 Results of the reliability analysis on

the vibration phenomenon

light of the fact that all non-compliance values lay within the 5 SUMMARY AND OUTLOOK
confidence bound of the Weibull straight line, it can be con-
cluded that each individual incident of non-compliance reveals In relation to the target objective of implementing an end-to-end
the same behavioural pattern. This deduction is supported by safeguarding process that covers the digital, test stand and street
the high correlation (98 %) of the non-compliance times with the phases, the concept outlined in this article highlights the efficiency
Weibull straight line. The reliability analysis for the maximum potentials that be achieved by implementing digital safeguarding
vibrational acceleration amax,st has shown that the non-compliance practises to ensure long-term quality. To this end, digital safe-
values simulated correspond to the failure behaviour anticipated guarding makes use of the evaluation criteria that were previously
for this characteristic value with respect to wear and that the reserved for the physical development phase.
failure incidents only occur when the vehicle has achieved a high The digital validation process described for safeguarding the
mileage figure. reliability of overall vehicle functions such as ride comfort thereby
44
gives rise to trial-testing scopes transitioning from test bench or
road-based environments to a digital validation system. This tran-
sition process ultimately leads to a reduced number of trial-testing
vehicles and, in light of new knowledge obtained during the digital
development phase, to a shorter physical or real-world develop-
ment phase.
The comfort phenomenon analysed merely represents one of
the elementary and relevant comfort aspects based on a selected
objectivity variable.
It is on this basis that additional comfort phenomena can be
analysed and digitally safeguarded in future so that further poten-
tials can be realised.

REFERENCES
[1] Breitling, T.; Großmann, T.; Zöller, A.: Digitale Prototypen unterstützen
Entwicklung. In: ATZextra (2009), No. 1, pp. 162-171
[2] Wattenberg, K.; Mahler, K.; Wiehe, L.; Breitling, T.; Dragon, L.; Bauer, W.:
Ohne Prototypen schneller zum Ziel. In: ATZextra (2012), No. 4, pp. 92-96
[3] Bertsche, B.; Lechner, G.: Zuverlässigkeit im Fahrzeug- und Maschinenbau.
Berlin: Springer, 2005
[4] [4] Heißing, B.; Brandl, H. J.: Subjektive Beurteilung des Fahrverhaltens.
Vogel, 2002
[5] Koller, O.; Mutter, K.; Uhl, M.; Zeiler, P.; Bertsche, B.: Simulation-Driven
Reliability Analysis in Automotive Applications. ARS 2013
[6] Morello, L.: The automotive body. Dordrecht: Springer, 2011
[7] Hieronimus, K.: Anforderungen an Schwingungs- und Akustikberechnungen
aus Sicht der Fahrzeugentwicklung. VDI-Bericht Nr. 186, Düsseldorf, 1990
[8] Mitschke, M.; Wallentowitz, H.: Dynamik der Kraftfahrzeuge. Berlin:
Springer, 1997
[9] Zeller, P.: Handbuch Fahrzeugakustik. Vieweg & Teubner, 2009
[10] Siebertz, K.; van Bebber, D.; Hochkirchen, T.: Statistische Versuchspla-
nung: Design of Experiments (DoE). Springer, 2010
[11] Rill, G.; Schaeffer, T.: Grundlagen und Methodik der Mehrkörpersimulation.
Vieweg & Teubner, 2010
[12] Jörißen, B.: Objektivierung der menschlichen Schwingungswahrnehmung
unter Einfluss realer Fahrbahnanregungen. Duisburg-Essen, Universität, Disser-
tation, 2012

03I2015 Volume 117 45

 RESE ARCH SIMUL ATION

AUTHORS
Knowledge Base for
Valid Driving Simulation

Driving simulators are being used increasingly as part of

Ilka Zöller, M. Sc.
is Scientist in the the vehicle research. In this regard, the validity of a driving
Research Group for Vehicle simulator represents a necessary condition in order to
Ergonomics, Institute of
Ergonomics and Human facilitate the transfer of the results to reality. The Institute
Factors at the Technische
Universität Darmstadt of Ergonomics & Human Factors (IAD) of the Technische
(Germany). Universität Darmstadt addressed this topic of validation of
driving simulators, as part of a two-year DFG project.

Nicole Mautes, B. Sc.

is Master’s Course Student
of Mechanical Engineering
and Student Assistant at
the Institute of Ergonomics
& Human Factors at the
Technische Universität
Darmstadt (Germany). L FOR SCIENTIFI
O VA CA
PR RT
AP I

CL
LO

ES
SEA

IN A
THE

TZ.
PEER REVIEW
|

|
REV RECEIVED 2014-07-14

Y.
REVIEWED 2014-09-10

STR
IEW
ACCEPTED 2014-12-09

DU
ED

Wen Ren, M. Sc.

IN
Y N
B

EX

D
is Master‘s Course PER HA
TS FR RC
OM RESEA
Graduate in Computational
Engineering at the
Technische Universität
Darmstadt (Germany).

Dr.-Ing.
Bettina Abendroth
is Research Group Leader
for Vehicle Ergonomics and
Deputy Manager of the
Institute of Ergonomics &
Human Factors at the
Technische Universität
Darmstadt (Germany).

46
1 MOTIVATION therefore physical reality. In the research, derivation of design rec-
2 BENCHMARKS FOR SELECTED DESIGN PAR AME TERS ommendations from a driving simulation valid for driver behaviour
3 VALIDIT Y OF DRIVER BEHAVIOUR AS A RESE ARCH FIELD is mostly yet to be achieved.
4 MODEL DE VELOPMENT
5 SOF T WARE TO OL
3 VALIDIT Y OF DRIVER BEHAVIOUR AS
6 SUMMARY
A RESEARCH FIELD

At the IAD, the question of validity of driver behaviour in driving

simulators has been explored in detail. Besides study of the litera-
ture, the project included the implementation of new series of
investigations in the field and in the simulator as well as ultimately
derivation of essential design recommendations for a valid (in the
following, the term validity refers to, unless explicitly contradicted,
1 MOTIVATION the driver behaviour related correspondence between real und
simulated environment of investigation) driving simulation.
The use of driving simulators as a study environment for driving The analysis of reference literature shows that are already some
tests has become a common practice. This is mainly due to the studies which are dedicated to the validation of driving simulators
fact that the influence of disturbance variables can be controlled [2, 12]. The research efforts to date, however, raise critical points
in a driving simulator [1] and a secure testing environment is that are essential for further investigation [13]. Validation studies
ensured for the subjects [2]. Nevertheless, a simulated test envi- lack a clean methodology and presentation of results. So as to
ronment has certain disadvantages too. Driving simulators can exclude the influence of individual performance requirements, it
never reflect the reality to 100 % [3]. To examine if the results is important to have recourse to different series of experiments on
obtained in a driving simulator are all the same transferable to identical group of subjects; however, this is still absent. In addi-
reality, it is necessary to validate the driving simulator. The aspects tion, there is often a lack of a complete presentation of the results,
of validity can be differentiated into two [4, 5], namely into physi- together with important statistical test values. Due to this fact, it
cal and behavioural correspondence. Physical correspondence is hardly possible to compare the findings of different studies.
describes the differences in the physical characteristics and the Also, there is a gap with regard to the objectives of the investiga-
external form between the driving simulator and the real vehicle, tions considered in the validation studies. Thus, only two studies
whereas the behavioural correspondence refers to the balancing are known [13] that examined the suitability of driving simulators
of the driver behaviour in simulated and real investigation environ- as a study environment for functional analysis of driver assistance
ment and is assumed to be present, provided that no statistically systems. However, this question turns out to be important if par-
significant difference exists [6]. ticularly for such study objectives mostly experiments are con-
ducted in simulated environment. Also the validation of driving
simulators for night driving has rarely been conducted, although
2 BENCHMARKS FOR SELECTED DESIGN PARAMETERS
under such adverse visibility conditions high accident rates prevail
There is a range of different driving simulators, which are differ- which accordingly results in a need for research.
ent with regard to their design parameters. The equipment qual- As part of the project, corresponding driving tests were con-
ity contributes directly to the physical validity [7]. In the litera- ducted in the field and in the static IAD driving simulator (180°
ture, there are benchmarks for each of the parameters which are FOV horizontally, sound simulation, force feedback). It is found
presented as excerpts in the following. What is particularly impor- that partially valid results can be achieved [14] with a driving sim-
tant is the visual perception. The horizontal viewing angle (field ulator of this equipment in investigations under night vision con-
of view, FOV) affects visual perception significantly. According ditions. When testing critical braking manoeuvres with active brake
to [8] an angle of 50° horizontally is acceptable as the minimum intervention, however, a static driving simulator has its limitations
FOV; however, depending on the driving situation, a FOV of 180° [15] due to lack of vestibular feedback. In this regard, an acoustic
may be advisable [9]. For a correct perception of speed, at least feedback could create a more realistic impression and provide
120° is required [10]. The vertical FOV plays a lesser role; in this essential stimuli for the perception of the acceleration forces.
respect, [8] considers 40° to be adequate [9]. The physical valid- Also a systemic study of the influence of independent variables
ity of a driving simulator, besides the visual, inter alia, is also on the validity remained until now largely incomplete. [13] pro-
influenced by the proprioceptive perception. For driving simula- vide an overview of the main influencing factors: task-related and
tors, there are several possibilities of simulating motion, for environment-related stress factors and individual conditions of
example hexapod, rail system, turntable, and vibration actuators. performance. In a study, mostly a specific driving simulator is
A 100-% representation of acceleration in a driving simulator is examined with regard to validity. What remains disregarded is
usually not possible; therefore, it is reduced by a scaling factor. the fact that the individual design components of a driving simu-
For lateral movements, according to [11], a value of 0.5 to 1.0 lator as well as the investigation scenario have an impact on the
and for longitudinal movements the order of less than 0.05 is results. Accordingly, a study was conducted at IAD, which exam-
possible. ined the influence of motion simulation and the FOV horizontally
The benchmarks presented provide design suggestions for a with respect to validity. For this purpose, the dynamic driving
physically valid driving simulation. The driver behaviour related simulator belonging to the Fraunhofer IGD was used. The test
correspondence is not necessarily achieved by the fact that the section consisted of city, highway and rural road passages. In
components of a driving simulator indicate a high quality and the literature there are already few studies that explore the influ-
03I2015 Volume 117 47
RESE ARCH SIMUL ATION

ence of FOV horizontally or also of motion simulation [10, 16] in for the driver behaviour validity. In addition to action, also the other
more detail. However, a comprehensive study that considers sys- phases of information processing act on the stress a [1, 22]
tematically varied both design parameters and also a wide range because they “demand a workload” from the human. The detec-
of validity parameters in various scenarios, is not available till tion leads, on a long-term basis, to a stress in the form of a fatigue
now. Mostly seen was a significant influence of design parame- effect, since it only involves the direct reception of stimuli by sense
ters on the subjective perception of reality of the subjects. The organs and does not result in memory balancing or processing.
results of the objective characteristics of this study are integrated The phases of cognition and decision-making, however, act quickly
into the software tool described later and elaborated there on the stress. There is also a backlash of the stress on the phase
systematically. of detection [1], since a high stress affects the stimulus percep-
tion negatively seen from a long-term point of view. Besides the
human-related effect of action on the stress, there is also a back-
4 MODEL DEVELOPMENT
lash of action on the environment (scenario) as well as the driving
From the previous observation, it is not clear how the design simulation itself (hardware). Therefore, the subject can defuse the
parameters affect the driver behaviour and thus the validity dangerous situation, for example, by pressing the brake pedal
depending on a scenario. This will be explained in more detail in (hardware) in case of a critical rear-end collision and thus influ-
the following. FIGURE 1 shows a model-based analysis of the inter- ence the scenario. As is evident in this model, the design param-
dependency between the independent variables of scenario and eters of the driving simulator (hardware/software) as well as the
hardware/software and the dependent variables of action, stress scenario selection will influence the validity. Therefore, it is impor-
and subjective evaluation. According to the methodology of [17], tant to adapt the design of a driving simulator and its components
the validation of a simulator involves a comparison this same to the respective target of investigation.
dependent variables in the field and in the simulator. The aim of
the model is to explain the relationship between dependent and
5 SOFTWARE TOOL
independent variables based on the information processing of
humans. The scenario of a driving test series goes down as an The software tool SimuVal was developed at the IAD to provide
environmental variable in the man-machine interaction; it is mod- assistance to researchers in the configuration of a driving simula-
elled in the simulator software. The hardware design parameters tor, so that it is optimally suitable for the respective target of
of the driving simulator pass on this scenario information, for investigation. Since the optimal form of the design parameters
example, about the imaging medium to the human. The human depends on the considered characteristic values as well as the
receives the stimuli through his senses and processes them. investigation scenario, a guide in the classical sense was out of
According to [18] the information processing can be divided into the question. In order to meet this complexity, the systematised
steps of detection, recognition, decision and action. In the pro- findings were combined into a software tool. The software tool
cessing of information, there is a balancing with the sensory mem- was developed in C++. The tool offers the possibility to system-
ory as well as the short and long-term memory [19]. In this, there atically look for study results concerning driver behaviour validity
is also a comparison of the stimuli presented in the simulator with through a graphical user interface, to add new entries or modify
reality. This closeness to reality of the simulated environment as existing ones, FIGURE 2.
perceived by the subject acts indirectly on his decision and action
and thus on the validity. This validity aspect can also be referred
6 SUMMARY
to as fidelity of a driving simulator [20] and, according to [7], in
analogy to driving behavioural validity. At the end of information In summary, it can be said that the DFG project at IAD has deliv-
processing is the action of man. It is directly related to stress [21]. ered a comprehensive insight into the question of the validity of
In the area of action as well as in the area of stress, there are vari- driver behaviour validity, systematised previous study results in
ous validity parameters that allow a comparison of driver behaviour this context, revealed key factors affecting the driver behaviour
between the field and the driving simulator and provide evidence validity and quantified their effect on a number of validation char-

Human

Stress

Detect Recognise Decide Action

Memory
Driver behaviour
Sense organs validity
Subjective evaluation (according to
Brown1975)
Fahrerverhaltensvalidität (i.A. an Brown 1975)

Hardware
FIGURE 1 Relationships of effect between independent
Software
Driving simulation and dependent variables of the driver behaviour validity
(solid arrows = short-term effect; dotted arrows = long-
Scenario Environment term effect)

48
 FIGURE 2 Screenshot of the
software tool SimuVal

acteristics in the context of a comprehensive study. The results In: 60. Kongress der Gesellschaft für Arbeitswissenschaft, 12-14 March 2014,
Munich
from the trials as well as from the literature are systematically pro-
[15] Zöller, I.; Betz, A.; Mautes, N.; Scholz, L.; Abendroth, B.; Bruder, R.; Win-
cessed in the software tool SimuVal according to the influencing ner, H.: Valid representation of a highly dynamic collision avoidance scenario in
factors and shown graphically to the user. a driving simulator. In: Transportation Research Part F [accepted for publication]
[16] Colombet, F.; Reymond, D. M. G.; Pere, C.; Merienne, F.; Kemeny, A.: Motion
cueing: What is the impact on the driver’s behavior? In: Proceedings Driving Sim-
REFERENCES ulator Conference (2008), pp. 171-182
[1] Vollrath, M.; Schießl, C.: Belastung und Beanspruchung im Fahrzeug – [17] Brown, J.: Visual elements in flight simulator. In: Report of Working Group
Anforderungen an Fahrerassistenz. In: VDI-Berichte (2004), No. 1864, pp. 34 (1975), Washington DC: Committee on Vision, Assembly of Behavioural and
343-360 Sociel Sciences. National Research Council, National Academy of Sciences
[2] Blana, E.: Driving simulator validation studies: A literature review. Institute [18] Luczak, H.: Untersuchungen informatorischer Belastung und Beanspru-
for Transport Studies, University of Leeds, Vol. Working Paper 480, 1996 chung des Menschen. Darmstadt, TU, Dissertation, 1975
[3] Knappe, G.: Empirische Untersuchungen zur Querregelung in Fahrsimula- [19] Schlick, C. M.; Bruder, R.; Luczak, H.: Arbeitswissenschaft. Heidelberg:
toren. Vergleichbarkeit von Untersuchungsergebnissen und Sensitivität der Springer, 2010
Messgrößen. Erlangen-Nürnberg, Universität, Dissertation, 2009 [20] Blana, E.; Golias, J.: Differences between vehicle lateral displacement on
[4] Mudd, S.: Assessment of the Fidelity of Dynamic Flight Simulators. In: the road and in a fixed-base simulator. In: Human Factors 44 (2002), No. 2,
Human Factors 10 (1986), No. 4, pp. 351-358 pp. 303-313
[5] McCormick, E.: Human Factors Engineering. New York: McGraw-Hill, 1970 [21] Rohmert, W.: Das Belastungs-Beanspruchungs-Konzept. In: Zeitschrift für
[6] Blaauw, G.: Driving Experience and Task Demands in Simulator and Instru- Arbeitswissenschaft 38 (1984), No. 4, pp. 193-200
mented Car: A Validation Study. In: Human Factors 24 (1982), No. 4, pp. [22] Handschmann, W.; Voss, M.: Informationsaufnahme und -verarbeitung
473-486 durch den Kraftfahrzeugführer. Teil 2: Beanspruchung durch Informationsbelas-
[7] Godley, S.; Fildes, B.; Triggs, T.: Driving simulator validation for speed tung. In: ATZ 81 (1979), No. 2, pp. 67-70
research. In: Accident Analysis and Prevention 34 (2002), pp. 589-600
[8] Haug, E.: Feasibility study and conceptual design of a national advanced
driving simulator. Iowa City: University of Iowa, College of Engineering, Center for
Simulation and Design Optimization of Mechanical Systems
[9] Padmos, P.; Milders, M.: Quality criteria for simulator images: A literature
review. In: Human Factors 34 (1992), No. 6, pp. 727-748
[10] Jamson, H.: Image characteristics and their effect on driving simulator
validity. In: Proceedings of the International Driving Symposium on Human
Factors in Driver Assessment, Training and Vehicle Design (2001), Colorado
[11] Greenberg, J.; Blommer, M.: Physical fidelity of driving simulators. In:
Handbook of Driving Simulation for Engineering, Medicine, and Psychology,
Boca Raton, Florida, CRC Press, 2011
[12] Mullen, N.; Charlton, J.; Devlin, A.; Bédard, M.: Simulator validity: Behav- THANKS
iors observed on the simulator and on the road. In: Handbook of Driving Simula-
tion for Engineering, Medicine, and Psychology, Boca Raton, Florida, CRC Press, The authors would like to thank the Institute of Automotive Engineering of
2011 Technische Universität Darmstadt (Professor Hermann Winner) and the
[13] Zöller, I.; Diederich, C.; Abendroth, B.; Bruder, R.: Fahrsimulatorvalidität – Fraunhofer Institute for Computer Graphics Research (IGD) in Darmstadt
Systematisierung und quantitative Analyse bisheriger Forschungen. In:
(Professor André Stork) for providing the test vehicle and the driving simulator
Zeitschrift für Arbeitswissenschaft 67 (2013), No. 4, pp. 197-206
[14] Zöller, I.; Yang, N.; Abendroth, B.; Bruder, R.: Zur Validität des Blickbewe- respectively within the DFG project Determination and Quantification of Design
gungsverhaltens unter Nachtsichtverhältnissen in einer Fahrsimulatorstudie. Features of a Realistic Driving Simulation.

03I2015 Volume 117 49