1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Development of Virtual Closure Slam Test

Autors: Dr.-Ing. Joachim Liebers

Dipl.-Ing. Manfred Saul

Company: Adam Opel GmbH, Rüsselsheim

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 1
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Outline
1. Introduction
2. Methods of slam simulations – comparative reflection
3. Similarities and differences between side door-, hood-
and tailgate/decklid slam simulation
4. Investigations on a front side door of a passenger car
5. Conclusion
6. Future outlook

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 2
 1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Introduction
• hang on parts (side doors, rear end closures and hoods) have to fulfil diverse
requirements over their complete life time

• many loadcases were developed for dimensioning

(linear
stiffness- and stress targets, nonlinear
e.g. tolerable plastic deformations
under misuse, dynamic
eigenfrequencies, mobilities and also transient events)

• the main function of hang on parts is to open and close the car

• the closing process requires a robust design to fulfil all functions over the
complete life of a car

• in the past many extensive tests were necessary (time- and cost intensive)

• the goal is to reduce the development time, save money, built better cars and
most importantly obtain more customer satisfaction

we must improve our virtual development process and substitute more

and more hardware tests by confident simulations
GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 3
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Development of Virtual Closure Slam Test

1. Introduction
2. Methods of slam simulations – comparative reflection
3. Similarities and differences between side door-, hood- and
tailgate/decklid slam simulation
4. Investigations on a front side door of a passenger car
5. Conclusion
6. Future outlook

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 4
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Methods of slam simulations – comparative reflection

Methods of slam-simulation

1. Linear analysis (inertia relief)

2. Explicit nonlinear transient analysis
3. Implicit nonlinear transient analysis

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 5
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Methods of slam simulations – comparative reflection

1. Linear analysis (inertia relief)

step 1
determination of the latch- and rubber bump stop forces via experiment

step 2
the determinated latch- and rubber bump stop forces are the external forces
and the mass inertia of the door, tailgate or hood represent the counterforce

sealings are not included

we get displacements and a stress distribution

the aim of dimensioning is a homogeneous stress distribution without
stress peaks above defined target values

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 6
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Methods of slam simulations – comparative reflection

2. Explicit nonlinear transient analysis

Model with all nonlinear properties (material, also with sealings and contacts)
inclusion of damping properties possible
the analysis start with an initial velocity at the point of the first contact
between the door, tailgate respectively hood with the latch, rubber bump stop
or sealing
it‘ s possible, that the evaluation includes the rebound
we get the displacements and the strain- respectively stress distribution as a
function of time as results

therefore a fatigue-analysis is possible

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 7
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Methods of slam simulations – comparative reflection

3. Implicit nonlinear transient analysis
Model with all nonlinear properties (material, also with sealings, contacts)
inclusion of damping properties possible
the analysis starts with an initial velocity at the design position
the initial velocity at the design position is lower than at the moment of first
contact between the door, tailgate respectively hood with the latch, rubber bump
stop or sealing, because transmutations of energy at the sealing, bump stops and
the latch
but there is enough experience for good estimation of initial velocity
it‘ s also possible to get the initial velocity from measurement e.g. at the
predecessor or at similar hang on parts
it‘ s possible, that the evaluation includes the rebound
we get as results the displacements and the strain- respectively stress
distribution as a function of time

therefore a fatigue-analysis possible

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 8
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Methods of slam simulations – comparative reflection

Pros and cons of the methods
Linear analysis (inertia relief) Explicit nonlinear transient Implicit nonlinear transient
analysis analysis
no inclusion of nonlinear effects (no inclusion of all nonlinear effects inclusion of all nonlinear effects
sealing, no latch- and bump stop
characteristic and no contacts)

no rebound included rebound included rebound included

latch force must be known
test the initial velocity at the design position
necessary must be estimated
there is enough
experience for good estimation
no dynamic effects included, only one dynamic effects included during the dynamic effects included during the
strain/stress peak evaluated complete time complete time
local mesh refinement without important limited possibility of local mesh local mesh refinement without important
evaluation time disadvantages possible refinement, reasion is the correlation evaluation time disadvantages possible
(for fatigue-Analysis essential) between element size and time step (for fatigue-Analysis essential)
in principle short evaluation time
drastic influence on the evaluation time

limited precision and capability
result evaluation good supported by additional data translaters necessary result evaluation good supported by
fatigue tools fatigue tools

same model can be used as in many more effort for adaption from the crash same model can be used as in many
other simulations (Stiffness, NVH) model (e.g. sealings) other simulations (Stiffness, NVH)

decision for implicit nonlinear transient analysis

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 9
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Development of Virtual Closure Slam Test

1. Introduction
2. Methods of slam simulations – comparative reflection
3. Similarities and differences between side door-, hood- and
tailgate/decklid slam simulation
4. Investigations on a front side door of a passenger car
5. Conclusion
6. Future outlook

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 10
 1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Similarities and differences between side door-, hood- and

tailgate/decklid slam simulation
Slam simulation for hood
• model:
hood in white with all added masses for trim – spotwelds and clinch points as hexahedra with
RBE3-elements, average element size 7mm
latch
hinge assembly
body in white front structure – cut line with SPC
nonlinear characteristics of latch and rubber bump stops
contact on important regions
nonlinear material for high loaded areas
• start point of simulation: design position
• initial velocity: calculated from height of fall (depends on model - with or without gasstrut)
• calculation time: approximately up to 2 ms after reversal point
• results und evaluation:
maximum v. Mises stress and plastic strain target during the complete time
overshooting and clearance to the headlight
no fatigue evaluation

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 11
 1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Similarities and differences between side door-, hood- and

tailgate/decklid slam simulation
Slam simulation for decklid / tailgate
• model:
decklid / tailgate (with glas) in white with all added masses for ligths and windshield motor
spotwelds and clinch points as hexahedra with RBE3-elements, average element size 7mm
latch and hinge assembly
body in white rear structure – cut line with SPC
latch and rubber bump stop nonlinear characteristic with damping properties
sealing with nonlinear characteristic with damping properties
contact at important reagions
no trim (like test)
• start point of simulation: design position
• initial velocity on the lower edge:
derived from test procedure (1,7m/s)

1,2m/s for simulation (transmutations of energy up to the design
position at the sealing, bump stop and the latch)
• calculation time: 50 ms
• results and evaluation: maximum v. Mises stress target during the complete time
fatigue evaluation for 40000 events
GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 12
 1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Similarities and differences between side door-, hood- and

tailgate/decklid slam simulation

Slam simulation for side doors

• model:
door in in white with all added masses for ligths, loudspeaker and mirrow - spotwelds and
clinch points as hexahedra with RBE3-elements, average element size 7mm
hinge assembly
window regulator assembly
glass with run channels and seals
latch and body side striker
body in white side structure – cut line with SPC
latch and puffer nonlinear characteristic with damping properties
all sealings with nonlinear characteristic with damping properties
contact on important regions
no trim (like test)

3 models with 3 different window positions (up, middle, down) same like in test
Important: all elements and nodes of interest, from which a fatigue evaluation will be made,
must have the same number and be located in the same component in
every model.

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 13
 1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Similarities and differences between side door-, hood- and

tailgate/decklid slam simulation

Slam simulation for side doors

start point of simulation: design position

initial velocity on the rear edge:
derived from test procedure (1,5m/s)

1,0m/s for simulation (transmutations of energy up to the design
position on the sealing and the latch)
calculation time: 50 ms
results und evaluation:
maximum v. Mises stress target during the complete time for every window position separate
fatigue evaluation results for every window position separate
fatigue evaluation results all 3 window positions overlaid

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 14
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Development of Virtual Closure Slam Test

1. Introduction
2. Methods of slam simulations – comparative reflection
3. Similarities and differences between side door-, hood- and
tailgate/decklid slam simulation
4. Investigations on a front side door of a passenger car
5. Conclusion
6. Future outlook

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 15
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Investigations on a front side door of a passenger car

FE model

average element size 7mm

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 16
 1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Investigations on a front side door of a passenger car

Target for front side doors

no failure (cracks, affected spotwelds or other lost functions)

20000 repeats 30000 repeats 50000 repeats

total 100000 repeats

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 17
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Investigations on a front side door of a passenger car

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 18
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Investigations on a front side door of a passenger car

Life in repeats
(base = 100.000 cycles

1 life repeat complete test)

spotweld 15mm moved

0,066
0,022
0,040
0,015
0,027
0,009
0,133
0,046
7,55
21,8 improvement up to 288 %
spotweld 20mm moved
damage upper window position 0,234
0,038
damage middle window position 0,139
0,024
damage lower window position 0,080
0,013
damage all 3 window positions overlaid 0,451
0,076
Life repeats complete test 2,22
13,2 improvement up to 594 %
GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 19
 1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Investigations on a front side door of a passenger car

alternative
2-bolt design with
basic version with 3 bolts experiment for cost saving
2-bolt design
2 spotwelds added
min life cycles – 27.140.000 repeats min life cycles – 41.920 repeats
min life cycles – 248.500

Test result – 49900 repeats

2 spotwelds added

Life in repeats
Target life cycles – 100.000 repeats
GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 20
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Development of Virtual Closure Slam Test

1. Introduction
2. Methods of slam simulations – comparative reflection
3. Similarities and differences between side door-, hood- and
tailgate/decklid slam simulation
4. Investigations on a front side door of a passenger car
5. Conclusion
6. Future outlook

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 21
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Conclusion

• it‘s possible to find critical regions

• the method is already applicable to optimisations at this stage
• it‘s difficult to include manufacturing tolerances
• today, dimensioning without hardware demands inacceptable securities,
which results in a heavy and expensive design
today it‘ not possible to
abstain from additional tests
• slam simulation is an excellent tool to support the improvement process
after failure during the test
• it‘ s also feasible to quantify an improvement
• the use of implicit nonlinear transient analysis method has the potential for
a lot of improvements to continue the reduction of test numbers

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 22
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Development of Virtual Closure Slam Test

1. Introduction
2. Methods of slam simulations – comparative reflection
3. Similarities and differences between side door-, hood- and
tailgate/decklid slam simulation
4. Investigations on a front side door of a passenger car
5. Conclusion
6. Future outlook

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 23
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Future outlook
for the aim – development without hardware – future method
advancements will be implemented
• improved information about dynamic characteristics
of sealings and rubber bump stops (stiffness and damping)
• improved modeling of bolt connections (possibly with preload)
• mesh refinement in critical regions
• improved formulations of material characteristic (cyclic stress-strain
curve)
• not only strain-life-curves for different materials - also for different units

accumulate experiences in future projects

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 24
1st European HyperWorks Technology Conference, Berlin, October 23-24, 2007

Thank you for your attention!

GME Engineering
Vehicle Performance Simulation Joachim Liebers, Manfred Saul 25