Full vehicle simulation - verified road loading for vehicle in design phase Otmar Gattringer, 2015
Introduction Passenger car
Transfer of signals
Measurement signals FEMFAT LAB vi – virtual iteration ?Introduction
?Measurement signals
?Full vehicle simulations based on
road load data (RLD)
–Existing vehicle
–New vehicle (virtual prototype) ?Conclusion
Time plot black…measurement
red…...simulation
Full vehicle simulation Content
Goal
?Verified road loading for vehicle in design phase
Initial situation
?Existing vehicle –RLD, measured on test tracks –Simulation model –Virtual results based on RLD (dynamics, durability,…) ?New vehicle (succeeding model)
–No physical prototype – no RLD available
–Simulation model exists, different changes
–Load of simulation model is missing
?Transfer of existing RLD to new vehicle
–Load of MBS model of new vehicle computed by virtual iteration
Introduction Full vehicle simulation
Typical changes
?
Wheel base ?
Springs ?
Dampers ?
Bushings ?
Mass ?…
Existing vehicle ?MSC.ADAMS/car model including radial
tire stiffness for vertical direction
?Measured signals of rough road - RLD
?Virtual iteration of road surface, 4-poster
?Drive (load)
–Vertical poster displacements (road
surface) ?Response (desired)
–Spring displacements - used for iteration
–Wheel center accelerations - used for
iteration
–WFT (wheel force transducer) signals -
model check
Displacement applied by MOTION
Iteration process
u n+1 = u n + F -1 ( y Desired - y n )
3. Transfer function
1. Noise – drive signal
2. Response of noise 4. Drive signal
5. Response
6. Response = desired
Spring displacement front left Spring displacement front left – interval of 1 second Results: 8th iteration, time domain
black…measurement
red…...simulation
WC acceleration Z rear left WC acceleration Z rear left – interval of 1 second
WFT FZ rear right (model check) WFT FZ rear right– interval of 1 second
New vehicle ?Modified MSC.ADAMS/car model –Spring rate, damper and bushing
characteristics
–Wheel base
–Center of gravity, mass and mass of
inertia of body
?Road surface of existing model will be used for vertical direction (invariant signals) –Front axle: signals unchanged
–Rear axle: signals time shifted according
new wheel base and vehicle speed
Existing vehicle New vehicle
black…road surface existing vehicle
red…...recomputed, new vehicle
Time shift of rear axle
signals
New vehicle
?WFT signals are used for remaining
directions, applied at wheel centers
–Front axle: signals are scaled regarding
new wheel load
–Rear axle: signals are scaled regarding
new wheel load and shifted according
new wheel base
?Simulation of rough road maneuver ?Results e.g. internal forces for durability
analysis
Displacement
applied by MOTION
Forces applied by
GFORCE
Additional comparison
Results of VI approach compared with
simple simulation using WFT signals only
?No tires are included
?Scale WFT signals regarding changed
wheel loads
?Model excited by scaled WFT signals in all
directions
?Weak springs in vertical direction are used
for stabilization (acting between “ground”
and wheel hub)
?Simulation of rough road maneuver
All directions are applied by GFORCE
Comparison
?Interface forces of body
(important for durability analysis of body)
?Representation in relative damage values
(damage content of signal - no relation to
damage of structure)
?Selection of points and directions based
on high forces
Conclusion of comparison ?Vertical interface forces overall to small
(usually critical at damper dome areas)
?Other directions fit better because same
signal is applied at wheel center,
difference comes from vertical influence
green……longitudinal forces
blue……https://www.doczj.com/doc/fc7853355.html,teral forces
r ed………vertical forces
?Full vehicle simulation based on RLD can be done in different ways, depending on parts of interest
?For parts located far away from excitation, measured internal signals should correlate with high quality to ensure accurate internal forces
(e.g. the damper force and spring displacement should be accurate at damper dome, on the other hand WFT FZ can be less accurate)
?Virtual iteration approach shows excellent convergence between measurement and simulation
?Virtual iteration
–Efficient method to generate absolute displacements (e.g. tire patch, frame
movement for add on parts like cab, tank, engine, exhaust systems)
–Model verification and trimming by additional checking signals
–Subsystems and test rigs can be simulated based on real RLD using simple and cheap measurements
–Efficient parameter studies and transfer to similar vehicles (invariant load)
–Method is automated for
?ADAMS
?SIMPACK
?MOTIONSOLVE
?RECURDYN
and implemented in the software FEMFAT Lab