© Fraunhofer ITWM1
CDTire: State-of-the-art Tire Models forFull Vehicle Simulation
2012 Americas HyperWorks Technology Conference
Dr. Manfred Bäcker, Axel Gallrein
© Fraunhofer ITWM2
Fraunhofer Society
Karlsruhe
DarmstadtWürzburg
Jena
Stuttgart
Duisburg
Oberhausen Dortmund
München
Saarbrücken
St. Ingbert
Magdeburg
Halle
Dresden
Leipzig
Ilmenau
Cottbus
Braunschweig
Berlin
PotsdamTeltow
Aachen
Schmallenberg
Sankt Augustin
Erlangen
Nürnberg
Freising
Holzkirchen
Pfinztal
Freiburg
Efringen-Kirchen
RostockItzehoe
Hannover
Bremen
EuskirchenChemnitz
FürthKaiserslautern
Paderborn
Schkopau
Lübeck
Erfurt
Oberpfaffenhofen
Bremerhaven
� Fraunhofer Society 2010
� Non-profit
� 60 institutes
� 18 000 employees
� 1.65 bill. Euro
� Fraunhofer ITWM
� 220 employees
� 18 mill. Euro
� Transport, fluid and material simulation
� Dynamics and durability
� Image processing
� Optimization
� Finance
� High performance computing
© Fraunhofer ITWM3
Agenda
� Overview
� CDTire Model Family
� Parameterization
� How CDTire works with MotionSolve
� Introducing new models
� CDTire 30/HPS
� CDTire 50
� On-going developments
� CDTire NVH
� Tire / soil interaction
© Fraunhofer ITWM4
Tire Simulation in Vehicle Development
Driver model
� Open loop
� Closed loop
� Optimal control
Vehicle model
� MBD model
� FEA model
� Subsystems
Tire model
� MBD model
� FEA model
� Linearization
Road model
� Rigid surface
� Soil
� Multi pass
© Fraunhofer ITWM5
MBD Tire Models - - - - vs. - - - Application Fields
Typical number of sim
ulationsTy
pical computatio
nal e
ffort
100
101
102
103100
101
102
103
104
105
DOFs
Empirical
Frequency-based
Rigid ringEmperical contact
Flexible beltBrush-type contact
FEA
Handling
NVH
Active safety
Ride/Comfort
Durability
Crash
REPs
© Fraunhofer ITWM6
MBD Tire Model Family CDTire
Typical number of sim
ulationsTy
pical computatio
nal e
ffort
100
101
102
103100
101
102
103
104
105
DOFs
Empirical
Frequency-based
Rigid ringEmperical contact
Flexible beltBrush-type contact
FEA
Handling
NVH
Active safety
Ride/Comfort
Durability
Crash
REPs
CDTire 20
CDTire 30CDTire 40CDTire MC
TireTool
© Fraunhofer ITWM7
� Model 20
� Rigid ring
� Physical tangentialcontact formulation
� Model 30
� Flexible ring
� Brush type contact
� Model 40
� Flexible Belt
� Independent sidewalls
� Model MC
� Large belt curvature
� Large inclination angles
MBD Tire Model Family CDTire
� Adaptive models(automatic switching)
� Model 2030
� Model 2040
© Fraunhofer ITWM8
CDTire 20
Longitudinal direction
Lat
eral
dir
ecti
on
Sheardeformation
XcYc
Belt velocity
Belt point
Surface pointTreadheight
� Model details
� Rigid ring
� Physical tangentialcontact formulation
� Road surfaces
� Arbitrary 3D
� Long wavelength
� Typical applications
� Handling (Durability)
� Active safety
� Real-time factor
� 3 - 6
© Fraunhofer ITWM9
CDTire 20: Application Scenario Handling
� Publication
Yang, X., Medepalli, S.: Comfort and Durability Tire Model Validation, Tire Science and Technology, TSTCA, Vol. 37, No. 4, October – December 2009, pp. 302-322.
© Fraunhofer ITWM10
CDTire 30
pppp� Model details
� Flexible ring
� Brush type contact formulation
� Accumulated sidewall
� Road surfaces
� Trackwise 2D
� Arbitrary wavelength
� Typical applications
� Ride & Comfort
� Durability
� Real-time factor
� 10 - 30
© Fraunhofer ITWM11
CDTire 30: Application Scenario Durability
� Publication
Sawa, N., Nimiya, Y., Kubota, Y., Itsubo, T. et al., "Fatigue Life Prediction on Rough Road Using Full Vehicle Co-simulation Model with Suspension Control," SAE Technical Paper 2010-01-0952, 2010, doi:10.4271/2010-01-0952.
Full vehicle cleat run
Full vehicle rough road run
© Fraunhofer ITWM12
CDTire 40
� Model details
� 3D structure
� Brush type contact formulation
� Independent sidewalls
� Road surfaces
� Arbitrary 3D
� Typical applications
� Ride & Comfort
� Durability
� Real-time factor
� 40 - 100
© Fraunhofer ITWM13
CDTire 40: Application Scenario Special Events
� Publication
Baecker, M., Gallrein, A., and Haga, H., "A Tire Model for Very Large Tire Deformations and its Application in Very Severe Events," SAE Int. J. Mater. Manuf. 3(1):142-151, 2010, doi:10.4271/2010-01-0373.
© Fraunhofer ITWM14
CDTirePI: Parameterization via Measurements
� Measurements
� Geometry (cross section, foot print)
� Modal analysis
� Static (stiffness’s, on cleat)
� Steady-state (long. + lateral slip)
� Transient (90°+45° cleat runs)
� Optimization schemes
� Local
� Fuzzy (expert knowledge buildinto rule sets)
© Fraunhofer ITWM15
TireTool: Parameterization via FEA tire model
Layer 1: Exp. Data
Layer 3: Application
Geometry FootprintsStiffness Cleats
Ground-outModal AnalyzeMaterial
Crash NVHMisuse Handling
DurabilitySpecial Events
Local structureproperties
Virtual measurement
Layer 2: ParameterIdentification
© Fraunhofer ITWM16
TireTool: Parameterization via FEA tire model
2D modelCross section 3D model Virtualmeasurements
�Cross section scan�Contour measurement
�CT-Scan cords�Tire weight�Shore A
�Modes and frequencies�Static stiffness’s
Fine tune
�Cleat runs�Ground-out�…
© Fraunhofer ITWM17
TireTool: Parameterization via FEA tire model
� Measurements
� Geometry (cross section,foot print)
� Modal analysis
� Static (vertical stiffness)
� Material
� CT
� Output is Abaqus model
© Fraunhofer ITWM18
Tire-Tool: Parameterization via FEA tire model
90° cleat run 40x40mm, 11 km/h, CDT40 validation
Longitudinal Force
-5000
-4000
-3000
-2000
-1000
0
1000
2000
3000
4000
5.1 5.15 5.2 5.25 5.3 5.35 5.4 5.45 5.5
Time [s]
For
ce [
N]
Measurement
Simulation
Vertical Force
0
5000
10000
15000
20000
25000
30000
5.1 5.15 5.2 5.25 5.3 5.35 5.4 5.45 5.5
Time [s]
Forc
e [N
]
Measurement
Simulation
� LDE test rig@ Fraunhofer
� Kistler P530 hub
� Max. 12 km/h
� Max. Fz 30 kN
� Up to rim contact
© Fraunhofer ITWM19
Tire-Tool: Parameterization via FEA tire model
90° cleat run 40x40mm, 11 km/h (FEA validation), 30 and 40 km/h (prediction)
© Fraunhofer ITWM20
TireTool: Parameterization via FEA tire model
3D continuum Reissner-Mindlinhypothesis 2D shell
FEM
Reduction of degrees-of-freedom (dofs)
� Proper orthogonal decomposition (POD)
� BMBF joint project „Multi-disciplinary Simulation and Non-linrear Model Reduction (SNiMoRed)
FEM
(CDTire 50)
© Fraunhofer ITWM21
How CDTire works with MotionSolve
� Cooperation Altair / Fraunhofer
� Altair is exclusive reseller forCDTire4MotionSolve
� Distributed with HyperWorks 11.0
� Integrated visual support in development
� Tire models
� 20 / 30 / 40 / 2030 / 2040
� Road models
� RSM1000 (parametric obstacles)
� RSM1002 (drum model)
� RSM2000 (large digital data)
� RSM3000 (OpenCRG)
� RSM1100 (User API)
� Full .adm compatibility
� Full CDTire.ini support (parallelization)
© Fraunhofer ITWM22
� Model 30/HPS
� C/C++ High Performance Solver
� Real-time capable
� Model 50
� Materialized sidewalls
� Geometrical non-linearelastic shell discretization
� Model NVH
� Flexible rim
� Cavity mode
Introducing new models
New developments
� Model 20
� Rigid ring
� Physical tangentialcontact formulation
� Model 30
� Flexible ring
� Brush type contact
� Model 40
� Flexible Belt
� Independent sidewalls
� Model MC
� Large belt curvature
� Large inclination angles
© Fraunhofer ITWM23
� Model 30/HPS
� C/C++ High Performance Solver
� Real-time capable
� Model 50
� Materialized sidewalls
� Geometrical non-linearelastic shell discretization
� Model NVH
� Flexible rim
� Cavity mode
Introducing new models
New developments
� Model 30/HPS
� MIL / SIL / HIL cascade
� Vehicle performance optimization
� Model 50
� Large deformation, ground-out
� Handling on rough road
� Model NVH
� Up to 300 Hz
Application scenarions
© Fraunhofer ITWM24
CDTire 30/HPS
� Master integrator(typically larger time steps)
� Iteration scheme
� Step size control
� Slave integrator(typically smaller time steps)
� Iteration schemes(HPS: PECE, Newmark)
� Master step size control
� Tunable controls:Master iterationSlave iterationSlave step size
timet t+Dt
Secured time stepnew time step to be calculated by MBS-solver
Estimated rim states at t+Dt
MBS-solver � � Tire model solver �Tire forces at t+Dt
Dt = MBS step size
Integrates tire states from t -> t+Dt using much smaller minor steps dt
dt = minor step of tire model integrator
dt
© Fraunhofer ITWM25
CDTire 30/HPS� Scenario: 4 cleats (20x20mm), 40 km/h, 3 sec
� C/C++ High Performance Solver
� Efficient memory storage and access
� Efficient implicit Newmark with special projector in Newton-Iteration
� Choice of deterministic integrator settings (fixed step, fixed iterations)
� Parallelizable incl. road surface models (OpenCrg, RSM2000, RSM1000)
CDT30 (explicit integrator) Real time factor = 1.8
CDT30/HPS (explicit integrator)Real time factor = 1.1
CDT30/HPS (implicit integrator, var. step) Real time factor = 0.6
CDT30/HPS (implicit integrator, determ.)
Real time factor = 0.8
0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6-5000
0
5000
10000
Fx,
N
New CDT30 model. 100 MP
0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.65000
6000
7000
8000
9000
Fz,
N
Time, s
© Fraunhofer ITWM26
CDTire 30/HPS: Application Scenario Test Rig� Scenario MIL / SIL / HIL: Implementation used as component on real test rigs
� Road profile and driving rule can be used as excitation of the test rig
� No test track measurement needed (usually used as target for iteration process)
© Fraunhofer ITWM27
CDTire 30/HPS: Application Scenario Drive Simul.� Scenario MIL / SIL / HIL: Component in Drive Simulator
� VAR-Simulator (Versatile Augmented Reality)
Real-time PC
HMI datarobot states
Vehicle CAD data
3D terrain data
3D scenery
Warping, edge blending, color &
contrast calibration
18 synchronal WUXGA projectors
Motiontracking
Graphics clusterwith scene graph
Fraunhofer ITWMFraunhofer ITWM Fraunhofer IGD & ITWM Fraunhofer FIRST Fraunhofer ITWM
Spherical screen 10m diameter
300° * 110° FOV
© Fraunhofer ITWM28
BeltSidewall
Rim
CDTire 50
� Functional element modeling
� Element stacking in pre-processing
� FD shell discretization
© Fraunhofer ITWM29
CDTire 50� Scenario: 18 cleats (10x10,…,20x20mm), 40 km/h, 10 sec
� C/C++ High Performance Solver
� Efficient memory storage and access
� Efficient implicit Newmark with special projector in Newton-Iteration
� Choice of deterministic integrator settings (fixed step, fixed iterations)
� Parallelizable incl. road surface models (OpenCrg, RSM2000, RSM1000)
CDT50 (explicit integrator)50 cross sections12 segmentsReal time factor = 49
© Fraunhofer ITWM30
CDTire NVH
Step: Loading−0Increment 22: Step Time = 0.9650
Deformed Var: U Deformation Scale Factor: +1.000e+00
apply vertical load on inflated tireODB: 03_tire_loading.odb Abaqus/Standard 6.10−EF2 Fri Jul 08 10:49:32 CEST 2011
XY
Z
� Method NVH – FEA tire
� Steady state transport analysis (relative kinematic ALE description)
� Rim and road (footprint) excitation
� Condensation of road excitation: Track-wise time delay as phase shift
� Internal Fraunhofer research project (MEF)
� Cavity possible
� Flexible rim possible
© Fraunhofer ITWM31
CDTire NVH
Linearization aroundStationary rolling state
AddGyroscopic forces
LinearState space equation
LaplaceTransform
TransientEquations of motion
Condensation ofRoad excitation
� Internal Fraunhofer research project (MEF)
� Solver independence
� Linearization and
� Transient
© Fraunhofer ITWM32
CDTire NVH
� Switch to „eulerian“ description by adding gyroscopic forces
� Leads to frequency split for rotating tires
� Example is free hanging tire, but rotating
Residual force
© Fraunhofer ITWM33
CDTire NVH
� Simulated experimental modal analysis
� Spacial acceleration measurement
� Free rim, constant rot. velocity w = 85 rad/s
Counter Inphase
© Fraunhofer ITWM34
No deformationRigid surface
Local deformation, coupledBulldozing effect
Terrain compactionMultipass
Tangential mass transportSlip sinkage
Local deformation, decoupledTerrain response
Diskrete Element MethodDEM
Tire / Soil interaction
© Fraunhofer ITWM35
Tire / Soil interaction
� Simple pressure-sinkage relations
� Bekker: p(z) = (kc/b + kΦ)∙zn
� Reece: p(z) = (c κc + γs b κΦ)∙(z/b)n
(kc,kΦ,κc,κΦ,n: pressure-sinkage parameters,γs: weight density, c: cohesion of terrain)
� Instantaneous soil deformation
� No sinkage “propagation”
� Localized parameter, i.e. k = k(x,y), n = n(x,y)
� Tangential frictional contact
� Implementation in C/C++
� Python interface for rapid evaluation of alternative sinkage models
© Fraunhofer ITWM36
Spreading the Application Range
Typical number of sim
ulationsTy
pical computatio
nal e
ffort
100
101
102
103100
101
102
103
104
105
DOFs
Empirical
Frequency-based
Rigid ringEmperical contact
Flexible beltBrush-type contact
FEA
Handling
NVH
Active safety
Ride/Comfort
Durability
Crash
REPs
CDTire NVH
CDTire 20
CDTire 30CDTire 40CDTire MCCDTire HPS
CDTire 50
TireTool
© Fraunhofer ITWM37
CDTire: State-of-the-art Tire Models forFull Vehicle Simulation
2012 Americas HyperWorks Technology ConferenceDr. Manfred Bäcker, Axel [email protected]
Thank you
for your attention !