Tyre Model
Tyre Model
The present contribution is concerned with the semi-physical tyre model TMeasy for vehicle
dynamics and handling analyses. Even in the case of more or less weak testing input data,
the effort for the application of TMeasy remains limited due to its consequent easy to use
orientation. One particular feature of TMeasy is the widely physical meaning of its smart
parameter set, which allows to sustain the identification process even under uncertain conditions.
After a general introduction, the modeling concept of TMeasy is compactly described in this
contribution. Taking the interface STI to MBS software into account, the way how to apply
TMeasy is briefly shown. This includes selected examples of application.
1 Introduction
1.1 Overall concept of TMeasy
TMeasy is a semi-physical tyre model for low frequency applications in vehicle dynamics. It has
consequently been following an easy to use strategy which takes the existing insufficiencies in
the availability of reliable testing data into account. In order to fulfil this aim, the number of
model parameters of TMeasy is rather limited according to the limited availability and accuracy
of the basic experimental data. On the other hand the more or less direct physical meaning
of the TMeasy model parameters enables their identification also in the case of unsure or even
incomplete measurement data sets.
Within TMeasy the contact force characteristics in longitudinal and lateral direction are described by a handful of physical parameters, which takes the degressive influence of decreasing
tyre load into account. The load influence can be easily decsribed by providing these parameters for the nominal payload and its double value. The combined force characteristics are then
directly generated via a generalised slip approach which does not need any additional fitting
1 Introduction
parameters. TMeasy also computes the tyre load as the third component of the tyre contact
forces.
The self aligning torque is approximated by the product of lateral force and the pneumatic
trail, which again is described by few physical parameters. TMeasy also includes the calculation
of the rolling resistance torque, the tipping torque and the bore (turn) torque around the normal
axis. The approximative calculation of the contact geometry delivers the camber angle which
influence on the contact forces and torques is considered within a certain range of application.
As long as the contact patch remains as a closed area TMeasy can handle uneven roads too.
Typical applications are wave tracks, longitudinal grooves and change of road inclination, which
are offered as special road types additionally to the standard even road including an arbitrary
subarea of deviating friction.
In order to model the time delay at transient force changes TMeasy includes nonlinear first
order dynamics for the longitudinal and lateral tyre forces as well as for the aligning torque and
the bore torque. The latter allows to consider the steering effort while parking. Again there are
no additional parameters spent for the first order filters, whereas the variable relaxation lengths
are directly calculated from known physical tyre properties.
1.2 History
TMeasy was developed under the primary aspect of practical applicability in vehicle dynamics.
The first version was published in [8]. It has to be mentioned that from that time some TMeasy
derivatives are widely in use which are also named TMeasy. Known reproductions are used by
dSPACE in the Vehicle Dynamics Simulation Package ASM [17] and in the program for accident
reconstruction PC-Crash [16]. Another implementation of TMeasy into Dymola is known, [18].
Of course, these externally created versions are excluded from the official TMeasy software
maintenance.
Based on increasing experiences from passenger car and truck applications, an improved version 2.0 was created. The steady state part is published in [3] and the modeling concept for
dynamic tyre forces and torques can be found in [9]. In 2003, enhancements in the discretisation
of the contact geometry led to version 3.0. Besides some enhancements with respect to MBS
integration via the Standard Tyre Interface (STI), the version TMeasy 4.0 includes a parking
torque model. This actual version was applied for the Tyre Performance Test programme TMPT
which was organised by the Vienna University of Technology in 2007, [4].
A new version 5.0 is currently in preparation which include a sophisticated model for the
computation of the dynamic tyre forces which provides a smooth transition from stand still to
normal driving situations.
1.3 Applications
For many years MAN Nutzfahrzeuge AG have been using TMeasy within the simulation system
SIMPACK [20] for the investigation of the dynamics and safety of heavy trucks, [5], [14]. In the
meantime, TMeasy has been offered as an official, optional tyre model for vehicle dynamics in
SIMPACK.
Furthermore, TMeasy is integrated as the standard tyre model into the simulation system
veDyna [21], which is widely applied for off-line and on-line simulation of road vehicles. Recently,
the implementation of TMeasy into the MBS system Adams [15] was successfully carried out
and validated, [1] [7] [2].
2 Modeling Concept
2 Modeling Concept
2.1 Contact Geometry
2.1.1 Local Track Plane and Geometric Contact Point
In any point of contact between the tyre and the road surface normal and friction forces are
transmitted. The effect of the contact forces can be fully described by a resulting force vector
applied at a specific point of the contact patch and a torque vector.
undeflected
tire contour
M
x +x
+y y
uneven
road
undeflected
tire contour
en
Q2
P Q1
longitudinal
inclination
lateral
inclination Q3
uneven
road
en
P
Q4
rQ2 Q1 rQ4 Q3
.
| rQ2 Q1 rQ4 Q3 |
(1)
As in reality, sharp bends and discontinuities, which will occur at step- or ramp-sized obstacles,
are smoothed by this approach.
The rim center plane is defined by the unit vector eyR into the direction of the wheel rotation
axis, Fig. 2. The unit vector ex into the direction of the intersection line of the local track plane
and the rim center plane defines the direction of the longitudinal tyre force. The direction of the
lateral tyre force is described by the unit vector ey which is mutual perpendicular to the track
normal en and the unit vector ex into the direction longitudinal tyre force. The tyre camber
angle
= arcsin eTyR en
(2)
describes the inclination of the wheel rotation axis eyR against the local track normal en . The
point P on the intersection line with the shortest distance to the wheel center M serves as
geometric contact point. Its location is described by the vector
r0P = r0M + rM P ,
(3)
where r0M defines the momentary position of the rim center M with respect to the earth-fixed
reference frame 0 and the vector from the rim center to the geometric contact track point can
be written as
rM P = rS ezR ,
(4)
2 Modeling Concept
rim
centre
plane e zR
tire
e yR
rMP
e yR
en
wheel
carrier
ex
b
P0 a
ey P
en
P0
z0
P*
x0
road: z = z ( x , y )
y0
x0
y0
z0
rS
b/2
en
P
ey
z
r0-rSL
en
Fz
A
ey
r0-rSR
Q
Fz
yQ
wC
wC
b2 tan
12 4z cos
(5)
will hold, where b is the width of the tyre, 4z denotes the tyre deflection and names the tyre
camber angle. If the cambered tyre has only a partial contact to the road then, according to
2 Modeling Concept
the deflection area a triangular pressure distribution will be assumed. Now, the location of the
static contact point Q is given by
1
b
yQ =
wC
,
(6)
3
2 cos
where wC defines the actual width of the contact patch and the term b/(2 cos ) describes the
distance from the geometric contact point P to the outer corner of the contact patch. The plus
sign holds for positive and the minus sign for negative camber angles. The static contact point
described by the vector
r0Q = r0P + yQ ey
(7)
represents the contact patch much better than the geometric contact point.
2.1.3 Contact Point Velocity
The absolute velocity of the static contact point will be obtained from
v0Q,0 = r0M,0 + rM Q,0 ,
(8)
where r0M,0 = v0M,0 denotes the absolute velocity of the wheel center and rM Q describes the
position of static contact point Q relative to the wheel center M . The vector rM Q contains the
tyre deflection 4z normal to the road and it takes part on all those motions of the wheel carrier
which do not contain elements of the wheel rotation. Hence, its time derivative can be calculated
from
rM Q,0 = 0R,0
rM Q,0 + 4z en,0 ,
(9)
is the angular velocity of the wheel rim without any component in the direction of
where 0R
the wheel rotation axis, 4z denotes the change of the tyre deflection, and en describes the road
normal. As the static contact point Q lies on the track, v0Q,0 must not contain any component
normal to the track
eTn,0 v0Q,0 = 0
or
eTn,0 v0M,0 + 0R,0
rM Q,0 + 4z eTn,0 en,0 = 0 .
(10)
As en,0 is a unit vector, eTn,0 en,0 = 1 will hold, and then, the time derivative of the tyre deformation
is simply given by
(14)
2 Modeling Concept
where the constants a1 and a2 may be calculated from the radial stiffness at nominal and double
payload. The parabolic approximation in Eq. (14) fits very well to the measurements, [3]. The
dynamic part is roughly approximated by
FzD = dR 4z ,
(15)
where dR is a constant describing the radial tyre damping, and the derivative of the tyre deformation 4z is given by Eq. (11).
The lateral shift of the vertical tyre force Fz from the geometric contact point P to the static
contact point Q is equivalent to a force applied in P and a tipping torque Tx acting around a
longitudinal axis in P , Fig. 4.
en
ey
en
ey
P Q
y
en
P
Fz
Fz
ey
Q
P
Tx
Fz
(vx rD )
rD ||
and sy =
vy
rD ||
(16)
appear simultaneously. Here, vx and vy are the components of the contact point velocity in
longitudinal and lateral direction, describes the angular velocity of the wheel and rD is the
dynamic rolling radius. Both slips can vectorially be added to a generalised slip
s
2
q
2
2
2
sx
sy
s =
+
=
sN
+ sN
,
(17)
x
y
sx
sy
N
where a normalization was performed, sx sN
x and sy sy , in order to achieve a nearly equally
weighted contribution to the generalised slip. If the wheel locks, the average transport velocity
will vanish, rD || = 0. Hence, longitudinal, lateral, and generalised slip will tend to infinity,
N
s . To avoid this problem, the normalised slips sN
x and sy are modified to
sN
x =
sx
(vx rD )
=
sx
rD || sx
and
sN
y =
sy
vy
=
sy
rD || sy
sN
x =
sN
y =
(vx rD )
rD || sx + vN
vy
.
rD || sy + vN
(18)
(19)
2 Modeling Concept
When choosing small values vN > 0 the singularity at rD || = 0 is avoided. In addition, the
generalised slip points then into the direction of the sliding velocity for a locked wheel. In
normal driving situations, where rD || vN holds, the differences between the primary and
the modified slips are hardly noticeable.
Fx
Fx
Fx
Fy
0
dF x
sx
sM
x
sSx
Fy Fy
dF 0
FS
FM
Fy
F(s)
Fx
sy
dF y
sSy
sS
sM
sy
sM
y
sx
(20)
sN
F N
y
Fy = F sin = F
=
sy = f sN
(21)
y ,
s
s
where f = F/s describes the global derivative of the generalised tyre force characteristics.
The generalised tyre force characteristics F = F (s) is now approximated in intervals by
appropriate functions, Fig. 6. In the first interval a rational fraction is used which is defined by
the initial inclination dF 0 and the location sM and the magnitude F M of the maximum tyre
force. Then, the generalised tyre force characteristics is smoothly continued by two parabolas
until it finally reaches the sliding area, were the generalised tyre force is simply approximated
by a straight line.
2 Modeling Concept
FM
parabola
FS
parabola
dF0
straight
line
rational
function
s*
sM
sS
n/L
(n/L)0
(n/L)0
s0y
sSy
sy
s0y
sy
2 Modeling Concept
ex
B
dr
F
r
L
ey
RP
circular
approximation
r n
rD ||
(24)
where r n describes the sliding velocity in the patch element and the term rD || represents the
average transport velocity of the tread particles. By setting r = RB we get the average bore slip
sB =
RB n
,
rD || + vN
(25)
where similar to (18) and (19) the artificial velocity vN 0 was added in the denominator
in order to avoid numerical problems at a locked wheel. Now, the bore torque can simply be
approximated by
TB = RB dF 0 sB .
(26)
Via the initial inclination dF 0 and the bore radius RB the bore torque TB automatically takes
the actual tyre properties into account. The bore torque is limited by its maximum value,
| TB | TBmax which is defined by (23).
2 Modeling Concept
In order to respect this fact in a tyre model, the characteristic data for two nominal wheel loads
and 2 FzN will be provided. From this data the initial inclinations dFx0 , dFy0 , the maximal
forces FxM , FyM and the sliding forces FxS , FyS for arbitrary wheel loads Fz are calculated by
quadratic functions. For the maximum longitudinal force it reads
Fz
1 M
1 M
M
N
N Fz
M
N
M
N
Fx (Fz ) = N 2 Fx (Fz ) 2 Fx (2Fz ) Fx (Fz ) 2 Fx (2Fz ) N .
(27)
Fz
Fz
FzN
M
S
S
The location of the maxima sM
x , sy , and the slip values, sx , sy , at which full sliding appears, are
defined as linear functions of the wheel load Fz . For the location of the maximum longitudinal
force this will result in
F
z
M
M
N
M
N
M
N
sx (Fz ) = sx (Fz ) + sx (2Fz ) sx (Fz )
1 .
(28)
FzN
The self-aligning torque is modeled via the lateral force and the dynamic tyre offset. The
characteristic curve parameters describing the dynamic tyre offset will be provided for the single
and double pay load too. Similar to Eq. (28) the parameters for arbitrary wheel loads were
calculated by linear inter- or extrapolation.
2.6.2 Coefficient of Friction
The tyre characteristics are valid for one specific tyre road combination only. Hence, different
tyre road combinations will demand for different sets of model parameter. A reduced or changed
friction coefficient mainly influences the maximum force and the sliding force, whereas the initial
inclination will remain unchanged. So, by setting
sM
L M
L M
s , FM
F ,
0
0
sS
L S
L S
s , FS
F ,
0
0
(29)
the essential tyre model parameter which primarily depend on the friction coefficient 0 are
adjusted to the new or a local friction coefficient L .
If the road model will not only provide the unevenness information z = fR (x, y) but also
the local friction coefficient [z, L ] = fR (x, y) then, braking on -split maneuvers can easily be
simulated, [10].
2.6.3 Camber
At a cambered tyre, Fig. 9, the angular velocity of the wheel has a component normal to the
road
n = sin ,
(30)
where denotes the camber angle. Now, the tread particles in the contact patch have a lateral
velocity which depends on their momentary position. At the contact point it vanishes whereas
at the end of the contact patch it takes on the same value as at the beginning, however, pointing
into the opposite direction. Assuming that the tread particles stick to the track, a parabolic
deflection profile will be generated.
The lateral displacements of the tread particles caused by a cambered tyre are compared now
with the ones caused by pure lateral slip. For small lateral slips the equivalent camber slip is
given by
1
sy = s .
(31)
3
10
2 Modeling Concept
en
rim
centre
plane
eyR
rD ||
ex
v()
y()
ey
Fy =
s ,
(32)
sy sy =0 y
where Fy F M limits the camber force to the maximum tyre force. By replacing the partial
derivative of the lateral tyre force at a vanishing lateral slip by the global derivative of the
generalised tyre force the camber force will be automatically reduced when approaching the
sliding area. By introducing a load dependent weighting factor in Eq. (32) the camber force can
be adjusted to measurements.
dx
Fx
cx
rim
dy
cy
tire
tire
xe vx - rD
Fy
ye
vy
and
11
Fy
Fy (vy + y e ) Fy (vy ) +
y e ,
vy
|
{z
}
| {z }
FyD
FyS
(33)
2 Modeling Concept
where xe and ye name the longitudinal and the lateral tyre deflection. In steady state the
longitudinal tyre forces FxS and FyS will be provided by Eqs. (20) and (21) as functions of the
N
normalised slips sN
x and sy . Their derivatives with respect to the components of the contact
point velocity are given by
FxS
FxS sN
FxS
1
x
=
=
N
N
vx
sx vx
sx rD ||
sx + vN
(34)
FyS
FyS sN
FyS
1
y
=
=
vy
sN
sN
sy + v N
y vy
y rD ||
(35)
where the definition of the normalised longitudinal slip in Eqs. (18) and (19) were used to generate
the derivatives of the slips with respect to the components of the contact point velocity. Corresponding to the first order approximations in Eq. (33) the partial derivatives of the steady state
tyre forces with respect to the normalised slips will be approximated by their global derivatives
FxS
FxS
f sN
x
=
= f
N
N
N
sx
sx
sx
and
FyS
FyS
f sN
y
N = N = f,
N
sy
sy
sy
(36)
1
x e
rD ||
sx + vN
and FyD f sN
y + f
1
y e ,
rD ||
sy + vN
(37)
where according to Eqs. (20) and (21) the steady state tyre forces FxS and FyS were replaced by
N
the terms f sN
x and f sy . On the other hand, the dynamic tyre forces can be derived from
FxD = cx xe + dx x e
and FyD = cy ye + dy y e ,
(38)
where cx , cy and dx , dy denote stiffness and damping properties of the tyre in longitudinal and
lateral direction. Inserting the normalised longitudinal slips defined by Eqs. (18) and (19) into
the Eq. (37) and combining them with Eq. (38) yields two first order differential equations for
the longitudinal and lateral tyre deflection
(vT x dx + f ) x e = f (vx rD ) vT x cx xe ,
vT y dy + f y e = f vy vT y cy ye ,
(39)
(40)
and vT y = rD || sy + vN
(41)
12
2 Modeling Concept
contact ring
c
RB
d
wheel rim
W
RP
(42)
Similar to Eqs. (39) and (40) this model approach results in a first order differential equation for
the tyre twist angle
2
2
dF0 RB
+ rD || d = dF0 RB
W rD || c .
(43)
At stand still ( = 0) the simple differential equation
= W
(44)
remains here which means that the torsional tyre deflection is increased or decreased as long
as steering motions W 6= 0 are performed. But, the differential equation (44) is only valid as
long as the resulting bore torque does not exceed the maximum value. To take this effect into
account at first the steady state torque is limited |c | TBmax . Then, adhesion is assumed
which is described by
2
dF0 RB
W + rD || TBst
.
(45)
A =
2
dF0 RB
+ rD || d
The resulting dynamic bore torque
TBD = c + d A
now allows to check for sliding which finally is done by
D
max
A if |TB | < TB
=
0 if |T D | T max
B
B
(46)
(47)
This model approach provides a continuous transition from stand still, rD || = 0, to normal
driving situations, rD || > 0, [9].
13
STI
Standard Tyre
Interface
Slip
Camber
Deflect.
P2
F, M
Wheel
motion
F, M
Model param.
USRMOD
P2
ROAD
TMroad
Tyre type 1
P1
TMeasy
x, y
Internal
standard
tyre model
Simulation
host
[Model param.]
z, (x, y)
P1
Road param.
Road type R
14
15
4000
2000
Fn=1800 N
Fn=3200 N
Fn=4600 N
Fn=6000 N
Approx 1
Approx 2
Approx 3
Approx 4
0
2000
4000
6000
20
15
10
10
15
20
adhesion
adhesion/
sliding
full sliding
FySnom
dFySnom
syS nom
syMnom
sy
16
zR
yR
C
z0
z
O
xR
0
y0
C
yR
y
x0
O
17
zC
C
xC
yC
Z
18
19
References
References
[1] L. Herak. Verification of semiphysical model of tyre TMeasy (Adams Implementation).
Master thesis, Slovak University of Technology in Bratislava, 2008.
[2] W. Hirschberg, F. Palcak, G. Rill and J. Sotnik. Reliable vehicle dynamics simulation in spite
of uncertain input data. EAEC Conf. 2009 Europe in the Second Century of Auto Mobility,
Bratislava: (CD) 2009.
[3] W. Hirschberg, G. Rill and H. Weinfurter. User-appropriate tyre-modeling for vehicle dynamics in standard and limit situations. Vehicle System Dynamics, 38(2): 103125, 2002.
[4] W. Hirschberg, G. Rill and H. Weinfurter. Tire Model TMeasy. Vehicle System Dynamics,
Vol. 45, Supplement 1: 101119, 2007.
[5] W. Hirschberg, H. Weinfurter and C. Jung. Ermittlung der Potenziale zur LKWStabilisierung durch Fahrdynamiksimulation. VDI-Berichte 1559, D
usseldorf: 167188, 2000.
[6] P. van der Jagt. The Road to Virtual Vehicle Prototyping; new CAE-models for accelerated
vehicle dynamics development. ISBN 90-386-2552-9 NUGI 834, Tech. Univ. Eindhoven, 2000.
[7] P. Kintler. Validierung des Reifenmodells TMeasy mittels eines Vollfahrzeugmodells.
Master thesis, Slovak University of Technology in Bratislava, 2009.
[8] G. Rill. Simulation von Kraftfahrzeugen. Vieweg Verlag, ISBN 3-528-08931-8, Braunschweig/Wiesbaden, Deutschland, 1994.
[9] G. Rill. First order tire dynamics. In Proc. of the III European Conference on Computational
Mechanics Solids, Structures and Coupled Problems in Engineering, Lisbon, Portugal, 2006.
[10] G. Rill and C. Chucholowski. Modeling concepts for modern steering systems. In ECCOMAS
Multibody Dynamics, Madrid, Spain, 2005.
[11] G. Rill and C. Chucholowski. Real time simulation of large vehicle systems. In ECCOMAS
Multibody Dynamics, Milano, Italy, 2007.
[12] J.J.M. van Oosten et al. Tydex Workshop: Standardisation of Data Exchange in Tyre
Testing and Tyre Modelling. Proc. 2nd Int. Colloquium on Tyre Models for Vehicle Dynamic
Analysis, Swets&Zeitlinger, Lisse 1997.
[13] S. R. Waser. Applikation des Reifenmodells TMeasy f
ur den Tyre Model Performance Test
TMPT. Diploma thesis, Graz University of Technology, Austria, 2005.
[14] H. Weinfurter, W. Hirschberg and E. Hipp. Entwicklung einer Storgrossenkompensation f
ur
Nutzfahrzeuge mittels Steer-by-Wire durch Simulation. In VDI-Berichte 1846, D
usseldorf:
923941, 2004.
[15] http://www.mscsoftware.com/, dated Aug 27, 2010.
[16] http://www.dsd.at/, dated Aug 27, 2010.
[17] http://www.dspace.de/ww/de/gmb/home/products/sw/
automotive simulation models.cfm/, dated Aug 27, 2010.
20
References
[18] http://www.dynasim.com/documents/0709 dysm fly 04vehi cl lo.pdf/,
dated Sept 11, 2007.
[19] http://www.mathworks.com/, dated Aug 27, 2010.
[20] http://www.simpack.de/, dated Aug 27, 2010.
[21] http://www.tesis.de/, dated Aug 27, 2010.
21