Development Tyres
Thermal and Mechanical
Tyre Modelling for Handling
Simulation
TaMeTirE is a new thermo-mechanical tyre model for handling simulation
developed by Michelin. Its main benefit compared to mathematical fit mod-
els is its capability to extrapolate tyre force and moment to the actual driv-
ing conditions of the vehicle by taking both speed and tyre temperature
effects into account and showing their significant influence on the lateral
force in pure cornering conditions. TaMeTirE can accurately predict the
lateral force especially when the tyre is at the limit of friction.
ATZ 05I2008 Volume 110
1 Introduction tion. In cornering conditions, the side- Authors
walls, belt and tread deformations lead to
The ability to predict tyre characteristics a non-uniform contact patch shape and
that are relevant to the actual driving condi- load distribution, Figure 1. An accurate Dr.-Ing.
tions for the tyre and the vehicle combined modelling of the mechanical behaviour of Pierre Février
remains an important means of improving the tyre can be addressed with Finite Ele- is technical manager
tyre models for handling and vehicle dy- ment approaches. Such models are used to for tyre modelling and
namics. Many test results show how tyre gain and understanding of tyre design and traction research at
force and moment characteristics are de- tyre mechanics. However, for full vehicle Michelin, Clermont-
pendent on the test conditions under which dynamics simulation, tyre modelling sim- Ferrand (France).
they are obtained [1]. This is mainly due to plifications are necessary in order to keep
thermal effects and thermo-dependent elas- the model as simple and as fast as possible.
Autorenfoto sehr blass!
tic and viscous properties of the rubber Basically, the TaMeTirE model is based on
compound. The current state of the art for three models: one for the tyre mechanics, Ing. Gérard Fandard
tyre modelling is based on mathematical one for the compound characteristics is tyre performance
formulas that are fitted to indoor test ma- (shear modulus and friction) and one for expert at Michelin, Cl-
chine measurements. These can represent the tyre temperatures, Figure 2. ermont-Ferrand
the tyre characteristics for a wide variety of (France).
tyre types and sizes, but the forces and mo-
ments they describe are only representative 2.1 Mechanical Model
of indoor measurement conditions that the The mechanical model is based on the
model is based on, and the quality of the well-known “brush element” approach
model prediction is completely linked to [3-13]. The total force is the sum of the
the measurement protocol. Recently, new tread elementary shear forces in the ad-
developments regarding the measurement herent and sliding areas of the contact loads or high inflation pressures, the
protocol have been carried out to bring the patch (CP). Moreover, the total lateral pressure distribution along the CP is par-
tyre’s thermal states as close as possible to force causes a CP displacement as well as abolic. In contrast, at high loads or low
the situations encountered on a track by belt bending and casing twisting, thus inflation pressures, the pressure distri-
the vehicle’s tyre [2]. However, because in- modifying the tread shear force genera- bution is more uniform. The tyre force
door tests cannot cover all the actual driv- tion. The following stiffnesses are intro- derivation is detailed in [14]. The force
ing conditions, such an approach that disre- duced: distribution is calculated along the CP
gards the influence of thermal effects re- according to a one-dimensional brush
mains limited to a few vehicle manoeuvres. K T Sidewall twisting stiffness, element approach (that is one rib). The
The aim is to develop a tyre force and mo- 1/S2 Belt bending stiffness, effect of the non-uniform CP shape is in-
ment model that takes thermal effects into KX, KY Tread stiffness in the longitudinal troduced by modifying the rib boundary
account for vehicle dynamic simulation and lateral directions. conditions. For calculation time efficien-
tools and for any vehicle manoeuvres. The cy, it is convenient to consider that there
mechanical and thermal tyre model TaMe- In cornering conditions, Figure 3, the self- is a unique point b in the contact patch
TirE calculates the longitudinal and lateral aligning torque MZ leads to a sidewall that corresponds to the transition be-
force as well as the self-aligning torque in a twisting mechanism that reduces the ef- tween adherent and sliding conditions.
combined slip angle and slip ratio, at differ- fective slip angle δ′ of the CP, Eq. (1): When the shear force is balanced by the
ent loads, inflation pressures, camber an- friction force, the contact remains slid-
M
gles, travelling speeds and road and ambi- δ′ = δ + δ0 + ___
K Z Eq. (1)
T
ent temperatures. The first part of this arti-
cle presents the modelling approach and In addition, the lateral force induces an
the equations. This is followed by an analy- in-plane belt deflection (that is bending
sis of data from flat-track measurements in the X-Y plane) that reduces the shear
and simulation results in pure cornering deformation of the tread. This deflection
conditions for a Michelin Primacy (225/55 is approximated by a quadratic formula,
R16) tyre. in which the curvature ρ of the belt at
the CP centre is given by Eq. (2):
2 Thermal and Mechanical Tyre ρ = S2 Fy Eq. (2)
Model Development
The CP width Ly and length Lx are func-
The complex non-uniform and non-linear tions of the load and inflation pressure. Figure 1: Tyre contact patch and pres-
deformations of the tyre structure have a The pressure distribution along the CP is sure distribution in cornering condition
significant effect on the tyre force genera- given by a polynomial equation. At small from Finite Element simulations
ATZ 05I2008 Volume 110
Development Tyres
ing until the exit of the contact patch.
The longitudinal and lateral forces are
given by Eq. (3) to Eq. (5):
Ly/2 a Ly/2 b
∫ ∫
FX = K ∫ ∫ (r τ) r+ τ(1+τ) β
X(XK–XN)dS + ____________
K
_____________
–Ly/2 b –Ly/2 –a √ 2
K
2 2
µ(p,Vg,T)p dS Gl. (3)
b
Ly/2 a Ly/2
∫
( 1+τ) β
∫ ∫
FY = K ∫ (r τ) + (1+τ) β
Y(YK–YN)dS + ____________
–a
_____________
–Ly/2 b –Ly/2 √ K
2 2 2
µ(p,Vg,T)p dS Gl. (4)
1 KX
β = δ′– __ __
2 S2FY (a+KN), rK = K , a = Lx/2Gl. (5)
Y
The first term of Eq. (3) and Eq. (4) is the
total force in the adherent area of the CP.
This force contribution is governed by the Figure 2: TaMeTirE modelling approach. 1. Mechanical model based on the “brush element”
tread stiffness and CP dimensions. The approach. 2. Rubber compound characteristics (shear modulus; friction coefficient evolution
second term of Eq. (3) and Eq. (4) is the to- with the contact temperature and sliding speed). 3. Thermal models for the contact tempera-
tal friction force in the sliding area. ture as well as for the surface and internal temperatures of the tread over a period of wheel
rotation
2.2 Rubber Shear Modulus Model and
Friction Model
Tread rubber characteristics play an im-
portant role in the tyre force generation,
as they govern the tread stiffness, the
friction properties and the heat genera-
tion. The energy loss and modulus vary
not only from one compound to another
but depend also on the stress frequency
and the temperature for a given com-
pound. The tread rigidity is directly re-
lated to the rubber compound shear
modulus G, Figure 2. As the rubber be-
comes softer when its bulk temperature
Ti increases, the tread rigidity will de-
pend on the tyre warm-up. As pointed
out in [1], higher cornering stiffness is
obtained at a lower tread temperature.
Accordingly, the influence of the ambi-
ent and road temperatures on the cor- Figure 3: Tyre structure deformations in cornering conditions: sidewall twisting due to the
nering stiffness is significant, as well as self-aligning torque and belt bending due to the lateral force
the operating condition history of the
tyre. As far as friction is concerned, two
stress mechanisms are involved in the
relative slippage between the rubber elastic properties of the rubber, and par- sliding speed, the peak value of the µ(T)
compound and the road surface. The ticularly its hysteresis at a high frequen- curve is shifted towards a higher temper-
first mechanism is the frequency excita- cy, play an important role. As a result, ature. Finally, as pointed out in [15], the
tion of the material by the road texture. the friction properties of a tread material friction coefficient µ tends to decrease
The rubber is distorted when it slips over sliding on a road surface are a function with increasing contact pressure.
the rough spots on the road, thus lead- of the contact pressure, the sliding speed
ing to a loss of energy, the so-called in- and the contact temperature, Figure 2. 2.3 Thermal Models
dentation mechanism. The second mech- Generally speaking, for a given sliding The friction force is coupled to the con-
anism is molecular adhesion, which speed Vg , the friction curve µ(T) shows a tact temperature evolution along the CP,
comes into play at a scale of one hun- peak value that can be significantly Figure 2. The contact temperature is esti-
dredth of a micron, and which is ampli- greater than the friction at a very low or mated according to the theory of two
fied by slippage. In both cases, the visco- very high temperature. With increasing uniform semi-infinite media in a perfect
ATZ 05I2008 Volume 110
thermal contact. In the sliding area of is more and more important. The belt gles are shown in Figure 6. The results cor-
the CP, the dissipation of energy due to compliance S2 tends to reduce the tread respond to slip angle sweeps from 0° to -
friction leads to a rise in the contact tem- shear force, while sidewall twisting leads 15° (ramp-up) and from -15° to 0° (ramp-
perature. The dissipation is modelled by to a reduction in the effective slip angle down), at 3000 N and 7800 N tyre loads
a heat flux at the interface between the of the CP. Therefore, Dz is found to de- respectively. At a high slip angle, the lat-
rubber and the road surface, Eq. (6): crease at a very high load. Moreover, the eral force is mainly governed by the fric-
casing and especially the sidewalls be- tion force. A significant tyre surface tem-
ϕ = µpVg Gl. (6) come stiffer at a higher inflation pres- perature is built up due to the friction
sure, so that Dz increases with the infla- heat flux. Because of the thermal inertia
In addition, the surface and internal tion pressure at a high load. It is worth of the rubber, the surface temperature
temperatures Ts and Ti of the tread are noting that the model can predict Dz as during the ramp-down is higher than
calculated from the one-dimensional well as the other tyre stiffnesses Gz, Hz during the ramp-up. Accordingly, the re-
thermal equation in the tread thickness, and Jz with a very good accuracy. sults show a loss of about 10 % of the
averaged over the tyre width and over a maximum lateral grip between the ramp-
period of rotation, Figure 4. The heat 3.2 Influence of the Tyre Temperature up and the ramp-down. The TaMeTyre
transfer at the tyre surface is governed by Measurement data and model predic- model can predict the influence of the
conduction to the track, as well as by tions of the lateral force at high slip an- temperature on the lateral grip evolution
convection to the air and heating due to
friction. Compression and shearing in
the contact patch are responsible for an
additional heating in the core of the Figure 4: Thermal modelling
tread due to the rubber hysteresis at the of the tyre tread tempera-
tyre rolling frequency [16]. ture over a period of wheel
rotation
3 Application to a Passenger Car Tyre
Measurements on a flat-track machine
were carried out for a Michelin Primacy
passenger car tyre (size 225/55 R16). The
test protocol consisted of several sequenc-
es: cornering stiffness at three inflation
pressure and slip angle sweeps between -
15°, 15° at different loads and travelling
speeds (40 km/h, 80 km/h and 160 km/h).
The surface temperature was measured
at the top of the tyre using an infrared
thermal camera.
3.1 Tyre Stiffness Prediction
The cornering stiffness (Dz), self-aligning
torque stiffness (Gz), camber force stiff-
ness (Hz) and self-aligning torque stiff-
ness due to camber (Jz) at the three infla-
tion pressures are shown, Figure 5. Both
measurement data and simulations are
depicted. The cornering stiffness Dz is
one of the main tyre characteristics gov-
erning vehicle dynamics at a small later-
al acceleration. At a small load, Dz is
mainly governed by the CP dimensions
and the tread stiffness. The influence of
the casing parameters S2 and KT is negli-
gible. When either the load increases or
the inflation pressure decreases, the con- Figure 5: Tyre stiffness evolution with load at three inflation pressures (1.8 b, 2.4 b and 3.0 b):
tact patch becomes longer and wider, cornering stiffness (Dz), self-aligning torque stiffness (Gz), camber force stiffness (Hz) and
thus leading to an increase in Dz. At a self-aligning torque due to camber (Jz). The dots correspond to the measurement data and
higher load, the influence of the casing the lines to the TaMeTirE simulations
ATZ 05I2008 Volume 110
Development Tyres
at the different loads with a good accu-
racy. It is worth noting that the influ-
ence of the load on the lateral grip is also
mainly due to thermal effects. Indeed, at
a high load, the CP is longer. The heating
of the rubber due to friction is therefore
more important. This leads to a lower lat-
eral grip at a load of 7800 N than at 3000
N, Figure 6.
3.3 Influence of the Travelling Speed
The measurement data and the simula-
tions show a decrease in lateral grip
when the travelling speed increases, Fig-
ure 7 and Figure 8. Between 40 km/h and
160 km/h, the loss of lateral grip can be
higher than 10 %. As the travelling speed
increases, the sliding speed inside the CP
increases in the same proportion. Fur-
thermore, as one of the main mecha-
nisms governing friction is the frequency
excitation of the material by the road Figure 6: Influence of the thermal effects on the lateral grip (Fy/Fz) during a slip angle sweep
texture, the primary effect of the sliding at 3000 N and 7800 N tyre load respectively. Symbols: measurement data; lines: model predic-
speed on the grip is to change the fre- tions. The circle and solid line corresponds to the 3000 N tyre load. Crosses and dashed line
quency excitation of the rubber. The sec- corresponds to the 7800 N tyre load
ond effect, which is at least as important
as the frequency, is the friction reduc-
tion due to higher contact temperatures.
Indeed, the friction heat flux increases
with the sliding speed, Eq. 6. the sliding area of the contact patch. This fast lap times in which the evolution of
enables fast calculations of tyre forces the tyre surface temperature is signifi-
and moments to be performed for vehi- cant. At small accelerations, the vehicle
4 Conclusion cle dynamic simulation with a capability dynamics are mainly governed by the
for real-time simulation. Measurements tyre stiffness Dz, Gz, Hz, Jz. These charac-
TaMeTirE, the new physical tyre model, on a flat-track machine show the strong teristics depend on the tread stiffness
has been developed to improve tyre force influence of the thermal effects and trav- and, accordingly, on the internal tem-
and moment predictions. The model cal- elling speed on the lateral force. The vari- perature of the tread. Therefore, the use
culates the longitudinal and lateral forc- ations in the lateral grip can be higher of TaMeTirE should also improve vehicle
es in combined conditions according to a than 10 %. Such evolutions are complex simulations in the range of low lateral
one-dimensional brush-element ap- and depend on the rubber operating con- accelerations. The TaMeTirE coefficients
proach. The forces are coupled to the tyre ditions in the contact patch: the contact can be obtained from specific tyre meas-
temperatures by introducing the rubber temperature, the sliding speed and the urements on a flat-track machine. Me-
compound properties of the tread. The contact pressure. These depend on the chanical, friction and thermal parame-
shear modulus decreases with increas- tyre design parameters, such as the tread ters are derived from the tyre force and
ing internal temperature and the friction and casing stiffnesses and on the CP di- moment characteristics and from the
is governed by the contact temperature, mensions. They also depend on the tyre tyre temperature measurements. The ap-
sliding speed and contact pressure. An operating condition history, which gov- proach is currently under validation for
analytical thermal model based on the erns the tyre temperatures. The main a wide range of tyre types and sizes.
theory of two semi-infinite media in per- benefit of TaMeTirE is its ability to pre-
fect contact is used for the contact tem- dict the tyre force and moment in ac-
perature. In addition, the average surface cordance with the actual driving condi- References
and internal temperatures of the tread tions of the vehicle. As temperature and [1] Février, P.; Fandard, G.: Thermal and mechanical
over a period of wheel rotation are calcu- speed effects are taken into account, tyre tyre force and moment modelling, Tyre.Wheel.
Tech, TÜV Süd Automotive Gmbh Congress,
lated from the one-dimensional thermal simulations can be carried out for a wide
Munich December 5th-6th 2006
equation in the tread thickness. The non- range of vehicle manoeuvres. Vehicle [2] Buisson, J.: Michelin indoor characterization for
linear thermal and mechanical equa- simulations can be improved especially handling applied mathematical formulae, 15.
tions are derived assuming a single tran- for manoeuvres with high lateral accel- Aachener Kolloquium Fahrzeug- und Motorentech-
sition point between the adherent and eration, such as safety manoeuvres and nik, Aachen October 9th-11th 2006
ATZ 05I2008 Volume 110
[8] Miyashita, N.; Kawazura, T.; Kabe, K.: Analytical
model of mu-S curve using generalized skewed-
parabola, JSAE Review 24 (2003), pp. 87-92
[9] Pauwelussen, J. P.: The local contact between tyre
and road under steady state combined slips,
Vehicle System Dynamics, Vol. 41, 2004
[10] Gäfvert, M.; Svendenius, J.: A novel semi-empiri-
cal tyre model for combined slips, Vehicle System
Dynamics, Vol. 43, 2005
[11] Velenis, E.; Tsiotras, P.; Canudas-de-wit, C.; Sorine,
M.: Dynamic tyre friction models for combined lon-
gitudinal and lateral vehicle motion, Vehicle
System Dynamics, Vol. 43, 2005
[12] Holtschulze, J.; Goertz, H.; Hüsemann, T.: A simpli-
fied tyre model for intelligent tyres, Vehicle System
dynamics, Vol. 43, 2005
[13] Miyashita, N.; Kabe, K.: A new analytical tyre
model for cornering simulation. Part II: Cornering
force and self-aligning torque, Tyre Science and
Technology, Vol. 34, April-June 2006
[14] Février, P.; Fandard, G.: A new thermal and me-
chanical tyre model for handling simulation, 11th In-
ternational VDI congress, Tyres-Chassis-Road,
Hanover October 23rd and 24th 2007
[15] Guo, K.; Zhuang, Y.; Lu, D.; Chen, S. K.; Lin, W.: A
Figure 7: Influence of the travelling speed on the lateral grip (Fy/Fz) at a 3000 N tyre load. study on speed-dependent tyre-road friction and its
Symbols: measurement data; lines: model predictions effect on the force and the moment, Vehicle Sys-
tem Dynamics, Vol. 43, 2005
[16] Février,P.; Le Maitre, O.: Tyre temperature model-
ling: application to race tyres, proceedings VDI 13th
International Congress numerical analysis and
simulation in vehicle engineering, Würzburg Sep-
tember 27th-28th 2006
Figure 8: Influence of the travelling speed on the lateral grip (Fy/Fz) at 7800 N tyre load.
Symbols: measurement data; lines: model predictions
[3] Clark, S. K.: Mechanics of Pneumatic Tyres, U.S. edition, SAE Inc 1996
Department of Commerce, National Bureau of [6] Hallum C. Understanding Race Tyres, Motorsports
Standards, November 1971 Engineering Conference Proceedings, Vol. 1:
[4] Sakai, H.: Theoretical and experimental studies on Vehicle Design and Safety, SAE 1998 Download des Beitrags unter
the dynamic properties of tyres. Part I: Review of [7] Lacombe J.: Tyre model for simulations of vehicle www.ATZonline.de
theories of rubber friction. International Journal of motion on high and low friction road surfaces,
Vehicle Design, Vol. 2, No. 1, 1981 Proceedings of the 2000 Winter Simulation Confer- Read the English e-magazine.
Order your test issue now:
[5] Dixon, J.C.: Tyre Suspension and Handling, 2nd ence 2000 [email protected]
ATZ 05I2008 Volume 110
