Вы находитесь на странице: 1из 8

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/268808529

Interaction models

Article · November 2014

CITATIONS READS

0 53

1 author:

Flavio Farroni
University of Naples Federico II
33 PUBLICATIONS 138 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Tire/Road Interaction Analyisis, Modelling and Simulation View project

MegaRide - applied vehicle research - an UniNa spin-off company View project

All content following this page was uploaded by Flavio Farroni on 27 November 2014.

The user has requested enhancement of the downloaded file.


VISIT WWW.VEHICLEDYNAMICSINTERNATIONAL.COM FOR EXCLUSIVE NEWS AND FEATURES

VEHICLE
INTERNATIONAL
www.VehicleDynamicsInternational.com

MX-5
Will the next generation
of Mazda’s iconic roadster
be another dynamic
benchmark?

Interaction models Critical condition Gravitational effects Non-linear modeling


A grip and thermodynamics- Exploring new methods for Unique insight into the relationship Innovative non-linear modeling,
sensitive tire-road interaction defining and detecting critical between vehicle control and the virtual sensing and control
force characterization procedure situations in vehicle dynamics nature of Earth’s gravity technologies for dynamics

in this issue
 



!LSBOPQLOV
Mazda MX-5
 Ahead of the fourth generation’s arrival in 2015,
we take a closer look at the car’s technology with
its program manager, Nobuhiro Yamamoto



Product &
service profiles
46 High-performance brakes
50 Effective code generation
What’s new? Technical papers
 Volkswagen Passat  Detecting critical situations 52 Hydraulic steering assist
One of the most important cars for the Volkswagen Sébastien Varrier from lcis and Damien Koenig 54 Future AWD technologies
Group enters its eighth generation and John J Martinez from GIPSA-Lab present their
findings on new methods for detecting critical 56 Model-based engineering
 Land Rover Discovery Sport situations in vehicle lateral dynamics
The new, small SUV from Land Rover makes 58 Sensor solutions
use of an innovative multilink rear axle  Interaction models 60 Virtual development
An analysis of the results provided by a grip and
Columns thermodynamics-sensitive tire/road interaction
force characterization procedure. By Flavio Farroni,
63 ADAS testing
 On the job: John Miles University of Naples
64 High-quality bearings
John considers the many highs and lows 67 High-temperature testing
of Lotus Cars’ 62-year history  Gravitational effects
A study into vehicle control and the innate nature 68 Vibration analysis
 Home truths: John Heider of gravity, by S2AB
As autonomous vehicles near reality, John 70 Autonomous braking
contemplates the future role of dynamicists  Non-linear modeling
Innovative non-linear modeling, virtual sensing
and control technologies for vehicle dynamics
Regulars applications, by Modelway
 Dynamic Legends
The original MX-5, the car that revived the two-
seater, attainable sports car segment, and which
remains a dynamics benchmark to this day

"OOVBM4IPXDBTFt7FIJDMF%ZOBNJDT*OUFSOBUJPOBMDPN

tire/road interaction

Interaction
models
AN ANALYSIS OF THE RESULTS PROVIDED BY A GRIP AND THERMODYNAMICS-SENSITIVE TIRE/ROAD
INTERACTION FORCE CHARACTERIZATION PROCEDURE. BY FLAVIO FARRONI, UNIVERSITY OF NAPLES

FIGURE 1: MODEL INTEGRATION The automotive sector is


SOLUTIONS looking for the optimal
solution in modeling
FIGURE 2: G-G DIAGRAM REALIZED
BOTH WITH EXPERIMENTAL DATA AND
and understanding
WITH RESULTS OF A SIMULATION tire behavior in experimental
PERFORMED WITH STARTING and simulation environments. 1,2,3
MF-TIRE PARAMETERS SET The studies and tools described
here represent a new approach
FIGURE 3: DETAIL OF THE FRONT
in tire characterization and vehicle
(LEFT) AND REAR (RIGHT) TIRE SLIP
ANGLES, BOTH FROM EXPERIMENTAL simulation procedures, leading to
DATA AND FROM OUTPUTS OF A the complete reproduction of the
SIMULATION PERFORMED WITH dynamic response of a tire and of
STARTING MF-TIRE PARAMETERS SET FIGURE 1
its frictional and thermodynamic
behavior simply by means of specific history and containing force and slip account the combined slip condition
track sessions and a few laboratory estimations useful to provide tire and the effects of vertical load and
measurements. This represents interaction characteristics. camber angle on the overall grip.
a bridge between a robust and Complete and detailed studies TRT (Thermo Racing Tyre)7 is
widespread approach, like Pacejka’s, of tires in a wide range of working an analytical-physical thermal tire
and purely physical modeling, that conditions are commonly carried model developed with the aim of
satisfies predictive requests and the out by means of complex, bulky predicting temperature with a high
need for deeper knowledge about and expensive test benches.6 The degree of accuracy and able to
complex phenomena. proposed procedure means the simulate the high-frequency thermal
vehicle can be employed as a moving dynamics characterizing high-
The tools lab, easily applying experimental performance systems. The model can
The final product is composed of and processing techniques. estimate the temperature distribution
the following four tools, which can TRIP-ID (Tyre/Road Interaction of even the deepest tire layers,
cooperate to form a multitude of Parameters Identification) provides usually not easily measurable online,
solutions. an innovative procedure to identify to predict the effects that fast
TRICK (Tyre/Road Interaction the Pacejka coefficients, starting temperature variations induce in the
Characterization & Knowledge)4 is from the experimental tests carried behavior of viscoelastic materials,
basically composed of a vehicle model out to measure global vehicle data and to take into account the
able to process experimental signals during outdoor track sessions. The dissipative phenomena related
acquired from the vehicle’s CANbus procedure collects and processes the to tire deformations.
and from additional instrumentation data provided by TRICK, eliminating GrETA (Grip Estimation for Tyre
(DATRON5) to estimate sideslip angle, the outlier points, discriminating Analyses)8 is a tire/road friction
providing a sort of virtual telemetry, between the various tire wear and physical model, developed to respond
based on the acquired signals’ time thermal phenomena, and taking into to the needs of race teams and tire
FIGURE 2 FIGURE 3

7FIJDMF%ZOBNJDT*OUFSOBUJPOBMDPNt"OOVBM4IPXDBTF

tire/road interaction

manufacturers, able to provide FIGURE 4: PLOT F (LEFT): STARTING


FIGURE 4
an effective calculation SET, FRONT TIRE, PURE LATERAL
INTERACTION. PLOT F (RIGHT):
of the power dissipated by road
IDENTIFIED FINAL SET, FRONT TIRE,
asperities indented in the tire tread, PURE LATERAL INTERACTION. PLOT
taking into account the phenomena H (LEFT): STARTING SET, REAR TIRE,
involved with adhesive friction, PURE LATERAL INTERACTION. PLOT
expressed by means of an original H (RIGHT): IDENTIFIED FINAL SET,
formulation whose parameters are REAR TIRE, PURE LATERAL
INTERACTION
identified thanks to dedicated
experimental tests. FIGURE 5: (RIGHT) G-G DIAGRAM
These tools are able to describe and REALIZED BOTH WITH
analyze aspects of phenomena EXPERIMENTAL DATA AND WITH
concerned with tire/road interaction, RESULTS OF A SIMULATION
PERFORMED WITH THE FINAL
but their cooperation can constitute
IDENTIFIED MF-TIRE PARAMETER
an even more powerful instrument SET. (LEFT) FIGURE 2 FOR
to extend the comprehension of such COMPARISON
a complex theme.
A general overview of the
developed models and procedures
is shown in Figure 1, in which it is
possible to observe the connections
that link the models, providing
different solutions for employment.
Finally, there is an absence of grip addition to grip overestimation
TRICK and TRIP-ID and stiffness variations due to (and consequent to it), a continuous
TRICK and TRIP-ID were developed thermal effects. and massive heat generation at the
with the initial aim of increasing the In Figure 2, a G-G diagram, contact interface, which increases
confidence of car makers in adopting a classic and simple instrument tire temperature. As is well known,
the Magic Formula in virtual drive employed to evaluate global vehicle one of the main effects of
modeling and vehicle dynamics performance,9 is plotted (in temperature on tires is stiffness
models employed for predictive nondimensional form, as are all variation2 (increasing temperature
performance analyses. One of the the following ones, for reasons of causes decreasing stiffness),
main advantages of the tool is the industrial confidentiality), comparing particularly evident in the front
ability to validate Pacejka coefficients on a reference track lap the measured tires, which in high-performance
provided by tire makers, or even vehicle accelerations with applications are typically narrower
do without their contribution, accelerations exported as output and less thermally inert than the
identifying coefficients after from a commercial, highly validated rears. Figure 3 focuses on these
a proper vehicle characterization vehicle simulation model employed considerations, highlighting the
and a specific track session. in a virtual driving simulation unsatisfactory results obtained with
The weak points of the initial MF- environment that has been adopted respect to front slip angles employing
Tyre parameter set identified by tire for the specific MF-Tyre parameter the starting tires’ parameter set; the
companies via bench procedures, and set provided by the tire maker. imbalances caused by poor estimation
highlighted by data analysis, are High grip levels reached by bench- of slip angles act on the whole
as follows: First, there is too much tested tires are often due to the vehicle’s tendency to understeer
grip in the longitudinal and lateral testing countersurface being abrasive or oversteer.10
interaction, due to the differences paper (or rough material characterized The identification of the optimal
between real roads and the belts by low macro-roughness), that is able parameter set by means of the TRIP-
employed for testing. to maximize the contact patch’s ID procedure also enables us to solve
Next, there is a lower likelihood effective area, providing an interface the simulated vehicle driveability
than in reality for the driver to be better than a real road. problems linked to the shape of the
able to stabilize the vehicle after the The employment of abrasive paper tires’ starting set. The mentioned
limit of adhesion has been crossed. and severe testing cycles causes, in lower-than-reality likelihood of the
driver returning the vehicle to a
FIGURE 5 stable condition after the limit of
adhesion has been crossed is due
to two factors: an excessively peaky
trend of lateral interaction curves
and a too-sharp decrease of cornering
force in a combined interaction at
increasing values of slip ratio. The
improvement that the characterized
tires have represented with respect
to the cited effects can be observed
in Figure 4, which compares the
starting set’s pure lateral interaction
curves (on the left, in black) with

"OOVBM4IPXDBTFt7FIJDMF%ZOBNJDT*OUFSOBUJPOBMDPN

tire/road interaction
FIGURE 6: DETAIL OF TIRE SECTIONS
CUT ALONG THE MERIDIAN PLANE

FIGURE 7:
(A) THE STABILITE 2017 ARGON-
FIGURE 6
KRYPTON LASER
(B) THE PHOENIX THERMOGRAPHIC
those of the optimal identified set CAMERA
(on the right, colored). (C) THE TI-45 THERMOGRAPHIC
It can be seen that data collected CAMERA
during an experimental session and (D) THE LASER SPOT ON THE TIRE
processed by the TRICK tool is able EXTERNAL SURFACE
(E) A THERMOGRAPHIC CAMERA
to provide information useful in
IMAGE OF THE LASER SPOT ON
modifying the starting set, obtaining THE TIRE EXTERNAL SURFACE
an identified set that results in good (F) A THERMOGRAPHIC CAMERA
agreement with the drivers’ requests IMAGE OF THE LASER SPOT ON
and with the objective data acquired THE TIRE INTERNAL SURFACE FIGURE 7

by equipping the vehicle with


FIGURE 8: COMPARISON OF STORAGE
measurement instruments. Figure 5 conductivity and specific heat to road asperities, optimizing contact
MODULUS (E’) BETWEEN A COMMON
shows the results of the simulations PASSENGER TIRE AND A GT SPORT
measurements. Tire layers need to be area at the price of a lower wear
performed with the identified tire TIRE characterized from a thermodynamic resistance.
set, comparing them with the ones point of view, focusing in particular Passenger tires are more stable and
relative to the starting set, shown FIGURE 9: COMPARISON OF TAN 6 on conductivity and specific heat able to offer good adhesion levels
on the left. BETWEEN A COMMON PASSENGER measurements. A standard test even at very low temperatures, being
TIRE AND A GT SPORT TIRE
procedure is carried out employing adapted to the widest possible range
TRICK and TRT a Stabilite 2017 argon-krypton laser of working conditions. Figure 9 shows
TRICK and TRT can be successfully (Figure 7a) pointed at the whole tire in a clear plot the possible reason
employed together, providing an or on specimens of each layer and for the so-called ‘feeling the grip’
instrument able to provide tire emitting a beam of given power. phenomenon. Sport tires, as
thermal analyses, useful to identify Knowing the specimen thickness and distinct from passenger ones, are
the range of temperature in which measuring temperature of the two characterized by a clear relative
grip performances are maximized surfaces by means of two maximum at about 42°C and by
and enabling the optimal tires and thermographic cameras (a Flir higher values of tan b at the usual
vehicle setup to be defined. Phoenix, Figure 7b and a Fluke Ti-45, usage temperatures.
The test procedures adopted to Figure 7c), it is possible to provide Specifying that the DMA test has
characterize the tires, obtaining an effective estimate of the desired been carried out at a frequency of
data useful to initialize the models parameters, validated thanks to 1Hz, notably different from common
properly, can be schematically comparison with tests carried out tread stress frequencies, a quick
divided into two main subcategories with a COND1 device, following calculation, hypothesizing an average
– destructive and non-destructive. certified procedures.11 road macro-roughness wavelength h
To the first belongs meridian plane The third component is DMA equal to 0.01m and an average
section analysis. This kind of test viscoelastic characterization. Tests sliding speed Vs of 5m/sec, enables
consists of the observation and carried out on sport tires have the real tire temperature at which
measurement of the thickness of highlighted interesting aspects, in the tan b maximum can be
the layers constituting the meridian particular comparing results with experienced by the driver to be
section. In Figure 6, it is possible common passenger tire ones. Figure 8 estimated. Applying a simplified
to distinguish the tread layer, shows that, as expected, sport tires version of the WLF equation,8 it
characterized by an evident and deep are characterized by lower storage is possible to obtain, which, added
pattern, the bulk layer, in which modulus values in their optimal to the starting 42°C, gives a
steel cord plies are clearly observable, thermal working range (35°C and temperature of 63.6°C, in accordance
and the innerliner, which is very above) that enable them to offer with the experimental value shown
thin and impermeable. better adhesion and to adapt better in the analyses already presented.
The second component of the
destructive subcategory is thermal

FIGURE 8 FIGURE 9
PASSENGER TIRE
GT TIRE PASSENGER TIRE
GT TIRE

"OOVBM4IPXDBTFt7FIJDMF%ZOBNJDT*OUFSOBUJPOBMDPN

tire/road interaction

FIGURE 11

FIGURE 10 FIGURE 12

Non-destructive testing procedures coefficient and temperature are TRT output as an input for GrETA, FIGURE 10: SCANS OF A GT TIRE
The first of these to be applied is based on the only thermal data which can be used to introduce into CONTACT PATCH UNDER DIFFERENT
TESTING CONDITIONS AT ZERO
contact patch analysis. A specific experimentally available, i.e. the the Pacejka interaction model the
CAMBER ANGLE. IT IS NOTICEABLE
test bench12 is used to apply a static tire’s external (and in a few cases, dependencies on temperature, tire THAT AT INCREASING LOAD THE
vertical load to the tested tires, internal) temperature, measured working variables, road roughness CONTACT AREA INCREASES,
analyzing contact patch extension using a great variety of techniques. and compound characteristics. PROGRESSIVELY INSERTING
and pressure distribution. It is A typical correlation between lateral The advantages coming from the SHOULDERS IN THE INTERACTION
possible to interpose pressure- grip and measured temperature cooperation of these models can be ZONE. AT HIGH INFLATION PRESSURE
THE CENTRAL RIB IS MORE EXTENDED,
sensitive Prescale sheets between the appears like that shown in Figure 13, summarized in the following three
WHILE LOWER PRESSURE TENDS
tire and the flat steel countersurface, from which very little information points, which have already been TO OVERLOAD THE SHOULDERS
and to plan tests at different loads, can be deduced. exploited (further application
inflation pressures and camber Thanks to the availability of possibilities are clearly available). FIGURE 11: INSTALLATION OF
angles. In Figure 10, the results the bulk temperature, it is possible The first is the prediction of tire INFRARED THERMAL SENSORS
of a zero-camber testing session to provide much more useful behavior on the various tracks of AND LOCALIZATION INSIDE VEHICLE
WHEELHOUSES
on a front tire are reported. correlations, such as the ones shown a racing championship, each
The second is track thermal tests. in Figure 14, from which an optimal characterized by different roughness FIGURE 12: TRT RESULTS EVALUATION
These sessions are carried out to thermal range can be identified. The (previously measured) and weather FOR FRONT AND REAR TIRES
a specifically developed procedure, reason why the bulk temperature conditions.
with the aim of collecting tire data offers better results can be attributed The next is a performance
under various thermal conditions. to the fact that the surface evaluation of the characteristics of
In order to acquire tire temperature, temperature varies with very fast various compounds, which enables
the vehicle is equipped with infrared dynamics but it is not possible to a dialog with the tire makers to
sensors installed in the wheelhouses modify the polymers’ characteristics be established, directing tire
(Figure 11) and directed on the tread quickly enough to see the response construction and compound
surface. The signals are acquired by of the whole tire’s frictional behavior. development to the achievement
Dewesoft hardware. Each tire tread Bulk temperature, on the other hand, of a common aim.
is interrogated by two different can be considered to be the tread’s The third is the definition of an
measurements, particularly useful core temperature, more resistant to optimal vehicle setup in terms of
for front tires, which when steering fast variations and directly connected wheel angles, load balance and tire
could be characterized by to the rubber’s viscoelastic state. inflation, and of driving strategies
discontinuous temperature profiles. As a further validation of the that are able to reach optimal grip/
After carrying out the track described procedure, it can be seen thermodynamic conditions.
experimental session and acquiring that the optimal temperature value Figures 15 and 16 show the
data to be processed by the TRICK is in good agreement with the differences between force data from
procedure, a ‘virtual telemetry’ theoretical result seen in Equation 1 telemetry and from the Pacejka
is generated. (see page 29), confirming that the model, whose inputs are the
Speed, slip, camber and force thermal model can be employed as a measured slip, load and camber. In
channels are used as input for TRT, predictive instrument to investigate the first case the calculated forces
whose results are compared with performance optimization strategies are reported as scaled by a Coulomb
the measured surface temperatures and that a proper knowledge of friction model, always equal to one
(Figure 12), delivering good polymer characteristics can be except for the static value (which
correspondence with available a useful starting point to a better means using a standard Pacejka
data and, very usefully for the understanding of the dynamics output, with no further processing).
grip analysis discussed in the of tire-surface interaction. In the second case, the forces are
following, an estimation of tire- processed with GrETA friction scaling
bulk temperature. TRIP-ID, TRT and GrETA factors, taking into account
Common analyses concerning the The thermal and grip models can phenomena neglected in the first
relationship between the tire friction usefully cooperate, employing the case. It can be seen that employing

"OOVBM4IPXDBTFt7FIJDMF%ZOBNJDT*OUFSOBUJPOBMDPN

tire/road interaction

FIGURE 13

FIGURE 15 FIGURE 14
FIGURE 13: FRONT AND REAR LATERAL the grip model produces better the skilled and stimulating academic temperature distribution, Meccanica,
GRIP REPORTED AS A FUNCTION results, particularly with respect environment of the Industrial Vol. 49, N. 3, pp707-723 (2014)
OF MEASURED TIRE SURFACE to longitudinal interaction in the Engineering Department of the 8) Farroni F, Russo M, Russo R,
TEMPERATURE
traction phase, which is thermally University of Naples Federico II and Timpone F, A physical-analytical
FIGURE 14: FRONT AND REAR GRIP stressful for high-performance tires the support of several companies, model for a real-time local grip
REPORTED AS A FUNCTION OF TIRE and able to generate heat for the motorsport teams and research estimation of tire rubber in sliding
BULK TEMPERATURE ESTIMATED BY friction power mechanism, which institutes who are all experts in contact with road asperities,
MEANS OF TRT. BELL-SHAPE CURVES induces noticeable effects in tire/ their fields. Proceedings of the Institution of
HAVE BEEN DRAWN TO HIGHLIGHT road interaction modeling. Mechanical Engineers, Part D: Journal
THE TRENDS
References of Automobile Engineering, Feb 14,
FIGURE 15: COMPARISON BETWEEN Physical interaction model 1) Guiggiani M, The Science of Vehicle 2014, DOI: 10.1177/
LONGITUDINAL TIRE FORCES and further developments Dynamics, Springer (2014) 0954407014521402
MODELED BY MF WITH A COULOMB The Pacejka model is not the most 2) Gent A N, Walter J D, The 9) Milliken W F, Milliken D L, Race Car
FRICTION LAW AND WITH GRETA flexible and detailed method to Pneumatic Tire, Vehicle Dynamics, SAE International
FRICTION MODEL
describe local phenomena of tire/ University of Akron, NHTSA (2005) (1995)
FIGURE 16: COMPARISON BETWEEN
road interaction, but represents 3) Pacejka H B, Tire and Vehicle 10) Farroni F, Russo M, Russo R, Terzo
LATERAL TIRE FORCES MODELED BY a very robust and intuitive solution Dynamics, Butterworth-Heinemann M, Timpone F, A Combined Use of
MF WITH A COULOMB FRICTION LAW to obtain the barely achievable aim (2006) Phase Plane and Handling Diagram
AND WITH GRETA FRICTION MODEL of modeling tire tangential forces. 4) Vehicle Dynamics International, Method to Study the Influence of Tire
For this reason, further May/June 2014, UKIP Media & Events and Vehicle Characteristics on
developments of the activities Ltd (2014) Stability, Vehicle System Dynamics,
discussed in the present work will 5) Bevly D M, Daily R, Travis W, Vol. 51, N. 8, pp1265-1285 (2013)
focus on the realization of a fully Estimation of Critical Tire Parameters 11) Buonanno G, Carotenuto A,
physical interaction model that, Using GPS Based Sideslip Giovinco G, Massarotti N, Experimental
starting from the knowledge acquired Measurements, Proc. SAE Dynamics, and Theoretical Modeling of the
about the topic by the vehicle Stability, and Controls Conf., Novi, Effective Thermal Conductivity of
dynamics research group,13,14 will be USA, pp87-94 (2006) Rough Steel Spheroid Packed Beds,
able to interact deeply with the other 6) Cabrera A, Ortíz A, Simón A, García Journal of Heat Transfer, Vol. 125,
models, creating an analytical and F, Pérez La Blanca A, A Versatile Flat N.4, pp693-702 (2003)
predictive instrument that can be Track Tire Testing Machine, Vehicle 12) Farroni F, Rocca E, Timpone F,
employed in a wide range of System Dynamics, Vol. 40, N.4, A Full Scale Test Rig to Characterize
automotive applications. pp271-284 (2003) Pneumatic Tire Mechanical Behavior,
7) Farroni F, Giordano D, Russo M, International Review of Mechanical
Acknowledgements Timpone F, TRT: thermo racing tyre Engineering (I.RE.M.E.), Vol. 7,
The author wishes to acknowledge a physical model to predict the tire N. 5, pp841-847 (2013)
13) Capone G, Giordano D, Russo M,
FIGURE 16
Terzo M, Timpone F, Ph.An.Ty.M.H.A.:
a Physical Analytical Tire Model for
Handling Analysis - the Normal
Interaction, Vehicle System Dynamics,
Vol. 47, N. 1, pp15-27 (2009)
14) Calabrese F, Farroni F, Timpone F,
A Flexible Ring Tire Model for Normal
Interaction, International Review on
Modeling and Simulations (I.
RE.MO.S.), Vol. 6, N. 4 (2013)

7FIJDMF%ZOBNJDT*OUFSOBUJPOBMDPNt"OOVBM4IPXDBTF

View publication stats

Похожие интересы