Editorial Type:
Article Category: Research Article
 | 
Online Publication Date: 01 Jun 2010

The Impact of Tire Measurement Data on Tire Modeling and Vehicle Dynamics Analysis

and
Page Range: 155 – 180
DOI: 10.2346/1.3428967
Save
Download PDF

Abstract

Computer aided engineering tools play an important role in today’s vehicle development process. Today, overall vehicle dynamics analysis and chassis component fatigue resistance investigations can be carried out without the need for existing prototype hardware versions of the corresponding vehicle.

An accurate tire model is a key element in precise modeling of the vehicle and its components. All forces acting on the vehicle (except for aerodynamic forces) are transferred via the tires. Therefore, the tire and its modeled characteristics have a major influence on the results of vehicle dynamics analysis.

At present, many tire simulation models are available for application in vehicle dynamics analysis. To obtain the best possible performance from these models, a number of different tire measurements are required to support the tire model parameter identification process.

This paper presents a review of different tire simulation models and their required tire measurements. Depending on the test rigs used and the measurement procedures applied, the tire measurement results may be somewhat different. What is the impact of these differences on the tire modeling performance and the vehicle dynamics analysis output? This paper gives an answer.

Keywords: tire force and moment measurements; tire test rigs; tire simulation; tire modeling; vehicle dynamics analysis; vehicle modeling; tire measurements
Copyright: The Tire Society
FIG. 1
FIG. 1

Typical full-vehicle simulation model (the picture shows a MSC ADMAS/Car model) with the main groups of chassis parts and components that have to be considered in order to obtain a realistic representation of the real world vehicle [8]. One important component is the tire simulation model.

FIG. 2
FIG. 2

Variation in tire longitudinal force transfer characteristics on different dry road surfaces at indoor test rigs (left) and on public roads outdoor (right); the presented brake slip characteristic curves were determined by Roth and Bachmann, TU Darmstadt [14,15]. The diagrams very clearly shows the great difference of possible maximum tire forces even if the road surface of interest is identified as dry.

FIG. 3
FIG. 3

Typical input signals and output data of a mathematical-empirical or semiphysical tire simulation model [1].

FIG. 4
FIG. 4

Selection of tire test rigs as applied for the F&M measurements on tracks with different surface curvatures: one flat belt machine and two test rigs with outer drum tracks. Flat belt as well as the outer drum track surfaces are coated with safety walk sandpaper with grit P80/P120 to simulate real road friction and roughness characteristics. All F&M measurements were carried out under dry conditions.

FIG. 5
FIG. 5

Tire F&M characteristics. Comparison between test rig measurements and model calculations from RWTH Aachen tire model TiMiS by Holtschulze [12,13]. In clockwise order, beginning top left: lateral force vs side slip angle, brake force vs brake slip, combined lateral-longitudinal slip, aligning torque vs side slip angle.

FIG. 6
FIG. 6

Input signals and output data of the complex 3D tire simulation model FTire by Professor Dr. Michael Gipser, Esslingen [10].

FIG. 7
FIG. 7

Curves Produced by the Sine and Cosine Versions of Pacejka’s Magic Formula [1,7,16,18,19].

FIG. 8
FIG. 8

Analytical physical approach to calculate tire F&M characteristics based on tire tread block shear deformation inside the contact patch. These sketches represent the semiphysical tire model from Holtschulze [12].

FIG. 9
FIG. 9

Vertical stiffness (left) and contact patch extension (right; at a wheel load of 5200N) of the Continental Premium Contact 2 205/55 R16 tire at 2.4 bar as measured on the three different tire test rigs shown in Fig. [4]. Vertical stiffness and contact patch extension grow with increasing drum diameter (infinite drum diameter for flat track) The contact patch extensions were determined applying the contact patch evaluation shown in Fig. [10].

FIG. 10
FIG. 10

Tire contact patch characteristics evaluation of the Continental Premium Contact 2 205/55 R16 tire at 2.4 bar. A pressure sensitive paper or film is used in order to determine the tire contact patch dimensions and pressure distribution. This film delivers a grayscale footprint bitmap. Its size and dimensions as well as the measured wheel load are used to calibrate the pressure distribution.

FIG. 11
FIG. 11

Lateral slip characteristics of the Continental Premium Contact 2 205/55 R16 tire at 2.4 bar as measured on the three different tire test rigs shown in Fig. [4]: Lateral force and aligning torque versus side slip angle.

FIG. 12
FIG. 12

Lateral slip characteristics of the Continental Premium Contact 2 205/55 R16 tire at 2.4 bar as measured on the three different tire test rigs shown in Fig. [4]: cornering stiffness and maximum lateral force coefficient reach their maximum values on a flat track surface.

FIG. 13
FIG. 13

Evaluation of contact patch wheel load dependencies for length, width, area size and average ground pressure for the Continental Premium Contact 2 205/55 R16 tire at 2.4 bar as measured on different track surface curvatures (see also Fig. [10]).

FIG. 14
FIG. 14

RWTH Aachen test track at ika: The test track is paved with a high level grip asphalt surface. The vehicle dynamics test drives took place there. Test drive car Audi A6 4.2l quattro equipped with Continental Premium Contact 2 205/55 R16 tires at 2.4 bar.

FIG. 15
FIG. 15

Matlab/SIMULINK model of the vehicle that was used for handling dynamics test drives. This model allows use of measured driver commands such as steering wheel angle and vehicle speed as input signals for model simulations. In addition different parameter files can be applied for the implemented tire model Magic Formula DelftTyre 5.2 [9,28,29].

FIG. 16
FIG. 16

Determination of important vehicle mass and inertia parameters on ika test facilities. The identification of these parameters is essential for the Matlab/SIMULINK vehicle model parameterization and a basis for reliable simulation results. Furthermore all chassis kinematics and compliance characteristics were measured on a corresponding test rig at ika [9].

FIG. 17
FIG. 17

Steady state cornering and self-steering properties: steering wheel angle versus lateral vehicle acceleration; comparison between measured vehicle test drive data and corresponding vehicle model simulation results with tire models based on measurement data from the different test rigs shown in Fig. [4]. The flat track tire model shows best results even though no big difference to the tire model based on drum D=2.55 m measurements can be observed. But still test drive results are not matched perfectly.

FIG. 18
FIG. 18

Double lane change maneuver at 60 km/h: comparison between measured vehicle test drive data and vehicle model simulations results with tire models based on measurement data from the different test rigs shown in Fig. [4]. Even when using tire data based on flat track measurements still a difference between simulation results and test drive data can be observed. Again no big difference between the flat track and the drum D=2.55 m tire models can be observed.

FIG. 19
FIG. 19

Indoor test rig F&M tire measurements as basis for tire model parameterization should be supported by suitable tire rubber—road surface friction coefficient measurements from test track and tire type that were used during vehicle dynamics test drives. This can be achieved by applying additional test devices such as a tread block friction measuring machine or a mobile tire test truck.

  • Download PDF