Tyre Sensing for Tyre Model Parametrization (TyreSens)

Forskningsprojekt, 2016 – 2018

Research Motivating Problem:

•The present tire model parameterization used in CAE suffers inaccuracy, making function
verification/validation on complete vehicle level difficult
•Current tyre parameterization based  on flat-track measurement that do not represent the used (asphalt) road surface
•Fast development cycles, including tires, make processes involving third parties
(measurement labs) impractical

Expected result:

A method to identify and parameterize a tyre model for vehicle dynamics CAE simulations by,

•Using an instrumented standard vehicle
•Perform a set of pre-defined test maneuvers in the data acquisition
•Extract tyre parameters from the data

Research questions:

•Which data is required to parametrize a given tyre model? (variations in camber, vertical force, carcass temperature etc.)
•Which is the minimum set of maneuvers for a given data set?
•Is controlled excitation useful/required, e.g. controlling the applied wheel torque through brake interface?
•Can flat-track measurements be used to improve the model fitting when data from the in-vehicle tests do not cover the relevant operating space of the vehicle?

Deliverables to which results will contribute:

•Proposed method  (maneuvers & excitation scheme) for tyre parameter identification
•Test vehicle for tyre parameter identification
•1 VehDyn PhD graduated

 ===

Tuning and verification of active safety systems, such as Electronic Stability Control (ESC) and Roll Stability Control (RSC), is currently a time consuming task that requires extensive testing on test tracks. In order to reduce the process time for development, tuning and verification of these systems there is currently a trend within the automotive industry to move towards virtual tuning and verification. Complexity of verification of autonomous driving functions also imply the necessity for simulation.

Virtual verification of safety functions and autonomous vehicles rely on accurate vehicle simulation models. Today, the correlation between simulations and test track measurements on a complete vehicle level is not good enough to be used with confidence in a verification context.

The lack of correlation between simulation and measurements are mainly explained through inaccurate tyre model parameters. The current state-of-the-art process used at car manufacturers for identifying tyre model parameters is through third party test houses and measurement from a flat-track tyre testing machine. Real road surfaces are not represented with sufficient accuracy for these types of measurements leading to inaccurate models. In addition, the process is expensive for the car manufacturers as well as inflexible during the fast turns in the vehicle development.

The project will investigate the possibility for an in-house process to determine tyre characteristics and model parameters through equipment and vehicles already existing at the car manufacturers. The potential benefits are flexibility, efficient use of resources and accuracy compared to state-of-the-art. The project will focus on the development of a method to collect measurement data and derive tyre model parameterizations. A substantial part of the project will investigate the validity of the proposed method.