Tyre Friction Potential Estimation by Aligning Torque and Lateral Force Information

The modern active systems and the advanced driver assistance systems have certainly reduced the amount of accidents. However these systems are still lacking one major information, which is the friction between the tyre-road interface. The performance of these systems could be enhanced substantially if this information was known. Rather simple methods can be used to determine the friction coefficient at the current operating point of the tyre. The challenge arises from the friction potential, which is the maximum friction coefficient that the tyre-road interface can produce. The objective of this master's thesis is to study the feasibility of estimating the friction potential by using the information of lateral force and the self aligning torque of the tyre. The well-known bicycle model of the vehicle is used to determine the lateral force of the front axle, which is divided to left and right side in accordance with the normal load distribution of the wheels. The normal loads of the wheels are calculated from the angle sensors, which are assembled in the transverse control arms of the vehicle. The same sensors are also exploited to determine the inclination angles of the steering axes. The heart of the estimation method is the brush tyre model. It receives the self aligning torque, lateral and normal force of the tyre as inputs and outputs the friction potential. The proving ground tests are done with a typical small family estate car (VW Golf Variant Mk5). Two distinguishing proving grounds are chosen for illustrating the operation of the estimation method on high and low friction level surfaces. Tests also in μ-split conditions and in situation where the vehicle is travelling from high friction level surface to low friction level surface are implemented. Results from both the high and low friction level surface illustrate that the estimation method is able to detect the friction states of the front tyres in steady state cornering manoeuvres. Remarkable and interesting results are found from the μ-split and surface transition tests were the estimation method distinguishes the difference between the highand low friction level surfaces. The presented friction estimation method has potential. However lots of work and further studies have to be conducted before this method can assist the modern active safety systems and the upcoming ADAS systems. Date: 24.11.2010 Language: English Number of pages: 137