Friction Coefficient Estimation Considering The Carcass Deflection

Nowadays, the sensors installed inside the tire have attracted more and more attentionin the domestic and overseas. Compared with the traditional body sensors, the spatialdistance to the tire of the new sensors is minimized, and a more accurate estimation of thekinematic conditions of the tire, dynamic properties and the adhesion characteristics of theroad would be obtained. And how to get this information from the output of the sensorsbecomes important and difficult.Considering the maximum friction coefficient, the existing research would choose totake the brush model as the estimator to deal with the corresponding tire forces. The brushmodel is simple and convenient, but when the sliding region of the whole contact area issmall, the results are undesirable. So this paper is aimed to find a new model to establish themaximum friction coefficient estimator to get a reliable estimator. The thesis is organized asfollows:The first chapter is the introduction part and describes the research background and itsmeaning. The existing two main methods to estimate the friction coefficient are alsodescribed. One is based on the force-slip curve and the other is based on the tire forces. Thepaper compares the pros and cons of the two methods. Besides, the structure and arrange arepresented in this chapter.The following chapter is the basis of the paper. That is, the tire forces are acquired withthe tire sensor and the corresponding carcass deflection. Firstly, the paper directly gets therelation of the carcass deflection and the tire forces with the FEA software Abaqus. Then thetheoretical relation is obtained with the TreadSim tire model.The third chapter introduces the brush model and estimates the maximum tire-roadfriction coefficient with the corresponding estimator. From the simulation, the estimation isgood when the sliding region is large, and bad when small. A sudden change would occur inaddition. And furthermore, the phenomena are explained theoretically. The forth chapter shows the brush model considering the carcass elasticity, and gets therelevant estimator. The simulation confirms a better estimation than the traditional brushmodel and avoids the sudden change.The fifth chapter considers the existence of the camber angle and the load change of thetire when the suspension presents different steric configuration. In other words, the tireforces are composited by the camber forces. A suspension-tire model is established withAdams/Car and the camber angles and load with different suspension configuration is alsoobtained. And then the chapter compares the estimation with the camber forces coupled anddecoupled. When the camber angle is large, the results of the latter are better with theestimator considering the carcass deflection.In view of the two models mentioned before are limited to steady state and theestimation is not as well as expectation, the sixth chapter explores the LuGre model. Inaddition to the lateral forces and aligning torque in pure side slip, an important conclusion isalso deduced theoretically. That is, the sliding is in presence for the whole contact areaexcept for the leading point, which could be potential to avoid the inaccuracy when thesliding region is relative small. It is possible to see the quantity describing the Stribeck effectas the estimator. Also the LuGre model considering the carcass deflection is established inthis chapter.The last chapter is the conclusions of the whole paper.