An inverse neural network model of disc brake performance at elevated temperatures

The demands imposed on a braking system, under wide range of operating conditions, are high and manifold. Improvement and control of automotive braking systems’ performance, under different operating conditions, is complicated by the fact that braking process has stochastic nature. The stochastic nature of braking process is determined by braking phenomena induced in the contact of friction pair (brake disc and disc pad) during braking. Consequently, the overall braking system’s performance has been also affected especially at high brake interface temperatures. Temperature sensitivity of motor vehicles brakes has always been an important aspect of their smooth and reliable functioning. It is particularly related to front brakes that absorb a major amount (up to 80%) of the vehicle total kinetic energy. The friction heat generated during braking application easily raises temperature at the friction interface beyond the glass transition temperature of the binder resin and often rises above decomposition temperature. The gas evolution at the braking interfaces because of pyrolysis and thermal degradation of the material results in the friction force decreasing. At such high temperatures, friction force suffers from a loss of effectiveness. This loss of effectiveness (brake fading) cannot be easily predicted due to subsequent thermo-mechanical deformation of disc and disc pad (friction material) which modifies the contact profile and pressure distribution, altering the frictional heat. The instability of the brake’s performance after a certain number of brake applications is common and depends on braking regimes represented by application pressure, initial speed, and brake interface temperature. Therefore, the most important issue is related to investigation of possibilities for control of brake performance, especially at elevated temperatures, in order to be stabilized and kept on some level. The control of motor vehicle brakes performance needs a model of how braking regimes, before all application pressure, affecting their performance for the specific friction pair characteristics. Analytical models of brakes performance are difficult, even impossible to be obtained due to complex and highly nonlinear phenomena involved during braking. That is why, in this chapter artificial neural network abilities have been used for modelling of the disc brake performance (braking torque) against synergy of influences of application pressure, initial speed, and brake interface temperature. Based on that, an inverse model of the disc brake performance has been developed able to predict the value of brake’s application pressure, which, for current values of brake interface temperature and initial speed, provides wanted braking torque. Consequently, the brake’s application pressure could be adjusted to keep the disc brake performance (braking torque) on some wanted level and prevent its decreasing during braking at elevated temperatures