Friction material development using artificial inteligence

Abstract

Friction material characteristics defined by synergy of brake pad constituent's properties can be significantly influenced by different manufacturing and testing parameters. Up to now, development of the new friction materials has been mostly done empirically because very little is known about very complex and highly nonlinear phenomena that are involved in the contact during braking. Top-down approach to the friction material development demands new sophisticated technologies that are able to model complex input ( friction material formulation, manufacturing, and testing) and output (changes of the friction coefficient and wear) relationship. The artificial intelligence can play a significant role in modelling the tribological systems for which physical descriptions are currently inaccurate or even unavailable. That is why in this paper will be analyzed possibility for friction material developing by means of artificial neural networks.