ABSTRACT
Friction models based on rst principles are far from mature. Unlike, say, heat transfer and uid dynamics, tribology is not very amenable to mathematical modelling. Ludema (1996) noted that over the then preceding 40 years more than 300 equations for the prediction of friction had been published, ‘but even the best has very limited use’, the key reasons being:
1. Most authors select a very narrow range of variables and then assume that friction varies monotonically over these ranges.
