Parameter optimization for a laboratory friction tester to predict tire ABS braking distance using design of experiments

Abstract

Predicting tire friction in a laboratory environment is a complicated issue. Varying one factor at the time for experimenting to obtain insight into tire friction does not necessarily provide a proper judgment. It is needed to take all interactions between the operating factors into account to determine the optimized test conditions to predict tire friction. The research methodology of the Design of Experiments (DoE) is particularly suited for such a system influenced by multi-factors; a technique that constructs a predictive mathematical model to describe the relationship between responses and critical factors. With the aid of DoE, the influence of various operating parameters of the Laboratory Abrasion Tester (LAT100) as a tribometer on tire friction is investigated: load, speed, and slip angle. Two counter-surfaces are employed: a sandpaper- and a corundum-based substrate. A predictive mathematical model is extracted from the DoE and suggests various optimal test conditions to predict the tire friction. It provides new insights into the impact of factor interactions on the coefficient of friction of tire tread compounds. The variation of the coefficient of friction vs speed on the corundum-based substrate is in good agreement with the ABS braking distance data of actual tires on the road.