Statistical Methods to Formulate Belt Drive CVT Fluids with High Friction Coefficients

In the development of next generation continuously variable transmission fluids (CVTF) where high metal-to-metal friction is required, statistical methods play an important role in finding and understanding chemical additive effects on friction coefficient. In this study, the iterative use of statistical experimental design is applied, first with a Plackett-Burman design to screen many component effects, and then with mixture experiments to further study selective component and interaction effects, so that optimal fluid performance can ultimately be achieved. Friction data generated from continuously variable transmission (CVT) push belt element, bench screen-tests is utilized to make inferences about formulation behavior.