Optimizing brake cup formulation using mixture experiments and principal component regression model

Abstract

This study focuses on developing an optimal formulation to improve reliability of brake cup for motors. Designs mixtures experiments were performed based on quadratic models. Principal component regression was then utilized to resolve collinearity problem among designed ingredients to establish response surface model for each physical property. Multiple physical properties of brake cup were simultaneously considered to determine optimal formulation using desirability function to enhance reliability. A real case demonstrates effectiveness of successfully developed high reliability brake cup at low manufacturing cost.