Multi-objective optimization of the tooth surface in helical gears using design of experiment and the response surface method

The design of tooth surface for low gear noise under various load conditions is very important, and gear noise is highly related to transmission error. Optimal tooth surface for reduction of transmission error is very difficult to analytically determine due to nonlinearity of transmission error and the need to satisfy multiple load conditions. Satisfying design variables in multiple load conditions leads to the Pareto optimum of multi-objective optimization. There, the method to determine optimal lead curve and robust tooth surface design is proposed, using the response surface method and multi-objective optimization. The effect of transmission error on the candidate design variables by a screening experiment has been investigated using analysis of variance. Design variables are likewise selected. The fitted regression model of transmission error is built with the statistic validation of the representation. The model with constraints is solved to obtain optimum lead curve design and robust design for the tooth surface under multiple loads.