Evolutionary elements on composite ascent algorithm for multiple response surface optimisation

Currently, competitive markets place a great emphasis on various quality characteristics to customers. In the past, competition on the manufacturing of the bearing outer rings has primarily focused on the product hardness. Other quality characteristics of stain yield and roundness performance have recently become equally critical to fulfill customer satisfaction. To simultaneously meet all requirement criteria, the multiple response surface optimisation has focused on the use of desirability function approach to combine multiple response criteria into a single performance measure and to determine the optimal process design parameters of a pre-heat and heat chamber heaters, number of vacuum, gas fan speed in quenching chamber and partial gas pressure. This study proposed integrated algorithms with Taguchi orthogonal design and analysis and desirability function approach in forms of the path of composite ascent. There are three phases which use the conventional aspect of the steepest ascent via influential parameters, designed for moving toward the optiumum, use the ant colony optimisation (ACO) and hunting search (HuS) mechanism to present getting stuck at the local optimum. Results demonstrate that the integrated algorithms with the ACO and HuS elements are superior to the use of conventional composite ascent algorithm on average of 15.11 and 16.43 % improvement, respectively, for the overall responses. The confirmation tests on the actual manufacturing process based on the predicted operating conditions were verified and carried out to show statistically significant improvement at 95 % confidence interval.