In this study a technique has been addressed in order to model and optimize AWJM process. The required data for modeling and optimization purposes has been achieved using orthogonal array Taguchi (OA-Taguchi), D-optimal techniques and their combination based on design of experiments (DOE) approach. Water pressure, abrasive flow rate, machining speed, and machining gap are the process variables considered in this study. To evaluate the process, surface roughness (SR) has been taken into account as the process characteristic. Regression modeling using which has been turned into a widely used method has been employed to establish a relationship between process input variables and output characteristic. Analysis of variance (ANOVA) has been employed to evaluate the adequacy of the proposed models among which the most fitted and proper ones selected as the authentic representative of the process and considered as the process objective function to be optimized. Next, to optimize the objective function in order to get the desired characteristic (SR), the proposed model has been embedded into simulated annealing (SA) algorithm. Based on the computational results (less than 4% error), the proposed procedure is quite effective in modeling and optimization of the process.
Read full abstract