A New Optimization Technique in Examining the Machinability of Sverker 21 Steel: Gray Relational Analysis-Based Genetic Algorithm

Abstract

In this study, to determine the optimum parameters of the WEDM method, Taguchi experiment design, gray relational analysis (GRA) and gray relational analysis-based genetic algorithm (GRABGA) methods were used. In experimental studies, first, microstructure, microhardness and conductivity examinations of, both commercial and heat treated, Sverker 21 cold tool steel were performed. Then, the commercial and heat-treated Sverker 21 cold tool steel was processed with the WEDM method at the determined parameters. The effects of processing parameters on average surface roughness (Ra), kerf, and vibration were evaluated separately using Taguchi test design, GRA and GRABGA methods. It was determined that the homogeneous distribution of the carbides increased, the microhardness increased and the conductivity increased in the microstructure of the heat-treated sample compared to the commercial sample. In the processing of materials with WEDM, while the Taguchi test design method was used to evaluate the effects of processing parameters on the results were evaluated separately, GRA method was used to determine the grade of relationship between the results. In addition, the GRABGA method enabled the optimum values of the processing parameters to be determined not only on the basis of the level, but also on the intermediate values. It was found that the GRA relationship value was 0.8564 whereas the GRABGA relationship value was 0.8977. It was concluded that the GRABGA method provided better results than the GRA method in the correlation analysis of the results obtained in processing D2 cold tool steel with the WEMD method.