Magnetic-assisted method and multi-objective optimization for improving the machining characteristics of WEDM in trim cutting magnetic material

摘要

In wire electrical discharge machining (WEDM), increasing the machining characteristics is always a research hot spot. Now, magnetic-assisted method cannot directly improve the performance of WEDM during machining magnetic material. In addition, due to the defects of convergence precision, convergence speed, global optimization, and subjectivity, the application of individual parameter optimization algorithm is limited. In order to address these problems, this paper proposes magnetic-assisted method and modified non-dominated neighbor immune algorithm to ameliorate the performance of WEDM in trim cutting magnetic material. Firstly, the force analysis of debris and the analysis of discharge current waveform are carried out to clarify what setting condition of the magnetic field generator is proper. It is indicated that the magnetic-assisted method maybe improves the machining characteristics of WEDM when the Lorentz force has the same direction with dielectric flashing force during trim cutting magnetic material. Additionally, a set of trim cutting Taguchi experiment is accomplished to investigate the effect of magnetic-assisted method on material remove rate (MRR), surface roughness (SR), and recast layer thickness (RU). Thereinto, it can be found that the machining characteristics of WEDM are obviously improved by magnetic-assisted method. More specifically, under magnetic-assisted method, MRR, SR, and RLT are improved by an average of 18.45, 17.86, and 17%, respectively. Then, on the basis of game theory, the modified nondominated neighbor immune algorithm (NNIA) is proposed to select best process parameters for further increasing the machining characteristics of magnetic-assisted WEDM (MA-WEDM). Eventually, the comparative results of three algorithms and verification experiment demonstrate that this modified NNIA exhibits better optimization performance than NSGA-II and NNIA. Specifically, by modified NNIA, the largest increase of MRR and the greatest decrease of SR and RLT are 22.37, 44.59, and 15.07%, respectively.