Analysis And Optimizition Of Process Parameters For Monocrystallime Silicon By WEDM

Monocrystalline silicon has good physical and mechanical properties, but because of the special structure of the material led to its high hardness, brittle. Making the process of Monocrystalline silicon difficult, at present, most monocrystalline silicon processing by wire sawing cutting. However, due to the existence of macro forces in the cutting process, making the surface of the slice surface warping and scratches can not be avoided. This paper studies the process of cutting monocrystalline silicon by wire electrical discharge machining (WEDM),aiming at improving the surface quality and cutting efficiency of monocrystalline silicon.On the basis of the analysis of WEDM monocrystalline silicon process. The voltage, pulse on time, pulse interval, wire saw speed as the main impact parameters on process objectives of WEDM monocrystalline silicon ,the single factor and multi-factor experiments have designed to explore the effect of process parameters on the process performance, then analyze the law of influence of process parameters on performance according to experiments date. Carried out single factor experiments,determine the range of individual parameters when the process objective is optimal .Carried out multi-factor experiments, the second regression model of the process performance and process parameters is established on the basis of RSM and the least square is used to fitted model equation. The variance of analysis of the model equation is carried out to determine the validity of the model equation of the process parameters and process objectives.The objective function of the multi - objective optimization model is established by the single - objective binary regression equation, A multi - objective optimization algorithm suitable for the objective function of this paper is selected. The non-dominated sorted genetic algorithm-? (NSGA-?) with elite retention strategy makes it possible to optimize many multi-objective models, so NSGA-? is introduced to optimize the multi-objective process model in this paper. The Pareto solution set about the cutting efficiency and the surface roughness is obtained by the evolution of 100 generations. The theoretical solution is compared with the experimental data to verify the feasibility of the theoretical solution.Strengthen Pareto Evolutionary Algorithm-II(SPEA-II) as its unique selection strategy and evolutionary strategy make its Pareto solution set more stable and uniformity than the solution set of other algorithms, So this algorithm is applied to multi-objective optimization of surface quality and cutting efficiency. The Pareto solution set obtained by SPEA-II are compared with NSGA-II, and they verifies mutually.