Multi-objective Optimization Of Plasma Spray Welding Process Parameters

Plasma spray welding technology is a very important surface repair technology in the remanufacturing industry.In the complex spray welding process,reasonable determination of process parameters is extremely critical to improve the quality of the spray welding layer and extend the life of remanufactured parts.Good quality spray welding layer usually has the characteristics of high micro hardness and strong wear resistance.Therefore,in this paper,the micro hardness,wear amount and dilution rate of the spray welding layer are used as response indicators to multi-objective optimize the plasma spray welding process.Using H13 steel as the base material and Ni60 B,TiC and TaC powder as the spray welding material,the preparation process of the spray welding layer is introduced in detail.Compared multiple test design methods,after determining the spray welding current,spray welding speed and powder feed flow as the optimization parameters,and the micro hardness,wear amount and dilution rate as the response targets,the two common multi-target optimization schemes are compared.First of all,the traditional multi-objective optimization scheme is used.The response surface method and the central composite method are combined and tested.After obtaining the test data,conduct a variance analysis on the data results through the Design-expert platform,establish a regression prediction model and verify the reliability,and obtain the influence rule of each optimization parameter and its interaction on the response index and the relative error of the model in the plasma spray welding process,and set the parameters in the optimization module according to the demand to obtain the final result.Then carry on the intelligent multi-objective optimization plan,choose the combination of neural network and genetic algorithm to optimize.Comparing the accuracy of the two common approximate models,the radial basis function(RBF)method was selected to build an approximate model of micro hardness,wear amount and dilution rate through the MATLAB platform.By comparing the advantages and disadvantages of the four common genetic algorithms,the NSGA-? multi-objective genetic algorithm is selected to performmulti-objective optimization on the established approximate model,and the approximate model-genetic algorithm multi-objective optimization method is proposed,finally,the pareto optimal solution set is obtained.Finally,the response surface method and the approximate model-genetic algorithm method are compared and tested to verify the feasibility of the scheme.It is concluded that using the combination of approximate model and genetic algorithm can significantly improve the quality of spray welding layer...