Robust Optimization Of Injection Mechanism And Process Parameters Based On The Kriging Metamodel

The thermal deformation of the shot-sleeve and punch will cause the change of the fit clearance of the injection mechanism,affect the movement of the mechanism and the flow of liquid metal,and then reduce the forming quality of the parts and the service life of the equipment.if the pouring temperature,shot-sleeve wall thickness,initial clearance and other structural and process parameters are reasonably optimized,this kind of problems can be effectively improved.However,the uncertainty of the parameters is widespread,which has a great influence on the change of the fit clearance and cause the fluctuation of the injection process.In this paper,considering the uncertainty of structure and process parameters,a multi-objective robust optimization design method of injection mechanism and process parameters is proposed and studied systematically.According to the thermal-force coupling model,the irradiation mechanism test device is developed,and the test is carried out on a 100 T press.Verify the accuracy of the finite element model by measuring the deformation and temperature data of different locations.At the same time,the influence of different process parameters on the filling process of the injection mechanism is analyzed based on the orthogonal test method,which involves the significance of a single process parameter and the cross-influence between different process parameters.In order to reduce the computational complexity of numerical simulation,a Bayes-Kriging approximation system based on sparse samples is proposed.In the effective range of process parameters,sparse sampling is carried out,and a multi-input and multi-output Kriging metamodel is established to accurately predict the performance target of the injection mechanism.Combined with Bayesian statistical inference,the Kriging model is evaluated quantitatively,and the influence of different threshold ? on the reliability of the metamodel is analyzed.The results show that the Kriging model under sparse samples is reliable enough to replace numerical simulation for optimization,simplify the iterative process and improve the optimization efficiency.The robust optimization model of injection mechanism is established.Based on NSGA-? algorithm and 6? criterion,the main parameters affecting the deformation and fit clearance of injection mechanism are robust optimized,and the optimal solution is obtained.Through model prediction,simulation and experiments,the robustness,reliability and sensitivity of the optimal solution under different strategies are compared.The results show that the robust optimization design considering the uncertainty of process parameters can effectively improve the stability of the injection mechanism.It shows that the robust optimization of the injection mechanism is effective when considering the uncertainty of process parameters.Taking the 2500 KN squeeze casting machine as the analysis object,the robust optimization design of the injection mechanism of the squeeze casting machine is carried out according to the method proposed in this paper.By setting different population sizes and generations,comparing Pareto optimal set and Pareto optimal solution under different combination values,and then selecting reasonable combination values to improve the accuracy of NSGA-? algorithm in solving the optimization results of injection mechanism.The comparison with the results of deterministic optimization shows that the robust optimization results can minimize the change of the gap between the punch and the shot-sleeve,and at the same time,under the influence of uncertainty,the clearance between the punch and the shot-sleeve remains relatively stable.