Optimization of the Warping Deformation of Mid-frame Plastic Components Based on BP-SSA Neural Network Mode
To mitigate the warping deformation of the mid-frame in the injection molding process of the wall-mounted fan,Moldflow software was used to simulate the mid-frame.An orthogonal experimental design was employed for six process parameters(mold temperature,melt temperature,filling time,solidification time,holding time,and holding pressure).Based on the experimental results,the back-propagation(BP)neural network model between process parameters and warping deformation was established.The model was optimized using the sparrow search algorithm(SSA)for global parameter optimization.The results indicate that when the mold temperature is 80℃,the melt temperature is 250℃,the holding pressure is 82 MPa,the holding time is 20 s,the injection time is 3 s,and the cooling time is 26.25 s,the warping deformation is minimized and its predicted value is 2.483 mm.Validation of the optimized process parameters using Moldflow software reveals a simulated warping deformation of 2.449 mm,resulting in a discrepancy of 1.3%between the predicted and actual values.The parameters obtained from the optimizations were verified by the test mold.The error between the result of the test mold warping and the result of the optimization algorithm is2.6%,which proves the accuracy of the optimization algorithm.The appearances of the injection parts have no defects such as flash shrinkage,and the assembly effects meet the expected requirements.The research findings suggest the feasibility of utilizing the BP neural network in conjunction with the sparrow search algorithm to optimize process parameters.
万方数据
维普
