Research On Optimization Of Structure And Injection Process Parameters Of Automotive Plastic Front-end Framework

Automobile fuel consumption accounts for a large part of global oil consumption,and it is urgent to reduce automobile fuel consumption in light of the energy crisis and environmental degradation.At present,the development of new-energy vehicle battery technology has encountered bottlenecks and it is difficult to improve the fuel efficiency of automobile engines.Therefore,the development of vehicle lightweight technology is the best way to save energy,protect the environment and reduce consumption.Plastic instead of steel technology is one of the important means of vehicle lightweight,Under the condition that the performance of the product is not changed,replacing the steel parts with plastic can greatly reduce the weight of the car.For injection molding parts,it is difficult to achieve the ideal situation by using traditional production experience and manufacturing technology,but with the rapid development of computer science and technology,CAE injection molding technology arises at the right moment.CAE injection molding technology can find hidden defects in the injection molding simulation process,and timely improve them,greatly saving production costs and shortening production time.This dissertation takes automobile front-end framework as the research object and Hypermesh software was used to design the structure of plastic front-end framework.The mechanical properties of the front-end framework were verified by finite element analysis with Hypermesh.The optimal runner systems of the front-end framework was obtained by using Moldflow,the uniform design test was designed and completed,and the linear regression equation were established to optimize test results.Genetic algorithm was used to optimize the BP neural network to predict is minimum combination of evaluation indexes corresponding process parameters.The main contents of this dissertation are as follows:(1)The corresponding eigenvalues were obtained by topological optimization of each subtarget with Hypermesh.The integrated response function was obtained from the unified programming of six sub-objectives by using weighted compromise programming.Taking the minimization of the comprehensive response function as the goal,the topology optimization of the automobile front-end framework was carried out.According to the actual production situation,the final structure of the automobile front-end framework was obtained.(2)Hypermesh software was used to extract and repair the middle surface of the front-end framework,mesh middle surface,and set the material and thickness of the mesh.The static analysis and modal analysis of the mesh model were carried out,and the fatigue analysis of the front-end framework was carried out by N-code to obtain the life and damage results.(3)The middle surface was divided into triangular meshes,and the influence of 5mm,7mm,9mm,11mm,13mm and 15mm meshes on the running speed and analysis results of Moldflow was analyzed.The runner system corresponding to gates 6,7 and 8 was designed,and the best runner system was obtained by evaluating the three runner systems with the comprehensive scoring method.(4)The uniform design test with 5 factors and 21 levels was designed and completed.The uniform design test result was obtained by simulating the uniform design test with Moldflow,and the linear regression analysis was carried out on the uniform design test table with SPSS,and the linear regression equation of the two evaluation indexes was obtained respectively.The second time uniform design test was conducted around the factors of the best evaluation index,and the regression equation was used to predict the parameters combination corresponding to the minimum evaluation index.(5)MATLAB was used to describe the evaluation index and process parameters between the three dimensional diagram,set up five parameters as the input,two evaluation indexes for the output of BP neural network model,and use genetic algorithm to optimize the weights and threshold of each node in the BP neural network model,and the optimization results were assigned BP neural network model,then the neural network model was used to predict the process parameters corresponding to the optimal evaluation index.Main innovations of this dissertation:(1)According to the results of single-objective optimization,entropy weight method was used to calculate the weight coefficient of lock limit collision load,buffer pad area load,pedestrian protection impact load and the low-order modals of the automobile plastic front-end framework.(2)A comprehensive scoring method is proposed to select the evaluation indexes that can reflect the advantages and disadvantages of the runner system.The runner systems were scored according to the performance of each runner system under different evaluation indexes.the better the performance,the higher the score.Finally,add up the scores of all evaluation indexes to obtain the total score,and the runner system with the highest total score was regarded as the optimal runner system.(3)The injection molding model of the plastic front-end framework of the automobile was designed and completed a comprehensive test,and the BP neural network was constructed based on the test results.The weights and thresholds of the nodes of the BP neural network model were optimized by genetic algorithm,and the optimized BP neural network model was used to predict the optimal evaluation index and corresponding forming process parameters.