Optimization Of Forming Process Parameters For Hydroforming Of Special-Shaped Parts

At present,the traditional plastic sheet metal processing is difficult to meet the stringent requirements in the industrial fields,such as automobile and aviation.In order to solve the problem of hard forming of part,hydroforming technology is used to form a part that made of aluminum alloy 2A12 and BP neural network and multi-objective genetic algorithm is used to optimize the forming process parameters of hydroforming.To measure the mechanical properties of material,the universal testing machine is used by the uniaxial tension test.The important process parameters of hydroforming(chamber pressure,drawing force and blank holder force),causes and preventive measures of forming defects are analyzed.Based on the structure of the part and material properties,forming process of the part is analyzed and the main process parameters of hydroforming are figured out.The finite element analysis software Dynaform is used to simulate the forming process,and the forming defects of the parts are analyzed in the postprocessing,the dangerous zone of wrinkling and the dangerous zone of fracture are determined.Three groups of forming process parameters are selected for numerical simulation,and the effect of forming process parameters on the quality of the forming parts is preliminarily analyzed according to the calculation results.Because the forming process of hydroforming has multi factors and multi level values and there is coupling effect among the factors,it is difficult to obtain the optimal process parameters by using the numerical simulation.And the forming process of hydroforming is highly nonlinear,it is difficult to optimize the process parameters by establishing a mathematical model.BP neural network is introduced to establish a network model between optimization objectives and optimization parameters.However,the establishment of neural network requires sample data for online learning,and the traditional comprehensive test scheme is not applicable to multi factor and multi-level optimization problems.Orthogonal test is used to select a representative test plan by orthogonal array,and the data of the test is used as the sample data of BP network to establish network model.According to the established network model,the multi-objective genetic algorithm is used to optimize the forming parameters,and a set of approximate Pareto solutions is obtained.A set of parameters is selected from the Pareto solutions for numerical simulation,and the optimization results are analyzed.According to the optimized process parameters,a set of die is designed and tested on a single acting hydraulic press.Finally,the qualified parts are made out.