| With the development of artificial intelligence and industrial Internet,traditional production lines are undergoing intelligent upgrading.Intelligent production lines need to incorporate more automatic and intelligent control methods such as recognition,process decision-making and quality feedback.In sheet stamping production line on-line recognition is the precondition of the subsequent production process planning and quality control,thus get rid of some laboratory testing methods,under the given performance recognition method based on online data of press line automation intelligent upgrade play a good role in promoting,therefore,this article is based on bending and bulging deformation of two kinds of inverse recognition of material parameters of the research is of great significance.In this paper,two kinds of prediction methods are proposed by using bending and bulging deformation to study the self-recognition and fusion recognition of material performance parameters.Two recognition methods are proposed:(1)correlation method:building simulation material,fitting relevant prediction equation,and reverse predicting the real material performance;(2)Optimization method: deduce analytical model,obtain experimental data,and reverse calibrate performance parameters with the minimum residual square sum of experimental data and analytical data as the optimization objective.The main contents of this paper are as follows:(1)A bending and bulging test device with adjustable die size was designed to extract the bending data of real sheet metal on-line and identify the elastic-plastic properties of materials based on the data.The influence of die size on forming curve is analyzed,and the relationship between yield load and yield strength is constructed.It is determined that the window vector method proposed in this paper is better for identifying yield load.An analytical bending model was established,and the least square sum of residual error between predicted and experimental data was taken as the optimization objective.The elastic-plastic parameters of materials were identified efficiently by optimization method.At the same time,under the assumption of nonlinear unloading elastic modulus,the recognition method of unloading performance parameters of plate bending is determined,and the unloading displacement load curve of different materials under bending cyclic loading is analyzed.(2)The bulging deformation is used to identify the hardening parameters and tensile strength of the material.The influence of the ratio of punch to die size and strain energy on the parameter β was obtained,and the correlation prediction model of tensile strength including the ratio of strain energy to punch to die size was established,which further improved the recognition accuracy of tensile strength.The bulging analytical model is established.The optimization objective is the least square sum of residual error between bulging analytical prediction data and experimental data.The large strain hardening parameters of materials are identified by optimization method.The results show that the bulging optimization large strain hardening parameters have good convergence consistency.(3)Aiming at the problem that the universality of self-recognition material parameters in bending and bulging is weak,the performance strategies of sequential fusion identification and deep fusion recognition are proposed,and two groups of recognition parameters are obtained respectively.It is verified that the performance parameters of deep fusion recognition group have stronger universality in the forming process of workpiece under different stress forming states.Finally,aiming at the problem of fracture recognition after large deformation,the GTN model is used to describe the fracture process of materials,and the recognition method of fracture performance parameters based on genetic algorithm is constructed,which verifies the convergence and efficiency of genetic algorithm in optimizing fracture performance parameters. |
PDF Full Text Request