Numerical Simulation For The Process Of Low Pressure Casting And The Application And Deplore Of ANN-Optimization Design System

Filling and solidification process of low pressure casting is a complex and dynamic process with huge heat dissipation. Whether the design of gating system and process parameters is reasonable or not, they decide the quality of a casting in a great extent. In order to reduce production costs and increase products quality, it is quietly desirable to utilize CAE numerical simulation technology to analyze and optimize parameters design in design stage. Though the application of numerical simulation technology has made active effect, there exist many problems in practical application, and these problems restrict the development and application of CAE technology. For examples: there is the lack of alloy thermal physical parameters; numerical simulation software by means of finite different method usually distorts casting model; CAE data results can not predict or/and diagnose the effect of casting technology automatically and intelligently.Based on above considerations, in this paper, thermal physical parameters of casting alloy were tested firstly to assure the acquirement of reliable simulation results. The fluid flow and heat transfer of casting were investigated by the way of numerical simulation with the finite element analysis software ProCAST for the purpose of improvingengineering application and the technological design of casting. After the description of two kinds of design projects based on CAE simulation results, ANN-Optimization design was used to optimize CAE simulation results of casting process. And then the technological design optimization model of shell casting by low pressure die casting process was constructed. Depending on software Matlab for the developing platform, ANN-Optimization design system was schemed and developed by means of Object Oriented Design(OOD) method combined with the knowledge of foundry. The system can not only realize prediction, diagnosis and optimization of technological parameters of shell casting by the training of CAE results data, and also train and simulate general sample data. The outstanding characteristics of specialization, practicability, and customization of the system make it convenient to use. The research shows that CAE technology can realize the optimization of technological design of casting and has active instructional significance for actual engineering application. The design method of modern artificial intelligent introduced to the field of casting design makes the best use of CAE results data, enhances the efficiency of numerical simulation, and decreases the operating requirement of tools. Besides, the combination of traditional orthogonal and neural network design method can grasp the inner rule of various factors with the least experimentations, obtain accurate result of prediction and optimization, and have comprehensive practical application value.