Automatic Optimal Riser Design System Of Steel Casting Using Genetic Algorithm

Riser design has taken a very important ratio in the design of the entire casting process. A well designed riser can help the casting avoid from shrinkage cavity and porosity, and improve the performance. In the casting numerical simulation software, however, the riser still needs to be added by means of the experience of people to judge and analyse, and then be finalized. This process uses the trial and error method, which requires large mount of manual operation, also needs the staff with high profession and experience, not only increases the labor cost, but also decreases the work efficiency, however, still cannot get the best design. Therefore, the traditional riser design methods must be combined with optimization algorithms. The most perfect result is the full realization of the riser's automatic generation. The combination of riser technics with the CAD/CAE, optimization method and artificial intelligence makes the development of riser design towards the automatic and intelligent direction. Some of the mainstream of casting simulation software (such as MAGMA by Germany, ProCAST by France, etc.) combines with numerical optimization techniques has developed the simulation system containing aided optimization function one after another.This paper is committed to developing a set of riser automatic optimization system which unites optimization algorithm with riser's traditional design process. For simple steel castings, it achieves the research object of directly getting the optimized riser used mathematical model, genetic algorithms and OpenGL technology based on MFC.In this paper, a mathematical model of riser has been established firstly. Through analyzing the technics of feeder design, and according the principles and the traditional experience requirements, the article concretely brings forward the mathematical model of riser which is also programmed.After that, with a in-depth study of the principles and content of genetic algorithms, according to the characteristics of genetic algorithms, genetic algorithm optimization process has been developed based on VC++ 6.0 platform by using object-oriented program technique language C++.Thirdly, by analyzing OpenGL technology, the characteristics of STL data structure and the triangle grid algorithm, the geometric modeling of three-dimensional entities of riser based on parameters has been realized. The program has been done.Finally, this paper combines the mathematical model of riser and genetic algorithm, implements the design of automatic optimal feeder system based on genetic algorithm. Where after, the article uses the system to optimize the casting of the ideal, gets the size of the optimal riser automatically, models and then carries through the analysis of solidification simulation to verify the effect of auto-optimization. The article gets the satisfactory proof.To some extent, this study provides a sound theoretical basis and practical proof for the future optimization of the automatic riser system. It lays a foundation of the further development of more complete and complex optimization for the casting process, which has important research and practical prospects.