Abstract
The trial-and-error method is widely used for the current optimization of the steel casting feeding system, which is highly random, subjective and thus inefficient. In the present work, both the theoretical and the experimental research on the modeling and optimization methods of the process are studied. An approximate alternative model is established based on the Back Propagation (BP) neural network and experimental design. The process parameters of the feeding system are taken as the input, the volumes of shrinkage cavities and porosities calculated by simulation are simultaneously taken as the output. Thus, a mathematical model is established by the BP neural network to combine the input variables with the output response. Then, this model is optimized by the nonlinear optimization function of the genetic algorithm. Finally, a feeding system optimization of a steel traveling wheel is conducted. No shrinkage cavities and porosities are induced through the optimization. Compared to the initial design scheme, the process yield is increased by 4.1% and the volume of the riser is decreased by 5.48×106 mm3.
Article PDF
Similar content being viewed by others
Avoid common mistakes on your manuscript.
References
China Foundry Association. 2013 China foundry industry statistical data officially released. Foundry Technology, 2014 (6): 11. (In Chinese)
Foundry Institution of Chinese Mechanical Engineering Society. Casting Manual. Beijing: China Machine Press, 2011. (In Chinese)
Liang Chen. Quality-Efficient-Green Foundry Industry Development - Interview of China Foundry Association of Vice Chairman Huang Tianyou. Aviation Manufacturing Technology, 2011, (z1): 68–70. (In Chinese)
Liu Baicheng. The trend of development of casting technology and computer simulation. Foundry Technology, 2005 (7): 611–617. (In Chinese)
Ransing R S and Sood M P. Optimization in castings-An overview of relevant computational technologies and future challenges. Metallurgical and Materials Transactions B, 2006, 37(6): 905–911.
Gao Shangshu and Yu Zhiya. Three equation method to calculate the riser. Foundry, 2000 (9): 533–537. (In Chinese)
Pan Feng. Research on the model of combination approximation and its application in lightweight design of car body. Ph.D. Thesis. Shanghai: Shanghai Jiao Tong University, 2011. (In Chinese)
Buscema M. Back propagation neural networks. Substance Use & Misuse, 1998, 33(2): 233–270.
Li Chenxi. Casting Process Design and Casting Defects Control. Beijing: Chemical Industry Press, 2009. (In Chinese)
Li Zhifeng. Mechanical optimization design. (Beijing: Higher Education Press, 2011: 267. (In Chinese)
Niyama E, Uchida T, Morikawa M, et al. A method of shrinkage prediction and its application to steel casting practice. International Journal of Cast Metals Research, 1982, 7(3): 52–63.
Marler R T and Arora J S. Survey of multi-objective optimization methods for engineering. Structural and Multidisciplinary Optimization, 2004, 26(6): 369–395.
Yang De. Experimental Design and Analysis (The first edition). (Beijing: China Agriculture Press, 2002: 508. (In Chinese)
Wang Xiaoping and Cao Liming. Genetic Algorithm: Theory, Application and Software Realization. (Xi'an: Xi'an Jiao Tong University Press, 2002. (In Chinese)
Author information
Authors and Affiliations
Corresponding author
Additional information
Dun-ming Liao Male, born in 1973, Professor, Ph. D., doctoral supervisor. His research mainly focuses on digital studies of the casting process. His academic research has led to the publication of more than 45 papers.
This research is financially supported by the Program for New Century Excellent Talents in University (Nos. NCET-13-0229, NCET-09-0396), the National Science & Technology Key Projects of Numerical Control (Nos. 2012ZX04010-031, 2012ZX0412-011) and the National High Technology Research and Development Program (“863” Program) of China (No. 2013031003).
Rights and permissions
About this article
Cite this article
Gong, Xd., Liao, Dm., Chen, T. et al. Optimization of steel casting feeding system based on BP neural network and genetic algorithm. China Foundry 13, 182–190 (2016). https://doi.org/10.1007/s41230-016-6008-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s41230-016-6008-8