Artificial neural networks in steel-mushy aluminum pressing bonding①
(1. Department of Mechanical Engineering, Tsinghua University, Beijing100084, P.R.China;
2. Department of Metal Forming, University of Science and Technology Beijing,
Beijing100083, P.R.China;
3. Department of Metal Forming, Northeastern University, Shenyang110006, P.R.China;
4. Anshan Automobile Fittings Factory, Anshan 114014, P.R.China)
2. Department of Metal Forming, University of Science and Technology Beijing,
Beijing100083, P.R.China;
3. Department of Metal Forming, Northeastern University, Shenyang110006, P.R.China;
4. Anshan Automobile Fittings Factory, Anshan 114014, P.R.China)
Abstract: Artificial neural networks were successfully used to research the modeling of aluminum solid fraction, preheat temperature of steel plate, preheat temperature of dies, free diffusing time before pressing and the interfacial shear strength in steel-mushy aluminum pressing bonding. Further more, the optimum bonding parameters for the largest interfacial shear strength were also optimized with a genetic algorithm.
Key words: artificial neural networks; steel-mushy aluminum bonding; genetic algorithm