Improving the Tensile Shear Load of Al–Mg–Si Alloy FSLW Joint by BPNN–GA

• Published in: Transactions of the Indian Institute of Metals Vol. 74; no. 6; pp. 1521 - 1528
• Main Authors: Guo, Ruixiu, Hu, Wei, Song, Qi, Ji, Shude, Qi, Weiwei, Yu, Haisheng
• Format: Journal Article
• Language: English
• Published: New Delhi Springer India 01.06.2021; Springer Nature B.V
• Subjects: Aluminum base alloys; Artificial neural networks; Back propagation networks; Chemistry and Materials Science; Corrosion and Coatings; Crack propagation; Friction stir welding; Genetic algorithms; Lap welding; Load; Magnesium; Materials Science; Metallic Materials; Neural networks; Optimization; Original Article; Process parameters; Shear; Silicon; Tribology; Small pin penetration; Tensile shear load; Friction stir lap welding; Parameters optimization; Back propagation neural network; Genetic algorithm
• ISSN: 0972-2815; 0975-1645
• DOI: 10.1007/s12666-021-02240-7

In order to obtain the joint with a high tensile shear load, the small penetration of tool pin into the lower sheet was applied during friction stir lap welding. The tensile shear loads of the Al–Mg–Si aluminum alloy joint under different process parameter combinations were optimized by combining back propagation neural network and genetic algorithm. The result showed that the hook bent down and the height of cold lap was small. Under the external tension load, the crack propagated along the lap interface or extended downward after reaching the highest point along the cold lap. With the increase of heat input, the tensile fracture mode of joint was more easily obtained. The highest tensile shear load of the joint reached 12.45kN, which was increased by 6.9% than the maximum value before optimization.