Parameters optimization of laser brazing in crimping butt using Taguchi and BPNN-GA |
| |
| Affiliation: | 1. State Key Lab of Digital Manufacturing Equipment and Technology, School of Mechanical Science and Engineering, Huazhong University of Science and Technology, Wuhan, China;2. Mechanical Science and Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, USA;3. College of Materials Science and Engineering, Huazhong University of Science and Technology, Wuhan, China;1. School of Laser Technology and Photonics, Suranaree University of Technology, Thailand;2. Department of Urban Environment Systems, Faculty of Engineering, Chiba University, Japan;1. School of Mechanical Engineering, Jiangsu University, Xuefu Road, Zhenjiang 212013, China;2. School of Mechanical Engineering, Shanghai Jiao Tong University, Dongchuan Road, Minhang District, Shanghai 200240, China;1. State Key Laboratory of Pulp and Paper Engineering, South China University of Technology, Guangzhou, 510640, China;2. Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, China;1. Centro de Investigaciones en Optica A.C., Loma del Bosque 115, 37150 Leon Guanajuato, Mexico;2. Centro de Investigacion en Matematicas A.C., Jalisco S/N Col. Valenciana, 36240 Guanajuato, Guanajuato, Mexico;3. Facultad de Matematicas, Universidad Autonoma de Yucatan, Cordemex 172, 97110 Merida, Yucatan, Mexico |
| |
| Abstract: | The laser brazing (LB) is widely used in the automotive industry due to the advantages of high speed, small heat affected zone, high quality of welding seam, and low heat input. Welding parameters play a significant role in determining the bead geometry and hence quality of the weld joint. This paper addresses the optimization of the seam shape in LB process with welding crimping butt of 0.8 mm thickness using back propagation neural network (BPNN) and genetic algorithm (GA). A 3-factor, 5-level welding experiment is conducted by Taguchi L25 orthogonal array through the statistical design method. Then, the input parameters are considered here including welding speed, wire speed rate, and gap with 5 levels. The output results are efficient connection length of left side and right side, top width (WT) and bottom width (WB) of the weld bead. The experiment results are embed into the BPNN network to establish relationship between the input and output variables. The predicted results of the BPNN are fed to GA algorithm that optimizes the process parameters subjected to the objectives. Then, the effects of welding speed (WS), wire feed rate (WF), and gap (GAP) on the sum values of bead geometry is discussed. Eventually, the confirmation experiments are carried out to demonstrate the optimal values were effective and reliable. On the whole, the proposed hybrid method, BPNN-GA, can be used to guide the actual work and improve the efficiency and stability of LB process. |
| |
| Keywords: | Laser brazing Bead geometry Crimping butt BPNN-GA Taguchi method |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
