A defect-responsive optimization method for the fiber laser butt welding of dissimilar materials. (15th January 2016)

• Record Type: Journal Article
• Authors: Ai, Yuewei; Jiang, Ping; Shao, Xinyu; Wang, Chunming; Li, Peigen; Mi, Gaoyang; Liu, Yang; Liu, Wei
• Is Part Of: Materials & design. Volume 90(2016)
• Page Start: 669
• Page End: 681
• Publication Date: 2016-01-15
• Subjects: LBW laser butt welding -- LP laser power -- WS welding speed -- FP focal position -- BPNN back propagation neural network -- PSO particle swarm optimization -- GA genetic algorithm -- S/N signal to noise -- ANOVA analysis of variance -- FZ fusion zone -- LD longitudinal defects -- BH bead height -- TD transverse defects -- WFL left front width -- WFR right front width -- WF front width -- WB back width -- T thickness -- HD height difference -- WD width difference -- HAZ heat-affected zone -- FAG fine equiaxed grains; Laser butt welding -- Welding defects reducing -- BPNN -- PSO -- GA; Materials -- Periodicals; Engineering design -- Periodicals; Matériaux -- Périodiques; Conception technique -- Périodiques; Electronic journals; 620.11
• DOI: 10.1016/j.matdes.2015.10.160
• Languages: English
• ISSNs: 0264-1275

Abstract

Abstract: Laser butt welding (LBW) of dissimilar materials has received great attention in automotive, power, chemical, nuclear and aerospace industries. The quality of welded joints is significantly affected by the generated defects in the dissimilar materials welding process. This paper proposes a defect-reducing optimization method that considers the geometric features of weld bead as evaluation indexes of welding defects and process parameters effect on the responses. The former aims to reduce welding defects for welding operations, and the latter seeks to identify the extent of contribution of actual process parameters on welding defects. The particle swarm optimization and back propagation neural network (PSO-BPNN), which has proved to be good modeling for no-linear problems, are utilized to establish the mathematical model and the defects reduction objective is combined with weld area. The genetic algorithm (GA) is adopted to solve the model. The effect of significant factors on the responses is identified based on the calculation of signal to noise (S/N) ratio and analysis of variance (ANOVA). The proposed method is evaluated by macro weld profile, microstructure and mechanical properties in confirmation tests. The results show that the proposed method is effective at reducing weld defects for dissimilar materials welding in practical production. Graphical abstract: Highlights: A defect-reducing optimization method for dissimilar laser welding is proposed. The welding defects generated in the welded joints are analyzed and evaluated. The optimal problem is modeled by process parameters and defect evaluation indexes. The results can help making decision on defect-free welded joints production.