Multi optimization of weld bead characteristics during GTAW of Inconel 825 using teaching learning based optimization

Welding of higher strength and higher corrosion resistant materials has always been a difficult task for the researchers. The higher strain hardening tendency and higher dynamic shear strength of the materials make it challenging for fabrication. Gas tungsten arc welding (GTAW) is a popular welding technique used for fabricating superalloys with high quality and low distortion. In this work, GTAW experiments were carried out on 2.5 mm thick Inconel 825 using an L-27 Box- Behnken experimental design. Welding current (I), welding speed (V), arc length (L) and gas flow rate (G) were considered as weld parameters to investigate welding characteristics viz., penetration (P) and width of the weld (W). Response surface modeling (RSM) was developed to predict the process responses. For both the responses, welding current is found as the most influential process parameter followed by welding speed. Increase in I increases P and W and decreases with an increase in V. Multi optimization using teaching learning based optimization (TLBO) algorithm results in optimum fitness value of 0.1930 at parameter setting of I = 87.682 A, V = 200.97 mm/min, G = 6.829 l/min, L = 2 mm. The corresponding optimum values of P and W are 2.202 mm and 4.039 mm respectively. The TLBO algorithm found effective and obtain the result with faster convergence (takes only 2-5 iterations) for obtaining the global optimum solution. (C) 2021 Elsevier Ltd. All rights reserved.