Applied Mathematics & Information Sciences
Comparison of Response Surface Methodology and Genetic Algorithm in Parameter Optimization of Laser Welding Process
This paper presents the comparative studies between Response Surface Methodology (RSM) and Genetic Algorithm (GA) in parameter optimization of laser welding process. Bead-on-plate weld was carried out on low carbon steel plate using diffusion cooled CO2 laser welding system. Weld bead geometry and heat affected zone were modelled and optimized as function of three laser welding process parameters namely laser power, welding speed and focal position using RSM based historic data design. The effects of the various laser welding parameters on weld bead geometry and Heat Affected Zone (HAZ) were studied using contour plots. The interaction effect of process parameters on the responses were studied using analysis of variance. Optimum welding parameters to obtain desire quality of joint were determined by numerical optimization module in RSM. Thereafter, priori approach of GA was also applied to determine the optimum welding parameter. The predictive ability of both the methodologies was compared. RSM yield better result than GA.
Digital Object Identifier (DOI)
Vijayan, K.; Ranjithkumar, P.; and Shanmugarajan, B.
"Comparison of Response Surface Methodology and Genetic Algorithm in Parameter Optimization of Laser Welding Process,"
Applied Mathematics & Information Sciences: Vol. 12:
1, Article 24.
Available at: https://digitalcommons.aaru.edu.jo/amis/vol12/iss1/24