A new multiobjective optimization with elliptical constraints approach for nonlinear models implemented in a stainless steel cladding process

This paper proposes a new multiobjective optimization with elliptical constraints approach for nonlinear models implemented in a cladding process of ABNT 1020 carbon steel plate using austenitic ABNT 316L stainless steel cored wire. Stainless steel stands out among the cladding materials as it allows obtaining surfaces with determined desirable characteristics from lower cost materials. This kind of process may be controlled by a relatively small number of input variables, i.e., the wire feed rate (WF), voltage (V), welding speed (WS), and the distance from the contact tip to the workpiece (N). Besides that, many outputs can be evaluated and simultaneously optimized. In the present paper, dilution (D), yield (Y), convexity index (CI), and penetration index (PI) were investigated. In order to consider the problem’s multivariate nature, techniques such as factor analysis and Bonferroni’s multivariate intervals were applied combined with elliptical constraints. The response variables were mathematically modeled using Poisson regression, and the obtained results were satisfactory since accurate models were achieved. The normal boundary intersection (NBI) method produced a set of viable configurations for the input variables that allows the experimenter to encounter the best system setup regarding the importance level of each response. Feasible and non-dominated solutions were found which means that the best possible scenario considering all the constraints was reached.