Modeling and multi-objective optimization of the milling process for AISI 1060 steel

Abstract

In this work, an experimental study was carried out to investigate the influence of the cutting parameters namely cutting speed $({V}_{c})$, feed per tooth $({f}_{z})$, and depth of cut $({a}_{p})$ on three machining performance aspects, including cutting temperature $({Q}_{s})$, surface roughness $({R}_{a})$, and microhardness $(H)$ when milling of AISI 1060 steel. Response surface methodology $({\text{RSM}})$ was used for evaluating and predicting the impact of the considered cutting parameters on the selected machining characteristic indices. The results revealed that the error rates of the developed models compared to experimental ones were found to be as follows: 3.19% for ${Q}_{s}$, 5.32% for ${R}_{a}$, 1.63% for $H$, These error rates underscore the robustness and reliability of the developed models in accurately predicting the respective machining characteristics. Moreover, this study stands out in its approach by leveraging the experimentally developed RSM models as constraints within the optimization framework, providing a more precise and tailored approach compared to relying on generalized empirical models commonly found in industry handbooks. This is why a multi-objective optimization using genetic algorithm (GA) was performed to minimize both the production time and the production cost per unit by defining the problem with three key cutting parameters and utilizing the experimentally derived response surface methodology (RSM) models as constraints. For instance, the $({R}_{a})$ and $H$ -based RSM model served as constraints ensuring surface roughness and microhardness values below a specified threshold while satisfying other machining constraints (tool life, cutting force, cutting power). As a result, a set of optimal solutions of combinations of cutting parameters is achieved for simultaneously minimum production time and cost per unit.