Otimização robusta do fresamento de topo do aço inoxidável duplex UNS S32205

Abstract

The study on the end milling of duplex stainless steel has been gaining prominence due to the challenges encountered in its low machinability. This occurs, therefore, these materials have low thermal conductivity, high tenacity and high rate of hardening. However, to ensure final quality in manufactured products, it is important that the milling process is well planned, aiming at lower mill wear during the process and at the same time presenting good productivity indicators. Thus, this work aims to perform a robust multi-objective optimization in the end milling process of the duplex stainless steel UNS S32205. Experiments were carried out following a central composite planning combining the control variables: cutting speed, feed per tooth, depth of cut and milled width and noise variables: flank wear (vb), fluid flow (Q) and tool balance (lt0). The response variables evaluated were the surface roughness and the material removal rate. The methodologies of response surface, robust parameter design, mean quadratic error and normal boundary intersection were applied. So, the effects of control and noise variables were analyzed and discussed, as well as their interactions on the response characteristics. Tool wear was the variable that most influenced the surface roughness. The material removal rate was influenced by all controllable process variables. The values obtained for Ra roughness ranged from 0.24 to 1.10 μm and the material removal rate ranged from 40.39 to 187.52 mm³/s. The Mean and variance optimization of each characteristic of interest was performed, as well as the mean quadratic error optimization. Thus, 21 Pareto-optimal solutions were obtained, contributing to the improvement of surface quality and productivity of the milling process. In the confirmation tests a Taguchi orthogonal arrangement (L9) were used where the optimal setups were obtained capable of mitigating the influence of noise variables, which corroborated the good adequacy of the proposed methodology.