Resumo:
The improvement of processes and products has increased substantially, due to the competition established between organizations, expanding quality control and the search for process optimization. Among these processes, in manufacturing, the turning of super duplex stainless steel UNS S32750 (SDSS) stands out, which has deserved special attention due to its excellent properties, mechanical and corrosion resistance, allowing its employability in diverse tasks, especially in the petrochemical industries and in processes with incidence of chlorides. Understanding controllable parameters is essential for improving quality and represents a gap in the literature and a research opportunity. Thus, in this work, it is proposed to combine known statistical tools to obtain optimization of multiple responses, in order to identify these parameters, seeking optimal initial conditions: o Design of Experiments, using a central composite design (CCD); principal component analysis (PCA) for dimensionality reduction; and the Desirability function, in determining a trade-off between the response variables. For processes of this type, it is common to evaluate the quality by the surface finish resulting from the machined parts. The study was conducted from an experimental design involving 3 control variables (xi): depth cut, feed rate and speed rate. As result, it was established that the optimal conditions for the process would be ap = 0.63 mm, f = 0.08 mm/rev and vc =178.5 m/min with calculated roughness equal to Ra = 0.235 μm, Ry = 1.084 μm and Rt =2.379μm. At the end, a confirmation experiment was carried out, in order to demonstrate the ability to obtain parts with acceptable levels of roughness, from the process parameters obtained by the optimization procedure. The results of the experimental confirmation rounds point to the good adequacy of the proposal.
