Modélisation du processus d’usinage et prédiction des performances technologiques du matériau usine.

Abstract

Stainless steels are widely used in industries such as aerospace, biomedical and automotive due to their corrosion resistance and excellent mechanical properties. However, their machinability remains a major challenge, requiring advanced techniques to optimize the machining process. The use of traditional cutting fluids poses risks to health and the environment, hence the growing interest in alternative solutions such as minimum quantity lubrication (MQL). An even more promising approach is to use nanofluid-assisted MQL (NF-MQL) to improve machining performance. In this study, we used the response surface (RSM) and artificial neural networks (ANN) methodology to establish models for predicting system behavior. Single and multi-objective optimization methods, such as the Taguchi approach, the desirability function (DF), as well as the NSGA-II and TOPSIS algorithms, were used to determine optimal machining conditions. The experiment was carried out on machining AISI 316L stainless steel in dry conditions, MQL and NF-MQL, evaluating key parameters such as surface roughness (Ra), cutting temperature (Tc), flank wear (Vb) and cutting force (Fz). The results showed that optimizing cutting parameters had a significant impact on process performance, with reduced tool wear, lower energy consumption and improved quality of machined parts. The integration of nanofluids into MQL has proven to be particularly effective, providing better lubrication, reduced cutting temperature and less tool wear, while promoting sustainable and environmentally friendly machining.