Enhanced sustainable turning of martensitic stainless steel incorporating vortex tube and minimum quantity lubrication for improved performance efficiency and environmentally friendly manufacturing processes

This study proposes a sustainable turning approach for martensitic stainless steel X20Cr13 using minimum quantity lubrication and a vortex tube to improve machining performance while reducing cutting-fluid consumption. A Box–Behnken design is employed to investigate the effects of machining parameters on surface roughness and tool life under optimized cooling and lubrication conditions. Experimental input data are obtained from a CNC tool-wear dataset. To accurately model machining behavior, a Continual Spatio-Temporal Graph Convolutional Network (CSTGCN) is developed to capture complex spatial and temporal relationships among machining parameters and adapt to changing cutting conditions. The proposed CSTGCN is used to predict surface roughness and is evaluated against existing methods using Python-based implementation. The results demonstrate that the proposed approach achieves 98.5% prediction accuracy with a 1.04% lower error loss, confirming its effectiveness for accurate and sustainable machining performance prediction.