Intelligent digital twin process for the end-milling simulation combined with a multi-physical model

Abstract

The digital twin platform has particularly attracted attention in machining processes. Recent studies on digital twins have revealed a significant reliance on artificial intelligence, which lacks interpretability. Consequently, the integration of physics-based equations has become essential to enhance transparency and reliability. This study focuses on the digital twin platform during the subtractive manufacturing process, specifically predicting cutting force, machining temperature, and tool wear through a material removal simulation and analytical-physical methods. The semi-Oxley cutting force model was utilized, incorporating information about the tool geometry, machining conditions, shear angle, material properties, and material removal volume. The estimation result of the cutting force achieved a maximum and minimum accuracy of 91.5% and 77.2%. Moreover, end-milling experiments were conducted to evaluate tool wear, with Usui's tool wear model employed to predict tool life during simulation. Chip analysis was conducted to determine shear and friction angles, which were subsequently used to predict cutting forces. Cutting temperature was estimated using an energy consumption equation based on the predicted cutting force, achieving a minimum accuracy of 82.6%. A simulation of aerospace component machining was performed using the digital twin platform, presenting results for cutting forces and tool wear.