Evaluation of different flow stress models for machining simulations of medium carbon steels

A reliable simulation of cutting forces and chip formation that requires only a few experiments for model calibration is crucial for cost-effective machinability assessment, cutting data optimization, and tool design. To achieve this goal, selecting a suitable flow stress model to represent the behavior of the workpiece material under extreme deformation conditions is the first step to establishing realistic machining simulations. This study demonstrates the applicability of an inverse approach for the parameter identification of phenomenological flow stress models with different levels of complexity, including the combined effects of strain, strain-rate and thermal softening/hardening. The 2D finite element simulations affirmed the capability of the presented approach for identifying the flow stress parameters regardless of the complexity of the implemented model. It is also observed that models with improved thermal softening and strain/strain-rate hardening terms provide better predictive capabilities for estimating the cutting forces and chip thickness in 2D, indicating an improvement between 7-15% compared to the Johnson-Cook model. However, the 3D finite element simulation of the cutting process using the calibrated models showed significantly larger deviations from the experimental measurements. An effort was made to explain the reasons for this discrepancy.