Inverse identification of Johnson–Cook flow stress parameters for Ti6Al4V

Determining material properties for machining simulations is challenging because direct measurement methods cannot reproduce the conditions of machining. Instead, an inverse parameter identification is used in this work to determine the material parameters for the Johnson-Cook model for Ti6Al4V (3.7165, Grade 5). A numerical simulation model using the smoothed particle hydrodynamics code mfree_iwf is used to recalculate an orthogonal cutting experiment. Due to GPU-acceleration the computational time is less than 5 min per simulation. Three different optimization algorithms (Simplex, Bayes, Differential Evolution) are used for the identification of the material parameters by minimizing the process force errors between experiment and simulation. Best results are obtained with the Differential Evolution algorithm. The sensitivity of material model parameters to the computed process force errors are shown and reveal for some of the material parameters adverse effects on these errors. Recomputations of experiments at different process conditions with the identified material parameters show good agreements in terms of process forces and the chip segmentation behaviour can be reproduced in high resolution simulations.