A novel genetic algorithm-based calibration framework for crystal plasticity parameters in DP780 steels using multiscale mechanical testing

This study presents a robust genetic algorithm (GA)-based framework for calibrating crystal plasticity (CP) model parameters in dual-phase (DP) steels using multiscale mechanical testing. To overcome challenges associated with phase-specific parameter identification and temperature dependence, a two-stage calibration strategy is developed. In the first stage, nanoindentation tests at multiple temperatures are employed to determine the CP parameters of ferrite by matching simulated and experimental force-displacement curves. In the second stage, uniaxial tensile data are used to calibrate martensite parameters via representative volume elements (RVEs). The GA efficiently explores the high-dimensional parameter space and ensures fast convergence while maintaining physical consistency. Comparative results show that the GA-calibrated parameters outperform those obtained by conventional trial-and-error methods, with better alignment to experimental data. The proposed framework enables accurate and scalable CP calibration across different temperatures and offers broad applicability to multiscale modeling and alloy design.