A GND-based back stress model for reverse loading in metal sheets with consideration of GNB

Abstract

Accurate prediction of springback and formability in sheet metal forming requires understanding reverse loading behavior under complex loading path changes, such as tension followed by compression. However, for ultra-thin sheets experimental characterization of such behavior is difficult due to compressive instability like plastic buckling. This study presents a crystal plasticity finite element method (CPFEM) incorporating a physically motivated back stress model based on geometrically necessary dislocations (GNDs) and boundaries (GNBs). The model captures grain size effects, including the Hall-Petch and Bauschinger effects, through a single grain size-dependent back stress parameter, enabling reverse loading prediction using only tensile data from specimens with different grain sizes. The back stress parameter was calibrated by fitting tensile stress-strain curves from two microstructures - one as-received and one annealed. Without using Tension-Compression (T-C) data for calibration, the model accurately predicted reverse loading behavior in low-carbon steel (0.64 mm thick) and Tension-Bending (T-B) responses in ultra-thin SUS316 (0.083 mm thick) when the developed theory was incorporated to an upscaled anisotropic hardening model. Identifiability analysis confirmed that the model parameters are uniquely determined by the available data. This physically interpretable framework provides an efficient and robust means to predict reverse loading in thin metal sheets, overcoming experimental limitations.