Advanced Constitutive Modeling of Triaxiality-Dependent Phase Transformation in Metastable Austenitic Steels: VUMAT Implementation and Multi-Model Validation
摘要
Metastable austenitic stainless steels, such as 304L, are widely used for their transformation-induced plasticity (TRIP) effect; yet, the precise influence of stress triaxiality, though stress state conditions, on the kinetics of martensitic transformation remains a critical challenge for accurate structural modeling. This study investigates the mechanical response and phase evolution of 304L TRIP steel under varying constraint conditions. We performed uniaxial tensile tests and double-edge notched tensile (DENT) tests with varying notch radii to induce initial stress triaxiality levels ranging from 0.33 to 0.91. Martensite volume fractions at fracture were quantified using x-ray diffraction (XRD) refined by a custom Python processing workflow. A constitutive model, integrating a modified Iwamoto kinetic law with a triaxiality correction term, was implemented via a VUMAT subroutine in Abaqus/Explicit. Experimental results reveal that higher stress triaxiality significantly accelerates transformation kinetics, increasing the martensite fraction from 40% in normalized specimens to over 63% in highly constrained geometries. The numerical model successfully captured the complex strain hardening behavior and phase evolution across all stress states, yielding a root mean square error (RMSE) below 10%. Furthermore, a comparative analysis of exponential, hill, and logistic models identifies the logistic formulation as the most robust mathematical framework for predicting triaxiality-dependent martensite saturation. These findings provide a generalized predictive law for estimating microstructural evolution in TRIP steels under multiaxial loading, essential for the design of high-performance structural components.