Open-source thermo-mechanical modelling of WAAM 316 L stainless steel T-joints: Validation of hardness, distortion and residual stress

Wire Arc Additive Manufacturing (WAAM) of 316L stainless steel enables monolithic T-joint walls but also induces distortion, spatial hardness variations, and banded residual-stress (RS) fields. We combine the contour method, neutron diffraction, and microhardness mapping with full-coupled thermo-mechanical finite-element simulations in an open-source framework (MOOSE). A model-validation process is developed that leverages a 33-layer thick wall (contour method) and a 3-layer thin wall (neutron diffraction and hardness) to cross-validate full-field predictions across thickness scales. After calibration to embedded thermocouples, the model reproduces peak temperatures, plate bending, and wall curvature. Mechanistically, we identify substrate restraint, cumulative plastic shakedown at mid-height, and top-surface relaxation as the drivers of the canonical tensile-compressive-tensile (T-C-T) topology in tall WAAM walls. For the 3-layer wall, predicted <i>σ_xx, σ_yy, and σ_zz</i> match neutron data along multiple lines within 20-50MPa. A strain-based hardness surrogate recovers the measured trends between base plate, wall core, edges, and the top layers. In the 33-layer wall, the simulations capture the T-C-T shape; sensitivity analyses indicate that incorporating kinematic/combined hardening is necessary to deepen the mid-wall compression in line with measurements, while isotropic<i> J₂</i> plasticity already provides robust trend fidelity and cross-code consistency. Strong-scaling tests to 120 cores show that the open-source workflow is computationally efficient for parametric studies. The validated dual-thickness benchmarks and open-source pipeline provide a transferable basis for microstructure-aware constitutive upgrades that link WAAM process parameters to residual stresses, distortion, and hardness in T-joint geometries.