A mesoscale simulation approach of microstructural evolution and anisotropic shrinkage during sintering of binder jet printed 17-4PH stainless steel

• Particle deletion of DEM packing successfully replicated BJT-printed anisotropy. • Modified KMC model effectively managed the non-isothermal sintering conditions. • DEM depicted anisotropic packing-modified KMC ensured precise sintering simulation. • DEM-KMC also captured the transition from anisotropic to isotropic shrinkage. • The feasibility of 3D DEM-KMC-FEM framework was proven. As an inevitable post-treatment of the binder jetting (BJT) process, sintering strongly influences the dimensional accuracy of the final components. The main challenge for precise dimension control lies in the anisotropic shrinkage during densification which is prompted by printing-induced heterogeneous particle configurations. To address this, we investigated anisotropic sintering behavior of BJT-printed 17-4PH stainless steel at the mesoscale by integrating simulated anisotropic powder configurations with a modified kinetic Monte Carlo (KMC) Potts model. Anisotropic representatives of printed parts were generated by rational particle deletion of isotropic discrete element method (DEM) packings. The obtained anisotropic packings were incorporated into a KMC Potts model modified with an Arrhenius-type temperature-dependent probability, enabling tracking of powder and pore evolutions during sintering. It was revealed that porous channels govern the anisotropic shrinkage and elevated temperature accelerates the transition from anisotropic to isotropic sintering. The predicted pore morphologies agree well with those under optical microscope, which validates the modelling strategy. Furthermore, as an initial feasibility test toward cross-scale sintering modeling, stereology-derived constitutive parameters like sintering stress and effective viscosity extracted from the KMC results were supplied to macroscale finite element method (FEM) modelling. The feasibility was proved by an accurate prediction of the relative density evolution