Computational evaluation of viscosity under microstructural uncertainties for the continuum model of sintering

The Continuum Theory of Sintering (CTS), which models the sintering body as a viscous medium, is widely used to simulate the behaviour of materials during sintering. However, its application in optimizing industrial sintering processes is often limited by the lack of key material parameters, particularly the viscosity of the sintering body. Most available viscosity models are also derived by considering idealized microstructures, limiting the predictive capability of the CTS. In this work, we present a computational framework for evaluating the viscosity of sintering bodies by explicitly incorporating their evolving microstructure during the sintering process. The approach integrates discrete element method (DEM) simulations with microstructural characterization, reconstruction, and homogenization. Simulated microstructure at a given density is characterized using multiple state variables, which are used to generate statistically equivalent microstructures. These are then used to define Boundary Value Problems (BVPs) for evaluating the effective properties and their associated uncertainties. Application of the method to the sintering of 316 L stainless steel powders demonstrates its ability to capture experimental data, while also enabling the assessment of existing viscosity models, particularly in relation to the evolving particle size. The results underscore the impact of pore morphologies, which can lead to variations in macroscopic viscosities by up to 145 % compared to existing models. The proposed framework links microstructural evolution of sintering bodies to their macroscopic properties, offering an efficient alternative to costly experimental methods. It also enables the routine application of the CTS for optimizing sintering cycles, thereby contributing to the development of advanced materials.