Discrete-element modeling of viscous flow sintering based on machine-learning-aided 3D microstructure reconstruction

This study aims to develop a numerical model devoted to the simulation of the early stages of viscous flow sintering of glassy materials using the discrete element method (DEM). An amorphous and spherical silica powder is selected as model glassy material. The green compacts were shaped using cold uniaxial pressing. A key innovation is the creation of an accurate digital twin of the microstructure, uniquely reconstructed in 3D from real FIB/SEM images using a machine learning algorithm for particle recognition. New DEM contact laws, derived from finite element analysis, incorporate attractive forces, viscosity, and neck growth. The DEM simulation's macroscopic shrinkage trajectory deviates by less than 1% from experimental dilatometry data during the isothermal dwell, validating the model's ability to describe the initial and intermediate sintering stages. • Granular skeleton generation using FIB/SEM. • Skeleton digitalization using machine learning algorithm. • DEM sintering simulation on amorphous silica glass.