Studying grain growth and densification during sintering with a discrete element model
PAREDES GOYES B. 1, JAUFFRES D. 1, MISSIAEN J. 1, MARTIN C. 1
1 Univ. Grenoble Alpes, CNRS, Grenoble INP, SIMaP, Grenoble, France
The sintering of ceramic powders is a high temperature process occurring below the melting point. Consolidation, densification and grain growth compete during sintering. Grain growth kinetics is classically described by a power law in terms of the mean grain size. However, a local analysis of grain growth is richer and can help to understand the complex interplay between grains during coarsening. Here, for the first time, we use the Discrete Element Method (DEM) to propose a grain growth model during sintering at the grain scale. Unlike other methods, the straightforward DEM framework allows studying packings with a large number of particles while considering the main physical phenomena. As in the literature, grain boundary migration is implemented as a mechanism for grain growth. However, evidences suggest that the favorable thermodynamic conditions for migration are reached after inter-particle mass transport.
We choose surface diffusion for this mechanism since the interest here is in oxide ceramics. In DEM simulations, large particles grow at the expense of smaller ones. This leads to the disappearance of the smallest particles and grain growth. An accurate DEM densification and consolidation model for particles of different size based on grain boundary and surface diffusion is proposed and detailed [1]. First, the DEM model for a system of two particles is validated against phase field and surface evolution models, which have a more detailed description of the physics of sintering. The simulations of the sintering of large packings are compared with alumina experimental data [2]. For initial and intermediate stages, using material parameters from the literature, the model correctly reproduce the experimental sintering trajectory. Furthermore, we take advantage of the ability of DEM to consider realistic initial size distributions. The comparison with the experimental size distribution evolution, which is a very severe criterion, shows good agreement with experimental data. Therefore, the model shows the ability of studying the microstructural evolution caused by densification and grain growth during conventional sintering. Simulations also show that for realistic particle size distributions, grain growth can affect the microstructure evolution early in the first stages of sintering.
The influence of the initial size distribution, of the grain-boundary mobility on grain growth kinetics is evaluated. Our simulations show that broader particle size distributions accelerate grain growth. Finally, a comparison with sintering trajectories of alumina experiments with different sizes and sintering temperatures is carried out. We confirm and analyze the proportionality of the relative density with the inverse of the square of the mean grain size. Finally, the conditions for triggering abnormal grain growth in the model are discussed.
References
[1] B. Paredes-Goyes, D. Jauffres, J.-M. Missiaen, and C. L. Martin, “Grain growth in sintering:
a discrete element model on large packings,” Acta Materialia, vol. 218, p. 117182, 2021.
[2] I. Nettleship, R. J. McAfee, and W. S. Slaughter, “Evolution of the grain size distribu-
tion during the sintering of alumina at 1350?c,” Journal of the American Ceramic Society,
vol. 85, no. 8, pp. 1954–1960, 2002.