We critically review and discuss the main glass-sintering models: Frenkel, Mackenzie-Shuttleworth, Scherer and the recently developed Clusters model, and focus on the problem of sintering with concurrent crystallization. The Clusters model is tested under various practical conditions. Isothermal tests are carried out on a widely polydispersed alumino-borosilicate (ABS) glass having jagged particles, which is stable against devitrification, and on a soda-lime-silica (SLS) glass with a narrow spherical particle distribution, which crystallizes easily. The algorithm for non-isothermal processes is also tested with two distinct systems: the same ABS glass and a narrow-sized cordierite glass, which is devitrification-prone. In addition to physical parameters such as viscosity, surface tension, particle-size distribution, crystal growth rate and number of nucleation sites, microscopic-particle-packing data are introduced into the model and it is demonstrated that the evolution of both density and pore size distribution can be reasonably predicted. All the results are discussed taking into account the assumptions made in the derivations and other complicating factors, such as irregular particle shape, compositional shifts due to crystallization, temperature gradients and degassing during sintering. Finally, we discuss the physical and processing parameters that determine whether sintering will be favorable over crystallization. We demonstrate that the Clusters model and related algorithm provide a powerful simulation tool to design the isothermal or non-isothermal densification of devitrifying or stable glass compacts with any particle-size distribution, thus minimizing the number of time-consuming laboratory experiments.

Cet article rapporte les principales théories relatives au frittage du verre en relation avec la cristallisation. Différents modèles sont testés en regard de plusieurs paramètres physiques, comme la viscosité, la tension superficielle, la taille et l'empilement des particules, la croissance et le nombre de nucléi. On montre que l'évolution de la densité et de la porosité est prévisible.