An Analysis of Four Different Approaches to Predict and Control Sintering

Abstract

Understanding and predicting sintering, which have been goals since the first attempts to mathematically describe the sintering process in the 1950s, are necessary to eliminate machining after sintering and to reliably predict and control the sintered microstructure and the resultant mechanical and other desired properties. In this study, four different sintering models are evaluated relative to one another and the experimental data, revealing their attributes, deficiencies, and modifications/improvements in order to facilitate their application, including the following: (i) a microstructure‐based model for solid state sintering, mainly developed by Riedel and Svoboda (RS); (ii) a viscous sintering (SOVS) model developed by Skorohod and advanced by Olevsky; (iii) a Kinetic Monte Carlo (KMC) model provided by Tikare; and (iv) the master sintering curve (MSC) approach introduced by Johnson et al. For different reasons, all four models have deficiencies that preclude achieving the most challenging goal of being able to comprehensively understand and predict sintering behavior: (i) the RS and the KMC models are complicated and difficult to use; (ii) the SOVS model cannot predict microstructure evolution; and (iii) the KMC model and the MSC have no stresses in their mathematical description, so they cannot simulate the effects of external forces. Each model also has attributes: (i) the KMC model allows one to follow the evolution of mesostructure; (ii) the MSC concept and the RS model are suitable for predicting densification curves for a wide variety of temperature–time profiles; and (iii) the SOVS and the RS models, which are implemented into finite element codes, can be used to predict density gradients and the warping of complex shape parts. Individually and together, the MSC, KMC, SOVS, and RS models can be useful tools to advance the fundamental understanding and improve the control of sintering.