ABSTRACT Despite numerous contributions made to the modelling of precipitation kinetics, no theory seems to have unified all the concepts required to simulate the nucleation, growth, and ripening processes as one continuous process. In the present work, a novel approach was used to reshape the master equation calculating the evolution of the size distribution, so that the zero-flux condition can represent a steady-state evolution. This modification was made by introducing explicitly the probability for clusters to grow or to dissolve in each size class. The expression found to calculate this probability allowed the calculation of the entire evolution of the size distribution from the first step of nucleation to the final stages of the ripening process. The winners-losers concept also allowed for a better definition of the average condensation and evaporation rates, which were found by applying the thermodynamic extremum principle and assuming that a population of clusters of a given size behaves on the average as one cluster of the same size and submitted to the same driving forces. Furthermore, the thermodynamic extremum principle allowed us to show that the subcritical growth regime must be interface controlled to maximise the dissipation rate of energy. The computational scheme based on the new theory was applied to calculate the size evolution of precipitates in the Al-Cu system. The influence of aging temperature on the evolution of the number density of precipitates during homogeneous nucleation was found to agree with experimental estimations of number densities made on uniformly distributed θ′-Al2Cu precipitates.
Read full abstract