An Alternative Approach to Modeling the Grain Structure of Castings

摘要

During the past 25 years there has developed increasing interest in the possibility of being able to predict the overall solidification process in metal castings. This has obvious applications in the automation of foundry practice, improved quality control, and for innovative design. To model the entire process, numerically, involves the two major steps of: (a) calculating the macroscopic heat flow and local temperature changes within the casting, and (b) matching this with a description of the pattern of microstructural development and latent heat evolution; the former has become routine but the latter remains less certain and open to discussion. The success of such models can be assessed by how accurately they predict the measured cooling curves and how closely they predict the corresponding microstructures, particularly the measured grain size. A majority of models have been applied to the near isothermal solidification of grey cast irons which develop a radial, duplex cell or grain structure; these involve semi-empirical models for heterogeneous nucleation on unspecified substrates of variable potency and population density. Considerable success has been achieved in fitting these modeling results to thermal analyses or to measured cell/grain counts, but not necessarily to both at once. In the absence of specific heterogeneous nucleating agents, such as may be added deliberately as grain refiners, it has long been recognized that during casting, and perhaps with initial columnar grain growth, the bulk liquid typically contains a remarkably high population density of dendritic or other fragments of the base material, variously introduced. The presence of such crystal fragments may readily be observed in transparent analogue models based on aqueous or organic systems, i.e., by experimental rather than numerical modeling. It is suggested that an alternative, more physically realistic modeling route for equiaxed grain formation may be developed in terms of the production, distribution, survival and growth of these inherent crystal fragments. Some success has been achieved in predicting the equiaxed grain size of continuous steel castings, using such an alternative model, and the outline of an experimental and numerical modeling program is suggested. Some of the points made in the text will be reinforced during presentation by the use of a video recording of fluid flow and crystal transport in aqueous analogue castings.