Abstract
A simple model describing the splat cooling of liquids in a two-piston apparatus is discussed, and the approximations involved are elucidated. Approximate laws, giving the evolution of the temperature and the mean cooling rate during a splat cooling experiment, are deduced. These relationships, which depend on the nature of the pistons, the quenched alloy and the sample thickness, agree with experiment. It is further shown that the thickness of the quenched platelets is always larger than a `critical value' attained in an `ideal' process. In such a process the solidification of the melt is the stage which stops the movement of the pistons, thereby fixing the final thickness. This fact is used in the approximate calculation of the critical thickness. Satisfactory agreement with the experimental values is obtained. In the same manner, approximate laws for the evolution of the temperature, the mean cooling rate and the sample thickness obtained in a `ski jump' apparatus for ultrarapid quenching are established.