Modeling of the Microchemistry for Diffusion of Selected Impurities in Uranium

Abstract

Unalloyed metallic uranium used in some work done at Y-12 contains small quantities of impurities, the three most significant of which are carbon, iron, and silicon. During metallurgical processing, as the metal cools from a molten condition towards room temperature, the metallic matrix solution becomes supersaturated in each of the impurities whose concentration exceeds the solubility limit. Many impurity atoms form compounds with uranium that precipitate out of the solution, thus creating and growing inclusions. The objective of the present work is to study the distribution of impurity atoms about some of the inclusions, with a view toward examining the effect of the interaction between inclusions on the impurity atom distribution. The method used is time-dependent mass diffusion from the supersaturated solution to the surfaces of the inclusions. Micrographs of metal samples suggest that the inclusions form in successive stages. After each inclusion forms, it begins to draw impurity atoms from its immediate vicinity, thus altering the amounts and distributions of impurity atoms available for formation and growth of later inclusions. In the present work, a one-dimensional spherical approximation was used to simulate inclusions and their regions of influence. A first set of calculations was run to simulate the distribution of impurity atoms about the largest inclusions. Then, a second set of calculations was run to see how the loss of impurity atoms to the largest inclusions might affect the distribution of impurity atoms around the next stage of inclusions. Plots are shown for the estimated distributions of impurity atoms in the region of influence about the inclusions for the three impurities studied. The authors believe that these distributions are qualitatively correct. However, there is enough uncertainty about precisely when inclusions nucleate and begin to grow that one should not put too much reliance on the quantitative results. This work does provide a framework and an advance toward a comprehensive model of uranium metal microchemical distributions.