Characterization of intergranular fission gas bubbles in U-Mo fuel.

Abstract

This report can be divided into two parts: the first part, which is composed of sections 1, 2, and 3, is devoted to report the analyses of fission gas bubbles; the second part, which is in section 4, is allocated to describe the mechanistic model development. Swelling data of irradiated U-Mo alloy typically show that the kinetics of fission gas bubbles is composed of two different rates: lower initially and higher later. The transition corresponds to a burnup of {approx}0 at% U-235 (LEU) or a fission density of {approx}3 x 10{sup 21} fissions/cm{sup 3}. Scanning electron microscopy (SEM) shows that gas bubbles appear only on the grain boundaries in the pretransition regime. At intermediate burnup where the transition begins, gas bubbles are observed to spread into the intragranular regions. At high burnup, they are uniformly distributed throughout fuel. In highly irradiated U-Mo alloy fuel large-scale gas bubbles form on some fuel particle peripheries. In some cases, these bubbles appear to be interconnected and occupy the interface region between fuel and the aluminum matrix for dispersion fuel, and fuel and cladding for monolithic fuel, respectively. This is a potential performance limit for U-Mo alloy fuel. Microscopic characterization of the evolution of fission gas bubbles is necessary to understand the underlying phenomena of the macroscopic behavior of fission gas swelling that can lead to a counter measure to potential performance limit. The microscopic characterization data, particularly in the pre-transition regime, can also be used in developing a mechanistic model that predicts fission gas bubble behavior as a function of burnup and helps identify critical physical properties for the future tests. Analyses of grain and grain boundary morphology were performed. Optical micrographs and scanning electron micrographs of irradiated fuel from RERTR-1, 2, 3 and 5 tests were used. Micrographic comparisons between as-fabricated and as-irradiated fuel revealed that the site of first bubble appearance is the grain boundary. Analysis using a simple diffusion model showed that, although the difference in the Mo-content between the grain boundary and grain interior region decreased with burnup, a complete convergence in the Mo-content was not reached at the end of the test for all RERTR tests. A total of 13 plates from RERTR-1, 2, 3 and 5 tests with different as-fabrication conditions and irradiation conditions were included for gas bubble analyses. Among them, two plates contained powders {gamma}-annealed at {approx}800 C for {approx}100 hours. Most of the plates were fabricated with as-atomized powders except for two as-machined powder plates. The Mo contents were 6, 7 and 10wt%. The irradiation temperature was in the range 70-190 C and the fission rate was in the range 2.4 x 10{sup 14} - 7 x 10{sup 14} f/cm{sup 3}-s. Bubble size for both of the {gamma}-annealed powder plates is smaller than the as-atomized powder plates. The bubble size for the as-atomized powder plates increases as a function of burnup and the bubble growth rate shows signs of slowing at burnups higher than {approx}40 at% U-235 (LEU). The bubble-size distribution for all plates is a quasi-normal, with the average bubble size ranging 0.14-0.18 {micro}m. Although there are considerable errors, after an initial incubation period the average bubble size increases with fission density and shows saturation at high fission density. Bubble population (density) per unit grain boundary length was measured. The {gamma}-annealed powder plates have a higher bubble density per unit grain boundary length than the as-atomized powder plates. The measured bubble number densities per unit grain boundary length for as-atomized powder plates are approximately constant with respect to burnup. Bubble density per unit cross section area was calculated using the density per unit grain boundary length data. The grains were modeled as tetrakaidecahedrons. Direct measurements for some plates were also performed and compared with the calculated quantities. Bubble density per unit grain boundary surface area was calculated by using the density per unit grain boundary length data. These data were used as input for mechanistic modeling described in section 4. Volumetric bubble density was calculated by using density per unit grain boundary surface area. Based on these data, bubble volumetric fraction was calculated. Bubble volume fraction was also calculated by using the density per unit cross section area. Bubble volume fraction was also directly measured for some plates. These three results are comparable although the direct measurement data are slightly larger than the others. Bubble volume fraction increased as a function of burnup, reaching {approx}2% of fuel volume at 3 x 10{sup 21} f/cm{sup 3}. Fission gas bubble swelling is minor compared to that of solid fission product swelling.