When UO 2 and (U,Pu)O 2 fuels locally reach high burn-up, a major change in the microstructure takes place. The initial grains are replaced by thousands of much smaller grains, fission gases form micrometric bubbles and metallic fission products form precipitates. This occurs typically at the rim of the pellets and in heterogeneous MOX fuel Pu rich agglomerates. The high burn-up at the rim of the pellets is due to a high capture of epithermal neutrons by 238 U leading locally to a higher concentration of fissile Pu than in the rest of the pellet. In the heterogeneous MOX fuels, this rim effect is also active, but most of the high burn-up structure (HBS) formation is linked to the high local concentration of fissile Pu in the Pu agglomerates. This Pu distribution leads to sharp borders between HBS and non-HBS areas ( Fig.1).
Fig. 1: Sharp limit between HBS and non-HBS areas in a 55 GWd/t HM MOX MIMAS fuel (from [1]).In these MOX fuels, the HBS forming at various radial positions, and not only at the rim, it was shown that the size of the new grains, of the bubbles and of the precipitates increase with the irradiation local temperatures. Other parameters have been shown to have an influence on the HBS initiation threshold, such as the irradiation density rate, the fuel composition with an effect of the Pu presence, but also of the Gd concentration in poisoned fuels, some of the studied additives, like Cr, and, maybe some of the impurities. However, not all the differences in the UO 2 HBS rim extent measured by different teams on various fuels have been explained [2][3]. The effect of impurities may be the main reason for these differences, but it has not been documented enough yet. It has been shown recently, with examinations of a UO 2 fuel in which 235 U was heterogeneously distributed, that a high Pu concentration is not mandatory for HBS formation [4].Several changes in the fuel behavior occur concomitantly with the HBS formation. An increase of the fission gas release and an increase in the fuel swelling rate are measured [3]. It was shown by indirect and direct approaches that HBS formation was not the main contributor to the increase of fission gas release at high burn-up [1,[5][6]. Indeed, studying the Kr/Xe ratio or some isotopic ratio in the gases collected during rod puncturing and using the differences in the gas productions as a function of the radial position, it was shown that the HBS areas were not the main source of the released gases. SIMS measurements of the local retention confirmed this trend. Nonetheless, the formation of a strong bonding between the fuel pellet periphery and the inner surface of the cladding, with the formation of the inner zirconia layer induces tensile stresses in the fuel during power