Hot-isostatic pressing of U-10Zr by grain boundary diffusion and creep cavitation. Part 2: Theory and data analysis

Abstract

Uranium-10 wt % zirconium (U-10Zr) is a fuel alloy that has been used in the Experimental Breeder Reactor-II (EBR-II). The high burnup that was desired in this fuel system made high demands on the mechanical compatibility between fuel and cladding both during normal operation and during safety-related transients when rapid differential expansion may cause high stresses. In general, this mechanical stress can be reduced by cladding deformation if the cladding is sufficiently ductile at high burnup, and/or by fuel hot-pressing. Fortunately, the fuel is very porous when it contacts the cladding, but this porosity gradually fills with solid fission products (primarily lanthanides) that may limit the fuel`s compressibility. If the porosity remains open, gaseous fission products are released and the porous fuel creeps rather than hot-presses under contact stresses. If the pores are closed by sintering or by solid fission products, the porous fuel will hot-isostatic press (HIP), as represented by the models to be discussed. HIP experiments performed at 700 C on U-10Zr samples with different impurity phase contents (Part 1) are analyzed in terms of several creep cavitation models. The coupled diffusion/creep cavitation model of Chen and Argon shows good quantitative agreement with measured HIP rates for hydride- and metal-derived U-10Zr materials, assuming that pores are uniformly distributed on grain boundaries and are of modal size, and that far-field strain rates are negligible. The analysis predicts, for the first time, an asymmetry between HIP and swelling at identical pressure-induced driving forces due to differences in grain boundary stresses. The differences in compressibility of hydride- and metal-derived U-10Zr can be partially explained by differences in pore size and spacing. The relevance of the experiments to description of in-reactor densification under external pressure or contact stress due to fuel/cladding mechanical interaction is discussed.