Modelling the effect of different mechanical restraints on PWR fuel rod deformation under conditions relevant to the NRU MT-3 experiment, using the MABEL-2D code

The MABEL-2D code has been used to study the effect of different mechanical restraint conditions on the ballooning response of fuel rods in the MT-3 experiment. Changing the restraint conditions from those considered appropriate to MT-3, where the rods pushed apart before flattening, to those appropriate to a large, uniformly ballooning array, was predicted to lead to increases in the axial extent of ballooning, in rupture times and in rupture strains, for a given pellet eccentricity within the cladding. At an eccentricity of 0.35, corresponding to a constant minimum to maximum pellet/clad gap ratio of about 1:2 the time to rupture was increased from 180 s to 185 s and the rupture strain from 43% to 45%. The increases in rupture strain and time were generally the same or smaller than differences, calculated in earlier work, resulting from the movement of rods relative to each other during ballooning. Pellet stack asymmetry within the cladding remains the dominant source of azimuthal temperature gradients. It is concluded that the absence of a model for lateral outmovement of neighbouring rods in earlier analyses of MT-3 would not, in the majority of cases, have significantly affected the prediction of rupture strains as a function of eccentricity. The predicted tendency of mechanical restraint to increase cladding strains agrees with the limited amount of experimental data. Good general agreement was found between azimuthal wall thickness distributions measured at the rupture locations and calculated using MABEL-2D with a constant pellet eccentricity, providing validation of the 2D strain model, and showing that the experimental distributions were consistent with a cosine azimuthal variation of temperature. Maximum strains were consistent with a local ductility rupture criterion. Paper to be presented to the IAEA Specialists Meeting on Water Reactor Fuel Element Performance Modelling, Windermere, Cumbria, UK, 9-13 April 1984. (author)