A survey is made of integral experiments useful for testing thorium and /sup 233/U nuclear data in thermal reactor applications. Emphasis is on homogeneous /sup 233/U--H/sub 2/O criticals and simple, water-moderated /sup 233/U--thorium and /sup 235/U--thorium lattice experiments. Thorium--/sup 233/U-graphite experiments are also discussed briefly. Although the available experiments provide a fairly consistent test of important nuclear data, their accuracy and scope leave much to be desired. In detailed Monte Carlo analyses, ENDF/B-IV data are found to perform reasonably well. Adequate (though partly fortuitous) agreement is found with integral measurements of thorium resonance capture in lattices. A new, harder fission spectrum for /sup 233/U can correct the principal discrepancy observed with ENDF/B-IV, a bias trend in K/sub eff/ attributed to an underprediction of leakage.

A model was developed to predict in-pile fission gas swelling, gas release, and densification in oxide fuels. This model considers fission gas behavior at the grain interior, on the grain boundaries, and at grain boundary edges under conditions of total gas bubble destruction by fission fragments and partial gas bubble destruction. When gas bubble swelling on grain edges reaches 5 percent, it is assumed that gas tunnels form along the edges. Gas release takes place by migration of the gas in the grains and on the grain boundaries to the edge tunnels. Intergranular and intragranular densifications are considered. Densification takes place by vacancy boil-off due to thermal excitation and vacancy knockout by the passage of fission fragments through the pores. The migration rates of both vacancies and interstitials to pores are also calculated. Comparisons are made between the model and experimental data for swelling, gas release, and densification and found to be in reasonable agreement in most cases.