Thermal-shock experiments for separate-effects validation of UO2 fuel fracture models

Abstract

Because of the important role that fracture plays in the behavior of ceramic UO2 fuel in a nuclear reactor environment, fracture models are a major component of fuel performance codes. As with any aspect of fuel performance, it is crucial to validate these fracture models against experimental data; however, obtaining well-controlled data for conditions representative of a reactor environment is difficult. Quenching is proposed here as a relatively simple approach for using in a laboratory environment to achieve conditions that approximate those of a reactor environment. In this paper, an experimental apparatus containing a single instrumented fuel pellet in a sealed section of copper tubing is developed and applied to a series of seven experiments in which the apparatus is first heated to a high temperature (580–680 °C) by immersion in a molten salt bath, then quenched in a cold bath (-10–4 °C). Development of these experiments was guided by numerical simulations, and post-test simulations were performed to aid in understanding the experimental behavior and assessing the accuracy of the predictions of fracture initiation and propagation. In addition, similar experiments were performed on solid copper rods to provide temperature-dependent heat transfer coefficients for use in simulations of these experiments. This study demonstrates that quenching is a viable approach for generating thermal gradients representative of those in prototypical light-water reactor conditions at powers of about 5–10 kW/m. Fracture is expected to begin at these power levels, and moderate amounts of radial and axial cracking was observed in these quenching tests.