Status of models describing the mechanical performance of coated particle fuels

Abstract

In the evolution of coated particle stress models recent progress has been limited to the following achievements: 1. (1) Progression from the description of single-layer shells to complex multi-layer shells and the inclusion of kernel coating interactions and the fabrication phase (essentially a progression towards more complicated boundary conditions). 2. (2) Increase in physical accuracy by the inclusion of irradiation induced dimensional changes, creep (steady-state and transient) and kernel swelling in addition to elastic deformation and thermal expansion (essentially an extension of the constitutive equation). 3. (3) Improved mathematical methods employed to solve the basic differential equations, resulting in drastically reduced computing times. The basic physical assumptions underlying all models published so far have nevertheless remained the same and limit their area of applicability (agreement with experimental observations) to medium fast neutron dose levels and steady-state irradiation conditions. The most serious limitation is due to the assumption of mutual independence of the various physical processes which are combined in the constitutive equation. Irradiation induced dimensional change data which have been obtained for an unrestrained specimen are assumed to apply in exactly the same way to all finite zones within a PyC coating layer, while experimentally it has been shown that zones of originally identical PyC characteristics develop different microstructures (anisotropies) according to their individually different irradiation (stress-strain) histories. It is therefore proposed to include a feed-back loop containing a PyC model in the structure of the stress models, so that the material data for individual zones can be updated according to their irradiation histories. A considerable amount of materials data for PyC would have to be obtained to make such an approach quantitative, but already a very simple rheological PyC model is expected to eliminate the present restrictions and improve greatly the logical structure of the models. Further improvement is also required with respect to the failure criteria which are employed to predict fracture. In addition to the commonly used ‘limiting hoop stress’ and ‘limiting radial stress’ failure criteria, some experimental evidence for ‘limiting shear stress’ and ‘limiting creep strain’ failure criteria is available.