Improved reduced order modelling for time-dependent fission gas diffusion in columnar grains tailored for fuel performance applications

At the beginning of irradiation, fast reactors experience a radical alteration of fuel micro-structure driven by high temperatures and steep temperature gradients, leading to the formation of columnar grains. It is common in fuel performance codes to model fission gas diffusion in the columnar grain zone by approximating fuel grains as spherical, using a representative radius for the cylindrical geometry and assuming a uniform temperature, thereby neglecting the temperature gradient. However, the distinct diffusion kinetics induced by the temperature gradient render this classical approach unsuitable, resulting in inaccurate estimations of gas release. To account for the non-spheroidal shape and the effects of the temperature gradient in fission gas modules of conventional fuel performance codes, a previous work by the same authors proposed a physicsbased model to solve the gas diffusion problem in cylindrical grains using reduced order modelling techniques. Building on this prior framework, the present work introduces an advanced reduced order modelling approach that addresses specific limitations of the earlier model. Results highlight the accuracy of the proposed model in both transient and steady-state conditions, while maintaining a computational cost comparable to that of grain-scale fission gas behaviour modules, making it suitable for integration into fuel performance simulations.