Predicting the hydride rim by improving the solubility limits in the Hydride Nucleation-Growth-Dissolution (HNGD) model

•The HNGD model was found to be unable to predict the hydride rim thickness under temperature gradient.•Two hypotheses, targeting the supersolubility and solubility, were proposed to address this issue.•The modified HNGD model accurately predicts the thickness of the hydride rim in a sample submitted to a temperature gradient.•The modified HNGD model predicts the hydride distribution in the case of a Zircaloy cladding with an inner Zr liner.•The modified HNGD model has been implemented and verified and is now available for general use in Bison. [Display omitted] During operation of a light water reactor, waterside corrosion of the Zircaloy nuclear fuel cladding causes hydrogen pickup. The absorbed hydrogen can redistribute in the cladding driven by existing concentration, stress, and temperature gradients. When the concentration reaches the solubility limit, hydrides precipitate. These hydrides can be more brittle than the Zircaloy matrix, so they can endanger the cladding integrity during a transient if their concentration is too high. In recent years, extensive efforts have been made to understand hydrogen behavior and to develop simulation tools able to predict hydrogen diffusion and hydride precipitation and dissolution. These efforts led to the development of the Hydride Nucleation-Growth-Dissolution (HNGD) model and its implementation into the nuclear fuel performance code Bison. While it offers a significant improvement and accurately predicts the amount of precipitates, this model fails to predict the thickness of the hydride rim under a temperature gradient. The current work presents the limitation of the HNGD model and proposes two hypotheses to improve the model’s accuracy. The first hypothesis introduces a time dependency to the supersolubility to reduce the nucleation barrier as hydrogen atoms find more favorable nucleation sites. The second one introduces a hydride content dependency to the solubility. These hypotheses were validated and implemented into Bison and are now available to the user community. The modified HNGD model accurately predicts the hydride rim thickness, and it was demonstrated that this updated model can be used in Bison to model Zircaloy cladding with a zirconium inner liner. Finally, potential experimental and numerical methods are discussed to further validate these hypotheses.