Numerical study of heat diffusion controlled bubble growth in A pressurized liquid

For water-cooled nuclear reactors an ability to predict various aspects of boiling is important for both safety and design. Practical component-scale modelling is necessarily done by semi- mechanistic, semi- Empirical approaches, generally based on heat flux partitioning. This needs to be informed by multiple detailed experimental observations of boiling behavior. Such measurements are difficult to make, especially at conditions of relevance to nuclear reactors, and now much effort is being applied to microscopic computational modelling of the detailed boiling process. Such analyses are difficult, with interface motion, heat transfer, phase change and mass transfer, alongside 'normal' fluid mechanics. They are also hard to validate. In this paper we attempt to predict one of the few cases for which there is an analytical solution of a reasonably representative model; the case of an isolated bubble growing in superheated fluid. We employ the interface-tracking code PSI-Boil (Y. Sato, B. Niceno, Journal of Computational Physics, Volume 249, 15 September 2013, Pages 127-161). Good results are obtained, but the work did highlight issues in studies of these microscopic phenomena, when the vapour flows associated with the mass transfers become the dominant flow. Spurious currents are generated associated with imperfect modelling of the high-curvature geometries. Eliminating these imposes particularly stringent requirements on aspects of the modelling. This problem is identified, and methods to eliminate the spurious currents under the relatively simple conditions of the present study are developed. These methods in turn provide pointers to a treatment in more general cases.