Abstract
To employ liquefied hydrogen (LH2) for large-scale storage and transportation of low-carbon energy, boil-off prediction in cryogenic tanks is crucial. Boil-off is driven by heat ingress from the ambient, which also leads to pressurization and thermal stratification in the ullage. In this work, a sharp-interface model is implemented in OpenFOAM and used to predict pressurization, stratification and boil-off rates at three different fill levels of an LH2 tank. A buoyancy source term is also implemented and added to existing RANS models. This source term is shown to completely suppress turbulence in the ullage. We have performed a series of simulations, predicting pressurization rates and thermal stratification. The results were compared to the K-Site experiments and for the first time, reasonable agreement was obtained across all fill levels. The boil-off rate showed qualitative agreement with the experimental estimates but was underpredicted in magnitude. The effects of heat-flux distribution, thermal contact resistance and variable wall thickness are also assessed, and their impact was found to be significant. This motivates the need for improved thermal boundary conditions, as well as possible coupling with resolved thermal models of the tank structure and insulation.