The Resolved and Unresolved Issues and Phenomena in Numerical Modelling of Air Cavity Closure Using a Scale-Resolving Method

The current International Maritime Organization (IMO) Greenhouse Gas (GHG) strategy for the shipping industry aims to reach net-zero GHG emissions from international shipping close to 2050. One of the technologies that can help reduce the fuel consumption of ships is by reducing the frictional resistance. For a typical sea-going cargo vessel the frictional resistance is in the order of 70% of the total ship resistance. One of the suitable options to realise such resistance reduction is by means of air lubrication. Zverkhovskyi (2014) concludes that the frictional drag can be reduced by some 60% on a flat plate, leading to a gross resistance reduction for a full block ship of some 12-17% of the total resistance.
A numerical model simulating the flow around air-lubricated vessels would be a valuable tool for ship designers and operators to predict the required propulsion power and operational characteristics of these vessels during the design phase. Despite the many numerical models available for successfully predicting viscous flows around ships, these standard models are not sufficient to accurately solve the flow around air-lubricated ships. From a practical point of view, Reynolds-averaged Navier-Stokes (RANS) methods combined with a Volume of Fluid (VOF) method for modelling the two-phase flow at the required high Reynolds numbers are of interest.
Previous publications showed that RANS models are capable of correctly model the global air cavity profiles, however the air loss mechanisms at the cavity closure were not or only partially captured (Rotte et al. (2020)). Since some of the air loss mechanisms were hypothesised to be governed by waves formed by turbulence structures hitting the interface, Rotte et al. (2018) examined hybrid RANS-LES simulations for modelling external air cavity flows. It was found that these simulations showed a more realistic cavity interface and closure behaviour than RANS simulations. However, the cavity profile was very sensitive to the upstream velocity profile of the boundary layer. And the way the hybrid model switches between RANS and LES inside the air cavity also posed problems, known as the Grey Area problem. Mukha and Bensow (2020) conducted simulations of an internal air cavity using LES and VOF interface capturing, and showed that the flow in the cavity closure is highly unsteady and turbulent, with air-water mixing and periodic shedding of bubble swarms.
This current paper discusses the capabilities and limitations of Large Eddy Simulation (LES) models applied to external air cavity flows. It focuses on identifying various air loss mechanisms, and the physical conditions under which they occur.