Evaluation of a Subgrid-Scale Computational Fluid Dynamics Model for Ice Roughness

Abstract

In the present work, LEWICE-based ice-shape prediction results are presented when coupled to a computational fluid dynamics (CFD) model using a discrete-element roughness method (DERM) prediction of heat transfer. The DERM is a subgrid-scale model for CFD that accounts for the momentum and heat transfer aspects of large-scale roughness that displays an improvement of the heat transfer predictions beyond those of traditional sand-grain-roughness (SGR) models. The CFD-DERM approach is used to replace the built-in heat transfer prediction module of LEWICE for a multistep ice-shape prediction. Comparisons of ice-shape predictions and aerodynamics are made between the experiment, SGR-LEWICE, and DERM-LEWICE to evaluate the benefit of an improved heat transfer prediction methodology. The results indicate that the DERM model provides an improved prediction of heat transfer relevant to ice roughness. Additionally, ice-shape predictions in the glaze-icing regime are shown to be sensitive to the convective heat transfer prediction method. However, the improved heat transfer prediction does not necessarily correlate to an improved ice-shape prediction.