Effect of Triggering Turbulence on Large-Eddy Simulation of Iced-Wing Aerodynamics

Analyzing iced-wing aerodynamics is important for aircraft safety. However, simulating iced configurations accurately and efficiently is challenging due to the complex geometries and flow topologies involved. To investigate these challenges, wall-modeled large-eddy simulation of a swept wing with leading-edge ice is performed using an unstructured Voronoi mesh paradigm and compared to experimental results. The study focuses on a swept wing featuring high-fidelity and smooth ice shapes at a Mach number of 0.18 and a Reynolds number per mean aerodynamic chord of [Formula: see text]. For the high-fidelity ice shape, good agreement with the experiment is obtained using a moderate level of mesh resolution. For the smooth ice shape, a much finer mesh is required to obtain similar accuracy. However, comparable accuracy is achieved at similar mesh resolutions for the high-fidelity ice and smooth ice when roughness is added to generate resolved turbulence near the leading edge. For the smooth ice, using an equivalent sand-grain roughness estimate to mimic experimental grit roughness gives a factor of approximately 30 reduction in computational cost for similar accuracy compared to simulating the smooth ice without roughness. Finally, the predicted aerodynamic degradation due to icing is quantified using simulations of the tripped wing without ice.