Performance assessment of high-order large eddy simulation and immersed boundary method for rotorcraft hover

Large eddy simulations (LES) of rotorcraft hover are performed using a multiblock immersed boundary method (IBM). The compressible Navier-Stokes equations are solved on a fixed structured Cartesian grid using eigth-order energy-stable summation-by-parts (SBP) finite-difference schemes. The stretched vortex subgrid-scale turbulence model is used. Multiresolution is employed and both interface and characteristic-based boundary conditions are enforced using the simultaneous approximation term (SAT) approach. The SBP-SAT numerical scheme, LES model, and IBM approach are validated against bench- mark test cases. The Rotor Body Interaction (ROBIN) fuselage test case (1979) is also presented for validation purposes. The Knight and Hefner (1938) rotor hover case is used for both validation and to study rotor wake flow physics. Finally, a coaxial rotor hover case is studied with comparisons to measurements of Ramasamy (2013). Comparisons to similar results from the HPCMP CREATE™-AV Helios with NASA OVERFLOW and NASA FUN3D as near-body solvers using unsteady RANS are presented to address both accuracy and efficiency issues associated with the new LES-IBM approach. Predictions from the new code are in excellent agreement with measured data and Helios for the mean wall pressure coefficient on the surface of the ROBIN fuselage. Predictions for the 2- and 5-bladed Knight-Hefner rotor hover cases are in excellent agreement with the measured data and Helios for the figure of merit. Blade load predictions between the two codes are also in excellent agreement. Regarding the visualization of tip-vortices in the rotor wake, differences between Helios and the new LES code may be attributable to the fact that Helios uses URANS and only predicts phase-averaged flow fields. Phase-averaging of the LES data leads to better agreement of the wake structure between both solver.