A novel sharp interface immersed boundary framework for viscous flow simulations at arbitrary Mach number involving complex and moving boundaries

摘要

This work presents a robust and efficient sharp interface immersed boundary (IBM) framework, which is applicable for all-speed flow regimes and is capable of handling arbitrarily complex bodies (stationary or moving). The work deploys an in-house, parallel, multi-block structured finite volume flow solver, which employs a 3D unsteady Favre averaged Navier Stokes equations in a generalized curvilinear coordinate system; while we employ a combination of HCIB (Hybrid Cartesian Immersed boundary) method and GC(Ghost-cell) for solution reconstruction near immersed boundary interface. A significant difficulty for these sharp interface approaches is of handling sharp features/edges of complex geometries. In this study, we observe that apart from the need for robust node classification strategy and higher order boundary formulations, the direction in which the reconstruction procedures are performed plays an important role in handling sharp edges. Taking this into account we present a versatile interface tracking procedure based on ray tracing algorithm and a novel three step solution reconstruction procedure that computes pseudo-normals in the regions where the normal is not well-defined and reconstructs the flow field along those directions. We demonstrate that this procedure enables solver to efficiently handle and accurately represent sharp-edged regions. A fifth-order weighted essentially non-oscillatory (WENO) scheme is used for capturing shock-induced discontinuities and complex fluid-solid interactions with high resolution. The developed IBM framework is applied to a wide range of flow phenomena encompassing all-speed regimes (M=0.001 to M = 2.0). A total of seven benchmark cases (three stationary and four moving bodies) are presented involving various geometries (cylinder, airfoil, wedge) and the predictions are found to be in excellent agreement with the published results. (C) 2020 Elsevier Ltd. All rights reserved.