Implementation and validation of infinite swept flow solvers in OpenFOAM

Analysing three-dimensional (3D) turbulent flows with the Reynolds-averaged Navier-Stokes (RANS) equations is a widely adopted routine in many industrial and research activities. However, in many practical cases computational campaigns based on 3D simulations may yield prohibitive computational costs. For instance, the characterization of a device operating in a wide envelope of flow conditions would require a very large number of simulations. Likewise, the multidisciplinary optimization of a device may require several geometrical iterations to reach a converged solution. Finally, multi-physics phenomena (among others, flow control, aeroelasticity and ice accretion) are characterized by larger computational times per simulation. In all these cases, researchers rely on exploiting flow symmetries to simplify the numerical setup and reduce as much as possible the computational cost of the analysis. In many cases of theoretical and industrial interest, the operation of the aerodynamic system can be analyzed using a two-dimensional, three-component (2D/3C) flow model. The basic idea behind this approach is to find a coordinate system in which the flow still has three non-zero velocity components, but all flow quantities are invariant along a certain direction (which is denoted as "spanwise"). In this case, any fluid-dynamic variable can be completely expressed as a function of the remaining ("in-plane") coordinates. This approach is suitable for the analysis of swept wings with high aspect ratio, yawed bluff bodies, three-dimensional shock-wave/boundary layer interactions with small sweep angles and also in some three-dimensional internal flows. The 2D/3C approach leads, for low-Reynolds-number constant-property flows, to the independence principle: the in-plane velocity components develop independently from the spanwise one. However, this is not the case for high-speed flows, as well as for turbulent flows, where the velocity components are mutually coupled. Yet, in those cases the 2D/3C flow model leads to a simplified set of equations, which can be solved on a two-dimensional grid. The present work deals with the OpenFOAM implementation of numerical solvers for the solution of turbulent flows within the 2D/3C assumption. Two different solvers, sweptFoam and compressibleSweptFoam, are introduced. Specifically, sweptFoam is a steady-state solver for the incompressible RANS equations, whereas compressibleSweptFoam deals with unsteady, compressible turbulent flows within a density-based framework. The numerical software is extensively validated for several benchmark cases, including inviscid and viscous, turbulent flows over infinite swept wings. It will be shown that two-dimensional simulations adopting the 2D/3C flow model are able to faithfully reproduce the fully three-dimensional flow physics with a reduced computational cost.