摘要
A Large Eddy Simulation of the flow around a NACA 0012 airfoil at a Reynolds number of 500,000 is presented. At this Reynolds number, the boundary layers transition from an initially laminar state to a turbulent state before reaching the trailing edge. Results obtained in this LES show a well-placed transition zone, and turbulence levels in good agreement with experimental data. Furthermore, the radiated acoustic field is computed directly in the same computation, which should allow future detailed examinations of the noise radiated by such a flow configuration. Recent rapid advances in Computational Aeroacoustics (CAA) have greatly increased the scope of problems that can be tackled by numerical methods. This is particularly true of Direct Noise Computations (DNC), in which the sound waves generated by turbulent flows are obtained directly from an unsteady compressible simulation of the Navier-Stokes equations. Indeed, high-Reynolds-number flows are now increasingly attainable, thanks to the development of high-order LES simulations that manage to preserve the large-scale small-amplitude acoustic perturbations alongside the small-scale large-amplitude aerodynamic fluctuations. The field of jet aeroacoustics has been particularly active in the advancement of these techniques, and a number of high-Reynolds-number jet flow simulations can be found in the literature. 2, 5, 6, 29 More recently, work on high-accuracy computations around curved geometries has shown that these methods are not restricted to Cartesian simulations. 21, 22, 26, 27, 30 A small number of time-accurate numerical studies has been performed around airfoils, generally placed at a small angle of attack to the flow, using an acoustic analogy to obtain far-field sound characteristics. 17, 21, 25, 31 The current work describes the use of a parallel curvilinear solver, based on high-order methods, to study the compressible flow around a 3-D NACA 0012 airfoil at a chord-based Reynolds number of 500,000, placed parallel to the upstream flow. It focuses on showing that realistic boundary layer transition is captured without needing to trigger the transition artificially, even when the airfoil is parallel to the upstream flow. Both mean flow quantities as well as fluctuation statistics compare well with experimental data. The acoustic field generated by the airfoil is also examined.