D1100 CFD for foil and propeller. Literature review and recommended methods.

This report reviews the Computational Fluid Dynamics (CFD) tools used in literature for solving the flow past foils and propellers, with a particular attention to those used in predicting noise generation and propagation. Limiting the review to the numerical algorithms developed for incompressible fluids, the CFD methods can be divided into two categories, depending on whether they consider as the working medium an inviscid and irrotational fluid or a viscous fluid. In the first case, some simplifications of the governing equations are possible, and the resulting numerical methods are particularly fast and efficient, but they cannot take into account all those effects bound to the viscosity of the fluid, in particular the onset of turbulent motions. To the second category belong the viscous solvers that provide a more complete modelling of the fluid properties at the expense of a much higher computational cost. Consistently with the Reynolds number of the flows that will be considered in the project, two types of approach can be suggested: 1. For the design/optimization phase, when many simulations near the design point are to be performed, potential methods can deliver very fast and reasonably accurate solutions, that permit a good prediction of the tonal noise generated by the propeller. These methods require a relatively small computational effort and can be run on a desktop computer. 2. For an accurate a posteriori validation of the propeller characteristics, the best option is represented by either a hybrid Reynolds-Averaged Navier-Stokes/Detached Eddy Simulation (RANS/DES) or a wall-modelled Large Eddy Simulation (LES). Both these approaches permit to capture not only the tonal noise but also the broadband noise generated by turbulence. They require a significant computational workload (typically a small cluster of modern CPUs committed for several days) and should be used only for a few selected validation runs. This report will try to highlight the differences between the different approaches, motivating why some methods should be considered for the forthcoming computations, or are not considered suitable for the purposes of the project.