No abstract

No abstract

Computational aerodynamic and aeroacoustic analyses of a submerged air inlet are performed at a low Mach number. A hybrid method is used, in which the flow in the vicinity of the inlet is solved through detached eddy simulation (DES) and the acoustic pressure in the far-field is computed through the use of a Ffowcs Williams and Hawkings integral. Several surfaces of integration are used, both solid and permeable. The inlet design is based on an experimental inlet developed by the National Advisory Committee for Aeronautics (NACA). The flow is solved first through steady-state RANS simulation, then time-dependent DES is run from the converged results. The results from both RANS simulations and DES show good agreement with experimental data from NACA, both in terms of integral quantities and surface pressure coefficients. Pressure fluctuations are observed on both sides of the lip of the inlet, and are greater at low velocity ratios, with the velocity ratio defined as the ratio between the flow velocity at the duct entrance and in the free stream. A transition is observed between a quasi-laminar flow at a velocity ratio of 0.8 and a turbulent flow at velocity ratios of 0.6 and 0.4. This turbulent behaviour at low velocity ratios is associated with much higher acoustic levels in the far-field. At low velocity ratios, the acoustic spectra in the far-field exhibit a broadband character with maximum levels distributed around a characteristic frequency given by the ratio between the flow velocity at the duct entrance and the duct entrance depth. At high velocity ratios, the spectra show tonal characteristics with peaks at around 90 percent of this characteristic frequency and at the corresponding harmonics. A comparison between the spectra from solid and permeable surfaces reveals that volume sound sources are negligible at this low Mach number. A visualization of the integrands in the Ffowcs Williams and Hawkings integral show that sound sources are located on both sides of the lip of the inlet, at the position of impact of the vortices, and along the vortex wakes. Some observations regarding the use of solid and permeable surfaces of integration are made.

QC 20160523

No abstract