CFD Simulation of Aero-acoustic Noise Generated in Air Vent
Examensarbete för masterexamen
In Envac AB's automatic waste collection systems the air enters the pipe systems through air vents containing a nozzle. Unfortunately this air vent, and especially the nozzle, generates disturbing noise. This has as a consequence that silencers have been built which in turn leads to higher production costs and larger facilities for the air vents. The aim of the present work has been to evaluate a simulation method for studying these air vents and to use this method to analyze the generation of sound in the air inlet vents. The method studied in this thesis work is based on the simple idea that the experienced sound can be reduced by minimizing the sources of sound generation. The aeroacoustic sources represent the unsteady phenomena in the flow which creates the sound. The advantage of this method is that sound sources such as the Lilley and Linear Euler Equation sources can be calculated from a steady RANS solution. Compared to resolving the pressure fluctuations with an unsteady simulations this is a computationally less demanding method. In order to validate the proposed method CFD simulations has been performed. The first set of simulations is RANS calculations on round jets with different inlet velocities. The comparison between changes in calculated sound sources and theoretically derived changes in sound level [Ref 11] shows that a relationship exists between the predicted level of aeroacoustc sources and the theoretically generated sound. After this relation has been established simulations were performed on two test cases where the first one is a backward facing step and the second one is a similar geometry but with a bend before the step. These geometries are suitable for the present work because they include the same type of flow phenomena as the real air vent. Both unsteady LES and steady RANS calculations were performed in order to evaluate both turbulence models and aeroacoustic sources. The accuracy of the calculations on the backward-facing step has been validated through comparison with DNS data [Ref 9]. The simulations have given good results both in terms of recommending a turbulence model for geometries similar to the two test cases and in terms of evaluating the aeroacoustic method. It was shown that the k- SST model is the most suitable turbulence model for the two test cases and that the integral of Lilly's source term, the Linear Euler Equation source term and the Acoustic Power in the domain are good parameters to use when trying to minimize the generated sound.
Tillämpad matematik , Teknisk fysik , Applied mathematics , Engineering physics