Normalised spectrum for flow-generated noise prediction using computational fluid dynamics

Abstract

Flow-generated noise from in-duct strip spoilers was radiated from an open exhaust termination unit into a 70 m3 reverberation chamber and have been measured in 1/3 octave bands using sound power spectra. Computational fluid dynamics (CFD) software package was adopted to model the strip spoilers in the air duct. Based on the results of CFD simulation of relevant configurations, the technique of Mak and Au and Mak and Oldham was adopted to normalise the experimental data. A normalised spectrum has been produced for predicting the sound power level of flow-noise produced by the strip spoilers in a rectangular air duct. The data collapse for the strip spoilers is generally good at higher Strouhal numbers, but is less efficient at lower Strouhal numbers where considerable scatter is observed. Together with the normalised spectrum, the predictive equations of Mak and his co-investigators can be used to predict the flow-generated noise produced by in-duct strip spoilers in a ventilation system at design stage using CFD. Practical application: Together with the normalised spectrum, the predictive equations of Mak and his co-investigators can be used to predict the flow-generated noise produced by in-duct strip spoilers in a ventilation system at design stage using CFD.