Re-assessment of the Corcos Model for Computation of Flow-induced Noise and Vibration

Since Corcos published his model for spatial correlation of turbulence wall pressures in 1964, it has been widely used in the prediction of flow-induced noise and vibration. Some revisions to the model are widely accepted - such as the low frequency correction of decay coefficients to account for boundary layer thickness [Cockburn & Jolly, 1968]. However, there is much less agreement on what corrections to the low wavenumber turbulence spectrum might be required, for low speed flows. Disparities may be partly attributed to the fact that surface pressures at low wavenumber consist of both incompressible turbulence pressure fluctuations and acoustic wavefields generated by turbulence, as anticipated by Lighthill [1954] and Curle [1955]. Another contributor is the inherently low signal to noise ratio of measurements to quantify low wavenumber content. This paper will summarized the author's experience using high resolution surface pressure microphone arrays and using transient compressible CFD simulations of turbulent flows to quantify the low wavenumber turbulence pressure spectrum and predict vibration response. These observations are used as the basis for a re-assessment of the Corcos model.