Computational fluid dynamics analysis and noise modeling of jets with internal forced mixers

The goal of the current research work is to develop a stand-alone jet noise prediction methodology. The current project is focused on jets with internal forced mixers, which are used in regional jet aircraft. In the current approach a two-step method is adopted. First, the turbulence properties in the jet plume are determined from Computational Fluid Dynamics (CFD) analysis using the Reynolds Averaged Navier-Stokes (BANS) equations with a two-equation turbulence model. Second, the far-field noise spectrum is predicted using a noise model based on the combination of simple single stream jet components taken from an existing experimental database. The results of this study show that the CFD predictions of the mean velocity field in the jet plume are in good agreement with experimental particle image velocimetry data. It is also observed that the CFD analysis over-predicts the turbulence levels in a simple single jet shear layer. However, it is determined that the CFD analysis under-predicts the enhancement of the shear layer turbulence levels for the forced mixers. Despite this deficiency, it is seen that the trends in the peak turbulence levels for various mixer geometries are correctly predicted by the CFD analysis. In the current study the far-field noise spectra are predicted using a noise model based on the combination of simple single stream jet components. It is found that the CFD-based two-source model noise predictions are under-predicted when compared to experimental acoustic data. This under-prediction appears to result from the under-prediction of the enhanced turbulence levels in the plume of the jets with forced mixers. In addition to the two-source noise model, a new multi-source model is proposed and evaluated. This model, which has a more general form, takes into account additional mean flow information from the jet plume. As a result, this model should be applicable to a wider range of geometric configurations. The results using this new multi-source model show a slight improvement for the less aggressive mixers. Additional discrepancies seen in the high penetration mixers are believed to be due to the deficiencies in the GFD predicted turbulence levels