Methods for High-Speed Jet Exhaust Aero-Vibro-Acoustic Modeling in Engine Test Cells

Abstract This paper presents an aero-vibro-acoustic modeling strategy to characterize sea-level engine test cell interior acoustic environments by combining scale-resolving computational fluid dynamics (CFD) simulations, distributed equivalent acoustic source models, and boundary element methods (BEM) and statistical energy analysis (SEA). The work is motivated by the need to redesign an existing sea-level engine test cell to accommodate a higher thrust engine. Wall-modeled large eddy simulations (WMLES) in both free-field and enclosed domains are generated to target the acoustic frequencies associated with large-scale mixing and shock-turbulent interaction sources. Free-field WMLES is probed in the jet near-field to decompose the fluctuating pressure into a plurality of partial acoustic fields at each frequency of interest. Partial fields are then fit to an equivalent acoustic source model using an inverse methodology. Source directivity is further improved through the use of test data. These sources are introduced into an acoustic tool that predicts interior noise levels through a combination of BEM and gradient SEA methods. At low frequencies, the BEM pressure spectra are first corrected to a WMLES simulation of interior noise levels with no noise control treatments. This helps to account for any effects the facility has on the jet sound source. The BEM analysis is then repeated with the corrections to examine the impact of noise control treatments which are characterized through impedance boundary conditions (BCs). At high frequencies, a directional point source (DPS) is used in a gradient SEA model to capture high-frequency roll-off. Results demonstrate how the modification of acoustic impedance and absorptivity on the test cell walls may be used to quantitatively predict the effectiveness of noise control treatment (NCT) solutions. In addition to describing a novel but practical, CFD-based methodology to characterize acoustic environments and perform trade studies of noise control solutions in engine test cells, the paper discusses some of the inherent challenges associated with defining approximate acoustic source models using free-field CFD alone. Specifically, the paper will discuss the possibility of strong fluid dynamic-acoustic coupling, or “lock-in/super resonance,” which may result in catastrophic resonant feedback under certain conditions. The understanding of such resonant responses is crucial to test cell facility design, demonstrating the need for a multidisciplinary approach for the design of future test cells which combines aerodynamic and acoustic prediction methods, anchored to test measurements.