Nikolas S. Zawodny

A covariance fitting approach for correlated acoustic source mapping

Microphone arrays are commonly used for noise source localization and power estimation in aeroacoustic measurements. The delay-and-sum (DAS) beamformer, which is the most widely used beamforming algorithm in practice, suffers from low resolution and high sidelobe level problems. Therefore, deconvolution approaches, such as the deconvolution approach for the mapping of acoustic sources (DAMAS), are often used for extracting the actual source powers from the contaminated DAS results. However, most deconvolution approaches assume that the sources are uncorrelated. Although deconvolution algorithms that can deal with correlated sources, such as DAMAS for correlated sources, do exist, these algorithms are computationally impractical even for small scanning grid sizes. This paper presents a covariance fitting approach for the mapping of acoustic correlated sources (MACS), which can work with uncorrelated, partially correlated or even coherent sources with a reasonably low computational complexity. MACS minimizes a quadratic cost function in a cyclic manner by making use of convex optimization and sparsity, and is guaranteed to converge at least locally. Simulations and experimental data acquired at the University of Florida Aeroacoustic Flow Facility with a 63-element logarithmic spiral microphone array in the absence of flow are used to demonstrate the performance of MACS.

A Comparative Study of a 1/4-Scale Gulfstream G550 Aircraft Nose Gear Model

A series of fluid dynamic and aeroacoustic wind tunnel experiments are performed at the University of Florida Aeroacoustic Flow Facility and the NASA-Langley Basic Aerodynamic Research Tunnel Facility on a high-fidelity -scale model of Gulfstream G550 aircraft nose gear. The primary objectives of this study are to obtain a comprehensive aeroacoustic dataset for a nose landing gear and to provide a clearer understanding of landing gear contributions to overall airframe noise of commercial aircraft during landing configurations. Data measurement and analysis consist of mean and fluctuating model surface pressure, noise source localization maps using a large-aperture microphone directional array, and the determination of far field noise level spectra using a linear array of free field microphones. A total of 24 test runs are performed, consisting of four model assembly configurations, each of which is subjected to three test section speeds, in two different test section orientations. The different model assembly configurations vary in complexity from a fully-dressed to a partially-dressed geometry. The two model orientations provide flyover and sideline views from the perspective of a phased acoustic array for noise source localization via beamforming. Results show that the torque arm section of the model exhibits the highest rms pressures for all model configurations, which is also evidenced in the sideline view noise source maps for the partially-dressed model geometries. Analysis of acoustic spectra data from the linear array microphones shows a slight decrease in sound pressure levels at mid to high frequencies for the partially-dressed cavity open model configuration. In addition, far field sound pressure level spectra scale approximately with the 6th power of velocity and do not exhibit traditional Strouhal number scaling behavior.

Comparison of microphone array processing techniques for aeroacoustic measurements

This paper presents a systematic comparison of several prominent beamforming algorithms developed for aeroacoustic measurements. The most widely used delay-and-sum (DAS) beamformer is known to suffer from high sidelobe level and low resolution problems. Therefore, more advanced methods, in particular the deconvolution approach for the mapping of acoustic sources (DAMAS), sparsity constrained DAMAS (SC-DAMAS), covariance matrix fitting (CMF) and CLEAN based on spatial source coherence (CLEAN-SC), have been considered to achieve improved resolution and more accurate signal power estimates. The performances of the aforementioned algorithms are evaluated via experiments involving a 63-element logarithmic spiral microphone array in the presence of a single source, two incoherent sources with similar strengths and with different strengths, and two coherent sources. It is observed that DAMAS, SC-DAMAS and CMF provide the most reliable source location estimates, even at relatively low frequencies. Furthermore, the integrated levels obtained with the array processing algorithms are shown to agree with what a single reference microphone placed at the center of the array measures when the array is appropriately calibrated. It is also shown that, as expected, the aforementioned algorithms are unsuccessful in distinguishing coherent acoustic sources unless the frequency is relatively high. DAS and CLEAN-SC are shown to be around 2 to 90 times faster than the other three algorithms.

Shear layer time-delay correction using a non-intrusive acoustic point source

Microphone array processing algorithms often assume straight-line source-to-observer wave propagation. However, when the microphone array is placed outside an open-jet test section, the presence of the shear layer refracts the acoustic waves and causes the wave propagation times to vary from a free-space model. With a known source location in space, the propagation time delay can be determined using Amiets theoretical method. In this study, the effects of shear layer refraction are examined using a pulsed laser system to generate a plasma point source in space and time for several different test section flow speeds and configurations. An array of microphones is used to measure the pulse signal, allowing for the use of qualitative beamforming and quantitative timing analysis. Results indicate that Amiets method properly accounts for planar shear layer refraction time delays within experimental uncertainty. This is true both when the source is in the inviscid core of the open-jet test section, as well as when the source is located in different model wakes of varying complexity. However, the method breaks down where the thin layer assumption fails, such as in the region where the tunnel test sections open jet interacts with the facility jet collector.

Isolated Open Rotor Noise Prediction Assessment Using the F31A31 Historical Blade Set

In an effort to mitigate next-generation fuel efficiency and environmental impact concerns for aviation, open rotor propulsion systems have received renewed interest. However, maintaining the high propulsive efficiency while simultaneously meeting noise goals has been one of the challenges in making open rotor propulsion a viable option. Improvements in prediction tools and design methodologies have opened the design space for next generation open rotor designs that satisfy these challenging objectives. As such, validation of aerodynamic and acoustic prediction tools has been an important aspect of open rotor research efforts. This paper describes validation efforts of a combined computational fluid dynamics and Ffowcs Williams and Hawkings equation methodology for open rotor aeroacoustic modeling. Performance and acoustic predictions were made for a benchmark open rotor blade set and compared with measurements over a range of rotor speeds and observer angles. Overall, the results indicate that the computational approach is acceptable for assessing low-noise open rotor designs. Additionally, this approach may be used to provide realistic incident source fields for acoustic shielding/scattering studies on various aircraft configurations.

Measurement of Phased Array Point Spread Functions for Use with Beamforming

Microphone arrays can be used to localize and estimate the strengths of acoustic sources present in a region of interest. However, the array measurement of a region, or beam map, is not an accurate representation of the acoustic field in that region. The true acoustic field is convolved with the array s sampling response, or point spread function (PSF). Many techniques exist to remove the PSFs effect on the beam map via deconvolution. Currently these methods use a theoretical estimate of the array point spread function and perhaps account for installation offsets via determination of the microphone locations. This methodology fails to account for any reflections or scattering in the measurement setup and still requires both microphone magnitude and phase calibration, as well as a separate shear layer correction in an open-jet facility. The research presented seeks to investigate direct measurement of the arrays PSF using a non-intrusive acoustic point source generated by a pulsed laser system. Experimental PSFs of the array are computed for different conditions to evaluate features such as shift-invariance, shear layers and model presence. Results show that experimental measurements trend with theory with regard to source offset. The source shows expected behavior due to shear layer refraction when observed in a flow, and application of a measured PSF to NACA 0012 aeroacoustic trailing-edge noise data shows a promising alternative to a classic shear layer correction method.

Small Propeller and Rotor Testing Capabilities of the NASA Langley Low Speed Aeroacoustic Wind Tunnel

The Low Speed Aeroacoustic Wind Tunnel (LSAWT) at NASA Langley Research Center has recently undergone a configuration change. This change incorporates an inlet nozzle extension meant to serve the dual purposes of achieving lower freestream velocities as well as a larger core flow region. The LSAWT, part of the NASA Langley Jet Noise Laboratory, had historically been utilized to simulate realistic forward flight conditions of commercial and military aircraft engines in an anechoic environment. The facility was modified starting in 2016 in order to expand its capabilities for the aerodynamic and acoustic testing of small propeller and unmanned aircraft system (UAS) rotor configurations. This paper describes the modifications made to the facility, its current aerodynamic and acoustic capabilities, the propeller and UAS rotor-vehicle configurations to be tested, and some preliminary predictions and experimental data for isolated propeller and UAS rotor configurations, respectively. Isolated propeller simulations have been performed spanning a range of advance ratios to identify the theoretical propeller operational limits of the LSAWT. Performance and acoustic measurements of an isolated UAS rotor in hover conditions are found to compare favorably with previously measured data in an anechoic chamber and blade element-based acoustic predictions.

Characterization of Noise Generation on a Canonical Nose Landing Gear Sub-system

A series of aeroacoustic experimental and computational measurements were conducted for the case of flow around a complex, three-dimensional landing gear sub-system. The model under study is a canonical representation of a nose landing gear torque link apparatus that is present on numerous commercial aircraft. Experimental measurements consisted of model steady and unsteady surface pressures, far-field acoustic measurements, and three-dimensional flow field velocities using stereoscopic particle image velocimetry. The experiments were also accompanied by computational fluid dynamics simulations using PowerFLOW software, a lattice-Boltzmann solver. The computational aeroacoustic capabilities of this software were also utilized to predict the far-field acoustic radiation via a built-in Ffowcs-Williams and Hawkings solver. Excellent agreement between the simulation and experiments have been obtained for surface pressure measurements, mean flow field behaviors, and far-field acoustics. A modified linear stochastic estimation procedure was implemented on the SPIV data to develop a low-order time-resolved estimate of the acoustic source terms of the vortex sound analogy.

Variability in the Propagation Phase of CFD-Based Noise Prediction: Summary of Results From Category 8 of the BANC-III Workshop

The usage of Computational Fluid Dynamics (CFD) in noise prediction typically has been a two part process: accurately predicting the flow conditions in the near-field and then propagating the noise from the near-field to the observer. Due to the increase in computing power and the cost benefit when weighed against wind tunnel testing, the usage of CFD to estimate the local flow field of complex geometrical structures has become more routine. Recently, the Benchmark problems in Airframe Noise Computation (BANC) workshops have provided a community focus on accurately simulating the local flow field near the body with various CFD approaches. However, to date, little effort has been given into assessing the impact of the propagation phase of noise prediction. This paper includes results from the BANC-III workshop which explores variability in the propagation phase of CFD-based noise prediction. This includes two test cases: an analytical solution of a quadrupole source near a sphere and a computational solution around a nose landing gear. Agreement between three codes was very good for the analytic test case, but CFD-based noise predictions indicate that the propagation phase can introduce 3dB or more of variability in noise predictions.

Design-optimization of a broadband phased microphone array for aeroacoustic applications

Phased microphone arrays are commonly used in acoustic beamforming applications. While numerous beamforming algorithms have been proposed to alleviate deficiencies of the delay-and-sum approach, few studies have focused on the array design itself. In aeroacoustic applications, the most common designs are based on circularly symmetric spiral arrays devised by Underbrink (1995). The design of an array using such a method is complex and tedious due to the numerous design variables and corresponding trade-offs between resolution, sidelobe suppression, size, and cost. In this paper, a systematic design-optimization approach is described that offers several objective functions and constraints. Candidate arrays for use in the University of Florida Aeroacoustic Flow Facility (UFAFF) are designed for a broadband frequency range of 1 to 80 kHz. The results of these different cases will be compared to those of an existing array design currently used in the UFAFF. An optimized design is selected and fabricated for ch...