Enhanced Modeling of Aircraft Taxiway Noise, Volume 2: Aircraft Taxi Noise Database and Development Process (2013)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

D-1 APPENDIX D. GROUND VORTEX INGESTION D.1. Narrow Band Analysis of Taxiing Aircraft Noise Data to Identify Ground Vortex Ingestion In order to determine the sensitivity of noise emissions versus the operating state of an aircraft that is taxiing the measurements made on taxiway E at IAD in July 2010 were utilized. A description of the measurement setup is provided in Appendix B. In order to determine what the operating state of the engine was for the aircraft that taxied by the microphones, a narrow-band analysis was performed on the recorded signal during a taxi operation. Figure D 1 shows a comparison of the power spectral density for an A319 event compared to the ambient. FIGURE D-1 Power spectral density of A319 Passby (red) and ambient (blue). Information was found for the V2522-A5 engine as part of a Type Certificate Sheet E40NE indicating that the maximum permissible engine operating speeds for the low pressure rotor (N1) is 5650 rpm. The previous ACRP Taxi study conducted by Wyle indicated that the A319s in that study had an average of 19.42 % N1 while taxiing. This, combined with the fact that the engine’s main fan has 22 blades leads one to believe that the fan noise should be around 402 Hz. As can be seen in Figure D-1, there is a noticeable peak around that frequency; however, it is broad. By considering the evolution of the spectral density with time using the spectrogram in Figure D-2, and knowing the time the engine’s cowling crosses the middle of the video image (the closest point of approach – CPA - to the sensors) one can discern the Doppler shift of this peak along with what is assumed to be the non-Doppler shifted frequency of emission for the main fan at the zero of the horizontal axis. The frequency as the engine crosses the center of the video is approximately 410 Hz. 0 100 200 300 400 500 600 700 800 900 1000 30 35 40 45 50 55 60 65 70 Frequency (Hz) P S D ( dB re f (2 0  P a) 2 / H z)

D-2 FIGURE D-2 Spectrogram of an A319 taxiing by the center sensor. The speeds of ten A319s that passed by the center sensor location were calculated using the recorded video along with the geometry presented in Appendix B. The measured taxi speeds are presented in Table D-1 along with the frequency emitted from the main fan at the CPA gleaned using the process described above. The additional columns in Table D-1 come from using the noise source characterization of the A319 pass bys in the Advanced Acoustic Model (AAM) as discussed below. (Note: One additional A319 data point was discarded due to an anomalously high RPM and low speed.) TABLE D-1 Measured Taxi Speeds of A319 Pass Bys Taxi Speed (knots) Main Fan Freq. at CPA (Hz) Average Band 26 (dB) Lmax (dBA) SEL (dB) SEL (dBC) SEL (dBA) 16.9 367 45.9 79 95 95 91 10.5 370 45.3 78 94 94 89 15.3 377 50.0 81 96 95 92 13.4 372 50.7 84 97 96 94 14.4 367 51.2 81 97 96 93 17.8 375 50.9 81 96 96 93 19.0 397 51.3 83 98 98 94 18.2 372 54.9 83 98 98 95 17.6 367 49.8 84 98 97 94 27.3 362 48.4 82 97 97 94

D-3 By using the meteorological data and the taxi speeds, the A319s can be characterized as free-field noise sources using the Acoustic Repropagation Technique (ART). The resulting characterizations can then be used in AAM to simulate taxi pass bys at the same speed and atmospheric conditions. Figure D-3 shows the one-third octave band centered on 400 Hz produced by ART. This represents the sound recorded by the 1.5m high microphone at the center sensor position with atmospheric absorption negated and propagation effects removed except for spherical spreading to 1000 feet radius. Because it is not known which of the two engines may be operating during the taxi pass by, the source is assumed to be at 5.5 feet above the ground. Also shown in Figure D-3 is the average level for the 20 to 160 degree span. FIGURE D-3 Band 26 (400 Hz one-third octave band) directivity pattern for A319 pass by 20 to 160 degree angle measured from the noise. Dashed = average level. Band 26 was chosen because the majority of the fan noise seen in the power spectral density was at frequencies within the limits of that one-third octave band. The sound exposure level (SEL) for the pass by event as simulated in AAM using standard atmospheric conditions and an 18 knot taxi speed 160’ from a 5’ (1.5m) receiver can be seen in Table D-1. Figure D-4 shows a graph of the metrics versus the main-fan frequency. As the graph shows, there is a positive trend (albeit small) with the main fan’s rotation rate. Examining the trend for Average Band 26 and SELA the slopes are more distinct than the other metrics (.0918 and .0615 vs. ~.04).

D-4 FIGURE D-4 Variation of metrics from simulated pass-bys with main fan frequency. A positive trend can be observed between the measured speed of the aircraft and the metrics from the simulation. Because the simulation includes all the noise produced by the aircraft, it is more inclusive and does not rely upon any particular noise source on the aircraft. As can be seen in Figure D-5 there is a noticeable trend between the integrated metrics and the speed at which the A319 was traveling when it was characterized. A linear trend line is shown for the A-weighted SEL calculated from the simulated pass bys. Also of note is that the averaged levels for Band 26 (400 Hz one-third octave band) show a similar trend line to that of the A-weighted SEL. y = 0.0918x + 15.636 y = 0.0615x + 69.781 y = 0.0417x + 65.968 y = 0.0458x + 79.571 y = 0.039x + 81.6 40.0 50.0 60.0 70.0 80.0 90.0 100.0 360 365 370 375 380 385 390 395 400 Le ve l (d B r ef  20  P a) Frequency (Hz) Ave.Band26 SELA Lmax SEL SELC Linear (Ave.Band26) Linear (SELA) Linear (Lmax) Linear (SEL) Linear (SELC)

D-5 FIGURE D-5 Variation of metrics from simulated pass bys with measured taxi speed. In conclusion, the evidence does support the variation of noise levels with the fan RPM and with the speed of the taxiing aircraft. There is agreement with the %N1 found in a previous study for the A319 and the consistent appearances of something near the equivalent frequency in the power spectral density. D.2. Interpretation of Taxi Condition Inflow Distortion Effects Narrowband spectra from ground taxi tests were examined in an attempt to identify fan tones and frequencies. Most of these PSD spectra show a strong fan blade passing frequency (BPF), occurring at the frequency corresponding to about 20% of takeoff fan speed. During a taxi pass-by, it is not known whether one or both engines were running, or if one engine was at idle, and whether the pilot had changed throttle during the taxi pass-by, so it is difficult to determine precisely the BPF frequency from the measurements. However, the majority of the data examined suggests a strong BPF tone present. If, in fact, the fan on the engine being measured follows accepted design practice for acoustics, then the BPF tone should be totally cut-off or non-propagating at these low taxi fan speeds. When operating at design conditions, without any inflow distortion, these fan tones should not propagate in the forward direction. Since the BPF is present in the data, it suggests that inflow distortion of some kind is present, and most likely is caused by ground vortex ingestion. Further, it is also likely that, at very low fan speeds, say ~20% of takeoff speed, that even the second harmonic of fan tone rotor stator interaction noise is cut off or non-propagating. Again considering the difficulty of pin-pointing BPF exactly from the measured PSD spectra, there seems to be evidence of strong 2BPF in some of the pass-bys measured, especially when the aircraft is approaching the microphone or at the point of closest approach. As the aircraft approaches the microphone, the forward propagating direction is the only place one is likely to measure strong fan tones, and these are likely caused by distortion. y = 0.06x + 48.114 y = 0.1098x + 78.332 y = 0.0851x + 94.172 y = 0.0871x + 93.635 y = 0.1338x + 89.052 40.0 50.0 60.0 70.0 80.0 90.0 100.0 0 5 10 15 20 25 30 35 40 45 50 Le ve l (d B r ef  20  P a) Taxi Speed (ft/s) Ave.Band26 Lmax SEL SELC SELA Linear (Lmax) Linear (SEL) Linear (SELC) Linear (SELA) Linear (SELA)

D-6 These observations suggest that inflow distortion, probably due to ground vortex ingestion, is most always present during taxi maneuvers. Therefore, when attempting to model taxi noise using the ANOPP methodology, it is suggested that the inflow distortion module be turned on in the Fan Noise component. D.3. ANOPP Inflow Distortion Model The Inflow distortion models in ANOPP (Version L28V3, October 2010) do not have a factor which reflects the magnitude or azimuthal extent of the inflow distortion, being only a function of the fan geometric and operating parameters. The Heidmann method and the Small Fan method both compute inflow distortion tones as a separate source, and this source is added to the inlet tones produced by rotor- stator interaction. The Large fan method computes inlet tones assuming that inflow distortion is present, and then applies a “clean-up” effect to these tones which varies with tone BPF harmonic number, directivity angle, and whether the aircraft conditions represent a takeoff condition or an approach condition. It is difficult to know what the inflow distortion characteristics (e.g., magnitude, azimuthal extent) are for a given aircraft and engine combination without carrying out extensive diagnostic measurements, either in scale model simulations or in a wind tunnel. Even then, actual weather conditions and airport geological and contour environment can have drastic effects on distortions ingested by the engines during a takeoff or landing maneuver. For the aircraft taxi condition, where defining this effect may be most important, it seems possible that an important mechanism is ground vortex ingestion. The ANOPP inflow distortion models do not really address ground vortex ingestion per se, but use a generalized model which may include inflow atmospheric turbulence ingestion, inlet asymmetry static pressure distortions, fan duct pylon and strut back pressure distortions, as well as ground vortex ingestion. The Large Fan method for inflow distortions is primarily an Atmospheric Turbulence Ingestion model, as it is derived from engine static measurements with and without an inflow control device (Turbulence Control Screen). It is therefore suggested that an experimental program could be developed to address ground vortex ingestion. This could be carried out on an outdoor test stand, and utilize both “Ground Vortex Destroyers” and a Turbulence Control Screen, separately and together, to identify the vortex ingestion impact, and the results of this test could be used to develop an additional source model for the Fan Method of ANOPP.