Flyover Beamforming

Flyover Beamforming
Noise Source Identification During Flyover of Passenger Aircraft
• Noise source identification during flyover testing during R&D
and at the certification stage of a project
• Noise source identification on subsonic transport aircraft on
undercarriage, slats and flaps, and engines
• Useful for flyover altitudes of between 30 m and 300 m
• Records all signals during measurements
• Transient tracked analysis using flight track information via
IRIG-B time coded signal
• Deconvolution algorithms used to improve spatial resolution
relative to classical beamforming
• Post-processing of data
• Photographic validation of position of aircraft for on-the-spot
synchronisation with acoustical data (optional)
• Fast validation on site of beamforming calculations (optional)
The aircraft position during a flyover is measured with an
onboard GPS system, and synchronisation with array data is
achieved through recording of an IRIG-B signal together with the
array data and the GPS data on the aircraft. The Beamforming
calculation is performed with a standard tracking time-domain
Delay And Sum (DAS) algorithm [1], with the capability of
Diagonal Removal to suppress wind noise.
For each focus point in the moving system, FFT and averaging in
short time intervals is then performed to obtain spectral noise
source maps representing the aircraft positions at the middle of
the intervals. With sufficiently short averaging intervals, the array
beam pattern remains almost constant during the corresponding
sweep of each focus point. This means that a Deconvolution
calculation can be performed for each FFT frequency and for
each averaging interval in order to enhance resolution, suppress
sidelobes and scale the maps.
In relation to the Deconvolution, it is important to take into
account the frequency shift of the sidelobes in the calculation of
the Point Spread Function (PSF) [2]. The requirement for
compensation can, however, in many cases be avoided by the
use of a carefully selected FFT record length. For the
Deconvolution, a FFT-NNLS algorithm is used [3].
The array design and the use of a frequency-dependent smooth
array-shading function are inspired by Sijtsma and Stoker [1].
However, to support quick and precise deployment on the
runway, a star-shaped array geometry is used as illustrated in
Fig. 1. The full array consists of 9 identical line-arrays which are
joined together on a centre plate and with equal angular spacing
controlled by aluminium arcs. The 12 microphones on one 6 m
line array are clicked into an aluminium tube, rotated in such a
way around its axis that the ¼ microphones touch the runway.
Due to the turbulence-induced loss of coherence over distance, a
smooth shading function is used that focuses on a central subarray, the radius of which is inversely proportional to the
frequency [1]. At high frequencies only a small central part of the
array is therefore used, which must then have small microphone
spacing. To counteract the resolution loss at low-to-medium
frequencies resulting from the high microphone density at the
centre, an additional weighting factor is applied to ensure
constant effective weight per unit area over the active part of the
The effective frequency-dependent shading to be applied to each
microphone signal is implemented as a FIR filter, which is applied
to the signal before the Beamforming calculation. An important
benefit of using only a circular central sub-array at each
frequency (over which microphone signal coherence is not lost)
is that loss of coherence need not be modelled in the PSF used
for Deconvolution.
Fig. 1 Schematic diagram showing a typical customised system for Noise Source Identification during flyover of passenger aircraft – measurement and
data acquisition
Weather parameters
- Temperature
- Relative humidity
- Barometric pressure
- Wind speed
- Wind direction
time code
via USB
12 m
Microphone position
Typical System
For a frequency range of 500 Hz to 6 kHz, use a 12 m diameter
ground based circular array, which includes 9 spokes, each
containing 12 microphones (and a total of 108 channels) [4].
File interface for
import of flight path
from dedicated
3rd party subsystem
For improved low frequency resolution (200 Hz to 6 kHz), an
array with a diameter of 30 m, and 9 spokes each containing 18
microphones is recommended.
[2] Guérin, S. and Siller, H., “A Hybrid Time-Frequency Approach
for the Noise Localization Analysis of Aircraft Fly-overs”, 14th
AIAA/CEAS Aeroacoustics Conference, Vancouver (Canada),
5-7 May 2008, AIAA Paper 2008-2955.
[3] Ehrenfried, K. and Koop, L., “A comparison of iterative
deconvolution algorithms for the mapping of acoustic sources”,
12th AIAA/CEAS Aeroacoustics Conference, Cambridge,
Massachusetts (USA), 8-10 May 2006, AIAA Paper 2006-2711.
[4] Jørgen Hald, Yutaka Ishii, Tatsuya Ishii, Hideshi Oinuma,
Kenichiro Nagai, Yuzuru Yokokawa and Kazuomi Yamamoto,
“High-resolution Fly-over Beamforming Using a Small Practical
Array”, AIAA/CEAS Aeroacoustics Conference, Colorado Springs
(USA), 4-6 June 2012, AIAA Paper 2012-2229.
Brüel & Kjær reserves the right to change specifications and accessories without notice. © Brüel & Kjær. All rights reserved.
HEADQUARTERS: Brüel & Kjær Sound & Vibration Measurement A/S · DK-2850 Nærum · Denmark
Telephone: +45 7741 2000 · Fax: +45 4580 1405 · · [email protected]
Local representatives and service organisations worldwide
BU 3085 – 12
[1] Sijtsma, P. and Stoker, R., “Determination of absolute
contributions of aircraft noise components using fly-over array
measurements”, 10th AIAA/CEAS Aeroacoustics Conference,
Manchester (UK), 10-12 May 2004, AIAA Paper 2004-2958.
Fig. 2 Results for the 2 kHz octave band when the nose of the aircraft is exactly over the array centre (x=0) under the following conditions: Landing
configuration, Level flight, Engine idle, Altitude 59 m, Speed 57 m/s. Left: DAS; Centre: DAS + Shading; Right: DAS + Shading + Deconvolution
(Measurement data obtained during joint research work between JAXA (Japan Aerospace Exploration Agency) and Brüel & Kjær)
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF