American Institute of Aeronautics and Astronautics
092407
1
Experimental Assessment of Low Noise
Landing Gear Component Design
Werner Dobrzynski
1
Deutsches Zentrum für Luft- und Raumfahrt (DLR), 38108 Braunschweig, Germany
Leung Choi Chow
2
Airbus, Filton, Bristol, BS99 7AR, Great Britain
Malcolm Smith
3
ISVR, University of Southampton, Highfield Hampshire, Great Britain
Antoine Boillot
4
Messier-Dowty SA, 78142 Velizy Villacoublay, France
Olivier Dereure
5
Messier-Bugatti SA, 78141 Velizy Villacoublay, France
and
Nicolas Molin
6
Airbus, 31060 Toulouse, France
Landing gear related airframe noise is one of the dominant aircraft noise components at
approach. It therefore is essential to particularly reduce landing gear noise. In the European
SILENCER project, advanced low noise gears had been designed and tested at full scale. In
the current European co-financed project TIMPAN (T
echnologies to IMProve Airframe
N
oise) still more advanced low noise design concepts were investigated and noise tested on a
¼ scaled main landing gear model in the German-Dutch Wind Tunnel. A variety of gear
configurations were tested including a new side-stay design, different modifications of bogie
inclination, wheel spacing, bogie fairings with different flow transparency, leg-door configu-
rations and brake fairings. The acquired farfield noise data are compared against the results
from a landing gear noise prediction model, transposed to full scale flight conditions and
compared against the full scale test data obtained for the SILENCER advanced A340 style 4-
wheel main landing gear. An optimal combination of tested gear modifications led to a noise
reduction of up to 8 dB(A) in terms of overall A-weighted noise levels relative to the
SILENCER reference gear configuration.
I. Introduction
ue to the advances in aircraft engine noise reduction, airframe noise became a major noise component during
approach and landing. For wide body aircraft in particular the dominant airframe noise sources are the landing
gears followed by aerodynamic noise originating from deployed high-lift devices.
1
Research Engineer, Institute of Aerodynamics and Flow Technology, Lilienthalplatz 7, 38108 Braunschweig,
Germany.
2
Engineer, Aerodynamics Department, Building 09B, Filton, Bristol, England BS99 7AR, Great Britain.
3
Research Engineer, University of Southampton, Highfield Hamphire, Great Britain.
4
Research Engineer, R&T Department, Zone Aéronautique Louis Breguet, 78142 Vélizy-Villacoublay, France
5
Research Engineer, R&T Department, Zone Aéronautique Louis Breguet, 78141 Vélizy-Villacoublay, France
6
Research Engineer, Acoustics and Environment Department, 316 route de Bayonne, 31060 Toulouse, France
D
AIAA/CEAS Aeroacoustics Conference, Miami, 2009, Paper No. 2009-3276
American Institute of Aeronautics and Astronautics
092407
2
Accordingly numerous research efforts were made to reduce landing gear noise either through dedicated wind tunnel
experiments or flight tests [1, 2]. Initial noise reduction solutions were the application of solid add-on fairings to
protect the complex landing gear structure from the flow [3 - 6]. Based on the results of these experiments it was
realized that flow displacement by such fairings could be detrimental regarding additional noise originating from
gear components adjacent to those which were faired. As a solution of this problem porous fairings were developed,
which would reduce the magnitude of flow displacement but still result in a sufficiently low wake flow velocity not
to generate high interaction noise levels with the downstream gear components [7 - 9].
While add-on solutions could be applied at short terms for current aircraft it was realized that low noise gears for
future aircraft can best be developed by accounting for noise aspects already in the design stage. A corresponding
effort was undertaken in the European research project SILENCER (“Significantly Lower Community Exposure to
Aircraft Noise”) [10].
Based on the results of this study, a combination of low noise gear component design and the application of po-
rous fairings [7] was considered to exploit the maximum possible noise reduction potential. This work was per-
formed in an European co-financed research project entitled “T
echnologies to IMProve Airframe Noise” (TIMPAN)
with partners from European aircraft industries, research establishments and academia and focused on an A340 style
4-wheel main landing gear which was originally used in SILENCER, but only limited noise reduction had been
achieved accompanied by some weight penalty.
The objective of the TIMPAN study therefore was to develop operational low noise main landing gear compo-
nents without weight penalties, taking into account modifications in the gear architecture (e.g. wheel spacing and
bogie angle) and to optimize and quantify the benefit from the application of porous fairings for various gear com-
ponents. As in SILENCER the design was based on the relevant constraints predefined by gear functionality and
safety for real aircraft application.
II. Main Landing Gear Configurations
The design work was based on the SILENCER advanced A340 type 4-wheel main landing gear configuration
and focused on
• a low noise design of individual gear components known to represent major noise contributors (e.g.
side-stay, various links, leg-door structure and brakes) and
• the noise-wise optimal arrangement of gear components to minimize the interaction of high speed tur-
bulent inflow with complex gear structures (e.g. variation of bogie angle, wheel spacing, placement of
fairings and additional ramp door).
Fig. 1
presents a comparison between the
SILENCER reference configuration and one of
the TIMPAN configurations to better understand
the design philosophy in TIMPAN. One of the
drawbacks of the SILENCER design was the ex-
cessive weight of the telescopic side-stay which,
however, allowed for a noise-wise optimal design
of the leg-door structure. In TIMPAN therefore a
new side-stay design also required a new low
noise leg-door design. As shown in Fig. 1 this is a
door which is articulated in a way to (once the
gear is deployed) protect the complex leg/ drag
stay structure from the high speed inflow. It
should also be noted (Fig. 1) that for both the
SILENCER and the TIMPAN gear design the
torque link is installed in front of the leg and is
protected through a fairing, while in the back only
a narrow slave link is attached to guide the dress-
ings.
Much effort was directed towards the development of a side-stay which could be almost as quiet as the
SILENCER clean circular telescopic stay. The final design is depicted in Fig. 2
. Compared with the current A340
design the major advantage is the integration of the down-lock springs into the stay to realize a comparatively
“clean” design of the components’ outer contours. In addition an upstream ramp was provided to, at the same time,
shield the still complex stays’ geometry, the upper leg area and the cavity aperture from the flow.
SILENCER Ref.
TIMPAN
Figure 1. Comparison of SILENCER and TIMPAN main
landing gear concepts
American Institute of Aeronautics and Astronautics
092407
3
Figure 3. Brake faring with mesh type insert
for brake coolin
g
Also advanced brake fairings were developed in TIMPAN. The brakes were partly recessed and completely
separated from the flow by a streamlined fairing. To allow for the necessary brake cooling a mesh-window was fore-
seen (Fig. 3
).
A low noise arrangement of selected bogie components
included:
• Variation of the bogie angle from 0° (reference) to -
15° toe down.
• Identification of a potentially optimal wheel spac-
ing (Fig. 4
) combined with different bogie- and
torque link fairings (solid and porous, respectively).
A “narrow” wheel spacing was defined as reduced spac-
ing by 50% of the tire width relative to the reference total tire
spacing and analogous “wide” as a 50% of tire width in-
crease in wheel spacing.
Examples of the application of solid or porous bogie fair-
ings in combination with a solid or porous torque link fair-
ings are presented in Fig. 5
.
Due to budget limitations in TIMPAN only scale model
tests were planned in order to make use of the already exist-
ing ¼ scaled SILENCER main landing gear high fidelity
mock-up. Accordingly all new gear components were manu-
factured at that scale to fit to the existing mock-up. The ad-
vantage of testing at model scale was that a wide range of
configurations could be tested due to correspondingly short
stopovers for gear modification.
III. Experiments
Noise measurements were performed in the DNW-LLF (German-Dutch Wind tunnel – Large Low Speed Facil-
ity) in its free-jet configuration with a nozzle cross section of 6 m by 6 m. The maximum wind speed for this tunnel
configuration is 78 m/s (152 kts), which is close to the typical landing/ approach speed for current commercial air-
craft. The anechoic test-hall (the lower limiting frequency is 80 Hz for broadband noise) allows farfield noise meas-
Figure 5. Bogie fairings in combination with
torque link fairings for different wheel spacing
Figure 2. TIMPAN side-stay design and upstream ramp
Ref.
Ref.
Flow
Narrow
Ref.
Narrow
Narrow
Wide
Narrow
Ref.
Ref.
Flow
Narrow
Ref.
Narrow
Narrow
Wide
Narrow
Figure 4. Schematic of selected combinations
of forward and rear wheels’ spacing
American Institute of Aeronautics and Astronautics
092407
4
urements outside the flow field at lateral distances up to about 18 m from the very landing gear which is well in both
the acoustic and geometric farfield.
A. Wind Tunnel Test Set-up
As in SILENCER the model gear was installed
in a side-wall of 7 m length, which forms an exten-
sion of one side of the wind tunnel nozzle (Fig 6
).
In x-direction (i.e. streamwise) the gear was in-
stalled at a distance of about 5 m from the nozzle
exit plane. The height of the side-wall is 8 m at the
nozzle and 9 m at its trailing edge (accounting for
free jet spreading). Those areas along the wall sur-
face (upper and lower edge areas), which are ex-
posed to the wind tunnel shear layer flow, are
treated with absorptive material to minimize flow
noise generation and radiation from the side-wall.
For the same reason the wall‘s trailing edge features
a saw-tooth shape.
The side-wall arrangement was used to simulate
the “in-flight” geometric/acoustic environment (re-
flection geometry from the wing surface) and to
reduce flow noise radiation from the support struc-
ture. In order to simulate the actual in-flight lower
wing surface boundary layer thickness the wind tunnel boundary layer was “peeled off” by means of a scoop which
is installed along the side-wall‘s leading edge. In TIMPAN the existing DNW test set up was adapted to install the ¼
scale mock-up of an A340 type main landing gear for later comparison with the SILENCER WP 2.3 advanced full
scale main landing gear test results [10].
During landing/ approach the A340 aircraft typically operates at a characteristic angle-of-attack with respect to
the inflow direction. Since in the test set-up the flow direction has to be parallel to the surface of the side-wall this
difference between inflow direction and aircraft axis must be accounted for. Based on the gear installation angle in
the aircraft (and accounting for deviations of local flow- from flight-directions) a slight backward gear-leg orienta-
tion was decided upon for the wind tunnel set-up. On the aircraft the main landing gear-leg is (laterally) inclined
with respect to the lower wing surface. Therefore in the test set-up a corresponding inclination angle was realized
between the gear-leg and the surface of the side-wall.
The gear bay geometry was not exactly reproduced. Instead an almost rectangular bay was used, internally lined
with absorbing foam to avoid acoustic resonance phenomena. However, the bay aperture was exactly simulated.
B. Measurement Techniques and Data Analy-
sis
A similar measurement set-up was applied as in
the preceding SILENCER test, i.e. 2 microphone
arrays “looking” to the gear from two directions
and 4 different rows of microphones (with micro-
phones distributed in flow direction) were in-
stalled close to the wall (2 rows), the floor (1 row)
and the ceiling (1 row), respectively (Fig. 7
). That
way noise radiation both towards the “ground”
and in sideline directions were determined. In
each individual row 9 microphones were posi-
tioned at angular increments of about 10° or less
in streamwise direction, covering a range of polar
angles between about 60°
<ϕ<
x
125° (Fig. 7).
All farfield measurement positions were equipped
with 1/2“-diameter LinearX M51 type electret
Figure 6. Overview of the measurement set-up in the
DNW-LLF 6 m by 6 m open test section
View against flow direction:
6 m by 6 m nozzle
= 51° Outboard
Di
r
e
c
t
i
o
n
t
o
t
h
e
“
g
r
o
u
n
d
”
= - 35° Inboard
9 Floor Mics:
# 28, 29, 30, 31, 32, 33, 34, 35, 36
9 Ceiling Mics
:
# 19, 20, 21, 22, 23, 24, 25, 26, 27
12°
9 Wall Mics:
# 1, 2, 3, 4,
5, 6, 7, 8, 9
9 Wall Mics:
# 10, 11, 12, 13,
14, 15, 16, 17, 18
y
ϕ
y
ϕ
9°
View against flow direction:
6 m by 6 m nozzle
= 51° Outboard
Di
r
e
c
t
i
o
n
t
o
t
h
e
“
g
r
o
u
n
d
”
= - 35° Inboard
9 Floor Mics:
# 28, 29, 30, 31, 32, 33, 34, 35, 36
9 Ceiling Mics
:
# 19, 20, 21, 22, 23, 24, 25, 26, 27
12°
9 Wall Mics:
# 1, 2, 3, 4,
5, 6, 7, 8, 9
9 Wall Mics:
# 10, 11, 12, 13,
14, 15, 16, 17, 18
y
ϕ
y
ϕ
9°
Figure 7. Selected farfield microphone positions on the
wind tunnel side wall, the ceiling and the tunnel floor
American Institute of Aeronautics and Astronautics
092407
5
freefield microphones. Acoustic data were acquired up to a frequency of 40 kHz.
The analysis and reduction of farfield noise data aimed at the determination of noise level spectra and radiation
directivities for different landing gear configurations at different flow velocities. If this basic information is at hand,
the measured data may ultimately be extrapolated towards the operational conditions as specified for approach noise
certification. To obtain the true source characteristics from wind tunnel out-of-flow acquired acoustic data, sound
pressure levels have to be corrected for insufficient signal-to-noise ratio, for the effects of shear-layer refraction [11]
(including wave convection), microphone directivity, atmospheric absorption [12] and for the effect of convective
amplification (assuming dipole type sources). All farfield noise data were normalized towards a constant propaga-
tion radius and will be presented in terms of 1/3-oct. band levels.
To visualize local flow conditions at selected gear components tufting tests were performed. Pictures from two
different view angles were recorded by means of two video cameras.
IV. Test Results
To ensure the quality of the data, the test started with a background noise measurement for the clean side-wall,
i.e. without gear and closed gear cavity. The test matrix comprised a total number of 47 gear configurations, result-
ing from different combinations of individual gear component designs. In order to save measurement time, different
from the “standard” procedure of testing for all configurations at 3 different speeds, the majority of configurations
were tested at 2 speeds only (i.e. 78 and 62.5 m/s).
To enable an extrapolation of noise data towards speeds beyond the measurement range respective scaling laws
must be defined. As one result of previous landing gear noise tests, dipole type noise source mechanisms were found
to dominate. Therefore the following velocity scaling of levels and frequencies pertain:
()
6
ref
vvlog10L ⋅=Δ (1)
based on an arbitrary reference speed v
ref
. From measured frequencies f and flow velocities v the relevant non-
dimensional Strouhal number St can be calculated as
.const
v
sf
St =
⋅
=
(2)
with s as characteristic length scale or scale factor.
In order to allow for the comparison of earlier results from full scale tests in SILENCER with the current scale
model data in TIMPAN the identical length scale s can be applied, but multiplied by the relevant scale factor thus
automatically accounting for 4 times higher frequencies in this model experiment.
To finally present source noise level directivities and account for the source size (model scale factor) at the same
time, all data will be referenced towards a constant propagation distance
ref
r
based on spherical sound attenuation
relative to the measurement distance r and accounting for the model (source) size through
()
(
)
sslog20rrlog20LL
refrefref
⋅
+⋅
+
= (3)
The data from microphones at similar streamwise (
x
ϕ
) positions but for slightly different azimuthal angles in
the range of
°<ϕ<° 123
y
(corresponding to the two rows of microphones on the test hall wall) were averaged and
considered to represent the noise characteristic for radiation direction towards the “ground”. This was considered
reasonable since the respective level spectra show similar and systematic variations for all tested gear configurations
in the order of less than 1 dB.
In order to check the validity of the anticipated scaling laws to account for the effect of flow speed on broadband
landing gear noise, in the following spectra are presented in a non-dimensional form based on Equ. (3) to normalize
levels and Equ. (2) to calculate Strouhal numbers from measured frequencies.
Prior to any comparison of noise spectra for different gear configurations as studied in TIMPAN it is worthwhile
to check how well the noise spectrum for the original SILENCER full scale advanced main landing gear compares
to the noise spectrum as obtained for the ¼ scale gear in its SILENCER reference configuration after transposition to
full scale conditions according to Equs. (2) and (3). This comparison is depicted in Fig. 8
and shows a surprisingly
good agreement except for a level peak at a frequency of about 1 kHz. It is interesting to note that both at model
scale and at full scale a similar level peak occurs, but with a much higher level for the scale model gear. The com-
parison of broadband noise levels is excellent for forward arc radiation directions while, compared to the full scale
build, the model gear turns out to be about 1 to 2 dB noisier for rear arc radiation directions.
