Content uploaded by Hadar Ben-Gida
Author content
All content in this area was uploaded by Hadar Ben-Gida on Apr 04, 2021
Content may be subject to copyright.
Noise Reduction of a UAV Propeller Using Grit–Type
Boundary Layer Tripping
H. Ben-Gida∗, M. Faran†. T. Kogan‡
Israeli Air Force
O. Stalnov§
Faculty of Aerospace Engineering, Technion - Institute of Technology, Haifa, 32000, Israel
The noise reduction of a small size UAV propeller, operating at
Reynolds numbers of 104, was studied by implementing grit–type bound-
ary layer tripping technique. Static noise measurements were conducted
in the anechoic chamber for several modified propeller configurations.
The effect of different grit particles, strip lengths and rotational veloc-
ities of the propeller were investigated on the noise spectrum, where
the acoustic results of the modified propeller were compared with those
of the un–tripped propeller. Although results showed the grit–type
particles generated higher broadband noise levels for high–frequencies,
the low–frequency tones (in particular the first harmonic) were reduced
by 2–6dB for a wide range of propeller RPMs. Reducing these low–
frequency tones can present a significant advantage in acoustic stealth
of aircraft flying at high altitudes.
I. Introduction
Unmanned aerial vehicles (UAVs) are being extensively developed for the past twenty
years and are playing increasingly important role in many fields (civilian or military).
Recent increase in the number of UAVs, combined with the integration of quiet electric
motors, have resulted in propeller noise being of major concern, particularly for military
applications. Thus, reducing propeller noise has an important role in the survivability
of the UAVs, as they become more vital in surveillance and reconnaissance missions [1].
Therefore, currently both the aeronautic and acoustic communities are investigating novel
approaches in reducing the noise emitted by propellers.
Throughout the history of mankind, humans have always looked at nature for solutions
to their problems. Such an approach, also known as biomimetics, has given rise to novel
technological solutions that imitate models, systems or elements of nature. In the context
of aeroacoustic noise reduction, one can imitate the silent flight of owls. Over millions of
years of evolution, owl species has produced many specialized configurations. Yet, almost
every member of the clan has common characteristic of silent flight, which is not audible
∗Officer at the Aerodynamic Group, IAF, bengida1989@gmail.com
†Officer at the Signature Group, IAF, michaelfaran@gmail.com
‡Head of the Loads and Flutter Group, IAF, tuvikog@gmail.com
§Assistant Professor, Faculty of Aerospace Engineering, oksana.s@tx.technion.ac.il
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
to man and, more important, to their prey [2,3]. This unique characteristic was found to
be related to the owl feathers, as first reported by Graham in 1934 [4]. In his study, he
identified three features of the owl feathers that are presumed responsible for the silent
flight: the leading–edge serrations, the fringes in the wing trailing–edge, and the velvet
upper surface of the wing.
Inspired by Graham’s insights on the silent flight of owls, Soderman [5] investigated
the possibility of reducing a two–bladed rotor noise by implementing leading–edge ser-
rations. He investigated rotational velocities that correspond to tip Reynolds numbers,
Rect(based on the tip chord length, ct), of 105–106. The highest noise reductions of 4
to 8dB in Overall–Sound–Pressure–Level (OASPL) were measured at Rect
∼105, with
no degradation in the rotor performance. It is noteworthy that owls achieve their silent
flight mode in similar Reynolds numbers. Hersh et al. [6] later postulated that attaching
leading–edge serrations to the rotor blades generate chordwise vortices on the suction
surface [7], and also trip the laminar boundary layer. Thereby, the tones, generated by
wake vortex shedding, change from periodic, or semi–periodic, to broadband. It should
be noted that trailing–edge serrations were also recently implemented on wind turbine
blades, and showed a reduction in broadband noise [8,9].
Today, many of the propellers designed for the new UAVs are required to operate at
Reynolds numbers (based on the chord length, c) of Rec∼104–105; similar to the range
occupied by owls. Thereby, a laminar boundary layer is expected to develop over most
of the propeller sections, commonly accompanied with the formation of a Laminar Sepa-
ration Bubble (LSB) on the suction surface. In such conditions, significant performance
drop and narrow band tones are expected [10, 11]. Although integrating leading–edge
serrations on UAV propellers might be an adequate solution for the aforesaid problems,
the implementation of such technology on propellers is a delicate procedure that might
result in maintenance issues, particularly for military applications.
A more practical approach is tripping the boundary layer on the suction surface of the
propeller blades. Overall, boundary layer trips are implemented on the leading–edge of
the wing to generate an early transition of the laminar boundary layer, which can result
in increased performance [12]. In this process, the LSB is eliminated, along with its
associated narrow band tones [10, 11,13]. Instead, a broadband noise spectra is formed,
due to the presence of turbulent boundary layer [13].
Leslie et al. [14] have further investigated the possibility of reducing this broadband
noise on a small size propeller (Master Airscrew 10x5) operating at Reynolds numbers
up to Rec= 2 ·104. They have studied two leading–edge boundary layer trips (alu-
minum and scotch tapes) that were implemented on the suction surfaces of the propeller
blades at range of chordwise locations x= 0.05 −0.2c, with heights in the range of
0.3–0.8δ(δ- boundary layer thickness). Measurements (with and without free stream)
have shown these boundary layer trips can produce a large reduction in broadband noise
(roughly above 1.5kHz), with no evidence of performance degradation. Moreover, higher
broadband noise was correlated with lower trip heights.
Reducing the high–frequency broadband noise spectra can be beneficial for low alti-
tude surveillance missions of UAVs; where these high–frequencies can be audible to anyone
on the ground. However, at high altitudes, where the high frequencies are attenuated by
the atmosphere, the low–frequency tones play a dominant role, thus endangering UAVs
during surveillance missions. Hence, development of boundary layer tripping techniques
for reducing the low–frequency tones of UAV propellers is of high significance.
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
A. Current Research
In the current paper, we investigate the possibility of reducing the low–frequency tones
of a small size propeller, operating at Reynolds numbers up to Rec= 2 ·104, by imple-
menting grit–type boundary layer tripping technique. Far–field noise measurements of
the propeller were performed in an anechoic chamber. A parametric investigation was
conducted to determine the effect of grit size, distribution (along the span of the propeller
blades) and the propeller rotational velocity on the noise spectrum. The acoustic results
of the modified propeller are compared with those of the un–tripped propeller.
II. Experimental Setup
A. Testing Facility
The experiments were conducted at the anechoic chamber of the Israel Defense Forces
(IDF). The chamber walls are specially designed to absorb sound waves and prevent their
reflection, thus allowing the propeller noise measurements to be conducted with negligible
background noise. The chamber dimensions are 3.5m in length, 3.5m in width and 1.5m
in height.
B. Propeller and Motor Unit
The propeller model used for the noise experiments is a two–bladed Master Airscrew G/F
3 Series KN 1160 propeller. It has a radius of R= 0.14m and a pitch of 0.15m. Figure 1
depicts the variation of the chord length and the twist angle along the propeller blade
span. This propeller was chosen due to the simplicity of its platform shape and airfoil
sections, thus allowing a simple implementation and investigation of the boundary layer
tripping. The measured noise data of this propeller can give significant insight, which
later may define the grit tripping characteristics for reducing the noise of UAV propellers
that operate in the Reynolds numbers of 104−105.
Figure 1: Spanwise variation of the chord length, c, and the twist angle, β, along the
radial distance from the blade root of a Master Airscrew G/F 3 Series KN 1160 propeller.
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
The propeller was operated using an OMA–3820–1200 electrical motor model, where
its rotational velocity was set with a wireless controller. The propeller–motor system
was mounted to a 0.5m height tripod stand, as depicted in Figure 2, in order to avoid
flow reflections from the chamber floor. Before the start of each noise measurement, the
rotational velocity of the propeller was measured with infra–red based RPM meter that
was placed on top of the tripod stand (see Figure 2). The measured values were compared
to the ones set by the wireless controller in order to verify the correct propeller rotational
is set. The motor power was supplied by three to four LiPo (Lithium polymer) battery
cells (each supplying 3.7V).
Figure 2: The propeller and motor system mounted on a tripod stand inside the anechoic
chamber.
C. Grit Particles
In the current study, we used SiC (Silicon Carbide) grit boundary layer tripping technique
to achieve transition. To generate the boundary layer transition on the propeller blades
in the range of Reynolds numbers investigated, we used different grit particles with an
average diameter in the range of d=0.25mm–0.56mm, as summarized in Table 1. These
grit types were chosen based on the empirical relation given by Barlow et al. [15], who
studied the effectiveness of grit–type boundary layer tripping in a wide range of Reynolds
numbers.
The grit particles were placed on a strip on both the suction and pressure surfaces.
According to Barlow et al. [15], in order for the grit–type strips to be effective, they need
to be placed at a distance of 0.05–0.1cfrom the leading–edge. As one would expect,
the tripping location on the wing can have a significant effect on the acoustic spectra.
Therefore, such location should be chosen carefully. Herr [16] has reported the low–
frequencies contributions decreased when the tripping location is further away from the
leading–edge, while high–frequencies contributions increased. Since we aim to reduce the
low–frequencies, we placed the grit strips at a distance of roughly 0.1cfrom the blade
leading–edge, which is the highest distance from the leading–edge that is efficient in
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
tripping the boundary layer [15]. The width of each strip was 3mm, where its height
can be estimated by the average particles diameter d(see Table 1). Two configurations
of the grit–type strips were tested in the current study; the first configuration included
a grit–type strip along the full span of the propeller, while in the second configuration
the strip was placed only on the outer semi–span of the blade. Figure 3 shows these two
configurations.
Table 1: The grit–type boundary layer tripping particles used in the experiments.
Grit No. d[mm]
#30 0.56
#36 0.48
#46 0.36
#60 0.25
(a) (b)
(c) (d)
Figure 3: The tested grit–type configurations in the anechoic chamber. (a) Grit No. #60
placed on the suction surface and along the full span of the propeller blade; (b) Grit No.
#60 placed on the pressure surface and along the full span of the propeller blade; (c) Grit
No. #36 placed on the suction surface and along the outer semi–span of the propeller
blade; (d) Grit No. #36 placed on the pressure surface and along the outer semi–span of
the propeller blade.
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
D. Far–field Noise Measurements
The propeller noise was measured using four 1/2” free–field B&K microphones (model
4189) and B&K data acquisition system (3050–A–040 model) operated at 16.4kHz. The
four microphones were placed around the propeller–motor system that was mounted on
a tripod and located at the center of the anechoic chamber. Three of the microphones
were placed ahead of the propeller, at a distance of 0.5m; two of the microphones were
placed at angles of 45oand 315oto the propeller, while the third microphone was placed
directly in front of the propeller (at an angle of 0o). The fourth microphone was placed
behind the propeller (at an angle of 180o), at a distance of 0.5m. Figure 4 depicts the
microphones setup in the anechoic chamber.
Figure 4: The microphone setup used inside the anechoic chamber. Three microphones
were placed ahead of the propeller, and one behind it.
E. Experimental Apparatus
The test conditions included five rotational velocities of the propeller at 4,000, 5,000,
6,000, 7,000 and 8,000 RPM. For each rotational velocity, a total of seven propeller
configurations were tested (one clean and six modified propeller configurations) at static
conditions. The objective of this measurements is to investigate the effect of grit–type
size and span variation on the noise spectrum. Table 2 summarizes the different propeller
configurations. For each configuration, experiments were repeated twice for validation.
F. Uncertainties and Limitations
The first Blade Passing Frequency (BPF) tone of the propeller that expected to appear
in the spectrum is calculated by the following equation:
f=NRP M
60 ,(1)
where Nis the number of blades. For a propeller rotating at 4,000 RPM, the fundamen-
tal frequency of tonal noise, according to Eq. 1, is equal to 133Hz. Thereby, harmonics
below this frequency should not appear for a perfectly balanced propeller. However,
tones of lower frequency might appear in the spectrum if the propeller is slightly out–of–
balance [17]. In the experiments conducted during this study, we encountered a slightly
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
Table 2: Propeller configurations that were tested in the anechoic chamber for each RPM.
Configuration
No. Grit No. Strip length
1 - none
2 #30 full span
3 #30 outer semi–span
4 #36 full span
5 #36 outer semi–span
6 #46 full span
7 #60 full span
imbalance rotating propeller. This could be due to several reasons. First, the rotating
axis of the engine, on which the propeller was installed, was not perfectly straight; thus,
causing some low–frequency vibrations. In addition, the grit–type particles were dis-
tributed randomly along each strip, which might lead to an asymmetry between the two
blades. Moreover, during the rotation, we observed random shedding of grit particles,
due to centrifugal forces.
III. Results and Discussion
A. Data Analysis Method
After reaching a steady RPM, the noise signals were acquired for at least 60sec. Each
recording was presented in its spectrogram, which reveals the Sound–Pressure–Level
(SPL) dependence on time/frequency. A Hanning window type was used, with 2048
samples and 80% overlap. The noise spectra presented in the current study are those
that correspond to the 80% percentile of the SPL signal over time, per each unit fre-
quency of the spectrogram. We used the 80% percentile, instead of the average 50%
percentile, since it is more commonly used in applications where rigorousness is essential.
B. Directivity Analysis
In the literature, studies have reported that the azimuthal directivity of the broadband
noise and the tonal loading noise for propeller at static conditions (without the free
stream velocity present) are the highest along its rotation axis (rear and forward of
the propeller) [18–20]. Using the grit–type tripping technique, one can alter this noise
directivity. Figure 5 depicts the spectra obtained from the four microphones at 5,000
RPM and with #36 grit–type strip placed on the full span of the propeller.
It is shown that microphones placed in front and behind of the propeller (Mic. 1 and 2,
respectively), have the highest tonal SPL (for balanced harmonics that are in multiples of
the fundamental rotating frequency), as expected. However, microphone 2 also recorded
the noise generated by the propeller wake, which can be identified as a broadband noise
in the low–frequency range. Therefore, the signal recorded by microphone 1 was chosen
for the acoustical post–analysis performed in this study. Unless otherwise indicated, we
present only measurements recorded by the microphone directly in front of the propeller
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
Figure 5: Comparison of the spectra obtained by the four microphones at 5,000 RPM
with #36 grit–type strip placed on the full span of the propeller.
(Mic 1; see Figure 4), for clarity.
C. Repeatability of the Results
To increase the validity of the results reported herein, we investigated the repeatability of
acoustic measurements. For this purpose, we repeated the noise measurements of several
configurations twice, at the same conditions. Figure 6 depicts spectra obtained from two
different experiments with similar propeller configurations at 5,000 RPM. Three different
sizes of grit–type particles were investigated (#36, #46 and #60), as strips placed along
the full span of the propeller blades. One may notice there is a negligible difference in
SPL between each pair of spectra. Thereby, the noise measurements are shown to be
consistent and repeatable.
D. Strip Length Effect
The following subsection presents the effect of strip length on the acoustic spectra. Two
strip length configurations of the grit–type particles were studied: (i) full span and (ii)
outer semi–span coverage. For each strip length, two rotational velocities were studied,
the 4,000 and 5,000 RPM. Figure 7 depicts a comparison of full and outer semi–span
coverage of #36 grit–type particles strip. It is assumed that the BPF tonal noise and the
broadband noise should be similar for the two strip length coverages studied. This is due
to the fact that propeller noise sources from the outer semi–span (both broadband and
tonal) contribute significantly more to the total acoustic radiation.
As depicted in Figure 7, similar acoustic spectra were found with the modified pro-
peller at 4,000 RPM and 5,000 RPM. A slight difference in the tonal noise can be observed
between the two strip lengths, while the resulted high–frequency broadband noise remain
unaltered. The propeller configuration with the outer semi–span coverage is shown to
have a slightly increased BPF harmonic SPL than the full span covered blades. This
difference appears to grow with increasing harmonic order. One may attribute this phe-
nomenon to the different lift distribution forming on the inner part of the propeller blades,
which are not covered with grit particles. The effect of different strip lengths on the pro-
peller noise should be examined further and more extensively to gain a better insight on
the aforesaid phenomenon.
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
(a)
(b)
(c)
Figure 6: Comparison of two noise measurements of the same propeller configuration at
5,000 RPM, with grit particles placed as a strip along the full span. (a) #36 grit; (b)
#46 grit; (c) #60 grit.
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
(a) 4,000 RPM
(b) 5,000 RPM
Figure 7: Comparison of full and outer semi–span coverage of #36 grit–type particles
strip placed on a propeller rotating at 4,000 and 5,000 RPM.
E. Grit Size Effect on Tonal Noise
The effect of the grit–type particles size on the noise spectrum was examined by placing
different grit–type strips (#30, #36, #46 and #60) on the full span of the propeller
blades, rotating at 4,000 RPM. As depicted in Figure 8, the harmonic of the imbalance
propeller appears in all of the spectra. This low–frequency harmonic is below the first
harmonic, as predicted by Eq. 1. Moreover, narrow–band tones that do not satisfy the
frequencies estimated by Eq. 1 appear in all the spectra shown in Figure 8. The spectra of
the clean propeller and the modified propellers with #30 and #36 grit–type strips show
these harmonics are almost negligible (in SPL) compared to the BPF harmonics. How-
ever, for the modified propellers with the #46 and #60 grit–type strips, these harmonics
are showing to have higher SPL than the BPF harmonics. We presume these behavior
originated from some periodic disturbance in the flow; such that is not correlated with
the cyclic BPF of the blades.
As depicted in Figure 8, the spectra of all the modified propeller configurations are
characterized with a clear increase in high–frequency broadband noise. This increase in
broadband noise suggests a turbulent boundary layer was successfully generated for all
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
modified propeller configurations. It is noteworthy that the broadband noise level for all
modified propeller configurations is higher than the clean propeller configuration. This
phenomenon is in contrast to the one described by Leslie et al. [14], who reported lower
broadband noise levels when tripping the boundary layer. However, Leslie et al. [14] used
trip tapes with heights that are almost an order of magnitude lower than the size of the
grit particles used in the current study. Therefore, introducing lower turbulence levels,
that are being manifested by the lower broadband noise. It appears that using grit–type
particles for tripping the boundary layer on the propeller can result a rather different
spectrum with higher levels of high–frequency broadband noise. Further theoretical and
experimental studies are required to gain insight on the difference in spectra between
these two tripping methods.
One may also notice that the broadband noise level increases as the implemented
grit particles are smaller in size. Presumably, the lower grit particles generate smaller
turbulent structures, which in turn increases the broadband noise level, as shown in Fig-
ure 8. However, if smaller grit–type particles are used, the tripping of the boundary layer
would not occur. Thus, the broadband noise would disappear along with the masking of
the high–frequency tonal noise of the propeller. As a result, only narrow range of grit
particles will be beneficial for a certain propeller, where experiments can be performed
to determine what is the optimum grit size.
Although high broadband noise levels exist, results for the modified propellers with
#30 and #36 grit–type strips revealed the first BPF harmonic is lower (in comparison to
the clean configuration), by approximately 6dB. The spectrum obtained for the modified
propeller with the #30 grit–type strip is lower in SPL for all respective BPF harmonics,
when compared to the spectrum of the clean propeller. A reduction of the SPL in the
low–frequency range is of high importance for acoustic stealth; this is because the high–
frequency content of the signal is attenuated much faster than the low–frequencies by the
atmosphere. Further research should verify that this comparative conclusion is invariant
in free stream conditions. As the size of the grit particles decreased, the bandwidth
of each harmonic is increased. This might occur due to the variation of the turbulent
fluctuations, which in turn alter the lift coefficient in time.
Figure 8: The effect of the grit particles size on the noise spectrum of the propeller at
4,000 RPM. Grit particles were placed as a strip along the full span of the propeller
blades.
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
(a) 4,000 RPM
(b) 7,000 RPM
(c) 8,000 RPM
Figure 9: The effect of the propeller RPM on the noise spectrum for a clean configuration
and configuration with a 36# grit–type strip placed along the outer semi–span of the
blades.
Figure 9 depicts the effect of the propeller RPM on the noise spectrum of both a clean
and a modified configuration. The modified configuration corresponds to a #36 grit–type
strip placed along the outer semi–span of the propeller blades. It is observed that the
harmonic of imbalance appears in all the spectra. As depicted in Figures 9b–9c, all the
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
BPF harmonics in the spectra of the modified propeller decreased compared to the clean
configuration by approximately 2–6dB. Of special interest is the first BPF harmonic,
which decreased by 5dB. However, the high–frequency domain in the spectrum of the
modified propeller is characterized with a higher broadband noise, in comparison to the
clean configuration. The broadband content of the signal in high–frequencies masks the
BPF harmonics of the modified propeller. Thus, probably producing ineffective results
of acoustical stealth for low altitude flight vehicles.
It is relevant to note that for 4,000 RPM (see Figure 9a), the BPF harmonics SPL
of the modified propeller are higher than those of the clean configuration; excluding
the first BPF harmonic, which is still lower by almost 5dB (as shown for 7,000 and
8,000 RPM). Presumably, this phenomenon is caused due to the incapability of the #36
grit–type particles to eliminate the LSB and generate boundary layer transition for the
propeller rotating at 4,000 RPM. It appears that the separation location along the upper
surface of the propeller blades is not delayed further downstream. Thus, leading to
a different behavior of loading distribution along the blades, which results in higher
narrow–band tones SPL. Placing larger grit–type particles can generate earlier boundary
layer transition, which may result lower BPF harmonics SPL; as shown for the #30
grit–type particles in Figure 8. Further theoretical parametric research might verify
this conclusion and reveal the equations relating all the physical parameters in question
(RPM, grit distribution along the propeller blades, size of the grit particles and blade
geometrical parameters).
IV. Conclusions
In the current study we investigated the possibility of implementing grit–type bound-
ary layer tripping on a two–bladed UAV propeller (Master Airscrew G/F 3 Series KN
1160 model) operating in the low Reynolds regime (Rec= 2 ·104), for reducing the low–
frequency noise. Noise measurements of the propeller at static conditions (without free
stream velocity) were performed by using four microphones in an anechoic chamber. For
this study, we used SiC grit–type particles with an average diameter of d=0.25mm–
0.56mm [15], which were placed on a 3mm width strip on both the suction and pressure
surfaces of the propeller blades. Following Barlow et al. [15] and Herr [16] we decided
to place the grit strips at a distance of 0.1cfrom the leading–edge of the blade, in order
to maximize the reduction of the low–frequency tonal noise. Different combinations of
grit sizes (#30–#60), strip lengths (full–span and outer semi–span) and propeller RPMs
(4,000–8,000 RPM) were examined to determine their effect on the noise spectrum. For
each modified propeller configuration tested, experiments were repeated twice for valida-
tion.
Results reveal the placement of grit–type particles on the propeller blades increases
the high–frequency broadband noise, resulting it to be higher than the clean propeller
configuration. This might present an evidence for the turbulent boundary layer to suc-
cessfully generated by the grit particles. Although trip tapes were found to reduce the
high–frequency broadband noise of a propeller [14], it appears that by using grit–type
particles one can result higher levels of high–frequency broadband noise. It is shown
that smaller grit particles generate higher broadband noise level; presumably, due to the
generation of smaller turbulent structures. Moreover, the strip length showed no effect
on the high–frequency broadband noise of the modified propeller.
Although high broadband noise levels exist, results for the modified propellers also
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
showed a reduction of the low–frequency BPF harmonics by approximately 2–6dB; where
the highest reduction (6dB) was found at the first BPF harmonic of the acoustic spectra
obtained for the modified propeller with the #30 and #36 grit–type strips. Moreover,
using a full-span strip length resulted in slightly lower BPF harmonic SPL than the
outer semi–span covered blades. This phenomenon could be related to the different lift
distribution generating on the propeller blades, for each strip length used. It is also
shown that smaller grit particles resulted in larger harmonic bandwidths, which might
originated from variation of the turbulent fluctuations that alter the lift coefficient in
time. A reduction of the SPL in the low–frequency range is of high importance for
acoustic stealth. Since the high–frequency content of the signal is attenuated much faster
than the low–frequencies by the atmosphere.
The ability of the grit particles to be effective in reducing the low–frequency tones
was found to be highly dependent on the RPM of the propeller. It appears that for low
propeller RPMs the grit particles might not eliminate the LSB and generate boundary
layer transition. For such cases, the separation location along the upper surface of the
propeller blades might not delayed further downstream, thus resulting a different behavior
of loading distribution along the blades, manifested by higher narrow–band tones SPL.
Further theoretical and experimental studies are required to verify the aforesaid con-
clusions. The insight gained might reveal the physical mechanism relating the different
parameters (RPM, grit distribution along the propeller blades, size of the grit particles
and blade geometrical parameters), thus suggesting an efficient UAV propeller configura-
tion for reducing the low–frequency tones.
Acknowledgments
The authors would like to acknowledge the Marie Sklodowska–Curie Individual Grant
(No. 658846) for financial supporting OS. Moreover, special thanks to the UAV school in
the Israeli Air Force (IAF) for supporting the experiments and the Wind Tunnels Center
in the Israel Aerospace Industries (IAI) for supplying the grit–type particles used in this
study.
References
1. Massey, K. and Gaeta, R., “Noise Measurements of Tactical UAVs,” Proceeding
16th AIAA/CEAS Aeroacoustics Conference, Stockholm, Sweden, 2010.
2. Neuhaus, W., Bretting, H., and Schweizer, B., “Morphologische und funktionelle
Untersuchungen ¨uber den lautlosen Flug der Eule (Strix aluco) im Vergleich zum Flug
der Ente (Anas plathyrhynchos),” Biol Zentr Bl, Vol. 92, 1973, pp. 495–512.
3. Bachmann, T., Kl¨an, S., W., W. B., Klaas, M., Schr¨oder, W., and Wagner, H.,
“Morphometric Characterisation of Wing Feathers of the Barn Owl Tyto Alba Pratincola
and the Pigeon Columba Livia,” Frontiers in Zoology , Vol. 4, No. 23, 2007.
4. Graham, R. R., “The Silent Flight of Owls,” Journal of the Royal Aeronautical
Society, Vol. 38, No. 286, 1934, pp. 837–843.
5. Soderman, P. T., “Leading edge serrations which reduce the noise of low-speed
rotors,” NASA TN D–7371, 1973.
6. Hersh, A. S., Soderman, P. T., and Hayden, R. E., “Investigation of Acoustic
Effects of Leading-Edge Serrations on Airfoils,” Journal of Aircraft, Vol. 11, No. 4, 1974,
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
pp. 197–202.
7. Arndt, R. E. and Nagel, R. T., “Effect of Leading Edge Serrations on Noise
Radiation from a Model Rotor,” AIAA, , No. 72–655, 1972.
8. Oerlemans, S., Schepers, J. G., Guidati, G., and Wagner, S., “Experimental
Demonstration of Wind Turbine Noise Reduction Through Optimized Airfoil Shape and
Trailing–Edge Serrations,” National Aerospace Laboratory NLR–TP–2001–324, 2001.
9. Ning, Z. and Hu, H., “An Experimental Study on the Aerodynamics and Aeroa-
coustic Characteristics of Small Propellers,” Proceeding of 54th AIAA Aerospace Sciences
Meeting, AIAA, San Diego, CA, USA, 2016.
10. McAlpine, A., Nash, E. C., and Lowson, M. V., “On the Generation of Discrete
Frequency Tones by the Flow Around an Aerofoil,” Journal of Sound and Vibration,
Vol. 222, No. 5, 1999, pp. 753–779.
11. Nash, E. C., Lowson, M. V., and McAlpine, A., “Boundary–Layer Instability
Noise on Aerofoils,” Journal of Fluid Mechanics, Vol. 382, 1999, pp. 27–61.
12. Seshagiri, A., Cooper, E., and Traub, L. W., “Effects of Vortex Geneartors on an
Airfoil at Low Reynolds Numbers,” Journal of Aircraft, Vol. 46, No. 1, 2009, pp. 116–122.
13. Oerlemans, S., “Wind Tunnel Aeroacoustic Tests of Six Airfoils for Use on Small
Wind Turbines,” National Renewable Energy Larboratory, Rept. NREL/SR–500–35339,
Golden, CO, USA, 2004.
14. Leslie, A., Wong, K. C., and Auld, D., “Broadband Noise Reduction on a Mini-
UAV Propeller,” Proceeding of 29th AIAA Aeroacoustics Conference , AIAA, Vancouver,
British Columbia, Canada, 2008.
15. Barlow, J. B., Rae, W. H., and Pope, A., Low–Speed Wind Tunnel Testing , Wiley–
Interscience, 1999.
16. Herr, M., “On the Design of Silent Trailing-Edges,” Notes on Numerical Fluid
Mechanics and Multidisciplinary Design, Springer Berlin, Heidelberg, 2008, pp. 430–437.
17. Hubbard, H. H., “Aeroacoustics of Flight Vehicles - Theory and Practice,” NASA
Reference Publication, 1258, 1991.
18. Hanshin, S., Jung-Chun, S., and Soogab, L., “Development of Hybrid Method for
the Prediction of Underwater Propeller Noise,” Journal of Sound and Vibration, Vol. 288,
2005, pp. 345–360.
19. Stuff, R., “Propellerl¨aerm bei Unterschallblattspitzenmachzahlen Umfangskraft
und Axialkraft,” Rept. Dfvlr–Mitt–82–17 Dfvlr, G¨ottingen, West Germany, 1982.
20. Muller, G. and M¨oser, M., Handbook of Engineering Acoustics, Springer–Verlag
Berlin Heidelberg, 2013.
Proceedings of the 57th Israel Annual Conference
on Aerospace Sciences, Tel-Aviv & Haifa, Israel,
March 15-16, 2017
WeL2T4.3
