Proceedings of Global Power and Propulsion Society

ISSN­Nr: 2504­4400
GPPS Xi’an21
18th –20th October, 2021

www.gpps.global

GPPS­TC­2021­0310
NOISE OF FUTURE AIRCRAFT PROPULSION: SOURCES AND DESIGN
CONSIDERATIONS

Georgios Protopapadakis Vasilis Gkoutzamanis

Aristotle University of Thessaloniki Aristotle University of Thessaloniki
gprotopa@auth.gr vgkoutzam@auth.gr
Thessaloniki, Greece Thessaloniki, Greece
Stavros Vouros Konstantinos Kyprianidis Anestis I. Kalfas
Mälardalen University Mälardalen University Aristotle University of Thessaloniki
stavros.vouros@mdh.se konstantinos.kyprianidis@mdh.se akalfas@auth.gr
Västerås, Sweden Västerås, Sweden Thessaloniki, Greece

ABSTRACT
This paper attempts to reduce the gap, when it comes to future aircraft concepts noise and aeroacoustics, and at the
same time introduce basic noise aspects for people unfamiliar to it. This is performed by taking into account the following
four stages: Identification of conventional aircraft noise sources, overview of the aeroacoustics of future aircraft config­
urations, derivation of conceptual design considerations for noise issues, and finally, insight on the impact of technology
advancements on certification. More specifically, a holistic overview of the aeroacoustics of aircraft was performed. The
noise sources were described and made available for those unacquainted with the subject. Aeroacoustic behavior of Contra­
Rotating Open Rotors, Boundary Layer Ingestion and Distributed Propulsion concepts was studied trough the most recent
and relevant literature. The phenomena causing noise were highlighted, and the differences between conventional and fu­
ture designs detected. It was found that generally noise sources are similar, but the intensity of some can be significantly
altered. Early design phase guidelines were extracted for each of the concepts. The design considerations are grouped into
a matrix showing the relative effect of each on noise emissions. With these guidelines, it is possible to perform design
choices in the conceptual level of fidelity for future aircraft concepts. Furthermore, certification aspects and observations
considering its future development are presented. Noise emissions are expected to change, and thus certification must be
adapted to it. Furthermore, in this quick advancing technology field, virtual certification is advised to be used.

INTRODUCTION
The Advisory Council for Aviation Research in Europe (ACARE) has set targets in its “Flightpath 2050” aviation report
(Darecki et al., 2011), with the long­term objective to reduce CO2 , NOX and noise emissions. Multiple stakeholders from
the aviation industry work towards climate neutral aviation. An example, amongst many, is the Clean Aviation Partnership,
discussing electric, hybrid­electric and hydrogen possibilities for the future of aviation (CleanAviation, 2020). To date,
many concepts, based on the aforementioned technologies, are under assessment and expected to bring many new features
and differences in operation, with the main objective to reach the environmental and socio­economic goals. Consequently,
it is considered useful to account for noise emissions as a criterion early in the design process.
In this work, attention is given mainly changes that propulsion electrification may bring to noise emissions. An interest­
ing review of different concepts and ongoing research for commuter, short­ and long­haul aircraft employing electrification
can be found in (Gkoutzamanis et al., 2021) and (Graham et al., 2014), respectively. In literature there are many works de­
scribing aeroacoustics and noise for conventional aircraft (Filippone, 2014; Casalino et al., 2008). Information on noise for
future design is very limited. Noise sources relative importance and intensities are expected to be modified. The motivation
of this work is to close the gap of knowledge in that field, by reviewing the new landscape of noise emissions and providing
design considerations for new concepts. Consequently, a literature review on aircraft noise sources is presented. Further
deepening in Boundary Layer Ingestion (BLI), Distributed Propulsion (DP) and Contra­Rotating Open Rotors (CROR) is
performed. Both sections are highlighting important phenomena causing noise, and the differences between conventional
and future designs. Finally, reflections on design considerations and noise certification are presented.

This work is licensed under Attribution­NonCommercial­NoDerivatives 4.0 International (CC­BY­NC­ND)

See: https://creativecommons.org/licenses/by­nc­nd/4.0/legalcode
SOURCES OF AERODYNAMIC NOISE
Aircraft noise can be grouped into categories. Diverse classification structures have been adopted so far. For example
Bertsch et al. (2019) identify the airframe, the engine and the integration related noise, while Filippone (2014) follows a
same approach but distinguishes thermal engine and propeller noise. On the other hand, Casalino et al. (2008) mention that
the fan and jet noise are much more significant than core. All of these review works cover into a great extend all relative
aircraft noise sources. The following classification is deployed in this work: airframe noise, propulsion noise. Figure 1
summarizes all the noise sources that will be presented bellow.

Airframe noise
For the first time in late 60’s, the airframe noise came into light (Hubbard, 1991). It is mostly significant during
the approach phase of the flight(Hileman et al., 2007), since the engine is operating at a lower power setting, reducing the
emitted noise, and all landing gears are extended. Noise produced by landing gear and high­lift devices are mainly produced
due to aerodynamic interactions.
All generating mechanisms for airframe are identified in one of the first reviews by NASA on aeroacoustics of flight
vehicles in early 90’s (Hubbard, 1991). A more recent review by Casalino et al. (2008), has also defined the most consider­
able airframe sources. Namely there is the trailing­edge, flap side­edge, under­carriage gear noises, cavity noise, turbulent
boundary layer, panel vibration and powered lift configuration noises. Both works agree on the noise emitting methods,
leading to the conclusion that few things have changed in that period of approximately 20 years.
Trailing­edge noise is produced by three phenomena. The first of which takes place when unsteady, turbulent, flow
interacts with the trailing edge of the airfoil. There is also noise caused by vortex shedding and finally aeroacoustic feedback
loop (Doolan et al., 2012). However, more research is needed to fully understand the feedback loop mechanism (Doolan
et al., 2012). Broadband noise is generated by the first mechanism, while the other two produce tonal noise.
Flap side­edge noise is generated by fluctuations that are induced by the free­shear layer at the flap bottom edge but
also vortices generated on the tip of the airfoil and merge downstream to produce acoustic disturbances (Filippone, 2014;
Kolb et al., 2020).
As all high­lift devices, vortex shedding and cavity noise are some of the slat noise sources (Filippone, 2014). Accord­
ing to Knacke and Thiele (2013), who did a numerical investigation, the noise principally occurs from the interaction of
turbulence with the upper slat trailing edge. The impingement of the shear layer on the pressure side surface is, according
to Terracol et al. (2016), the main noise generation mechanisms in slat cavities.
The complex geometry of landing gear impedes the evaluation of noise. The broadband noise is caused by turbulent
flow separation and the interactions of the turbulent wake with other components downstream. Tonal noise is mainly
generated by open ended pin holes and flow­excited cavities on the landing gear (Molin, 2019).

Propulsion system noise

Many components compose an engine, and many of them can produce noise in different ways. Such components are
the fan, the compressors, the combustor, the turbines and the nozzles. For each of these components an introduction to their
noise mechanisms is presented. Note that for a turboprop engine, instead of the fan, the propeller noise must be studied.
Since high bypass ratio (HBR) engines came to the foreground, the fan became one of the major sources of the engine
noise (Krejsa and Stone, 2014). The fan noise can be perceived either at the inlet of the engine or at the bypass nozzle outlet
(Dahl, 2012). The first cause is the aerodynamic loading of rigid surfaces, fan or stator blades, and it is a discrete tone.
Broadband noise is generated due to flow turbulence, or rotor wake turbulence, flow fluctuations in other words, impinging
on blades downstream. The third noise source is a Multiple Pure Tone noise (or buzz­saw) of a pressure spectrum that is
produced due to variation of shock strengths, at the supersonic tip speed of the fan (Krejsa and Stone, 2014).
Although the propeller is also used for propulsion, it has different design features. For propellers, other than the loading
noise that is present to the fan, there is also the thickness noise, which is corresponding to the displacement of a volume of
air caused by the motion of the blade (Hubbard, 1991). These are known as the tonal noises of a propeller. Non­linear and
unsteady noise can be produced in the rare case when tip speeds are transonic and the propeller is set with a certain angle
of attack (Hubbard, 1991). Broadband noise, as for the fan, is caused by the interaction with turbulent inflow and due to
turbulent boundary layer which leads to fluctuating load at the trailing edge.
Two types of noise mechanisms originate from a combustor, the direct and indirect noise, both caused by the unsteady
combustion flame (Dahl, 2012). The first is generated due to expansion of the gas volume through unsteady heat release, and
flow velocities that are perturbated and become turbulent. The transportation of non­uniformities of entropy (temperature)
through strong pressure gradients generate the indirect noise. Entropy waves are produced and propagate through the turbine
and nozzle to result into noise. Generally, indirect noise dominates the direct combustion noise, however they both have
the same basic frequency, but often with different spectral shapes.
The turbine contributes mainly with the low­pressure stages to the far­field noise (Dahl, 2012). The turbine noise is
noticed during approach conditions and low power, as a result of low jet and fan noise (Ali et al., 2016). The turbine,
has two kinds of noises, the broadband and discrete tones. The tones are produced due to interactions of rotor wakes that
meet downstream rotors and stators. The noise propagates away from the generating blade or vane row, with the flow

2
 Figure 1 Aircraft noise sources

having a helical spinning configuration (Dahl, 2012). Tones, as for compressor and fans, issue for harmonics of blade
passage frequencies (Barbosa and Dezan, 2013). The broadband noise on the other hand, is caused by the oscillating load
on the rotors and stators. This noise is related with random unsteady disturbance and turbulence within the flow (Barbosa
and Dezan, 2013). Such fluctuations can be the blade vortices and wakes, the boundary layer scattering but also geometric
asymmetry and perturbations. Someone can correlate this kind of noise with the airframe noise generation. Furthermore, the
broadband noise is dominated by a phenomenon so called haystacking. Haystacking is a noise generated by flow turbulence
causing lift fluctuations and scattering of the turbine tones into sideband noise in the external exhaust shear layers (Dahl,
2012). In other means, it is the redistribution of acoustic power caused by tonal noise into broadband noise. Other sources
can be present, such as instabilities in the shroud exit cavity, which mostly affect low frequencies (Rebholz et al., 2016).
For compressor rotors the loading and thickness noise are recognized as major noise sources (Tyler and Sofrin, 1962).
Other sources emerging from vortex shedding, unsteady loading and blade vibrations are present but less important. As
for the rotor­stator interaction, three sources are present: cutting of stator wake by the rotors, impingement of rotor wakes
dowstream and distortion of rotating pressure fields due caused by the proximity of reflecting objects. However, noise
is also caused by fan / propeller wake arriving at core inlet guide vanes (Agarwal et al., 2004). Nonetheless, part of the
aforementioned sources are expected to decay or to be cutoff from compressor design, and thus not propagating towards
the far­field (Tyler and Sofrin, 1962; Agarwal et al., 2004).
Jet noise is generated when part of the flow energy is converted to random velocity fluctuations and/or dissipated to
heat due to viscous effects (Dahl, 2012). Jet noise of a turbofan engine contributes nearly half of the overall acoustic energy
during take­off. Mixing, shock and screech noises are produced in the nozzle due to the jet, as reported in (Dahl, 2012).
Screech noises, according to the same literature, are not existent anymore in recently developed engines. Shock noise,
which is a broadband type noise, is present at supersonic nozzles and caused by the interaction of turbulent boundary layer
and the shock waves (Barbosa and Dezan, 2013). Sometimes it can be present to subsonic nozzles when the jet is not fully
expanded. Finally, mixing noise, also a broadband type, is generated by small­scale turbulence and large­scale eddies. More
dominant is the sound originating from the small­scale turbulence, but with the eddies having a role in turbulence mixing and
jet spreading. Mixing noise has three sub­categories, Large, Small and Transitional scale mixing noise. Large scale noise
is the product of large­scale turbulent structures, low frequency noise, while small­scale is due to small turbulent mixing
in the initial shear layer, high frequency noise. Transitional noise, is middle­to­high frequency noise and very difficult to
characterize. More information about jet noise can be found in (Smith, 1989).
Up to date, engines were usually positioned under the wing such that they have free­stream, uniform, intake flow. This
greatly reduced the turbulence at the intake, and thus turbulence related noise and aerodynamic behaviour deterioration.
However, novel technologies are in the horizon, with the aim to increase propulsive efficiency, and at the same time obtain
a reduction in noise emissions. The future preserves a more integrated airframe­engine configuration in order to take
advantage of some phenomena, such as boundary layer ingestion, and shielding. That is further discussed in more details
in the following section, focusing on the propulsion system, and more specifically, the thrust generating fan/propeller.

AEROACOUSTICS OF FUTURE CONCEPTS

Electrification brings many opportunities in aircraft noise reduction. A preliminary noise assessment using empirical
correlations for electric motor engines was performed by Huff et al. (2016), and showed that electric engine is quieter than

3
the other propulsion noise sources. More specifically, motors of 1 MW and 13.8 MW can be 8 to 20 and 17 to 29 dB SPL
equivalently quieter than the fan they drive, respectively. The same trend is identified also by Riboldi et al. (2020), who
proposed a source­blending method making use of existing noise models.
Of course, with the technological state of batteries, electric aircraft have increased Operating Empty Weight (OEW),
and thus, necessitate larger propellers or fans to produce the increased amount of thrust needed. This implies also higher
noise emissions, since a bigger propulsion system is needed. In addition, larger high­lift devices will be necessary for
a heavier aircraft. Overall, this induces larger noise issuance (Riboldi et al., 2020). However, electrification gives more
freedom in designing concepts that make use of innovative technologies for increasing propulsive efficiency and reducing
carbon footprint. Fan and propeller are usually the dominant noise sources, and thus the following section concentrates
in fan/propeller related concepts (Bertsch et al., 2019). Such are BLI, DP, while CROR could be also of interest for an
implementation with electric motors.

Boundary Layer Ingestion

The logic behind the boundary layer ingestion configuration is wake filling. Accelerating the flow near the trailing edge,
in case of a wing, or the end of fuselage, will conduct to a more uniform velocity and thus reducing the flow dissipation losses
(Steiner et al., 2012). That said, the ingestion of the boundary layer, and hence inlet flow distortion and non­uniformities,
is decreasing the efficiency of turbomachinery components (Romani et al., 2020). A benefit is the detracted wetted area,
and thus drag, by removing the nacelle and the pylon of typical wing mounted engines. Hendricks has cited also changes in
inlet total pressure recovery, in fuselage flow characteristics due to suction and in engine inlet flow (Hendricks, 2018). A
reduced inlet velocity is detected, leading to a lower acceleration required for a given thrust. Overall, a trade off between
external airplane aerodynamics and turbomachinery efficiency must take place. Generally, it is a promising configuration
with merits for further investigation.
It is known from jets, that noise is mainly produced from mixing of flows of different velocities, see section 2. So one
should expect a noise reduction with the use of BLI. However, the boundary layer disturbs the inflow of the turbomachinery
which is no longer ingesting a uniform free stream flow. Larger unsteadiness is now producing phenomena that must be
accounted as noise sources. Therefore, this section is trying to answer the question if a boundary layer ingestion system
will lead to a more elevated noise or to a quieter aircraft.
Defoe and Spakovszky (2013) studied an S­shaped (serpentine) inlet boundary layer ingestion fan. Attention was given
at shock rotor noise generation and propagation. The mechanism affecting the shock noise and its amplitude for different
boundary layer characteristic was studied, giving special care to vortex generating methods. A Reynolds Averaged Navier­
Stokes (RANS) informed body force, in an unsteady full­domain Euler solver was used for near­field (Defoe et al., 2010).
For far­field, the integral Ffowcs­Williams Hawkings (FW­H) method is used with permeable surfaces. Through the study,
co­swirling and counter­swirling vortices were detected, which can lead to changes in sound generation and propagation.
Altough the noise generation is slightly affected, an attenuation of sound power for higher frequencies is expected by the
former, while the opposite effect is detected from the latter. The serpentine geometry is of great importance for both vortex
generation and boundary layer characteristics, which affect the fan noise level. Overall the quantified extra shock noise
produced due to boundary layer ingestion is from 13 to 18 dB of far­field OASPL.
Schwartz et al. (2019) studied a conventional fuselage BLI propulsion. The tonal noise sources were considered in
this work, namely the rotor­alone tones, the rotor­stator interaction tones and rotor­inflow interaction tones. For rotor­
alone and rotor­inflow interaction, a solver based on Goldstein formulation of FW­H was used, while for the rotor­stator
interaction the V072 from NASA along with a multimode radiation model (Topol and Mathews, 2010). The results show
that all configurations are noisier than the conventional under­wing setup. Note that according to the results of Schwartz
et al., when the flow is axisymmetric, the dominating source is rotor­stator noise and when there is no axisymmetry, the
rotor­inflow noise is dominating the first and second Blade Passing Frequency (BPF). Furthermore, for non­axisymmetric
flows, the first and second BPF are overtoping by far the third tone.
Clark et al. (2018) work was on NASA’s BLI D8 double­bubble concept . Two phenomena are studied, the ingestion
of turbulence and the non­uniform flow intake. Turbulence ingestion, according to Clark et al. (2018), can produce tonal
components caused by blade passage through turbulence. At the same time, random fluctuations caused by turbulence
generate unsteady loading leading to broadband noise. On the other hand, boundary layer profile induce a tonal component
by the blade passing one per revolution through a non­uniform flow. The aforementioned noise sources are modelled within
the known tool Aircraft NOise Prediction Program (ANOPP) (Lopes and Burley, 2016). Results show that BLI increases
noise up to 16.1 dB of cumulative EPNL, with a stronger impact on sideline noise (94.2 instead of 87.6 EPNdB).
Investigation of Rotor­stator interaction tonal noise was conducted by Petrosino et al. (2021). First, a 3D RANS
simulation was examined. Necessary information is extracted, such as turbulent flow quantities, and then the aerodynamic
excitation is reconstructed. The noise is computed based on analytical theory utilizing circumferential modes in frequency
domain. The propagation of the modes in the engine is modelled, while the coupling with the far­field propagation is done
via acoustic pressure boundary conditions. There was a free­field and a scattering propagation model based on Boundary
Element Method. Finally, the ground propagation is based on Quasi­Static Acoustic Approach. All methods used are cited

4
by Petrosino et al. in their work. The results showed generally lower noise emitted due to BLI. However, higher frequencies
are more pronounced in the BLI compared to free­field. Furthermore, it is shown that scattering effect mainly affects the
higher frequencies, and thus more dominant for BLI. Overall, BLI is quieter but due to scattering losses some of the benefits.
It is important to highlight the fact that the fan tested was initially designed for BLI.
In another paper of Spakovszky (2019), the BLI effect on a Hybrid­wing body is studied. This concept is a lot different
compared to the others, since the fuselage is a hybrid­wing body with integrated engines at the back. Similar to the problem
of Defoe presented previously, a RANS informed body force, in an unsteady full­domain Euler solver was used for near­
field. For far­field the integral FW­H method is used with permeable surfaces. According to the Spakovszky (2019),
Presence of turbulence at the inlet has a more pronounced effect on shock strength, compared to stagger angle, rising the
amplitude of shaft­order components. High frequencies, more than 11 times the shaft rpm, are probably cut off frequencies
since they are not present at the serpentine inlet, as mentioned also by Defoe and Spakovszky (2013). Lower frequencies
are more elevated though, leading to a 7 dB increase in average linear OASPL and to a 3 dB increase of the A­weighted
OASPL. Note that the hybrid­wing body configuration permits more noise shielding that should be studied.
Finally, Romani et al. (2020) studied the effect of boundary layer ingestion. Lattice Boltzmann and Very Large Eddy
Simulation approach was used to assess near­field noise and FW­H for far­field. It is another different concept, with the
engine mounted on the side of the fuselage, towards the tail. This work covers the aerodynamics and noise sources into
great extent for this configuration. Increased turbulence and non­axisymmetric inlet flow caused by the ingestion of the
boundary layer cause a dominating rotor­stator noise. This occurs by the unsteady loading of the stator, due to periodic
fluctuations of the wake of the rotor provoked by the boundary layer. Furthermore, broadband dominates the BLI noise and
tones are masked, results show. Overall, the BLI is 4 EPNdB (front side) to 18 EPNdB (aft side) noisier than the isolated
engine. A peak of 108 EPNdB compared to 98 EPNdB for the isolated is observed at takeoff flight path.
In table 1 one can find works that recently tried to assess boundary layer ingestion fan noise. From the aforementioned,
although the magnitude cannot be quantified since it can vary from application to application, it can be concluded that the
BLI is increasing the noise generated by a given engine. This is expected, considering that turbulence generally leads to an

Authors Methods Integration type Effect

Romani LBM and VLES near field and Faras­ Conventional fuselage, From 4 EPNdB (front side) to 18 EP­
et al., 2020 sat 1A for far­field acoustics exterior. NdB (aft side) increase in noise.
Clark et al., Empirical and semi­empirical Double bubble fuselage Broadband and tone effects included,
2018 ANOPP: Modelling the tonal and with the engine on top of levels are increased by 16.1 EPNdB
broadband noise due to turbulence it. above the baseline configuration (3.5
and tonal noise due to different dB at approach, 6.6 dB at sideline and
loading once per revolution. 6 dB at cutback).
Schwartz Rotor­only is modelled with Gold­ Four configurations Classification in terms of SPL emis­
et al., 2019 stein formulation of FW­H. The shown of fuselage sions: conventional < no wing BLI
rotor­stator interaction is modelled propulsion: a conven­ < high wing (no scoop) BLI < low
with Nasa V072 cut­on/off modes tional fuselage with wing BLI < high­wing (scoop) BLI.
with semi­empirical method and fi­ high­wing / low­wing / A difference of up to 41 dB SPL is
nally the inflow interaction with an no wing / conventional detected at one meter from the source
in­house FW­H to account for BLI. turbofan. for the first BPF.
Petrosino Using aerodynamic excitation and The engine is positioned BLI gives generally lower noise
et al., 2021 based on analytical theory utilizing at the back of the fuse­ (SPL) at low frequencies compared
circumferential modes in frequency lage of a conventional to free stream (fan designed for BLI).
domain, the prediction of rotor­stator aircraft with a T­tail. Installation and scattering is more
interaction is performed. pronounced for high frequencies.
Spakovszky, Multiple pure tone noise is studied. Hybrid­wing body with High frequencies (11*shaft revs per
2019 RANS informed body force imple­ integrated engines at the second) are cutoff at the inlet, but
mented in full­domain unsteady Eu­ back with serpentine in­ lower frequencies are elevated lead­
ler and the far field (outside of the let. ing to a 7 dB increase in average lin­
canal) via FW­H integral method us­ ear OASPL and to a 3 dB increase of
ing permeable surfaces. the A­weighted OASPL.
Defoe and RANS informed body force imple­ No specific position, S­ Rotor shock noise can produce from
Spakovszky, mented in full­domain unsteady Eu­ shaped inlet BLI. 13 up to 18 dB more OASPL at the
2013 ler and the far field (outside of the far field when BLI. Strongly affected
canal) via FW­H integral method us­ by vortex lift­of.
ing permeable surfaces.

Table 1 Boundary Layer Ingestion works

5
 Figure 2 BLI main noise sources

increase of a propeller or fan broadband noise (Sharpf and Mueller, 1995). The shock related noise is significant, while the
Rotor­Stator interaction is the dominating noise for subsonic fans. The main noise sources of BLI can be seen also in Figure
2. However, if a fan is specially designed for BLI, when it is exposed to free stream it can produce more noise (Petrosino
et al., 2021). Even when the fan is specially designed for BLI, special attention has to be given to the integration, either for
intake geometry or position for scattering effect.

Distributed Propulsion
Electric engines require less maintenance and have better scaling capabilities than jet engines. Therefore, it is easier to
distribute the propulsion to multiple small engines. They can be installed as under­wing propellers, embedded in the wing,
or installed for BLI at the trailing edge etc. (Wick et al., 2015). Wick et al. (2015) reported a 8% efficiency improvement
in transonic conditions, compared to conventional under wing engines with an equivalent propulsive area. However, the
fundamental benefit of distributed propulsion, is the possibility to increase the equivalent bypass ratio. This allows a higher
mass flow, leading to a reduced flow acceleration required, and thus lower tip speeds. According to Berton and Nark (2019),
electric engines provide maximum power to a wider range of rotating speeds, and coupled with variable pitch, it will allow
for lower tip speeds, and thus, quieter takeoff conditions. Overall, a diminution of emitted noise by the propulsion system
is expected (Synodinos et al., 2018a).
Synodinos et al. (2018a) studied the noise emissions for an A320 aircraft configured with multiple electrically driven
fans. The noise assessment was done following a previous work, implementing Heidmann’s work for the fan, Fink’s for
the airframe and Lighthill acoustic analogy for the jet (Synodinos et al., 2018a,b). The drag difference occurring from the
number and size of the engines was considered negligible and was ignored. The weight of the aircraft, depending on the
propulsion system was taken into account. It was concluded that increasing the size of propulsor array, and thus the number
of propulsors, reduces the total weight of aircraft. Weight is important since it affects the flap deflection needed, and thus, the
approach noise. The same tendency can be observed in figure 6. Νoise power level was calculated approximately 2dB lower
at the source when distributed propulsion was implemented, according to Synodinos et al. (2018a). Noise­power­distance
(NPD) curves show significant changes with the number, and thus size, of propulsors. Smaller fans will spin at higher rpm,
and the tonal noises will be in different spectral range. Note that higher frequencies dissipate faster in the atmosphere. Thus,
Increasing the number of propulsors to 6 seems to reduce up to 2km the range of the 90 dB sound exposure level contours
(or NPD) aft to the takeoff region runway for the turboelectric configuration compared to the conventional A320. Note that
higher fidelity simulations could be used to verify these results.
Moore and Ning (2019) addressed differently the distributed propulsion concept. The main question they tried to
answer is how the takeoff distance and noise are associated with the number of propellers and some of their geometric
characteristics such as blade twist, chord, radius and number of blades. The propeller aerodynamics were studied with
a blade element momentum theory method code CCBlade (Ning, 2014), with airfoil data given from XFOIL. The stall
delay was computed with Airfoilpreppy (Ning, 2013). The assessment was validated with experimental data, showing good
agreement. The noise was computed using the semi­empirical BPM equations (Brooks and Marcolini, 1986). The coupling
of the propeller to the wing and the aerodynamics were computed with vortex lattice method. The design parameters of
the propellers are the number of blades, radius, twist, blade chord. The airfoil profile is kept constant. The 16 propeller
configuration showed the best result, reducing to half the takeoff distance. However, when the noise was constrained, the 8
propeller configuration showed the optimum takeoff distance with a reduction of 43% compared to conventional two prop
case. Furthermore, according to the results presented in the work of Moore and Ning (2019), the following conclusion can
be drawn: the 2 propeller configuration will always be noisier than the 4 or 8 propeller for a given thrust to weight ratio.
However, results are questionable since the work focused more on aerodynamics for takeoff distance reduction and only a
low fidelity semi­empirical model was used for noise assessment.
Bernardini et al. (2020) have studied the effect of interaction between rotors of a distributed propulsion array. Par­
ticularly, the distance between consecutive rotors, with specific geometry and a radius of 1.48 meters, was considered.
Boundary element method was used for the aerodynamic assessment of the propeller providing input to a Farassat 1A
boundary integral formulation of the FW­H. Results show significant aerodynamic interaction up to a distance (tip­to­tip)

6
of one diameter, however, noise is affected only for tip distance below a quarter of a diameter. The loading and thickness
noise components were considered for this work. For two adjacent propellers, a new loading noise component emerge, with
a frequency of 1/rev. Furthermore, the rotating direction determines the directivity of noise in the plane perpendicular (ver­
tical) to propellers; larger noise values are present in the forward direction. Finally, higher harmonics are mostly affected,
with an increase of up to 20dB for the 3*BPF.
Rizzi et al. (2017) did an experimental research on annoyance of distributed propulsion for community noise. A
series of 5­blade propellers, having the same size and not overlapping, is considered. The number of propellers is a design
parameter (6, 12, 18) and determines the diameter (0.678, 0.339, 0.226 meters) of the propeller for a given wingspan. All
propellers, for all diameters, are designed for low­tip speed of 137 m/s for noise reduction purposes, leading to a given
rotational speed (1931, 3863,5794 RPM). Another parameter considered, is frequency difference between propellers, with
a range of 5 Hz in BPF. The propellers were aerodynamically assessed with XROTOR, surface pressures extracted from
PAS and the results were then used for noise calculation with ANOPP with a Farassat 1A FW­H solver (Lopes and Burley,
2016; Farassat, 2007). Note than no interaction noises were taken into account, only multiple “single” propellers were
considered. The results are given in hemispheres and show that a change of 1Hz in frequency between adjacent propellers
can give a significantly different sound pressure radiation pattern. A 10 dB decrease of SPL along the center­line is observed,
however other regions experience higher SPLs. For both cases, the total value of sound power is the same when integrated
spatially. For the spread frequency design, a fixed frequency difference of 1 Hz is selected. An experiment took place
with the auralization of the previous results, considering the number of propellers and the frequency difference. The motor
controller error and turbulence was take also into account for the auralization. NASA personnel served as test subjects
and had to rate from “Not At All” annoyed to “Extremely” annoyed. The results of the experiment showed an annoyance
increases with the number of propellers, as the total radiated sound power and BPF increase. Something expected, according
to Rizzi et al. (2017), since the tip speed is held constant and is independent from the number of propulsors. Finally, the
outcome of the experiments shows that there is no change in annoyance from spreading the frequency.
Zhou et al. did an experimental research on rotor­rotor interactions for two abutting counter rotating propellers (Zhou
et al., 2017). More specifically, small propellers of 0.24 meters of diameter (D), 3 Newton thrust and 4860 RPM were tested
in laboratory for various tip­to­tip gaps. Noise was measured in an anechoic chamber with five microphones placed every
30°, starting from the front, at a distance equal to 6D. For the single propeller, the results showed maximum SPL at front
and aft directions and minimum noise at 90°. Furthermore, results of 120°and 60°were very close, a behavior similar to
dipole distribution, expected for propellers (Zhou et al., 2017). Once the two propellers were positioned within 0.2D an

Authors Methods Integration type Effect

Synodinos The noise prediction methods em­ No aerodynamic instal­ Results are presented in noise–
et al., 2018a ployed are the semi­empirical (Syn­ lation effect considered. power–distance curves (NPD), and
odinos et al., 2018b). seem to be quieter with up to 2 km
reduction of 90dB NPD.
Moore and Blade element momentum theory No aerodynamic instal­ The configurations giving the short­
Ning, 2019 coupled with Brooks, Pope and Mar­ lation effects consid­ est takeoff length for the lower possi­
colini (BPM) equations. ered. ble noise is the one with 8 propulsors.
Bernardini Boundary element method coupled No aerodynamic instal­ For distances smaller than 1 diameter
et al., 2020 with FW­H Farassat A1 formulation. lation effects consid­ the noise directivity patterns are sig­
ered. nificantly affected.
Rizzi et al., Aerodynamics pressures from XRO­ Isolated propellers, no Small changes of rpm across the pro­
2017 TOR feed the ANOPP FW­H Faras­ propeller to propeller in­ peller array can alter the numerical
sat A1 formulation solver. Results teraction noise consid­ results but is not perceived by the
are later used for auralization and ex­ ered. subjects of the experiment. If fixed
periments. tip speeds, a larger number of pro­
pellers produce more noise.
Zhou et al., Experimental in anehoic chamber, Tested isolated and Interaction can increase noise in all
2017 PIV and Stereoscopic PIV, and Dy­ two counter rotating azimuthal angle and produce thrust
namic loading tests. propellers. fluctuations and turbulent kinetic en­
ergy increase.
Alvarez Compared URANS and VPM cou­ Tested isolated and Interaction of tip­to­tip distance, is
et al., 2020 pled with FW­H and(BPM) equa­ two counter rotating showing decrease in noise with in­
tions. Validated against experiments propellers. creasing distance. The same applies
of Zhou et al.(Zhou et al., 2017). for downstream gap, until the expan­
sion of stream tubes.

Table 2 Distributed propulsion works

7
increase in noise was observed. For a distance equal to 0.05D an average increase of 3 dB SPL (in the range of 68 to 75
dB) is measured across the azimuthal angle (0<β <120°). The spectrum analysis indicates an augmentation in all spectrum
and especially in BPF harmonics. However, dynamic load measurements also took place, showing only 2% reduction in
normalized thrust coefficient. There were also thrust fluctuation leading to a normalized thrust standard deviation 2.5 times
larger at a distance of 0.05D, compared to single rotor. In addition, free­run PIV and Stereoscopic PIV revealed higher
turbulent kinetic energy which implies also an increase of broadband noise. Overall, as mentioned in previous paragraphs,
reducing the distance provokes disturbances and interactions that lead to reduced efficiency and higher noise emissions.
Alvarez et al. studied the tip distance, downstream spacing and tip loading for two adjacent propellers (Alvarez et al.,
2020). The propeller itself consist of two blades of approximately 10 inches diameter and a tip Mach number of 0.2. URANS
and Vortex Particle Method (VPM) were coupled with FW­H solver and BPM equations for broadband noise. The URANS
setup had coarse discretization in both space and time, with 14 millions cells and 3°of rotation per time step with a total of
10 revolutions. VPM is a meshless solver, and its time discretization was 5°per time step and a total of five revolutions.
Results of the two methods were tested and showed similar behavior, with VPM having the potential to capture some higher
frequency interactions. Although, the results maybe caused by numerical noise according to Alvarez et al. (Alvarez et al.,
2020). In terms of acoustics results, as mentioned in previous paragraphs, bringing closer the tips of the propellers increases
the noise, especially in upstream and downstream of the rotors. The same response is perceived for downstream distance.
However, the gap must be carefully selected to be in the stream tube contraction region. Further downstream, the flow
will start to expand due to leapfrogging and tip­vortices breakdown, increasing the interaction between the rotors. A 12 dB
A­weighted SPL increase was calculated when taking into account the interactions, compared to interaction­free studies,
indicating the importance of interaction noise.
In table 2 one can find the list of the aforementioned works. The majority state that multiple small propulsors generally
reduces noise compared to bigger engines. However, propeller interaction must be carefully studied. Installation spacing
was found to be important, nonetheless aspects such as frequency differences etc., that were not extensively covered in this
work, should be considered. If spacing is correctly selected, tip­to­tip distance and downstream gap as shown in figure
3, a noise reduction can be achieved. As an extra comment from the authors: the ducted e­fans could help reduce noise,
by reducing interaction and increasing shielding, however the trade off with potential weight and drag increase should be
further studied.

Contra­Rotating Open Rotors

Contra­Rotating Open Rotors (CROR) came to the surface for the first time in the 70s and 80s. The reason was the
expected increased efficiency (Van Zante, 2015). CROR engines consist of two contra­rotating fans mounted on the outside
of the nacelle. They can be driven either from a gearbox or directly from contra­rotating power turbines. As the flow passes
through the first fan, velocity components in the tangential direction appear. Passing through the second fan, which rotates
in the opposite direction, swirling is reduced and the flow becomes more axial. The energy and velocity of the flow is thus
better exploited. As a result, it has higher propulsive efficiency and therefore lower fuel consumption and reduced gaseous
emissions. However, the simultaneous appearance of strong rotor­rotor interactions and noise shielding reduction derived
from the nacelle removal, makes CROR noisier than turbofans (Kyprianidis et al., 2015). This is what actually blocked the
research back in the 80s. Recently, CROR were brought back to life, and it is considered useful to briefly cover the topic.
Although it would be interesting to cover noise sources and modelling techniques for CROR, they were covered in the
recent PhD thesis of Smith (2020), and therefore is not covered in this section. However, review of contemporary studies
covering CROR noise is discussed.
In the paper of Fenyvesi et al. (2019), an experimental investigation of pylon­engine interaction was conducted, com­
paring the free and the integrated (pusher) configuration. The acoustic measurements were performed and processed with a

(a) (b)
Figure 3 Distributed propulsion additional source and important parameters: (a) Front face (b) Side face

8
Delay­and­sum beamforming in the frequency domain (Mueller, 2002). The interaction and BPF tones at Mach radius are
the coherent rotating noises identified, while blade­wake interaction is considered a stationary coherent noise. Shaft order,
or one­per­revolution, and rotating broadband are the incoherent rotating noises. Comparing uninstalled and installed en­
gines, major differences were distinguished. More specifically, for the uninstalled, all first five BPF are dictated by coherent
rotating noise (interaction and BPF tones), while for the installed, four out of five are dominated by the blade­wake inter­
action coherent stationary noise. As for the broadband noise, only small effects are detected when the pylon is considered.
Overall, the wake­blade interaction of the first rotor is more pronounced than aft rotor wake­blade interaction, since the
turbulence of the wake from the pylon decays with distance.
In continuation to the previous paragraph, the typical noise sources considered for broadband noise are described.
Horváth (2015) cited three major mechanisms. First is the trailing­edge, that can be found also as rotor or boundary layer
self­noise. Trailing­edge noise is generated by turbulent boundary layer passing through the trailing­edge and causing
loading fluctuations. Another source is tip noise, or vortex self­noise, occurring by the interaction of the tip vortex with the
trailing edge. Last, the leading­edge noise produced by the inflow turbulence interacting with the blade.
Kennedy et al. (2018) did an experimental parametric study of CROR noise . The integration was the main subject,
and the parameters included were: the different kind of tails, engine pylon elongation, engine pylon installation angle,
and variable wing to engine distance. Tractor and pusher configurations were considered, for different angle of attack and
velocities. The CROR itself consisted of 12 blades for both front and aft rotor, spinning at the same rotating speed. A
series of microphones is used for near­field, around the front blade, and far­field measurements, at top, lateral and front
positions. For the pusher configuration, the approach is dominated by the 2*BPF noise. Increasing velocity makes the
BPF more pronounced, and sometimes stronger than 2*BPF. On the other hand, for takeoff, BPF is clearly more distinct
compared to approach, with implies that the interaction with pylon wake is more energetic. This trend is not reproduced for
the tractor configuration, which according to the author, the results are closer to isolated engines but also further confirms
the connection of strong BPF tone with pylon wake at the pusher configuration. A tonal noise reduction is possible with
elongated and with larger pylon angle (with respect to wings) leading a significant decrease in 2*BPF but a small increase
in BPF for approach. Exactly the opposite applies for takeoff. Note that at the majority of directivities, the 2*BPF is the
dominating noise. The same applies for tractor configurations, elongated pylons and at higher angles, lead to a reduction
of the dominant tones of 2*BPF, while there is a negligible increase in the BPF tones. For takeoff it is different, since
4*BPF is also important, and there is no clear indication for overall noise reduction. Finally, there is broadband noise also,
which was isolated by removing the tones with a peak­finding algorithm. For pusher, the broadband levels are lower for the
takeoff than for approach, with the exception of elongated pylon with low angles. For tractor, there is no significant impact
either on noise emission of the CROR, with effects less than 1 dB across almost all configurations and emission angles. The
conclusion is that the pusher configuration with elongated and higher angle pylon, gives the best overall noise reduction.
In terms of numerical parametric works, Kopiev et al. (2017) studied the community noise of CROR engines. The
methodology followed was high fidelity simulations. More specifically, hybrid RANS – DDES (Delayed Detached Eddy
Simulations) simulations were used to compute near field noise and FW­H for far field propagation (Shur et al., 2015).
The design parameters considered for this comparison study were the number of blades, the aft rotor diameter, the distance
between the rotors and blade chord width. The initial configuration was an 8 front and 7 aft blades of the same diameter.
Results show that the most effective way to reduce EPNdB noise for flyovers is increasing the number of aft rotor from
7 to 9 blades, keeping thrust constant. Reducing aft rotor diameter also showed a reduction in noise, while rotor distance
issues a reduced noise in both directions. According to Kopiev et al., although not optimized, this configuration is better
compared to the initial turbofan configuration, stating however that airframe interaction is not taken into account.
Kritikos et al. (2012) also investigated numerically an isolated CROR. Specifically a non­linear harmonic approach
for unsteady solution was used (Vilmin et al., 2006). Then FW­H approach was used for calculation of noise at specific
points for approach, takeoff and sideline, by computing the EPNL. For approach, the pitch angle was variated for achieving
the thrust requirements keeping constant the RPMs of the CROR. The parameters under investigation were the Front Rotor
diameter, Front Rotor tip speed, Clipping, Spacing, RPM ratio, Front Rotor blades, Rear Rotor blades and incidence. A
design of experiment took place including the eight parameters, constrained with an upper and lower limit. In total 102
cases were computed, 51 for takeoff and 51 for approach and response surfaces were created using the Kriging interpolation
method. According to the author, Front Rotor Tip Speed, Speed Ratio and the Number of Blades have a direct and relative
effect on noise emissions, while Front Rotor Diameter and incidence do not give a clear response. Finally, with increase in
Clipping and Spacing, a relaxation in interaction occurs, and thus, a reduction in noise generation.
Dürrwächter et al. (2019) studied the installation of a CROR pusher engine and the impact of the tail type to the overall
tonal noise emissions. 3D URANS simulations are driven to extract the flow characteristics at the integration surfaces.
The acoustic extrapolation take place with ACCO, a tool that includes both FW­H and Mohring analogies, showing similar
results. A boundary element method was implemented for scattering and the effect of a background mean flow is taken
into account by applying a Lorentz transform. A study on including the pylon in the integration surface or using a the
BEM technique, showed similar results indicating that the pylon only introduce scattering effects and no new source noises.
However, an increased noise caused by unsteady loading provoked by the pylon wake is found, and agrees with literature

9
 Figure 4 CROR additional noise sources

according to the author. Three tail configurations were examined, a T­tail, a U­tail and a L­tail. The two latter configurations
showed a decrease in noise of about 7 dB noise level in the downward aera, due to shielding, compared to T­tail and no­tail.
However, an increase in the upward aera due to scattering is present. Again results agree with literature according to the
author. In the lateral direction, shielding and scattering neutralize each other, giving the same noise as a no­tail configuration.
Overall shielding benefits seems to peak for the U­tail at­2.4EPNdB with respect to the case without a tail. Final remarks
were that these results are case specific and that new noise sources could be emerging from the effect of the engine on flow
characteristics of the fuselage and the tail.
Smith et al. (2020) studied CROR with a locked row. The three following configurations were compared: a conven­
tional CROR, the front rotor locked and the aft rotor locked. The CROR consist of a two rotors with 4 blades of 3 meters of
diameter each, rotating at the same speed and with a thrust output of 6 kN at takeoff conditions (M=0.2 for free stream and
M=0.46 at the tip). URANS is used for the flow field and the acoustic pressure is computed on the blade with a FW­H code
based on Farassat 1A formulation. The aerodynamic analysis shows an efficiency reduction of 5.5% and 6.9% for locked
front and aft blade row accordingly. However a reduction in both maximum A­weighted SPL and EPNL is of 3.5 dBA and
7 EPNLdB for the front locked, 7.9 dBA and 12 EPNLdB for aft locked blade row, compared to baseline. Note that for
some directivity angles, the front locked can produce more noise than baseline, while for aft locked all directivity angles
are below baseline. This trend is mainly due to the fact that interaction noises are reduced when the aft is locked, compared
to the other two configurations, since it is the blade row that sees the most disturbed flow. However, the blade rows were
not optimized, and changes such as tip clipping could further improve or change the results.
Grasso et al. (2018) did a multidisciplinary optimization of a CROR considering wake­interaction noise and trailing­
edge noise. The approached for the computation of the noise sources is the following: A RANS simulation is performed,
then the mean velocity disturbance is extrapolated using the proposed method by Jaron et al. (2014), and the description of
noise modelling tools based on analytical formulations is given by Grasso (2017). The optimization framework was CADO
(Verstraete, 2010). The outcome of the optimization was a pareto front, and some design considerations. More specifically,
for most cases an increase of tip distance by applying lean and sweep lead to de­phasing and noise reduction. However, a
drop in efficiency was also perceived for many cases leading to the conclusion that aerodynamic efficiency and noise are
two antagonistic objectives.
Falissard et al. (2018) discussed the effect of deformation during the different operating conditions of the blade with
respect to the designed one. Aerodynamics and acoustics are assessed and compared to quantify the difference that can occur
by not considering in terms of thrust and acoustic emissions. Strain Pattern Analysis is used for assessing the deformation
through experiments for takeoff, cruise and approach condition (Geeraert and Stephan, 2015). Changes in sweep and twist
can be observed for all phases, but more pronounced especially for cruise but also for takeoff. Generally, the front blade
sees more geometrical changes. CFD simulations are performed with a phase/lag chorochronic approach. Porous surface
is used for FW­H that takes into account all sources and nonuniform propagation. A sensitivity analysis was conducted
to determine the appropriate location of that surface. The results overall considering the deformation, give aerodynamic
results closer to the experimental, while acoustics are also enhanced for cruise condition and at the same level for the other
two working conditions. As a final comment, at takeoff flow separation takes place at the leading edge near the tip of both
rotors, and the deformation of the blades can lead to very different results.
Colin et al. (2014) studied three configurations, an isolated, with pylon and with airframe. The numerical investigation
was performed using CFD and FW­H. The first case, which is axisymmetric, both passage and full 360 degrees were
evaluated. Results where quite identical, with small differences for interaction noise. For the configuration with the pylon,
there is an increase of BPF tone in the upstream direction and emergence of third tones for the front and aft rotor. This

10
occurs because of the incidence variation provoked by the pylon wake (pusher configuration). Furthermore, for the airframe
configuration, a 360 degree CFD and Fast Fourier Transformation (FFT) for noise assessment was compared with the typical
CFD and FW­H. The CFD with FFT although takes into account more physical phenomena of the flow, it shows high
dissipation when the grid is not fine enough. The opposite is true for the CFD and FW­H configuration. Results were
validated with measurements, exhibiting good agreement according to the author. As a last comment, the author propose a
CFD – Unsteady Euler for wave propagation for potentially better results.
Throughout this section, various works on CROR were presented. Since the engines studied do not have a geometric
or power consistency, a quantified comparison is not possible. However, the research outcomes and findings agree in a great
extent. New significant noise sources are tip vortex and rotor­rotor interaction. Finally, it can be deduced that rotor­rotor
interaction is often the major source for isolated engines, and wake­rotor interaction for installed pusher configurations. A
summary of noise sources for CROR, additional to that of a typical fan, is depicted in Figure 4.

DESIGN CONSIDERATIONS
Νoise modelling comprises two parts: (1) aerodynamic modelling (wake and blade response) and (2) aeroacoustic mod­
elling. Depending on the methods used in (1), interactions and/or unsteady effects can be represented. Step (2) comprises
two substeps: (2a) sourcenoise modelling, and (2b) farfield propagation. The majority of works included here, used un­
steady fluid dynamic for aerodynamics and FW­H for aeroacoustics modelling. This demonstrates that for the assessment
of future integrated concepts, computationally expensive, higher fidelity methods are most often used. A representative
example is the work of Romani et al., of an integrated BLI via Lattice Boltzmann method and permeable FW­H, which
required 6 hours per fan revolution solved on a cluster of 1000 cores (Romani et al., 2020). However, in a conceptual level
of design, this kind of assessments are almost prohibitive. Therefore, the following section is dedicated in giving some
design considerations, to help decision making for new concepts, reducing the necessity for high fidelity assessment at that
stage. Table 3 consists of the concise information provided in the section.

Boundary layer ingestion:

Engine inlet positioning (S­shaped or not) can be challenging. In addition to the boundary layer, wing wake could also
be interacting with the rotor, leading to further noise increase. Especially for small aircraft, the type of wings used, such as
high, low – wing etc may play a considerable role. For an aircraft of similar size to Cessna 208 Caravan, a high wing and

Configuration Parameter Change Noise Effect on

De­phasing of aft rotor sources
Sweep and lean ∼ ↓
Reduced R­R interactions (increased tip distance)
Blade­to­blade distance ↑ ↓ Reduced R­R interactions
CROR Tip Clipping ↑ ↓ Tip vortex­aft rotor interaction
Number of blades (front) ↑ ↑ ∼
Number of blades (rear) ↑ ↓ ∼
Aft rotor locked ∼ ↓ Reduced R­R interactions

Reduced downstream wake interactions

tip­to­tip distance ↑ ↓
Reduced unsteady loading
Propeller tip loading ↑ ∼ ∼
DP Downstream distance ↑ ↓ Reduced R­R interactions
Number of propellers ↑ ↓ Possibility for reduced tip speeds
Propeller diameter ↑ ↑ Lower freq.: less atmospheric dissipation
Frequency spreading ∼ ∼ Affects numerical results but not the experiments

AR → 1
BLI S­shaped inlet ↓ No vortex lift­off
OR → 0

Scattering and shielding effects

Position with respect to tale ∼ ∼
Engine­tale aerodynamic interaction
All concepts
Position with respect to wing ∼ ∼ Incoming wing wake
Blade deformation ∼ ∼ Flow field changes

Table 3 Concise table with design considerations

11
no scoop configuration is better compared to low wing or high wing with scoop (Schwartz et al., 2019).
When using a S­shaped inlet for BLI implementations, it is very essential to design the duct such that the duct outlet
(before fan) to inlet area ratio (AR = D( out)/D( in)) and offset ratio (OR = OF/D( out)) are minimized as much as possible,
see figure 2. This will reduce the chance of vortex lift off. Fan shock noise is only little altered by vortex liftoff, however,
the decay rate is reduced and the upstream propagation is changed (Defoe and Spakovszky, 2013; Spakovszky, 2019).
For the case that the BLI or the engines are positioned near the rear tale, special intention must be given to the prop­
agation effects. The geometry of the tail must be carefully selected to improve shielding and reduce scattering. That is
because engine bellow tale can increase scattering towards the ground (Clark et al., 2018). Furthermore, when positioned at
the back, it should be avoided to have tail support geometry in front of the inlet of BLI engines, since it produces unsteady
loading due to non axisymmetric flow (Petrosino et al., 2021) .

Distributed propulsion:
When noise constraints are imposed, multiple smaller propellers allow for smaller tip speeds, still being able to comply
with thrust requirements (Moore and Ning, 2019). Reducing tip speed has a beneficial effect on noise, as already mentioned.
However, thrust–to–weight ratio must be taken into consideration when choosing the number of propellers, because a large
amount of propellers rotating at lower speeds will have a significant amount of weight penalty. Moore and Ning (2019) in
their study, found that the optimum propeller number is 8, for an aircraft of the size similar to Tecnam p2006t. Another ben­
efit of multiple small propellers, is that they have higher BPF, which have larger sound power dissipation in the atmosphere
(Synodinos et al., 2018a). A lower impact at the far field is expected, and community noise level reduced.
However, the positioning of the propellers must be carefully designed. A tip­to­tip distance, L, smaller than the quarter
of the propeller diameter, (0.25D), may produces more aerodynamic interactions and thus noise (Bernardini et al., 2020).
The directivity of noise will also be affected, emitting stronger towards the front direction. Furthermore, dowstream spacing
can help reduce interactions (Zhou et al., 2017). Specifically, Alvarez et al. (2020) found that the optimum downstream
distance is half the diameter (0.5D) for their test case, noting that for larger distance, the streamtube is no longer contracting.
Downstream and tip­to­tip distance, will also define the maximum number of propellers that can be fitted in a given wing.
Finally, higher tip loading increases aerodynamic interactions, but seems not to affect noise results (Alvarez et al., 2020).
A way to reduce potential noise emissions for distributed propulsion, is varying the propellers speeds, such that the
emit in different frequencies. Although operating in different rpms showed lower noise in numerical investigation, no effect
on annoyance was detected by the subjects at the experiment produced with the auralization of numerical results (Rizzi
et al., 2017). In conclusion, it seems that there are no aeroacoustic benefit, that would suggest frequency spreading.

Contra­Rotating Open Rotors:

Generally, distance between the two rotors will have a beneficial effect on noise. It is possible to obtain a bigger
distance by positioning the blades apart, or by using sweep and lean to estrange blade tips (Kritikos et al., 2012; Grasso
et al., 2018). However, noise and aerodynamic efficiency seem to be antagonistic. De­phasing of the rear blade noise
sources can be also obtain by lean and sweep, reducing the sound pressure levels (Grasso et al., 2018). Tip clipping is
another measure that is often used to reduce CROR noise, by avoiding the interaction of front rotor tip vortex with the rear
blade (Kritikos et al., 2012). Front rotor tip speed, number of front rotor blades (constant number for aft rotor) produce a
proportional effect on noise, as well as tip speeds, as already mentioned (Kritikos et al., 2012). Increase of aft rotor blades
(for given front row) can help reduce noise at sideline and takeoff (Kopiev et al., 2017; Kritikos et al., 2012).
For installed configurations, the positioning is equally important as for BLI engines. Installed configuration CROR
shows increased blade wake interaction noise in the front rotor due to the pylon, or aerodynamic load variation when in
tractor configuration (Kennedy et al., 2018). These effects must be considered, and the distance between pylon and blade
rows must be carefully selected. The engine must be far enough from the aircraft body, so that there is no unsteady loading
and turbulence intake caused by boundary layer, if BLI is not the purpose. Scattering effects, as for BLI, must also be taken
into account, having even more importance since the engine is not shielded by a nacelle or the body of the aircraft.
Finally, a method to reduce noise emissions during takeoff is by locking the aft blade row (working as a stator) (Smith
et al., 2020). This will decreases the noise generated by the aft rotor, in all directivities. Small increase of front rotor noise is
expected in some directions. Efficiency drop is present, but for long flights were the cruise phase is particularly important,
the benefit of noise overcomes the drawback of lower efficiency during takeoff.
Another critical point is blade deformation. Significant blade deformation during operation, can change the flow
field around the engine and affect both efficiency and noise emissions (Falissard et al., 2018). Similar to blade deformation,
manufacturing tolerances should be taken also into account for final noise outcome. A method for predicting the sensitivities
of manufacturing of turbine blades, which could also be applied for deformation, is presented by Zamboni et al. (2016).

CERTIFICATION
The overall effect of noise generated by an aircraft is measured as a function of its impact on the community. The term
“community” describes both the passengers flying during an aircraft’s mission, as well as residential areas in the vicinity
of airports. In an attempt to constrain the emitted external noise levels, specific standards and recommended practices are

12
 Figure 5 Single landing and take­off 80 dB noise contours for four hypothetical aircraft that just meet the noise
limits of the various ICAO Annex 16 Volume 1 Chapters (EASA, 2019)

adopted. In particular, the noise levels are established in compliance with the applicable noise certification standards as
defined in the first volume of ICAO, in Annex 16 (ICAO, 2014). It is out of the scope of this work to analyze in detail
this document. In brief, there are two major axes upon which it is built. The first defines the type of aircraft based on
its maximum take­off mass and cruise velocity, ranging from subsonic jet airplanes and propeller driven airplanes, up to
helicopters and novel rotorcraft concepts. In order to gain a better understanding of the different levels of noise for each
category, Figure 6 summarises the noise levels for certified aircraft configurations, based on EASA’s publicly available
databases. The second classification considers the increasingly stringent standards, with type certificates submitted before
1977, between 1977 and 2006, between 2006 and 2017, as well as 2017 and beyond. A representative illustration can be
observed in Figure 5 , showing the change in the emitted noise levels for the aforementioned time periods, for a variety of
aircraft classes.
To date, when it comes to certification of aircraft noise, specific points are used to evaluate the emitted noise levels,
based on distances which are aircraft­type specific (ICAO, 2014). The metric used for noise in certification is Perceived
Noise Level (PNL) – based; a full description of the metric can be found in appendix 2 of Annex 16 volume 1 of ICAO
(2014). For environmental assessments, Sound Pressure Level (SPL) is often used. SPL is the logarithmic perception of
sound, defined as the logarithmic fraction of the measured noise with respect to the reference pressure threshold of human
hearing. More metrics exist, able to consider parameters such as the distance, the duration of exposure, the sound intensity
etc. but all having as base measurement the pressure fluctuations.
The forecast growth in air traffic, along with the increasingly stringent environmental legislation, suggest the intro­
duction of radical propulsion architectures in the future civil aircraft fleet. Νovel configurations may have different aircraft
shapes, such as hybrid­wing bodies, or propulsion system position and type, which may bring unusual noise sources or
directivity angles. For example, DP is expected to have a larger number of sources, emitting with a lower power. Another
example, is the change of emitting frequencies due to different propulsor sizes. Engine position will affect the directivity
of sound, making some community regions quieter, and some other noisier. Further research is necessary in the orientation
of noise source identification for the new types of aircraft for even better understanding. The above elements depicts the
need to invest in research and transformation of noise certification landscape, in order to satisfy noise reduction targets.
Furthermore, for the envisaged designs, under the set time­frames, all the product development stages shall be acceler­
ated. In this context, the impact of conceptual design changes on aircraft certification capability shall be brought up earlier
in the development process. Snider et al. (2013) developed a method for rotorcraft broadband noise prediction related to
civil noise certification. Vouros et al. (2021) discussed the impact of morphing rotor optimization on ground noise levels of
rotorcraft following the EASA descent certification trajectory (ICAO, 2008). A numerical procedure for the prediction of
certification noise generated by aircraft landing gear was presented by Casalino et al. (2012). As depicted by Nöding and
Bertsch (2021), inclusion of full­aircraft virtual certification in the conceptual design stage of future aircraft will enable the
holistic design space exploration and evaluation of concept proofs.

Figure 6 Noise levels on approved aircraft configurations based on EASA’s database

13
CONCLUSIONS
Throughout this work, an overview of the aeroacoustics of aircraft was performed. The noise sources were exam­
ined and made available to the reader. Aeroacoustic behavior of Boundary Layer Ingestion, Distributed Propulsion and
Contra­Rotating Open Rotors concepts was studied and presented through various works in the literature. Furthermore,
conceptual design phase guidelines were provided for each of the concepts. Finally, certification aspects and observations
were discussed, considering future changes.
Noise sources for BLI, DP and CROR do not differ from that of a typical propeller or fan in terms of physical phenom­
ena. However, they can be more pronounced due to different operating conditions. It was derived that BLI tends to increase
noise compared to the conventional under­wing engine positioning. Special attention must be paid in reducing the AR and
OR parameters in case of S­shaped inlet. The main reason for noise augmentation is the non­uniform and turbulent inlet
flow usually encountered by the fan, and possible vortex lift off. On the other hand, distributed propulsion showed a leaning
towards noise reduction. The main reason is the possibility for tip speed reduction and larger atmospheric dissipation for
higher frequencies. However, if positioned improperly, the interaction noise can be significant. Tip­to­tip and downstream
distances must be at least 0.25D to reduce to a maximum noise. Tip loading and frequency spreading did not seem to have
any effect in noise emissions for distributed propulsion. For CROR, it was concluded that increasing the blade­to­blade
distance, tip clipping, number of rear rotor blades reduces the noise. Sweep and lean can be used to reduce noise, while
aft rotor lockage can reduce noise for take­off. Finally, integration plays a major role, since positioning with respect to
wing and tale can enhance noise caused turbulent intake, but scattering effects as well. The majority of works included here
were based on the relatively common aerodynamics assessment coupled with acoustic analogy setup. Something expected
since noise emerging from interactions is very difficult to be evaluated. Finally, the landscape of certification is expected
to change, in order to comply with noise source changes but also the requirement to be virtually applied in an early design
phase.
An overview for conceptual design noise sources, design considerations and certification for novel aircraft engine
architecture has been proposed. In the future, emerging works could be included to make the design considerations even
more complete, and the certification aspects more clear.

ACKNOWLEDGMENTS
Part of the work described in this article was done within the project HECARRUS, which has received funding from
the Clean Sky 2 Joint Undertaking (JU) under the European Union’s Horizon 2020 Research and Innovation Programme
under Grant Agreement number 865089.

REFERENCES
Agarwal, N., Ganz, U. and Premo, J. (2004). Compressor Noise Contribution to Inlet Noise. In 10th AIAA/CEAS Aeroa­
coustics Conference, doi:10.2514/6.2004­2913.
Ali, F., Ellbrant, L., Elmdahl, D. and Grönstedt, T. (2016). A Noise Assessment Framework for Subsonic Aircraft and
Engines. Turbo Expo: Power for Land, Sea, and Air, doi:10.1115/GT2016­58012, v001T01A039.
Alvarez, E., Schenk, A., Critchfield, T. and Ning, A. (2020). Rotor­on­rotor aeroacoustic interactions of multirotor in hover.
Journal of the American Helicopter Society (in review) .
Barbosa, J. R. and Dezan, D. J. (2013). Turbojet Engine Noise Prediction Utilizing Empirical Methods. Turbo Expo: Power
for Land, Sea, and Air, doi:10.1115/GT2013­95274, v002T07A024.
Bernardini, G., Centracchio, F., Gennaretti, M. et al. (2020). Numerical characterisation of the aeroacoustic signature of
propeller arrays for distributed electric propulsion. Applied Sciences 10: 2643, doi:10.3390/app10082643.
Berton, J. J. and Nark, D. M. (2019). Low­noise operating mode for propeller­driven electric airplanes. Journal of Aircraft
56: 1708–1714, doi:10.2514/1.C035242.
Bertsch, L., Snellen, M., Enghardt, L. and Hillenherms, C. (2019). Aircraft noise generation and assessment: executive
summary. doi:10.1007/s13272­019­00384­3.
Brooks, T. F. and Marcolini, M. A. (1986). Airfoil tip vortex formation noise. AIAA Journal 24: 246–252, doi:10.2514/3.
9252.
Casalino, D., Diozzi, F., Sannino, R. and Paonessa, A. (2008). Aircraft noise reduction technologies: A bibliographic review.
Aerospace Science and Technology 12: 1–17, doi:10.1016/j.ast.2007.10.004.
Casalino, D., Noelting, S., Fares, E., VandeVen, T., Perot, F. and Bres, G. (2012). Towards Numerical Aircraft Noise Cer­
tification: Analysis of a Full­Scale Landing Gear in Fly­Over Configuration. doi:10.2514/6.2012­2235.
Clark, I., Thomas, R. H. and Guo, Y. (2018). Aircraft System Noise Assessment of the NASA D8 Subsonic Transport
Concept. In 2018 AIAA/CEAS Aeroacoustics Conference, doi:10.2514/6.2018­3124.
CleanAviation (2020). An ambitious vision for clean aviation. https://clean­aviation.eu/#home.
Colin, Y., Wlassow, F., Caruelle, B. et al. (2014). Installation effects on contra­rotating open rotor noise at high­speed. In
20th AIAA/CEAS aeroacoustics conference, 2971, doi:10.2514/6.2014­2971.
Dahl, M. D. (2012). Assessment of nasa’s aircraft noise prediction capability. https://ntrs.nasa.gov/citations/20120012957.
Darecki, M., King, I., Edelstenne, C. et al. (2011). Flightpath 2050 Europe’s Vision for Aviation. doi:10.2777/50266.
Defoe, J., Narkaj, A. and Spakovszky, Z. (2010). A Body­Force Based Methodology For Predicting Multiple­Pure­Tone
Noise: Validation. In 16th AIAA/CEAS Aeroacoustics Conference, doi:10.2514/6.2010­3747.

14
Defoe, J. J. and Spakovszky, Z. S. (2013). Effects of Boundary­Layer Ingestion on the Aero­Acoustics of Transonic Fan
Rotors. Journal of Turbomachinery 135, doi:10.1115/1.4023461, 051013.
Doolan, C., Moreau, D., Arcondoulis, E. J. G. and Albarracin, C. (2012). Trailing edge noise production, prediction and
control. New Zealand Acoustics 25: 22–29.
Dürrwächter, L., Keßler, M. and Krämer, E. (2019). Numerical assessment of open­rotor noise shielding with a coupled
approach. AIAA Journal 57: 1930–1940, doi:10.2514/1.J057531.
EASA, E., EEA (2019). European Aviation Environmental Report 2019. Tech. rep.
Falissard, F., Boisard, R., Gaveriaux, R. et al. (2018). Influence of blade deformations on open­rotor low­speed and high­
speed aerodynamics and aeroacoustics. Journal of Aircraft 55: 2267–2281, doi:10.2514/1.C034676.
Farassat, F. (2007). Derivation of Formulations 1 and 1A of Farassat. Tech. rep.
Fenyvesi, B., Tokaji, K. and Horváth, C. (2019). Investigation of a pylons effect on the character of counter­rotating open
rotor noise using beamforming technology. Acta Acustica united with Acustica 105: 56–65, doi:10.3813/AAA.919287.
Filippone, A. (2014). Aircraft noise prediction. Progress in Aerospace Sciences 68: 27–63, doi:10.1016/j.paerosci.2014.02.
001.
Geeraert, A. and Stephan, C. (2015). CROR blade deformation, part 1: Experimental results by strain pattern analysis. In
IFASD 2015.
Gkoutzamanis, V. G., Kavvalos, M. D., Srinivas, A., Mavroudi, D., Korbetis, G., Kyprianidis, K. G. and Kalfas, A. I. (2021).
Conceptual Design and Energy Storage Positioning Aspects for a Hybrid­Electric Light Aircraft. Journal of Engineering
for Gas Turbines and Power 143, doi:10.1115/1.4050870, 091019.
Graham, W., Hall, C. and Vera Morales, M. (2014). The potential of future aircraft technology for noise and pollutant
emissions reduction. Transport Policy 34: 36–51, doi:10.1016/j.tranpol.2014.02.017.
Grasso, G. (2017). Development of hybrid methods for the computation of tonal and broadband fan noise source and prop­
agation. Ph.D. thesis, Université de Sherbrooke ­ von Kármán Institute for Fluid Dynamics.
Grasso, G., Moreau, S., Christophe, J. and Schram, C. (2018). Multi­disciplinary optimization of a contra­rotating fan.
International Journal of Aeroacoustics 17: 655–686, doi:10.1177/1475472X18789000.
Hendricks, E. S. (2018). A review of boundary layer ingestion modeling approaches for use in conceptual design.
https://ntrs.nasa.gov/citations/20180005165. Report, NASA Glenn Research Center .
Hileman, J., Spakovszky, Z., Drela, M. and Sargeant, M. (2007). Airframe Design for ”Silent Aircraft”. In 45th AIAA
Aerospace Sciences Meeting and Exhibit, doi:10.2514/6.2007­453.
Horváth, C. (2015). Beamforming investigation of dominant counter­rotating open rotor tonal and broadband noise sources.
AIAA Journal 53: 1602–1611, doi:10.2514/1.J053465.
Hubbard, H. H. (1991). Aeroacoustics of flight vehicles: theory and practice, 1. NASA Office of Management, Scientific
and Technical Information Program.
Huff, D. L., Henderson, B. S. and Envia, E. (2016). Motor Noise for Electric Powered Aircraft. In 22nd AIAA/CEAS Aeroa­
coustics Conference, AIAA 2016­2882, doi:10.2514/6.2016­2882.
ICAO (2008). Procedures for the noise certification of aircraft. environmental technical manual, sgar 1.
ICAO (2014). Annex 16, icao volume 1, aircraft noise.
Jaron, R., Moreau, A. and Guerin, S. (2014). RANS­informed fan noise prediction: separation and extrapolation of rotor
wake and potential field. In 20th AIAA/CEAS aeroacoustics conference, 2946, doi:10.2514/6.2014­2946.
Kennedy, j., Eret, P. and Bennett, G. (2018). A parametric study of airframe effects on the noise emission from installed
contra­rotating open rotors. International Journal of Aeroacoustics 17: 624–654, doi:10.1177/1475472X18789003.
Knacke, T. J. and Thiele, F. (2013). Numerical Analysis of Slat Noise Generation. In 19th AIAA/CEAS Aeroacoustics Con­
ference, doi:10.2514/6.2013­2162.
Kolb, A., Mancini, S., Rossignol, K.­S. and Ewert, R. (2020). Flap Side­Edge Noise Simulation Using RANS­based Source
Modelling. In AIAA AVIATION 2020 FORUM, doi:10.2514/6.2020­2581.
Kopiev, V., Shur, M., Travin, A. et al. (2017). Assessment of community noise for a medium­range airplane with open­rotor
engines. Acoustical Physics 63: 723–730, doi:10.1134/S1063771017060069.
Krejsa, E. A. and Stone, J. R. (2014). Enhanced fan noise modeling for turbofan engines.
https://ntrs.nasa.gov/citations/20150000884.
Kritikos, K., Giordano, E., Kalfas, A. I. and Tantot, N. (2012). Prediction of Certification Noise Levels Generated by Contra­
Rotating Open Rotor Engines. Turbo Expo: Power for Land, Sea, and Air, 239–247, doi:10.1115/GT2012­69232.
Kyprianidis, K. G., Gkoudesnes, C. and Camilleri, W. (2015). Gas Turbines for Power and Propulsion. American Cancer
Society. 1–25, doi:10.1002/9781118991978.hces140.
Lopes, L. and Burley, C. (2016). Anopp2 user’s manual. https://ntrs.nasa.gov/citations/20160014858 .
Molin, N. (2019). Airframe noise modeling and prediction. CEAS Aeronautical Journal 10: 11–29, doi:10.1007/
s13272­019­00375­4.
Moore, K. R. and Ning, A. (2019). Takeoff and performance trade­offs of retrofit distributed electric propulsion for urban
transport. Journal of Aircraft 56: 1880–1892, doi:10.2514/1.C035321.
Mueller, T. (2002). Aeroacoustic measurement(book). Experimental Fluid Mechanics Series, Berlin, Germany, Springer­
Verlag GmbH, 2002 doi:10.1007/978­3­662­05058­3.
Ning, S. A. (2013). Airfoilprep. py documentation: Release 0.1. 0. Tech. rep., National Renewable Energy Lab.(NREL),
Golden, CO (United States), doi:10.2172/1260130.

15
Ning, S. A. (2014). A simple solution method for the blade element momentum equations with guaranteed convergence.
Wind Energy 17: 1327–1345, doi:10.1002/WE.1636.
Nöding, M. and Bertsch, L. (2021). Application of noise certification regulations within conceptual aircraft design.
Aerospace 8, doi:10.3390/aerospace8080210.
Petrosino, F., Barbarino, M. and Staggat, M. (2021). Aeroacoustics assessment of an hybrid aircraft configuration with
rear­mounted boundary layer ingested engine. Applied Sciences 11: 2936, doi:10.3390/app11072936.
Rebholz, P. S., Krebietke, S., Abhari, R. S. and Kalfas, A. I. (2016). Turbine aerodynamic low­frequency oscillation and
noise reduction using partial shrouds. Journal of Propulsion and Power 32: 1067–1076, doi:10.2514/1.B36031.
Riboldi, C. E., Trainelli, L., Mariani, L., Rolando, A. and Salucci, F. (2020). Predicting the effect of electric and hybrid­
electric aviation on acoustic pollution:. Noise Mapping 7: 35–56, doi:doi:10.1515/noise­2020­0004.
Rizzi, S. A., Palumbo, D. L., Rathsam, J., Christian, A. W. and Rafaelof, M. (2017). Annoyance to Noise Produced by a Dis­
tributed Electric Propulsion High­Lift System. In 23rd AIAA/CEAS Aeroacoustics Conference, doi:10.2514/6.2017­4050.
Romani, G., Ye, Q., Avallone, F., Ragni, D. and Casalino, D. (2020). Numerical analysis of fan noise for the nova boundary­
layer ingestion configuration. Aerospace Science and Technology 96: 105532, doi:10.1016/j.ast.2019.105532.
Schwartz, K., Burdisso, R., Witcher, B. et al. (2019). Development of a Ducted Propulsor for BLI Electric Regional Aircraft
­ Part II: Aeroacoustic Analysis. In AIAA Propulsion and Energy 2019 Forum, doi:10.2514/6.2019­3854.
Sharpf, D. and Mueller, T. (1995). An experimental investigation of the sources of propeller noise due to the ingestion of
turbulence at low speed. Experiments in Fluids 18: 277–287, doi:10.1007/BF00195098.
Shur, M. L., Spalart, P. R., Strelets, M. K. and Travin, A. K. (2015). An enhanced version of des with rapid transition from
rans to les in separated flows. Flow, turbulence and combustion 95: 709–737, doi:10.1007/s10494­015­9618­0.
Smith, D. A. (2020). Aerodynamic and Aeroacoustic Analysis of Counter Rotating Open Rotors. Ph.D. thesis, University
of Manchester, doi:10.13140/RG.2.2.23515.59682.
Smith, D. A., Filippone, A. and Barakos, G. N. (2020). Acoustic analysis of counter­rotating open rotors with a locked blade
row. AIAA Journal 58: 4401–4414, doi:10.2514/1.J059273.
Smith, M. J. T. (1989). Aircraft Noise. Cambridge Aerospace Series. Cambridge University Press, doi:10.1017/
CBO9780511584527.
Snider, R., Samuels, T., Goldman, B. and Brentner (2013). Full­scale rotorcraft broadband noise prediction and its relevance
to civil noise certification criteria. In 69th American Helicopter Society International Annual Forum 2013, 2, 1293–1307.
Spakovszky, Z. (2019). Advanced low­noise aircraft configurations and their assessment: past, present, and future. CEAS
Aeronautical Journal 10: 137–157, doi:10.1007/s13272­019­00371­8.
Steiner, H., Seitz, A., Wieczorek, K., Plötner, K., Isikveren, A. and Hornung, M. (2012). Multi­disciplinary design and
feasibility study of distributed propulsion systems. In 28th International Congress of the Aeronautical Sciences, 23–28.
Synodinos, A., Self, R. and Torija, A. (2018a). Preliminary Noise Assessment of Aircraft with Distributed Electric Propul­
sion. In 2018 AIAA/CEAS Aeroacoustics Conference, doi:10.2514/6.2018­2817.
Synodinos, A. P., Self, R. H. and Torija, A. J. (2018b). Framework for predicting noise–power–distance curves for novel
aircraft designs. Journal of Aircraft 55: 781–791, doi:10.2514/1.C034466.
Terracol, M., Manoha, E. and Lemoine, B. (2016). Investigation of the unsteady flow and noise generation in a slat cove.
AIAA Journal 54: 469–489, doi:10.2514/1.J053479.
Topol, D. A. and Mathews, D. C. (2010). Rotor wake/stator interaction noise prediction code technical documentation and
user’s manual. https://ntrs.nasa.gov/citations/20110000525.
Tyler, J. M. and Sofrin, T. G. (1962). Axial Flow Compressor Noise Studies. In Pre­1964 SAE Technical Papers 620532.
SAE International, doi:10.4271/620532.
Van Zante, D. E. (2015). Progress in Open Rotor Research: A U.S. Perspective. Turbo Expo: Power for Land, Sea, and Air,
doi:10.1115/GT2015­42203, v001T01A003.
Verstraete, T. (2010). CADO: a computer aided design and optimization tool for turbomachinery applications. In Proceed­
ings of the 2nd international conference on engineering optimization, Lisbon, Portugal, 69, 69.
Vilmin, S., Lorrain, E., Hirsch, C. and Swoboda, M. (2006). Unsteady Flow Modeling Across the Rotor/Stator Inter­
face Using the Nonlinear Harmonic Method. Turbo Expo: Power for Land, Sea, and Air, 1227–1237, doi:10.1115/
GT2006­90210.
Vouros, S., Polyzos, N. D., Goulos, I. and Pachidis, V. (2021). Impact of optimized variable rotor speed and active blade
twist control on helicopter blade–vortex interaction noise and environmental impact. Journal of Fluids and Structures
104: 103285, doi:10.1016/j.jfluidstructs.2021.103285.
Wick, A. T., Hooker, J. R. and Zeune, C. H. (2015). Integrated Aerodynamic Benefits of Distributed Propulsion. In 53rd
AIAA Aerospace Sciences Meeting, doi:10.2514/6.2015­1500.
Zamboni, G., Banks, G. and Bather, S. (2016). Gradient­Based Adjoint and Design of Experiment CFD Methodologies
to Improve the Manufacturability of High Pressure Turbine Blades. Turbo Expo: Power for Land, Sea, and Air, doi:
10.1115/GT2016­56042, v02CT39A001.
Zhou, W., Ning, Z., Li, H. and Hu, H. (2017). An Experimental Investigation on Rotor­to­Rotor Interactions of Small UAV
Propellers. In 35th AIAA Applied Aerodynamics Conference, doi:10.2514/6.2017­3744.

16