sensors

Article Measurement of the Flow Field Generated by † Multicopter Propellers

Zbigniew Czy˙z 1,*, Paweł Karpi ´nski 2 and Wit Stryczniewicz 3

1 Aeronautics Faculty, Military University of Aviation, 08-521 D˛eblin,Poland 2 Department of Thermodynamics, Fluid Mechanics and Aviation Propulsion Systems, Faculty of Mechanical Engineering, Lublin University of Technology, 20-618 Lublin, Poland; [email protected] 3 Aerodynamics Department, Łukasiewicz Institute of Aviation, 02-256 Warsaw, Poland; [email protected] * Correspondence: [email protected] This paper is an extended version of the published conference paper “Czy˙z,Z.; Siadkowska, K. † Measurement of Air Flow Velocity around the Unmanned Rotorcraft. In Proceedings of the 2020 IEEE International Workshop on Metrology for AeroSpace, Pisa, Italy, 22–24 June 2020.”



Received: 4 August 2020; Accepted: 24 September 2020; Published: 27 September 2020


Abstract: This paper presents the results of research on the airflow around a multirotor aircraft. The research consisted of the analysis of the velocity field using particle image velocimetry. Based on the tests carried out in a wind tunnel, the distribution of the velocity and its components in the vertical plane passing through the propeller axis were determined for several values of the angle of attack of the tested object for two values of airflow velocity inside the tunnel, i.e., vwt = 0 m/s and vwt = 12.5 m/s. Determining the velocity value as a function of the coordinates of the adopted reference system allowed for defining the range of impact of the horizontal propellers and the fuselage of the research object itself. The tests allowed for quantitative and qualitative analyses of the airflow through the horizontal rotor. Particular attention was paid to the impact of the airflow and the angle of attack on the obtained velocity field distributions.

Keywords: autogyro; flow velocity; gyrocopter; gyroplane; hybrid propulsion; multicopter; PIV; particle image velocimetry; wind tunnel

1. Introduction Several techniques for measuring and calculating aerodynamic properties and the phenomena occurring during airflow are used in aircraft aerodynamics studies. One of the trends involves the use of synergies of experimental research in wind tunnels and numerical calculations. In the case of numerical tests, computational fluid dynamics (CFD) is used at the design stage. Thus, it is possible to validate the assumptions of the conducted projects and avoid the costly preparation of models or prototypes for bench tests [1–4]. This study is an extended version of the paper entitled “Measurement of Air Flow Velocity around the Unmanned Rotorcraft,” presented during the 2020 IEEE International Workshop on Metrology for Aerospace in Pisa, Italy [5]. The numerical calculation stage is followed by experimental tunnel testing, in which visualization testing plays an important role. From the beginning of aeronautical research, visualization studies have been used to understand the phenomena occurring in all phases of aircraft flight. Currently, one of the most commonly used methods is particle image velocimetry (PIV), which is an optical method used for the qualitative visualization and quantitative measurement of flow velocity, e.g., the air around the examined object. The result of the measurement is a vector field of temporary velocity values. Based on this, it is possible to describe

Sensors 2020, 20, 5537; doi:10.3390/s20195537 www.mdpi.com/journal/sensors Sensors 2020, 20, 5537 2 of 22 Sensors 2020, 20, x FOR PEER REVIEW 2 of 22 thecontains phenomena small particles occurring (tracer during particles) the measurement. that act as markers The flowing [6]. While medium maintaining contains relatively small particles small (tracerparticle particles) sizes in the that main act as fluid, markers it is [6possible]. While to maintaining accurately relativelyreflect the smalldynamics particle and sizes flow in path. the main The fluid,fluid with it is possiblethe suspended to accurately fine particles reflect is the illuminated dynamics such and that flow the path. particles The fluidare visible with thein one suspended selected fineplane. particles Particle is image illuminated velocimetry such thatinvolves the particles photographing are visible suspended in one selected particles plane. placed Particle in a flow image and velocimetrycorrelating the involves obtained photographing images. Based suspended on the moveme particlesnt placed of the in tracer a flow particles, and correlating a quantitative, the obtained two- images.dimensional Based flow on theimage movement is obtained, of the tracerwhich particles,contains avelocity quantitative, vectors two-dimensional that specify the flow coordinate image is obtained,values and which their containsdirection. velocity vectors that specify the coordinate values and their direction. Figure1 1 shows shows aa diagram diagram explaining explaining how how the the PIV PIV technique technique works. works. AA fluidfluid withwith disperseddispersed particles flowsflows through thethe researchresearch domain.domain. In thethe measurement space, there is a laser whose beam intersects the fluidfluid inin thethe selectedselected plane.plane. The laser light falls on particlesparticles suspendedsuspended in thethe flowingflowing medium suchsuch thatthat theythey cancan bebe observed. observed. A A camera camera placed placed perpendicularly perpendicularly to to the the laser laser cutting cutting plane plane is usedis used to to observe observe the the particles particles in in real real time. time. In In order order to to obtain obtainthe the velocityvelocity fieldfield inin the examined space, a simple physical relationshiprelationship isis used,used, i.e.,i.e., vv == dx//ddtt.. The image obtainedobtained atat timetime tt11 is compared with the image at timetime tt2 in terms of the temporary positions of the particles suspended in in the the medium. medium. Knowing the time didifferencefference and distance didifferencefference for a givengiven set ofof particles,particles, it is possible to calculate the velocity at a given point of thethe measuringmeasuring space at aa givengiven moment.moment. For this purpose, post-processing isis used, used, which which involves involves performing performi ang cross-correlation a cross-correlation process. process. The presented The presented analysis isanalysis carried is out carried for the out vertical for the and vertical horizontal and horizontal velocity components.velocity components.

Figure 1. Diagram illustrating the principle of the particle image velocimetry (PIV) method.

If therethere isis aa didifferencefference in in the the density density of of the the medium medium and and the the particles particles sent sent into into it, it, it isit necessaryis necessary to taketo take into into account account the the velocity velocity lag lag in the in constantlythe constant acceleratingly accelerating fluid. fluid. This This situation situation occurs occurs in the in case the ofcase gases, of gases, where where the density the density of particles of particles is much is much higher higher than than the densitythe density of fluid of fluid [7]. The[7]. The velocity velocity lag valuelag value can can then then be calculated be calculated using using the formula:the formula:

𝜌 −𝜌 𝑢 =𝑢 −𝑢=𝑑 𝑎, 18𝜇 (1) where: up—velocity of the tracer particles (m/s), u—fluid velocity (m/s), ρp—density of the tracer particles (kg/m3). In the PIV method, Brownian motion plays an important role, which involves the chaotic movements of particles in a fluid caused by the collisions of a suspension with fluid particles. The Sensors 2020, 20, 5537 3 of 22

2 ρp ρ us = up u = d − a, (1) − p 18µ where: up—velocity of the tracer particles (m/s), u—fluid velocity (m/s), 3 ρp—density of the tracer particles (kg/m ).

In the PIV method, Brownian motion plays an important role, which involves the chaotic movements of particles in a fluid caused by the collisions of a suspension with fluid particles. The motion of a spherical particle, which the Stokes law can be applied to, is characterized by the diffusion coefficient introduced by Einstein [7]:

KT D = a , (2) 3πµ dp · where: dp—particle diameter (m), K—Boltzmann constant (J/K),

Ta—absolute fluid temperature (K), µ—dynamic viscosity of the fluid (Pa s). · The PIV technique is currently widely used in aerodynamic research, both in the field of basic sciences and commercial research, for low and high flow rates. The application of PIV to helicopter rotor aerodynamic measurements was described in detail in Reference [8]. In Reference [9], on the basis of the conducted analyses, an image of the vortex ring on a helicopter’s rotor was obtained, and in Reference [10], a study of the temporal and spatial turbulence scales using a PIV system with a high measurement frequency was presented. A vortex ring was also examined using this method in Reference [11]. The rotor wake in ground effect and its investigation in a wind tunnel using the PIV technique was presented in Reference [12]. References [13,14] provide examples of the application of the PIV technique for testing transonic and supersonic flows, where the accurate representations of the shock wave position and the wave shape were of great importance. In Reference [15], calibration tests of the nozzle in the tunnel were performed with Mach numbers equal to 3.5 and 4.5. The measurements using modern PIV systems are characterized by an increasing frequency and a spatial resolution, thus making it possible to examine complex flow phenomena [16]. The development of PIV systems is a consequence of the progress in laser and computer technologies, which facilitate the implementation of advanced image processing algorithms [7,17,18]. In this work, the PIV technique was used to determine the aerodynamic interference of rotor streams of an innovative multirotor aircraft. The results obtained can be used to optimize the new design. The PIV method can be used to analyze interference or aerodynamic interaction of objects. An example is Reference [19], in which the authors examined the interaction between a helicopter rotor and a ship hull on its board during the landing process. The PIV technique has been constantly improving, as confirmed by several scientific papers devoted to this topic. References [20,21] present a new system for large-scale tomographic PIV in low-speed wind tunnels. For tracer particles, sub-millimeter helium-filled soap bubbles were used. It was proved that using this solution significantly improved the light scattering compared to micron-sized droplets. The authors of Reference [22] used a pulse-burst laser as part of a time-resolved PIV method. This made it possible to visualize the flow under high-speed and turbulence conditions. PIV tests can be combined with simulation tests performed using the CFD method. An example is Reference [23], in which a vertical axis wind turbine of the Savonius type was studied. The simulation tests were validated through wind tunnel tests using the PIV method. A similar approach was used in Sensors 2020, 20, 5537 4 of 22

Reference [24], in which a cross-ventilation flow for different isolated building configurations using CFD and PIV methods was studied. This method is often used with other measuring methods, e.g., embedded laser doppler velocimetry (ELDV). This approach was used by the authors of Reference [25], where the boundary layer profile and the characteristics of the flow velocity distribution, close to the leading edge of a helicopter blade profile, were examined using both techniques. The authors of Reference [26] used the same methodology in order to examine vortex shedding and the wake–wake interaction in a transonic turbine stage. A comprehensive analysis of the application of the PIV method for studying the aerodynamic loads of airfoils and aircraft propellers is presented in Reference [27]. The authors focused on the method of determining the pressure field using the flow velocity field obtained by the PIV method, among other aspects. The research in this area is also presented in Reference [28]. Furthermore, Reference [29] presented various ways of using the PIV technique for the rotor aerodynamics analysis. The PIV method is used to study the distribution of flow velocities outside and inside an aircraft. In Reference [30], the authors tested tactical jet aircraft nozzles, while in Reference [31], they analyzed the air distribution inside the passenger cabin. In Reference [32], the distribution of the flow velocity around the helicopter structure was examined in 37 different regions for two model configurations (isolated fuselage and fully-equipped model) in four different flight conditions. These studies are more widely presented in Reference [33]. Interesting research on helicopter aerodynamics using the PIV method is presented in Reference [34]. The wake of a full-scale UH-60A rotor was tested in forward flight in a wind tunnel. Flow field tests are extremely important from the point of view of analyzing the phenomenon of aerodynamic interference. Research on this subject is carried out in the largest research and scientific centers in the world [35]. It should be noted that flow visualization obtained as a result of computer model solutions is one of the most important instruments for analyzing the solutions obtained, and is also a tool for the in-depth analysis of the impact of flow interference phenomena on the resultant loads of rotorcraft. The variability of the flow field in time, the interpenetration of vortex traces, and the pulsations of the disturbance field can be assessed both qualitatively and quantitatively by observing the image of this phenomenon on the computer screen, which involves introducing the selectivity of interference components where possible. Based on the tests carried out using PIV image anemometry, airflow velocity profiles were developed as a function of the distance from the propellers. On this basis, it is possible to determine the range of the additional multirotor drive system. The analysis of the velocity field around the tested object allows for the selection of the installation location (height) of both the rotor working in autorotation relative to the fuselage and other components to minimize the impact of the additional propulsion system.

2. Research Object and Methodology

2.1. Research Object The research object was a model of an aircraft equipped with a hybrid propulsion system. The aircraft is an innovative structure combining the features of a multirotor and a gyroplane. The concept combined the advantages of a gyroplane (a light machine with a simple design and was cheaper to operate) with the advantages of a multirotor design that allowed for shortening the take-off distance or even vertical take-off and landing. The use of additional propulsion ensured flight stabilization at low speeds and increased the safety of the entire aircraft. The main rotor rotated, as in a classic gyroplane, due to the phenomenon of autorotation, while the set of additional propellers was driven by brushless electric motors. The combination of these two different propulsion systems resulted in an innovative means of individual air transport and a machine capable of performing various types of unmanned missions. The structure of the tested aircraft consisted of four horizontal propellers, a pushing propeller, and a two-blade main rotor. The basic technical parameters of the research object are presented in Table1, while the design of the discussed concept is shown in Figure2. Sensors 2020, 20, 5537 5 of 22

Table 1. Technical parameters of the tested aircraft.

Parameter Value Rotor diameter 1100 mm Body dimensions with stabilizers (l h w) 633 178 294 mm × × × × Diameter of the three-blade push propeller 228.6 mm Diameter of horizontal propellers 254.0 mm ROXXY BL-Outrunner 3548/05 Push propeller engine (Robbe, Inzersdorf im Kremstal, Austria) T-MOTOR MT2814 770 KV Sensors 2020, 20, x FORHorizontal PEER REVIEW propeller engine 5 of 22 (T-Motor, Nanchang, Jiangxi, China) Sensors 2020, 20, x FOR PEER REVIEW 5 of 22

Figure 2. View of the model of the tested multirotor aircraft. Figure 2. ViewView of the model of the tested multirotor aircraft. The geometric model of the aircraft was based on the structure of the Taifun gyroplane, which was Thedesigned geometricgeometric by modelthemodel company of of the the aircraft aircraftAVIATION was was based basedARTUR on on the TRENDAKthe structure structure of (Jaktorów-Kolonia, theof the Taifun Taifun gyroplane, gyroplane, Poland). which which wasThe designedwasgyroplane designed by model the by company wasthe companytransferred AVIATION AVIATION for ARTURthe purposes ARTUR TRENDAK of ongoingTRENDAK (Jaktor worksów-Kolonia, (Jaktorów-Kolonia, due to the Poland). cooperation ThePoland). gyroplane between The modelgyroplaneentities. was Figures model transferred 3 was and transferred 4 for show the the purposes for characteristic the purposes of ongoing dime of worksongoingnsions dueof works the to fuselage the due cooperation to withthe cooperation the betweentail boom, between entities. which Figuresentities.was used3 Figures and to 4develop show 3 and the the4 characteristicshow design the of characteristic the dimensions hybrid airc dime ofraft. thensions These fuselage of are the with the fuselage real the taildimensions with boom, the whichtail of boom, the was Typhoon which used towasgyroplane. develop used to the Thedevelop design model the of used thedesign hybridfor theof the aircraft.research hybrid Thesewas airc maderaft. are theThese on real a scaleare dimensions the of real1:8. dimensions of the Typhoon of the gyroplane. Typhoon Thegyroplane. model usedThe model for the used research for the was research made on was a scalemade of on 1:8. a scale of 1:8.

Figure 3. Geometric dimensions (mm) of the test object (front view). Figure 3. Geometric dimensions (mm) of the test object (front view). Figure 3. Geometric dimensions (mm) of the test object (front view).

Figure 4. Taifun gyroplane-geometrical dimensions (mm) of the object (view from the left). Figure 4. Taifun gyroplane-geometrical dimensions (mm) of the object (view from the left). The drive unit of the examined object consisted of several cooperating key elements, i.e., propellers,The drive engines, unit controlof the regulators,examined powerobject supplyconsisted (battery), of several flight cooperating controller, and key radio-controlled elements, i.e., propellers,(R/C) devices. engines, A set control of propellers regulators, and engines power supply were responsible (battery), flight for generating controller, thrust. and radio-controlled Blades with the (R/C)NACA-9H-12 devices. A profile set of propellerswere used. and Each engines set generated were responsible about 1800 for ggenerating of thrust fromthrust. each Blades engine. with Thisthe NACA-9H-12value was required profile to were lift the used. aircraft, Each including set generated sufficient about reserves 1800 g toof performthrust from maneuvers each engine. during This the value was required to lift the aircraft, including sufficient reserves to perform maneuvers during the Sensors 2020, 20, x FOR PEER REVIEW 5 of 22

Figure 2. View of the model of the tested multirotor aircraft.

The geometric model of the aircraft was based on the structure of the Taifun gyroplane, which was designed by the company AVIATION ARTUR TRENDAK (Jaktorów-Kolonia, Poland). The gyroplane model was transferred for the purposes of ongoing works due to the cooperation between entities. Figures 3 and 4 show the characteristic dimensions of the fuselage with the tail boom, which was used to develop the design of the hybrid aircraft. These are the real dimensions of the Typhoon gyroplane. The model used for the research was made on a scale of 1:8.

Sensors 2020, 20, 5537 6 of 22 Figure 3. Geometric dimensions (mm) of the test object (front view).

FigureFigure 4.4. Taifun gyroplane-geometricalgyroplane-geometrical dimensionsdimensions (mm)(mm) ofof thethe objectobject (view(view fromfrom thethe left). left).

TheThe drivedrive unit unit of theof examinedthe examined object object consisted consis of severalted of cooperatingseveral cooperating key elements, key i.e.,elements, propellers, i.e., engines,propellers, control engines, regulators, control power regulators, supply power (battery), supply flight (battery), controller, flight and controller, radio-controlled and radio-controlled (R/C) devices. A(R/C) set of devices. propellers A set and of propellers engines were and responsible engines were for responsible generating thrust.for generating Blades withthrust. the Blades NACA-9H-12 with the profileNACA-9H-12 were used. profile Each were set generated used. Each about set 1800generated g of thrust about from 1800 each g of engine. thrust Thisfrom value each wasengine. required This tovalue liftSensors thewas aircraft, 2020 required, 20, x includingFOR to PEER lift REVIEWthe su aircraft,fficient reservesincluding to sufficient perform maneuversreserves to perform during the maneuvers flight, based during6 of 22 on thethe assumption that the main rotor did not generate lift. The additional drive units were equipped with T-MOTORflight, based MT2814 on the 770 assumption KV engines. that Thethe main design rotor of did the not engines generate used lift. inThe combination additional drive with units 10-inch propellerswere equipped allowed forwith generating T-MOTOR the MT2814 assumed 770 thrust. KV engines. The design of the engines used in combination with 10-inch propellers allowed for generating the assumed thrust. 2.2. Methodology 2.2. Methodology The presence of additional propellers affects the airflow around the fuselage, which may result in The presence of additional propellers affects the airflow around the fuselage, which may result the deterioration of the aerodynamic properties of the entire structure. Therefore, the research was in the deterioration of the aerodynamic properties of the entire structure. Therefore, the research was aimedaimed at analyzing at analyzing the the impact impact of of additional additional horizontalhorizontal propellers propellers on on the the velocity velocity field field around around the the hybridhybrid aircraft. aircraft. The The tests tests were were carried out out in in a T-1 a T-1 wind wind tunnel tunnel (Figure (Figure 5). This5). is Thisa tunnel is a with tunnel a low- with a low-speedspeed flow flow and and continuous continuous operation. operation. The The device device has an has open an openmeasuring measuring space with space a withdiameter a diameter of of 1.51.5 m m and and a a length length of of 2.02.0 m.m. TheThe maximum air air velocity velocity in inthe the tunnel tunnel equals equals 40 m/s, 40 mwhereas/s, whereas the the minimumminimum is equal is equal to 10to 10 m /m/s.s. The The airflow airflow isis generatedgenerated by by rotating rotating adjustable adjustable fan fanblades blades driven driven by a by a 55 kW55 electrickW electric motor. motor. During During the the measurements, measurements, th thee turbulence turbulence intensity intensity was was limited limited to 0.5% to 0.5% for the for the emptyempty measuring measuring space. space. The The test test stand stand isis partpart of the tunnel tunnel measuring measuring and and control control system, system, which which is is described in more detail in Reference [36]. described in more detail in Reference [36].

Figure 5. View of the used T-1 wind tunnel at the Polish Institute of Aviation [37]. Figure 5. View of the used T-1 wind tunnel at the Polish Institute of Aviation [37].

FigureFigure6 shows 6 shows a test a test stand stand for for visualizing visualizing thethe airflowairflow around around a amultirotor multirotor aircraft. aircraft. A detailed A detailed descriptiondescription of the of the measuring measuring system system is is presented presented in References References [9,38]. [9,38 The]. The velocity velocity field field visualization visualization systemsystem was was based based on theon PIVthe PIV technique technique and and consisted consisted of of two two key key components: components: aa twin-headtwin-head solid- solid-state state laser from Litron (Litron, Rugby, Warwickshire, England) (ND: YAG neodymium laser) and a PCO HiSense digital camera (PCO AG, Kelheim, Germany). The beam in each head was triggered using a given frequency and the time interval between the heads. Then, the beam was formed as a luminous cutting plane due to the cylindrical lenses. The laser light was introduced into the measuring chamber due to a special mounting arm. The tracer particles were carried out in the form of dispersed oil droplets generated on an air generator at a pressure of 1.5 bar. In this way, particles with an average size of 2 μm were obtained, which were then introduced into the tunnel channel through a nozzle. The Stokes criterion was verified. In the experiments, the DHES (dioctyl sebacate, di(2-ethylhexyl) sebacate) particles were used with a characteristic time of 2 μs. The selected particles followed the investigated flow based on their Stokes parameter value of less than 10−1 [7,39–41]. The aircraft model was placed in the tunnel in such a way that the axis of the tunnel coincided with the plane of symmetry of the fuselage of the research object. For the 0° angle of attack, the normal vector to the main rotor plane was perpendicular to the air velocity vector in the wind tunnel. The Sensors 2020, 20, 5537 7 of 22 laser from Litron (Litron, Rugby, Warwickshire, England) (ND: YAG neodymium laser) and a PCO HiSense digital camera (PCO AG, Kelheim, Germany). The beam in each head was triggered using a given frequency and the time interval between the heads. Then, the beam was formed as a luminous cutting plane due to the cylindrical lenses. The laser light was introduced into the measuring chamber due to a special mounting arm. The tracer particles were carried out in the form of dispersed oil droplets generated on an air generator at a pressure of 1.5 bar. In this way, particles with an average size of 2 µm were obtained, which were then introduced into the tunnel channel through a nozzle. The Stokes criterion was verified. In the experiments, the DHES (dioctyl sebacate, di(2-ethylhexyl) sebacate) particles were used with a characteristic time of 2 µs. The selected particles followed the 1 investigated flow based on their Stokes parameter value of less than 10− [7,39–41]. The aircraft model was placed in the tunnel in such a way that the axis of the tunnel coincided with the plane of symmetry of the fuselage of the research object. For the 0◦ angle of attack, the normal vectorSensors to the 2020 main, 20, x rotorFOR PEER plane REVIEW was perpendicular to the air velocity vector in the wind tunnel.7 Theof 22 tests were carried out for several defined attack angle values. Samples were taken at various blade positions tests were carried out for several defined attack angle values. Samples were taken at various blade around the blade revolution. The inverted vertical and horizontal velocity values at the advanced positions around the blade revolution. The inverted vertical and horizontal velocity values at the bladeadvanced might be blade a result might of be the a experimentalresult of the experiment data corruptional data corruption by light-sheet by light-sheet reflections. reflections.

FigureFigure 6. Diagram 6. Diagram of theof the test test stand stand for for visualization visualization of of the the flow flow around around the the research research object object (own (own study study basedbased on Reference on Reference [37 ]).[37]).

The cameraThe camera was positionedwas positioned in such in such a way a way that thethat observed the observed area area was 700was 700700 × mm,700 mm, which which included × the mainincluded rotor, the the main additional rotor, the propeller additional at the propelle rear ofr at the the body, rear theof the fuselage, body, andthe fuselage, the stabilizers. and the It did stabilizers. It did not cover the whole frame of the camera due to the angle of propagation of the not cover the whole frame of the camera due to the angle of propagation of the cutting plane being too cutting plane being too small. The generation of the air velocity vectors using the PIV method was small.performed The generation with a ofresolution the air velocity of 5 mm. vectors The univer usingsal the outlier PIV methoddetection was technique performed was used with for a resolution PIV of 5 mm.data Thevalidation universal [42]. outlier The median detection test technique was modi wasfied usedby introducing for PIV data a validationsingle threshold [42]. Theto the median test wasnormalized modified vector by introducing residuals. The a single resulting threshold algorithm to the is universal normalized for vectora variety residuals. of different The flow resulting algorithmconditions is universal and characteristics. for a variety In this of study, different the normalized flow conditions residuals and of characteristics.the vectors were calculated In this study, the normalizedfrom 5 × 5 mm residuals adjacent of vectors. the vectors For the were selected calculated interrogation from 5 window5 mm adjacentsize of 32 vectors.× 32 px, the For resulting the selected × interrogationvector velocity window field size had of 16,129 32 32vectors px, the placed resulting in a regular vector grid velocity of 127 field columns had 16,129 and 127 vectors rows. placedThe in × a regularpresentation grid of 127of such columns a dense and vector 127 velocity rows. Thefield presentationin this paper was of such difficult a dense because vector the vectors velocity have field in this paperto be very was small diffi cultor they because will overlap. the vectors For the have clarity to of be the very presentation, small or theyevery will second overlap. row and For column the clarity of the vector field is presented. This allowed for presenting sufficiently large vectors for a clear of the presentation, every second row and column of the vector field is presented. This allowed for presentation and to avoid overlapping. The presented velocity field was created by averaging an presentingensemble suffi ofciently 150 instantaneous large vectors velocity for a clearfields. presentation The size of the and ensemble to avoid was overlapping. chosen based The on presented the velocityresults field of wasprevious created studies. by averaging The convergence an ensemble of the control of 150 points instantaneous located in velocityfour different fields. locations The size of the ensemblewas verified: was P1—in chosen the basedfreestream on the above results the model, of previous P2—in front studies. of the The main convergence rotor above the of themodel control pointsfuselage, located P3—behind in four diff theerent main locations rotor, and was P4—in verified: the small P1—in horizontal the freestream propeller abovewake. The the location model, P2—inof frontthe of thecontrol main points rotor is abovepresented the in model Figure fuselage, 7. The red P3—behind dashed line in the Figure main 7 rotor,corresponds and P4—in to the theRMS small horizontalvelocity propeller profile for wake. the selected The location x-coordinate. of the control points is presented in Figure7. The red dashed line in Figure7 corresponds to the RMS velocity profile for the selected x-coordinate. Sensors 2020, 20, x FOR PEER REVIEW 8 of 22

Sensors 2020, 20, 5537 8 of 22 Sensors 2020, 20, x FOR PEER REVIEW 8 of 22

FigureFigure 7. Location7. Location of of the the controlcontrol points in in the the measurement measurement space. space.space.

InIn orderorderIn order to to present presentto present the the flow the flow fieldflow field aroundfield aroundaround the research the researchresearch object, object, object, the velocity the the velocity velocity distribution distribution distribution was analyzedwas was analyzed for the selected values of the y-coordinate marked in red broken lines in Figure 8. The foranalyzed the selected for the values selected of thevaluesy-coordinate of the y-coordinate marked in marked red broken in red lines broken in Figure lines8. in The Figure coordinate 8. The coordinate system for the measurement space was defined by two axes: x and y. The values of the x- systemcoordinate for thesystem measurement for the measurement space was definedspace was by twodefined axes: byx twoand axes:y. The x and values y. The of the valuesx-coordinate of the x- coordinate increased toward the front of the aircraft. The y-axis was perpendicular to the x-axis. Due increasedcoordinateto the towardanalysis increased of the a towardfragment front of th of thee thefront aircraft. measurement of the The aircraft. yarea,-axis Thethe was values y-axis perpendicular on was the perpendicular x-axis did to thenot startx -axis.to thefrom x Due -axis.zero. to Due the analysisto the analysis of a fragment of a fragment of the of measurement the measurement area, thearea, values the values on the onx-axis the x did-axis not did start not fromstart from zero. zero.

Figure 8. The adopted coordinate system of the examined area around the aircraft, along with the cross-sections considered (horizontal axis—x, vertical axis—y).

Figure 8.8. The adopted coordinate system of the examined area around the aircraft,aircraft, alongalong withwith thethe cross-sectionscross-sections consideredconsidered (horizontal(horizontal axis—axis—xx,, verticalvertical axis—axis—yy).). Sensors 20202020,, 2020,, xx FORFOR PEERPEER REVIEWREVIEW 99 ofof 2222 Sensors 2020, 20, 5537 9 of 22

Due to the fact that for a configuration of active propellerspropellers withwith aa verticalvertical thrustthrust vector,vector, thethe verticalDue airflow to the factis forced, that for and a configuration thus the selected of active coordinates propellers on withthe vertical a vertical y-axis-axis thrust werewere vector, analyzed.analyzed. the vertical NearNear theairflowthe activeactive is propellers,propellers, forced, and thethe thus distancedistance the selected ofof thethe analyzedanalyzed coordinates sectionssections on the waswas vertical 4040 mm,mm,y-axis whilewhile were awayaway analyzed. fromfrom them,them, Near itit waswas the equalactive to propellers, 80 mm. The the propeller distance ofax theisis ofof analyzed thethe additionaladditional sections drivedrive was unitunit 40mm, passedpassed while throughthrough away coordinatecoordinate from them, x it == was 880880 mm.equal The to 80 propeller mm. The operating propeller plane axis of at the α == additional 0°0° waswas atat drivethethe heightheight unitpassed dedescribedscribed through withwith coordinate thethe coordinatecoordinatex = 880 y == mm. 140140 mm. The position of the main rotor head was described using coordinates x = 1000 mm and y = 380 mm.The propellerThe position operating of the planemain rotor at α = head0◦ was was at described the height using described coordinates with the x coordinate= 1000 mm yand= 140 y = mm.380 mm,The position while the of horizontal the main rotor stabilizer head was was located described at a using height coordinates of approximatelyx = 1000 y == mm 200200 and mmmmy ((α= ==380 0°).0°). mm, The number of samples used for averaging was sufficient to reach a steady value, as it can be whileThe the number horizontal of samples stabilizer used was locatedfor averaging at a height was ofsuff approximatelyicient to reachy a= steady200 mm value, (α = 0as◦). it can be seenseen inThein thethe number convergenceconvergence of samples studiesstudies used resultsresults for presentedpresented averaging inin was Figures suffi 9cient and 10 to [43,44]. reach a Figure steady 9 shows value, the as itmean can velocitybe seen as in a the function convergence of the number studies of results samples presented forfor thethe inindicatedindicated Figures controlcontrol9 and 10 pointspoints[ 43, 44P1–P4.P1–P4.]. Figure TheThe 9assumedassumed shows numberthe mean of velocitysamples aswas a functionequal to 150, of the which number allowed of samples for obtaining for the convergence indicated controlfor the mean points velocity P1–P4. inThein thethe assumed controlcontrol numberpoints.points. Moreover,Moreover, of samples thethe was convergenceconvergence equal to 150, ofof which thethe RMSRMS allowed velocitiesvelocities for obtaining atat pointspoints P1–P4 convergenceP1–P4 waswas verified.verified. for the RMS Themean obtained velocity results in the controlare presented points. Moreover,in Figure 10. the The convergence RMS velocity of the magnitude RMS velocities vRMS atwaswas points calculatedcalculated P1–P4 using: wasusing: verified. The obtained results are presented in Figure 10. The RMS velocity magnitude vRMS was calculated using: 𝑣 = q𝑣 +𝑣 (3) = 2 + 2 (3) vRMS vxRMS vyRMS (3)

FigureFigure 9. 9. ConvergenceConvergence for for increasing increasing numbers numbers of vectorvector fieldsfields ofof the the mean mean velocity velocity at at the the control control points. points.

Figure 10. ConvergenceConvergence for the RMS velocity at the control points.

The uncertainty in the vectors of the instantaneous velocity field maps was estimated using the primary peak ratio technique [45]. The uncertainty of an individual velocity vector in the freestream did not exceed 0.5 m/s, while in the main rotor proximity,imity, itit waswas lessless thanthan 11 m/s.m/s. AnAn exceptionallyexceptionally Sensors 2020, 20, 5537 10 of 22

The uncertainty in the vectors of the instantaneous velocity field maps was estimated using the primarySensors 2020 peak, 20, xratio FOR PEER technique REVIEW [45 ]. The uncertainty of an individual velocity vector in the freestream10 of 22 didSensors not 2020 exceed, 20, x FOR 0.5 PEER m/s, REVIEW while in the main rotor proximity, it was less than 1 m/s. An exceptionally10 of 22 high value of uncertainty of up to 2 m/s m/s was estimated in the the wake wake of of the the small small horizontal horizontal propellers. propellers. high value of uncertainty of up to 2 m/s was estimated in the wake of the small horizontal propellers. This might have have been been caused caused by by a aloss loss of ofparticle particle pairs pairs due due to the to thestrong strong motion motion out of out the of plane. the plane. The This might have been caused by a loss of particle pairs due to the strong motion out of the plane. The Theuncertainty uncertainty of the of instantaneous the instantaneous velocity velocity measurement measurement at given at given x- andx- and y-coordinatesy-coordinates propagated propagated to uncertainty of the instantaneous velocity measurement at given x- and y-coordinates propagated to tothe the uncertainty uncertainty of ofthe the averaged averaged results results [43,44]. [43,44]. The The standard standard deviation deviation of of the the uncertainty uncertainty velocity the uncertainty of the averaged results [43,44]. The standard deviation of the uncertainty velocity magnitude waswas equalequal toto 0.14470.1447 m m/s/s at at P1 P1 and and 0.2397 0.2397 m m/s/s at at P4. P4. According According to to the the methodology methodology given given in magnitude was equal to 0.1447 m/s at P1 and 0.2397 m/s at P4. According to the methodology given Referencein Reference [46 [46],], the the uncertainty uncertainty of theof the mean mean velocity velocity of 150 of 150 samples samples was was equal equal to 0.0118 to 0.0118 m/s atm/s P1 at and P1 in Reference [46], the uncertainty of the mean velocity of 150 samples was equal to 0.0118 m/s at P1 0.0196and 0.0196 m/s atm/s P4. at The P4. selection The selection of points of points P1 and P1 P4 and for analyzingP4 for analyzing the uncertainty the uncertainty velocity velocity was dictated was and 0.0196 m/s at P4. The selection of points P1 and P4 for analyzing the uncertainty velocity was bydictated the fact by that the extremefact that flowextreme conditions flow conditions occurred occurred at these points. at these At points. point P1At (thepoint furthest P1 (the fromfurthest the dictated by the fact that extreme flow conditions occurred at these points. At point P1 (the furthest testfrom object), the test the object), lowest the uncertainties lowest uncertainties were expected, were andexpected, at point and P4 at (below point theP4 propeller),(below the thepropeller), highest from the test object), the lowest uncertainties were expected, and at point P4 (below the propeller), uncertaintiesthe highest uncertainties were expected. were expected. the highest uncertainties were expected. Figure 1111 presentspresents thethe RMS RMS velocity velocity profile profile as as a functiona function of of the thex-coordinate x-coordinate for forv = v12 = m12/ s,m/s,α = α0 ◦=, Figure 11 presents the RMS velocity profile as a function of the x-coordinate for v = 12 m/s, α = and0°, andy = y720 = 720 mm. mm. Figure Figure 12 shows12 shows the thevRMS vRMSprofiles profiles along along the thex-coordinate x-coordinate for for the the selected selected values values of 0°, and y = 720 mm. Figure 12 shows the vRMS profiles along the x-coordinate for the selected values theof they-coordinate. y-coordinate. of the y-coordinate.

Figure 11. RMS velocity profile as a function of the x-coordinate (vx—green line, vy—blue line). Figure 11. RMS velocity profileprofile as a function of thethe xx-coordinate-coordinate ((vvxx—green line, vy—blue line).

FigureFigure 12.12. TheThe vvRMSRMS profilesprofiles along along the the xx-coordinate-coordinate for for y = 100,100, 200, 200, 300, 300, and and 400 400 mm. mm. Figure 12. The vRMS profiles along the x-coordinate for y = 100, 200, 300, and 400 mm.

The experimental data acquisition and evaluation of the particle image recordings were The experimental data acquisition and evaluation of the particle image recordings were performed according to best practices (i.e., the selection of tracer particles, imaging optics, optimal performed according to best practices (i.e., the selection of tracer particles, imaging optics, optimal time between laser pulses, and seeding density listed in Reference [7]). The laser plane lens was time between laser pulses, and seeding density listed in Reference [7]). The laser plane lens was located outside the measurement space above the research object. Such a location caused the area to located outside the measurement space above the research object. Such a location caused the area to be obscured (covered), e.g., under the engine or (partially) by the propeller itself. At the stage of be obscured (covered), e.g., under the engine or (partially) by the propeller itself. At the stage of Sensors 2020, 20, 5537 11 of 22

The experimental data acquisition and evaluation of the particle image recordings were performed according to best practices (i.e., the selection of tracer particles, imaging optics, optimal time between laser pulses, and seeding density listed in Reference [7]). The laser plane lens was located outside the measurement space above the research object. Such a location caused the area to be obscured (covered), e.g., under the engine or (partially) by the propeller itself. At the stage of planning the experiment, alternative locations of the laser were considered. Locating the laser behind the research object would also lead to certain areas being obscured (in this case before the engine and the propeller). In addition, the laser plane lens would be placed inside the measurement space where the air flows along with the tracer particles. After analyzing possible variants of the laser plane lens settings, the variant described in this paper was chosen. The analysis of the images obtained during the tests was carried out using the DynamicStudio 6.11 software from Dantec Dynamics (Skovlunde, Denmark). The vector velocity fields for a pair of particle images were generated using the adaptive correlation function, which was based on the correlation analysis [18]. The final image size, after three steps of reducing the grid, was 64 64 px × with a 50% overlap of the image position in the grid. The obtained vector fields of the displacement in pixels were scaled using a calibration procedure using calibration photos. Therefore, it was possible to convert from pixels per second to meters per second. Post-processing was used to remove the individual vectors with values that exceeded physically obtainable speeds under the given conditions and to supplement places in the vector field where the fluid velocity was not determined. Thanks to the median and averaging filter, it was possible to conduct the data processing procedure.

3. Results and Analysis The tests were carried out for the following aircraft configuration: fuselage, without pusher propeller, with the main rotor, and with 1” diameter horizontal propellers operating at 100% of the available power (the measurement plane passed through the propeller axes with the vertical thrust vector). The front propeller was also working during the PIV measurements. The results were visualized on a plane passing through the axes of additional multirotor aircraft propellers. Figures 13–15 present the visualization of the aircraft flow around in the configuration described above for the angles of attack α equal to 0 , 4 , 8 , 12 , and 13.5 . The presented results are ◦ − ◦ − ◦ − ◦ − ◦ qualitative, while the quantitative analysis is presented in a further part of this paper. In the first phase of the tests, the distribution of velocities around the aircraft was checked for a 0◦ angle of attack in the absence of airflow in the wind tunnel and with airflow at a velocity of 12.5 m/s (Figure 14). In both cases, the propellers of the additional propulsion system operated at 9240 rpm. Figure 13 shows a view of the propeller located at the rear of the fuselage. In the case of no airflow for the 0◦ angle of attack, asymmetry was observed in the velocity values under the tested rotor disk. Due to the local flow disturbance, among others, resulting from the presence of the fuselage, higher velocity values in the stream behind the horizontal rotor occurred in the zone closer to the stabilizers. The flow through the rotor was almost vertical. The presence of airflow at a velocity of 12.5 m/s inside the tunnel meant that the stream behind the horizontal propeller was curved in the normal direction to the airflow direction at an angle of approximately 45◦. Similarly to the first case, asymmetry of the velocity value in the stream behind the horizontal propeller was observed. The next part of the research focused on analyzing the velocity distribution for several defined values of the angle of attack in the presence of airflow at a velocity of 12.5 m/s inside the wind tunnel. This angle had a negative value, which means that the body was pitched forward, just as during the forward flight scenario without an active main rotor. Sensors 2020, 20, x FOR PEER REVIEW 11 of 22

planning the experiment, alternative locations of the laser were considered. Locating the laser behind the research object would also lead to certain areas being obscured (in this case before the engine and the propeller). In addition, the laser plane lens would be placed inside the measurement space where the air flows along with the tracer particles. After analyzing possible variants of the laser plane lens settings, the variant described in this paper was chosen. The analysis of the images obtained during the tests was carried out using the DynamicStudio 6.11 software from Dantec Dynamics (Skovlunde, Denmark). The vector velocity fields for a pair of particle images were generated using the adaptive correlation function, which was based on the correlation analysis [18]. The final image size, after three steps of reducing the grid, was 64 × 64 px with a 50% overlap of the image position in the grid. The obtained vector fields of the displacement in pixels were scaled using a calibration procedure using calibration photos. Therefore, it was possible to convert from pixels per second to meters per second. Post-processing was used to remove the individual vectors with values that exceeded physically obtainable speeds under the given conditions and to supplement places in the vector field where the fluid velocity was not determined. Thanks to the median and averaging filter, it was possible to conduct the data processing procedure.

3. Results and Analysis The tests were carried out for the following aircraft configuration: fuselage, without pusher propeller, with the main rotor, and with 1’’ diameter horizontal propellers operating at 100% of the available power (the measurement plane passed through the propeller axes with the vertical thrust vector). The front propeller was also working during the PIV measurements. The results were visualized on a plane passing through the axes of additional multirotor aircraft propellers. Figures 13–15 present the visualization of the aircraft flow around in the configuration described above for the angles of attack α equal to 0°, −4°, −8°, −12°, and −13.5°. The presented results are qualitative, while the quantitative analysis is presented in a further part of this paper. In the first phase of the tests, the distribution of velocities around the aircraft was checked for a 0° angle of attack in the absence of airflow in the wind tunnel and with airflow at a velocity of 12.5 m/s (Figure 14). In both cases, the propellers of the additional propulsion system operated at 9240 rpm. Figure 13 shows a view of the propeller located at the rear of the fuselage. In the case of no airflow for the 0° angle of attack, asymmetry was observed in the velocity values under the tested rotor disk. Due to the local flow disturbance, among others, resulting from the presence of the fuselage, higher velocity values in the stream behind the horizontal rotor occurred in the zone closer to the stabilizers. The flow through the rotor was almost vertical. The presence of airflow at a velocity of 12.5 m/s inside the tunnel meant that the stream behind the horizontal propeller was curved in the Sensorsnormal2020 ,direction20, 5537 to the airflow direction at an angle of approximately 45°. Similarly to the first case,12 of 22 asymmetry of the velocity value in the stream behind the horizontal propeller was observed.

Sensors 2020, 20, x FOR PEER REVIEW 12 of 22

Figure 13. Velocity field around the horizontal propeller for the following configuration: angle of

attack α = 0°, without airflow (on the left) and with airflow vwt = 12.5 m/s (on the right).

The next part of the research focused on analyz ing the velocity distribution for several defined valuesFigure of the 13. angleVelocity of attack field around in the the presence horizontal of airf propellerlow at for a thevelocity following of 12.5 configuration: m/s inside angle the wind of attack tunnel. This angle had a negative value, which means that the body was pitched forward, just as during the α = 0◦, without airflow (on the left) and with airflow vwt = 12.5 m/s (on the right). forward flight scenario without an active main rotor. Figure 14 shows thethe distributiondistribution ofof thethe velocityvelocity fieldfield forfor thethe angle angle of of attack attackα α= = −44° andandα α= = −88° − ◦ − ◦ with airflow airflow inside inside the the wind wind tunnel. tunnel. In In the the case case of ofα =α −=4° compared4 compared to the to the0° angle 0 angle of attack, of attack, the − ◦ ◦ maximumthe maximum velocity velocity area area shifted shifted to tothe the left left toward toward th thee stabilizers. stabilizers. At At the the end end of of this this zone, zone, a a turbulent turbulent area was createdcreated inin whichwhich thethe velocityvelocity vectors vectors were were not not arranged arranged in in an an orderly orderly manner. manner. An An increase increase in inthe the angle angle of of attack attack up up to toα =α = 8−8°resulted resulted in in a a disappearance disappearance ofof thethe characteristiccharacteristic areaarea of maximum − ◦ velocity. However,However, therethere remained remained a a high-velocity high-velocity area area between between the the end end of theof the propeller propeller blade blade and and the thestabilizers stabilizers of the of aircraft.the aircraft.

Figure 14.14. VelocityVelocity field field around around the the horizontal horizontal propeller propelle forr thefor followingthe following configuration: configuration: angle angle of attack of attackα = 4 α (on= −4° the (on left) the and left) angle and of angle attack ofα attack= 8 α(on = − the8° right),(on the with right), an airflowwith an velocity airflow ofvelocityvwt = 12.5 of v mwt/ s.= − ◦ − ◦ 12.5 m/s. Figure 15 shows a distribution of the velocity field for increased values of the angle of attack, α = Figure12 and 15α shows= 13.5 a distributionrelative to the of the airflow velocity inside field the for tunnel. increased For an values angle of the 12 ,angle the characteristic of attack, α − ◦ − ◦ ◦ =area −12° of and increased α = −13.5° velocity relative again to the appeared airflow behindinside the the tunne end ofl. For the an rear angle blade. of 12°, The the formation characteristic of an areaadditional of increased area of velocity increased again velocity appeared in the behind form of the a parallel end of zonethe rear near blade. the engine The formation head should of an be additionalemphasized. area An of analysis increased of thevelocity velocity in the distribution form of a for parallel the angle zone of near attack th equale engine to head13.5 shouldindicated be − ◦ emphasized. An analysis of the velocity distribution for the angle of attack equal to −13.5° indicated an increase in the maximum velocity value behind the propeller. In the front of the rotor disk (viewed from the front of the fuselage), the velocity was close to the velocity of the airflow. This meant that the largest velocity field gradient occurred toward the stabilizers. In both considered cases, a turbulent area was formed at the end of the stream. Sensors 2020, 20, 5537 13 of 22 an increase in the maximum velocity value behind the propeller. In the front of the rotor disk (viewed from the front of the fuselage), the velocity was close to the velocity of the airflow. This meant that the largest velocity field gradient occurred toward the stabilizers. In both considered cases, a turbulent areaSensors was 2020 formed, 20, x FOR at PEER the endREVIEW of the stream. 13 of 22

Figure 15. Velocity fieldfield around thethe horizontalhorizontal propellerspropellers forfor thethe followingfollowing configuration:configuration: angle of attack α = −12°12 (on(on the the left) left) and and angle angle of of attack attack α α= =−13.5°13.5 (on(on the the right), right), with with an anairflow airflow velocity velocity of v ofwt − ◦ − ◦ v=wt 12.5= 12.5m/s. m /s.

Based onon thethe teststests performed performed using using the the PIV PIV technique technique for for the the considered considered configuration configuration of the of testthe object,test object, the airflow the airflow velocity velocity profiles profiles were developed were developed as a function as a function of the distance of the fromdistance the propellers. from the Onpropellers. this basis, On it this was basis, possible it was to conductpossible a to quantitative conduct a quantitative analysis aimed analysis at determining aimed at thedetermining impact range the ofimpact an additional range of an multirotor additional propulsion multirotor system. propulsion The system. analysis The of theanalysis velocity of the field velocity around field the around tested objectthe tested made object it possible made it to possible determine to thedetermine mounting the locationmounting (height) location of the(height) main of rotor the workingmain rotor in autorotationworking in autorotation relative to the relative fuselage to the and fuselage other components and other components to minimize to the minimize impact the of the impact additional of the propulsionadditional propulsion system. Figures system. 16 –Figures23 show 16–23 the course show ofthe the course resultant of the airflow resultant velocity airflow and velocity its horizontal and its andhorizontal vertical and components vertical components around the around considered the considered object in object the tested in the aircraft tested configurationaircraft configuration for the selectedfor the selected values values of the angleof the ofangle attack of attackα equal α equal to 0◦, to –4 0°,◦, and –4°, –12and◦ .–12°. The The air velocity air velocityv in v the in the tunnel tunnel for thefor the presented presented configurations configurations was was 0 m /0s m/s and and 12.5 12.5 m/s. m/s. When analyzing the the resultant resultant velocity velocity for for the the angle angle of of attack attack α α= =0°0 in◦ inthe the absence absence of ofairflow airflow in inthe the tunnel, tunnel, it was it was observed observed that that the thehighest highest velocity velocity value value of approximately of approximately 21 m/s 21 moccurred/s occurred for the for thecoordinate coordinate x = 800x = mm,800 mm,i.e., for i.e., the for center the center of the of length the length of the ofrear the blade. rear blade.The highest The highest values for values the forconsidered the considered value of value x were of recordedx were recorded for y, ranging for y, from ranging 20 to from 140 20mm. to For 140 the mm. advancing For the advancing blade, the blade,maximum the maximumvelocity value velocity in the value area in just the areabelow just its below center its of center length of lengthwas lower was lowerand amounted and amounted to 18 tom/s. 18 This m/s. Thisconfirmed confirmed the asymmetry the asymmetry of the of thevelocity velocity field field distribution distribution in in the the stream stream behind behind the horizontal propeller described above.above. For the distance y = 140140 mm, there was a sudden change in the resultant velocityvelocity comparedcompared to,to, e.g.,e.g.,y y= = 20, 60, or 8080 mm.mm. It is worth mentioning thatthat yy == 140 mm was the levellevel atat which which the the propeller propeller was was located. located. The The highest highest flow flow velocity velocity value value was achieved was achieved by the streamby the locatedstream approximatelylocated approximately 20 mm from 20 themm plane from of the the propeller.plane of Forthe thepropeller.x-coordinate For correspondingthe x-coordinate to thecorresponding rotor axis, a to reduction the rotor in axis, the a resultant reduction velocity in the valueresultant was velocity observed. value was observed. Sensors 2020, 20, x FOR PEER REVIEW 14 of 22 Sensors 2020, 20, 5537 14 of 22 Sensors 2020, 20, x FOR PEER REVIEW 14 of 22

Figure 16. Resultant airflow velocities around the research object for the following configuration: Figureangle of 16. attack ResultantResultant α = 0° andairflow airflow airflow velocities velocities in the aroundwind around tunnel the the research vwt = 0 m/s. object for the following configuration: configuration:

angle of attack α = 0°0◦ andand airflow airflow in in the the wind wind tunnel tunnel vvwtwt == 0 0m/s. m/s. The presence of the forced airflow in the tunnel with a velocity of 12.5 m/s resulted in a significantThe presencepresence change of inof the thethe forced recordedforced airflow airflow velocity in the in tunnelprofilesthe tunne with (Figurel awith velocity 17).a velocity In of 12.5the mareaof/s 12.5 resulted in frontm/s in resultedof a significantthe tested in a changesignificantpropeller in ( the xchange from recorded 1000 in the velocityto 1100recorded mm) profiles velocityand (Figure behind profiles 17 it). near In (Figure the the area stabilizers 17). in front In the ( ofx from thearea tested 550in front to propeller 650 of mm), the ( xtested therefrom 1000propellerwas a to strong 1100 (x from mm)flow 1000disturbance and behindto 1100 zone, itmm) near asand the seen behind stabilizers by the it flneaructuations (x from the stabilizers 550 in to the 650 velocity ( mm),x from therevalues. 550 wasto At650 aa strongsufficientlymm), there flow disturbancewaslarge a distancestrong zone, flow from asdisturbance seenthe plane by the zone,of fluctuations the as te seensted by inpropeller thethe velocityfluctuations of the values. aircraft, in the At velocity a i.e., suffi yciently > values. 200 largemm, At distanceathe sufficiently resultant from thelargevelocity plane distance was of the approximately from tested the propeller plane 15 of m/s. thethe A aircraft,te slightlysted propeller i.e., highery > 200 value of mm, the than theaircraft, resultantthe set i.e., air velocityy velocity > 200 wasmm, inside approximately the the resultant tunnel 15velocitywas m /dues. A was slightlyto approximatelyinterference higher value with 15 thanm/s.the A theaircraft slightly set air fusela velocityhigherge value insideelements than the theand tunnel set the air was presence velocity due to interferenceinsideof a localthe tunnel withflow thewasdisturbance. aircraft due to fuselage Nevertheless,interference elements with in and the the the area aircraft presence of x fuselaranging of a localge fromelements flow 600 disturbance. mmand tothe 750 Nevertheless,presence mm, two of characteristic ina local the area flow of xdisturbance.extremesranging exceeding from Nevertheless, 600 mm 20 m/s to 750 werein mm,the distinguished. area two characteristicof x ranging They corresponded extremesfrom 600 exceedingmm to to a cutting750 20 mm, m/ splane were two at distinguished.characteristic distances of Theyextremes20 mm corresponded and exceeding 60 mm. to20This am/s cutting means were plane distinguished.that the at distances area of They the of 20maximumcorresponded mm and airflow 60 mm.to a cuttingvelocity This means plane shifted that at distancesfrom the area the ofx- the20coordinate mm maximum and corresponding 60 airflowmm. This velocity means to the shifted centerthat the fromof thearea the rear ofx -coordinatetheblade maximum to the correspondingcoordinate airflow velocitycorresponding to the shifted center to from ofa distance the the rear x- bladecoordinateof 50 to to 100 the corresponding mm coordinate behind correspondingits to end. the center toof athe distance rear blade of 50 to to the 100 coordinate mm behind corresponding its end. to a distance of 50 to 100 mm behind its end.

Figure 17. Resultant airflow airflow velocities velocities around around the the research object for the following configuration:configuration: angleFigure of 17. attack Resultant α = 0°0◦ andairflowand airflow airflow velocities in in the the windaround wind tunnel the research vwt == 12.512.5 object m/s. m/s. for the following configuration:

angle of attack α = 0° and airflow in the wind tunnel vwt = 12.5 m/s. The horizontal velocity component for the angle of attack α = 4 with the airflow inside the The horizontal velocity component for the angle of attack α = −−4°◦ with the airflow inside the tunnel is shown in Figure 18. At a sufficiently large distance from the tested propeller, the absolute tunnelThe is horizontalshown in Figure velocity 18. component At a sufficiently for the large angle distance of attack from α =the −4° tested with propeller,the airflow the inside absolute the tunnel is shown in Figure 18. At a sufficiently large distance from the tested propeller, the absolute Sensors 2020, 20, 5537 15 of 22 Sensors 2020, 20, x FOR PEER REVIEW 15 of 22 value of this flowflow velocity component was close to 15 m/s. m/s. Simila Similarly,rly, as in the cases described above, thethe didifferencefference betweenbetween thethe given velocity of 12 mm/s/s was due to the eeffectffect of aerodynamic interference withwith thethe restrest ofof thethe aircraftaircraft parts.parts. As in the case of a 00°◦ angle of attack for the x-coordinate in the range fromfrom 550550 toto 650650 mmmm andand fromfrom 10001000 toto 11001100 mm,mm, therethere was a strong flowflow disturbancedisturbance that resulted in thethe fluctuationfluctuation ofof thethe measuredmeasured componentcomponent value.value. In the vicinity of the propeller (x equal to 800 to 950 mm),mm), thethe horizontal horizontal velocity velocity component component values values for for the the cutting cutting planes planes located located a short a short distance distance from thefrom propeller the propeller plane plane (y = 140–160 (y = 140–160 mm) mm) were were recorded recorded as decreasing as decreasing to approximately to approximately zero. zero. This This was duewas todue the to intense the intense propeller propeller action, action, which which generated generated a flow a flow perpendicular perpendicular to the to mainthe main flow flow in the in tunnel.the tunnel. The The flow flow generated generated by by the the propeller propeller in in this this region region was was perpendicular perpendicular to to thethe measurementmeasurement plane,plane, thus a significantsignificant loss of the particle pairs occurred. At the same time, for these cross-sections, an asymmetryasymmetry ofof thisthis velocityvelocity component component was was observed observed relative relative to to the the propeller propelle center,r center, i.e., i.e., relative relative to theto the coordinate coordinatex = x880 = 880 mm. mm.

Figure 18.18.Horizontal Horizontal components components of theof airflowthe airflow velocities veloci aroundties around the research the research object for object the following for the

configuration:following configuration: angle of attack angleα of= attack4 and α = airflow −4° and in airflow the wind in tunnelthe windvwt tunnel= 12.5 v mwt /=s. 12.5 m/s. − ◦ The vertical component for the angle of attack α = 4 with the airflow in the tunnel is shown The vertical component for the angle of attack α = −4°− with◦ the airflow in the tunnel is shown in inFigure Figure 19. 19 Similar. Similar to tothe the horizontal horizontal component, component, at atthe the beginning beginning and and end end of ofthe the analyzed analyzed area, area, a acharacteristic characteristic flow flow disturbance disturbance zone zone was was distinguished, distinguished, in in which which the the velocity velocity values for individual components werewere subject subject to intenseto intense fluctuations. fluctuations As. the As cutting the cutting planes planes moved awaymoved from away the from propeller the plane,propeller the plane, velocity the componentvelocity component value approached value appr zero.oached The zero. zero The value zero was value not was reached not reached due to due an aerodynamicto an aerodynamic interference interference with the wi aircraftth the aircraft fuselage fuselage components. components. The characteristic The characteristic shape had shape a velocity had a forvelocity the cutting for the plane cutting at aplane distance at a ofdistancey = 140 of mm, y = corresponding140 mm, corresponding to the plane to the of the plane tested of the propeller. tested Apropeller. change A in change the sign in ofthe the sign velocity of the velocity value for value this for plane this plane was observed was observed after after passing passing through through the propellerthe propeller axis, axis, i.e., thei.e., coordinate the coordinatex = 880 x = mm. 880 Initially,mm. Initially, a positive a positive value ofvalue 6 m/ sof decreased 6 m/s decreased and reaches and a velocity of 13 m/s on the other side of the rotor disk. Extreme velocity values were recorded for the reaches a velocity− of −13 m/s on the other side of the rotor disk. Extreme velocity values were recorded cross-sections y = 20 mm and y = 60 mm, which were approximately 13.5 m/s. These values occurred for the cross-sections y = 20 mm and y = 60 mm, which were approximately− −13.5 m/s. These values approximatelyoccurred approximately 100 mm behind 100 mm the behind end of the the end rear of blade. the rear blade. Sensors 2020, 20, 5537 16 of 22 Sensors 2020, 20, x FOR PEER REVIEW 16 of 22

Figure 19. Vertical components of the airflowairflow velocities around the research object for the following configuration:configuration: angle of attack α = −4°4◦ andand airflow airflow in in the the wind wind tunnel tunnel vvwtwt = =12.512.5 m/s. m/ s. − With both components, it was possible to determine the resultant velocity for the considered configurationconfiguration (Figure 20).20). The The velocity values obtained as a result of the calculations had a positive value for the whole considered rangerange ofof xx-- andand yy-coordinates.-coordinates. ExtremeExtreme valuesvalues occurredoccurred forfor yy == 20 mm and y = 60 mm and were equalequal toto 22.522.5m m/s/s and and 23 23 m m/s,/s, respectively. respectively. The The lowest lowest velocity velocity values values close close to zeroto zero were were recorded recorded for for the the area area corresponding corresponding to the to propellerthe propeller axis, axis, i.e., xi.e.,from x from 850 to 850 900 to mm 900 andmm forandx equalfor x equal to 1000 to to1000 1100 to mm,1100 i.e.,mm, for i.e., the for area the just area before just before the end the of end the frontof the blade. front blade. Similarly Similarly for α = for4 α, for x equal to 1000 to 1100 mm, i.e., for the area just before the end of the front blade. Similarly for− α◦ as= − in4°, theas in case the ofcaseα = of0 ◦α, = the 0°, maximum the maximum value value was was reached reached at a at distance a distance from from the the propeller propeller equal equal to approximatelyto approximately 80–120 80–120 mm. mm.

Figure 20.20. ResultantResultant airflowairflow velocities velocities around around the the research research object object for thefor following the following configuration: configuration: angle ofangle attack of attackα = 4 α =and −4° airflow and airflow in the in wind the wind tunnel tunnelvwt = v12.5wt = 12.5 m/s. m/s. angle of attack− α◦ = −4° and airflow in the wind tunnel vwt = 12.5 m/s. Figure 21 presents the horizontal components of the airflow velocities around the research object Figure 21 presents the horizontal components of the airflow velocities around the research object for the angle of attack α = 12◦ and the airflow velocity in the tunnel vwt = 12.5 m/s. Similarly to a for the angle of attack α = −−12° and the airflow velocity in the tunnel vwt = 12.5 m/s. Similarly to a smaller angle of attack in the area between the propeller and the stabilizers, there was a strong flow smaller angle of attack in the area between the propeller and the stabilizers, there was a strong flow disturbance area in which the component of the horizontalhorizontal velocity had a negative sign and locally x exceeded thethe absoluteabsolute value value of of 20 20 m m/s/s (for (forequal x equal to 550to 550 to 700 to 700 mm). mm). The The velocity velocity value value fluctuated fluctuated from from zero to −15 m/s. A strong flow disturbance also occurred in the area in front of the propeller (x Sensors 2020, 20, 5537 17 of 22

Sensors 2020, 20, x FOR PEER REVIEW 17 of 22 zero to 15 m/s. A strong flow disturbance also occurred in the area in front of the propeller (x equal to − 1000equal to to 1100 1000 mm), to 1100 except mm), that except the velocity that the value velocity oscillated value from oscillated5 to 5 from m/s. At−5 ato greater 5 m/s. distance At a greater from − thedistance rotor from (y > 200the mm)rotor the(y > velocity 200 mm) stabilized the velocity and wasstabilized approximately and was approximately15 m/s. The slightly −15 m/s. higher The − velocityslightly higher than the velocity set airflow than velocitythe set airflow was due velocity to the was effect due of localto the air effect acceleration of local air in theacceleration area of the in aircraftthe area fuselage. of the aircraft The increase fuselage. in The aircraft increase pitch in (reduction aircraft pitch of the (reduction angle of attack of the from angle 4of toattack12 from) did − ◦ − ◦ not−4° qualitativelyto −12°) did changenot qualitatively the distribution change of the the distributi vertical velocityon of the component vertical velocity in the areas component behind andin the in frontareas ofbehind the propeller. and in front In both of the cases, propeller. a characteristic In both cases, flow disturbancea characteristic zone flow was disturbance present. The zone largest was significantpresent. The di largestfference sign occurredificant fordifference the cross-section occurred fory = the20 mmcross-section in the immediate y = 20 mm vicinity in the ofimmediate the rotor diskvicinity for ofx fromthe rotor 700 todisk 950 for mm. x from In this700 to case, 950 the mm. increase In this incase, the the pitch increase angle in of the the pitch aircraft angle caused of the a decreaseaircraft caused in the horizontala decrease component in the horizontal from 15 component m/s to 8 mfrom/s, which −15 m/s indicated to −8 am/s, decrease which in indicated the impact a − − ofdecrease the forced in the airflow impact on of the the value forced of airflow the horizontal on the componentvalue of the inhorizontal this zone. component in this zone.

Figure 21.21.Horizontal Horizontal components components of theof airflowthe airflow velocities veloci aroundties around the research the research object for object the following for the configuration:following configuration: angle of attack angleα of= attack12 and α = airflow−12° and in airflow the wind in tunnelthe windvwt tunnel= 12.5 v mwt /=s. 12.5 m/s. − ◦ In thethe casecase of of the the vertical vertical velocity velocity components components (Figure (Figure 22), 22), there there were were also twoalso characteristictwo characteristic areas ofareas flow of disturbanceflow disturbance occurring occurring in front in front of and of behindand behind the propeller. the propeller. At a At greater a greater distance distance from from the propellerthe propeller (y > (y200 > 200 mm), mm), the the velocity velocity stabilized stabilized for for almost almost the the entire entirex x-coordinate-coordinate range. range. TheThe furtherfurther away from the the propeller, propeller, the the closer closer to to zero zero the the velo velocitycity value value became. became. A slight A slightlyly negative negative value value of the of thecomponent component (from (from −1 to −15 tom/s)5 resulted m/s) resulted from the from induction the induction of velocity of velocity due to an due interference to an interference with the − − withaircraft the fuselage. aircraft fuselage.Compared Compared to the configuration to the configuration without airflow without in airflowthe wind in tunnel the wind for α tunnel = 0°, the for αextreme= 0◦, the velocities extreme in velocities the stream in the behind stream the behind horizontal the horizontal propeller propeller ranged from ranged −13 fromto −10 13m/s. to Around10 m/s. − − Aroundthe propeller the propeller axis (x = axis880 mm), (x = 880 the mm),velocity the component velocity component decreased decreased and oscillated and oscillatedaround −5 aroundm/s. A particular5 m/s. A particulartrend was trendobserved was for observed the coordinate for the coordinate y = 140 mm,y = which140 mm, corresponded which corresponded to the propeller to the − propellerplane. As plane.the x-coordinate As the x-coordinate increased (moving increased toward (moving the towardfront of the the frontaircraft), of the the aircraft), vertical thecomponent vertical componentchanged sign changed after passing sign after through passing the through propeller the axis. propeller The positive axis. The values positive were values the result were of the the result flow ofdisturbance the flow disturbance present on the present propeller on the blade. propeller The increa blade.se The in the increase pitch inangle the pitchof the angle aircraft of theresulted aircraft in resultedan increase in an in increasethe vertical in the component vertical component fluctuations fluctuations for the cross-sections for the cross-sections close to the close plane to theof the plane rotor of thedisk rotor in the disk propeller in the propellerregion (x regionfrom 550 (x fromto 750 550 mm). to 750This mm). effect This was e identicalffect was to identical the effect to theobserved effect observedfor the horizontal for the horizontal component. component. In addition, In addition, a significant a significant change change in this in velocity this velocity component component was wasobserved observed for the for cross-section the cross-section y = 140y = mm,140 mm,which which corresponded corresponded to the to initial the initial plane plane of the of rotor the rotordisk. disk.The velocity The velocity on the on side the of side the of front the front blade blade was wasthenthen closer closer to zero. to zero. For Forthis thisregion, region, the thevelocity velocity for forcross-sections cross-sections in the in theclose close vicinity vicinity of the of thepropeller propeller also also slightly slightly decreased, decreased, i.e., i.e., for fory = y60= mm60 mm and and y = y100= 100mm. mm. Sensors 2020, 20, 5537 18 of 22 Sensors 2020, 20, x FOR PEER REVIEW 18 of 22

Sensors 2020, 20, x FOR PEER REVIEW 18 of 22

FigureFigure 22. Vertical22. Vertical components components of of the the airflow airflow velocityvelocity around around the the research research object object for forthe thefollowing following configuration: angle of attack α = −12° and airflow in the wind tunnel vwt = 12.5 m/s. configuration: angle of attack α = 12◦ and airflow in the wind tunnel vwt = 12.5 m/s. Figure 22. Vertical components −of the airflow velocity around the research object for the following Finally,configuration:Finally, the the resulting resulting angle of airflow attack airflow α velocity= −velocity12° and wasairflowwas calculatedcalculated in the wind for fortunnel the the vconfigurationwt configuration = 12.5 m/s. considered considered below below (Figure 23). Comparing the obtained values for the angle of attack α = −4° and α = −12°, their greater (Figure 23). Comparing the obtained values for the angle of attack α = 4◦ and α = 12◦, their greater dynamicsFinally, was the observed resulting in airflowthe case velocity of a larger was pitc calculatedh angle offor the the aircraft. configuration− Therefore, considered the− curves below had dynamics was observed in the case of a larger pitch angle of the aircraft. Therefore, the curves (Figureirregular 23). shapes Comparing with numerous the obtained local values extremes. for the Th angleis was of attackthe effect α = − of4° andswirling α = − 12°,due theirto the greater flow had irregular shapes with numerous local extremes. This was the effect of swirling due to the flow dynamicsseparation was at an observed increased in angle the case of attack. of a larger The maximum pitch angle velocity of the valuesaircraft. did Therefore, not change the significantly curves had separationandirregular ranged at anshapes from increased 22.5with m/s numerous angle to 23.5 of attack. m/slocal for extremes. Thethe area maximum inTh theis wasstream velocity the behind effect values theof swirling didhorizontal not changedue propeller. to the significantly flow The and rangedlargestseparation difference from at an 22.5 increased was m observed/s to angle 23.5 forof m attack. the/s forcross-section The the maximum area y in = 100 the velocity mm, stream invalues which behind did an not extreme the change horizontal value significantly of about propeller. The largest22and m/s ranged was difference observedfrom 22.5 was form/s observed a tolarger 23.5 m/spitch for for theangle the cross-section areaof the in aircraft.the streamy =The100 behind velocity mm, the inin horizontal whichthe area an of propeller. extremethe blade Thevalue tip of aboutalsolargest 22 mintensified,/ sdifference was observed i.e., was for observed x for = 750 a larger mmfor the for pitch cross-section the anglecross-section of y the = 100 aircraft.y = mm, 20 mm. in The which This velocity isan seen extreme in inthe Figure value area 15of of inabout the the blade tip alsoform22 intensified,m/s of was an additional observed i.e., for forhigh-velocityx a= larger750 mm pitch area for angle that the was cross-sectionof the not aircraft. clearlyy visibleThe= 20 velocity mm.for the Thisin smaller the is area seen value of in the of Figure theblade angle 15 tip in the formalsoof of attack an intensified, additional (Figure i.e.,14). high-velocity forIn additionx = 750 mm to area the for above-mentioned that the cross-section was not clearly fluctuations,y = 20 visible mm. This forthe thelargest is seen smaller velocity in Figure value differences 15 of in the the angle of attackformwere (Figure ofobserved an additional 14 for). Incross-section addition high-velocity to fragments the area above-mentioned that runn wasing not directly clearly fluctuations, under visible the for propeller, the the smaller largest i.e., value for velocity x offrom the di 750anglefferences to 1000of attack mm (Figure and y from 14). In20 additionmm to 60 to mm. the above-mentionedThe increase in th efluctuations, pitch angle theof the largest propeller velocity plane differences resulted were observed for cross-section fragments running directly under the propeller, i.e., for x from 750 to inwere the observed most significant for cross-section change in fragments flow conditions running in directly the area under just below the propeller, the rotating i.e., blades.for x from 750 to 1000 mm1000 andmm yandfrom y from 20 mm 20 mm to 60 to mm.60 mm. The The increase increase in in the the pitchpitch angleangle of of the the propeller propeller plane plane resulted resulted in the mostin the significant most significant change change in flow in flow conditions conditions in thein the area area just just below below the the rotatingrotating blades. blades.

Figure 23. Resultant airflow velocities around the research object for the following configuration: angle of attack α = −12° and airflow in the wind tunnel vwt = 12.5 m/s. FigureFigure 23. Resultant 23. Resultant airflow airflow velocities velocities around around the the research research object object for for the the following following configuration: configuration: angle of attackangleα of= attack12 andα = − airflow12° and airflow in the windin the tunnelwind tunnelvwt = v12.5wt = 12.5 m/s. m/s. − ◦ Sensors 2020, 20, 5537 19 of 22

4. Discussion and Conclusions This paper presents the results of field velocity tests around a hybrid aircraft. The PIV technique was used to record the velocity values. The configurations with airflow forced by a fan inside the wind tunnel and without airflow were analyzed. The range of impact of one of the additional horizontal propellers and the impact of the aircraft fuselage on the recorded velocity field were analyzed. In addition, how the angle of attack affected the vertical and horizontal components and the resultant air velocity around additional propellers installed on the aircraft was also investigated. A significant impact of the airflow on the velocity distribution around the tested propeller was observed. In the absence of airflow, there were two characteristic extremes surrounded by the centers of the front and rear blades. The resultant velocity value for the front blade was lower than the value for the rear blade by approximately 3 m/s. The observed asymmetry was the result of an aerodynamic interference with individual elements of the aircraft. In the case of airflow in the zone in front of the propeller and in the stream behind it, a strong flow disturbance zone was created, which caused fluctuations of the resultant velocity value for all considered cross-sections. Importantly, the area of the highest velocity was shifted from the center of the blades toward the stream behind the horizontal propeller. The increase in the angle of attack did not significantly affect the vertical and horizontal components. As for the 0◦ angle of attack, the presence of the airflow caused a strong flow disturbance zone in front of and behind the propeller. The largest differences, in the form of an intensification of the velocity value fluctuations, occurred in the case of cross-sections located close to the rotor disk (y equal to 20 mm and 60 mm) in the stream behind the horizontal propeller (x equal to 550 to 750 mm) and in the area directly below the rotor disk (x equal to 750 to 900 mm). It should be mentioned that the y position of the disk changed with the angle of attack. The set flow velocity in the tunnel (12 m/s) significantly differed from the velocity recorded in undisturbed cross-sections (15 m/s). This was the effect of aerodynamic interference, which should be taken into account when analyzing the velocity and pressure field around the research object. The impact of the propeller ahead of its disk was definitely weaker than behind it. At a distance of 40 mm behind the propeller, the air reached a velocity of approximately 19.5 m/s, while at the same distance before it, the velocity was 7.5 m/s. When the velocity behind the propeller at a distance of 80 mm increased to 21 m/s, then the velocity in front of the propeller decreased to 5 m/s. At a distance of 120 mm behind the propeller, the maximum velocity value was also approximately 21 m/s, while at the same distance in front of the propeller, the maximum velocity value was 4 m/s. Up to this distance, the velocity behind the propeller increased with the distance, whereas in front of the propeller, this relationship was the opposite. A gradual decrease in velocity was observed as the distance increased. This was an obvious tendency, but without research, it is impossible to quantify the range of this impact. At a height of y = 380 mm, the maximum resultant velocity was less than 2 m/s. This was the height at which the main rotor was mounted (according to the additional reference system). This situation occurred without the main airflow in the wind tunnel. In the presence of airflow in the tunnel, the impact was further reduced before the propeller. This was a positive effect because it confirmed the validity of the concept of the main drive unit supported by additional propellers operating in the horizontal plane. The main rotor operation was not disturbed by additional propellers.

Author Contributions: Conceptualization, Z.C.; methodology, Z.C., W.S. and P.K.; software, W.S. and Z.C.; validation, Z.C. and W.S.; formal analysis, Z.C. and P.K.; investigation, Z.C. and P.K.; resources, Z.C. and W.S.; writing—original draft preparation, Z.C. and P.K.; writing—review and editing, Z.C. and P.K.; visualization, Z.C., W.S. and P.K.; supervision, Z.C.; project administration, Z.C.; funding acquisition, Z.C. All authors have read and agreed to the published version of the manuscript. Funding: This work was financed by the Polish National Centre for Research and Development under the LIDER program, grant agreement no. LIDER/27/0140/L-10/18/NCBR/2019. Sensors 2020, 20, 5537 20 of 22

Acknowledgments: We would like to offer our sincere thanks to the Institute of Aviation in Warsaw (Poland) for agreeing to conduct tests in their wind tunnel. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Czy˙z,Z.; Łusiak, T.; Magryta, P. Numerical analysis of the horizontal stabilizer influence on the Gyroplane Aerodynamic characteristics. Trans. Inst. Aviat. 2013, 232, 3–14. [CrossRef] 2. Bogda´nski,K.; Rodzewicz, M.; Ruchała, P. Characteristics of locked and free-wheeling ducted fan based on wind tunnel test and CFD analyses. In Proceedings of the 5th CEAS Air & Space Conference, Delft, The Netherlands, 7–11 September 2015. 3. Versteeg, K.K.; Malalasekera, W. An introduction to Computational Fluid Dynamics; Person Education Limited: London, UK, 2007. 4. Wilox, D.C. Turbulence Modeling for CFD, 3rd ed.; DCW Industries: La Canada, CA, USA, 2006. 5. Czy˙z,Z.; Siadkowska, K. Measurement of Air Flow Velocity around the Unmanned Rotorcraft. In Proceedings of the 2020 IEEE International Workshop on Metrology for AeroSpace (MetroAeroSpace), Pisa, Italy, 22–24 June 2020. 6. Hamdi, M.; Havet, M.; Rouaud, O.; Tarlet, D. Comparison of different tracers for PIV measurements in EHD airflow. Exp. Fluids 2014, 55, 1702. [CrossRef] 7. Raffael, M.; Willert, C.E.; Scarano, F.; Kaehler, C.J.; Wereley, S.T.; Kompenhans, J. Physical and Technical Background. In Particle Image Velocimetry, 3rd ed.; Springer: Berlin/Heidelberg, Germany, 2018; pp. 33–111. 8. Raffel, M.; Bauknecht, A.; Ramasamy, M.; Yamauchi, G.; Heineck, J.; Jenkins, L. Contributions of Particle Image Velocimetry to Helicopter Aerodynamics. AIAA J. 2017, 55, 1–16. [CrossRef] 9. Stryczniewicz, W.; Surmacz, K. PIV measurements of the vortex ring state of the main rotor of a helicopter. Trans. Inst. Aviat. 2014, 2, 17–27. [CrossRef] 10. Tanahashi, M.; Hirayama, T.; Taka, S.; Miyauchi, T. Measurement of fin scale structure in turbulence by time-resolved dual-plane stereoscopic PIV. Int. J. Heat Fluid Flow 2008, 29, 792–802. [CrossRef] 11. Surmacz, K.; Ruchała, P.; Stryczniewicz, W. Wind tunnel tests of the development and demise of Vortex Ring State of the rotor. In Advances in Mechanics: Theoretical, Computational and Interdisciplinary Issues, Proceedings of the 3rd Polish Congress of Mechanics, PCM and the 21st International Conference on Computer Methods in Mechanics, CMM, Gdansk, Poland, 8–11 September 2015; Taylor & Francis Ltd: London, UK, 2016. 12. Nathan, N. The Rotor Wake in Ground Effect and its Investigation in a Wind Tunnel. Ph.D. Thesis, University of Glasgow, Glasgow, UK, 2010. 13. Chen, F.; Liu, H.; Yang, Z.; Hu, H. Tracking characteristics of tracer particles for PIV measurements in supersonic flows. Chin. J. Aeronaut. 2017, 30, 577–585. [CrossRef] 14. Scarano, F. Overview of PIV in supersonic flows. In Particle Image Velocimetry; Springer: Berlin/Heidelberg, Germany, 2007; pp. 445–463. 15. Haertig, J.; Havermann, M.; Rey, C.; George, A. Particle image velocimetry in Mach 3.5 and 4.5 shock-tunnel flows. AIAA J. 2002, 40, 1056–1060. [CrossRef] 16. Stryczniewicz, W. Application of an inverse method for improvement of the measurement capability of the particle image velocimetry method. Trans. Inst. Aviat. 2016, 2, 93–107. [CrossRef] 17. Pust, O. PIV: Direct cross-correlation compared with FFT-based cross-correlation. In Proceedings of the 10th International Symposium on Applications of Laser Techniques to Fluid Mechanics, Lisbon, Portugal, 10–13 June 2000; p. 114. 18. Stryczniewicz, W. Development of Particle Imagine Velocimetry Algorithm. Probl. Mechatroniki 2012, 3, 41–54. 19. Nacakli, Y.; Landman, D. Helicopter downwash/frigate airwake interaction flowfield PIV surveys in a low speed wind tunnel. In Annual Forum Proceedings; American Helicopter Society: Virginia Beach, CA, USA, 2011; Volume 4, pp. 2988–2998. 20. Caridi, G.C.A.; Ragni, D.; Sciacchitano, A.; Scarano, F. HFSB-seeding for large-scale tomographic PIV in wind tunnels. Exp. Fluids 2016, 57, 190. [CrossRef] 21. Scarano, F.; Ghaemi, S.; Caridi, G.C.A.; Bosbach, J.; Dierksheide, U.; Sciacchitano, A. On the use of helium-filled soap bubbles for large-scale tomographic PIV in wind tunnel experiments. Exp. Fluids 2015, 56, 42. [CrossRef] Sensors 2020, 20, 5537 21 of 22

22. Beresh, S.; Kearney, S.; Wagner, J.; Guildenbecher, D.; Henfling, J.; Spillers, R.; Pruett, B.; Jiang, N.; Slipchenko, M.; Mance, J.; et al. Pulse-burst PIV in a high-speed wind tunnel. Meas. Sci. Technol. 2015, 26, 095305. [CrossRef] 23. Dobrev, I.; Massouh, F. CFD and PIV investigation of unsteady flow through Savonius wind turbine. Energy Procedia 2011, 6, 711–720. [CrossRef] 24. Ramponi, R.; Blocken, B. CFD simulation of cross-ventilation flow for different isolated building configurations: Validation with wind tunnel measurements and analysis of physical and numerical diffusion effects. J. Wind Eng. Ind. Aerod. 2012, 104, 408–418. [CrossRef] 25. Raffel, M.; Favier, D.; Berton, E.; Rondot, C.; Nsimba, M.; Geissler, W. Micro-PIV and ELDV wind tunnel investigations of the laminar separation bubble above a helicopter blade tip. Meas. Sci. Technol. 2006, 17, 1652–1658. [CrossRef] 26. Göttlich, E.; Woisetschläger, J.; Pieringer, P.; Hampel, B.; Heitmeir, F. Investigation of vortex shedding and wake-wake interaction in a transonic turbine stage using laser-doppler-velocimetry and particle-image-velocimetry. J. Turbomach. 2006, 128, 178–187. [CrossRef] 27. Ragni, D. PIV-Based Load Determination in Aircraft Propellers. Ph.D. Thesis, Delft University of Technology, Delft, The Netherlands, 2012. 28. Ragni, D.; van Oudheusden, B.W.; Scarano, F. PIV-load determination in aircraft propellers. In Proceedings of the 16th International Symposium on Applications of Laser Techniques to Fluid Mechanics, Lisbon, Portugal, 9–12 July 2012. 29. Raffel, M.; Merz, C.; Bauknecht, A.; Schwermer, T. High-speed PIV, deformation and BOS measurements for rotor aerodynamics. In Proceedings of the AFDAR Workshop on Application of Advanced Flow Diagnostics to Aeronautics, Brunswick, Germany, 16–17 October 2013. 30. Bridges, J.; Wernet, M.; Frate, F. PIV measurements of chevrons on F400-series tactical aircraft nozzle model. In Proceedings of the 49th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition, Orlando, FL, USA, 4–7 January 2011; p. 1157. 31. Cao, X.; Liu, J.; Pei, J.; Zhang, Y.; Li, J.; Zhu, X. 2D-PIV measurement of aircraft cabin air distribution with a high spatial resolution. Build. Environ. 2014, 82, 9–19. [CrossRef] 32. De Gregorio, F.; Pengel, K.; Kindler, K. A comprehensive PIV measurement campaign on a fully equipped helicopter model. Exp. Fluids 2012, 53, 37–49. [CrossRef] 33. De Gregorio, F. Flow field characterization and interactional aerodynamics analysis of a complete helicopter. Aerosp. Sci. Technol. 2012, 19, 19–36. [CrossRef] 34. Wadcock, A.; Yamauchi, G.; Solis, E.; Pete, A. PIV measurements in the wake of a full-scale rotor in forward flight. In Proceedings of the 29th AIAA Applied Aerodynamics Conference, Honolulu, HI, USA, 27–30 June 2011; p. 3370. 35. Houston, S.; Thomson, D. On the modelling of gyroplane flight dynamics. Prog. Aerosp. Sci. 2007, 88, 43–58. [CrossRef] 36. Ruchała, P. The measurement & control system in the T-1wind tunnel. Trans. Inst. Aviat. 2013, 5, 65–80. 37. Ruchała, P.; Bogda´nski,K.; Stryczniewicz, W. Analyses of airflow around ducted fan propulsion system in pusher configuration using PIV method. In Proceedings of the 30th Congress of International Council of the Aeronautical Sciences, Daejeon, Korea, 25–30 September 2016. 38. Czy˙z,Z.; Stryczniewicz, W. Investigation of aerodynamic interference in a multirotor by PIV method. ASTRJ 2018, 12, 106–114. [CrossRef] 39. Yoon, S.; Ross, J.; Mench, M.; Sharp, K. Gas-phase particle image velocimetry (PIV) for application to the design of fuel cell reactant flow channels. J. Power Source 2006, 160, 1017–1025. [CrossRef] 40. Hosokawa, M.; Kiyoshi, N.; Naito, M.; Yokoyama, T. Nanoparticle Technology Handbook, 3rd ed.; Elsevier: Amsterdam, The Netherlands, 2007; pp. 109–168. 41. Pescini, E.; Francioso, L.; De Giorgi, M.; Ficarella, A. Investigation of a Micro Dielectric Barrier Discharge Plasma Actuator for Regional Aircraft Active Flow Control. IEEE Trans. Plasma Sci. 2015, 43, 3668–3680. [CrossRef] 42. Westerweel, J.; Scarano, F. Universal outlier detection for PIV data. Exp. Fluids 2005, 39, 1096–1100. [CrossRef] 43. Ruisi, R.; Zare-Behtash, H.; Kontis, K.; Erfani, R. Active flow control over a backward-facing step using plasma actuation. Acta Astronaut. 2016, 126, 354–363. [CrossRef] Sensors 2020, 20, 5537 22 of 22

44. Pescini, E.; De Giorgi, M.G.; Suma, A.; Francioso, L.; Ficarella, A. Separation control by a microfabricated SDBD plasma actuator for small engine turbine applications: Influence of the excitation waveform. Aerosp. Sci. Technol. 2018, 76, 442–454. [CrossRef] 45. Charonko, J.; Vlachos, P. Estimation of uncertainty bounds for individual particle image velocimetry measurements from cross-correlation peak ratio. Meas. Sci. Technol. 2013, 24, 065301. [CrossRef] 46. Sciacchitano, A.; Wieneke, B. PIV uncertainty propagation. Meas. Sci. Technol. 2016, 27, 084006. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).