Final Report Summary - ANCORA (ANotec-COmoti Rotorcraft Acoustics initiative for preliminary acoustic flight tests for the tuning of simplified rotorcraft noise models)
Executive Summary:
One of the objectives of GRC5 is to implement a tool for the minimisation of noise impact on the ground, capable of being executed on-board “on-the-fly”, providing flight directives to the FMS of the helicopter. To this end reliable and fast noise predictions will have to be made, based on actual flight conditions. The envisaged semi-empirical model to be used for this purpose requires information to be derived from experimental data.
The main objective of ANCORA was therefore to determine the transfer function between the noise measured on-board the helicopter, close to the noise sources, and the noise received on the ground by a grid of microphones, during a flight test campaign.
In the first phase of the project a test campaign was performed with the on-board microphones to establish the optimum microphone positions for the final flight tests. However, due to the long-term unavailability of the helicopter in the configuration required for these final tests, it was decided to use an alternative test vehicle. Although the flight tests were performed on a small propeller aircraft, maximum commonality was pursued with the originally envisaged flight plan in terms of measurement systems, flight procedures and data analysis. The main objective of ANCORA was therefore adapted accordingly: validate the full measurement system (both on-board and ground-based) and the software needed for the data analysis and the determination of the transfer functions, developed in ANCORA. This will allow performing the originally envisaged flight tests with the helicopter at a later stage (beyond the ANCORA project) with low risk.
ANCORA demonstrated the feasibility of the application of surface microphones on the helicopter fuselage and on a small turboprop aircraft and will subsequently use this knowledge for a flight test campaign with the same helicopter at a later stage.
ANCORA delivered a robust and reliable mobile noise measurement system, easily scalable and optimised for minimum deployment time and cost. During the test campaign a large number of steady-state conditions were flown over a grid consisting of 30 microphones.
ANCORA developed an advanced method for the determination of the transfer functions between on-board and ground microphones.
All results from the flight tests and data analysis (raw data, 1/3 octave spectra, transfer functions) are available through a data repository.
Based on the above it can in general be concluded that the hardware and software tools developed in ANCORA are valid and that thus the main objective of the ANCORA project has been achieved.
Project Context and Objectives:
One of the objectives of GRC5 is to implement a tool for the minimisation of noise impact on the ground, capable of being executed on-board “on-the-fly”, providing flight directives to the FMS of the helicopter. To this end reliable and fast noise predictions will have to be made, based on actual flight conditions. The envisaged semi-empirical model to be used for this purpose requires information to be derived from experimental data.
The main objective of ANCORA was therefore to determine the transfer function between the noise measured on-board the helicopter, close to the noise sources, and the noise received on the ground by a grid of microphones, during a flight test campaign.
In the first phase of the project a test campaign was performed with the on-board microphones to establish the optimum microphone positions for the final flight tests. However, due to the long-term unavailability of the helicopter in the configuration required for these final tests, it was decided to use an alternative test vehicle. Although the flight tests were performed on a small propeller aircraft, maximum commonality was pursued with the originally envisaged flight plan in terms of measurement systems, flight procedures and data analysis. The main objective of ANCORA was therefore adapted accordingly: validate the full measurement system (both on-board and ground-based) and the software needed for the data analysis and the determination of the transfer functions, developed in ANCORA. This will allow performing the originally envisaged flight tests with the helicopter at a later stage (beyond the ANCORA project) with low risk.
Somewhat more concrete, the principal objectives for ANCORA were:
• To demonstrate the feasibility of the application of surface microphones on the helicopter fuselage and on a small turboprop aircraft to subsequently use this knowledge for a flight test campaign with the same helicopter at a later stage.
• To deliver a robust and reliable mobile noise measurement system, easily scalable and optimised for minimum deployment time and cost.
• To develop an advanced method for the determination of the transfer functions between on-board and ground microphones.
• To design a data repository containing the measured data and the relevant results of the data analysis, for further processing beyond ANCORA.
Project Results:
The three main innovations delivered by ANCORA are the following:
Development of an on-board measurement system consisting of a boom microphone and surface microphones on a helicopter fuselage and on a small turboprop aircraft, to determine the noise close to the source
Development of a robust and reliable mobile noise measurement system, easily scalable and optimised for minimum deployment time and cost.
Development of an advanced method for the determination of the transfer functions between on-board and ground microphones.
In the following more details are provided on each of these innovations and their validation through flight tests.
Development of an on-board microphone system
Until now microphones on the exterior of helicopters have mainly been used to measure the influence of airborne noise (in particular boundary layer noise) on interior cabin noise or as error signal sensors in active rotor noise control systems. In these applications, there are quite relaxed requirements for the microphone type and its location.
In the application in the ANCORA project, however, the transfer function between on-board and ground noise will have to be determined in order to provide accurate information for the validation of prediction models. This requires a correct measurement of the absolute noise levels on-board over a wide frequency range. This implies a much more strict selection of the microphone type to be used and its location on the helicopter exterior, avoiding areas where important influence of rotor flow or forward speed is encountered. In ANCORA the use of surface microphones was envisaged for this purpose, together with a boom microphone with nose cone.
Before the full flight test campaign with both ground and on-board microphones, test flights were performed to select the optimum microphone positions. This section describes the process by which the selected microphones were validated for this purpose, the initial test flights on the helicopter and to conclude the application on the small propeller aircraft, used for the final flight test campaign.
Definition of the on-board microphones chain and initial validation tests
Since the on-board microphone system has two different models of surface microphones, the goal of the first test was to check these acoustic transducers under the same conditions. Thanks to the courtesy of INCAS (National Institute for Aerospace Research "Elie Carafoli") from Romania, these surface microphones have been checked inside their subsonic wind tunnel (see Figure 1).
A second test using both surface microphones consisted in measuring the flow generated noise when they are installed on an automobile in two different positions, one on the front wing and the other on window after mirror in order to measure also in turbulent conditions (see Figure 1).
Figure 1: Validation tests of the surface microphones in a wind-tunnel and on a car
In addition, a test was designed to check the boom microphone installed on a dedicated device on an automobile, to test at various angles of attack with two different nose-cone geometries (ogive nose-cone B&K UA0386 and parabolic nose-cone B&K UA051), compatible with the free-field microphone type 40AE from GRAS. Using the ogive and parabolic nose-cones for the boom microphone at 0 and 10 degree, the aerodynamic noise looks similar in the frequency domain, whereas from 20 degree onwards the parabolic design starts to show significantly more flow induced noise. Based on this analysis, and also taking into account that during the helicopter flight tests, higher speeds will be encountered, it was concluded that the ogive nose-cone is better for the project application.
Pre-tests of the on-board microphone system on a helicopter
The main objective of the on-board pre-tests was to determine the best position for both surface microphones and to validate their use for the purpose of the project. In addition, these tests were used as a risk reduction exercise with respect to the on-board noise data acquisition system, which exhibits some new functionalities never tested before in-flight. During the pre-tests the boom microphone was not mounted and thus it could not been tested.
AgustaWestland performed some initial tests with a borrowed surface microphone and their own noise data acquisition system. The microphone was placed on the fuselage at a position selected mainly by intuition. The main conclusion from this test was that above around 80kts the flow induced broadband noise starts to mask the rotor tones.
In the frame of the ANCORA project specific flight tests were performed on the AW139 P1 to establish the optimal position for the surface microphones. Based on aerodynamic calculations performed by AgustaWestland, 3 new positions were selected in addition to the position used in the first AgustaWestland test. All positions tested were on the left-hand side of the aircraft, mainly due to the existence of a hole in one of the windows which allowed passing the cables in an easy manner.
Two flight tests were performed with the helicopter. In the first flight the surface microphones were mounted at position P1 (same as AgustaWestland initial test) and in the new position P2. In the second flight both microphones were moved to positions P3 and P4 respectively (see Figure 2).
All data was analyzed in narrowband. The main results are provided in Figure 3 and 4 for positions P1-P2 and P3-P4 respectively. Points P3 and P4 appear the best positions for the surface microphones, with rotor tones clearly visible over an acceptable speed range.
Figure 2: Positions of surface microphones
Figure 3: Noise signal at P1 (top) and P2 (bottom) for various flight speeds
Figure 4: Noise signal at P3 (top) and P4 (bottom) for various flight speeds
Tests of the on-board microphone system on a small propeller aircraft
Although the experience gained from the study described in 2.1.2 could be exploited for the final flight tests on the small propeller aircraft, the study on finding good positions for the on-board microphones had obviously to be repeated. In addition during these tests the boom microphone was also deployed, which made an additional investigation on its behavior under flight conditions necessary.
To this end some initial test flights were performed during which the surface microphones were located on the lower wing (pressure side, with more laminar flow) at a section about 2/3 of the wing span. One microphone was placed close to the leading edge of the RH wing (P39, see Figure 5), whereas the second surface microphone was placed at 80 cm from the leading edge (P38). These positions were selected based on recommendations made by the aeroacoustic department of ONERA (France), who kindly supported the request for support on this topic. It is noted that, taking into account the propeller rotation (clockwise when looking in direction of flight), higher acoustic levels are to be expected on the lower LH wing. However, due to obstructions present, it was considered not safe to locate the microphones and their cables on the LH wing.
After having gathered sufficient data the noise signals of these microphones were analysed. From this analysis it was concluded that both surface microphones appeared to have a rather low tone level as compared to the broadband noise (See Figures 6 and 7), probably due to the relatively long lateral distance to the propeller axis. Therefore it was decided to locate both surface microphones more inwards to the fuselage. In order to capture the asymmetry effect caused by the rotation of the propeller, both surface microphones were located symmetrically on either side of the fuselage (P40 and P41).
Figure 5: Details of the microphone positions
As can be seen from Figure 6, the noise signal at these positions is significantly better as compared to the first configuration tested. The main part of the flight test program has therefore been flown with the microphones in these positions.
Figure 6: Comparison of noise signals at positions P38-P40 (left) and p39-P41 (right)
During the initial flight tests with the AW139 helicopter the boom microphone could not be deployed. Therefore the present flight tests were the first opportunity to test the behavior of the boom microphone under actual flight conditions.
Due to safety reasons the only position for fixing the boom microphone, was under the Venturi tube at 1500 mm from the propeller plane and 140 mm from the fuselage (Figure 7).
Figure 7: Boom microphone installation
Based on the experience from windtunnel testing and also considering the results of the tests performed in ANCORA with a car (see 2.1.1) the behavior of a boom microphone appears to depend on the angle of attack of the incoming air. Although for the current test aircraft no information was available on angle of attack, it may be assumed that this angle will be significantly different between the level flight (no flap, high speed) and approach (full flap, low speed) conditions. Figure 12 shows a comparison between 2 test runs performed at the same flight conditions, for level flight and approach respectively. It can be seen that for both flight conditions the signals are matching very well, thus demonstrating the repeatability over a range of inflow conditions.
Figure 8: Noise signal from boom microphone for 2 runs at same flight conditions
High speed level flight (left) and Approach (right)
Based on this result it can be concluded that the measurement of tonal noise sources with a boom microphone under flight conditions is feasible. It can also be concluded that for the final flight tests the position of the boom microphone at P37 was adequate.
Conclusions on the use of the on-board microphone system
Based on the research performed in ANCORA it can be concluded that the use of an on-board microphone system, consisting of surface microphones and a boom microphone, is feasible.
Appropriate locations for the microphones may be selected by means of a semi-empirical approach. A proper position is required to guarantee adequate tonal information up to sufficiently high speeds. Repeatable results may be obtained under a range of operating conditions.
Development of a robust and reliable mobile noise measurement system
The main challenge for the development of the ground-based measurement system was to provide a robust and reliable noise measuring system, comprising of 31 microphones, quickly deployable in the field.
Noise Measuring System (NMS)
After investigating a variety of options to optimise the system and minimise the risk, a solution was found based on the latest generation of sound level meter, developed by Svantek (see Figure 9). The 979 model selected contains in a single unit all major system elements (microphone, real time analyser, digital recorder, gps, modem wi-fi). Integration in a single unit practically eliminates the need for any connections to be made in the field, thus significantly reducing any problems in this respect. Each unit is located in the vicinity (< 10 m) of the microphone position, thus avoiding long cables, another important source of problems and long mounting times.
Figure 9: Svantek noise measurement station
Communication within the ground-based measurement system is performed by means of Zigbee 2.4 GHz radio modules. Each NMS is equipped with a modem which provides bi-directional data communication with the central ground station (CGS). The modems are configured in mesh mode, rather than providing a point-to-point connection. In this manner the distance between NMS and CGS becomes practically irrelevant. In addition this configuration is more robust. In order to avoid saturation of the network when sending real-time noise data to CGS, a total of 4 independent networks has been deployed, each with maximum 8 connected units.
Each measurement station is powered by an internal battery, providing an autonomy of at least 8 hours under normal operating conditions. In practice this means that a unit can be used for several days before recharging is required.
During each measurement the 1/3 octave band spectra are determined each 500 ms and sent to CGS. Simultaneously the pressure time history is stored in a 24-bit wav file at a 48kHz sampling rate.
The internal clock of each sound level meter is synchronised with gps time by means of the integrated gps receiver. In this manner all stations and the aircraft are synchronised against the same time source.
Central Ground Station (CGS)
CGS is the neuralgic center of the flight tests. It consists of the Anotec mobile laboratory van with the central processing unit, where the data from all ground-based systems are received and presented to the CGS operator. Power supply for all CGS systems is based on standard 12 Vdc car batteries, allowing for continuous operation during a full day in any remote environment and for easy replacement in case of failure.
In order to monitor test progress and check proper system functioning, a specific quick-look application was developed. Noise data and status from all 31 NMS units are received continuously and presented on the central display. From here the start and stop commands of the measurements are given. An indication is given if sufficient noise data has been received at each station (by means of a check on the so-called 10dB down time) (see Figure 10).
Figure 10: Central control display in CGS
Directly after each measurement the data from noise stations and flight track system are automatically processed and first results are displayed, so as to be able to get a quick overview of the validity of the measurement.
System tests
Even though various components of the measuring system are adaptations to well-tested applications, the deployment of new hardware always implies a risk. In order to minimise this risk, several tests have been performed with increasing complexity, ahead of the actual test campaign.
The noise measuring system was successfully deployed during some UAV flight tests, during which especially the communication between CGS and stations was tested (see Figure 11).
Figure 11: NMS tests during UAV flight test campaign
Integration tests were performed in two stages. In a first test all aircraft based equipment was built into the Anotec laboratory van, acting as the aircraft. Also the Anotec differential GPS was deployed, in order to provide source position information, needed for software checks. The full ground-based system was deployed alongside a road with the central control built in a rental van. The “aircraft van” was driven along the array of microphones using its claxon, so as to simulate an overflight with some distinctive tones (see Figure 12). In this manner all systems could be checked.
Figure 12: Simulated flight tests with a van
In order to check the quick-look capabilities, a second test was performed at Granada airport (GRX). The full ground-based system was deployed close to the flight track (see Figure 13). The noise of actual aircraft operations at the airport was recorded, together with their flight tracks, emitted by the aircraft through their ADS-B system. This test was also used to train the mounting of the systems so as to minimise the time required for actual field deployment.
Figure 13: System test at Granada airport
During the actual flight test campaign at Trebujena airfield in the south of Spain, the full noise measurement system was deployed and successfully tested under actual flight test conditions. An array of microphones spanning more than 1 km was used (Figure 14).
Figure 14: Flight tests at Trebujena airfield
Conclusions on the mobile noise measurement system
Based on the research performed in ANCORA it can be concluded that the mobile noise measuring system developed for the flight tests is a robust and reliable solution that can easily be deployed in the field.
Its architecture makes a very flexible microphone layout in the field possible, thus facilitating the use of complex flight test procedures.
Development of an advanced method to determine transfer functions
The main objective of ANCORA is to validate algorithms to determine the transfer function between the noise measured on-board the helicopter and the noise measured on the ground. This transfer function can then be used in the tuning of a semi-empirical rotorcraft noise prediction model, being developed for the implementation in an on-board system of an AgustaWestland helicopter. This on-board system will be able to give indications of the flight path that the pilot needs to follow for the minimisation of the noise impact “on-the-fly”.
The method used to determine the mentioned transfer function (TF) using on-board and ground noise measurements is described hereafter. For the purpose of ANCORA this method has been implemented in a Matlab script.
Original signals
As a first step the sensor time series as stored in the wav files, generated during the measurements by both the on-board and ground systems, are converted to pressure time signals [Pa], applying the sensitivity determined during the pistonphone calibration.
As an example, the spectrograms of some measured noise signals are plotted for on-board and ground microphones in figures 15 and 16 respectively.
Figure 15 Spectrogram of signal measured on-board
Figure 16 Spectrogram of signal measured on ground
Processed ground signals
The measured time series are then divided into blocks. Each block is labelled with the relevant information on the aircraft position at the time instant corresponding to the beginning of the block. This information is obtained by interpolating in the measured flight path and recalculating distances and directivity angles () at interpolated flight path points and the travel time.
The above process allows sampling of the aircraft trajectory synchronized with the resolution applied in the noise signal processing. With this labeling it is possible to align the blocks of the ground and on-board signals, shifted due to the travel time of the sound from source to observer. In addition, with this labelling the spectral shift due to the Doppler effect can be compensated.
Once the blocks are aligned, a correction for the corresponding Doppler shift is applied on the blocks for the ground signal. This correction will be different for each location, depending on the relative velocity of the helicopter, and is realised by means of an interpolated delay line, using fractional delay filters.
An example of the processed signal for the same ground microphone is given in Figure 17.
Figure 17 Spectrogram of processed ground signal (Flight 06 Run 13 NMS 26)
Transfer Function
For each of the blocks the cross-correlation of the time signals on the ground and on-board and the corresponding spectrum is determined. See Figure 18 for an example of the frequency response of the transfer function H1 for the microphone pair mentioned above.
Figure 18 Frequency response of the transfer function H1 for a single microphone pair
The above process can be repeated for the other two on-board microphones correlated with the same ground microphone, thus obtaining a Cross-Spectral-Density-Matrix for each time block. With this, it will thus be possible to estimate the frequency response of the system (H1) based on the cross-spectral densities helicopter-ground (Ghg) and the auto-spectrum of the signal on the ground (Ggg):
H_1≅(∑G_hg )/G_gg
Figure 19 presents the resulting frequency response of the averaged transfer function H1.
Figure 19 Frequency response of the transfer function H1 for 3 microphone pairs averaged
Conclusions on Transfer Function
Based on the research performed in ANCORA it can be concluded that it is feasible to derive Transfer Functions between the noise measured on-board the helicopter and the noise measured on the ground.
Actual flight test data has been used to validate the algorithms and to produce an extensive database.
Potential Impact:
The main objective of ANCORA was to validate algorithms to determine the transfer function between the noise measured on-board the helicopter and the noise measured on the ground. This transfer function can then be used in the tuning of a semi-empirical rotorcraft noise prediction model, being developed for the implementation in an on-board system of an AgustaWestland helicopter. This on-board system will be able to give indications of the flight path that the pilot needs to follow for the minimisation of the noise impact “on-the-fly”.
Successful completion of ANCORA is thus decisive for enabling future application of the above described development.
An important condition for the accomplishment of the abovementioned objective was the availability of validated noise measurement systems for both on-board and ground applications.
The development and validation of an on-board microphone system for the measurement of noise signals close to the source was therefore an important enabler for the main objective of ANCORA.
Another result of ANCORA is the availability to the EU aircraft industry of a robust and reliable mobile noise measurement system, easily scalable and optimised for minimum deployment time, at a cost substantially lower than that of systems currently used. Having such a system readily available for flight tests anywhere, greatly enhances the possibility of its use in research projects with relatively limited budget, thus allowing for a significantly increased knowledge of rotorcraft (and also fixed wing aircraft) noise, one of the main objectives of the Clean Sky JTI.
Data from the ground-based noise measurements can be processed with software developed in ANCORA, so as to provide a HELENA compatible database. This will improve the capabilities of the prediction tool, available to the HELENA consortium (i.e. all mayor European industries and research institutes) and, in the public version, outside the consortium. Enhancing the HELENA database with new helicopter types is an important step towards the acceptance of HELENA as the European helicopter noise model for strategic noise mapping in the frame of the European environmental noise directive, as envisaged in the CNOSSOS-EU project.
List of Websites:
For information please contact:
Nico van Oosten
Anotec Engineering SL
c/ Rector Jose Vida Soria, 2
Urb. Terrazas de Playa Granada, portal 7-2ºC
18613 Motril (Granada)
Spain
tel/fax (+34) 958 620 631
email nico@anotecengineering.com
One of the objectives of GRC5 is to implement a tool for the minimisation of noise impact on the ground, capable of being executed on-board “on-the-fly”, providing flight directives to the FMS of the helicopter. To this end reliable and fast noise predictions will have to be made, based on actual flight conditions. The envisaged semi-empirical model to be used for this purpose requires information to be derived from experimental data.
The main objective of ANCORA was therefore to determine the transfer function between the noise measured on-board the helicopter, close to the noise sources, and the noise received on the ground by a grid of microphones, during a flight test campaign.
In the first phase of the project a test campaign was performed with the on-board microphones to establish the optimum microphone positions for the final flight tests. However, due to the long-term unavailability of the helicopter in the configuration required for these final tests, it was decided to use an alternative test vehicle. Although the flight tests were performed on a small propeller aircraft, maximum commonality was pursued with the originally envisaged flight plan in terms of measurement systems, flight procedures and data analysis. The main objective of ANCORA was therefore adapted accordingly: validate the full measurement system (both on-board and ground-based) and the software needed for the data analysis and the determination of the transfer functions, developed in ANCORA. This will allow performing the originally envisaged flight tests with the helicopter at a later stage (beyond the ANCORA project) with low risk.
ANCORA demonstrated the feasibility of the application of surface microphones on the helicopter fuselage and on a small turboprop aircraft and will subsequently use this knowledge for a flight test campaign with the same helicopter at a later stage.
ANCORA delivered a robust and reliable mobile noise measurement system, easily scalable and optimised for minimum deployment time and cost. During the test campaign a large number of steady-state conditions were flown over a grid consisting of 30 microphones.
ANCORA developed an advanced method for the determination of the transfer functions between on-board and ground microphones.
All results from the flight tests and data analysis (raw data, 1/3 octave spectra, transfer functions) are available through a data repository.
Based on the above it can in general be concluded that the hardware and software tools developed in ANCORA are valid and that thus the main objective of the ANCORA project has been achieved.
Project Context and Objectives:
One of the objectives of GRC5 is to implement a tool for the minimisation of noise impact on the ground, capable of being executed on-board “on-the-fly”, providing flight directives to the FMS of the helicopter. To this end reliable and fast noise predictions will have to be made, based on actual flight conditions. The envisaged semi-empirical model to be used for this purpose requires information to be derived from experimental data.
The main objective of ANCORA was therefore to determine the transfer function between the noise measured on-board the helicopter, close to the noise sources, and the noise received on the ground by a grid of microphones, during a flight test campaign.
In the first phase of the project a test campaign was performed with the on-board microphones to establish the optimum microphone positions for the final flight tests. However, due to the long-term unavailability of the helicopter in the configuration required for these final tests, it was decided to use an alternative test vehicle. Although the flight tests were performed on a small propeller aircraft, maximum commonality was pursued with the originally envisaged flight plan in terms of measurement systems, flight procedures and data analysis. The main objective of ANCORA was therefore adapted accordingly: validate the full measurement system (both on-board and ground-based) and the software needed for the data analysis and the determination of the transfer functions, developed in ANCORA. This will allow performing the originally envisaged flight tests with the helicopter at a later stage (beyond the ANCORA project) with low risk.
Somewhat more concrete, the principal objectives for ANCORA were:
• To demonstrate the feasibility of the application of surface microphones on the helicopter fuselage and on a small turboprop aircraft to subsequently use this knowledge for a flight test campaign with the same helicopter at a later stage.
• To deliver a robust and reliable mobile noise measurement system, easily scalable and optimised for minimum deployment time and cost.
• To develop an advanced method for the determination of the transfer functions between on-board and ground microphones.
• To design a data repository containing the measured data and the relevant results of the data analysis, for further processing beyond ANCORA.
Project Results:
The three main innovations delivered by ANCORA are the following:
Development of an on-board measurement system consisting of a boom microphone and surface microphones on a helicopter fuselage and on a small turboprop aircraft, to determine the noise close to the source
Development of a robust and reliable mobile noise measurement system, easily scalable and optimised for minimum deployment time and cost.
Development of an advanced method for the determination of the transfer functions between on-board and ground microphones.
In the following more details are provided on each of these innovations and their validation through flight tests.
Development of an on-board microphone system
Until now microphones on the exterior of helicopters have mainly been used to measure the influence of airborne noise (in particular boundary layer noise) on interior cabin noise or as error signal sensors in active rotor noise control systems. In these applications, there are quite relaxed requirements for the microphone type and its location.
In the application in the ANCORA project, however, the transfer function between on-board and ground noise will have to be determined in order to provide accurate information for the validation of prediction models. This requires a correct measurement of the absolute noise levels on-board over a wide frequency range. This implies a much more strict selection of the microphone type to be used and its location on the helicopter exterior, avoiding areas where important influence of rotor flow or forward speed is encountered. In ANCORA the use of surface microphones was envisaged for this purpose, together with a boom microphone with nose cone.
Before the full flight test campaign with both ground and on-board microphones, test flights were performed to select the optimum microphone positions. This section describes the process by which the selected microphones were validated for this purpose, the initial test flights on the helicopter and to conclude the application on the small propeller aircraft, used for the final flight test campaign.
Definition of the on-board microphones chain and initial validation tests
Since the on-board microphone system has two different models of surface microphones, the goal of the first test was to check these acoustic transducers under the same conditions. Thanks to the courtesy of INCAS (National Institute for Aerospace Research "Elie Carafoli") from Romania, these surface microphones have been checked inside their subsonic wind tunnel (see Figure 1).
A second test using both surface microphones consisted in measuring the flow generated noise when they are installed on an automobile in two different positions, one on the front wing and the other on window after mirror in order to measure also in turbulent conditions (see Figure 1).
Figure 1: Validation tests of the surface microphones in a wind-tunnel and on a car
In addition, a test was designed to check the boom microphone installed on a dedicated device on an automobile, to test at various angles of attack with two different nose-cone geometries (ogive nose-cone B&K UA0386 and parabolic nose-cone B&K UA051), compatible with the free-field microphone type 40AE from GRAS. Using the ogive and parabolic nose-cones for the boom microphone at 0 and 10 degree, the aerodynamic noise looks similar in the frequency domain, whereas from 20 degree onwards the parabolic design starts to show significantly more flow induced noise. Based on this analysis, and also taking into account that during the helicopter flight tests, higher speeds will be encountered, it was concluded that the ogive nose-cone is better for the project application.
Pre-tests of the on-board microphone system on a helicopter
The main objective of the on-board pre-tests was to determine the best position for both surface microphones and to validate their use for the purpose of the project. In addition, these tests were used as a risk reduction exercise with respect to the on-board noise data acquisition system, which exhibits some new functionalities never tested before in-flight. During the pre-tests the boom microphone was not mounted and thus it could not been tested.
AgustaWestland performed some initial tests with a borrowed surface microphone and their own noise data acquisition system. The microphone was placed on the fuselage at a position selected mainly by intuition. The main conclusion from this test was that above around 80kts the flow induced broadband noise starts to mask the rotor tones.
In the frame of the ANCORA project specific flight tests were performed on the AW139 P1 to establish the optimal position for the surface microphones. Based on aerodynamic calculations performed by AgustaWestland, 3 new positions were selected in addition to the position used in the first AgustaWestland test. All positions tested were on the left-hand side of the aircraft, mainly due to the existence of a hole in one of the windows which allowed passing the cables in an easy manner.
Two flight tests were performed with the helicopter. In the first flight the surface microphones were mounted at position P1 (same as AgustaWestland initial test) and in the new position P2. In the second flight both microphones were moved to positions P3 and P4 respectively (see Figure 2).
All data was analyzed in narrowband. The main results are provided in Figure 3 and 4 for positions P1-P2 and P3-P4 respectively. Points P3 and P4 appear the best positions for the surface microphones, with rotor tones clearly visible over an acceptable speed range.
Figure 2: Positions of surface microphones
Figure 3: Noise signal at P1 (top) and P2 (bottom) for various flight speeds
Figure 4: Noise signal at P3 (top) and P4 (bottom) for various flight speeds
Tests of the on-board microphone system on a small propeller aircraft
Although the experience gained from the study described in 2.1.2 could be exploited for the final flight tests on the small propeller aircraft, the study on finding good positions for the on-board microphones had obviously to be repeated. In addition during these tests the boom microphone was also deployed, which made an additional investigation on its behavior under flight conditions necessary.
To this end some initial test flights were performed during which the surface microphones were located on the lower wing (pressure side, with more laminar flow) at a section about 2/3 of the wing span. One microphone was placed close to the leading edge of the RH wing (P39, see Figure 5), whereas the second surface microphone was placed at 80 cm from the leading edge (P38). These positions were selected based on recommendations made by the aeroacoustic department of ONERA (France), who kindly supported the request for support on this topic. It is noted that, taking into account the propeller rotation (clockwise when looking in direction of flight), higher acoustic levels are to be expected on the lower LH wing. However, due to obstructions present, it was considered not safe to locate the microphones and their cables on the LH wing.
After having gathered sufficient data the noise signals of these microphones were analysed. From this analysis it was concluded that both surface microphones appeared to have a rather low tone level as compared to the broadband noise (See Figures 6 and 7), probably due to the relatively long lateral distance to the propeller axis. Therefore it was decided to locate both surface microphones more inwards to the fuselage. In order to capture the asymmetry effect caused by the rotation of the propeller, both surface microphones were located symmetrically on either side of the fuselage (P40 and P41).
Figure 5: Details of the microphone positions
As can be seen from Figure 6, the noise signal at these positions is significantly better as compared to the first configuration tested. The main part of the flight test program has therefore been flown with the microphones in these positions.
Figure 6: Comparison of noise signals at positions P38-P40 (left) and p39-P41 (right)
During the initial flight tests with the AW139 helicopter the boom microphone could not be deployed. Therefore the present flight tests were the first opportunity to test the behavior of the boom microphone under actual flight conditions.
Due to safety reasons the only position for fixing the boom microphone, was under the Venturi tube at 1500 mm from the propeller plane and 140 mm from the fuselage (Figure 7).
Figure 7: Boom microphone installation
Based on the experience from windtunnel testing and also considering the results of the tests performed in ANCORA with a car (see 2.1.1) the behavior of a boom microphone appears to depend on the angle of attack of the incoming air. Although for the current test aircraft no information was available on angle of attack, it may be assumed that this angle will be significantly different between the level flight (no flap, high speed) and approach (full flap, low speed) conditions. Figure 12 shows a comparison between 2 test runs performed at the same flight conditions, for level flight and approach respectively. It can be seen that for both flight conditions the signals are matching very well, thus demonstrating the repeatability over a range of inflow conditions.
Figure 8: Noise signal from boom microphone for 2 runs at same flight conditions
High speed level flight (left) and Approach (right)
Based on this result it can be concluded that the measurement of tonal noise sources with a boom microphone under flight conditions is feasible. It can also be concluded that for the final flight tests the position of the boom microphone at P37 was adequate.
Conclusions on the use of the on-board microphone system
Based on the research performed in ANCORA it can be concluded that the use of an on-board microphone system, consisting of surface microphones and a boom microphone, is feasible.
Appropriate locations for the microphones may be selected by means of a semi-empirical approach. A proper position is required to guarantee adequate tonal information up to sufficiently high speeds. Repeatable results may be obtained under a range of operating conditions.
Development of a robust and reliable mobile noise measurement system
The main challenge for the development of the ground-based measurement system was to provide a robust and reliable noise measuring system, comprising of 31 microphones, quickly deployable in the field.
Noise Measuring System (NMS)
After investigating a variety of options to optimise the system and minimise the risk, a solution was found based on the latest generation of sound level meter, developed by Svantek (see Figure 9). The 979 model selected contains in a single unit all major system elements (microphone, real time analyser, digital recorder, gps, modem wi-fi). Integration in a single unit practically eliminates the need for any connections to be made in the field, thus significantly reducing any problems in this respect. Each unit is located in the vicinity (< 10 m) of the microphone position, thus avoiding long cables, another important source of problems and long mounting times.
Figure 9: Svantek noise measurement station
Communication within the ground-based measurement system is performed by means of Zigbee 2.4 GHz radio modules. Each NMS is equipped with a modem which provides bi-directional data communication with the central ground station (CGS). The modems are configured in mesh mode, rather than providing a point-to-point connection. In this manner the distance between NMS and CGS becomes practically irrelevant. In addition this configuration is more robust. In order to avoid saturation of the network when sending real-time noise data to CGS, a total of 4 independent networks has been deployed, each with maximum 8 connected units.
Each measurement station is powered by an internal battery, providing an autonomy of at least 8 hours under normal operating conditions. In practice this means that a unit can be used for several days before recharging is required.
During each measurement the 1/3 octave band spectra are determined each 500 ms and sent to CGS. Simultaneously the pressure time history is stored in a 24-bit wav file at a 48kHz sampling rate.
The internal clock of each sound level meter is synchronised with gps time by means of the integrated gps receiver. In this manner all stations and the aircraft are synchronised against the same time source.
Central Ground Station (CGS)
CGS is the neuralgic center of the flight tests. It consists of the Anotec mobile laboratory van with the central processing unit, where the data from all ground-based systems are received and presented to the CGS operator. Power supply for all CGS systems is based on standard 12 Vdc car batteries, allowing for continuous operation during a full day in any remote environment and for easy replacement in case of failure.
In order to monitor test progress and check proper system functioning, a specific quick-look application was developed. Noise data and status from all 31 NMS units are received continuously and presented on the central display. From here the start and stop commands of the measurements are given. An indication is given if sufficient noise data has been received at each station (by means of a check on the so-called 10dB down time) (see Figure 10).
Figure 10: Central control display in CGS
Directly after each measurement the data from noise stations and flight track system are automatically processed and first results are displayed, so as to be able to get a quick overview of the validity of the measurement.
System tests
Even though various components of the measuring system are adaptations to well-tested applications, the deployment of new hardware always implies a risk. In order to minimise this risk, several tests have been performed with increasing complexity, ahead of the actual test campaign.
The noise measuring system was successfully deployed during some UAV flight tests, during which especially the communication between CGS and stations was tested (see Figure 11).
Figure 11: NMS tests during UAV flight test campaign
Integration tests were performed in two stages. In a first test all aircraft based equipment was built into the Anotec laboratory van, acting as the aircraft. Also the Anotec differential GPS was deployed, in order to provide source position information, needed for software checks. The full ground-based system was deployed alongside a road with the central control built in a rental van. The “aircraft van” was driven along the array of microphones using its claxon, so as to simulate an overflight with some distinctive tones (see Figure 12). In this manner all systems could be checked.
Figure 12: Simulated flight tests with a van
In order to check the quick-look capabilities, a second test was performed at Granada airport (GRX). The full ground-based system was deployed close to the flight track (see Figure 13). The noise of actual aircraft operations at the airport was recorded, together with their flight tracks, emitted by the aircraft through their ADS-B system. This test was also used to train the mounting of the systems so as to minimise the time required for actual field deployment.
Figure 13: System test at Granada airport
During the actual flight test campaign at Trebujena airfield in the south of Spain, the full noise measurement system was deployed and successfully tested under actual flight test conditions. An array of microphones spanning more than 1 km was used (Figure 14).
Figure 14: Flight tests at Trebujena airfield
Conclusions on the mobile noise measurement system
Based on the research performed in ANCORA it can be concluded that the mobile noise measuring system developed for the flight tests is a robust and reliable solution that can easily be deployed in the field.
Its architecture makes a very flexible microphone layout in the field possible, thus facilitating the use of complex flight test procedures.
Development of an advanced method to determine transfer functions
The main objective of ANCORA is to validate algorithms to determine the transfer function between the noise measured on-board the helicopter and the noise measured on the ground. This transfer function can then be used in the tuning of a semi-empirical rotorcraft noise prediction model, being developed for the implementation in an on-board system of an AgustaWestland helicopter. This on-board system will be able to give indications of the flight path that the pilot needs to follow for the minimisation of the noise impact “on-the-fly”.
The method used to determine the mentioned transfer function (TF) using on-board and ground noise measurements is described hereafter. For the purpose of ANCORA this method has been implemented in a Matlab script.
Original signals
As a first step the sensor time series as stored in the wav files, generated during the measurements by both the on-board and ground systems, are converted to pressure time signals [Pa], applying the sensitivity determined during the pistonphone calibration.
As an example, the spectrograms of some measured noise signals are plotted for on-board and ground microphones in figures 15 and 16 respectively.
Figure 15 Spectrogram of signal measured on-board
Figure 16 Spectrogram of signal measured on ground
Processed ground signals
The measured time series are then divided into blocks. Each block is labelled with the relevant information on the aircraft position at the time instant corresponding to the beginning of the block. This information is obtained by interpolating in the measured flight path and recalculating distances and directivity angles () at interpolated flight path points and the travel time.
The above process allows sampling of the aircraft trajectory synchronized with the resolution applied in the noise signal processing. With this labeling it is possible to align the blocks of the ground and on-board signals, shifted due to the travel time of the sound from source to observer. In addition, with this labelling the spectral shift due to the Doppler effect can be compensated.
Once the blocks are aligned, a correction for the corresponding Doppler shift is applied on the blocks for the ground signal. This correction will be different for each location, depending on the relative velocity of the helicopter, and is realised by means of an interpolated delay line, using fractional delay filters.
An example of the processed signal for the same ground microphone is given in Figure 17.
Figure 17 Spectrogram of processed ground signal (Flight 06 Run 13 NMS 26)
Transfer Function
For each of the blocks the cross-correlation of the time signals on the ground and on-board and the corresponding spectrum is determined. See Figure 18 for an example of the frequency response of the transfer function H1 for the microphone pair mentioned above.
Figure 18 Frequency response of the transfer function H1 for a single microphone pair
The above process can be repeated for the other two on-board microphones correlated with the same ground microphone, thus obtaining a Cross-Spectral-Density-Matrix for each time block. With this, it will thus be possible to estimate the frequency response of the system (H1) based on the cross-spectral densities helicopter-ground (Ghg) and the auto-spectrum of the signal on the ground (Ggg):
H_1≅(∑G_hg )/G_gg
Figure 19 presents the resulting frequency response of the averaged transfer function H1.
Figure 19 Frequency response of the transfer function H1 for 3 microphone pairs averaged
Conclusions on Transfer Function
Based on the research performed in ANCORA it can be concluded that it is feasible to derive Transfer Functions between the noise measured on-board the helicopter and the noise measured on the ground.
Actual flight test data has been used to validate the algorithms and to produce an extensive database.
Potential Impact:
The main objective of ANCORA was to validate algorithms to determine the transfer function between the noise measured on-board the helicopter and the noise measured on the ground. This transfer function can then be used in the tuning of a semi-empirical rotorcraft noise prediction model, being developed for the implementation in an on-board system of an AgustaWestland helicopter. This on-board system will be able to give indications of the flight path that the pilot needs to follow for the minimisation of the noise impact “on-the-fly”.
Successful completion of ANCORA is thus decisive for enabling future application of the above described development.
An important condition for the accomplishment of the abovementioned objective was the availability of validated noise measurement systems for both on-board and ground applications.
The development and validation of an on-board microphone system for the measurement of noise signals close to the source was therefore an important enabler for the main objective of ANCORA.
Another result of ANCORA is the availability to the EU aircraft industry of a robust and reliable mobile noise measurement system, easily scalable and optimised for minimum deployment time, at a cost substantially lower than that of systems currently used. Having such a system readily available for flight tests anywhere, greatly enhances the possibility of its use in research projects with relatively limited budget, thus allowing for a significantly increased knowledge of rotorcraft (and also fixed wing aircraft) noise, one of the main objectives of the Clean Sky JTI.
Data from the ground-based noise measurements can be processed with software developed in ANCORA, so as to provide a HELENA compatible database. This will improve the capabilities of the prediction tool, available to the HELENA consortium (i.e. all mayor European industries and research institutes) and, in the public version, outside the consortium. Enhancing the HELENA database with new helicopter types is an important step towards the acceptance of HELENA as the European helicopter noise model for strategic noise mapping in the frame of the European environmental noise directive, as envisaged in the CNOSSOS-EU project.
List of Websites:
For information please contact:
Nico van Oosten
Anotec Engineering SL
c/ Rector Jose Vida Soria, 2
Urb. Terrazas de Playa Granada, portal 7-2ºC
18613 Motril (Granada)
Spain
tel/fax (+34) 958 620 631
email nico@anotecengineering.com
