DEVICE FOR DETECTING THE APPROACH OF A VORTEX RING STATE, ROTARY-WING AERODYNE COMPRISING SAID DEVICE, AND ASSOCIATED METHOD
20220355919 · 2022-11-10
Inventors
- Laurent BINET (Grans, FR)
- Stéphane PEZET (Saint Mitre Les Remparts, FR)
- Marc PRUNEL (Aix-en-Provence, FR)
Cpc classification
B64C19/00
PERFORMING OPERATIONS; TRANSPORTING
International classification
Abstract
A device for detecting the approach of a vortex ring state for a rotary wing aerodyne, the detection device including a set of vibration sensors configured to be distributed in or on the aerodyne, and a data processing unit configured to receive in real time measurement data from the sensors, process the data in order to calculate in real time the vibration spectrum of the aerodyne, detect in real time, by vibration analysis, the approach of a vortex ring state as a function of the calculated vibration spectrum, and issue an alarm in the event of detection of the approach of a vortex ring state.
Claims
1. A device for detecting the approach of a vortex ring state for a rotary wing aerodyne, wherein the detection device comprises a set of vibration sensors configured to be distributed in the rotary wing aerodyne or on the rotary wing aerodyne and configured to detect vibrations at different points of the rotary wing aerodyne at least along the longitudinal and vertical axes of the rotary wing aerodyne, the detection device further comprising a data processing unit connected to the set of vibration sensors, the data processing unit being configured to: receive measurement data in real time from the set of vibration sensors, process the measurement data received so as to calculate in real time the vibration spectrum of the rotary wing aerodyne along at least one of the longitudinal and vertical axes of the rotary wing aerodyne on at least two distinct vibration sensors from the set of vibration sensors, detect in real time, by vibration analysis, the approach of a vortex ring state as a function of the calculated vibration spectrum, and issue an alarm in the event of detection of the approach of a vortex ring state.
2. The device for detecting the approach of a vortex ring state according to claim 1, wherein each vibration sensor is a three-axis accelerometer, namely a longitudinal axis X corresponding to the longitudinal axis of the rotary wing aerodyne, a transverse axis Y corresponding to the transverse axis of the rotary wing aerodyne and a vertical axis Z.
3. The device for detecting the approach of a vortex ring state of claim 2, wherein the vibration sensor set comprises a first three-axis accelerometer configured to be installed in the cabin of the rotary wing aerodyne, a second three-axis accelerometer configured to be installed on the main gearbox of the rotary wing aerodyne, and a third three-axis accelerometer configured to be installed on the engine deck of the rotary wing aerodyne proximate to the center of gravity of the rotary wing aerodyne.
4. The device for detecting the approach of a vortex ring state according to claim 3, wherein the data processing unit is configured to perform at least one of: a first vibration analysis on the measurement data streams corresponding to the longitudinal axis X and the vertical axis Z of the second three-axis accelerometer comprising, for each of the measurement data streams corresponding to the longitudinal axis X and the vertical axis Z of the second three-axis accelerometer, comparing the fundamental frequency of the main rotor of the rotary wing aircraft with the predominant frequency, namely corresponding to the maximum amplitude level, of the vibration spectrum derived from a discrete Fourier transform on the corresponding measurement data stream, and, if the difference between the predominant frequency and the fundamental frequency of the main rotor is greater than a predetermined frequency difference threshold, generating a detection signal associated with the corresponding measurement data stream; a second vibration analysis on the measurement data streams corresponding to the vertical axis Z of the first three-axis accelerometer and corresponding to the longitudinal axis X and vertical axis Z of the third three-axis accelerometer comprising, for each of the measurement data streams corresponding to the vertical axis Z of the first three-axis accelerometer and corresponding to the longitudinal axis X and vertical axis Z of the third three-axis accelerometer, digital filtering without phase shifting of the corresponding measurement data stream using a low-pass finite impulse response filter and a high-pass finite impulse response filter, calculating the RMS value of the signal from the low-pass filter and the RMS value of the signal from the high-pass filter, calculating an energy ratio corresponding to the RMS value of the signal from the high-pass filter divided by the RMS value of the signal from the low-pass filter, and, if the energy ratio is greater than a corresponding predetermined energy ratio threshold, generating a detection signal associated with the corresponding measurement data stream, and a third vibration analysis comprising: applying a Hilbert transform to the measurement data stream corresponding to the vertical axis Z of the first three-axis accelerometer to extend the real signal into the complex domain; applying a Fast Fourier Transform (FFT) to the Hilbert transformed signal; and filtering the obtained FFT signal by plus or minus 5 Hz around the fundamental main rotor frequency corresponding to the current main rotor speed, the fundamental main rotor frequency defined as (NR*b)/60, where NR is the main rotor speed, and b is the number of main rotor blades; reconstructing the filtered signal by applying an inverse Fourier transform; producing a spectrogram on the reconstructed signal and removing the negative frequency part of the spectrum; calculating the standard deviation of the amplitude of the spectrogram over a sliding window of a predefined duration; and, if the calculated standard deviation is above a predefined threshold, generating a detection signal.
5. The device for detecting the approach of a vortex ring state according to claim 4, wherein the data processing unit is configured to perform the first vibration analysis and the second vibration analysis simultaneously.
6. The device for detecting the approach of a vortex ring state according to claim 5, wherein the data processing unit is configured to detect the approach of a vortex ring state and to issue an alarm when the first vibration analysis and the second vibration analysis simultaneously generate at least one detection signal.
7. The device for detecting the approach of a vortex ring state according to claim 6, wherein the data processing unit is configured to detect the approach of a vortex ring state and generate an alarm when at least three detection signals from the first and second vibration analyses are generated simultaneously.
8. The device for detecting the approach of a vortex ring state according to claim 1, wherein the data processing unit is further configured to use at least one of the following measurements to detect the approach of a vortex ring state: measuring the indicated speed of the rotary wing aerodyne, measuring the vertical speed of the rotary wing aerodyne, and measuring the rotation speed of the main rotor of the rotary wing aerodyne.
9. The vortex ring state approach detection device according to claim 1, wherein the detection device further comprises a warning unit connected to the data processing unit and configured to generate a vortex ring state approach warning when the data processing unit issues an alarm, the vortex ring state approach warning being at least one of audible, visual and haptic.
10. A rotary wing aerodyne comprising a device for detecting the approach of a vortex ring state according to claim 1, the set of vibration sensors of the detection device being distributed in the rotary wing aerodyne or on the rotary wing aerodyne.
11. The rotary wing aerodyne according to claim 10, further comprising a flight control system connected to the data processing unit of the detection device and configured to modify, automatically or semi-automatically, the piloting of the rotary wing aerodyne when the data processing unit of the detection device issues an alarm, so as to prevent the rotary wing aerodyne from entering the vortex ring state.
12. A method of detecting the approach of a vortex ring state by the rotary wing aerodyne according to claim 10, the method comprising, by the data processing unit of the vortex ring state detection device: receiving in real time measurement data from the set of vibration sensors; processing the received measurement data so as to calculate in real time the vibration spectrum of the rotary wing aerodyne on at least two distinct vibration sensors among the set of vibration sensors; detecting in real time, by vibration analysis, the approach of a vortex ring state according to the calculated vibration spectrum; and issuing an alarm in case of detection of the approach of a vortex ring state.
13. The device for detecting the approach of a vortex ring state according to claim 2, wherein the data processing unit is further configured to use at least one of the following measurements to detect the approach of a vortex ring state: measuring the indicated speed of the rotary wing aerodyne, measuring the vertical speed of the rotary wing aerodyne, and measuring the rotation speed of the main rotor of the rotary wing aerodyne.
14. The device for detecting the approach of a vortex ring state according to claim 3, wherein the data processing unit is further configured to use at least one of the following measurements to detect the approach of a vortex ring state: measuring the indicated speed of the rotary wing aerodyne, measuring the vertical speed of the rotary wing aerodyne, and measuring the rotation speed of the main rotor of the rotary wing aerodyne.
15. The device for detecting the approach of a vortex ring state according to claim 4, wherein the data processing unit is further configured to use at least one of the following measurements to detect the approach of a vortex ring state: measuring the indicated speed of the rotary wing aerodyne, measuring the vertical speed of the rotary wing aerodyne, and measuring the rotation speed of the main rotor of the rotary wing aerodyne.
16. The device for detecting the approach of a vortex ring state according to claim 5, wherein the data processing unit is further configured to use at least one of the following measurements to detect the approach of a vortex ring state: measuring the indicated speed of the rotary wing aerodyne, measuring the vertical speed of the rotary wing aerodyne, and measuring the rotation speed of the main rotor of the rotary wing aerodyne.
17. The device for detecting the approach of a vortex ring state according to claim 6, wherein the data processing unit is further configured to use at least one of the following measurements to detect the approach of a vortex ring state: measuring the indicated speed of the rotary wing aerodyne, measuring the vertical speed of the rotary wing aerodyne, and measuring the rotation speed of the main rotor of the rotary wing aerodyne.
18. The device for detecting the approach of a vortex ring state according to claim 7, wherein the data processing unit is further configured to use at least one of the following measurements to detect the approach of a vortex ring state: measuring the indicated speed of the rotary wing aerodyne, measuring the vertical speed of the rotary wing aerodyne, and measuring the rotation speed of the main rotor of the rotary wing aerodyne.
19. The vortex ring state approach detection device according to claim 2, wherein the detection device further comprises a warning unit connected to the data processing unit and configured to generate a vortex ring state approach warning when the data processing unit issues an alarm, the vortex ring state approach warning being at least one of audible, visual and haptic.
20. A method of detecting the approach of a vortex ring state by the rotary wing aerodyne according to claim 11, the method comprising, by the data processing unit of the vortex ring state detection device: receiving in real time measurement data from the set of vibration sensors; processing the received measurement data so as to calculate in real time the vibration spectrum of the rotary wing aerodyne on at least two distinct vibration sensors among the set of vibration sensors; detecting in real time, by vibration analysis, the approach of a vortex ring state according to the calculated vibration spectrum; and issuing an alarm in case of detection of the approach of a vortex ring state.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
[0066] To better illustrate the object of the present invention, a preferred embodiment is described below, by way of illustration and not limitation, with reference to the attached drawings.
[0067] On these drawings:
[0068]
[0069]
[0070]
[0071]
[0072]
DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0073] Referring to
[0074] The rotary wing aerodyne 1 is equipped with a main rotor 2 which provides both lift and propulsion for the aerodyne 1, said main rotor 2 carrying a number of blades 3.
[0075] In
[0076] In the level ground/ascent condition, shown in the left-hand side view of
[0077] On the contrary, in the high descent rate condition, shown in the right-hand side view of
[0078] The vortex ring state (or vortex phenomenon) is a particular regime of main rotor 2 operation that can occur in descent flight at low forward speeds, typically less than 25 knots, and at moderate rates of descent, typically of the order of −1000 feet per minute. In this situation, shown in the central view of
[0079] The vortices thus created lead to an increase in vibrations, a loss of lift and a reduced controllability of the rotary wing aerodyne 1. This results in a significant loss of altitude, which can be fatal if the aerodyne 1 is close to the ground when the vortex phenomenon occurs, particularly during the approach phases.
[0080] Referring to
[0081] The duration of the test flight ranges from 0 s to 34 s, with times t1, t2, t3 and t4 between 0 and 34 s.
[0082] The vortex phenomenon or vortex ring state is due to the appearance around the rotor disc of vortex rings generated by the re-ingestion by the rotor 2 of its own flow, the latter being blocked by the air mass present under the rotary wing aerodyne 1.
[0083] The random formation and decay of these vortices over time creates a highly disturbed and unsteady flow around the rotor 2. Moving in this environment, the blades 3 are then subjected to a strongly disturbed aerodynamic field, which modifies in a significant way their local incidences, and thus their loads. Lift and drag are then strongly modified, resulting in a vibration level different from that measurable in normal flight conditions as can be seen in
[0084] The dotted curve in
[0085] Once entered into vortex ring state, the aerodynamics on blades 3 is so disturbed that a sudden loss of lift occurs, leading to an important loss of altitude that can be potentially dangerous. However, before this happens, the vortices begin to form and alter the aerodynamics on the blades 3, and thus alter the vibrational response of the main rotor 2, and then the vibrational response of the rest of the rotary wing aerodyne 1.
[0086] Referring to
[0087] The helicopter type rotary wing aerodyne 10 comprises an airframe 11 consisting of a fuselage 12 and a landing gear 13.
[0088] The rotary wing aerodyne 10 further comprises a main rotor 14 carrying a plurality of blades 15.
[0089] It should be noted that some parts of the aerodyne 10, such as the blades 15, the tail of the fuselage 12 and the cabin fairing in the fuselage 12, have not been shown in
[0090] The rotary wing aerodyne 10 according to the present invention further comprises a device for detecting the approach of a vortex ring state comprising three vibration sensors 16a, 16b, 16c arranged in the rotary wing aerodyne 1 and configured to detect at different points of the rotary wing aerodyne 10 vibrations at least along the longitudinal axis X and the vertical axis Z (shown in
[0091] As the main rotor 14 is mechanically linked to the fuselage 12 of the aerodyne 10, the vibrations of the main rotor 14 are transmitted to the fuselage of the aerodyne 10. These vibrations are thus measured in real time using the vibration sensors 16a, 16b, 16c.
[0092] It should be noted that the detection device could also comprise at least two vibration sensors 16a, 16b, 16c, without departing from the scope of the present invention.
[0093] Each vibration sensor 16a, 16b, 16c is a three-axis accelerometer, namely a longitudinal axis X corresponding to the longitudinal axis of the rotary wing aerodyne 10, a transverse axis Y corresponding to the transverse axis of the rotary wing aerodyne 10 and a vertical axis Z.
[0094] The three vibration sensors 16a, 16b, 16c thus consist of a first three-axis accelerometer 16a installed in the cabin 17 of the rotary wing aerodyne 10 under the pilot's seat 17a, a second three-axis accelerometer 16b installed on the main gearbox 18 of the rotary wing aerodyne 10, preferably on the lower fitting of the left rear suspension bar of the main gearbox 18, and a third three-axis accelerometer 16c installed on the engine deck 19 of the rotary wing aerodyne 10 near the centre of gravity of the rotary wing aerodyne 10.
[0095] The detection device further comprises a computer type data processing unit 20 connected to the three vibration sensors 16a, 16b, 16c in a wired or wireless manner.
[0096] The data processing unit 20 is configured to: receive in real time the vibration measurement data from the three vibration sensors 16a, 16b, 16c; process the vibration measurement data received so as to calculate in real time the vibration spectrum of the rotary wing aerodyne 10 along at least one of the longitudinal X and vertical Z axes of the rotary wing aerodyne 10 on at least two distinct vibration sensors among the three vibration sensors 16a, 16b, 16c; detect or not in real time, by vibration analysis, the approach of a vortex ring state according to the calculated vibration spectrum; and issue an alarm in the event of detection of the approach of a vortex ring state.
[0097] The vibration measurement data from the three vibration sensors 16a, 16b, 16c are preferably sampled at a frequency of 4000 Hz to ensure a frequency measurement bandwidth of up to 2000 Hz.
[0098] Based on the direct measurement of vibrations 21 due to the appearance of vortex rings 22 on the blades 15 of the main rotor 14, the detection device according to the present invention thus makes it possible to detect the approach of the vortex phenomenon in a reliable manner.
[0099] The appearance of a characteristic vibration spectrum in the rotary wing aerodyne 10 is due to the passage of the blades 15 of the main rotor 14 into the vortices 22 before these are sufficiently developed to cause the drop in lift. Detection of the approach of the vortex phenomenon by this means is therefore achieved moments before the actual entry into the vortex phenomenon, thus providing a potentially sufficient margin of time to warn the pilot or perform corrective actions through the flight control system of the rotary wing aerodyne 10, as will be described in more detail later.
[0100] In particular, the data processing unit 20 is configured to perform at least one of: [0101] a first vibration analysis on the measurement data streams corresponding to the longitudinal axis X and the vertical axis Z of the second three-axis accelerometer 16b comprising, for each of the measurement data streams corresponding to the longitudinal axis X and the vertical axis Z of the second three-axis accelerometer 16b: [0102] comparing the fundamental frequency of the main rotor 14 of the rotary wing aerodyne 10 with the predominant frequency, i.e. corresponding to the maximum amplitude level, of the vibration spectrum derived from a discrete Fourier transform on the corresponding measurement data stream, and [0103] if the difference between the predominant frequency and the fundamental frequency of the main rotor 14 is greater than a predetermined frequency difference threshold, generating a detection signal associated with the corresponding measurement data stream; and [0104] a second vibration analysis on the measurement data streams corresponding to the vertical axis Z of the first three-axis accelerometer 16a and corresponding to the longitudinal axis X and vertical axis Z of the third three-axis accelerometer 16c comprising, for each of the measurement data streams corresponding to the vertical axis Z of the first three-axis accelerometer 16a and corresponding to the longitudinal axis X and vertical axis Z of the third three-axis accelerometer 16c [0105] digital filtering without phase shifting of the corresponding measurement data stream using a low-pass finite impulse response filter and a high-pass finite impulse response filter; [0106] calculating the RMS value of the signal from the low-pass filter and the RMS value of the signal from the high-pass filter, [0107] calculating an energy ratio corresponding to the RMS value of the signal from the high-pass filter divided by the RMS value of the signal from the low-pass filter, and [0108] if the energy ratio is greater than a corresponding predetermined energy ratio threshold, generating a detection signal associated with the corresponding measurement data stream.
[0109] Of the nine measurement channels provided by the three three-axis accelerometers 16a, 16b, 16c, only five channels are thus used for data processing and vibration analyses, namely: [0110] for the first vibration analysis, the channels corresponding to the longitudinal axis X and the vertical axis Z of the second three-axis accelerometer 16b (hereinafter referred to as “channel 2” and “channel 3”, respectively); and [0111] for the second vibration analysis, the channel corresponding to the vertical axis Z of the first three-axis accelerometer 16a (hereinafter referred to as “channel 1”) and the channels corresponding to the longitudinal axis X and vertical axis Z of the third three-axis accelerometer 16c (hereinafter referred to as “channel 4” and “channel 5”, respectively)
[0112] The first vibration analysis concerns channels 2 and 3 of the accelerometer 16b, and consists in comparing the fundamental frequency of the main rotor 14 corresponding to the current speed of the main rotor (F.sub.rotor=(NR*b)/60, where F.sub.rotor is the fundamental frequency of the main rotor 14, NR is the speed of rotation of the main rotor 14, and b is the number of blades 15 of the main rotor 14) and the predominant frequency of the vibration spectrum resulting from the discrete Fourier transform carried out on the measurement data streams from channels 2 and 3. A weighting window of the “hanning” type is preferably used in the calculation of the discrete Fourier transform, this preferably being 2 seconds wide, equivalent to 8000 data samples, with a refresh rate of 0.0625 s.
[0113] Thus, during this first vibration analysis, if the predominant frequency is different (beyond the frequency difference threshold predetermined by the flight test analysis and which can be adapted according to the aerodyne 10) from the fundamental frequency of the rotor 14 F.sub.rotor, then a detection signal is generated by the first vibration analysis on the corresponding channel. This first vibration analysis can therefore potentially provide two detection signals, namely a possible detection signal associated with channel 2 and a possible detection signal associated with channel 3.
[0114] The second vibration analysis is performed on channels 1, 4 and 5 of accelerometers 16a, 16c. Over a range of data from these channels (e.g. N=8000 samples), digital filtering without phase shifting is performed, using low-pass (preferably cut-off frequency 19.5 Hz) and high-pass (preferably cut-off frequency 150 Hz) finite impulse response (FIR) filters (preferably of order 100), with each measurement data stream filtered in this way having two signals hereinafter denoted x.sub.low and x.sub.high.
[0115] The RMS value of each of the filtered streams is then calculated, for example for the signal from the high-pass filter:
where RMS.sub.high is the RMS value from the high-pass filter and N is the number of samples.
[0116] Then, an “energy ratio” corresponding to the RMS value of the signal from the high-pass filter divided by the RMS value of the signal from the low-pass filter is calculated: ER=RMS.sub.high/RMS.sub.low, where ER is the energy ratio, RMS.sub.high is the RMS value of the signal from the high-pass filter and RMS.sub.low is the RMS value of the signal from the low-pass filter
[0117] Each “energy ratio” of channels 1, 4 and 5 has a specific predetermined energy ratio threshold.
[0118] For a detection signal to be generated by the second vibration analysis, one of the energy ratios must therefore be above its predetermined specific energy ratio threshold.
[0119] It should be noted that the thresholds used in both vibration analyses could also be modified depending on the type of aerodyne 10 and/or the type of main rotor 14.
[0120] Preferably, the first vibration analysis and the second vibration analysis are performed simultaneously by the data processing unit 20. Indeed, since a significant increase in the energy of the vibration intensity at high frequencies is not a feature specific to the state of vortex rings, it follows that the combination of both vibration analyses is preferable in order not to generate false alarms.
[0121] Preferably, the data processing unit 20 is configured to detect the approach of a vortex ring state and issue an alarm when the first vibration analysis and the second vibration analysis simultaneously generate at least one detection signal, thereby avoiding the generation of false alarms.
[0122] More preferably, the data processing unit 20 is configured to detect the approach of a vortex ring state and generate an alarm when at least three detection signals from the first and second vibration analyses are generated simultaneously, thereby further avoiding the generation of false alarms and improving the reliability of detecting the approach of the vortex phenomenon.
[0123] The data processing unit 20 may be further configured to use at least one of the following measurements, provided by the basic instrumentation of the rotary wing aerodyne 10, to detect the approach of a vortex ring state: measuring the indicated speed of the rotary wing aerodyne 10, measuring the vertical speed of the rotary wing aerodyne 10, and measuring the rotation speed of the main rotor 14 of the rotary wing aerodyne 10, so as to obtain additional information about the flight domain in which the rotary wing aerodyne 10 is at the current time. This avoids false alarms if the vibration level increases significantly while the aerodyne 10 is clearly outside the vortex domain (for example, during dynamic manoeuvres at high forward speeds, or during a climbing flight).
[0124] In each of the first and second vibration analyses, the indicated speed measurement and the vertical speed measurement may, for example, be used by the data processing unit 20 in the following manner: generating the detection signal only if in addition the indicated speed measurement is less than 25 knots and the vertical speed measurement is less than −700 feet per minute.
[0125] The detection device according to the present invention may also comprise a warning unit (not shown in
[0126] The time margin available between the time of detection and the appearance of the vortex phenomenon allows to implement the function of warning the pilot of the aerodyne 10, so that the pilot can avoid the appearance of the vortex ring state by manually changing the piloting of the aerodyne 10.
[0127] The warning, generated in the event of approaching vortex ring state in order to alert the crew of the aerodyne 10, may thus be one of an audible warning from at least one loudspeaker disposed in the cabin 17 of the aerodyne 10, a visual warning from at least one light indicator and/or at least one display device provided in the cabin 17 of the aerodyne 10, and a haptic warning (e.g. by haptic feedback on the force feedback control stick of the aerodyne 10).
[0128] The rotary wing aerodyne 10 further comprises a flight control system (not shown in
[0129] The time margin available between the moment of detection and the appearance of the vortex phenomenon thus allows the implementation of a pilot assistance function via the flight control system of the aerodyne 10.
[0130] In the event of detection of the approach of the vortex ring state by the detection device installed in the aerodyne 10, the flight control system of the aerodyne 10 can thus automatically or semi-automatically modify the flight parameters of the aerodyne 10 in order to prevent the aerodyne 10 from entering the vortex ring state (by reduction of the rate of descent or actions on the swashplate (longitudinal or lateral), for example).
[0131] The data processing unit 20 may also be configured to perform a third vibration analysis comprising: applying a Hilbert transform to the measurement data stream corresponding to the vertical axis Z of the first three-axis accelerometer 16a to extend the real signal into the complex domain; applying a Fast Fourier Transform (FFT) to the Hilbert transformed signal; and filtering the signal obtained by FFT by plus or minus 5 Hz around the fundamental frequency of the main rotor 14 corresponding to the current speed of the main rotor 14 (Frotor=(NR*b)/ 60, where Frotor is the fundamental frequency of the main rotor 14, NR is the rotation speed of the main rotor 14, and b is the number of blades 15 of the main rotor 14); reconstituting the filtered signal by applying an inverse Fourier transform.
[0132] It is then possible to represent the energy distribution of the reconstructed signal in the time-frequency plane via the square of the modulus of the short-term Fourier transform, this representation is called a spectrogram. A spectrogram is therefore produced on the signal previously filtered around the main rotor frequency, from which the non-physical part of the spectrum (negative frequency part) is removed.
[0133] As the approach to the vortex state is characterized by an increase in vibration energy, the variation in the signal amplitude is analyzed. For this purpose, the square root of the variance or, equivalently, the root mean square of the deviations of the signal from its mean is calculated (standard deviation) over a sliding window of 0.7 s. A threshold is defined from the average of this standard deviation.
[0134] When the standard deviation of the signal is above this threshold, a detection signal is generated. It should be noted that the threshold used in this third vibration analysis could also be modified depending on the type of aerodyne 10 and/or the type of main rotor 14.
[0135] In order to make the detection more robust, it is possible to apply the third vibration analysis to a whole part of the measurement channels of the accelerometers 16a, 16b, 16c in order to correlate the analyses and avoid the generation of false alarms. Similarly, it is possible to apply the third vibration analysis to other frequency harmonics of the main rotor 14.
[0136] It should be noted that this third vibration analysis is totally independent of the first and second vibration analyses and can replace them. However, for greater robustness, this third vibration analysis can also be associated with the first and second vibration analyses in order to correlate the detections and potentially avoid false alarms.
[0137] It is understood that the particular embodiment just described is indicative and non-limiting, and that modifications may be made without departing from the present invention.