Blade structure health monitoring system

Abstract

A rotating system comprising two or more blades 3 mounted on a hub installed on a rotatable propeller shaft 1, each blade provided with a respective sensor 4 arranged to detect response of the respective blade to harmonic excitation; and the system further comprising means configured to compare the response of the respective blade to that of the other blade(s).

Claims

1. A method of monitoring deflection of a blade in a rotating system comprising two or more blades mounted on a rotatable blade shaft, the two or more blades subjected to a one per revolution, 1P, sinusoidal excitation, generated by an inflow angle of a flow field relative to a center line of rotation of the system; the method comprising: detecting response of the two or more blades to the forced one per revolution, 1P, sinusoidal excitation via signals output from two or more sensors, each sensor of the two or more sensors disposed at a blade of the two or more blades; performing a fast Fourier transform (FFT) on the signals output from the two or more sensors for at least one full revolution of the rotating system; computing an average amplitude for the two or more blades of the FFT harmonic corresponding to a frequency of 1P; comparing, for each blade of the two or more blades, a 1P harmonic amplitude of the blade with the average amplitude for the two or more blades; determining, for each blade of the two or more blades, a difference between the 1P harmonic amplitude of the blade and the average amplitude for the two or more blades; and determining, for each blade of the two or more blades, that the blade is faulty if the difference exceeds a predetermined threshold.

2. The method of claim 1, further comprising issuing a notification if the difference exceeds the predetermined threshold.

3. The method of claim 1, wherein the detecting is performed whilst the blades are rotating during operation of the rotating system.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 shows a side view of a propeller system with sensors provided for each blade.

(2) FIG. 2 is a schematic view of the system shown in FIG. 1.

(3) FIG. 3 is a schematic view of an alternative embodiment.

(4) FIG. 4 is a schematic view of a further embodiment.

(5) FIG. 5 is a flow chart showing the methodology of one embodiment.

(6) FIG. 6 is a flow chart showing a second embodiment.

(7) FIG. 7 is a flow chart showing an alternative embodiment.

(8) FIG. 8 is a flow chart showing another embodiment.

(9) FIG. 9 is a flow chart showing another embodiment.

DETAILED DESCRIPTION

(10) Referring to FIG. 1, a propeller comprises a propeller shaft 1 that rotates a hub 2 on which propeller blades 3 are mounted.

(11) In the embodiment shown in FIG. 1, sensors 4 (4.sub.1, 4.sub.2 . . . 4.sub.N for N blades as shown in FIG. 2) are mounted on or embedded in each blade 3 and are connected, e.g. by wiring 5, to a blade prognostic health monitoring (BPHM) control unit 6 arranged, in this case, on the hub.

(12) The BPHM control unit 6 is connected to an FADEC (full authority digital engine control) 7 via a brush block slip ring assembly 8. The FADEC can, instead, be an aircraft maintenance computer AMC.

(13) This arrangement is shown in block-diagram form in FIG. 2. In this arrangement, the detection and comparison computations are formed centrally in the BPHM control unit and fault messages are transmitted to the FADEC or AMC.

(14) In the embodiment shown in FIG. 3, each blade is provided with, in addition to a sensor, an embedded blade computing unit EBCU 9 to allow for a decentralised monitoring of the individual blade conditions, but the comparisons are formed centrally in the BPHM.

(15) In another embodiment shown in FIG. 4, each blade is provided with an embedded blade health monitoring computing unit 10 instead of a central BPHM computing unit and these units communicate with the FADEC or AMC via a communication bus, particularly a digital communication bus.

(16) In the embodiments of FIGS. 2 and 3, communication between the EBCUs and BPHM computing units and between the FADEC/AMC and the BPHM can be by means of an analogue or a digital communication bus.

(17) The health of the individual blades is, as mentioned above, determined based on a comparison of blade responses to harmonics. Different algorithms can be used to perform this comparison.

(18) FIG. 5 is a flow chart showing one example algorithm.

(19) For each blade (up to N blades), the sensor signal is acquired and a fast Fourier transform (FFT) is performed on the signals to produce data for one, two or more propeller turns.

(20) The average amplitude of the FFT first mode for all blades is then computed. FFT amplitudes of subsequent modes can also be used if necessary

(21) Whilst FFT computing is preferred, the average signals can also be derived without performing FFT.

(22) Then, for each blade, the first mode amplitude response is compared with the computed average amplitude.

(23) If the difference exceeds a predetermined threshold (in this example 5%), that blade is declared as faulty.

(24) The predetermined threshold of 5% is an example only and this may, for example, need to be larger to accommodate a sensor and processing errors as well as blade-to-blade scatter. The threshold can also be less than 5%.

(25) To ensure continuous monitoring of the blade health, the loop is repeated at determined intervals, for example each 100 ms . . . . Other intervals can be used, including intervals much longer than 100 ms.

(26) In an alternative algorithm, shown in FIG. 6, the rate of rotation of the propeller (the propeller RPM) is acquired from the FADEC.

(27) As with the example shown in FIG. 5, the average amplitude of the FFTs for all of the blades is computed but, in this example, the computation is of the average amplitude of the FFT harmonic corresponding to a frequency of 1P. This is determined from the acquired propeller RPM. 1P frequency is RPM/60.

(28) Then, for each blade, the 1 P harmonic amplitude is compared with the determined average. Again, if the difference exceeds a predetermined threshold, for example 5%, that blade is declared to be faulty. An advantage of this alternative algorithm is to provide a ‘filtering’ of the sensor signals that can eliminate noise and make the detection more accurate and robust.

(29) In an alternative embodiment shown in FIG. 7, the sensor signals for each blade are filtered and, at each blade, e.g. in an EBCU as shown in FIG. 3, an average amplitude for that blade is computed.

(30) An average amplitude for all of the blade sensor signals is then computed and the average amplitude determined for each blade is compared with the common blade average. Again, this is then compared with a predetermined threshold, e.g. 5%, and if the comparison exceeds the threshold, the blade is declared to be faulty.

(31) In the algorithm shown in FIG. 8, the sensor signals are acquired for each blade and the FFTs are determined for each blade as in the other examples. Then, essentially, the first mode amplitudes of the blades are compared pair-by-pair, for example the first mode amplitudes of the first and third blades are compared, the second and fourth blades, etc. If this comparison yields a difference exceeding a predetermined threshold, e.g. 5%, a comparison is then performed using different combinations of pairs of blades, e.g. blades 1 and 2, 3 and 4, etc. and the difference is again compared with the predetermined threshold. A blade with two detections exceeding the threshold is declared faulty.

(32) FIG. 9 shows an algorithm of an example having only two blades.

(33) Here, the FFTs for each blade are compared with each other for the first mode amplitudes and if the difference is greater than or equal to the predetermined threshold, e.g. 5%, then one of the blades is considered to be faulty and this triggers an inspection to determine which blade is faulty. Where only two blades are compared, in this algorithm, as there are only two blades to compare, it is not possible, in the algorithm, to isolate the faulty blade and this must be done by inspection.

(34) In the examples shown, the method is performed during operation of the rotating system/propeller i.e. whilst the blades are rotating during flight of an aircraft incorporating the propeller.