Title: Algorithms to monitor damaging events on a plane blade with an autonomous embedded microsystem

Authors:

Mathieu LASTAPIS1,2 Christophe ESCRIBA1,2 Jean-Yves FOURNIOLS1,2

1

CNRS; LAAS; 7 avenue du colonel Roche, F-31077 Toulouse Cedex 4, France Université de Toulouse; UPS, INSA, INP, ISAE ; UT1, UTM, LAAS ; F-31077 Toulouse Cedex 4, France 2

Tel: +33 561 336 494, Fax: +33 561 336 208, [lastapis, escriba, fourniols]@laas.fr

ABSTRACT Monitoring of aircrafts composites structures is part of today challenge named SHM, for Structural Health Monitoring. Turboprop blade surveillance is part of it. In this context we devised a microsystem embedded on these structures in order to monitor various blade parameters such as speed, vibrations or impacts. This paper intends to show results obtained in blade monitoring data processing. These works lead to a dedicated algorithm of surveillance based on a single accelerometer.

INTRODUCTION In several fields, manufacturers aim to monitor their structure, especially when they are in composites. A domain pools these wishes and is named SHM. These researches are dedicated to better plan maintenance calendars. Also manufacturers will be able to improve their products by adapting them to what they are subjected to during their life. The plane propeller blade is one of these aircrafts composites structures. Previously, we devised a microsystem able to monitor a propeller blade in real time. This monitoring improves blade safety and maintenance. Its design is based on accelerometers in order to detect various damaging events:   

impacts from birds happening at taking off or when landing, stones when plane uses unprepared runways, or bullets when plane is intended for military purposes, overspeeds, which can induce damaging aerodynamic loads, overvibrations, which can lead to cracks into the composite structure.

In order to diagnose blade structural health, our microsystem processes in real time signals from the sensor. Then detected event is recorded into memory. Data extraction is made through USB interface to be post-analysed. These devices are detailed in previous papers: one is devised for Transall blade [1] (Fig. 1) while another one targets drone blade [2]. This paper is dedicated to results based on aircraft blades. _____________ Mathieu Lastapis, Christophe Escriba, Jean-Yves Fourniols. CNRS; LAAS; 7 avenue du colonel Roche, F-31077 Toulouse Cedex 4, France. Université de Toulouse; UPS, INSA, INP, ISAE ; UT1, UTM, LAAS ; F-31077 Toulouse Cedex 4, France.

Fig. 1: Transall blade embedded board

Monitoring turboprop blades is a new challenge in research. Equivalent works can be found on helicopter [3] or more commonly on wind turbine blades [4-6]. A company has even emerged from this research field: Insensys [7]. Also, SmartFibres sell systems based on optical fibres to monitor such structures [8]. Unfortunately plane blades do not offer the same constraints. Volume is much lower and each blade must be equipped individually. That's why we have chosen a system based on a single accelerometer. This one is fitted on blade at a sensitive place whereas main system (signal conditioning, power, data processing...) is at blade bottom [2]. Blade monitoring process is based on a dedicated algorithm. First part concerns impact detection and localization. Then techniques to detect overvibrations and overspeeds are shown. Finally, a global algorithm is processed with previous analysis. This one is ready to be embedded for flight tests.

IMPACTS DETECTION AND LOCALIZATION Impacts on a blade are the only events which stimulate eigenfrequencies of the structure. So impacts detection is based on eigenfrequencies surveillance. After processing the signal to obtain a spectrum (Fig. 2), embedded algorithm looks for various proper modes of the blade. If they appear, the conclusion that an impact has happened can be done.

Fig. 2: Blade eigenfrequencies stimulated by an impact

A second important aim for the impact monitoring is localization. Indeed, blade is mechanically more or less sensitive among shock localization. An impact

at blade bottom is less critical than higher. So to achieve such a work, localization of these impacts is done by measuring differences between eigenfrequencies amplitudes (Fig. 3). This property comes from the half-constrained blade structure. Excited eigenfrequencies are different among impact position. A bottom impact engenders high frequencies while a top impact creates more low frequencies.

(a)

(b) Fig. 3: Relative amplitude shift depending on impact localization ((a) top impact and (b) bottom impact)

A lab demonstrator has been developed to show this behaviour (Fig. 4). On the blade of an ATR airplane, an accelerometer is mounted. Its signal is converted in digital through a National Instruments acquisition card. Then LabView processes signals among previously shown techniques. Finally it displays results by discriminating three zones on the blade (green spots on Fig. 3). Localization is reliable at 75 % (15 impacts well placed for 20 tries).

Fig. 4: screen capture of lab demonstrator of impact detection and localization

VIBRATIONS AND SPEED MONITORING Last points to monitor on a blade are vibrations and rotating speeds. These events can be damaging for the structure if they are above a certain level. Here, data in this paragraph comes from strain gauges because flight tests with the designed embedded system have not been run yet. While flying, blade vibrates almost like a pure sinusoidal signal. Depending on flight phases (take off, fast climbing ...), signal maxima vary. Knowing various constraints and normal vibrations, overvibrations are detected by watching if signal is over predefined thresholds (Fig. 5).

Fig. 5: Vibrations during a flight and illustrations of the technique to detect overvibrations

About overspeeds, monitoring technique is based on spectral analysis. During a flight, blade vibrates at its rotating speed. So measurement of this speed is achieved by processing a frequency analysis on the sensor signal (Fig. 5). Result is shown Fig. 6. It gives a main frequency at 14.04 Hz, which matches standard rotating speed of A400M propellers: 842 RPM.

Fig. 6: spectral analysis of a vibrating signal during a flight

120 samples acquisition (fs = 600Hz; Tacq = 200ms)

Eigenfrequencies?

Frequency analysis

Min and max measurement

Speed measurement

Above or under threshold? Y

Y

Localization by eigenfrequencies amplitude analysis

Overspeed? Y Y

Impact detection and localization

Overspeed detected

Overvibrations detected

Event memory storage with timestamping and stored by level

Fig. 7: Algorithm to monitor propeller blades

GLOBAL ALGORITHM PROCESSING By compiling all previous monitoring techniques, an algorithm can be processed. Its aim is to be embedded on the microsystem. So it has to be energy and time saving. Fig. 7 shows this algorithm. It starts by acquiring samples from accelerometer signal. Then the spectral analysis and the research of maxima/minima are processed. The second one gives result if overvibrations happen. Spectral analysis leads to speed measurement and thus overspeed detection. But it allows checking if eigenfrequencies appear in signal spectrum. If so, another algorithm localizes the impact on the blade. In any cases when a damaging event is detected, a record with time stamping is written in flash memory. To minimise data recording, event are classified among their damaging capabilities: for example overspeeds are marked "level 1" as light and possibly without consequences. "Level 5" represents an important overspeed and damages are very probable.

CONCLUSION This paper intends to show how to monitor a turboprop blade with a single accelerometer. Such a solution allows a maximal integration for a minimum power. It gives good results even if a long flight test campaign will check them soon.

REFERENCES [1]

[2]

[3] [4] [5] [6] [7] [8]

Lastapis, M., C. Escriba, Y. Grondin, G. Auriol, P. Amat, S. Andrieu, J. Strak, J.-L. Boizard, and J.-Y. Fourniols. 2010."Blade recorder microsystem design and validation for aeronautical structural health monitoring". in European Workshop on Structural Health Monitoring. Sorrento, Italy: DEStech publications. Lastapis, M., C. Escriba, G. Auriol, E. Albu, P. Berthou, J.-L. Boizard, J.-M. Dilhac, and J.-Y. Fourniols. 2010."Embedded blade microsystem and events recorder for drone structural health monitoring". in Asian-Pacific Workshop on Structural Health Monitoring. Tokyo, Japan. Roy, N. and R. Ganguli. 2005. "Helicopter rotor blade frequency evolution with damage growth and signal processing," Journal of Sound and Vibration, 283(3-5):821-851. Rumsey, M.A. and J.A. Paquette. 2008. "Structural health monitoring of wind turbine blades," Smart Sensor Phenomena, Technology, Networks, and Systems 2008, 6933:450. Ciang, C.C., J.R. Lee, and H.J. Bang. 2008. "Structural health monitoring for a wind turbine system: a review of damage detection methods," Measurement Science & Technology, 19(12):-. Ghoshal, A., M.J. Sundaresan, M.J. Schulz, and P.F. Pai. 2000. "Structural health monitoring techniques for wind turbine blades," Journal of Wind Engineering and Industrial Aerodynamics, 85(3):309-324. Insensys. 2010; Available from: http://www.insensys.com. Smart Fibres. 2010; Available from: http://www.smartfibres.com.