Title: Blade recorder microsystem design and validation for aeronautical structural health monitoring
Authors:
Mathieu Lastapis Christophe Escriba Yannick Grondin Guillaume Auriol Pascal Amat Stéphane Andrieu Jérôme Strak Jean-Louis Boizard Jean-Yves Fourniols
ABSTRACT Structural health monitoring through sensor network based on microsystems is an actual challenge in aerospace and especially for composite structure monitoring. As an example, we devise a blade microsystem recorder as a solution to improve predictive safety through a real time monitoring able to detect: shocks like stone, bullet or bird impacts, aerodynamic charges. In this paper, we present an integrated recorder electronic system able to determine whether the blade structure is damaged or not. Our approach is based on accelerometer detection even if the centrifugal acceleration can reach over 6000g taken in account huge temperature variations! This dedicated solution offers better diagnosis than strain gauges and the results demonstrate shocks, overspeeds and general structural health monitor with our sensor integrated in the blade. Energy harvesting is discussed either through engine sounds converted into electrical energy thanks to the combination of serial or parallel piezo nor de-ice blade system combined with induction magnetic coupling.
INTRODUCTION Structural health monitoring is an important domain of research to reduce maintenance and inspections costs of various structures. Monitoring a structure allows anticipating more serious damages and repairing or changing before it breaks. It applies in particular to composite structures which are more and more used in aeronautics [1, 2], or in energy fields [3-5]. _____________ Mathieu Lastapis, Christophe Escriba, Yannick Grondin, Guillaume Auriol, Jean-Louis Boizard, Jean-Yves Fourniols, Université de Toulouse, CNRS ; LAAS ; 7 avenue du colonel Roche, F-31077 Toulouse, France. Pascal Amat, Stéphane Andrieu, Jérôme Strak, Ratier-Figeac, route de Cahors, BP. 2, 46101 Figeac, France.
For example, A380 uses 25% and the next A350WB will be composed of 45% composite material [6, 7]. Blades are concerned by this new composite material whatever if it is helicopter blades [8], wind turbine blades [4] or aircraft propeller blades. About the last one, literature is poor. This paper will give a few ideas of what we can meet on an aircraft propeller blade.
1. MONITORING AN AIRCRAFT PROPELLER BLADE: A TRUE CHALLENGE A harsh environment for electronics Nowadays, aircraft propeller blades are made by composite fibres even on the next Airbus A400M: each 8-bladed propeller will be powered by one of the four engines. This composite structure induces fragility to vibrations, impacts or too heavy loads. Moreover, they can fly in a very harsh environment: harsh for them, but also for an embedded electronics: temperature can reach -55°C or climb up to +90°C. Even if electronics devices can face, it is more difficult to embed energy. At -55°C, a few batteries can work. We will talk about this point later in this paper. For the A400M aircraft, blades rotate at 842 rpm in normal conditions. It can rotate up to 1200 rpm if overspeed is needed. These rotational speeds create a huge centrifugal acceleration: it reaches over 4000g at the top of the blade. On an ATR 42500 blade for example, values of 6000g can happen! These accelerations imply to embed a minimum of weight and to be sure it is well-attached: in our approach we consider that to keep the blade balanced 200 grams is a reasonable weight to add inside the blade with a volume inferior to 10 cm3. In this available space, we have to insert the electronics, immune against strong electromagnetic waves and water/solvent resistant (field emission lower than 70dB (µV/m), according to MIL-STD461 E standards). The most destructive event for a blade, and so for embedded electronics, is the various impacts met by aircrafts: bird impacts happen generally at the take-off and can be very damaging. To bird (that concerns both civilian and military) we have to add bullet and stone impacts (when landing to or taking off from rough runway). Finally, this system must be autonomous in energy. Indeed, no wire can be driven to the blade to power an electronics system. The only wires presents on a blade are used for de-icer, and they bring energy temporarily. So it is an interesting energy scavenging source, but in any case, system must run with a minimum current consumption. An order of hundreds of µA seems to be honest for an embedded autonomous system. Technology choice: optical fibres, strain gauges, piezoelectric sensors or accelerometers? Several technologies are available for monitoring a blade. We will only focus on non-destructive testing (NDT) methods. Four main families can be identified [9, 10]. i) Optical fibres useful in many approaches [11], present two main problems: their current consumption and their processing tools needed.
ii) Strain gauges are very low cost and can give very precise measures [12]. In a laboratory environment, they are the perfect candidate for SHM. But in industry, their placement is a major drawback. It takes too much time and need a skilled technician to stick it. Moreover, due to sticking conditions, gages’ response can mismatch. iii) Piezoelectric sensors are the most widely used sensors for damage detection based on sound and ultrasound [9, 10]. They are very accurate to detect damages or fatigue cracks when used as a matrix. iv) Compared to piezoelectric sensors, accelerometers get several advantages: they are low cost, small sized, long life of use, light weight and able to measure 3 dimensions. Hermetically sealed in a package, accelerometers are quite resistant to impacts and can be immunized against electromagnetic waves if the package is metallic. Their only drawback concerns their placement when a huge centrifugal acceleration is present [13]. A misalignment between the axis supposed to measure centrifugal acceleration and this acceleration can lead to an undesirable offset on others axis, and even more to saturation. So a great care combined with specific calibration, must be brought to this point, even if laser placement may help a lot. We based our choice on accelerometer for designing our microsystem.
2. DEVELOPMENT OF AN INTEGRATED BLADE MICROSYSTEM RECORDER Our embedded system is based on an accelerometer whose outputs are recording in a memory through a microcontroller. For this prototype, energy is provided by a single Li-SoCl2 cell able to operate from -60°C to +85°C. Currently, data is extracted through an USB wire. A wireless link is being developed for a better use. A functional diagram is presented Figure 1.
Accelerometer
Li-SoCl2 cell
Microcontroller (energy management, turnon time, status, USB2.0, real-time clock, bandgap reference…)
8 Gbits flash memory
HMI microsystem PC
Figure 1 : Functional diagram of the microsystem (left) and picture of the microsystem 63mm*32mm*8mm without power cell or 18mm with the cell (right)
Able to record up to 9 channels at 2 kHz, this system (Fig. 1) requires 7 mW in a full-operational mode where we can record up to 8 Gbits. To ensure the robustness to temperature variations (-40°C to +85°C with our setup) a thermal test is detailed Figure 2, where we observe lower than 50mV variations.
2
100
1.9
1.8
1.7
1.6
Voltage (V)
Temperature (°C)
50
0 1.5
1.4
-50
0
100
200
300 Time (minutes)
400
500
1.3 600
Figure 2 : Thermal sensitivity of accelerometer (temperature in dash line)
3. FIRST RESULTS IN BLADE BEHAVIORAL ANALYSIS Our microsystem presents the capability to monitor and analyse several blade parameters: i) Vibrations detection as seen on Figure 3. This figure presents outputs of accelerometer during a mono-axial vibration test on blade. It results in a more responsive axis which is longitudinal to the excitation one. ii) Impacts are monitored by our microsystem as demonstrated on Figure 4. Here is presented a ping impact on a blade in static conditions (blade not rotating). iii) Overspeeds will be monitored using earth gravity field. Each rotation induces a plus or minus 1g on the accelerometer. Similar works has ever been done recently [14] but on wind turbine blades and without an embedded system. Monitoring these three parameters in real time allows us to know whether the blade has been damaged. A specific signal processing has been developed to extract this analysis. .
Sorties de l'accéléromètre "pale" en fonction du temps 1.8
X-axis Y-axis Z-axis
1.75
1.7
X
Z
1.65 Voltage (V)
Y
1.6
1.55
1.5
1.45
5.483
5.484
5.485
5.486 5.487 Time (seconds)
5.488
5.489
5.49
Figure 3 : Definition of the blade orthonormal basis (left) and vibrations detected by the accelerometer without centrifugal effort (right)
6
Acceleration (g)
4 2 0 -2 -4
0
0.1
Densité spectrale d'énergie(V².sec²)
-3
1
x 10
0.2
0.3
0.4
0.5 0.6 Time (seconds)
0.7
0.8
0.9
1
Figure 4 : Impact recorded by our microsystem (static test)
0.8
4. FIRST RESULTS IN ENERGY HARVESTING 0.6
The association of a harvesting solution to the battery has been investigated 0.4 through two approaches: i) 0.2Transformation of vibrations into electrical energy using piezoelectric sensors. Figure 5 shows results of the power available with 4 piezoelectric 0 10devices placed 10 10 10 10 kHz) associated 10 at 12cm from noise source10(92 dB, 3.7 Fréquence en Hz serial-parallel with the optimum configuration. With a load resistance of 1060 µW (Figure 5a). This power is efficient to switch on accelerometer, analog-to-digital conversion and realtime clock monitoring on our blade recorder. Only the memory could not be supplied. -1
0
1
2
800
900
750
800
700
Power (µW)
1000
700
5
650
600
600
500
550
400
4
850
1100
Power (µW)
3
1
2
3
4
5 6 Resistance (kohms)
7
8
9
10
500
1
2
3
4
5 6 Resistance (kohms)
7
8
9
10
Figure 5: Power retrieved with 4 piezo association
ii) De-ice blade system combined with induction magnetic coupling: results show that we can retrieve 25 µW at 400 Hz combining six coils of 100µH (Figure 6).
26
24
22
Power (µW)
20
18
16
14
12
10
8
1
2
3
4
5 6 Resistance (kOhms)
7
8
9
10
Figure 6: Power retrieved by magnetic induction
CONCLUSION In this paper we devised an embedded SHM system able to monitor key parameters in order to diagnose structural health monitoring inside the blade. Based on accelerometer, our system is capable of detecting vibrations, impacts or overspeeds. Estimation of eventual blade damages depends on the processing of these detections. Energy harvesting solutions using acoustic vibrations transformation through piezo sensors association in parallel and the defrost system of the blade has been presented.
REFERENCES 1. 2.
3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
Paget, C. 2008. "Bonded thin metallic foil sensor applications for structural health monitoring of aeronautical structures," Physical Mesomechanics, 11(5-6):308-313. Bovio, I., et al. 2003. "Damage detection and health monitoring based on vibration measurements and recognition algorithms in real-scale aeronautical structural components," Damage Assessment of Structures, 245(2):519-526. 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):-. 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. Ghoshal, A., et al. 2000. "Structural health monitoring techniques for wind turbine blades," Journal of Wind Engineering and Industrial Aerodynamics, 85(3):309-324. Staszewski, W.J., S. Mahzan, and R. Traynor. 2009. "Health monitoring of aerospace composite structures Active and passive approach," Composites Science and Technology, 69(11-12):1678-1685. Walz, M. 2006. "The dream of composites, R&D, Features – Advantage Business Media." Matsuzaki, R., M. Melnykowycz, and A. Todoroki. 2009. "Antenna/sensor multifunctional composites for the wireless detection of damage," Composites Science and Technology, 69(15-16):2507-2513. Shang, B., B. Song, and F. Shang. 2008."New Sensor Technologies in Aircraft Structural Health Monitoring". in International Conference on Condition Monitoring and Diagnosis. Prosser, W.H. 2003. "Development of structural health management technology for aerospace vehicles," JANNAF 39th. Ecke, W. and K. Schroder. 2008. "Fiber Bragg grating sensor system for operational load monitoring of wind turbine blades," Smart Sensor Phenomena, Technology, Networks, and Systems 2008, 6933:I9330-I9330. Haksoo, C., C. Sukwon, and C. Hojung. 2008."Structural Health Monitoring system based on strain gauge enabled wireless sensor nodes". in 5th International Conference on Networked Sensing Systems. Roundy, S. 2008."Energy harvesting for tire pressure monitoring systems: design considerations". in PowerMEMS. White, J.R., D.E. Adams, and M.A. Rumsey. 2009."Sensor acceleration potential field identification of wind turbine rotor blades". in 7th International workshop on structural health monitoring.
