Autonomous piezoelectric Structural Health Monitoring system for on-production line use. C. Delebarre, S. Grondel and F. Rivart. IEMN-DOAE UMR CNRS8520, Université de Valenciennes et du Hainaut Cambrésis, Le Mont-Houy, 59313 Valenciennes cedex 9. France Corresponding author email:
[email protected] Abstract. The aim of this study is to control aeronautical composite structures on production line since they are extremely vulnerable to impact damages due to tool drops for example. To detect such an event on the structure, a piezoelectric wireless sensor network has to be developed. Therefore the feasibility of an innovative technique based on a piezoelectric harvesting device is presented. The system carries out Structural Health Monitoring tasks using piezoelectric transducers bonded onto the structure and is also fully autonomous since the same transducers are used to convert the mechanical vibrations of the structure into electrical power. As vibrations are not available during the production process, the autonomy of the system is provided thanks to a Lamb waves emitter located in the sensor array. After measurements, the system is able to harvest 7.36 milli-watts for a 100 milli-watts mechanical power applied to the structure. This electrical power can be used both by the electronic detector and the WIFI transmitter for the detection of impacts of less than 1 Joule. Keywords: Structural Health Monitoring, Piezoelectric power harvesting, impact detection, Piezoelectric ceramic.
1 INTRODUCTION Since aeronautical structures tend to be more and more sophisticated today, it becomes necessary to develop in situ Non Destructive Testing techniques for the monitoring of last generation aircraft. This concept also called Structural Health Monitoring (SHM) has as main objective to detect and locate damages occurring within the aircraft structure during its flight. To ensure a good efficiency of this SHM system, a sensor network covering all the monitored areas has to be deployed
which induces at the same time numerous difficulties due to connection and wiring aspects. Consequently wireless communications as well as energy autonomy capabilities are required in order to satisfy the future aeronautical certification steps for SHM systems. Nowadays, the assembly of composite elements which constitute more than fifty per cent of aircraft structure requires also a particular care because these elements are very sensitive to damages. For example, a tool falling might create fibre cracks or delamination, despite the fact that non visible damage can be seen. So, it is essential to detect the different impacts not only in flight operation but likewise during the assembly process, i.e. immediately when they occur. The existing AIRBUS SHM technology [1] allows the automatic detection and location of the impact events. However, the monitoring of large structures like wing boxes involves a sensor networking as far as impact Lamb waves [9] to detect are very attenuated. Again the energy autonomy of the sensor network becomes a very strategic topic if one wishes to reduce the wire clutter while at the same time improve the monitoring capabilities. Several energy harvesting solutions are available and many authors investigated their efficiency with growing interests in the last few years. Hence, Mathuna et al. [2] followed by Harb [3] proposed a synthesis of these different energy harvesting solutions whereas Park et al. [4] particularly focused on those which could be used in SHM systems. Among the different potential ambient power sources for SHM applications, thermal energy can be conveniently transduced into electrical energy as demonstrated by Samson et al. [5] which presented a system based on thermal conversion using an independent energy source combined with a wireless sensor node. As the power available from the conversion of mechanical energy is abundant enough to be of use, Roundy et al. [6] investigated also the potential of electromagnetic, electrostatic and piezoelectric vibration conversion. Since the use of piezoelectric cantilever beams for the mechanical energy conversion is very promising but limited to their natural frequency, Ferrari et al. in [8] proposed to work with several piezoelectric cantilever beams on which different masses have been bounded to convert energy over an extended frequency range. From their side, Lefeuvre et al. [7] more specifically focused on the optimization of the conditioning circuit using nonlinear techniques to increase the power harvested level. In this study, the authors prefer to work with piezoelectric plate transducers to convert the mechanical vibrations into electrical power, and because no vibrations are available on production line, they propose an original technical solution to generate them. This innovative idea has led to an international patent [13] which provides a method for wireless power transmission between one or more piezoelectric ceramic harvesters and an emitting device arranged to generate vibrations in the structure. This method is called “active power harvesting technique” in contrast to the usual methods operating with ambient energy. The
main advantage of this solution is that the structural health monitoring can now operate both in flight and on the ground. The global principle of this new solution is reminded in lengthy details in section 2. In particular, this section describes first all the elements allowing on the one hand the structural health monitoring of the aircraft and on the other hand the supply of the sensor network. Then information is given on the method allowing an optimum power transmission frequency in accordance with a predetermined requirement. Section 3 is dedicated to the experimental setup as well as the results obtained with this technique to detect impact like tools fall during the manufacturing. Finally a conclusion to this work is given. 2 Description of the principle and the operation of the autonomous SHM system. In order to detect destructive events on the composite structure during the manufacturing process, and despite no mechanical vibrations are available on the inspected elements, it becomes necessary to apply some mechanical energy to the structure for the power supply of the sensors. This is done using a SHM transducer which generates some vibrations inside the plate which works as a power transmission line. The vibrations may be Lamb waves and may comprise one or more non-dispersive Lamb modes which propagate all over the structure toward the harvesters located on several strategic places. Before demonstrating the feasibility of this innovative solution, it is first essential to remind its principle and the key points of its operation. To this end, Fig. 1 gives a general overview of the aircraft with the structural health monitoring system and the different sensors networks covering its area. More precisely, the aircraft is provided with a structural health monitoring (SHM) system 105, which utilises a set of sensors 106 to collect SHM data representing acoustic emissions, in the form of guided Lamb waves, from the structure of the aircraft. The SHM data is here used to monitor the integrity of the aircraft structure while the sensors are arranged to communicate with the SHM system wirelessly and each of the sensors is autonomously powered via an integrated power-harvesting device. Due to the lack of ambient vibrations on the structure during the manufacturing process, it becomes necessary to apply some mechanical energy to the structure if one wants to supply the sensor network destined to detect the destructive events. For this purpose a set of power transmission (WPT) devices 107 are provided at selected points on the structure of the aircraft, each WPT device being powered by the aircraft systems and arranged to generate high frequency vibrations, i.e. Lamb waves in the structure of the aircraft at a predetermined frequency so as to transmit power in the form of kinetic energy through the structure. The waves propagate all over the structure and enable the surrounding group of sensors to receive the
transmitted power via their respective integrated power harvesting devices. The harvested power is then used for the operation of the sensor for collecting SHM data and communicating it wirelessly to the SHM system. Fig. 2 describes with further details the operation principle of the SHM system. As shown on this figure, each of the sensors 106 attached to the aircraft structure consists of a first piezoelectric element 202 which is connected to a power harvester module (P) 106 comprising a model LTC 3588-1 piezoelectric energy harvesting power supply from Linear Technology Corporation. The power harvested from this piezoelectric element by the power harvester module (P) is stored and used for powering sensor logic (S) and a wireless communications to module (C) of the sensor. In the present embodiment, these piezoelectric elements are 2 mm thick, 20 mm diameter circular piezoceramic sensors of type Pz27 from Ferroperm Piezoceramics A/S. They are specially selected with modes of resonant vibration in both the thickness extension mode and the radius extension mode, in other words, the directions respectively perpendicular and parallel to the plane of the plate.
It may be also noted from Fig. 2, that the WPT device 107 includes a power supply 206 arranged to draw power from the aircraft systems, a power management module 208, a second piezoelectric element 209 with similar characteristics to the first piezoelectric element and a wireless communications module 210. This power supply is arranged to provide a selectable range of frequencies of signal to the second piezoelectric element under the control of the power management module. In this case, signals at 10 volts peak in the selectable frequency range 100 kHz to 600 kHz with a sinusoidal waveform are used to excite the second piezoelectric element. Lamb waves are then produced in response to either mode of resonant vibration of the piezoelectric element. After propagation through the plate, they excite the first sensors and enable power harvesting by the respective power harvester modules (P). Since they are less attenuated by the structure than other Lamb waves modes or other waves at other frequencies, non-dispersive Lamb waves modes, such as S0 or A0, are preferably selected for power transmission. It should be mentioned here that the power management module 208 of the WPT device comprises an optimisation means in the form of an optimisation module arranged to control the power transmission frequency of the signal operating the second piezoelectric element so as to optimise the transmission of power from the WPT to device to the power harvesting modules of the sensors. This optimization module is designed to power the second piezoelectric element for a selected range of power transmission test frequencies within its output frequency bandwidth and monitor the effect of each transmitted frequency on the voltage generated by the power harvester modules of each of the sensors. The optimization module is then
arranged to analyse the resulting harvested voltages or voltage responses and to select an optimum power transmission frequency in accordance with a predetermined formula. A major benefit is that this method may be performed in response to a predetermined set of changes of state associated with the structure and more particularly to the mechanical loads applied to the structure for example. Another innovative aspect of this contribution lies in the fact that the transmission of the energy to the sensors is not performed using wires but using the structure which works as a waveguide providing mechanical energy to each sensor working both as a harvester and an impact detector. 3 Experimental setup and results. Fig. 3 shows now the layout of the proposed experimental set-up. One can see that, as soon as the mechanical Lamb waves propagates all over the structure, many sensors could be localized anywhere on the structure and all of them receive a part of the mechanical power applied to the plate. Because Lamb waves propagate over long distances, the number of sensors nodes which has to be used for the full monitoring of a four square meter wing box for example is less than 6.
This is why the Carbon-epoxy composite plate (450mm*180mm*3.2mm) is instrumented with three SHM piezoelectric transducers with same dimensions than in section 2 as shown in Fig. 4. The transducer located in the middle of the plate is used as an actuator (emitting device), while the two others are used as energy harvester and impact detector.
Fig. 4: Piezoelectric transducers localization.
Using the methodology proposed in section 2 for the choice of the optimum frequency, a harmonic excitation is carried out around the radial frequency in the 100-150 kHz range. This frequency range is mainly fixed by the radial mode of the sensor which allows the generation of the antisymetric A0 Lamb mode in the plate [10].
Fig. 5. Voltage response of the harvesters.
Fig. 5 gives the voltage harvested by different transducers bounded onto the plate. Each transducer is used alternatively as an emitting and then a receiving device. The EiRj denomination means Emitter I Receiver j. Using emitter 1 (Fig. 5-a), the optimal frequency is given at 113.7 kHz for a maximum voltage of 5.7 Volts. However at this frequency, receiver 3 does not give a significant response. Consequently, it is not interesting to select this frequency. If we consider now the configuration for which transducer 3 is used as the emitter, (Fig. 5-b), one can decide to select the working frequency at a 121.7 kHz value which provides 3.5 Volts and 3.6 Volts for both receivers, respectively. As a result, this test shows that, for a given structure to monitor, it could be more interesting to choose a working frequency which provides for all the receivers the maximum power. During the manufacturing process and in the absence of damaging events, the Power Management circuit stores the electrical scavenged energy by the harvester [11, 12] in a capacitor at a 3.6 Volts voltage. As already explained in section 2, the conditioning circuit used to harvest the DC voltage is a LTC3588 power management circuit from Linear Technology. The AC voltage provided by the receiver is rectified using a bridge semi-conductor rectifier and the electrical charges are accumulated in a capacitor. Consequently, the voltage across the capacitor gradually increases. When its value reaches an under voltage lockout rising threshold, the buck converter is enabled and the output voltage is in regulation. In our case, the harvested power value is 7.36 milli-Watts. If now an impact occurs, a part (electrode N°2 on Fig. 6) of the SHM sensor receives the signal corresponding to the impact because the Lamb waves are filtered by the electrode N°2.
Fig. 6. SHM sensor/harvester configuration.
To test the system, the authors have decided to realize one impact using a small metallic mass of 20 gram falling from a 20 cm high. Fig. 7 shows the signals corresponding to the impact and the comparator switch output respectively. In fact, the impact signal serves as an input to a latch component which provides the impact detection signal. Then, this impact detection signal is used to light a LED which simulates the consumption of a WIFI emission.
Fig. 7. Impact signal.
Fig. 8-a is an image of the first prototype which has been used to exhibit the feasibility of the application. Then a smaller prototype has been developed using Surface Mount Components (SMC) and a low power consumption 8051 microcontroller (Fig. 8-b).
Fig. 8-a. Patch photography and 8-b: SMC prototype
4 Conclusions. This paper presents an industrial solution for the SHM of composite structure during their assembly on a wing box manufacturing process for aerospace applications. A wireless sensor network able to detect impact like tools fall during the manufacturing of the full aeronautic structure has been developed. This is an innovative and promising technique [13] because a single transducer powered by a classical emitting source allows secondary transducers to harvest energy and detect damaging impacts. This method is called “active power harvesting technique” in contrast to the usual methods operating with ambient energy. The main advantage of this solution is that the structural health monitoring can now operate both in flight and on the ground. Work is now in progress to reduce the size of the electronic board and to install such a prototype on real structures. List of figures captions Figure 1: General overview of the Aircraft with the embedded structural health monitoring system [13]. Figure 2: Principle of operation [13]. Figure3: Experimental set-up description. Figure4: Piezoelectric transducers localization.
Figure 5: Voltage response of the harvesters. Figure 6: SHM sensor/harvester configuration. Figure7: Impact signal. Figure 8: Patch photography and 8-b: SMC prototype.
References 1.
2.
3. 4.
5.
6.
7.
8.
9.
C.A Paget, K. Tiplady, M. Kluge, T. Becker, J. Schalk, “Feasibility study on wireless impact damage assessment system for thick aeronautical composites”, Proceedings of the 5th European Workshop on Structural Health Monitoring, 2010. C.O. Mathuna, T. O’Donnell, R.V. Martinez-Catala, J.Rohan and B. O’Flynn, “Energy scavenging for long-term deployable wireless sensor networks”. Elsevier, Talenta, vol.75, n°3, pp. 613-623, 2007. A. Harb, “Energy harvesting: State-of-the-art”, Elsevier, Renewable Energy, vol.36, n°10, pp. 2641-2654, 2011. G. Park, C.R. Farrar, M.D. Todd, W. Hodgkiss and T. Rosing, “Energy Harvesting for Structural Health Monitoring Sensor Networks”, Los Alamos Laboratory, LA-14314-MS, 2007. D. Samson, M. Kluge, Th. Becker and U. Schmid, “Energy harvesting for Autonomous Wireless Sensor Nodes in Aircraft”, Eurosensor XXIV Conference, Elsevier, Procedia Engineering, vol.5, pp. 1160-1163, 2010. S. Roundy, P.K. Wright and J. Rabaey, “A study of low level vibrations as a power source for wireless sensor nodes”, Elsevier, Computer Communications, vol.26, n°11, pp. 1131-1144, 2003. E. Lefeuvre, A. Badel, C. Richard, L. Petit and G. Guyomar, “A comparison between several vibration-powered piezoelectric generators for standalone systems”, Elsevier, Sensors and Actuators A: Physical, vol.126, n°2, pp. 405416, 2006. M. Ferrari, V. Ferrari, M. Guizzetti, D. Marioli and A. Taroni, “Piezoelectric multi-frequency energy converter for power harvesting in autonomous microsystems”, Elsevier, Sensors and Actuators A: Physical, vol.142, n°1, pp. 329-335, 2008. I.A. Viktorov, Rayleigh and Lamb waves: physical theory and applications. Plenum Press, 1970.
S. Grondel, C.A. Paget, C. Delebarre, J. Assaad and K. Levin. “Design of optimal configuration for generating A0 Lamb mode in a composite plate using piezoceramic transducers”. J. Acoust. Soc. Am, vol. 112, n°1, pp. 84-90, 2002. 11. C. Delebarre, T. Sainthuile, S. Grondel and C. Paget. “Power harvesting capabilities of SHM ultrasonic sensors”, Smart Materials Research, vol. 2012, 2012. 12. T. Sainthuile, S. Grondel, C. Delebarre, S. Godts and C. Paget, “Energy harvesting Process Modelling of an Aeronautical Structural Health Monitoring System Using a Bond-Graph Approach”, International journal of Aerospace Sciences, vol. 1 (5), pp. 107 -115, 2012. 13. Patent US 20130236038 A1, “Wireless Power Transmission”, Sept 2013. C. Paget (Airbus), T. Sainthuile, C. Delebarre, and S. Grondel.
10.
