NOVEL PIEZOELECTRIC ENERGY HARVESTING DEVICES FOR UNMANNED AERIAL VEHICLES

Transcription

1 NOVEL PIEZOELECTRIC ENERGY HARVESTING DEVICES FOR UNMANNED AERIAL VEHICLES Steven R. Anton Faculty Advisor: Daniel J. Inman Center for Intelligent Material Systems and Structures Virginia Polytechnic Institute and State University, Blacksburg, VA 61 Abstract The development of small aircraft including unmanned aerial vehicles (UAVs) and micro air vehicles (MAVs) has gained tremendous interest in the research community. One particular area of interest involves creating innovative techniques to increase the flight time or endurance of UAVs and MAVs. Energy harvesting technology presents a potential solution for the improvement of flight times by converting ambient energy into electrical energy that can be used to power the aircraft. Adding the necessary components to perform energy harvesting can, however, add a significant amount of mass to the aircraft, which often have small payload capacities. The additional mass can hinder the performance of the aircraft and, in fact, result in a decreased overall endurance. A novel concept is presented in this paper involving the combination of piezoelectric devices and new thin-film battery technology to form multifunctional self-charging, loadbearing energy harvesting devices for use in UAV systems. The proposed self-charging structures contain both power generation and energy storage capabilities in a multilayered, composite platform consisting of active piezoceramic layers for scavenging energy, thinfilm battery layers for storing scavenged energy, and a central metallic substrate layer. The compact nature of the devices allows easier integration into UAV systems and their flexibility provides the ability to carry load as structural members. A potential application of the selfcharging structures is use in the wing spars of UAVs or as the entire wing of an MAV. This paper addresses several aspects of the development and experimental evaluation of the proposed self-charging structures. Details of the design and fabrication of the selfcharging structures and appropriate energy harvesting circuitry are given. Results of experimental testing to evaluate the energy harvesting performance of the devices under harmonic loading are also presented. Keywords: self-charging, energy harvesting, multifunctional, unmanned aerial vehicles, thin-film batteries, piezoelectric Introduction The topic of energy harvesting has attracted much attention in the research community in the last decade. In particular, an emphasis has been placed on vibrationbased energy harvesting using piezoelectric materials, and a significant amount of research has been conducted on this topic 1,. One particular area where piezoelectric energy harvesting proves useful is in scavenging vibration energy during flight of unmanned aerial vehicles (UAVs). The development of UAVs has been of interest for military applications for several decades. Focus has recently been placed on small UAVs that can be carried and deployed by soldiers in the field and used for surveillance purposes. A limitation of these vehicles is their short endurance or flight time. The use of piezoelectric energy harvesting has been previously proposed by the author as a means of scavenging ambient energy and supplementing the fuel supply of a small aircraft with harvested energy 3. It has been found that continuous power on the order of at least 1 μw can be generated during flight of an RC glider aircraft from piezoelectric fiber-based devices mounted at the root of the wings. The author s previous study investigated the implementation of piezoelectric energy harvesters installed as add-on components to UAV systems. Typical energy harvesting devices designed as add-on components have the effect of mass loading the host structure. In a UAV application, this mass loading is undesirable as it can decrease the overall efficiency of the aircraft, and in some cases, the losses due to the additional mass can be greater than the gain due to harvesting. In an effort to better integrate energy harvesting systems into host structures, a multifunctional approach is investigated in this study in which the harvesters can be used to scavenge and store Anton 1

2 electrical energy, and support structural loads. In this fashion, an energy harvesting device can be used to replace a structural member (or portion thereof) of the host system, thus reducing the total mass addition of the harvester. When considering the mass added to a host structure, all of the components of an energy harvesting system must be considered, including the active element, harvesting circuitry, and storage devices. Recent developments in the field of lithium-based batteries have given rise to flexible, light weight, thin-film batteries that are ideal for use in harvesting systems. The performance of these batteries under static flexural deflection and uniaxial pressure has been investigated in the research, 5. Additionally, the ability to embed thin-film lithium batteries into structural composites has been explored 6. It is proposed in this research to directly embed these thin-film batteries into a load bearing energy harvesting package to create selfcharging, structural devices with true multifunctionality. This paper focuses on the development and experimental evaluation of self-charging structures. The goal is to evaluate the feasibility and practicality of the self-charging structure concept. An evaluation is carried out on the thin-film batteries used in this research, and their electrical and mechanical performance is explored. Details on the fabrication of self-charging devices are presented including the various challenges involved. A brief discussion of appropriate energy harvesting circuitry is given. Lastly, the experimental performance of the self-charging structures in harvesting and storing energy is described. Self-Charging Structure Concept Typical energy harvesting systems are designed as addon components that can add a significant amount of mass to the host structure. Considering the importance of reducing the total mass addition of energy harvesting systems in UAV structures, a multifunctional, multilayered composite piezoelectric energy harvester is developed that is capable of generating electrical energy, storing that energy, and supporting mechanical loads. The proposed self-charging composite harvester, shown in Figure 1, can be integrated into the design of a structure such that it may replace an existing component, thus reducing the added mass to the host structure. It should be noted that the orientation of the piezoelectric and thin-film battery layers is arbitrary and the optimal configuration will depend on the application. The outermost layer will experience the Figure 1: Schematic of self-charging structure. largest amount of strain, thus the piezoelectric layers are shown as the outer layers in Figure 1 for maximum energy generation. The piezoelectric layers, however, may be the critical layer in terms of failure; therefore, the battery layers may be better suited as the outer layers. A critical element of energy harvesting systems is a storage device in which scavenged electrical energy can be placed temporarily before use. Most piezoelectric energy harvesters do not produce adequate energy levels for immediate use, thus the energy must be accumulated for some time prior to use. Typical storage elements include capacitors, supercapacitors, and batteries. A storage element must be considered as part of the harvesting system, and as such, they contribute to the total mass and volume consumed by the system. Often times, the storage device can be on the same order of size and mass as the active harvesting element, or perhaps larger. Selection of an appropriate storage device, therefore, becomes an important element in the design of any energy harvesting system. Recent developments in the area of lithium-based batteries yield the discovery of flexible thin-film lithium batteries with thicknesses on the order of millimeters and capacities in the milliamp-hour range for a 5. mm x 5. mm device. These thin-film batteries present a breakthrough in storage element technology as they are much smaller and lighter than conventional storage devices, and they are flexible. Additionally, the small capacities of these batteries are a good match for the low electrical output levels associated with piezoelectric energy harvesting. In order to best utilize these thin-film lithium battery cells, it is proposed to embed them directly into the harvesting device to create a compact, multifunctional device capable of both generating and storing electrical energy. As the device is excited mechanically, electric charge accumulates on the electrodes of the Anton

3 piezoelectric device due to the direct piezoelectric effect, and that energy is routed to the battery layers through conditioning circuitry for storage. Due to the flexibility of the thin-film batteries, they can withstand the mechanical loading and deformations imposed on the active piezoelectric harvesting element. Additionally, their flexibility is a necessary characteristic in order to be directly embedded into the harvester. Conventional storage elements are not appropriate for direct application to piezoelectric devices as their stiffness would absorb much of the input vibration energy needed to induce strain into the piezoelectric for energy generation. Mechanical failure is also a concern with conventional storage devices as they are not designed to support structural loads. A novel aspect of the self-charging structure concept is that the composite harvester can be used as a loadbearing member in a host structure. The harvester can be used in place of an existing component, thus reducing the total mass added to the host structure. Such a composite harvesting device, for example, can be considered as the wing (or a portion of the wing) of a small UAV or MAV (micro air vehicle) as shown in Figure. When integrating self-charging structures into a UAV wing spar as suggested, several design parameters should be analyzed including the structural stiffness of the spar, the allowable stress at most critical point of the spar, and the mass addition due to the device. Coupled with an appropriate piezoelectric device, the mechanical properties of the self-charging structure can be tailored to the meet the needs of a specific application. Piezoelectric devices range from stiff, brittle monolithic piezoceramic to flexible piezoelectric fiber-based transducers, thus a wide range of mechanical properties can be obtained. harvesting element to create a composite device capable of generating and storing electrical energy, and integrating an energy harvesting system into a host structure as a load bearing member both present novel concepts in piezoelectric energy harvesting. Selfcharging structures introduce the concept of multifunctionality in energy harvesting, which gives promise for more efficient and useful devices. Performance Evaluation of Thin-Film Batteries NanoEnergy thin film lithium rechargeable batteries manufactured by Front Edge Technology, Inc. (Baldwin Park, CA) are investigated in this research (Figure 3 (a)). The all solid-state energy cells utilize a ceramic electrolyte composed of lithium phosphorous oxynitride (LiPON), developed at the Oak Ridge National 7, 8 Laboratory, which eliminates any liquid components. The batteries consist of a lithium anode, LiPON electrolyte, and lithium cobalt dioxide cathode built on a top and bottom mica substrate and sealed between the substrate layers with a Surlyn (DuPont) sealant layer (Figure 3 (b)). Typical battery dimensions are 8.96 mm x 5. mm with a thickness of 15 μm and a mass around. grams. The nominal voltage of the cells is. V and their capacities are around mah. The manufacturer publishes that the batteries have more than 1 charge/discharge cycles at 1% depth of discharge, can be charged to 7% of rated (a) Directly bonding a storage device to the active (b) Mica Substrate Surlyn Sealant Anode - Lithium Electrolyte - LiPON Cathode - LiCoO Figure : Small UAV with self-charging structures embedded in wing spar. Figure 3: NanoEnergy thin-film batteries (a) Photograph; (b) Cross-sectional detail. Anton 3

4 capacity in just minutes, and can be discharged at rates up to 1C (1 times the rated capacity). The NanoEnergy thin-film batteries have a typical internal resistance on the order of 5 - Ω. This is extremely high compared to the internal resistances of most conventional alkaline, nickel-metal hydride, and lithium-ion batteries which are on the order of.1 Ω 1 Ω. When current flows through a battery, there is a voltage drop across the internal resistance of the battery which decreases the terminal voltage as well as the efficiency of charging and discharging. A large internal resistance is detrimental to battery performance because it causes a large voltage drop for large current loads. Due to the large internal resistance of the thin-film batteries, it is difficult to source high current levels and maintain a high voltage. In order to draw 1 C out of a mah thin-film battery with an internal resistance of Ω, for example, the terminal voltage drop would be V = I R =.A Ω=.8V (1) drop int therefore, the battery can only supply 1 C of current at voltages less than 3. V. Additionally, the manufacturer suggests that the batteries not be discharged below 3 V to prevent damage to the cells, therefore, a current of ma can only be sourced from the battery for a short period of time. In order to fully discharge the thin-film batteries, a large load resistance must be used to source a small amount of current over a large period of time, otherwise the terminal voltage will drop to 3. V before all of the energy is released from the battery, causing only a partial discharge. Prior to combining the thin film batteries with the piezoelectric devices to create a self-charging structure, the performance of the batteries is evaluated experimentally. The batteries are charged using an HP 685A power supply/amplifier and discharged through standard carbon film resistors. During charging and discharging, the battery voltage and the current flowing in/out of the battery are monitored and recorded using a National Instruments data acquisition system. The current measurements can be used to quantify the amount of energy flowing through the battery. Batteries capacities are rated in milliamp-hours (mah), which describes charge over time. The capacity achieved during charging and discharging can be calculated by performing numerical integration of the current measurement over time as follows: C = i dt () Typical voltage and current measurements during charging and discharging of the NanoEnergy batteries are shown in Figure. Charging of the batteries is performed using a constant-voltage method of supplying. V of potential to the battery and charging until the current input to the battery drops to 1. ma (Figure (a)). Discharging is accomplished by connecting a 981 Ω resistor to the battery until a terminal voltage of 3. V is obtained. For this particular battery, an initial voltage drop of.9 V (from.15 V to V) and a current of 3.9 ma are observed, as seen in Figure (b). The corresponding internal resistance of this battery can be calculated as: V = I R R drop int int Vdrop.9V = = = 7.36Ω I.39A (3) Based on the results shown in Figure, it can be seen that the initial surge current in charging is around.75 ma and then the battery maintains an input current around ma but begins dropping down to 1 ma after about 1 seconds. The voltage is held at. V throughout charging, however, the voltage drops to about.1 V after the power supply is disconnected. In discharging, a current of around ma is sustained for about seconds upon which the current begins to drop quickly. The battery voltage behaves similarly, maintaining a voltage of around 3.6 V for seconds upon which the potential drops quickly to 3. V before the load is disconnected. The battery voltage recovers to around 3.8 V after the load is removed. Carrying out the capacity calculation given in Equation (), the capacity in charging is calculated as 3.1 mah, and the capacity in discharging is.79 mah. It is expected that these capacities be reasonably close to one another, as is the case. Some energy is lost during charging, likely due to leakage effects, and this is reflected in the fact that the capacity in discharging is slightly less than the capacity in charging. It is worth noting that through handling and testing of the NanoEnergy thin-film batteries, it has been observed that the batteries are extremely fragile. Although they are flexible and can withstand being slightly bent and twisted without damage, the packaging of the cells is highly susceptible to damage. The mica substrate, for example, can peel away from the sealant layer when small shear loads are applied to Anton

5 8 6 Baseline Charge Profile of NanoEnergy Thin Film Batteries. found to be damaged after testing, causing the seal to fail allowing oxygen to interact with the lithium-metal and cause oxidation. Fabrication of Self-Charging Structures Current (ma) 3 1 Voltage (V) Assembly of the self-charging structures involves several steps including the selection of piezoelectric materials and substrate materials, bonding the battery, piezoelectric, and substrate layers, and connecting leads to the electrodes of the thin-film batteries. These steps are outlined in the following sections. Current (ma) Time (sec) Baseline Discharge Profile of NanoEnergy Thin Film Batteries Time (sec) Figure : Characteristic curves of NanoEnergy thin-film batteries in (a) charging and (b) discharging. the edges of the device. Additionally, the electrodes consist of 1 μm thick metal foils which can easily peel off of the substrate and become damaged. It is known in the research community that thin-film battery technology still faces several challenges 9. Previous research on NanoEnergy thin-film batteries has been conducted by Pereira et al., 5 in which their performance under flexural deflection and uniaxial pressure is investigated, and similar observations have been reported with respect to battery fragility. In static bending, it is found that both mechanical and electrical failure occur for flex ratios (defined as the deflection divided by the span length) greater than 1.3%, which is quite low 5. When subjected to uniaxial pressures, however, the batteries are found to fail at pressures above. MPa, which is reasonable. Under both types of loading, the Surlyn sealant and mica substrate were. 3 1 Voltage (V) Selection of piezoelectric and substrate materials A commercially available piezoelectric material will be used as the active energy harvesting element in this research. Several companies produce piezoelectric materials that can be considered for the self-charging devices. Piezoelectric materials can be categorized into three classes including monolithic piezoceramic, piezoelectric fiber-based, and piezoelectric polymer film devices. Piezoelectric polymer films such as polyvinylidene fluoride (PVDF) are extremely compliant with an elastic modulus around 3 GPa, and their power output is typically orders of magnitude less than monolithic and fiber based piezoelectric materials, therefore, they will not be considered for use in selfcharging structures. Piezoelectric fiber-based devices such as the Macro Fiber Composite (MFC) (Smart Material Corp., Sarasota, FL) or the Piezoelectric Fiber Composite (PFC) (Advanced Cerametrics, Inc., Lambertville, NJ) consist of monolithic piezoelectric fibers embedded in a polymer matrix with interdigitated electrodes. The structure of such devices allows flexibility and high energy generation capabilities. Monolithic piezoelectric ceramic materials exhibit high energy generation abilities compared with fiber and polymer based devices, however, they are brittle in nature and susceptible to failure under loading. As a compromise between high energy generation and strength under dynamic loading, QuickPack piezoelectric devices (Midé Technology Corp., Medford, MA) are selected for use in self-charging structures. QuickPack devices contain monolithic piezoceramic (PZT-5A) active elements bracketed in Kapton to protect the piezoelectric and give some robustness. The particular devices used are QP1N actuators which have overall dimensions of 5.8 mm x 5. mm x.381 mm, an active piezoelectric element of 5.97mm x.57 mm x.5 mm, and a mass of.67 g. The substrate layer used for the self-charging structure is 11-O aluminum alloy with dimensions of 63.5 mm x 5. mm x.17 mm and a mass of.555 g. Anton 5

6 The substrate is chosen to be slightly longer than the piezoelectric elements in order to provide a surface for attachment of electrical leads. Bonding the substrate, piezoelectric, and battery layers An appropriate bonding agent must be selected to bond the various layers of the self-charging structures. Several epoxies and super glues are evaluated, and 3M ScotchWeld DP6 -part epoxy is selected due to its high shear strength ( psi when bonded to Aluminum) and high volume resistivity (. x 1 1 ohm-cm). This particular epoxy is used to bond all of the layers of the structure. Each piezoelectric layer is first bonded to the substrate independently and then the thin-film battery layers are bonded to the piezoelectric layers. The thin-film batteries are selected as the outer most layers to facilitate attachment of their electrical leads. Bonding of the piezoelectric layers is achieved by applying a thin layer of DP6 epoxy onto one side of the QP1N device and then placing the device onto the substrate. Curing of the epoxy is achieved under vacuum using a vacuum pump and a vacuum bag setup (Figure 5). The layers are placed on a thin metal plate and held down with tape before being placed in the vacuum bag so that there is no slippage between layers when curing. The device is allowed to cure for 6 hours before being removed from vacuum. This process is repeated for both piezoelectric layers. Bonding of the NanoEnergy thin-film batteries is performed after both piezoelectric layers have been attached to the substrate using the same procedure as with the piezoelectric layers. The battery layers are placed towards the free end of the device to reduce the induced strain and help prevent electrical or mechanical failure of the batteries. DP6 epoxy is used between the battery and piezoelectric layers and is allowed to cure under vacuum for 6 hours. After each layer is bonded, the thickness of the structure is measured to determine the thickness of each epoxy layer. Vacuum curing is used to achieve thin, uniform bonding between each layer. The average epoxy layer thickness is around. mm, yielding good results. The overall thickness of the 13-layer structure is mm. Attaching leads to the piezoelectric and thin-film battery layers Electrical leads must be attached to both the QuickPacks and the NanoEnergy cells for testing. The QuickPack devices contain an electrical connector as shipped from the manufacturer, however, the connector increases the overall length of the device and has a mass of about.557 g, therefore, it is desirable to remove the connector. The connectors are removed, leaving flat electrodes covered in Kapton. In order to attach leads, a small section of the Kapton coating is removed using a razor blade to expose the electrodes. -gauge insulated and stranded wire is then soldered to the exposed electrodes to create an electrical connection. Figure 5: (a) Vacuum bag setup for curing epoxy layers; (b) Close-up view of selfcharging structure under vacuum. As discussed earlier, the NanoEnergy batteries have been found to be extremely fragile. Attachment of electrical leads requires a careful procedure. The battery electrodes consist of 1 μm thick metal foil which cannot be soldered to because the heat can damage the battery. Additionally, the electrodes peel Anton 6

7 off from the substrate quite easily, therefore, a conducting tape (such as copper tape) cannot be used because it can peel the electrodes off. Electrical connection between leads and the battery electrodes is accomplished using Model 181 Silver Paint (Ernest F. Fullam, Inc., Clifton Park, NY). -gauge insulated and stranded wire is used and the silver paint is applied to the wire and electrode area and placed in an oven at 65 C for 9 minutes to cure. Once the paint has cured, Loctite 3381 UV curable epoxy is applied over the electrode connection area to provide electrical isolation and mechanical support of the connection. The epoxy is cured under LED UV light for about 3 minutes. Once the leads are connected to both the piezoelectric and battery layers, the device is complete. A photograph of the completed self-charging device along with detail of the battery electrode connection is shown in Figure 6. Energy Harvesting Circuitry The voltage output of a vibrating piezoelectric device is an alternating (AC) signal. Typical electronic and storage devices in which piezoelectric energy is likely to be used require DC input signals, therefore, the output of a piezoelectric element requires conditioning before use, which can be accomplished using some sort of energy harvesting circuitry. There is a strong emphasis in the energy harvesting community on the development of highly efficient energy harvesting circuitry. Researchers have investigated various methods of increasing the percentage of energy generated by the piezoelectric element that is transmitted to the load. In particular, the research group at INSA-Lyon in France has investigated methods of synchronizing the extraction of electrical energy with the excitation frequency through switching circuits 1, and has also investigated buck-boost converters for piezoelectric energy harvesting 11. The aim of this study is to simply prove the concept of self-charging structures, therefore, a simple, nonoptimized energy harvesting circuit is employed. The circuit contains a full wave diode rectifier, a smoothing capacitor, and a Texas Instruments TPS7151 adjustable output voltage regulator. A schematic of the circuit is shown in Figure 7 (a) and representative voltage waveforms at each stage of the circuit are shown in Figure 7 (b). The rectifier transforms the AC signal into a completely positive signal, the smoothing capacitor turns the signal into quasi-dc, and the regulator drops the voltage down to a steady level acceptable by the thin film batteries. Experimental Performance of Self-Charging Structures Figure 6: (a) Complete self-charging device; (b) Electrode connection using silver paint and UV curable epoxy. The performance of the fabricated self-charging structure is evaluated experimentally by mounting the device in a cantilever fashion and subjecting it to base excitations. The self-charging structure is clamped to a small LDS electrodynamic shaker with an overhang length of 3.5 mm (Figure 8 (b)). Initially, frequency response measurements are recorded using a SigLab data acquisition system configured to record base acceleration, tip displacement, and voltage output of the energy harvester. The base acceleration is measured using a PCB U35C67 accelerometer and the tip displacement is measured using a Polytec PDV-1 Laser Doppler Vibrometer. The maximum amount of Anton 7

8 C1 33μ F C. μ F C 3.7 μ F v o R1.8 M Ω R 1. M Ω Figure 7: (a) Energy harvesting circuit; (b) Voltage waveforms at various stages in the circuit assuming a 5 volt PZT input and a regulator limit of. volts. power generated by the self-charging structure is obtained when exciting at the first bending mode where the strain induced in the piezoelectric layers is largest. The frequency response data is used to determine the resonant frequency at the first bending mode of the device such that subsequent energy harvesting experiments can be performed at resonance. The experimental setup is shown in Figure 8 (a). Experimental data is shown in Figure 9 for the voltage output to base acceleration frequency response functions (FRFs). Data is recorded both with the selfcharging structure connected electrically to the harvesting circuit and the thin-film battery, as well as for the open circuit condition with the circuit disconnected. In both cases, the two piezoelectric layers are connected in parallel to improve the current output. It can be seen from Figure 9 that the resonant frequency of the self-charging structure is 57.5 Hz. Additionally, the frequency response measurement with the circuit and battery connected tends to be similar to the open circuit condition. Figure 8: Overall experimental setup for FRF and energy harvesting experimentation; (b) Self-charging structure mounted on electromagnetic shaker. With the resonant frequency at the first bending mode of the self-charging structure determined, the energy harvesting performance of the device can be experimentally evaluated. Using an experimental setup similar to that shown in Figure 8, the self-charging structure is excited at 57.5 Hz and the output of the piezoelectric layers is fed into the harvesting circuit and used to charge the thin-film battery layers. For this experimentation, the two piezoelectric layers are connected in parallel for increased current output and used to charge a single battery layer. The input base acceleration amplitude is set to ±1.3 g, which was calculated to be the critical acceleration for the maximum allowable stress in the piezoelectric layers at the root of the beam considering a safety factor of. The device is excited for 8 hours and the battery voltage and current into the battery are measured throughout the Anton 8

9 1 FRF - Self-Charging Structure 1 1 Battery Connected Open Circuit.75 Charge Profile of Self-Charging Structures. Voltage FRF (V/g) Current (ma) Voltage (V) Frequency (Hz) Figure 9: Experimental Frequency Response Functions of self-charging structure. test using the National Instruments data acquisition hardware. Once the test is complete, the battery is discharged using a 371 Ω resistor. Results from both the charging and discharging tests are shown in Figure 1. From Figure 1 (a), it can be seen that the piezoelectric layers are able to supply an average of about.165 ma of current into the battery. The battery voltage slowly increases from about 3.7 V to 3.9 V. Using Equation (), the capacity during charging is found to be 1.38 mah. During discharging, the voltage on the battery is maintained around 3.7 V and the current output is held at 1.15 ma for 5 seconds. A capacity of.78 mah is found by integrating the current over time. There is a significant difference between the capacity calculated in charging and that calculated in discharging. It is likely that this is a leakage effect where some of the energy during charging is dissipated in the battery, thus there is a decrease in capacity when discharging. The charge/discharge results presented in Figure 1 prove the ability of the self-charging structures to both generate and store electrical energy in a multifunctional manner. The current of.165 ma during charging is a reasonable number for piezoelectric energy harvesting. The average power into the battery during charging is around.65mw, which is proven to be enough energy to support a load of 3.7 V and 1.15 ma for 5 seconds. Summary and Conclusions Time (sec) Current (ma) 3 1 Disharge Profile of Self-Charging Structures Time (sec) x 1 Figure 1: Experimental curves for selfcharging structure in (a) charging and (b) discharging. This paper focuses on the fabrication and experimental validation of self-charging structures in an effort to evaluate the feasibility and practicality of the selfcharging structure concept. Thin-film lithium-based batteries are investigated and found to have high energy densities and to couple well with piezoelectric energy harvesting, however, the batteries are observed to be quite fragile. Fabrication of the self-charging structures is proven successful by using epoxy cured under vacuum as the bonding layer material. Silver paint cured in an oven then coated in an epoxy sealant is found to be an effective electrode connection for the frail thin film battery terminals, where conventional solder is used on the piezoelectric layers. Experimental testing of the frequency response behavior has shown the natural frequency of the first bending mode of the self-charging structure to be 57.5 Hz. When excited at the first bending mode with ± 1.3 g base acceleration input, the device is able to generate an average of.65 mw of continuous power, supplying 1.38 mah of capacity to the thin film battery layer. Upon. 3 1 Voltage (V) Anton 9

10 discharging,.78 mah of capacity is drawn from the battery. This difference is attributed to leakage losses in the battery during charging. Overall, the results of this research have proven the concept of self-charging structures offering multifunctinality in piezoelectric energy harvesting. It has been shown feasible to integrate novel thin-film batteries along with piezoelectric devices to form load-bearing, selfcharging structures that can be embedded into the wings of UAVs and MAVs. Future work will consider performing detailed strength analysis and testing of the devices. Dynamic strength testing will be performed for both the thin-film batteries and the complete selfcharging structure to determine their robustness with regards to UAV applications. Experimental charge/discharge behavior will be analyzed at various input acceleration levels to determine performance under different excitation conditions. Additionally, efficient DC-DC converters will be investigated to maximize the efficiency of the energy harvesting circuit used. Acknowledgements The author would like to acknowledge the support of Na Kong from the Virginia Tech VLSI for Telecommunications (VTVT) laboratory at Virginia Tech. The author also gratefully acknowledges the support of the Air Force Office of Scientific Research MURI under Grant No. F Energy Harvesting and Storage Systems for Future Air Force Vehicles monitored by Dr. B. L. Lee. References [1] Anton, S. R. and Sodano, H. A., "A review of power harvesting using piezoelectric materials (3-6)," Smart Materials and Structures, 16(3) R1-R1 (7). [] Cook-Chennault, K. A., Thambi, N., and Sastry, A. M., "Powering MEMS portable devices-a review of non-regenerative and regenerative power supply systems with special emphasis on piezoelectric energy harvesting systems," Smart Materials and Structures, 17() 31 (33 pp.) (8). [3] Anton, S. R. and Inman, D. J., "Vibration energy harvesting for unmanned aerial vehicles," in SPIE's 15th Annual International Symposium on Smart Structures and Materials & Nondestructive Evaluation and Health Monitoring. Bellingham WA, USA, 698, 698, (8). [] Pereira, T., Scaffaro, R., Guo, Z., Nieh, S., Arias, J., and Hahn, H. T., "Performance of thin-film lithium energy cells under uniaxial pressure," Advanced Engineering Materials, 1() (8). [5] Pereira, T., Scaffaro, R., Nieh, S., Arias, J., Guo, Z., and Thomas Hahn, H., "The performance of thin-film Li-ion batteries under flexural deflection," Journal of Micromechanics and Microengineering, 16(1) (6). [6] Pereira, T., Guo, Z., Nieh, S., Arias, J., and Hahn, H. T., "Embedding thin-film lithium energy cells in structural composites," Composites Science and Technology, 68(7-8) (8). [7] Bates, J. B., Dudney, N. J., Gruzalski, G. R., Zuhr, R. A., Choudhury, A., Luck, C. F., and Robertson, J. D., "Electrical properties of amorphous lithium electrolyte thin films," Solid State Ionics, 53-56(Part 1) (199). [8] Yu, X., Bates, J. B., Jellison, G. E., Jr., and Hart, F. X., "Stable thin-film lithium electrolyte: lithium phosphorus oxynitride," Journal of the Electrochemical Society, 1() 5-53 (1997). [9] Patil, A., Patil, V., Wook Shin, D., Choi, J.-W., Paik, D.-S., and Yoon, S.-J., "Issue and challenges facing rechargeable thin film lithium batteries," Materials Research Bulletin, 3(8-9) (8). [1] Lallart, M., Garbuio, L., Petit, L., Richard, C., and Guyomar, D., "Double synchronized switch harvesting (DSSH): A new energy harvesting scheme for efficient energy extraction," IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 55(1) (8). [11] Lefeuvre, E., Audigier, D., Richard, C., and Guyomar, D., "Buck-boost converter for sensorless power optimization of piezoelectric energy harvester," IEEE Transactions on Power Electronics, (5) 18-5 (7). Anton 1