International Workshop SMART MATERIALS AND STRUCTURES

Transcription

1 Cansmart 2009 International Workshop SMART MATERIALS AND STRUCTURES October 2009, Montreal, Quebec, Canada POWER FOR WIRELESS SENSORS Nezih Mrad Defence R&D Canada (DRDC), Department of National Defence (DND) National Defence Headquarters, Ottawa, ON, K1A 0K2, Canada Tel: (613) , Fax: (613) Missy Alphonse Fabien and Farid Mabrouki Institute for Aerospace Research (IAR), National Research Council (NRC) 1200 Montreal Road, Bldg. M-14, Ottawa, ON K1A 0R6, Canada ABSTRACT The recent economic downturn, the high cost of energy and its environmental impact coupled with continued drive to operate efficient military and commercial infrastructure have intensified the development of several technologies including flexible electronics, wireless sensors, wireless sensor networks, and advanced power generation. Additionally, with the green energy revolution, and among several green energy initiatives, energy harvesting technology development has taken front stage particularly in the area of environmental energy harvesting. Energy harvesting is the process by which energy is captured and stored. A variety of different methods exist for harvesting energy, such as solar power, piezoelectricity, thermoelectricity, and physical motion. If pervasive networks of wireless sensors and communication devices are to achieve their full potential, practical solutions for self-powering these autonomous electronic devices must be sought. This document reports on an experimental parametric analysis of piezoelectric vibration-based energy scavenging technique to potentially power wireless sensors and sensor networks. The potential of this energy harvesting approach is demonstrated and experimental results are presented. Keywords: Wireless sensors, Sensor networks, Energy harvesting, Energy scavenging, Structural health monitoring. 177

2 INTRODUCTION In recent years, due to weight, cost, redundancy and remote access benefits, Wireless Sensors (WS) and Wireless Sensor Networks (WSN) have gained significant interest from both military and commercial sectors. A wireless sensor, also known as a mote, smart dust, smart sensor or sensor node, consists of a sensor and instrument packages that are microprocessor driven and include advanced features such as communication and limited data storage and processing. A wireless sensor network usually consists of a large number of small scale nodes and has limited data processing and storage capacity, wireless data communication and transmission ability, and advanced sensing capabilities. The WSN is also known as a network of Radio Frequency (RF) transceivers, sensors, machine controllers, microcontrollers and user interface devices with at least two nodes communicating by means of wireless transmission. Figures 1 and 2 present a simplified block diagram of a wireless sensor node and a conceptual configuration of wireless sensor network for aircraft Structural Health Monitoring (SHM). Applications that could benefit from such wireless technology include aerospace and civil structural health monitoring, environmental and industrial systems control, components and assets tracking and monitoring. It has been reported recently [1] that the wireless sensors and transmitters market is fast growing, with 200 % growth worldwide in the last 5 years and 45% projected growth in the next 3 years, seeing a market forecast reaching 1,800 million US dollars by 2012 with major market growth in North America (up to 600 million dollars), Europe (500 million dollars) and Asia Pacific (400 million dollars). With the continuous challenges facing the Canadian Forces (CF) in providing increased operational availability and reduced maintenance cost of it aircraft fleet, integration of WS and WSN is expected to reduce the on-going fiscal and operational pressures and provides the CF with on-demand decision making capabilities for its fleet s life cycle management. In fact it is expected that wireless technology will contribute significantly to addressing the Canadian auditor general recommendations on improving operational performance and asset tracking. While providing added flexibility for data, information, knowledge accumulation and decision making, power requirement and generation continue to be in the mind of the end-user as they consider integration of such wireless technology. A variety of different methods exist for power generation [2] exploiting the environment while protecting it. This work assesses the feasibility and suitability of one approach based on the use of piezoelectric material and ambient vibration for the generation (harvesting or scavenging) of energy to potentially power wireless sensors and sensor networks requiring power ranging from few μw to few mw (e.g. Ultra-low power consumption of 60µW to 300µW depending on the channel number). The potential of this energy harvesting approach is investigated and experimentally analysed. POWER SOURCES The conventional approach to power sources is the use of electrochemical batteries. These batteries can not only increase the size and weight of autonomous wireless sensors, but also suffer from the limitations of a brief service life and the need for costly replacement. Additionally, it has been experienced that at extreme temperatures (high), the use of these 178

3 storage devices is unsafe. The conventional approach is not highly practical or cost effective for many applications including those faced by the CF. Table 1[3] illustrates the characteristics of different commercial battery types; whereas; Figure 3 [4] shows the life cycle of rechargeable and non-rechargeable batteries. Sensor/ Interface Microcontroller Data Acquisition Data Processing Radio Transceiver Transmit Data Receive Data Supply Voltage Fig. 1: Simplified block diagram of a wireless sensor node Battery Type Fig. 2: A conceptual configuration of wireless sensor network for SHM Table 1: Characteristics of different commercial battery type Volumetric Gravitational Self-Discharge Energy Density Energy Density % per year (Wh/dm 3 ) (Wh/kg) Cycle Life No. Alkaline Ni-Cd Ni-MH Li-ion Days of Use at 100 μw Non-Rechargeable Lithium-Ion NMH Fig. 3: Life cycle of one cubic centimetres rechargeable and non-rechargeable batteries 179

4 Even though recent efforts have focused on the development of super and ultra capacitors, fuel cells, and other fixed energy alternatives with impressive large capacitance and higher number of cycles, these remain impractical for wireless devices with an expected lifetime of more than 10 years. For wireless sensors and communication nodes to achieve their full potential, practical options of power generation for these autonomous electronic devices must be investigated. A widely investigated attractive alternative is to use devices that generate power by scavenging ambient environment energy. Energy harvesting or scavenging is the process by which energy is captured and stored. Frequently this term is applied when speaking about small autonomous devices, like those used in wireless sensor networks. A variety of different methods and approaches exist for harvesting energy, such as solar power, ocean tides, piezoelectricity, thermoelectricity, and physical motion. Figure 4 [4] and Table 2 [5] show the output from a variety of energy harvesting schemes μ W/cm Solar (Outside) Solar (Inside) Vibration (@ 250 Hz) Vibration (@ 10 Hz) Human Shoe Human Breath (Theory) Thermocouple Fig. 4: Power output using different energy harvesting schemes Table 2: Estimated power and challenges for different energy sources Energy Source Challenge Estimated Power (in 1 cm 3 or 1 cm 2 ) Light Conform to small surface area 10µW-15mW Outdoors: 0.15mW-15mW; Indoors: <10µW Vibrations Variability of vibration 1µW-200µW Piezoelectric: ~ 200µW; Electrostatic: 50µW-100µW; Electromagnetic: <1µW Thermal Small thermal gradients 15µW; 10 C gradient ENERGY HARVESTING With the green energy revolution, energy harvesting technology development has intensified, in recent years [6]. The global market for energy harvesting devices for small electronic and electrical equipment is expected to reach 10,000 million devices by 2019 from the current (2009) half a million [7]. Figures 5 [7], illustrates the efficiency-power density and cost-life options for currently popular forms of energy harvesting approaches. Vibrationbased energy harvesting techniques has emerged as one of the promising approaches to address the needs for low power wireless sensors and sensor networks. Figure 6 illustrates the wide range of power output (10 μw to 1 kw) for vibration and heat based devices at 180

5 different frequencies. Piezoelectric vibration-based energy converters are shown to deliver the highest efficiency at lowest cost and increased life cycle. Power Density Watts/ cm Thermoelectric Electrodynamic Photovoltic Piezo Efficiency % years Electro dynamic Life 1 year High Thermoelectric Cost per watt Piezo Fig. 5: Cost-life options for currently popular forms of energy harvesting approaches Photovoltic Low Fig. 6: Power output for vibration-based devices ENERGY HARVESTING SYSTEM The Energy Harvesting (EH) system evaluated in this study is an advanced ceramitrics (ACI) energy harvesting system. The portable system consists of four main components as shown in Figure 7. The Piezoelectric Fiber Composite (PFC) beam, with wired electrodes and snap connector terminals is used to generate/harvest energy when exposed to vibration. The mechanical amplification device is employed to increase the output power of the PFCB composite through increased beam deflection. The energy harvesting circuit card (DB-2) with regulated (EH-R) or unregulated (EH-UR) power output is used to store the harvested energy. The ruggedized enclosure is used to house the EH system components. 181

6 This piezoelectric energy harvester employs a Pieozoelectric Fiber Composite (PFC) cantilevered beam that is developed from a technology called Viscose Suspension Spinning Process (VSSP) [8]. Conventional piezoelectric ceramic materials are rigid, heavy, and produced in block form. This low-cost technology process can produce fibers ranging in diameter from 10 microns to 250 microns. When formed into user defined shape, the ceramic fibers possess all the desirable properties of ceramics (electrical, thermal, chemical) but eliminate its detrimental characteristics (brittleness, weight). It has been reported [8] that the VSSP generates fibers with 20-30% more efficient energy conversion than traditional bulk ceramics. To put this into perspective, mechanical to electrical transduction efficiency can reach 70% compared with the 16-18% common to solar energy harvesting. These fine ceramic fiber composites provide increased specific strength over monolithic materials, as a result of the fiber load sharing mechanism. For improved toughness, durable polyethylene sheets are used when potted or laminating. The orthotropic nature of the unidirectional fibers and interdigital electrode permits effective design of modal actuators and sensors. The system presented in Figure 7 is not optimized for any specific self-powered application, but is suitable for applications where mechanical vibration exist (e.g. automobiles, aircraft, trucks, etc.). This system is used to conduct a systematic analysis of the feasibility and suitability of piezoelectric vibration-based energy harvesting approach for potential use with wireless sensors and other low power electronics within a military environment. Fig. 7a: Portable energy harvesting system DB-2 PCFC Fig. 7b: PCFC piezo beam and power conditioning and energy storage unit (DB-2) 182

7 SYSTEM ANALYSIS To enhance the understanding of the influencing parameters on the performance of vibration-based energy harvesting using piezoelectric materials, to contribute to future selection of optimum parameters for efficient EH system performance, and to asses the feasibility and the suitability of this technology for potential use with wireless low power electronics, an experimental parametric assessment is conducted. Using an electromechanical shaker, an accelerometer, a LabVIEW-based data acquisition (DAQ) system and the ACI energy harvesting system, the impact of the variation of excitation frequency, amplitude, material thickness and system load on the system s power output is assessed. Experimental results show that at a fixed excitation amplitude of 0.5 V a maximum power output was obtained at an optimal excitation frequency equals to that of the PFC beam s resonance frequency. According to the characteristics of the DB-2 power conditioning unit, when the capacitors are fully charged, the energy stored should be within 2.67 mj and 6.27 mj. The energy obtained during the experiment after time t = 167 seconds, for the same beam, is 2.9 mj (Table 3). As shown in Figure 8 it has been reported [8] that using random base excitation, a 1.44 mj can be stored in the capacitor after 33 seconds for material type 3 and 57 seconds for material type 1. Figure 8b illustrates the higher time required to store the same energy using the same materials in a different environment. Additionally, at the PFC beam s resonance frequency, a higher system power output was obtained with increased excitation amplitudes (Table 4), with linear and exponential increase for beam configurations with and without proof mass, respectively. Reduced capacitor charging time was also obtained for higher excitation amplitudes. Output saturation was experienced at excitation amplitude of 1.8 V (10 mj) due to the 10 V DAQ output limit. The energy increase with excitation amplitude provides flexibility in the design of the system for known energy need (e.g. specific energy needs results in specific excitation amplitude). Table 2: EH system output for excitation frequency variation at time 167 seconds Beam with proof mass Beam without proof mass Resonant frequency (Hz) Maximum energy stored (mj) Maximum power transferred (µw) At excitation frequency of 40 Hz and amplitude of 0.5 V, it was shown that with increased PFC beam thickness, the overall beam capacitance increased linearly, the material resistance decreased from GΩ to kω, and the energy stored in the capacitor increased exponentially with highest value obtained with 5-layers PFC beam (e.g. 1 mj at 8 Seconds). The PFC beams are stacked mechanically in series and electrically in parallel, resulting in Piezoelectric Multilayer Composites (PMC). Figure 9 illustrates the latter at time 10 seconds. Additional experiments, demonstrated that to obtain optimum power, a load with optimum resistance must be coupled to the capacitor. This resistance is expressed as R OPT, π R OPT =, where ω and C 0 are the resonant frequency and the capacitance of the 2C 0 ω piezoelectric material, respectively. For 2-PMC beam, an excitation frequency of 55 Hz, 183

8 excitation amplitude of 1 V, a capacitor capacitance of 7.62 nf and an optimal resistance of 600 kω are obtained with maximum power output of 75 μw. It is noted that the power output without the 600 kω load is 1.25 mw. Stored Energy (mj) mj Dodge Dakota on 29S Transducer Type1 Transducer Type2 Transducer Type Time (min) Stored Energy (mj) mj Honda Civic on 202N Transdycer Type1 Transducer Type4 Transducer Type2 Transducer Type Time (min) a b Fig. 8: Energy stored within capacitors in two environments under random vibrations Table 4: EH system output for excitation amplitude variation at time 167 seconds Beam with proof mass (W PM) Beam without proof mass (w/o PM) Resonant frequency (Hz) Maximum energy stored (mj) 10 * 4.35 Maximum power transferred (µw) ~ Optimal excitation amplitude None since energy and power linearly increase with an increase of amplitude *: DAQ range limitation +: After time t = 50 seconds at resonant frequency f = 55 Hz and amplitude of 1V ~: After time t = 6 minutes at resonant frequency f = 40 Hz and amplitude of 1V None since energy and power nolinearly increase with an increase of amplitude Additional experiments were conducted, using controlled random vibrations, and it was observed that (at time 110 seconds) a 115 μw is obtained at the capacitor in the unloaded configuration (No resistive load), whereas a 4.5 μw (at time 135 Seconds) is obtained when the capacitor is exposed to the optimal resistance of 600 kω. CONCLUSIONS The feasibility and the suitability of a vibration-based energy scavenging approach using piezoelectric material that can scavenge power from low-level ambient vibration sources were assessed. The analysis considered the impact of the variation of excitation frequency, amplitude, material thickness and system load on the system s power output. Given appropriate data acquisition system, efficient power storage and conditioning, suitable piezoelectric material, beam configuration, and innovative power amplification the resulting 184

9 vibration-based power source is sufficient to support networks of ultra-low-power and peerpeer wireless sensor nodes. Energy (m J) number of layers Fig. 9: Energy stored for different beam thickness ACKNOWLEDGMENTS This work is funded by the Director General Air Equipment Program Management (DGAEPM) of the Department of National Defence (DND). The technical contributions of Mr. Marc Genest of the Institute for Aerospace Research (IAR), the National Research Council (NRC) are acknowledged. REFERENCES 1. Sateesh, Wireless Sensor Networks - A Market Research Report, Dolcera Blogs, August 2009, 2. 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 Mater. Struct. Vol. 17, pp. 1-33, Mateu, L. and Moll, F., Review of Energy Harvesting Techniques and Applications for Microelectronics, VLSI Circuits and Systems II. Edited by Lopez, Jose F.; Fernandez, Francisco V.; Lopez-Villegas, Jose Maria; de la Rosa, Jose M. Proceedings of the SPIE, Vol. 5837, pp , Jones, M.H., Energy Scavenging for Wireless Sensor Networks, ELG7178F Topics in Communications II, Wireless Ad-Hoc Networking, November Torres, E.O. and Rincón-Mora, G.A. start: Energy-harvesting chips: The quest for everlasting life, Automotive Design Line, June 30, 2005 ( 6. Paradiso, J.A., Starner, J., Energy Scavenging for Mobile and Wireless Electronics, IEEE Computer Society, vol. 4 no. 1, pp , Harrop, P., Paybacks from Energy Harvesting, Printed Electronics World, 2009 ( 185

10 8. Advanced Cerametrics, Inc., Energy Harvesting Evaluation System, pp.1-8, 2006, ( 186