Project topics and materials
Get Complete Project Material File(s) Now! »
Energy Harvesting at Microscale
Throughout the history of technological advancements during human evolution, there have been attempts to obtain energy from the ambient atmosphere and surroundings. The cavemen used friction to create sparks and fire producing electric charge; the waterwheel was a main source of energy during the medieval periods and windmills were widely used to grind grains. As technology developed the devices became more efficient and smaller and their energy consumption also reduced proportionally. The recent developments in micro and nanoelectronics have resulted in the miniaturization of devices that have power requirements in the order of several microwatts. IoT is one of the recent technologies with a direct and evident influence on several domains such as industry, healthcare, agriculture, automobile and smart homes. In IoT, WSNs, consisting of microcontrollers, memory chips and frequency transmitters, play a key part that collecting and transmitting data and MEMS relays that consume power in the order of several microwatts [23], [24]. Therefore, along with the Internet of Things (IoT), and equally with the developments in artificial intelligence, the need of the day is to develop self-powered sensors and circuits that consume infinitesimal power and, also, that these devices and sensors are powered from ambient energy sources [2], [11]. It is equally important that these WSNs are independently powered and disconnected physically from the master nodes to ensure full autonomy of the IoT. In addition, the availability of sensors that use a minimum number of cables to interconnect the electrical architecture of the vehicle could prove to be a significant advantage for the automotive industry, as this could lead to saving of the carbon footprint is a key criterion in the future vehicle design. Therefore, the energy devices that can generate power from ambient energy sources and that can be integrated into the sensor systems, offer great scope for further development of autonomous sensor networks.
Power requirement of WSNs
Autonomous WSN is depicted in terms of four general functional boxes namely: (i) sensor section (ii) microcontroller section (iii) transmission section and finally the (iv) harvester section as shown in Figure 1.1. An IoT node will run through several operational cycles, which typically, involve several tasks such as sensing, actuation, processing, transmission, and consume similar amounts of energy at every operation sequence. Various types of sensors are used in condition monitoring such as accelerometers, thermal sensors, and bio-sensors for health monitoring, etc. [1], [2], [5] The power requirements for active sensing are generally on the micro-watt level [25]. However, the main power consumption of the system is from the microcontroller unit and the transmitter section. Therefore, to reduce the overall power consumption of the WSN, it is important to ensure that the microcontroller and the transmitter consume minimum power.
There has been a significant advancement in ultra-low power microcontrollers & RF Radio in recent years from the manufacturers. Some of the most popular microcontroller for low power applications are STM32L0-STMicroelectronics [12], PIC12LF1571-Microchip [26], MSP432P401-Texas Instruments [27] etc. For example, the STM32L0 series [12] microcontrollers from ST Microelectronics consume a quiescent current of only a fraction of microamperes while working in an operating voltage range of 1.8 V-3.6 V in the standby mode drawing power of few microwatts. STM32L0 is based on a 32-bit Arm Cortex processor designed for a broad range of embedded applications which work in three operation modes: main/active, low power and power down/standby mode. In standby mode the current consumption is less than 1µA, corresponding to a power consumption of ~3 µW. The active mode power consumption of these microcontrollers can be in the order of 50 µW-500 µW depending upon the processing needs and the duty cycle of operation. The total energy consumed by the processor is in the range of 50-500 µJ per transmission event and is controlled by the duration of active mode operation, which could last several milliseconds.
The total amount of energy required for the processing and the wireless transmission of the WSNs are typically in the order of 40-500 µJ that could change depending upon the RF protocol used for transmission. Bluetooth® Low Energy (BLE) consumes significantly less power during a transmission event compared to Wi-fi, LoRa and ZigBee. So, BLE is most suitable for applications requiring a transmission range less than 10 m, which is sufficient for IoT systems. For example, BLUENRG-2:Bluetooth® Low Energy wireless System-on-Chip (SoC) from ST Microelectronics [28], which is used in our work has an average advertising current consumption of about 15 µA during the transmitting phase while working at 2 V supply consuming power of 30 µW, at 8 dBm transmission power.
The total power consumption of all the components together (sensor, transmitter and microcontroller units) in a WSN during one duty cycle could be in the order of 100 µW to 1 mW [1] depending on the duty cycle of operation. The power consumption of WSNs in comparison with various typical low-power devices are given in Table 1.2.
The total power consumed by an IoT device PWSN in each cycle can be characterized by the power consumed by its components and their processes and expressed as: =+++ 1.1 where PSB is the standby power consumption, PMCU is the power consumed by the processor unit PS is the power consumed by the sensor, and PTX is the power consumed by the transmission section. Although the microscale ambient energy harvesters may not be able to provide the power for continuous operation, it is sufficient to power these WSN nodes in an intermittent manner. For instance, the system goes to a standby mode when the energy from the harvester is stored in a capacitor and can be used to discharge as soon as enough energy is stored for the operational mode. For an IoT sensor device in a standby state, its power drawn can be extremely low (few microwatts). Therefore, the WSN can be designed smartly with low power consumption while awake, and negligible power consumption during sleep mode.
Ambient power sources for Energy harvesting
The ambient energy available in the environment can be broadly classified into mechanical sources (vibrations, stress), radiant sources (solar, infrared, and RF) and thermal sources (temperature gradients or fluctuations). The range of available power densities expressed in (µW/cm2) of these sources is shown in Figure 1.2.
Although the power density of solar energy is high, harvesting is not possible in practical sensor application areas such as inside the machine, where the solar radiations cannot reach. Likewise, it is also difficult to harvest energy from thermal gradients where there is an absence of temperature gradient. However, for instance, automobile exhaust pipes can reach up to 700°C and can be used for harvesting thermal energy. Similarly, RF energy harvesting has also some limitations in the implementation in the machinery due to the strong absorption of RF energy by the metals, thus limiting the available power density. Therefore, the ambient energy harvesting is strictly application-specific, depending on the availability of the power source within the location of the installation target. Therefore, to power such sensor networks, that is to be installed especially in indoor applications, automotive and mechanical structures, the ambient mechanical energy can be a perfect choice. A comparison of the advantages and disadvantages of various microscale energy harvesting technologies are outlined in Table 1.3.
Vibration Energy Harvesting (ViEH)
Vibration Energy Harvesting is the technique to convert vibrations into usable electrical power by a two steps conversion. Initially, the vibration is transformed into a relative motion between two elements, with the aid of an equivalent mass-spring system as shown in Figure 1.3. The relative motion is then converted to electric power by a mechanical-to-electrical conversion mechanism (mainly piezoelectric, electromagnetic or electrostatic). The equivalent spring-mass system creates a phenomenon of resonance, amplifying the relative spring displacement z(t) compared to the input vibrations amplitude y(t), leading to an increasing harvested power at resonance [29].
In the equivalent mechanical model, a mass (M) is suspended in a frame by a spring (stiffness ks) and damped by dashpot C with an affective damping coefficient of dm+de, where dm is the mechanical and de is the electrical contribution to the damping respectively. The differential equation of the system can be described as [7]: ( ) + ( + ) ( ) + ( ) = − ( ) 1.2
The energy that can be extracted from the above system, attains a maximum when the excitation frequency ω of the is equal to the natural resonant frequency ωr of the spring-mass system.
The power converted by a vibration energy harvester to the electricity is equal to the power absorbed by the electrical induced damping de component of the dashpot C. where ζT is the total damping ratio, ζT = (ζm + ζe) = (dm+de)/2M Y0 is the amplitude of vibration, ζm is the mechanical and ζe is the electrical damping of the system respectively. The maximum electrical power that can be extracted at the resonant frequency ( = r) .
The several vibrating structures available in the environment present rich sources of usable energy for low power applications. Base accelerations of such structures vary from 0.01 to 10 g, with frequencies ranging between 1 and several kHz as shown in Table 2.3 [9].
The ambient mechanical energy can be transduced to electrical energy by main mechanisms namely piezoelectric, electrostatic, electromagnetic, and triboelectric energy harvesters. For example, the typing in keyboard can generate power of the order of few milliwatts, energy generation by the motion of upper limbs is about 10 mW. Walking can generate as much as 10 W. The microscale Electro-Magnetic Energy Harvesters (EMEHs) tend to produce very low AC voltages in the order of millivolts. Furthermore, the output voltage reduces as the size is reduced.
While comparing the output power of all the vibrational energy harvesting technologies, Electrostatic Energy Harvesters (EEHs) have a limitation of implementation that the device vibrates with a magnitude of several hundreds of microns while maintaining a minimum capacitive air gap of 0.5 mm to produce similar performance [32]. For both piezoelectric and electrostatic 13 generators, the output current will reduce with size due to the reduction of the capacitance of the device which decreases with decreasing size. Furthermore, piezoelectric generators have the advantage that it produces high voltages. Out of all these microscale vibration energy harvesting, PEHs, therefore, have an upper edge due to higher output voltage, relatively high output power density, and simplicity in design over their counterparts such as EMEHs and EEHs. Moreover, the piezoelectric transduction mechanism does not need any moving parts and, thus, is easy to maintain.
Electromagnetic Energy Harvesters (EMEHs)
EMEHs can be used to convert the vibration energy available in the ambient environment to electrical energy using the principle of Faraday’s law of electromagnetic induction. When a conductor move in a magnetic field (Figure 1.4), a voltage is induced across it. The conductor may be in the shape of a coil that is planar or wound. Generally, the magnet that is used to provide the field could be permanent magnets such as neodymium (NdFeB) [35] or Alnico (AlNiCo) [36].
The amount of induced voltage VEM (peak to peak) in the electromagnetic coil can be computed by: where, N is the number of turns, and is the total flux linkage between the magnet and the coil. The power generated by the EMEH is extracted by connecting a load resistance across the two ends of the coil. The energy harvested by the EMEH is influenced by the damping offered by the coil to the input vibrations. Also, EMEHs produce maximum power when the relative motion between the magnet and coil is maximum.
From an interesting application point of view as depicted in Figure 1.5a,b Gao et al. in 2017, proposed a novel Rail-Borne Energy Harvester for powering wireless sensor networks in the railway industry [37]. The authors proposed a magnetic levitation harvester that offers broadband harvesting at a low-frequency (3–7 Hz) for a given rail displacement of 0.6 to 1.2 mm. The harvester provided an output power of 119 mW and the output peak-peak voltage of 2.32 V is achieved with the rail displacement of 1.2 mm. Beeby et al. [35] reported a micro-EMEH of a cantilever type structure, with four magnets made from NdFeB bonded to the free end (Figure 1.5c). The coil is positioned between the two magnets, with the beam and magnets able to move relative to each other. The magnets move vertically relative to the coil that is adhesively bonded to the plastic base. The authors claimed to harvest output power of 46 µW at a relatively low acceleration of 0.06 g which is a large value compared to other EMEH reported so far. Also, the aeroelastic EMEH has received widespread attention in recent years with the advancement of MEMS technologies [14].
Table of contents :
General Introduction
Thesis Overview
Thesis Structure
Context
1 On the interest of ambient energy harvesting
1.1 Energy Harvesting at Microscale
1.2 Power requirement of WSNs
1.3 Ambient power sources for Energy harvesting
1.3.1 Vibration Energy Harvesting (ViEH)
1.3.2 Piezoelectric Energy Harvesting (PEH)
1.3.3 Piezoelectric Vibration Energy Harvesters
1.4 Power Management circuits for PEH
1.5 Summary
2 Electromechanical characterization of piezoelectric materials
2.1 Piezoelectric Materials
2.2 Piezoelectric constituent equations
2.3 Classification of Piezoelectric Materials:
2.4 Poling of Piezoelectric Materials
2.5 Ferroelectric Hysteresis Loop and Butterfly Loop
2.5.1 Ferroelectric Hysteresis Loop
2.5.2 Butterfly loop
2.6 Key Properties of Piezoelectric Materials
2.6.1 Piezoelectric coefficients
2.6.2 Youngs Modulus
2.6.3 Dielectric Permittivity
2.6.4 The electromechanical coupling factor, Kij
2.7 Towards Lead free Piezoelectric ceramics
2.8 Piezoresponse Force Microcopy (PFM)
2.8.1 Choice of AFM tip and operating frequency
2.8.2 Calibration of vertical deflection sensitivity
2.8.3 Measuring dij coefficients
2.9 PFM Characterization of Lithium Tantalate (LT) samples
2.9.1 On the interest of Lithium Tantalate
2.9.2 LT samples and topography
2.9.3 Domain polarization of LT samples
2.10 PFM study of Sodium Potassium Niobate (KNN) thin films
2.10.1 Introduction
2.10.2 KNN cantilever dimensions and topography
2.10.3 Piezoresponse characterization of KNN films by PFM
2.11 External Poling of KNN PTF
2.11.1 Poling process:
2.11.2 PFM study of the poled KNN PTF
2.12 PFM study of Sodium Potassium Niobate (KNN) thin films doped with Tantalum
2.12.1 Introduction
2.12.2 Piezoresponse characterization of KNNT film by PFM
2.12.3 Conclusion
2.13 PFM study of Sodium Potassium Niobate (KNN) and KNN-Ta doped (KNNT) fibres
2.13.1 Introduction
2.13.2 Experimental setup and procedure
2.13.3 Topography of KNN fibers
2.13.4 PFM study of the KNN fibres before poling
2.13.5 Radial Poling of KNN fibres
2.13.6 PFM study of KNNT (KNN doped with 10% Ta) Fibers
2.13.7 Conclusion
2.14 Comparison of the Figure of Merit of the cantilevers
2.15 Summary
3 Piezoelectric Microgenerator based on lead-free Lithium Niobate single crystal
3.1 Introduction
3.2 Microgenerator geometry and Fabrication process
3.3 Dynamic Electrical Characterization
3.4 Estimation of Mechanical Quality Factor (Qm) based on impedance curve
3.5 Electromechanical coupling coefficient (k33 and k31)
3.6 Design of Proof mass
3.7 Equivalent Circuit model of Piezoelectric Energy Harvester (PEH)
3.7.1 Experimental Parameter Identification
3.7.2 Analysis of the impedance of the equivalent circuit in SPICE
3.8 Estimation of the mechanical quality factor (Qm) and coupling coefficient K with proof mass
3.9 Optimal load resistance for maximum output power
3.9.1 Theoretical prediction
3.9.2 Experimental determination of optimal load
3.9.3 Optimal load by LT-SPICE
3.10 Output Voltage and RMS power of the harvester
3.11 Test Results and Analysis
3.12 Power harvesting circuit for Piezoelectric Energy Harvester
3.13 A novel three-terminal harvester concept with Maximum Power Point Tracking (MPPT) for ultra-low power applications
3.13.1 Introduction
3.13.2 Theory and working principle
3.13.3 Design considerations
3.13.4 The feasibility study of 3 terminal harvester
3.13.5 Bending-Mode Resonance Frequency
3.13.6 Finite element Analysis
3.13.7 Boundary Conditions
3.13.8 Experimental results
3.13.9 Measurement of Open circuit voltage
3.13.10 Conclusion
3.14 Summary
4 Energy Autonomous Wireless Vibration Sensor based on Piezoelectric Microgenerator
4.1 Introduction
4.2 System description
4.3 Piezoelectric Energy Harvester Design
4.4 Energy Autonomous Wireless Vibration Sensor Working Principle
4.5 Experimental Results
4.6 Conclusions
Conclusion and Perspectives
References
