Msc Mechatronics Engineering

DESIGN OF A UP-CONVERTED 2-DOF PIEZOELECTRIC ENERGY

HARVESTING PROTOTYPE BASED ON HUMAN HEART VIBRATIONS

By: Hygin Davidson MAYEKOL MAYECK

Reg. No: MT300-0001/2019

Supervisors: Prof. Ahmed Mohamed Rashad Fath El-Bab

Dr. Evan MURIMI

Research proposal for thesis to be submited in partial fulfilment of the requirement for the
award of degree of Master of Science in Mechatronic Engineering of Pan African University
Institute for Basic Sciences, Technology and Innovation

May 2020
Declaration
This research proposal is my original work and to the best of my knowledge, it has not been
presented for a degree award or any other purpose in this or any other university.

Hygin Davidson MAYEKOL MAYECK

Reg. No: MT300-0001/2019

Sign.......................... Date..........................

This research proposal is submitted to Pan African University Institute for Basic Sciences,
Technology and Innovation with our approval as the Institute supervisors:

Prof. Ahmed Mohamed Rashad Fath El-Bab

Sign.......................... Date..........................

Dr. Evan MURIMI

Sign.......................... Date..........................

i
Abstract
Energy harvesting is a technique of collecting energy from the ambient environment and trans-
form it into electrical energy. Recent technologies are more interested by developing au-
tonomous and self powered devices. These devices offer a lot of advantages and have a wide
rang of applications. One form of energy found in the environment is the kinematic energy,
and it is the most available.
Few devices are developed to operate at very low frequency. Even though some have tar-
geted the human motion vibration frequency range frome 1 to 10Hz, in practice the operation
frequency of their devices are from 2 to 6Hz, related to human limbs motion and used to power
hand-held devices. The small range from 1 to 2Hz is important since it covers the heart beat
frequency, and harvesters that can operate in this range may be used to power medical implants
placed near the heart, such as pacemaker.
This proposal presents a design of an energy harvesting device capable of transforming
human heart vibration frequencies into electrical energy. That energy harvester aim to power
pacemakers. The model presented in this work is a 2-DOF piezoelectric vibration absorber,
made with a cut-out beam, that is up-converted by two cantilevers facing each side of the 2-
DOF system. Classical piezoelectric up-converted devices produce output either from one or two
cantilevers. The proposed harvester will generate power from three beams, and the frequencies
are chosen to cover a wide bandwidth in the range of human body vibration frequencies.
The 2-DOF system will cover frequencies from 1 to 2Hz, which covers the frequencies from
human heart beat, and the up-converter beams will have resonant frequency from 5 to 6Hz,
which are the most dominants frequencies from human motion. The present harvester is ex-
pected to operate throughout the human body frequency range. The output power sould be
higher than the simple 2-DOF system, and better results are expected in terms of power den-
sity. After designing the model, an experimental work will be done in order to investigate the
system performance and validate the design methodologies.
This research may inspire further studies to impliment the proposed device in MEMS scale,
or to propose other designs of harvesters that operate at very low frequency.

ii
Contents

Declaration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii
List of tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
List of figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
List of abreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
List of symbol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

1 Introduction 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Problem statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3.1 Specific objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2 Literature review 4
2.1 Area of applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 Transduction techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2.1 Piezoelectric transduction . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.2 Electrostatic transduction . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.3 Electromagnetic transduction . . . . . . . . . . . . . . . . . . . . . . . . 10
2.3 2-DOF vibration energy harvester . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.3.1 Hybrid energy harvester . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.3.2 L-shaped energy harvester . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.3.3 Cut-out beam configuration . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.4 Non-resonant techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.4.1 Frequency Up-conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.4.2 Free moving mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.5 Research gaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

iii
3 Methodology 20
3.1 Energy harvester design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.1 Mathematical model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.2 Design layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.1.3 Frequency range selection . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.1.4 System parameters selection . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2 Electrical model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2.1 Basic configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2.2 Material properties and selection . . . . . . . . . . . . . . . . . . . . . . 25
3.2.3 Analytical model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.3 Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.3.1 3D model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.3.2 Virtual test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.4 Device fabrication and experimental test . . . . . . . . . . . . . . . . . . . . . . 30

4 Expected outcome and schedule 32

4.1 Expected outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.2 Budget . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.3 Source of Funds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.4 Work Schedule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

iv
List of Tables

2.1 Piezoelectric Material Properties [2] . . . . . . . . . . . . . . . . . . . . . . . . . 7

3.1 Comparison of PZT and PVDF properties [19] . . . . . . . . . . . . . . . . . . . 26

4.1 Budget of the research project . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

4.2 Work Schedule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

v
List of Figures

2.1 Size comparaison of pacemaker: (a) Traditional single chamber pacemaker VS

Micra, (b) Actual size of Micra pacemaker. Image from [Link]
com . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Micra pacemaker: (a) Illustration of the pacemaker, (b) Detailed illustration.
Image from [Link] . . . . . . . . . . . . . . . . . . . . . 6
2.3 Illustration of a piezoelectric bimorph [19] . . . . . . . . . . . . . . . . . . . . . 6
2.4 Comparison of output power of piezoelectric generator using different piezoelec-
tric materials [1] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.5 Piezoelectric constants in typical EH modes [2] . . . . . . . . . . . . . . . . . . 8
2.6 Circuit representation of a piezoelectric generator [1] . . . . . . . . . . . . . . . 8
2.7 Electrostatic generator: (a) in-plan gap overlap; (b) in-plan gap closing; (c)
out-of-plane gap closing [4] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.8 Circuit representation for electrostatic generator [4] . . . . . . . . . . . . . . . . 10
2.9 Electromagnetic generator: (a) Uniform magnetic field; (b) Variable magnetic
field [4] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.10 Electromagnetic generator equivalent circuit [4] . . . . . . . . . . . . . . . . . . 11
2.11 Scematic of the multimodal energy harvesting device [20] . . . . . . . . . . . . . 12
2.12 Illustration of L-shaped piezoelectric energy harvester [20] . . . . . . . . . . . . 12
2.13 2-DOF layout [22] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.14 (a). Shematic diagram of a vibration absorber, (b) Normalized amplitudes vs
normalized frequency [22] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.15 Normalised amplitude vs Normalised frequency [22] . . . . . . . . . . . . . . . . 15
2.16 Shematic structure of the frequency Up-conversion proposed by Halim and Park
[9] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.17 Impact vibration energy harvester [13] . . . . . . . . . . . . . . . . . . . . . . . 17
2.18 Frequency up-converter stopper: (a) Design 1, (b) Design 2 [14] . . . . . . . . . 17

vi
2.19 Schematic diagram of a scrape-through piezoelectric energy harvester [12] . . . . 18
2.20 Free moving mass: (a) free moving object, (b) free moving liquid [5] . . . . . . . 19

3.1 Operating mechanism of the proposed scrape-through piezoelectric energy har-

vester [12] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2 Circuit representation of piezoelectric bimorph [19] . . . . . . . . . . . . . . . . 26
3.3 Ciruit representation of a piezoelectric bimorph with load [19] . . . . . . . . . . 28
3.4 Equivalent circuit of a piezoelectric generator [24] . . . . . . . . . . . . . . . . . 29
3.5 Schematic diagram of the experimental setup . . . . . . . . . . . . . . . . . . . 31

vii
List of abreviations

EH Energy Harvesting
MEMS Micro-Electro-Mechanical-System
1-DOF One Degree of Freedom
2-DOF Two Degree of Freedom
PZT Piezoelectric Lead Zirconate Titanate
PZT-5H Piezoelectric Soft Lead Zirconate Titanate
PZT-5A Piezoelectric Hard Lead Zirconate Titanate
PVDF Polyvinylidene Fluoride
WSN Wireless Sensor Network
BSN Body Sensor Network
LF-S Low Frequency System
HF-S High Frequency System
FUC Frequency Up-Conversion
GEMC Generalized Electromechanical Coupling

viii
List of Symbols

Symbol Description
E Young’s Modulus [Pa]
k Stiffness
ω Frequency
m Mass
P Power
V Voltage
I Current
x Displacement
X Amplitude
F Force
I Moment of inertia
l Length
t Thickness
d31 Transverse-axial Piezoelectric Constant
ε33 Transverse Dielectric Coefficient
ξ Strain
δ Stress
Γ Generalized Electromechanical Coupling factor

ix
Chapter 1

Introduction

1.1 Background
Autonomous systems have inspired an important number of research. These autonomous de-
vices require mostly low power sources, with lifespan that match their work-time. Energy
harvesting is a method of transforming ambient energy to electrical energy in order to supply
other devices. This technique can produce power in the range of micro to milli-watts, which
constitute enough power to supply autonomous and micro electronics systems [1]. Among the
applications of energy harvesting there are wireless sensors, implants, embedded system, etc.
The forms of energy that existe in the environment and are exploited by energy harvesting
system are mostly thermal, kinetic, solar or electromagnetic. Solar power gives a better power
density in outdoor condition but become less suitable for embedded systems, indoor applications
or dirty environments where the photocells can be obscured [2]. The thermal energy require
a thermal source, and this condition limits its application. Such limitation is also noticed for
electromagnetic energy. Kinematic energy seems to be the most available form of energy. Also
defined as mechanical energy, it can be vibrations or motion. It can be found anywhere, this
makes it to be the most suitable source for energy harvesting. The range of frequency of the
excitation depend on the source. A human body in motion can produce frequencies from 1 to
10Hz [3], while a healthy human heart beats 60 to 100 times per minute, depending on the age
and fitness level, that is translated to 1-1.67Hz 1 .
Mechanical energy harvesting can be done trough piezoelectric, electromagnetic or electro-
static transduction. However most vibration power generators are resonant system, this means
they produce a maximum power when their resonant frequency matches the excitation fre-
1
Joseph Kuan (2017). [Link]
Acceced 19-4-2020

1
quency. The consequence of such system is that at any other frequency, an important decrease
in the power generated will be observed [4]. To improve the aspect of the harvester, there are
techniques for widening operating frequency (frequency tuning, multi-frequency and non-linear
system), multidirectional harvester and non-resonant approach (frequency-up conversion and
free moving mass) [5].
Systems that operate in wide frequency range show improvement in terms of power gener-
ation. The 2-DOF is a frequency tuning technique for widening the operating frequency of a
harvester. However it is a resonant system, which means that the power generated can still be
improved by applying the non-resonant approach.
Maamer et al. [5] made the assumption that non-resonant system approach is more apropri-
ate for low frequency energy harvesting. Reason is that the power generation performances of
resonant energy harvesters (cantilevers and spring-mass system) are limited when they operate
in low frequency and wideband applications, whatever the transduction mechanisms [6]. The
non-resonant system is obtain by creating a system that does not depend on a specific resonant
frequency. Such system may not have a mechanical stiffness (free moving mass) or transferring
the initial vibration to a higher frequency vibrator (Up-conversion).

1.2 Problem statement

Energy harvesting has become a solution to create autonomous system, by replacing batteries or
extend their life span. The applications are from industries, mobile devices or medical implants.
Vibration energy harvesters are most interesting, and piezoelectric transduction is the most
used technique when it comes to mechanical energy harvesting because it gives relatively high
electrical power.
It is important to develop devices capable of harvesting energy from very low frequencies,
like human body frequencies. This can be applied to mobile devices such as watches or medical
implants, and will reduce the dependency to battery that has to be replaced after a range of
time. The process of replacing the battery for medical implants is expensive and risky for the
patient. Therefore, an energy harvester that can harvest mechanical energy from human body
vibrations frequencies and transduce it into electrical energy could be a solution to improve
human wireless sensor technology and the performance of biomedical implants.
A pacemaker is a medical implant used to regulate the heart beat. Most pacemakers use
leadeds to flow signals between the heart and the device. A new generation of pacemaker,

2
which is leadless, developed by Medtronic can be placed directly inside the heart 2 . This new
implant is much smaller than the classical version (about -93%) and has many advantages in
termes of safety. However, both versions still depend on the battery lifespan. The dependance
to the battery means that the device has to be replaced after a certain range of time. The
cost of the device and the surgery is very high. However if the medical implant can be made
autonomous, the patiennt can avoid others surgeries and their risks.
It is then important to develop an energy harvester that can power pacemakers. To power
such devices, the harvester has to match the heart beat frequencies’ range of 1 to 1.67Hz and
satisfy the size and power requirement of the devices.
For devices operating in frequencies below 50Hz, like the human body which has very low
frequencies, it is preferable to harvest energy with non-resonant systems [5]. The combination
of 2-DOF vibration absorber and frequency Up-conversion must be one way to approach the
solution of this problem. The 2-DOF energy absorber technique is applied for widening the
operating frequency range, and the frequency up conversion will transfer that frequency to a
higher frequency vibrator in order to produce a higher output power.

1.3 Objectives
This work aim to develop a piezoelectric energy harvester based on human heart beat frequen-
cies.

1.3.1 Specific objectives

• To design a 2-DOF energy absorber that can operate at frequencies of 1 to 1.67Hz;

• To combine the 2-DOF energy absorber with a frequency up-conversion system;

• To develop a finite element model to investigate the system performance;

• To validate experimentally the performance of the harvester.

2
Medtronic (2020). [Link]
[Link]
Accessed 19-4-2020

3
Chapter 2

Literature review

Energy harvesting has been widely investigated. In the literature there are papers that had
studied different techniques and designs. The transducing techniques mostly used for mechan-
ical energy harvesting are piezoelectric, electromagnetic and electrostatic transduction. Many
researchers have showed their interest in considering human body vibrations [7]–[11] and MEMS
applications [12]–[14].

2.1 Area of applications

Paul D. Mitcheson et al. [15] reported how energy harvesting could be applied in wireless
sensor network (WSN) which include status monitoring of machinery, structural monitoring
of building, environmental monitoring of domestic environments, security, military tracking,
aircraft engine monitoring, wearable computing, personal tracking and recovery systems. This
systems require to harvest energy from human motion, machinery and structural vibrations.
The body sensor network (BSN) is a specific type of WSN and has a medical application. It
is the most likely to use body motion vibrations. The evolution of low power and micro-size
electronics has facilitated the potential application of energy harvesting in BSN. The amount
of power needed for the sensor node is of 10 to 20µW or less, and even though the power
needed for the interface circuit has to be considered, the required power will still be of a few
micro-watts.
Four main factors are important to harveste energy for portable electronic devices: the
typical power consumption of the device, the usage pattern, the device size (thus the harvester
allowable size), and the motion to which the device is subjected (specifically for motion harvest-
ing). This made wireless sensors and implantable devices appear to be the primary application
area for body motion harvesting [15].

4
 Energy harvester based on human body vibration had been applied mostly on hand-held
devices. Ju et al. [16] came up with a non-resonant magnetoelectric vibration energy harvester,
simulated both by hand-shaking and with a vibration exciter. When excited manually the device
produced 5 − 11.2V and 0.35 − 0.57µW , while with a rang of excitation frequency of 6 − 20Hz
and 1 − 3g acceleration it produced 1.43V for 19Hz, 2.2V and 0.045µW for 20Hz. Halim and
Park [9] reported that the hand-shaking vibration has frequencies of 2.5−6Hz and 15−20ms−2
amplitude. Halim et al. [11] used these characteristics to design a device that produces 110µW
for an average power of 15.4µW cm−3 .
Any energy harvester that aim to power a device must respect the power and size require-
ments of that device. The battery of a pacemaker represents 2/3 of the device volume and
1mW is reasonable to power it [17]. For an implant that is 42 × 51 × 6mm3 , the harvester size
should be less or equal to 27 × 27 × 6mm3 [17]. The Micra pacemaker is 93% smaller than
the classical type of pacemaker (Figure 2.1). The device has a volume of 0.8cm3 , its battery
constitute almost half of the total volume (Figure 2.2), and it require 1.5V to operate 1 . To
power it, the energy harvester must respect a very strict size requirement of 0.3cm3 [7].

(a) (b)

Figure 2.1: Size comparaison of pacemaker: (a) Traditional single chamber pacemaker VS
Micra, (b) Actual size of Micra pacemaker. Image from [Link]

2.2 Transduction techniques

These are techniques used to convert mechanical energy present in the environment into elec-
trical energy. They are three, piezoelectric transduction, electromagnetic transduction and
1
Medtronic (2020). [Link]
[Link]
Accessed 19-4-2020

5
 (a) (b)

Figure 2.2: Micra pacemaker: (a) Illustration of the pacemaker, (b) Detailed illustration. Image
from [Link]

electrostatic transduction.

2.2.1 Piezoelectric transduction

This transduction technique uses piezoelectric materials to transform mechanical energy into
electrical energy, and it is the most used in the literature. It has an important potential in
MEMS application for biomedical devices, and can produce relatively high power. Piezoelectric
materials are able to produce a potential difference when they are subjected to a deformation,
and deform when a potential difference is apply on them. This property is used to convert
vibration energy to electrical energy, and the technique offer a simple approach in the process.
The output voltage is > 5V but current is very low because the output impedance is typically
high, more than 100KΩ [18]. Figure 2.3 shows the shematic of a piezoelectric bimorph.

Figure 2.3: Illustration of a piezoelectric bimorph [19]

Piezoelectric materials are either single crystal (e.g. quartz), ceramics (e.g. PZT), thin-
film (sputtered zinc oxide, screen printable thick films based upon piezoceramic powders),

6
and polymer materials such as polyvinylidene fluoride (PVDF). The following table show the
parameter of piezoelectric material [2]. The generated output power depend on the type of
material used (Figure 2.4). PZT and PVDF materials are the commonly used materials for
piezoelectric transduction [4]. In MEMS, PZT materials have many advantages like providing
large motions and low hysteresis [14].

Table 2.1: Piezoelectric Material Properties [2]

Property Constant definition Units
√
Electromechanical (mechanical energy stored ÷ electrical energy applied)
√ -
coupling coefficient k (electrical energy stored ÷ mechanical energy applied)
strain ÷ applied field m/V
Piezoelectric constant d short circuit charge ÷ applied stress C/N
open circuit field ÷ applied stress V /N
Piezoelectric constant g strain developed ÷ applied charge density m/C

Figure 2.4: Comparison of output power of piezoelectric generator using different piezoelectric
materials [1]

The subscript of the parameters is determined by the direction of the applied force (Fig-
ure2.5). For a mechanical strain that exerce in the direction of polarisation, the constants
are denoted with the subscript 33 (e.g. d33 ). If the direction of application of the strain is
perpendicular to the polarisation, the constants are denoted with the subscript 31 (e.g. d31 ).
The electric circuit representation of a piezoelectric generator is given by Figure 2.6. Vo.c is
the open circuit voltage, RL and Rs are respectively the load and the piezoelectric material’s
resistances. C is the capacitance of the piezoelectric material between two electrodes. The
power generation and further explanation about this transduction technique will be given in
Section 3.2. The constitutive equations of a piezoelectric converter given by [1] are the following:

7
 σ
δ= +d·E (2.1)
Y
D =ε·E+d·σ (2.2)

Where:
δ is the mechanical strain
σ is the mechanical stress
Y is the Young’s modulus
d is the piezoelectric strain coefficient
E is the electric field
D is the electrical displacement or charge density
ε is the electric constant of the piezoelectric material

Figure 2.5: Piezoelectric constants in typical EH modes [2]

Figure 2.6: Circuit representation of a piezoelectric generator [1]

8
2.2.2 Electrostatic transduction

This transduction technique works like a variable capacitor where the total energy stored varied
depending to the external mechanical force. This method is mostly applied in micro scaled
systems. It produces a voltage of > 100V and a low output current [18]. Since the capacitor
has to be charged first to an initial voltage for it to start the conversion process, an external
voltage source is required. This condition is the main disadvantage of this technique [19].
There are three types of electrostatic configurations, the in-plane gap overlap (Figure 2.7.a),
the in-plan gap closing (Figure 2.7.b) and the out-of-plane gap closing (Figure 2.7.c). From
these three types, two mode of operations are possible, the voltage-constrained which provides
more energy; and the charge-constrained which generated energy can approach the one of the
first mode by adding a capacitor in parallel withe the harvesting capacitor [4]. The cicruit
representation of the electrostatic transduction is given by Figure 2.8.

Figure 2.7: Electrostatic generator: (a) in-plan gap overlap; (b) in-plan gap closing; (c) out-of-
plane gap closing [4]

9
 Figure 2.8: Circuit representation for electrostatic generator [4]

2.2.3 Electromagnetic transduction

The technique obey to Faradey’s law of electromagnetic induction. Permanent magnets are used
to produce a magnetic field, and coil as conductor. One of the components is fixed while the
other is attached to the mass. The electrical energy is generated by the relative displacement
of the movable part. The voltage across the coil is function of the magnetic field, the velocity of
the motion and also the number of turn of the coil. Two configurations are possible, the uniform
magnetic field (Figure 2.9.a) and the variable magnetic field (Figure 2.9.b) [4]. The electrical
circuit representation of electromagnetic generator is given in Figure 2.10. The presence of
the magnetic field make the technique less suitable for medical implant, and electromagnetic
harvesters have low power density [18].

Figure 2.9: Electromagnetic generator: (a) Uniform magnetic field; (b) Variable magnetic field
[4]

Electromagnetic transduction method offer a variety of configurations and produces a low

10
 Figure 2.10: Electromagnetic generator equivalent circuit [4]

voltage output for a high output current. However the presence of the magnetic field limits its
application. The amount of power produced is influenced by the magnetic field density and the
number of coil used. Compare to the piezoelectric transduction method, the power density for
high frequency systems is important while for low frequencies applications the piezoelectric is
more adapted [18].

2.3 2-DOF vibration energy harvester

This is a particular type of multimodal energy harvester in which the first two bending modes
are used. Different configurations are found in the literature.

2.3.1 Hybrid energy harvester

Tadesse et al. [20] presented a hybrid energy harvester composed of a permanent magnet
attached at the tip and a cantilever beam with bonded piezoelectric plates. The magnet oscil-
lated within a stationary coil fixed to the cage (Figure 2.11). The electromagnetic scheme was
efficient for the first mode while the piezoelectric was produicing considerable output for the
second mode. However the gap between the two resonant frequencies was too much (20Hz and
300Hz). Such bandwidth may only be useful for system that has very wide operation frequency
spectrum.

2.3.2 L-shaped energy harvester

Another configuration for 2-DOF vibration energy harvester is the L-shaped cantilever piezo-
electric structure 2.12. Ertuek et al. cited in [20] used this configuration. With proper selection

11
 Figure 2.11: Scematic of the multimodal energy harvesting device [20]

of the parameters, the second resonant frequency was almost double the first. However there
is a challenge in avoiding mode-shaped-dependent voltage cancelation.

Figure 2.12: Illustration of L-shaped piezoelectric energy harvester [20]

2.3.3 Cut-out beam configuration

Based on the principe of undumped dynamic vibreation absorber proposed by Hartog [21], it
consiste of a system with two masses (m1 and m2 ) and two springs (k1 and k2 ). Magdy et al.
[22] designed a piezoelectric energy harvester as a cut-out beam (Figure 2.13). This 2-DOF

12
configuration is compact and offer a minimum use of the harvester’s volume.

Figure 2.13: 2-DOF layout [22]

The system can be represented as shown in Figure 2.14. The dynamic model of the system
under an excitation force F = fo sin ωt is given by:
      
m1 0 x¨1 k1 + k2 −k2 x1 F
  +    (2.3)
0 m2 x¨2 −k2 k2 x2 0

f0 and ω are the amplitude and frequency of the exciting force, x1 and x2 are respectively
the displacements of m1 and m2 . The amplitudes X1 and X2 produced by each masses are
expressed as:

(k2 − m2 ω 2 )fo
X1 = (2.4)
(k1 + k2 − m1 ω 2 )(k2 − m2 ω 2 ) − k22
k2 fo
X2 = (2.5)
(k1 + k2 − m1 ω 2 )(k2 − m2 ω 2 ) − k22

From Equation 2.5 it can be observed that the amplitude X1 of the main system mass
√
cancels when the exciting frequency ω is equal to the value of ω22 = mk22 . At that moment,
the amplitude of the second mass becomes X2 = − kf02 . This condition allow the second spring-
mass system to operate as a vibration absorber [22].
The maximum energy obtained from this system is equivalent to the amount of the dissi-
pated energy by the correspondant damping ratio c2 due to transaction from the mechanical
energy to the electrical one. The output power is given by equation 2.6 and is function of the
difference between the two amplitude. The bandwidth depend also on the difference between
the two amplitudes as shown in Figure 2.15. There is then a tradoff between the bandwidth

13
Figure 2.14: (a). Shematic diagram of a vibration absorber, (b) Normalized amplitudes vs
normalized frequency [22]

and the generated power [22].

P = c2 ω 2 (X2 − X1 )2 (2.6)

2.4 Non-resonant techniques

The literature reported two main non-resonant techniques, the Frequency Up-conversion and
the Free moving Mass technique.

2.4.1 Frequency Up-conversion

It combines two subsystems, a low-frequency system (LF-S) and a high-frequency system (HF-
S). The first system absorbs low-frequency excitations and transfer it to the second one which
now transform it into electrical energy.
Quite a number of researchers had work on frequency up-conversion with different trans-
duction techniques and proposed their design. Up-conversion is obtained by impact between
the LF-S and the HF-S [9], [11], [13], [14], [16]. Halim and Park [9] proposed an electromagnetic

14
Figure 2.15: Normalised amplitude vs Normalised frequency [22]

15
energy harvester from human-body-induced vibration. The design of the harvester is made of a
free moving non-magnetic ball within a cylinder which hits periodically tow magnets suspended
on helical compression springs, and allow the magnets to vibrate with high frequency (Figure
2.16). The tests are made with hand-shake (2.5-6Hz). The same design in [9] is presented in
[11], and the harvester offered a power density of 0.33mW cm−3 what makes the EH suiteble
for hand-held smart devices since it requires a movement.
Ju et al. [16] replaced the non-magnetic ball by a permanant magnet, and the suspended
magnet by two aluminium springs. The low frequency cause the ball to move and hit periodically
both ends, achieving frequency up-conversion from low frequency input vibration. In order to
return the compressional energy stored in the cage during the impact, they used the trampoline
effect.

Figure 2.16: Shematic structure of the frequency Up-conversion proposed by Halim and Park
[9]

Frequency up-conversion by impact can also be applied in piezoelectric transduction. Lei

Gu [13] used a driving beam, working in ambient low frequency, hits high resonant frequency
beams (2 electricity-generating beam), thus up-converts the LF (Figure 2.17). The output
voltage is generated by the two HF beams with frequency superior to 100Hz. Liu et al. [14]
used only two cantilevers, one as excitation oscillator and the other as frequency up-converter
stopper. In their work, two design are proposed as shown in Figure 2.18.
An other configuration of frequency up-conversion is proposed by Liu et al. [12]. This time
the LF-S scrape-through HF-S. They used a LF PZT (36Hz) cantilever attached to a proof
mass that impacts and scrapes through the HF PZT (618Hz) cantilever (Figure 2.19). This
configuration allow the two beams to be free of bending. The two cantilevers have the same
resonant frequency but the proof mass drops considerably the one of LF PZT. Their design is

16
 Figure 2.17: Impact vibration energy harvester [13]

(a) (b)

Figure 2.18: Frequency up-converter stopper: (a) Design 1, (b) Design 2 [14]

17
not modelled as a spring mass system, rather they use the strain distribution approach. The
output voltage is function of the strain along the beam.

(a) (b)

Figure 2.19: Schematic diagram of a scrape-through piezoelectric energy harvester [12]

Using piezoelectric transduction, the output power and the power density are significantly
improved with frequency up-conversion compare to the conventional piezoeletric beam. Plus
the technique is suitable for MEMS implementation [12]–[14].

2.4.2 Free moving mass

The free moving mass method is divided into two types, the free moving object and the free
moving liquid. While the frequency up converter require at least one resonant system for
transduction, free moving mass devices are totally non-resonant and based on the arbitrary
motion of a mass. These are suitable for frequencies below 10Hz and imply 3D harvesting. A
free moving mass can be attached by a rope, a rod or rolling inside a cage (Figure 2.20.a). The
free moving liquid make use of ferromagnetic fluid (Figure 2.20.b) [5].
The free moving mass system does not imply a mechanical stiffness but involve the presence
of an electromagnetic flux [5], which make it applicable mostly with electromagnetic transduc-
tion [23]. The free movement of the object creates an electromagnetic flux across the coil.
The permanant magnets and the ferrofluid are responsible of a closed-loop magnetic circuit
[5]. The literature has shown that the frequency up-conversion is applied for both piezoelectric
and electromagnetic transduction [9], [12] while it is rare to find an example of the free moving
mass applied with a different transduction technique than the electromagnetic one.

18
 (a) (b)

Figure 2.20: Free moving mass: (a) free moving object, (b) free moving liquid [5]

2.5 Research gaps

Many research are still ongoing in energy harvesters capable of harvesting energy from ambiant
energy. From this literature review, the following gaps have been identified:

• Most of researcher have used the spring-mass model for system operating at low frequency,
bellow 50Hz, the case of human motion vibration;

• Most of human vibration based harvester exploite motion of human limbs. There is a
need to develop harvesters based on human heart vibrations;

• The non-resonant approach is not widely applied for low-frequency systems, especialy for
human vibrations based harvesters;

• Investigations are mostly theoretical and a few prototypes were made in micro-scale.

19
Chapter 3

Methodology

The methods used for this study must help us to achieve our objectives. The two first specific
objectives, which consiste of designing the 2-DOF and Up-conversion systems, are satisfied
by the first two section of this chapter. The 2-DOF system constitutes the LF-S of the up-
conversion. The LF-S will transfer its vibrations to the HF-S by physical contact. The HF-S is
constitute of two cantilevers that face each side of the 2-DOF system. Even though the LF-S
and the HF-S are two different systems, they are designed as one system.

3.1 Energy harvester design

3.1.1 Mathematical model

The mathematical model represent the system dynamic and help to determine the different
parameters of the system. The 2-DOF system is design using the following steps in Section
2.3.3 proposed by [22]. The frequency range and the two amplitudes of this 2-DOF system
are the needed parameters to design the HF-S. This is where the combination between the two
systems (2-DOF and up-converter) takes place. The frequency vibration of the 2-DOF system
is transfered to a higher frequency system, which is constituted of two cantilevers without
proof mass. The harvester become a combination of two frequency up-conversion systems. The
amplitudes X1 and X2 are important in determinig:

• The gap between the LF-S and the HF-S beams;

• The distance between the 2-DOF beam and the ground.

The proposed design will opperate at very low frequency, which must make it to be fabricated
as a MEMS structure. Such mechanical system should not be described by a simple spring-mass

20
system, because, as it was mentioned early in Section 1.1, the spring-mass system is a resonant
system and hens is not suitable for low frequency systems. The strain distribution approach
proposed an alternative to the spring-mass model. The mathematical equations were proposed
by [19], [24].
For a beam with a mass attached on it, it consider a concentration of the applied force at
the middle of the mass as follow:

( )
3tb 2lb + lm − 2x
ξ(x) = δm (3.1)
lb 4lb2 + 9lb lm + 6lm
2

Where tb , lb , lm , and δm are respectively the thickness of the beam, the lenght of the
beam, the lenght of the tip mass and the mass displacement. For a beam without mass, the
strain distribution is given by:
3tb (lb − x)
ξ(x) = δb (3.2)
2lb3
Where δb is the beam displacement. The open circuit voltage obtained depend on the
piezoelectric material properties and the strain distribution along that layer. It is given by:

∫ lb
d31 Y tc
Vo.c = ξ(x)dx (3.3)
ε33 lb 0

Where Y and tc are respectively the Young’s modulus and the thickness of the piezoelectric
material, d31 and ε33 are the transverse-axial piezoelectric constant and the transverse dielectric
coefficient. Therefore the open circuit voltage for the beam without mass (Equation 3.5) and
for the beam with mass (Equation 3.4) are:

−d31 Y tc 3(lb + lm )tb δm

[Link] = (3.4)
ε33 lb 4lb2 + 9lb lm + 6lm2

−d31 Y tc 3tb δb
[Link] = (3.5)
ε33 4lb2

The derived electrical power is given by:

2
1 Vo.c
Pmax = ZL (3.6)
2 (Zp + Zl )2
Where Zp is the piezoelectric impedance and ZL is the load impedance. To obtain the
maximum power, the piezoelectric and load impedances have to match.

21
3.1.2 Design layout

As explained earlier, the 2-DOF vibration absorber (Figure 2.13) constitutes the LF-S while the
HF-S is constituted of two piezoelectric cantilevers. A gap is predetermined between the two
systems to allow a transfer of energy from the low frequency system to the high frequency system
by scraping through it (Figure 2.19). When the vibration amplitude of the LF-S is sufficient,
it hit periodically the HF-S and the base. This results to a broadening of the bandwidth of the
LF-S and a transfer of vibration to the HF-S (Figure 3.1).

Figure 3.1: Operating mechanism of the proposed scrape-through piezoelectric energy harvester
[12]

The system parameters (m1 , m2 , k1 , k2 , kh1 , kh2 ) and dimensions (lenght and
thickness) will be selected to satisfy the following criteria:

• The resonant frequency of the 2-DOF system is selected to match the human heart beat
frequency range of 1-6Hz.

• The frequency bandwidth of the harvester should match the human frequency spectrum.

• The generated output should be as large as possible

• The device should give a good power density

• The device should be of low cost

• The device volume should not exeed 0.3cm3 .

22
3.1.3 Frequency range selection

The human heart beat frequency range is from 1 to 1.67Hz 1 . The main frequency is chosen
to be ω = 1.3Hz ≈ 8rad/s. In case the resonant frequency of the excitation is below 50Hz,
the resonant frequency of the FUC stopper should be around 5 to 10 times higher, and a high
acceleration is also needed to obtain a better performance (wider operating bandwidth and
higher power output) [12]. The FUC stopper frequency is selected to be 5 times higher than
the choosen main frequency i.e. 6.5Hz. This is near the most dominant frequency of the human
motion frequency range of 5Hz [8].

3.1.4 System parameters selection

The first step in the system parameter selection is the material selection. In MEMS applications,
PZT materials have a lot of advantages like providing large motions and low hysteresis [14]. But
they have high density, brittleness, low yield strains and high cost of raw material. However
PVDF materials have chemical inertness, useful strain values, flexiblility, robustness, low cost
of raw materials and are easy to manipulate [25]. The poly methyl methacrystal (PMMA) or
acrylic is a type of PVDF, and is selected because of its many advantages compare to the silicon
[26]. The stress in the beam k2 is calculated using the following equation:

F l2 Y
σ= (3.7)
I
and
3EId
F = m2 × a ≈ F = (3.8)
l23
Where:
Y is t2 /2 in which t2 is the thickness of the beam
E is the Young’s modulus
I is the moment of inertia
F is the effective force du to the acceleration of 1g [22], which is the most dominant acceleration
of the human motion
d is the allowable deflrction
l2 is the length of the secondary beam
b2 is the width of the secondary beam
1
Joseph Kuan (2017). [Link]
Acceced 19-4-2020

23
t2 is the thickness of the secondary beam
For continuously imposed load, accordung to the manufacturer’s data sheet, the maximum
stress on the PMMA should not exceed 10.4M P a [26]. Once the maximum stress have been
set, the mass m2 and the stiffness k2 can be calculated using the following equation [26]:

6m2 al2
σ= (3.9)
b2 t2
√2
k2
ω22 = (3.10)
m2

Where ω = ω22 = 1.3Hz ≈ 8rad/s is the chosen main frequency. The parameters of the
2-DOF system are related by the following equations [26]:

k1 k2
ω12 × ω22 = × (3.11)
m1 m2
k1 k2 k2
ω12 + ω22 = + + (3.12)
m1 m2 m1

The dimensions are determined using the following equation [26]:

3EI
k= (3.13)
L3
Where E is the Young’s modulus, I is the area moment of inertia, and L is the length of
the beam.

3.2 Electrical model

3.2.1 Basic configuration

The proposed design derives the strain distribution along the piezoelectric beam to determine
the voltage and power produced. This approach was proposed and used by [19], [24].
Figure 2.5 shows the different mode of operation of piezoelectric material. In 33 mode both
the stress and the voltage act in z direction (or direction 3) while in 31 mode the voltage act
in direction 3 and the stress in direction 1 (or x direction). The 31 mode require a small input
force and work with low frequency [19]. The constitutive equations 2.1 and 2.2 can be rewrite
as:

24
 ξ1 = sE E E
11 T1 + s12 T2 + s13 T3 + d31 E3 (3.14)

ξ2 = sE E E
12 T1 + s11 T2 + s13 T3 + d31 E3 (3.15)

ξ3 = sE E
13 (T1 + T2 ) + s33 T3 + d31 E3 (3.16)

D3 = d31 (T1 + T2 ) + d33 T3 + ε0 εT33 E3 (3.17)

Ti , ξi and sij are the components of the stress, strain vectors and the components of
compliance at a constant electrical field respectively. E3 and D3 are the components in the z
direction of the electric field and the electric displacement respectively. ε0 εT33 and d31 respec-
tively represent the dielectric permitivity in the z-direction and the piezoelectric constant.
The change in the modal frequency of the beam in vibration creats two seanarios of the load
resistance, short-circuit when the load is zero and open circuit when the load tend to infinity.
The short-circuit current and the open circuit voltage are determined by respectively setting
E3 = 0 or D3 = 0 in equation 3.17.

3.2.2 Material properties and selection

The important properties considered for this study are the piezoelectric strain coefficient, the
Young’s modulus, the dielectric constant and the coupling coefficient. The coupling coefficient
tells the transduction ability of the material. It is determined using the following equation:

√
Y
K= d (3.18)
ε
Where:
K is the coupling coefficient
Y is the Young’s modulus
ε is the dielectric constant
d is the piezoelectric strain coefficient
Another important property is the tensile strength which gives the maximum strain that can
be supported by the material. From table 3.1, the PZT has a higher coupling coefficient, thus
can produce higher output voltage and power. However for reason alredy given, the prototype
will be made using PVDF.

25
 Table 3.1: Comparison of PZT and PVDF properties [19]
Property Units PZT PVDF
−12
Strain coefficient (d31 ) 10 m/v 320 20
Coupling coefficient (K31 ) CV /N m 0.44 0.11
Dielectric constant ε/ε0 3800 12
10 2
Young’s modulus 10 N /m 5.0 0.3
7 2
Tensile strength 10 N /m 2.0 5.2

3.2.3 Analytical model

The piezoelectric transduction scheme can be divided in two parts, the mechanical part and
the electric one. On the mechanical side, the variable across is stress and the variable through
is strain. The damper coefficient and the stifness link stress to strain. While voltage is the
variable across and current the variable through, on the electrical side. The coupling factor,
which liks the two part is represented as a transformer with an equivalent turns ratio n∗ = −dY
[19]. Figure 3.2 shows the representation of the system explained above.

Figure 3.2: Circuit representation of piezoelectric bimorph [19]

The input acceleration generate stress in the beam. The next equation shows the relation-
ship between the input vibration and the related stress input.

m
σin = ÿ (3.19)
b∗∗
Where b∗∗ is the geometric constant that relates the bending stress to force at the end of the
beam, and ÿ is the input acceleration. The Kirchoff’s Voltage Law (KVL) and the Kirchoff’s
Current Law (KCL) applied on this circuit (Figure 3.2) give the following system of equations:

26
 −k bm b∗∗ kd31
δ̈ = δ− δ̇ + V + b∗ ÿ (3.20)
m m mtc
−Y dtc
V̇ = δ̇ (3.21)
ε

Where:
V is the output voltage
tc is the thickness of the piezoelectric layer
k is the spring constant of the beam
b∗ is the geometric constant relating the average strain to displacement at the end of the beam.
Let suppose that a load is applied on the system as a resistance. The cricuit shown in
Figure 3.2 changes to the one of Figure 3.3, and equations 3.20 and 3.21 become 3.22 and 3.23.

−k bm b∗∗ kd31
δ̈ = δ− δ̇ + V + b∗ ÿ (3.22)
m m mtc
−Y dtc 1
V̇ = δ̇ − V (3.23)
ε RCp

Where R represents the load resisteance and Cp the capacitance of the piezoelectric bender.
From equations 3.14, 3.15, 3.16 and 3.17, the short-circuit current is calculated by setting
E = 0, and the open circuit voltage is calculated by setting D = 0. The results are shown
respectively in equation 3.24 and 3.25.

∫ lb
Is.c. = 2πwte fr e31 (1 − νb ) ξ(x)dx (3.24)
( ) 0
∫ lb
−e31 tp
Vo.c. = (1 − νb ) ξ(x)dx (3.25)
ε0 εT33 lb 0

ξ1
Where νb = ξ2
is the Poisson’s ratio of the beam, tb is the thickness of the beam, lb is
the lenght of the beam, fr is the period of the beam vibration, wte represent the rectangular
segment at the top of electrode and e31 is a constant given by:

d31
e31 = (3.26)
sE
11 + sE12

kd
The electrical coupling term represented by mtc
V can be used to calculate the linear damping

27
 Figure 3.3: Ciruit representation of a piezoelectric bimorph with load [19]

ratio.

ωk 2
ζe = √ (3.27)
1
ω2 + (RC)2

The damping ratio ζe can be made equal to the machanical damping ratio ζ by proper
selectrion of the load resistance. When the input frequency matches the frequency of the
piezoelectric harvester, the power output can be derived as follow:

( ∗ )2
1 (RC)2 Y dtb
P = 2 ε
A2 (3.28)
ω (4ζ 2 + k 4 )(RCω)2 + 4ζk 2 (RCω) + 2ζ 2 in
Ain represents the acceleration magnitude of the input vibration. By differentiating and
solving equation 3.28 in respect of R, the optimal load resistance can be expressed as:

1 2ζ
Ropt = √ (3.29)
ωC 4ζ 2 + K 4

K in equation 3.29 represents the piezoelectric coupling coefficient. For very low coupling
coefficient (K 4 ≈ 0), the optimal load resistance will be Ropt = 1/ωC. By simplifying the ciruit
in Fiqure 3.3 to the one of Figure 3.4, the current through the load and the power output
become:

Vo.c.
Il = (3.30)
(Zp + Zl )
2
1 Vo.c.
P = Zl (3.31)
2 (Zp + Zl )

28
 Figure 3.4: Equivalent circuit of a piezoelectric generator [24]

With Zp and Zl the piezoelectric complex impedance and the load impedance respectively.

3.3 Simulation

3.3.1 3D model

Solidworks software is very handy when it comes to create 3D model. The software allow also
the material selection, and determine the type of joint between components. All the cantilever
beams are chosen to be of Acrylic glass. The masses are chosen to be copper and the ground
stopper is alluminium. The software define the material properties.

3.3.2 Virtual test

The virtual test of the energy harvester is done with COMSOL MultiPhysics software. The test
is done by creating a finite element model of the energy harvester, using the following steps:

• Upload the 3D model of the device in COMSOL

• Select ”structural mechanics with piezoelectric devices application mode” from ”Add
physics”

• Select ”eigen frequency” from ”Studies”

• Define the variables as follow:

– Exciting frequency range is set from 1 to 2Hz

– Exciting acceleration is set for 1g

29
 – The bottom of the harvester (ground) is considered fixed

• Set up physics

• Define the finite element meshing

• Simulate

The simulation results will give different beam displacement corresponding to the resonant
frequency and the produced electric voltage corresponding to beam length and thickness. Since
the up-convertion system frequency is in the range of human motion frequencies (≈ 6Hz),
the system should also respond to excitation from human motion vibrations. To confirm this
affirmation, the system will also be tested from 1 to 10Hz as excitation frequency range.

3.4 Device fabrication and experimental test

The prototype is made with PMMA. The cutting of the PMMA is made using a laser machine.
The different components (masses, beams and ground stopper) are screwed. The harvester will
be fixed on supporting frame to facilitate the experiment. The experimental test is important
to validate the simulated results. The needed equipments for the experimental test are:

• A signal generator (or function generator) which will power the source of vibration.

• A shaker to generate vibrations.

• An accelerometer fixed on the shaker to mesure the vibrations.

• Two power amplifiers, one for the shaker and another for the accelerometer.

• A holding frame to help fixing the harvester on the shaker.

• A digital oscilloscope to collect the data. The data collected are the acceleration mesured
by the accelerometer and the output voltage generated by the piezoelectric patch.

• A flash memory to save the data from the oscilloscope.

• A computer with Lab view software installed to analyse the data collected.

Figure 3.5 shows the shematic diagram of the experimental setup.

30
Figure 3.5: Schematic diagram of the experimental setup

31
Chapter 4

Expected outcome and schedule

4.1 Expected outcome

The result of this investigation must give an energy harvester able to supply low power elec-
tronics that can be near the heart such as peacemaker. The frequency range consider is from 1
to 2Hz in order to human heart beat vibration frequencies. Since the up-convertion system fre-
quency is in the range of human motion frequencies (≈ 6Hz), the system should also respond
to excitation from human motion vibrations. Thus system can react to a wide bandwudth
within the rang of 1 to 10Hz. However the system will be tested to confirm this hability. The
expected output must be in the order of a few milliwatt. The 2-DOF offers a optimum use of
the volume while the up-conversion increase the output power. The power will be generated
by three piezoelectric beams, each reacting to different resonant frequency. This will provide a
better power density. The ideal will be to fabricate a micro-scale prototype, but a macro-scale
prototype will also bring satisfaction in order to validate the simulated results by comparison
with the real model tests results.

4.2 Budget

4.3 Source of Funds

The funds for this research will be provided by Pan African University Institute for Basic
Science, Technology and Innovation.

32
 Table 4.1: Budget of the research project
Item Unit Quantity Unit Cost (USD) Total Cost (USD)
Laptop Piece 1 Available Available
Solidworks software Package 1 Available Available
COMSOL software Package 1 Available Available
Flash disk Number 1 20 25
Publication Number 1 500 500
Printing Papers Ream 4 10 40
Internet & Communication Months 12 50 600
Conference - - - 1000
Materials & Equipments - - - 2000
Total - - - 4165

4.4 Work Schedule

Table 4.2: Work Schedule

Period
2019 2020 2021
Activity Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr
Literature Review
Proposal Development
Submission of Proposal
Report Writing
Harvester Design
Simulation
Fabrication
Testing
Data acquisition
Publication
Conference
Interim presentation
Final Correction
Thesis Submission

33
References

[1] T. J. Kazmierski and S. Beeby, Energy harvesting systems. Springer, 2014.

[2] S. Beeby and N. M. White, Energy harvesting for autonomous systems. Artech House,
2010.

[3] D. Zhu, “Vibration energy harvesting: Machinery vibration, human movement and flow
induced vibration,” Sustainable Energy Harvesting Technologies-Past, Present and Future,
vol. 1, pp. 22–54, 2011.

[4] D. Zhu and S. Beeby, “Kinetic energy harvesting,” in Energy harvesting systems, Springer,
2011, pp. 1–77.

[5] B. Maamer, A. Boughamoura, A. M. F. El-Bab, L. A. Francis, and F. Tounsi, “A review on

design improvements and techniques for mechanical energy harvesting using piezoelectric
and electromagnetic schemes,” Energy Conversion and Management, vol. 199, 2019.

[6] L. Bu, X. Wu, X. Wang, and L. Liu, “Non-resonant electrostatic energy harvester for
wideband applications,” Micro & Nano Letters, vol. 8, no. 3, pp. 135–137, 2013.

[7] M. Colin, S. Basrour, and L. Rufer, “Design, fabrication and characterization of a very low
frequency piezoelectric energy harvester designed for heart beat vibration scavenging,” in
Smart Sensors, Actuators, and MEMS VI, International Society for Optics and Photonics,
vol. 8763, 2013.

[8] M. N. Andanje, B. W. Ikua, and A. M. F. Elbab, “Piezo-electric energy harvesting from

human motion based on 2-dof vibration absorber mode: Design methodology and exper-
imental validation,” International Journal of Mechanical & Mechatronics Engineering
IJMME-IJENS, vol. 19, no. 04, 2019.

[9] M. A. Halim and J. Y. Park, “A non-resonant, frequency up-converted electromagnetic

energy harvester from human-body-induced vibration for hand-held smart system appli-
cations,” Journal of Applied Physics, vol. 115, no. 9, 2014.

34
[10] C. Saha, T. O’donnell, N. Wang, and P. McCloskey, “Electromagnetic generator for har-
vesting energy from human motion,” Sensors and Actuators A: Physical, vol. 147, no. 1,
pp. 248–253, 2008.

[11] M. A. Halim, H. Cho, and J. Y. Park, “Design and experiment of a human-limb driven,
frequency up-converted electromagnetic energy harvester,” Energy Conversion and Man-
agement, vol. 106, pp. 393–404, 2015.

[12] H. Liu, C. J. Tay, C. Quan, T. Kobayashi, and C. Lee, “A scrape-through piezoelec-

tric mems energy harvester with frequency broadband and up-conversion behaviors,”
Microsystem technologies, vol. 17, no. 12, pp. 1747–1754, 2011.

[13] L. Gu, “Low-frequency piezoelectric energy harvesting prototype suitable for the mems
implementation,” Microelectronics Journal, vol. 42, no. 2, pp. 277–282, 2011.

[14] H. Liu, C. Lee, T. Kobayashi, C. J. Tay, and C. Quan, “Piezoelectric mems-based wide-
band energy harvesting systems using a frequency-up-conversion cantilever stopper,” Sen-
sors and Actuators A: Physical, vol. 186, pp. 242–248, 2012.

[15] P. D. Mitcheson, E. M. Yeatman, G. K. Rao, A. S. Holmes, and T. C. Green, “Energy

harvesting from human and machine motion for wireless electronic devices,” Proceedings
of the IEEE, vol. 96, no. 9, pp. 1457–1486, 2008.

[16] S. Ju, S. Chae, Y. Choi, S. Jun, S. Park, S. Lee, H. Lee, and C. Ji, “Frequency up-
converted low frequency vibration energy harvester using trampoline effect,” in Journal
of Physics: Conference Series, IOP Publishing, vol. 476, 2013.

[17] M. Amin Karami and D. J. Inman, “Powering pacemakers from heartbeat vibrations
using linear and nonlinear energy harvesters,” Applied Physics Letters, vol. 100, no. 4,
2012.

[18] M. Soliman, “Wideband micro-power generators for vibration energy harvesting,” Ph.D.
dissertation, University of Waterloo, 2009.

[19] S. J. Roundy, “Energy scavenging for wireless sensor nodes with a focus on vibration
to electricity conversion,” Ph.D. dissertation, University of California, Berkeley Berkeley,
CA, 2003.

[20] N. Elvin and A. Erturk, Advances in energy harvesting methods. Springer Science &
Business Media, 2013.

[21] J. P. Den Hartog, Mechanical vibrations. Courier Corporation, 1985.

35
[22] M. M. Magdy, A. M. F. El-Bab, and S. F. Assal, “Design methodology of a micro-scale
2-dof energy harvesting device for low frequency and wide bandwidth,” Journal of Sensor
Technology, vol. 4, pp. 37–48, 2014.

[23] A. Nammari, L. Caskey, J. Negrete, and H. Bardaweel, “Fabrication and characterization

of non-resonant magneto-mechanical low-frequency vibration energy harvester,” Mechan-
ical Systems and Signal Processing, vol. 102, pp. 298–311, 2018.

[24] T. Kamel, R. Elfrink, M. Renaud, D. Hohlfeld, M. Goedbloed, C. De Nooijer, M. Jam-

bunathan, and R. Van Schaijk, “Modeling and characterization of mems-based piezo-
electric harvesting devices,” Journal of Micromechanics and Microengineering, vol. 20,
no. 10, 2010.

[25] M. L. Thompson, “On the material properties and constitutive equations of piezoelectric
poly vinylidene fluoride (pvdf),” Ph.D. dissertation, Drexel University, 2002.

[26] N. A. Maurine, “Development of a 2-dof piezoelectric energy harvester prototype based

on human motion,” M.S. thesis, PAN African University Institute for Basic Sciences,
Technology and Innovation, 2019.

36