energies

Article
A Study on the Power Generation Capacity of
Piezoelectric Energy Harvesters with Different
Fixation Modes and Adjustment Methods
Zhixiang Li 1,2, , Gongbo Zhou 1,2,, *, Zhencai Zhu 1,2 and Wei Li 1,2
1 School of Mechanical and Electrical Engineering, China University of Mining & Technology, Xuzhou 221116,
Jiangsu, China; lixiang3700@sina.com (Z.L.); zhuzhencai@vip.163.com (Z.Z.); liwei_cmee@163.com (W.L.)
2 Jiangsu Key Laboratory of Mine Mechanical and Electrical Equipment,
China University of Mining & Technology, Xuzhou 221116, Jiangsu, China
* Correspondence: zhougongbo1985@163.com
These authors contributed equally to this work.

Academic Editor: Ling Bing Kong

Received: 8 November 2015; Accepted: 27 January 2016; Published: 19 February 2016

Abstract: The power generation capacity of piezoelectric energy harvesters (PEHs) is not only
related to the properties of the piezoelectric material, the vibration magnitude and the subsequent
conditioning circuit, but also to the fixation modes and adjustment methods. In this paper,
a commercial piezoelectric ceramic plate (PCP) in simply supported beam fixation mode and
cantilever beam fixation mode were analyzed through finite element simulations and experiments,
and furthermore, two ways of adjusting the natural frequency of PCP are studied and compared.
As a result, some guidelines are proposed for the application of PCPs according to the simulation and
experimental results which showed that: (1) the simply supported beam fixation mode is suitable
for environments in which the exciting frequency exceeds 50 Hz, while the cantilever beam fixation
mode fits the circumstance where the exciting frequency is below 50 Hz; (2) the maximum generation
power a PCP produces in simply supported beam fixation mode is larger than that in cantilever beam
fixation mode; (3) adjusting the weight of the mass block affixed on the PCP can change the natural
frequency of PCP more efficiently than length-width ratio does.

Keywords: wireless sensor node; piezoelectric energy harvester; fixation modes; adjusting methods;
finite element simulation

1. Introduction
In recent years, the fast development of microelectronics and wireless communication technology
has promoted the wide use of wireless sensor networks (WSNs), however, batteries are needed to
power the wireless sensor nodes, and must be recharged or replaced when their capacity is low,
which is a prominent drawback in the development of WSNs. Many scholars have prosed ways to
scavenge energy from the ambient environment to powerwireless sensor nodes, which represents
a reasonable solution to the problem. Nowadays, studies on ambient energy scavenging technology are
developing very fast [1]. These energy capture technologies include solar energy capture technology,
wind energy capture technology, electromagnetic energy capture technology and vibration energy
capture technology, etc. [2]. Among these technologies, solar energy capture and wind energy capture
technologies have been studied for decades, and great developments have been achieved [3,4], however,
these two important energy harvesting methods require certain conditions (i.e., sunlight or wind must
exist), besides, the volume of devices used by these two energy harvesting methods are relatively

Energies 2016, 9, 98; doi:10.3390/en9020098 www.mdpi.com/journal/energies

Energies 2016, 9, 98 2 of 14

large in size, causing inconvenience in some application areas, such as mine data acquisition and
other dark sealed places. Electromagnetic energy is spread almost everywhere around the world, but
its energy density is rather low and unstable [5,6]. Vibration energy harvesting is a fast developing
field due to the wide availability of vibration sources and various ways of harvesting them, and
relatively high conversion efficiency [7]. Nowadays, three methods to collect the vibration energy
are considered to be most promising, namely, the capacitive method, the electromagnetic induction
method and the piezoelectric method [8]. A comprehensive comparison of these three different
methods suggests that the piezoelectric method is simple, the devices are easy to manufacture and
they have high output voltages, which makes it the most reasonable choice in the vibration energy
harvesting field [9]. So far, many scholars have done a lot of work and made incredible achievements
in the piezoelectric energy harvesting field. As early as 2001, Shenck et al. [10] from the Massachusetts
Institute of Technology successfully developed electrical power generating sports shoes utilizing
a piezoelectric plate. Ottman [11] studied the performance of a variety of subsequent conditioning
circuits for piezoelectric energy harvesters. In 2010, the Linear Technology Company designed and
produced the integrated LTC3588 chip, which is specially designed for micro energy collection in
the ambient environment. Zhou et al. [1214] abandoned the conventional way of using mechanical
vibrations as the excitation source and adopted wind or a flowing fluid to excite PCPs, which expands
the PCP application field. Xu et al. [1517] adjusted the structure aspects of the PCP in order to
maximize the power generation capacity. Abdelkefi et al. [18] studied through theoretical analysis
and finite element method simulations a special electromechanical model in which the piezoelectric
sheet does not cover the whole substrate beam. By quantifying the compromise between the length
of the piezoelectric material, the load resistance, and the excitation frequency, the performance of
beam-mass system was optimized. Abdelkefi et al. also designed a new structure which can produce
both bending and torsion in PCPs to enhance their power generation, and the results showed that
the harvested power can be increased by as much as 30% compared with sedigns where beam only
undergoes bending [19]. Sharpes et al. [20,21] created a totally new zigzag-dimensional beam with
a low volume and compared its power generation capacity with that of a conventional beam. It turned
out that the new zigzag-dimensional beam can provide a significant increase in power. Tang and
Abdelkefi et al. [2224] proposed and studied a new PCP structure with a magnet. With the help of
the magnet, a broad operating bandwidth of the energy harvester was studied, especially in low
frequency. As shown above, no matter what the excitation source is, mechanic vibration or fluid
vibration, cantilever beams for PCPs are the most commonly used fixation mode, while the simply
supported beam fixation mode is rarely adopted. Gafforelli et al. [25] displayed a bridge-shaped
energy harvest which actually is a PCP in simply supported beam fixation mode, but they did not
study the influence(s) of the mass block and length-width ratio on the power generation of the PCP in
simply supported beam fixation mode, and no comparisons between bridge-shaped PCPs and ordinary
PCPs in cantilever beam fixation mode were discussed. Besides, adding a mass block on a PCP is
often adopted to change its natural frequency, but other methods such as changing the length-width
ratio of PCP are scarcely reported. Both the simply supported beam fixation and the influence of
length-width ratio have not been comprehensively studied, so in this paper finite element simulations
and experiments were performed to figure out how the fixation mode and the adjustment method
influence the power generation capacity of a PCP.

2. Model Parameters
A commercial PCP was adopted as the simulation model. As shown in Figure 1, this PCP is
composed of a copper plate and two piezoelectric patches. The two piezoelectric patches are affixed
on the two sides of the copper plate via glue, making the PCP produce twice the quantity of electrical
power as the design with only one piezoelectric patch does, thus the PCP parallel connection method
is often selected to produce large currents in practical applications.
Energies 2016, 9, 98 3 of 14

The parameters of each material in the PCP are shown in the Table 1. The copper plate material is
copper CW617N, and the mass block is nickel. The material parameters of the copper plate and mass
block include Youngs modulus, Poisson ratio and density, and the material parameters of piezoelectric
patchEnergies2016,9,98
include elastic coefficient matrix, piezoelectric matrix and dielectric constant matrix [26].3of14
Element
solid45 is taken to mesh the copper plate and mass block in ANSYS, meanwhile, coupled field element
matrix[26].Elementsolid45istakentomeshthecopperplateandmassblockinANSYS,meanwhile,
solid5coupledfieldelementsolid5isselectedtomeshthepiezoelectricpatch.
is selected to mesh the piezoelectric patch.

Figure 1. Structure of PCP.
Figure1.StructureofPCP.

Table 1. Parameters of the device materials.

Table1.Parametersofthedevicematerials.

Parameter CopperPlate PiezoelectricPatch MassBlock

Parameter Copper Plate Piezoelectric Patch Mass Block
Youngsmodulus(GPa) 110 180
Youngs modulus (GPa)
Poissonratio 110
0.324 180
0.3
Poisson ratio
Density(kg/m 3) 0.324
8300 0.3
8900
Density (kg/m3 ) 8300 8900
Elasticcoefficientmatrix D
Elastic coefficient matrix D
Piezoelectricmatrix e
Piezoelectric matrix e
Dielectricconstantmatrix
Dielectric constant matrix

InTable1,D,e,iselasticcoefficientmatrix,piezoelectricmatrixanddielectricconstantmatrix
In Table 1, D, e, is elastic coefficient matrix, piezoelectric matrix and dielectric constant matrix
respectivelyandtheirvaluesareasfollows:
13.2 their
respectively and 7.1 7.31
values are as follows: 0 0 5.2
0
13.2 7.3 fi 00 5.2 1100
5.2 .
fi
13.2 7.1 7.31 11.5 10 0 1100
0 0 15.1
D 13.2 7.3 ffi 10 Pa, e 0 0 (C/m
5.2
2 ffi
),
827

3.0
fi


ffi
0 0 0 ffi 1100

11.5 ffi
ffi 0
0 15.1 ffi
ffi
D 2.6 ffi 1010 Pa, e0 12.7 0 ffi pC/m2 q, 1100 fl .
ffi
3.0 ffi 0 0 0 ffi
2.6
fl 12.7
0
0 0 827
12.7 0 fl
ffi ffi
2.6
2.6 12.7 0 0
Inthesimulationpart,modalanalysiswascarriedoutfirsttogeteachordernaturalfrequency
ofthePCP,andthenharmonicresponseanalysiswasadoptedtogetthevoltagesgeneratedbythe
In the simulation part, modal analysis was carried out first to get each order natural frequency
PCPundertheharmonicforce.Beforestartingtheharmonicresponseanalysis,theharmonicforce
of theand
PCP, and then
frequency harmonic
range need toresponse analysis
be defined. was adopted
The harmonic to get
force was the0.5
set to voltages generated
N in every by the
harmonic
PCP under the harmonic force. Before starting the harmonic response analysis, the harmonic force
responseanalysisandfrequencyrangewaskeepingnearthenaturalfrequencyofPCPacquiredin
and frequency range need to be defined. The harmonic force was set to 0.5 N in every harmonic
modalanalysis.Withthehelpofharmonicresponseanalysis,theinformationofstress,displacement
response analysis and frequency range was keeping near the natural frequency of PCP acquired in
andvoltage,etc.atanypositionofthePCPcanbeobtainedaftercalculation.Voltagesonthetopside
modal ofthepiezoelectricpatcharerecordedduringeachharmonicresponseanalysisinthispaper.The
analysis. With the help of harmonic response analysis, the information of stress, displacement
processofacquiringvoltageisasfollows:first,eachsideofthepiezoelectricpatchiscoupledtobe
and voltage, etc. at any position of the PCP can be obtained after calculation. Voltages on the top side of
completelyinvoltage.Next,thevoltageonthebottomsideissetto0VasGND.Finally,thevoltage
the piezoelectric patch are recorded during each harmonic response analysis in this paper. The process
onthetopsideisextractedviacommandNSOL(acommandstreaminANSYS)aftercalculationand
of acquiring voltage is as follows: first, each side of the piezoelectric patch is coupled to be completely
displayedversusexcitingfrequency.
in voltage. Next, the voltage on the bottom side is set to 0 V as GND. Finally, the voltage on the top
side is3.SimulationofSimplySupportedBeamFixationMode
extracted via command NSOL (a command stream in ANSYS) after calculation and displayed
versus exciting frequency.
For the convenience of simulations and experiments, only one end of the PCP in simply
supported beam fixation mode is fully restricted, and the other end is restricted in the Y and Z
3. Simulation of Simply Supported Beam Fixation Mode
directionsinordertoproducelargerhorizontalmovementswhenthePCPdeforms.Figure2displays
For the convenience of simulations and experiments, only one end of the PCP in simply supported
themodelswitchofthesimplysupportedbeammodel.
Twoadjustmentmethods,massblockaffixedonthePCPandlengthwidthratioofthePCP,are
beam fixation mode is fully restricted, and the other end is restricted in the Y and Z directions in order
analyzedinoursimulationundersimplysupportedbeamfixationmode.Inthesimulationofsimply
to produce larger horizontal movements when the PCP deforms. Figure 2 displays the model switch
of thesupportedbeamfixationmode,thesizeofthepiezoelectricpatchis50mm50mm,anditremains
simply supported beam model.
constantduringthesimulationprocess.First,theweightofthemassblockaffixedonthecenterof
Energies 2016, 9, 98 4 of 14

Two adjustment methods, mass block affixed on the PCP and length-width ratio of the PCP, are
analyzed in our simulation under simply supported beam fixation mode. In the simulation of simply
supported beam fixation mode, the size of the piezoelectric patch is 50 mm 50 mm, and it remains
Energies2016,9,98
Energies2016,9,98 4of14
constant during the simulation process. First, the weight of the mass block affixed on the4of14 center of
PCP works as the variable to study the influence of the mass block. Then, the ength-width ratio
PCPworksasthevariabletostudytheinfluenceofthemassblock.Then,theengthwidthratioofthe of the
PCPworksasthevariabletostudytheinfluenceofthemassblock.Then,theengthwidthratioofthe
piezoelectric patch is taken as variable to study its influence, and during this step, no mass block is
piezoelectricpatchistakenasvariabletostudyitsinfluence,andduringthisstep,nomassblockis
piezoelectricpatchistakenasvariabletostudyitsinfluence,andduringthisstep,nomassblockis
affixed on the PCP. Each order natural frequency of PCP can be acquired through modal analysis and
affixedonthePCP.EachordernaturalfrequencyofPCPcanbeacquiredthroughmodalanalysisand
affixedonthePCP.EachordernaturalfrequencyofPCPcanbeacquiredthroughmodalanalysisand
the first
theorder
the first was
first order
order selected
was as the natural
was selected
selected as thefrequency
as the natural of PCP. Maximum
natural frequency
frequency of PCP.
of output voltages
PCP. Maximum
Maximum corresponding
output
output voltages
voltages
correspondingtospecificweightsofmassblockorlengthwidthratioscanbeobtainedviaharmonic
to specific weights of mass block or length-width ratios can be obtained via harmonic response analysis.
correspondingtospecificweightsofmassblockorlengthwidthratioscanbeobtainedviaharmonic
responseanalysis.TheharmonicforceexertedonthenodelocatedinthecenterofthePCPis0.5N
responseanalysis.TheharmonicforceexertedonthenodelocatedinthecenterofthePCPis0.5N
The harmonic force exerted on the node located in the center of the PCP is 0.5 N in the harmonic
intheharmonicresponseanalysisandexcitingfrequencyrangeissetnearthenaturalfrequencyof
intheharmonicresponseanalysisandexcitingfrequencyrangeissetnearthenaturalfrequencyof
response analysis and exciting frequency range is set near the natural frequency of PCP gained in
PCPgainedinmodalanalysis.Figure3showsthemodelsaftermeshinganddefiningconstraintsin
PCPgainedinmodalanalysis.Figure3showsthemodelsaftermeshinganddefiningconstraintsin
modal analysis. Figure 3 shows the models after meshing and defining constraints in simply supported
simplysupportedbeamfixationmode.
simplysupportedbeamfixationmode.
beam fixation mode.

YY YY

XX XX

Figure 2. Model switch of simply supported beam model.
Figure2.Modelswitchofsimplysupportedbeammodel.
Figure2.Modelswitchofsimplysupportedbeammodel.

(a)
(a) (b)
(b)
Figure3.Modelsaftermeshinganddefiningconstraintsinsimplysupportedbeamfixationmode:
Figure3.Modelsaftermeshinganddefiningconstraintsinsimplysupportedbeamfixationmode:
3. Models after meshing and defining constraints in simply supported beam fixation mode:
Figure(a)Withmassblock;(b)Withoutmassblock.
(a)Withmassblock;(b)Withoutmassblock.
(a) With mass block; (b) Without mass block.
FromFigure4,itisobviousthatbothofthenaturalfrequencyofPCPandexcitingfrequencyat
FromFigure4,itisobviousthatbothofthenaturalfrequencyofPCPandexcitingfrequencyat
maximum
maximum
From Figurevoltage
voltage decrease
4, it decrease
is obvious with
with the
the
that increase
increase
both ofweight
of
of the weightof
natural of massblock.
mass
frequencyblock. As
ofAsPCP shown
shown
and inin Figurefrequency
Figure
exciting 5,the
5, the
naturalfrequencyofthePCPandtheexcitationfrequencyatmaximumvoltagealsodroprapidly
naturalfrequencyofthePCPandtheexcitationfrequencyatmaximumvoltagealsodroprapidly
at maximum voltage decrease with the increase of weight of mass block. As shown in Figure 5, the
withthe
with theincrease
increaseofoflengthwidth
lengthwidthratio
ratiowhen
whenthe
thelengthwidth
lengthwidthratio
ratio isbelow
below1,
1,and
andflattens
flattensout
out
natural frequency of the PCP and the excitation frequency at maximumisvoltage also drop rapidly with
graduallyifthelengthwidthismorethan1.Figures4and5indicatethattheexcitingfrequenciesat
graduallyifthelengthwidthismorethan1.Figures4and5indicatethattheexcitingfrequenciesat
the increase of length-width
maximumvoltage
voltageof
ratio
ofeach
eachPCP
when the
PCPaffixed
length-width
affixedwith
withdifferent
ratio
differentmass
is below
massblock
1,
blockcoincide
and flattens
coincidewell
wellwith
out
withthe
gradually
thenatural
natural
if
maximum
the length-width is more than 1. Figures 4 and 5 indicate that the exciting frequencies at maximum
frequenciesofthePCPs.Therefore,theexcitingfrequencyshouldbesetnearbythenaturalfrequency
frequenciesofthePCPs.Therefore,theexcitingfrequencyshouldbesetnearbythenaturalfrequency
voltage of each PCP affixed with different mass block coincide well with the natural frequencies of
ofthePCPtoobtainmaximumvoltage.Figure6demonstratesthatwiththeriseofweightofthemass
ofthePCPtoobtainmaximumvoltage.Figure6demonstratesthatwiththeriseofweightofthemass
the PCPs. Therefore, the exciting frequency should be set nearby the natural frequency of the PCP
blockaffixedonthePCP,themaximumvoltagekeepsfalling.Conversely,maximumvoltagegoes
blockaffixedonthePCP,themaximumvoltagekeepsfalling.Conversely,maximumvoltagegoes
upwiththeincreaseoflengthwidthratioofPCP.Bycomparingtheinfluencesofweightofmass
to obtain maximum voltage. Figure 6 demonstrates that with the rise of weight of the mass block
upwiththeincreaseoflengthwidthratioofPCP.Bycomparingtheinfluencesofweightofmass
affixedblockandlengthwidthratioonpowergenerationperformanceofPCPinsimplysupportedbeam
blockandlengthwidthratioonpowergenerationperformanceofPCPinsimplysupportedbeam
on the PCP, the maximum voltage keeps falling. Conversely, maximum voltage goes up with
fixationmode,wecanconcludethat:(1)ItismoreeffectivetochangelengthwidthratioofthePCP
fixationmode,wecanconcludethat:(1)ItismoreeffectivetochangelengthwidthratioofthePCP
the increase of length-width ratio of PCP. By comparing the influences of weight of mass block and
thanweightofthemassblockforadjustingthenaturalfrequencyofthePCPinsimplysupported
thanweightofthemassblockforadjustingthenaturalfrequencyofthePCPinsimplysupported
length-width ratio on power generation performance of PCP in simply supported beam fixation mode,
beamfixation
beam fixationmode;
mode; (2)A Awider
widerrange
rangeof
ofgenerated
generatedvoltage
voltagecan
canbe
be obtainedby bychanging
changingthethe
we can conclude that: (1) It(2)
is more effective to change length-width ratioobtained
of the PCP than weight of the
lengthwidthratiocomparedwithalteringtheweightofmassblock.
lengthwidthratiocomparedwithalteringtheweightofmassblock.
mass block for adjusting the natural frequency of the PCP in simply supported beam fixation mode;
Energies 2016, 9, 98 5 of 14

(2) A wider range of generated voltage can be obtained by changing the length-width ratio compared
with altering the weight of mass block.
Energies2016,9,98 5of14
Energies2016,9,98 5of14

Natural frequency
Natural frequency
Energies2016,9,98 Frequency to maximum
maximum voltage
voltage 5of14
450
450 Frequency to

Natural frequency
400
400 Frequency to maximum voltage
(Hz)
450
(Hz) (Hz)
350
Frequency
350
400
Frequency

300
300
350
Frequency

250
300
250

200
250
200
00 55 10
10 15
15 20
20 25
25 30
30 35
35
Weight of
Weight of mass
mass block
block (g)
(g)
200
0 5 10 15 20 25 30 35
Weight of mass block
frequency (g) and exciting frequency
Figure4.WeightofmassblockversusnaturalfrequencyofPCPandexcitingfrequencyatmaximum
4. Weight of mass block versus natural of PCP at maximum
FigureFigure4.WeightofmassblockversusnaturalfrequencyofPCPandexcitingfrequencyatmaximum

voltageinsimplysupportedbeamfixationmode.
voltage in simply supported beam fixation mode.
voltageinsimplysupportedbeamfixationmode.
Figure4.WeightofmassblockversusnaturalfrequencyofPCPandexcitingfrequencyatmaximum
voltageinsimplysupportedbeamfixationmode.
3000
3000 Natural frequency
Natural frequency
Frequency
Frequency toto maximum
maximum voltage
voltage
3000
2500
2500 Natural frequency
Frequency to maximum voltage
2500
2000
(Hz)

2000
(Hz)(Hz)

2000
1500
Frequency

1500
Frequency

1500
Frequency

1000
1000
1000
500
500
500
00
0 11 22 33 44 55 66 77
00
0 1 2 3 4
Length-width 5
ratio 6 7
Length-width ratio
Length-width ratio
Figure5.Lengthwidth
Figure 5.Lengthwidthratioversus
ratioversusnatural
naturalfrequency
frequencyofPCP
ofPCPand
andexciting
excitingfrequency
frequencyat
atmaximum
maximum
5. Length-width ratio versus natural
naturalfrequency
Figurevoltageinsimplysupportedbeamfixationmode.
Figure 5.Lengthwidth ratioversus of PCP
frequency ofPCP and
and exciting
exciting frequency
frequency at maximum
at maximum
voltageinsimplysupportedbeamfixationmode.
voltagevoltageinsimplysupportedbeamfixationmode.
in simply supported beam fixation mode.
Length-width ratio
Length-widthratio
ratio
0 1 2 Length-width
3 4 5 6 7 8 9 10
00 11 2
2
3
3
4
4 5
5 6
6 7
7 8
8 9 9 10 10
20.0
6.0 20.0
20.0
6.0
6.0 17.5
(V)

(V)

17.5
(V)

17.5
voltage(V)
voltage(V)

Maximum voltage (V)

5.5
5.5
5.5 15.0
15.0
15.0
Maximumvoltage

Maximumvoltage
voltage

5.0 12.5
5.0
5.0 12.5
12.5
10.0
4.5 10.0
10.0
4.5
4.5
Maximum
Maximum

Maximum

7.5
7.57.5
4.0
4.0 5.0
4.0 5.05.0
2.5
2.52.5
3.5
3.5
3.5
0.0
0.00.0
0 5 10 15 2020 2525 30 30 35 35
0 55 10
10 15
15 20 25 30 35
Weight
Weightof
Weight ofmass
of massblock
mass block(g)
block (g)
(g)
Figure6.Weightofmassblockandlengthwidthratioversusmaximumvoltageinsimplysupported
Figure6.Weightofmassblockandlengthwidthratioversusmaximumvoltageinsimplysupported
Weight of mass block and length-width ratio versus maximum voltage in simply supported
FigureFigure6.Weightofmassblockandlengthwidthratioversusmaximumvoltageinsimplysupported
6.
beamfixationmode.
beamfixationmode.
beam beamfixationmode.
fixation mode.
Energies 2016, 9, 98 6 of 14

Energies2016,9,98 6of14
4. Simulation of Cantilever Beam Fixation Mode
4.SimulationofCantileverBeamFixationMode
The cantilever beam is a frequently used method compared with simply supported beam
Energies2016,9,98
as only
6of14
one endThecantileverbeamisafrequentlyusedmethodcomparedwithsimplysupportedbeamasonly
of it needs to be fixed; however, it places high demands on the material properties of the beam
oneendofitneedstobefixed;however,itplaceshighdemandsonthematerialpropertiesofthe
as it usually has a large deformation. In the simulation of simply supported beam fixation mode, the
4.SimulationofCantileverBeamFixationMode
beam as it usually has a large deformation. In the simulation of simply supported beam fixation
majority Thecantileverbeamisafrequentlyusedmethodcomparedwithsimplysupportedbeamasonly
of the settings are consistent with the settings in a simply supported beam fixation mode,
mode,themajorityofthesettingsareconsistentwiththesettingsinasimplysupportedbeamfixation
except that the mounting position of the mass block is adjusted to the unfixed end of PCP, and the
oneendofitneedstobefixed;however,itplaceshighdemandsonthematerialpropertiesofthe
mode,exceptthatthemountingpositionofthemassblockisadjustedtotheunfixedendofPCP,and
harmonic force
beam as of 0.5 N
it usually isaimposed
has on a node
large deformation. Infrom the edge of
the simulation center of supported
simply the unfixed endfixation
beam of the PCP.
theharmonicforceof0.5NisimposedonanodefromtheedgecenteroftheunfixedendofthePCP.
Figure 7 shows the models after meshing and defining constraints in cantilever beam fixation
mode,themajorityofthesettingsareconsistentwiththesettingsinasimplysupportedbeamfixation
Figure7showsthemodelsaftermeshinganddefiningconstraintsincantileverbeamfixationmode. mode.
mode,exceptthatthemountingpositionofthemassblockisadjustedtotheunfixedendofPCP,and
theharmonicforceof0.5NisimposedonanodefromtheedgecenteroftheunfixedendofthePCP.
Figure7showsthemodelsaftermeshinganddefiningconstraintsincantileverbeamfixationmode.

(a) (b)
Figure 7. Models after meshing and defining constraints in cantilever
(a) meshing beam fixation mode:
(b)
Figure 7. Models after and defining constraints in cantilever beam fixation mode:
(a)Withmassblock;(b)Withoutmassblock.
(a) With mass block; (b) Without mass block.
Figure 7. Models after meshing and defining constraints in cantilever beam fixation mode:
(a)Withmassblock;(b)Withoutmassblock.
Figure 8 indicates similar results that both of the natural frequency of PCP and excitation
Figure 8 indicates similar results that both of the natural frequency of PCP and excitation frequency
frequencyatmaximumvoltagedecreasewiththeincreaseofweightofthemassblock.Asrevealed
Figure 8 indicates similar results that both of the natural frequency of PCP and excitation
at maximum voltage decrease with the increase of weight of the mass block. As revealed in Figure 9,
inFigure9,thenaturalfrequencyofthePCPandtheexcitationfrequencyatmaximumvoltagedrop
frequencyatmaximumvoltagedecreasewiththeincreaseofweightofthemassblock.Asrevealed
therapidlywiththeincreaseoflengthwidthratiowhenthelengthwidthratioisbelow4,andgradually
natural frequency of the PCP and the excitation frequency at maximum voltage drop rapidly
inFigure9,thenaturalfrequencyofthePCPandtheexcitationfrequencyatmaximumvoltagedrop
reach
with the a plateauof
increase when the lengthwidth
length-width ratio whenratiothe
is greater than 4.ratio
length-width Figures 8 and4,
is below 9 and
illustrate
rapidlywiththeincreaseoflengthwidthratiowhenthelengthwidthratioisbelow4,andgradually
the same
gradually reach
conclusionthattheexcitingfrequenciesatmaximumvoltageofeachPCPaccordwellwiththenatural
a plateau when the length-width ratio is greater than 4. Figures 8 and 9 illustrate the same conclusion
reach a plateau when the lengthwidth ratio is greater than 4. Figures 8 and 9 illustrate the same
frequenciesofthePCPs.
that the exciting frequencies at maximum voltage of each PCP accord well with the natural frequencies
conclusionthattheexcitingfrequenciesatmaximumvoltageofeachPCPaccordwellwiththenatural
of thefrequenciesofthePCPs.
PCPs.
90
Natural frequency
90 Frequency to maximum voltage
80 Natural frequency
Frequency to maximum voltage
80
70
(Hz)

70
60
(Hz)

60
Frequency

50
Frequency

50
40
40
30
30
20
200 5 10 15 20 25 30 35
0 5 10 15 20 25 30 35
Weight of mass
massblock(g)
block(g)
Weight of
Figure8.WeightofmassblockversusnaturalfrequencyofPCPandexcitingfrequencyatmaximum
Figure8.WeightofmassblockversusnaturalfrequencyofPCPandexcitingfrequencyatmaximum
Figure 8. Weight of mass block versus natural frequency of PCP and exciting frequency at maximum
voltageincantileverbeamfixationmode.
voltageincantileverbeamfixationmode.
voltage in cantilever beam fixation mode.
Energies 2016, 9, 98 7 of 14
Energies2016,9,98 7of14

3000
Natural frequency
Energies2016,9,98 Frequency to maximum voltage 7of14
2500
3000
Natural frequency
2000
Frequency (Hz)
Frequency to maximum voltage
2500
1500
2000
Frequency (Hz)

1000
1500

500
1000

0500
0 1 2 3 4 5 6 7 8 9 10
0 Length-width ratio
0 1 2 3 4 5 6 7 8 9 10
Figure
Figure9. Length-width ratio
9.Lengthwidth ratioversus naturalLength-width
versus natural frequency
frequencyof ratio
PCP
ofPCPand
andexciting
excitingfrequency at
frequency atmaximum
maximum
voltage in cantilever beam fixation mode.
voltageincantileverbeamfixationmode.
Figure 9.Lengthwidth ratioversus natural frequency ofPCP and exciting frequency at maximum
voltageincantileverbeamfixationmode.
Figure10
Figure 10shows
shows that
that with
with the
the increase
increase in
in weight
weight of
of the
the mass
mass block
block affixed
affixed on
on the
the PCP,
PCP,the
the
maximum
maximum voltage
voltage keeps
keeps falling.
falling. The
The maximum
maximum voltage
voltage also
also rises
rises with
with the
the increase
increase of
of the
the
Figure 10 shows that with the increase in weight of the mass block affixed on the PCP, the
PCP
PCP
lengthwidthratio.Bycomparingtheinfluencesofmassblockandlengthwidthratioincantilever
length-width
maximum ratio. By comparing
voltage the
keeps falling. influences
The maximum ofvoltage
mass block and with
also rises length-width ratio
the increase ofin cantilever
the PCP
beamfixationmode,conclusionscanbesummarizedasfollows:
beam fixation mode, conclusions can be summarized as follows:
lengthwidthratio.Bycomparingtheinfluencesofmassblockandlengthwidthratioincantilever
beamfixationmode,conclusionscanbesummarizedasfollows:
(1) (1)
It isItismoreeffectivetochangelengthwidthratioofthePCPthanweightofthemassblock
more effective to change length-width ratio of the PCP than weight of the mass block for
(1) ItismoreeffectivetochangelengthwidthratioofthePCPthanweightofthemassblock
foradjustingthenaturalfrequencyofthePCPincantileverbeamfixationmode.
adjusting the natural frequency of the PCP in cantilever beam fixation mode.
foradjustingthenaturalfrequencyofthePCPincantileverbeamfixationmode.
(2) A widerwider
(2) A rangerange of generated
of generated voltagevoltage can be achieved
can be achieved by changing
by changing the lengthwidth
the length-width ratio
ratio compared
(2) A wider range of generated voltage can be achieved by changing the lengthwidth ratio
withcomparedwithalteringtheweightofmassblock.
altering the weight of mass block.
comparedwithalteringtheweightofmassblock.

Length-width ratio
Length-width ratio
0 0 1 1 22 33 44 5
5
6
6 7
7 8
8
140140 300300

135135 250250
Maximum voltage (V)
Maximum voltage (V)

Maximum voltage (V)

Maximum voltage (V)

130130 200200

125125 150150
120120
100
115 100
115
110 50
110 50
105 0
105 0 5 10 15 20 25 30 35 0
0 5 10 15 20 25 30 35
Weight of mass block (g)
Weight of mass block (g)
FigureFigure10.Weightofmassblockandlengthwidthratioversusmaximumvoltageincantileverbeam
10. Weight of mass block and length-width ratio versus maximum voltage in cantilever beam
fixationmode.
Figure10.Weightofmassblockandlengthwidthratioversusmaximumvoltageincantileverbeam
fixation mode.
fixationmode.
5.ExperimentalStudies
5. Experimental Studies
5.ExperimentalStudies
Experiments with PCPs affixed with a mass block in simply supported beam and cantilever
Experiments with PCPs affixed with a mass block in simply supported beam and cantilever beam
beamfixationmodewerecarriedoutbyusingavibrationtestrig.AsshowninFigure11,thevibration
Experiments
fixation mode were with PCPs
carried outaffixed with
by using a a mass block
vibration test in
rig.simply supported
As shown beam
in Figure 11,and
the cantilever
testrigiscomposedofuppercomputersoftware,signalcontroller,vibrationamplifierandvibrator. vibration
beamfixationmodewerecarriedoutbyusingavibrationtestrig.AsshowninFigure11,thevibration
test rig is composed
Among of upper
these parts, computer
the signal software,
controller signal
generates controller,
signals vibration
and transmits amplifier
them and vibrator.
to the vibration
testrigiscomposedofuppercomputersoftware,signalcontroller,vibrationamplifierandvibrator.
Among these parts, the signal controller generates signals and transmits them to the vibration
Energies 2016, 9, 98 8 of 14

Energies2016,9,98 8of14
Among these parts, the signal controller generates signals and transmits them to the vibration amplifier,
whichamplifier,whichwilldrivethevibratortowork.Theuppercomputersoftwarecansettheexciting
will drive the vibrator to work. The upper computer software can set the exciting frequency,
amplitude and amplitude
frequency, working mode of the mode
and working vibration test
of the rig andtest
vibration oscilloscope is adopted
rig and oscilloscope to observe
is adopted to the
voltage generated by PCP. Figure 12 shows the PCPs fixed in simply supported beam and cantilever
observethevoltagegeneratedbyPCP.Figure12showsthePCPsfixedinsimplysupportedbeam
beamandcantileverbeamfixationmode.
fixation mode.

Figure11.Vibrationtestrig.
Figure 11. Vibration test rig.

Figure 12. Piezoelectric energy harvester.
Figure12.Piezoelectricenergyharvester.

Theharmonicforcesettingsarealittlebitdifferentintheexperimentasthevibratortestrigcan
The harmonic force settings are a little bit different in the experiment as the vibrator test rig
can not produce a programmableharmonic
not produce a programmable harmonic force
forcemeasured
measured by by
value of the
value force.
of the If the
force. If settings of of
the settings
harmonic force in experiments strictly follow thesettings in simulations, the characteristics
harmonic force in experiments strictly follow the settings in simulations, the characteristics of the of the
PCPwouldbefewasthepositionwheretheforceexertedatisfixedwiththevibrator.Inordertoget
PCP would be few as the position where the force exerted at is fixed with the vibrator. In order to get
similarresultstothesimulation,wedirectlyequippedthePCPwiththevibratorinsimplysupported
similar results to the simulation, we directly equipped the PCP with the vibrator in simply supported
beamfixationmodeandcantileverbeamfixationmode.Thenthepiezoelectricmodelcanbeasimply
beamsupportedbeamorcantileverbeamagain.Datalikenaturalfrequencyandmaximumvoltagecanbe
fixation mode and cantilever beam fixation mode. Then the piezoelectric model can be a simply
supported beam or cantilever beam again. Data like natural frequency and maximum voltage can
acquiredagain.Afterseveraltries,theexcitingforceamplitudeissetto1mmintheexperimentsand
be acquired again. After several tries, the exciting force amplitude is set to 1 mm in the experiments
accompaniedbyawideexcitingfrequencyrangesincethenaturalfrequencyofthePCPmodelis
and accompanied by a wide exciting frequency range since the natural frequency of the PCP model is
unknowntousatfirst.
The electric
unknown to us at first. energy generated by a PCP usually cannot power ordinary loads (such as a
microcontrollerunit)duetoitshighvoltageandultralowcurrent,sothemethodsofusingaload
The electric energy generated by a PCP usually cannot power ordinary loads (such as
resistancetoconsumeelectricpowerorstoringelectricenergywithareservoirlikeacapacitorare
a microcontroller unit) due to its high voltage and ultra-low current, so the methods of using a load
thetwomaintechniquestocalculatethequantityofelectricity.Inthispaper,asupercapacitorwas
resistance to consume electric power or storing electric energy with a reservoir like a capacitor are
adopted to store the electric energy generated by PCP due to its high energy density, stable
the two main techniques to calculate the quantity of electricity. In this paper, a super capacitor
performance,fastchargingspeedandhighefficiency[27].Thealternatingcurrent(AC)needstobe
was adopted to store the electric energy generated by PCP due to its high energy density, stable
stabilizedbyrectifying,filteringandregularizingbeforechargingthesupercapacitor.Fortheaimto
performance, fast-charging speed and high efficiency [27]. The alternating current (AC) needs to be
reduceelectricenergyloss,thestepofregularizingtheACwasremoved.Figure13displaysthecircuit
stabilized by rectifying, filtering and regularizing before charging the super capacitor. For the aim to
schematicdiagramanditscircuitboard.Thecapacitanceoftheadoptedsupercapacitoris0.1F.
reduce electric energy loss, the step of regularizing the AC was removed. Figure 13 displays the circuit
schematic diagram and its circuit board. The capacitance of the adopted super capacitor is 0.1 F.
Energies 2016, 9, 98 9 of 14
Energies2016,9,98 9of14

(a) (b)
Figure 13. (a) Circuit schematic for rectifying and filtering AC; (b) Rectifying and filtering board,
Figure 13. (a) Circuit schematic for rectifying and filtering AC; (b) Rectifying and filtering board, super
supercapacitorboard.
capacitor board.

The way to calculate the electric energy stored in super capacitor are deduced as follows.
The way to calculate the electric energy stored in super capacitor are deduced as follows.
Equations(1)and(2)aretheexpressionofpowerandelectricenergy,respectively:
Equations (1) and (2) are the expression of power and electric energy, respectively:
dW (t )
p(t ) u (t )i (t ) (1)
dt
dWptq
pptq uptqiptq (1)
dt
t2
Wc (t ) pt2(t ) dt (2)
t1
Wc ptq pptqdt (2)
t1
Andtheexpressionofelectricenergycandeveloptoanotherformasfollows:
And the expression of electric energy tcan 2 develop t 2 to another form as follows:
Wc (t ) p ()d u ()i ()d
t1t2 t1
t2
Wc ptqt 2 t1du () t1 upqipqd
ppqd
t2 cu () dupqd (3)
t1 t1
cupq d d (3)
d
u ( t 2) 1
cc uup(tt1
upt2qudu c1[u (t22) u (t1)]
2 2
2
1) q udu 2 2 cru pt2q u pt1qs

() .
dQ dudupq
dQ
where: i ()
where: ipq c c .
d
d d d
As the capacitance of a super capacitor is constant, the quantity of electric energy that the capacitor
Ascan
stored thebecapacitance
obtained if of
theainitial
super capacitor
and is constant,
final voltages the quantity
are known. of electric energy
In the experiments, the timethat the
to charge
capacitorstoredcanbeobtainediftheinitialandfinalvoltagesareknown.Intheexperiments,the
the super capacitor was set to 2 min and the voltage of super capacitor was recorded at the beginning
timetochargethesupercapacitorwassetto2minandthevoltageofsupercapacitorwasrecorded
and the end of the test. Then the electric energy quantity that super capacitor stored in 2 min can be
atthebeginningandtheendofthetest.Thentheelectricenergyquantitythatsupercapacitorstored
acquired through calculation of Equation (3), which can also indicate the average generation power of
in2mincanbeacquiredthroughcalculationofEquation(3),whichcanalsoindicatetheaverage
PCP in the 2 min.
generationpowerofPCPinthe2min.
5.1. Simply Supported Beam Fixation Mode
5.1.SimplySupportedBeamFixationMode
In this part, we first set the vibration test rig to sweep the frequency working mode to obtain the
Inthispart,wefirstsetthevibrationtestrigtosweepthefrequencyworkingmodetoobtainthe
maximum voltage and its corresponding exciting frequency of each PCP in the simply supported beam
maximum voltageanditscorresponding
fixation mode. According to the simulation excitingfrequency of each
results, the exciting PCP in the
frequency valuesimply
is justsupported
the natural
beam fixation
frequency value mode.
of theAccording
PCP whento the
the simulation
generated results,
voltage the exciting
reaches its peak.frequency
In order tovalue is maximum
get the just the
naturalfrequencyvalueofthePCPwhenthegeneratedvoltagereachesitspeak.Inordertogetthe
power of each PCP, then we use the natural frequency (or exciting frequency that makes the voltage
maximumpowerofeachPCP,thenweusethenaturalfrequency(orexcitingfrequencythatmakes
reach its peak) of each PCP as the excitation frequency to excite each PCP for 2 min and the exciting
thevoltagereachitspeak)ofeachPCPastheexcitationfrequencytoexciteeachPCPfor2minand
force amplitude is set to 1 mm. With the help of a rectifying and filtering board and a super capacitors
theexcitingforceamplitudeissetto1mm.Withthehelpofarectifyingandfilteringboardanda
board, the maximum power generation of every PCP with different mass block weights is recorded.
superAscapacitors
displayed board, the maximum
in Figure power generation
14, the maximum voltages ofof12every
groups PCP
ofwith
PCPsdifferent massdifferent
affixed with block
weightsisrecorded.
mass blocks all remain between 18 V and 31 V. Compared with Figure 6 in Section 3, the voltages in
bothAsdisplayedinFigure14,themaximumvoltagesof12groupsofPCPsaffixedwithdifferent
the simulations and experiments decrease with the increase of weight of the mass block, though
massblocksallremainbetween18Vand31V.ComparedwithFigure6inSection3,thevoltagesin
the dropping speed of the maximum voltage in the experiments is slower than that in the software
boththesimulationsandexperimentsdecreasewiththeincreaseofweightofthemassblock,though
simulation. Figure 15 indicates that excitation frequencies at the maximum voltage generated by the
thedroppingspeedofthemaximumvoltageintheexperimentsisslowerthanthatinthesoftware
simulation.Figure15indicatesthatexcitationfrequenciesatthemaximumvoltagegeneratedbythe
Energies 2016, 9, 98 10 of 14

Energies2016,9,98
Energies2016,9,98 10of14
10of14
PCP decrease with the increase of weight of the mass block affixed on the PCP. Compared with Figure 4
PCP
PCPdecrease
in Section 3, these with
decrease withthe
the
results increase of
ofweight
increasewell
coincide with of
weight ofthe
eachthemass
massblock
other. blockaffixed
Figure affixed
15 alsoon
onthe
thePCP.
PCP.Compared
demonstrates Compared
that thewith
with
natural
Figure4inSection3,theseresultscoincidewellwitheachother.Figure15alsodemonstratesthatthe
Figure4inSection3,theseresultscoincidewellwitheachother.Figure15alsodemonstratesthatthe
frequency has a great impact on the maximum generation power of a PCP, because the maximum
natural
naturalfrequency
frequencyhas
hasaagreat
greatimpact
impacton
onthe
themaximum
maximumgeneration
generationpower
powerofofaaPCP,
PCP,because
becausethe
the
generation power of the PCP drops very rapidly with the decrease of natural frequency. As for the
maximumgenerationpowerofthePCPdropsveryrapidlywiththedecreaseofnaturalfrequency.
maximumgenerationpowerofthePCPdropsveryrapidlywiththedecreaseofnaturalfrequency.
unsmooth
As part unsmooth
of the maximum generation power when powerthe weight ofthethe mass of block mass
equals 15 g and
Asfor
forthe
the unsmoothpart
partof
ofthe
themaximum
maximumgeneration
generation powerwhen
when theweight
weight ofthe
the massblock
block
18 g in Figure 15, the glue between PCP and mass block could be the reason for that as it can changed
equals15gand18ginFigure15,thegluebetweenPCPandmassblockcouldbethereasonforthat
equals15gand18ginFigure15,thegluebetweenPCPandmassblockcouldbethereasonforthat
the characteristics of the PCP.
asitcanchangedthecharacteristicsofthePCP.
asitcanchangedthecharacteristicsofthePCP.

45
45

40
40
(V)
voltage (V)

35
35
Maximum voltage

30
30

25
25
Maximum

20
20

15
15

10
10
00 55 10
10 15
15 20
20 25
25 30
30 35
35 40
40
Weight
Weightof
ofmass
massblock
block(g)
(g)
Figure14.Weightofmassversusmaximumvoltageinsimplysupportedbeamfixationmode.
Figure14.Weightofmassversusmaximumvoltageinsimplysupportedbeamfixationmode.
Figure 14. Weight of mass versus maximum voltage in simply supported beam fixation mode.

130
130 10
10 (mW)
power (mW)
120
120 99

110
110 88
(Hz)
frequency (Hz)

generation power

77
100
100
Natural frequency

66
90
Maximum generation

90
55
80
80
44
70
70 33
Natural

60
60
Maximum

22
50
50 11
40
40 00
00 33 66 99 12
12 15
15 18
18 21
21 24
24 27
27 30
30 33
33 36
36
Weight
Weightof
ofmass
massblock
block(g)
(g)

Figure 15. Weight
Figure
Figure15. of mass
15.Weight
Weight of block
ofmass
mass versus
block
block natural
versus
versusnatural frequency
naturalfrequency andmaximum
frequencyand
and maximum power
power
maximum of
power of PCP
ofPCP
PCPin in simply
insimply
simply
supported beam fixation mode.
supportedbeamfixationmode.
supportedbeamfixationmode.

5.2.CantileverBeamFixationMode
5.2.CantileverBeamFixationMode
5.2. Cantilever Beam Fixation Mode
Settingsinthispartareconsistentwiththeabovesettings,maximumvoltageanditscorresponding
Settingsinthispartareconsistentwiththeabovesettings,maximumvoltageanditscorresponding
Settings in this part are consistent with the above settings, maximum voltage and its corresponding
exciting
excitingfrequency,
frequency,maximum
maximumgeneration
generationpower
powerare
areall
allobtained.
obtained.As
Asshown
shownin
inFigure
Figure16,
16,the
the
exciting frequency, maximum generation power are all obtained. As shown in Figure 16, the maximum
maximumvoltageoftheselectedcommercialPCPcanreach40Vandthevoltagesgeneratedby12
maximumvoltageoftheselectedcommercialPCPcanreach40Vandthevoltagesgeneratedby12
voltage of the selected commercial PCP can reach 40 V and the voltages generated by 12 groups of PCPs
groupsofPCPsaffixedwithdifferentmassblocksallrangedfrom35Vto40V.ComparedwithFigure
groupsofPCPsaffixedwithdifferentmassblocksallrangedfrom35Vto40V.ComparedwithFigure
affixed with different mass blocks all ranged from 35 V to 40 V. Compared with Figure 10 in Section 4,
10inSection4,theirvariationrangesarebothsmall.Asforthelowervoltageintheexperiments,
10inSection4,theirvariationrangesarebothsmall.Asforthelowervoltageintheexperiments,
theirmanufacturelimitationsandexperimentalerrorsarethemainreasons.Figure17demonstratesthat
variation ranges are both small. As for the lower voltage in the experiments, manufacture
manufacturelimitationsandexperimentalerrorsarethemainreasons.Figure17demonstratesthat
limitations and experimental errors are the main reasons. Figure 17 demonstrates that the excitation
theexcitationfrequenciesatmaximumvoltagegeneratedbythePCPsdecreasewiththeincreaseof
theexcitationfrequenciesatmaximumvoltagegeneratedbythePCPsdecreasewiththeincreaseof
frequencies at maximum voltage generated by the PCPs decrease with the increase of weight of the
mass block affixed on the PCP. Compared with Figure 8, their variation trends are almost the same.
Therefore, the results from experiments in cantilever beam fixation mode are similar to the simulation
 Energies2016,9,98
Energies2016,9,98 11of14
11of14
Energies 2016, 9, 98 11 of 14
weight
weightof
ofthe
themass
massblock
blockaffixed
affixedon
onthe
thePCP.
PCP.Compared
Comparedwith
withFigure
Figure8,
8,their
theirvariation
variationtrends
trendsare
are
almostthesame.Therefore,theresultsfromexperimentsincantileverbeamfixationmodearesimilar
almostthesame.Therefore,theresultsfromexperimentsincantileverbeamfixationmodearesimilar
results. At the same time, the excitation frequency corresponding to the maximum voltage in the
tothesimulationresults.Atthesametime,theexcitationfrequencycorrespondingtothemaximum
tothesimulationresults.Atthesametime,theexcitationfrequencycorrespondingtothemaximum
voltageintheexperimentcanbetreatedasthenaturalfrequencyofPCPmodelaswell.
experiment can be treated as the natural frequency of PCP model as well.
voltageintheexperimentcanbetreatedasthenaturalfrequencyofPCPmodelaswell.

60
60

50
50

(V)
voltage(V)
40
40
Maximumvoltage
30
30
Maximum

20
20

10
10

00
00 10
10 20
20 30
30 40
40 50
50
Weight
Weight of
of mass
mass block
block (g)
(g)
Figure16.Weightofmassversusmaximumvoltageincantileverbeamfixationmode.
Figure16.Weightofmassversusmaximumvoltageincantileverbeamfixationmode.
Figure 16. Weight of mass versus maximum voltage in cantilever beam fixation mode.

45
45

(mW)
power(mW)
0.5
0.5
40
40
(Hz)
frequency (Hz)

0.4

generation power
35 0.4
35
Natural frequency

30
30 0.3
0.3
Maximum generation
25
25 0.2
0.2
20
20
Natural

0.1
0.1
15
Maximum

15
0.0
0.0
10
10
00 33 66 99 12
12 15
15 18
18 21
21 24
24 27
27 30
30 33
33 36
36
Weight
Weight of
of mass
mass block
block (g)
(g)
Figure 17. Weight of mass block versus natural frequency and maximum power of PCP in cantilever
Figure17.WeightofmassblockversusnaturalfrequencyandmaximumpowerofPCPincantilever
Figure17.WeightofmassblockversusnaturalfrequencyandmaximumpowerofPCPincantilever
beambeamfixationmode.
fixation mode.
beamfixationmode.

In
In addition,
addition, comparing
comparing the
the maximum
maximum generation
generation power
power in
in simply
simply supported
supported beam and
In addition, comparing the maximum generation power in simply supported beambeam and
and cantilever
cantileverbeamfixationmode,thefollowingconclusionscanbemade:
cantileverbeamfixationmode,thefollowingconclusionscanbemade:
beam fixation mode, the following conclusions can be made:
(1)
(1) The
Thenaturalfrequency
naturalfrequencyof
ofthe
thePCP
PCPin
insimply
simplysupported
supportedbeam
beammode
modeislarger
islargerthan
thanthat
thatin
in
(1) The natural frequency of the PCP in simply supported beam mode is larger than that in cantilever
cantileverbeamfixationmode.
cantileverbeamfixationmode.
beam (2) Themaximumgenerationpowerismuchhigherthanthatincantileverbeamfixationmode.
fixation
(2) mode.
Themaximumgenerationpowerismuchhigherthanthatincantileverbeamfixationmode.
(2) (3)
(3) ThenaturalfrequencydeterminesthemaximumgenerationpowerofaPCP.
ThenaturalfrequencydeterminesthemaximumgenerationpowerofaPCP.
The maximum generation power is much higher than that in cantilever beam fixation mode.
(4)
(4) Duringtheexperiments,wenoticedthatthemaximumgenerationpowerandmaximum
Duringtheexperiments,wenoticedthatthemaximumgenerationpowerandmaximum
(3) The natural frequency determines the maximum generation power of a PCP.
voltageinsimplysupportedbeamfixationmodeareverysensitivetotheexactexcitation
voltageinsimplysupportedbeamfixationmodeareverysensitivetotheexactexcitation
(4) During the experiments, we noticed that the maximum generation power and maximum voltage
frequencyasasmalldeviationinexcitationfrequencymaycauseittogeneratefarfromthe
frequencyasasmalldeviationinexcitationfrequencymaycauseittogeneratefarfromthe
in simply supported beam fixation mode are very sensitive to the exact excitation frequency as
maximumpower,buttheexcitationfrequencyrangeincantileverbeamfixationmodeiswider.
maximumpower,buttheexcitationfrequencyrangeincantileverbeamfixationmodeiswider.
a small deviation in excitation frequency may cause it to generate far from the maximum power,
but the excitation frequency range in cantilever beam fixation mode is wider.

6. Comparison Results and Guidelines for Application of PCP

According to the simulation and experimental statistics described above, the natural frequency of
the PCP in simply supported beam fixation mode is higher; the natural frequencies of PCPs affixed
Energies 2016, 9, 98 12 of 14

with mass block range were 200400 Hz in simulations and 50120 Hz in experiments. Furthermore,
the deformations of PCPs in simply supported beam fixation mode are relatively small, causing the
maximum voltages generated by PCPs to be relatively low, but the maximum generation power in
experiments astonished us; the maximum generation power from every PCP in simply supported
beam fixation mode is larger than that in cantilever beam fixation mode. However, experiments
in simply supported beam fixation mode showed that maximum generation power and maximum
voltage are very sensitive to the exact excitation frequency, because a small deviation can result in
an ultra-low generation power compared with the maximum generation power.
As for the cantilever beam fixation mode, the natural frequencies of PCPs are relatively low.
Most of the natural frequencies of PCPs with different length-width ratios are no more than 50 Hz in
both simulations and experiments, nevertheless the deformations of PCPs are relatively large, resulting
in a great increase in generated voltage; the maximum voltages generated by PCPs can reach as high as
138 V in simulations and a peak voltage of 40 V in experiments. Beside, the excitation force frequency
for achieving maximum voltage and maximum power generation is wider. However, attention should
be paid on the strength of the PCP since a large deformation may break it. Compared with PCPs in
simply supported beam fixation mode, the maximum generation power by PCPs in cantilever beam
fixation mode is pretty low; most of the maximum generation powers in cantilever beam fixation mode
are no more than 0.01 mW, but the majority of maximum generation powers in simply supported beam
fixation mode exceed 1 mW.
Wireless sensor nodes could be placed anywhere in our daily life, for the aim to make it work
properly with a long working lifespan, fixation mode and adjusting method must fit the exciting source
to generate maximum capacity. As shown in Table 2, the frequencies of exciting source in popular use
range at 60200 Hz, and the accelerations range at 0.510 m2 /s, which can provide instructions for
practical applications of PCP [2].

Table 2. Parameters of each material.

Vibration Source Frequency (Hz)

Car engine compartment 200
Base of 3-axis machine tool 70
Blender casing 121
Clothes dryer 121
Person nervously tapping their heel 1
Car instrument panel 13
Door frame just after door closes 125
Small microwave oven 121
HVAC vents in office building 60
Windows next to a busy road 100
CD on notebook computer 75
Second story floor of busy office 100

Therefore, cantilever beam fixation mode is a reasonable choice in practical applications when
the excitation frequency is variable and below 50 Hz. The length-width ratio is a good parameter
to rectify the natural frequency of PCPs for the aim of achieving resonance and optimal power
generating capacity. At the same time, attention should be paid to the strength of the PCP as higher
power generation capacity means larger deformation and being easier to break. When the excitation
frequencies are higher than 50 Hz, the simply supported beam fixation mode is a better choice as
it is easier to resonate. Its output voltage may be low, but its maximum generation power is much
higher. Adjusting the length-width ratio can be also adopted to modify the natural frequencies of
PCPs to obtain the maximum power generation capacity. Furthermore, the deformation of the PCP
adopting simply supported beam fixation mode is relatively small, which provides a long-time and
stable operation. Thus, wireless sensor nodes with piezoelectric energy harvester in simply supported
Energies 2016, 9, 98 13 of 14

beam fixation mode can have a longer working lifespan. Also, we must pay attention to the excitation
force frequency, because a small deviation from the exact excitation frequency may cause unexpected
low power generation efficiency.

Acknowledgments: The material presented in this paper is based upon work supported by National Natural
Science Foundation of China (No. 51575513), Fundamental Research Funds for the Central Universities
(No. 2015QNA37), Program for Changjiang Scholars and Innovative Research Team in University (No. IRT1292),
as well as Priority Academic Program Development of Jiangsu Higher Education Institutions.
Author Contributions: All the authors have contributed their efforts to make sure the paper is qualified.
Gongbo Zhou designed and supervised the work. Zhixiang Li conducted the simulation and experiment parts.
Zhencai Zhu and Wei Li discussed the experiment results.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Zhou, G.; Huang, L.; Li, W.; Zhu, Z. Harvesting Ambient Environmental Energy for Wireless Sensor
Networks: A Survey. J. Sens. 2014, 2014. [CrossRef]
2. Roundy, S.; Wright, P.K.; Rabaey, J. A study of low level vibrations as a power source for wireless sensor
nodes. Comput. Commun. 2003, 26, 11311144. [CrossRef]
3. Souliotis, M.; Singh, R.; Papaefthimiou, S.; Lazarus, I.J.; Andriosopoulos, K. Integrated collector storage solar
water heaters: Survey and recent developments. Energy Syst. 2015, 7, 124. [CrossRef]
4. Leung, D.Y.C.; Yang, Y. Wind energy development and its environmental impact: A review. Renew. Sustain.
Energy Rev. 2012, 16, 10311039. [CrossRef]
5. Nishimoto, H.; Kawahara, Y.; Asami, T. Prototype implementation of ambient RF energy harvesting wireless
sensor networks. In Proceedings of the 2010 Sensors, Kona, HI, USA, 14 November 2010; pp. 12821287.
6. Jabbar, H.; Song, Y.; Jeong, T. RF energy harvesting system and circuits for charging of mobile devices.
IEEE Trans. Consum. Electron. 2010, 56, 247253. [CrossRef]
7. Mathuna, C.O.; ODonnell, T.; Martinez-Catala, R.V.; Rohan, J.; OFlynn, B. Energy scavenging for long-term
deployable wireless sensor networks. Talanta 2008, 75, 613623. [CrossRef] [PubMed]
8. Yang, B.; Lee, C.; Xiang, W.; Xie, J.; Han He, J.; Kotlanka, R.K.; Low, S.P.; Feng, H. Electromagnetic energy
harvesting from vibrations of multiple frequencies. J. Micromech. Microeng. 2009, 19. [CrossRef]
9. Anton, S.R.; Sodano, H.A. A review of power harvesting using piezoelectric materials (20032006).
Smart Mater. Struct. 2007, 16, R1R21. [CrossRef]
10. Shenck, N.S.; Paradiso, J.A. Energy scavenging with shoe-mounted piezoelectrics. IEEE Micro 2001, 21, 3042.
[CrossRef]
11. Ottman, G.K.; Hofmann, H.F.; Bhatt, A.C.; Lesieutre, G.A. Adaptive piezoelectric energy harvesting circuit
for wireless remote power supply. IEEE Trans. Power Electron. 2002, 17, 669676. [CrossRef]
12. Zhou, G.; Wang, H.; Zhu, Z.; Huang, L.; Li, W. Performance Analysis of Wind-Induced Piezoelectric Vibration
Bimorph Cantilever for Rotating Machinery. Shock Vib. 2015, 2015, 110. [CrossRef]
13. Zhang, C.; He, X.F.; Li, S.Y.; Cheng, Y.Q.; Rao, Y. A wind energy powered wireless temperature sensor node.
Sensors 2015, 15, 50205031. [CrossRef] [PubMed]
14. Song, R.; Shan, X.; Lv, F.; Xie, T. A study of vortex-induced energy harvesting from water using PZT
piezoelectric cantilever with cylindrical extension. Ceramics Int. 2015, 41, S768S773. [CrossRef]
15. Xu, J.W.; Shao, W.W.; Kong, F.R.; Feng, Z.H. Right-angle piezoelectric cantilever with improved energy
harvesting efficiency. Appl. Phys. Lett. 2010, 96. [CrossRef]
16. Ben Ayed, S.; Abdelkefi, A.; Najar, F.; Hajj, M.R. Design and performance of variable-shaped piezoelectric
energy harvesters. J. Intell. Mater. Syst. Struct. 2013, 25, 174186. [CrossRef]
17. Rosa, M.; De Marqui, C., Jr. Modeling and analysis of a piezoelectric energy harvester with varying
cross-sectional area. Shock Vib. 2014, 2014, 19. [CrossRef]
18. Abdelkefi, A.; Barsallo, N.; Tang, L.; Yang, Y.; Hajj, M.R. Modeling, validation, and performance of
low-frequency piezoelectric energy harvesters. J. Intell. Mater. Syst. Struct. 2013, 25, 14291444. [CrossRef]
19. Abdelkefi, A.; Najar, F.; Nayfeh, A.H.; Ayed, S.B. An energy harvester using piezoelectric cantilever beams
undergoing coupled bendingtorsion vibrations. Smart Mater. Struct. 2011, 20. [CrossRef]
Energies 2016, 9, 98 14 of 14

20. Sharpes, N.; Abdelkefi, A.; Priya, S. Comparative Analysis of One-Dimensional and Two-Dimensional
Cantilever Piezoelectric Energy Harvesters. Energy Harvest. Syst. 2014, 1, 209216. [CrossRef]
21. Sharpes, N.; Abdelkefi, A.; Priya, S. Two-dimensional concentrated-stress low-frequency piezoelectric
vibration energy harvesters. Appl. Phys. Lett. 2015, 107. [CrossRef]
22. Tang, L.; Yang, Y. A nonlinear piezoelectric energy harvester with magnetic oscillator. Appl. Phys. Lett.
2012, 101. [CrossRef]
23. Abdelkefi, A.; Barsallo, N. Comparative modeling of low-frequency piezomagnetoelastic energy harvesters.
J. Intell. Mater. Syst. Struct. 2014, 25, 17711785. [CrossRef]
24. Abdelkefi, A.; Barsallo, N. Nonlinear analysis and power improvement of broadband low-frequency
piezomagnetoelastic energy harvesters. Nonlinear Dyn. 2016, 83, 4156. [CrossRef]
25. Gafforelli, G.; Corigliano, A.; Xu, R.; Kim, S.-G. Experimental verification of a bridge-shaped, nonlinear
vibration energy harvester. Appl. Phys. Lett. 2014, 105. [CrossRef]
26. Wang, X.M.; Wang, J.Z.; Gao, Z.Y.; Zhu, X.J. Simulation and experimental study on active vibration control
of piezoelectric smart plate based on ANSYS. In Proceedings of the 2010 2nd International Conference on
Advanced Computer Control (ICACC), Shenyang, China, 2729 March 2010; pp. 257260.
27. Conway, B.E. Similarities and Differences between Supercapacitors and Batteries for Storing Electrical Energy.
In Electrochemical Supercapacitors; Springer US: New York, NY, USA, 1999; pp. 1131.

2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons by Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).