All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.

Information Report

Category:
## Documents

Published:

Views: 6 | Pages: 6

Extension: PDF | Download: 0

Share

Description

Download Current Applied Physics Volume 15 issue 6 2015 [doi 10.1016_j.cap.2015.02.009] Hwang, Sung Joo; Jung, Hyun Jun; Kim, Jeong Hun; Ahn, Jung Hwan; -- Designing and manufacturing a piezoelectric tile fo.pdf

Tags

Transcript

Designing and manufacturing a piezoelectric tile for harvesting energyfrom footsteps
Sung Joo Hwang
a
, Hyun Jun Jung
a
, Jeong Hun Kim
a
, Jung Hwan Ahn
a
, Daniel Song
a
,Yewon Song
b
, Hee Lak Lee
b
, Sung Pil Moon
c
, Hyeonsu Park
d
, Tae Hyun Sung
a
,
*
a
Department of Electrical Engineering, Hanyang University, Seoul 136-791, Republic of Korea
b
Department of Mechanical Engineering, Hanyang University, Seoul 136-791, Republic of Korea
c
Department of Physics, Hanyang University, Seoul 136-791, Republic of Korea
d
Department of Mathematics, Hanyang University, Seoul 136-791, Republic of Korea
a r t i c l e i n f o
Article history:
Received 11 November 2014Received in revised form22 January 2015Accepted 5 February 2015Available online 14 March 2015
Keywords:
Piezoelectric tileMechanical resonanceFrequency matchingImpedance matching
a b s t r a c t
The objective of this research is to design a piezoelectric tile for harvesting energy from footsteps and tooptimize the system for harvesting maximum energy. Because piezoelectric modules easily break whendirectly subjected to energy generated by human movements, we designed a tile that employs indirectenergy transmission using springs and a tip mass. We aimed at matching the mechanical resonancefrequency of the tile with that of the piezoelectric modules. The resonance frequency of a piezoelectricmodule with a 10-g tip mass was almost similar to the vibration frequency of the tile at 22.5 Hz whenwedropped an 80-g steel ball from a 1-m height. We performed impedance matching and realized amatching value of 15 k
U
. Under these optimal mechanical and electrical conditions, we harvested 770-
m
W RMS and 55-mW peak output power.
©
2015 Elsevier B.V. All rights reserved.
1. Introduction
Overthepast tenyears, researchstudies onpiezoelectric energyharvesting have been extensively conducted [1
e
18]. Among them,studies on piezoelectric energy harvesting from human force havealso been actively conducted. These studies include the energyharvested from the bending of elbow or
nger joints [19], implantsin the knee joints [20], electricity generated using polyvinylidenedi
uoride attached to bag straps [21], piezoelectric modulesinserted under the soles of shoes [22
e
25], or motion of the humanlimbs [26,27].The power generated from wearable devices such as shoes orbackpacks can be utilized as micro-electricity sources for auxiliarypower. Although the above mentioned cases cannot be consideredas macro-sources because of their limited installation area, inde-pendent units such as piezoelectric tiles can be planted over awider area; thus, they can be used as macro-power sources.Two ways of generating power from piezoelectric modules areavailable: hitting [28
e
31] and vibrating [32
e
42]. Hitting involvesthe transfer of energy directly to the piezoelectric modules, andthus, it generates more power than the vibrating method of generating power. However, because the hitting method of gener-atingpowercaneasilybreakthemodules,thevibratingmethodhasbeen studied more widely. Ceramic piezoelectric modules can beeasily broken when directly subjected to energy generated fromhuman movements. Therefore, we developed a piezoelectric tilethat employs an indirect method of energy transmission using aspring and a tip mass.This researchis aimed at matching the resonance frequencyof acantilevered piezoelectric module with the frequency of a piezo-electric tile. In addition, we optimize the circuits through imped-ance matching after mechanical optimization by frequencymatching.
2. Method
Fig. 1 shows the piezoelectric tile used in our experiment.Fig. 1(a) shows that it is modeled on a real tile, and its area is150
150mm
2
.Fig.1(b)showsthatthepiezoelectrictileconsistsof an upper plate that has to be directly stepped on, a middle platewhere the piezoelectric modules are set up, a bottom plate, andfour supporting springs. The middle plate is piezo installed layer
*
Corresponding author.
E-mail address:
sungth@hanyang.ac.kr (T.H. Sung).
Contents lists available at ScienceDirect
Current Applied Physics
journal homepage: www.elsevier.com/locate/cap
http://dx.doi.org/10.1016/j.cap.2015.02.0091567-1739/
©
2015 Elsevier B.V. All rights reserved.
Current Applied Physics 15 (2015) 669
e
674
which is attached upper plate. The length of the four springs is40 mm, and the thickness of the upper and bottom plates is 10 mmeach. Fig. 1(c) shows a detailed image of the cross section of themiddle part where the modules are placed. The piezoelectric ma-terial with the dimension of 47
32
0.2 mm
3
is placed on asubstrate of stainless steel plate having the dimension of 62
37
0.2 mm
3
. The thick
lm piezoelectric material is PZT-PZNM manufactured by TIOCEAN (Korea). Table 1 lists the prop-erties of the piezoelectric material.
2.1. Frequency matching between the piezoelectric tile and the piezoelectric module
Theoretically, resonance frequency of the cantilever beam relieson effective stiffness of the beam and effective mass of both beamand a tip mass [43].
u
beam
¼
ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ
K
beam
m
eff
s
(1)
Effective stiffness (
K
beam
), which can be written as
K
beam
¼
b
4
L
3
0@X
n
1
i
¼
1
n
i
E
i
h
3
i
þ
X
n
2
j
¼
1
n
j
E
j
h
3
j
1A
(2)
where
b
isthewidthofthebeam,
L
isthelengthofthebeam;n
1
and
n
2
are the numbers of piezoelectric and electrode layers;
E
i
and
h
i
aretheYoung'smodulusandheightofeachpiezoelectriclayer;and
E
j
and
h
j
are the Young's modulus and height of electrode layer,respectively. The effective mass of cantilever beam (
m
eff
) with thetip mass can be approximated as
m
eff
¼
m
t
þ
0
:
23
bL
0@X
n
1
i
¼
1
n
i
r
i
h
i
þ
X
n
2
j
¼
1
n
j
r
j
h
j
1A
(3)
where
m
t
is tip mass; and
r
i
and
r
j
are the densities of the piezo-electric and electrode plate, respectively [44].Inorderto
ndoutthenaturalfrequencyof thetile,wedroppedan 80-g steel ball from a height of 1 m for equal input energy. ThedisplacementsensorZS-HLDS10showninFig.2(a)ismanufacturedby Omron (Korea), and the DPO-4054B oscilloscope is manufac-tured by Tektronix (USA). The displacement sensor measured thedisplacement of the vibrating surface when the steel ball hit thesurface under free fall.Fig. 2(b) shows the experimental tools used for measuring theresonance frequency of the piezoelectric module with a tip mass.The function generator is model 33250A manufactured by Agilent(USA),thevibrationexciterismodel4809manufacturedbyBrüel
&
Kjær (Denmark), and the power ampli
er is model 2718 manu-factured by Brüel
&
Kjær (Denmark). We used the displacementsensor to measure the vibration displacement on the surface whenthe 80-g steel ball fell on it.UsingtheexcitershowninFig.2(b),wedeterminedthedifferentresonancefrequencies byvaryingthetip mass at0,10,30, and50g.Subsequently, we checked whether the output voltage was highnear the resonance point by measuring the voltage using differenttip masses.From the experiment described, we adopted the tip mass withthe most similar resonance frequency to the vibration frequency of the tile.
2.2. Impedance matching using resistive component andapplication to real conditions
Extracted power from piezoelectric module, which can bewritten as
Fig. 1.
Conceptual design of the piezoelectric and real tiles. (a) Real piezoelectric tilewith a real tile. (b) Illustration of the piezoelectric tile. (c) Piezo-installed layer.
Table 1
Material properties of the piezoelectric module.Piezoelectric material ValueDensity (g/cm
3
) 7.60Dielectric constants (
ε
33
T/
ε
0
) 21Piezoelectric charge constants (
10
12
m/V): d
33
, d
31
450,
200Piezoelectric voltage constants (
10
3
V m/N): g
33
,g
31
22.1,
11.1Elastic constants (
10
12
m
2
/N): S
E11
, S
D11
13.8, 11.8Stainless steel (SUS-304) ValueDensity (g/cm
3
) 8Young's modulus (GPa) 193
S.J. Hwang et al. / Current Applied Physics 15 (2015) 669
e
674
670
P
¼
V
S
Z
S
þ
Z
L
2
Z
L
(4)
where
Vs
is generated voltage from piezoelectric module,
Zs
is in-ternal impedance of piezoelectric module, and
Z
L
is load imped-ance. Maximum power was transferred from piezoelectric moduleto load when load impedance matched the internal impedance of piezoelectric module [45].We must perform impedance matching on all the inductanceand resistance components to extract the maximum power; how-ever, matching of the inductance components requires inductors of a very large size [45]. Therefore, we only used resistance compo-nents for performing impedance matching in this research.
P
max
¼
V
S
Z
S
þ
Z
L
;
opt
2
R
L
;
opt
(5)
¼
V
2
s
ð
R
s
þ
R
s
Þ
2
þð
X
s
X
s
Þ
2
!
2
R
s
¼
V
2
s
4
R
s
(6)
If weconsideronlyresistive load, thematching resistance of thefour parallel piezoelectricmoduleswill haveonlyaquarter valueof a single module.
P
rms
¼
V
rms
R
L
;
opt
2
(7)
We measured the output RMS voltage by oscilloscope duringfour seconds and calculated RMS power of one piezoelectricmodule using equation (7). In this case, we varied the resistancefrom10to100k
U
ata10-k
U
step.Whenweused four piezoelectricmodules connected in parallel, we varied the resistance from 3 to30k
U
ata3-k
U
step. Then,wecomparedtheRMSandpeakpowersof the four piezoelectric modules.Also, optimum resistance (R
L,opt
) can be obtained by
R
L
;
opt
¼
1
C
p
2
p
f
(8)
where
C
p
is the internal capacitance of piezoelectric material,
u
isthe angular frequency and f is vibration frequency of cantileverbeam [46]. And the capacitance is
C
p
¼
d
31
L
p
b
p
g
31
t
p
(9)
where
d
31
is the piezoelectric charge constant,
g
31
is the piezo-electric voltage constant, and
L
p
,
b
p
and
t
p
is the length, width andthickness of the piezoelectric material, respectively [47].After this process, we applied the piezoelectric tile to a realsituation. When a 68-kg man stepped on the tile once, wemeasured the RMS and peak values of the output power.
3. Result and discussion
3.1. Frequency matching between the piezoelectric tile and the piezoelectric module using FFT analysis and vibration exciter
The natural frequency for the total system was 23.5 Hz withoutany tip mass, and became smaller as the tip mass became heavier.With the 50-g tip mass which was the heaviest, we measured19.2 Hz for the natural frequency. Since the frequency rangedaround20Hz,itwasnecessarytotunethepiezoelectricmodulestothis value.To determine the appropriate tip mass that could resonate withthe piezoelectric tile, we calculated and measured the resonancefrequency with respect to the variation in the tip mass. We usedequation (2) for calculating effective stiffness (
K
beam
) of cantileverbeamandthevaluewas 146.2 N/m.Theeffectivemassof cantileverbeam (
m
eff
) with 0, 10, 30 and 50 g tip mass was calculated withequation (3). Each value was 1.4, 11.4, 31.4, and 51.4 g. With thesedata, the tip masses of 0, 10, 30, and 50 g resulted in theoreticalresonancefrequenciesof51.8,18,10.9,and8.5Hz,whilegivingveryclose measured values of 51.5, 20, 11, and 8.5 Hz, respectively, asshown in Fig. 3. Therefore, 10-g tip mass gave the most desirableresult.Wealsomeasuredthenaturalfrequencyofthetotalsystemwiththis tip mass. Fig. 4 shows that the main frequency is 22.5 Hz, asobtained from the fast Fourier transformation (FFT) analysis of thevibration displacement of the piezoelectric tile; the vibrationdisplacement was caused bytheimpact of the 80-g steel ball underfree fall.We conducted another experiment to determine the outputvoltage (open circuit) caused by the tip mass, by dropping steelballs with weights of 20, 40, 60, and80 g. Fig. 5 shows that a higherinput energy resulted in a higher output voltage for all cases. Thegenerationcapacity,however,wasnotproportionaltotheweightof the tip mass. The 10-g tip mass had a higher output than the 30- or
Fig. 2.
Experimental setup (a) for measuring the vibration frequency of the piezo-electric tile, (b) for measuring the resonance frequency of a piezoelectric module witha tip mass, composed of the function generator, vibration exciter, power ampli
er, andpiezoelectric module (made by TIOCEAN CO.).
S.J. Hwang et al. / Current Applied Physics 15 (2015) 669
e
674
671
50-g tip masses because the piezoelectric module with a 10-g tipmass had the most similar resonance frequency to the vibrationfrequency of the piezoelectric tile.Contrary tothe common idea that a heavier tip mass results in ahigher output power, the most desirable result was obtained whena10-g tip mass was used that realized a resonance frequencyof thepiezoelectric module that was similar to the vibration frequency of the piezoelectric tile. This mechanical resonance frequencymatching is a good initial step in the development of a macro-scalepiezoelectric tile.
3.2. Impedance matching for circuit optimization and application toreal condition
We could calculate the matched resistance at 20 Hz for onepiezoelectric panel with the equation (8), while the equation (9)
gave 135 nF for the capacitance. Therefore, the resistance became59 k
U
. For four piezoelectric panels, the
C
p
becomes four timeslarger, which caused the matched impedance to be four timessmaller than the one for single panel. Fig. 6 shows the resistiveimpedance matched at 15 k
U
, yielding an output power of 770
m
W(RMSvalue),andtheyarewell-measuredvaluescomparedwiththetheoretical ones.Therefore, by optimizing the system using the tip mass on thepiezoelectric module and the circuit through impedance matching,we could obtain the optimal conditions for the piezoelectric tile.Under these conditions, the output power-generation process
Fig. 3.
Theoretical and measured value of resonance frequency with respect to thevarious tip masses.
Fig. 4.
FFT analysis of the displacement vibration of the piezoelectric tile.
Fig. 5.
Output voltage with respect to the various tip masses.
Fig. 6.
Output voltage and power graph according to the load.
Fig. 7.
Output RMS power and peak power of the four piezoelectric modules withrespect to the load.
S.J. Hwang et al. / Current Applied Physics 15 (2015) 669
e
674
672

Recommended

Related Search

Applied PhysicsJournal of Theoretical and Applied physicsApplied And Interdisciplinary PhysicsApplied Mathematics and Theoretical PhysicsApplied Nuclear PhysicsApplied Mathematics \u0026 PhysicsApplied Mathematics and PhysicsIEEE 802.15.6Applied Mathematics to Biology, Physics, etclaser physics (DFB dye lasers), volume hologr

We Need Your Support

Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...Sign Now!

We are very appreciated for your Prompt Action!

x