Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Research ArticleA Broadband Vibration-Based Energy Harvester Using an Arrayof Piezoelectric Beams Connected by Springs
V Meruane and K Pichara
Department of Mechanical Engineering Universidad de Chile Beauchef 851 9370456 Santiago Chile
Correspondence should be addressed to V Meruane vmeruaneinguchilecl
Received 2 June 2015 Revised 25 August 2015 Accepted 27 August 2015
Academic Editor Carlo Trigona
Copyright copy 2016 V Meruane and K Pichara This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited
Piezoelectric cantilevered beams have been widely used as vibration-based energy harvesters Nevertheless these devices have anarrow frequency band and if the excitation is slightly different there is a significant drop in the level of power generated To handlethis problem the present investigation proposes the use of an array of piezoelectric cantilevered beams connected by springs asa broadband vibration-based energy harvester The equations for the voltage and power output of the system are derived basedon the analytical solution of the piezoelectric cantilevered energy harvester with Euler-Bernoulli beam assumptions To study theadvantages and disadvantages of the proposed system the results are compared with those of an array of disconnected beams (withno springs) The analytical model is validated with experimental measurements of three bimorph beams with and without springsThe results show that connecting the array of beams with springs allows increasing the frequency band of operation and increasingthe amount of power generated
1 Introduction
The idea of harvesting energy from the environment result-ing in self-powered monitoring systems is fascinating andhas attracted the attention of many researchers This isdemonstrated by the large numbers of investigations in thearea [1ndash8] Examples of remote sensors that could benefitfrom energy harvesting are weather monitoring sea surfacewater monitoring tactical military surveillance monitoringof equipment and structures oceanography and volcanoand seismic monitoring to name a few For most of theseapplications renewable power sources are needed for long-term remotemonitoring Some of the different power sourcesavailable are thermoelectric energy mechanical vibrationshydroelectric power wind energy solar power and ambientradio frequency
Mechanical vibrations are an attractive source due to theirhigh availability in certain environments The transductionmechanisms that can be used to convert ambient vibrationinto electrical energy are electromagnetic electrostatic andpiezoelectric Among these mechanisms piezoelectric trans-duction has received great attention in the last years due
to its ability to directly convert strain energy into usableelectric energy [1 9] In piezoelectric energy harvesting themechanical energy of vibrations is transformed by a piezo-electric material to electrical energy When a piezoelectricmaterial is deformed the central molecules in the crystalbecome polarized and form a dipole If the dipoles arearranged then two surfaces of thematerial become positivelyand negatively charged This property can be exploited totransform mechanical strain into electrical energy Figure 1presents a scheme of two cantilever piezoelectric powergenerators The generators are submitted to base excitationwhich causes the beams to vibrate inducing an alternatingvoltage The voltage generated by the piezoelectric elementscan be used to power a device directly or to charge arechargeable battery In the case of a unimorph beam alayer of piezoceramic material is attached to a metal layer(substructure) whereas in a bimorph configuration a metallayer is sandwiched between two piezoelectric layers In bothcases a proofmass is attached at the end to increase the strain
The main problem of conventional energy harvestersis their narrow frequency band of operation Conventionalenergy harvesters as the ones shown in Figure 1 have
Hindawi Publishing CorporationShock and VibrationVolume 2016 Article ID 9614842 13 pageshttpdxdoiorg10115520169614842
2 Shock and Vibration
Mt Rl (t)
+
minus
PiezoceramicSubstructure
(a)
Mt Rl (t)
+
minus
PiezoceramicSubstructure
(b)
Figure 1 Cantilever piezoelectric energy configurations (a) unimorph (b) bimorph (parallel connection)
an operating frequency equal to their resonance frequencyand if the excitation is slightly different there is a significantdrop in the level of power generated This problem has ledto the search for broadband energy harvesters Twiefel andWestermann [10] present a survey of broadband techniquesfor vibration-based energy harvesting These techniques areclassified as follows
(i) Linear generators arrays combined modes
(ii) Nonlinear generators bimultiple stable vibroimpactresonant tuning
(iii) Advanced electronic networks switching networkssynthetic impedance
According to the researchers there is no universal solutiondepending on the application one technique may be betterthan the others For example linear generators are moreefficient for stochastic excitations whereas nonlinear tech-niques are more suitable for harmonic excitations or very lowexcitation frequencies
In the case of linear generators recent research hasbeen focused on developing systems that exhibit multipleresonances that are closely spaced The simplest solution isto use multiple cantilevers The cantilevers are built withdifferent geometries (large tip mass or width) so each beamhas a slightly different resonant frequency thus increasing thebandwidth of the full system A first design of a cantileverarray was presented by Shahruz [11 12] He concluded thatif the dimensions of the beams and masses are chosenappropriately then the system performs as a broadbandharvester However the maximal frequency band is limitedand is independent of the beams dimensions This approachhas been studied extensively since then [13ndash15]
Another method is to use structures with two or moredegrees of freedom resulting in two or more resonantfrequencies that are closely spaced [16ndash19] Z Yang andJ Yang [18] analyzed the power generated by the flexuralvibrations of two piezoelectric beams connected by springsthey showed that this structure can be used as a broadbandenergy harvester Nevertheless the model presented by theauthors is restricted to two beams and the extension tomore than two beams is not trivial Springs have also beenused in MEMS-based energy harvesters where it has been
demonstrated that the use of nonlinear springs increases thegenerated power [20ndash22]
The research community related to piezoelectric energyharvesting is very wide and includes researchers frommechanical electrical materials and civil engineering Thishas led to the development of many different models for thesame problem a piezoelectric cantilevered beam Most ofthese models use simplifications such as oversimplifying ofthe piezoelectric coupling as viscous damping [23] using ofthe static deflection of the beam in a dynamic model [24 25]no inclusion of the piezoelectric effect in the mechanicalmodel [26] or approximate solutions as the Rayleigh-Ritzapproximation [27 28] Erturk and Inman [29] providea comprehensive discussion of the different simplificationsfound in literature to model piezoelectric beams highlight-ing the common indiscretions and providing correctionsand necessary clarifications for researchers from differentengineering fields The best solution can be obtained bysolving the continuous electromechanical model analyticallyDifferent approaches for solving the continuousmodel can befound in literature [30ndash33]
In this paper we present a general approach to character-ize multiple piezoelectric beams connected with springs Theanalytical solution of the cantilevered piezoelectric energyharvester with Euler-Bernoulli beam assumptions proposedby Erturk and Inman [33] is extended to the case of an arrayof piezoelectric beams connected by springs The proposedconfiguration is attractive since it combines the properties ofbeam arrays and of systems withmultiple degrees of freedomFurthermore the results demonstrate that connecting anarray of beams with springs allows increasing the frequencyband of operation and increasing the amount of powergenerated when compared with an array of disconnectedbeams
The remainder of this work is organized as followsSection 2 presents the structure under investigation andthe methodology to compute its modes shapes and naturalfrequencies Section 3 provides the derivation of the elec-tromechanical equations of a bimorph cantilever beam InSection 4 the equations of the output voltage and power ofan array of bimorph cantilever beams connected by springsare derived Section 5 presents analytical and experimentalresults for different combinations of tip masses and springsconstants comparing the results with those of an array ofdisconnected beams (with no springs) Finally conclusionsand forthcoming work are presented in Section 6
Shock and Vibration 3
Mt1
Mt2
MtN
k1
k2
kN
(a) Full system
kiminus1
kiwi(x t)
Mtiminus1
Mti
Mti+1
X
(b) Scheme of the 119894th beam
Figure 2 Structure consisting in an array of piezoelectric beams connected by springs
2 Structure and Vibration Modes
The structure under investigation is an array of piezoelectricbeams connected by springs as illustrated in Figure 2(a) Allbeams are clamped at the left-end to a vertical wallThe right-end of each beam is connected to a concentrated tip massand to its neighbors beams by springs The inertia of theconcentrated masses is neglected All beams have identicaldimensions and the only differences are the tip masses
To derive the vibration modes of this system it isnecessary to study the boundary conditions of each beam sep-arately Looking at the 119894th beam as shown in Figure 2(b) andassuming a separation-of-variables solution the transversedisplacement of the 119894th beam can be expressed as 119908
119894(119909 119905) =
120601119894(119909)120578(119905) The function 120601
119894(119909) represents the mode shape and
depends on the boundary conditions In an Euler-Bernoullibeam the general solution for 120601
119894(119909) is
120601119894 (119909) = 120582
119894
1sin (120573119909) + 120582119894
2cos (120573119909) + 120582119894
3sinh (120573119909)
+ 120582119894
4cosh (120573119909)
(1)
The boundary conditions at the left-end are
120601119894 (0) = 0
120597120601119894 (0)
120597119909= 0
(2)
which implies that
120582119894
4= minus120582119894
2
120582119894
3= minus120582119894
1
(3)
Therefore the mode shape can be written as
120601119894 (119909) = 120582
119894
1[sin (120573119909) minus sinh (120573119909)]
+ 120582119894
2[cos (120573119909) minus cosh (120573119909)]
(4)
At the right-end the bending moment and shear forcemust satisfy the following conditions
1198841198681205972119908119894
1205971199092= 0
120597
120597119909[119884119868
1205972119908119894
1205971199092] = minus119872
119905119894
1205972119908119894
1205971199052minus 119896119894(119908119894minus 119908119894+1)
minus 119896119894minus1(119908119894minus 119908119894minus1)
(5)
where 119884 is Youngrsquos modulus 119868 is the area moment of inertiaof the cross section 119872
119905119894is the 119894th tip mass and 119896
119894is the
stiffness of the 119894th spring Replacing that 119908119894(119909 119905) = 120601
119894(119909)120578(119905)
and 1205972120578(119905)1205971199052 = minus1205962119899120578(119905) 120596
119899being the natural frequency
yields
1205972120601119894 (119871)
1205971199092= 0
1198841198681205973120601119894 (119871)
1205971199093= 1205962
119899119872119905119894120601119894 (119871) minus 119896119894 [120601119894 (119871) minus 120601119894+1 (119871)]
minus 119896119894minus1[120601119894 (119871) minus 120601119894minus1 (119871)]
(6)
4 Shock and Vibration
where 119871 is the length of the beam Replacing the definition of120601119894(119909) from (4) the following equations are obtained
[
0 0 1205721
1205722
0 0
1205723(119896119894minus1) 1205724(119896119894minus1) 1205725(119872119905119894 119896119894 119896119894minus1) 1205726(119872119905119894 119896119894 119896119894minus1) 1205723(119896119894) 1205724(119896119894)]
[[[[[[[[[[[[
[
120582119894minus1
1
120582119894minus1
2
120582119894
1
120582119894
2
120582119894+1
1
120582119894+1
2
]]]]]]]]]]]]
]
=
[[[[[[[[[[[
[
0
0
0
0
0
0
]]]]]]]]]]]
]
(7)
with
1205721= sin (120573119871) + sinh (120573119871)
1205722= cos (120573119871) + cosh (120573119871)
1205723(119896119894) = 119896119894(sinh (120573119871) minus sin (120573119871))
1205724(119896119894) = 119896119894(cosh (120573119871) minus cos (120573119871))
1205725(119872119905119894 119896119894 119896119894minus1)
= (119896119894+ 119896119894minus1minus1198721199051198941205962
119899) (sin (120573119871) minus sinh (120573119871))
+ 1198841198681205733(cos (120573119871) + cosh (120573119871))
1205726(119872119905119894 119896119894 119896119894minus1)
= (119896119894+ 119896119894minus1minus1198721199051198941205962
119899) (cos (120573119871) minus cosh (120573119871))
+ 1198841198681205733(sinh (120573119871) minus sin (120573119871))
(8)
Equation (7) can be obtained for each of the119873 beams inthe structure Particular cases are the first and last beams Inthe case of the first beam there is no previous neighbor thusthe following equation is obtained
[
1205721
1205722
0 0
1205725(1198721199051 1198961 0) 120572
6(1198721199051 1198961 0) 120572
3(1198961) 1205724(1198961)]
[[[[[[
[
1205821
1
1205821
2
1205822
1
1205822
2
]]]]]]
]
=
[[[[[
[
0
0
0
0
]]]]]
]
(9)
whereas for the last beam there is no next neighbor and theresulting equations are
[
0 0 1205721
1205722
1205723(119896119873minus1) 1205724(119896119873minus1) 1205725(119872119905119873 0 119896119873minus1) 1205726(119872119905119873 0 119896119873minus1)]
[[[[[[
[
120582119873minus1
1
120582119873minus1
2
120582119873
1
120582119873
2
]]]]]]
]
=
[[[[[
[
0
0
0
0
]]]]]
]
(10)
Coupling all equations yields a system of 2times119873 equationswhich can be written as the following matrix equation
C120582 = 0 (11)
where C is a coefficient matrix and 120582 = [12058211 1205821
2 1205822
1 1205822
2
120582119873
1 120582119873
2]119879 This equation can have nonzero solution for 120582 only
if the determinant of the coefficient matrix vanishes (ie thecoefficient matrix is singular) Setting the determinant of Cequal to zero yields the characteristic equation that is satisfiedby an infinite number of choices for 120573 denoted by 120573
119903 The
vector of unknowns 120582 is calculated by replacing 120573119903in the
coefficient matrix C and then computing its null space Foreach value of 120573
119903a different vector 120582
119903is obtained
Given 120573119903and 120582
119903 the mode shapes are given by
120601119903119894 (119909) = 120582
119894
1199031[sin (120573
119903119909) minus sinh (120573
119903119909)]
+ 120582119894
1199032[cos (120573
119903119909) minus cosh (120573
119903119909)]
(12)
The undamped natural frequency associated with the 119903thmode is given by
120596119899119903= 1205732
119903radic119884119868
119898 (13)
where119898 is the mass per unit length As an example Figure 3illustrates the vibration modes of three elastically connected
Shock and Vibration 5
120573rL = 1536
(a) Mode 1
120573rL = 1573
(b) Mode 2
120573rL = 1640
(c) Mode 3
120573rL = 4192
(d) Mode 4 (e) Mode 5 (f) Mode 6
(g) Mode 7 (h) Mode 8 (i) Mode 9
Figure 3 First nine vibration modes of three elastically connected beams with identical tip masses and spring constants (119872119905119894=13
=
03119898119871 11989612= 03119864119868119871
3)
beams In this example all beams have a mass tip of 03119898119871and all springs have a stiffness of 031198841198681198713
3 Electromechanical Model of a BimorphPiezoelectric Beam
31 Mechanical Equation Let us consider the bimorph can-tilever beam shown in Figure 4 The bimorph beam consistsin two layers of piezoelectric material (PZT) bonded to asubstructure layer Assuming an Euler-Bernoulli beam theequation of motion can be written as [33]
1205972119872(119909 119905)
1205971199092+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
= minus [119898 +119872119905120575 (119909 minus 119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(14)
b
x(t)
Mt
hp
hp
hs
PZT
PZTSubstructureY0e
i120596t
Figure 4 Scheme of the bimorph piezoelectric beam
where 119908rel(119909 119905) is the transverse deflection of the beamrelative to the base 119908
119887(119905) is the base displacement119872(119909 119905) is
the internal bending moment 119888119904is the equivalent coefficient
of strain rate damping 119888119886is the viscous air damping 119868 is the
area moment of inertia of the cross section119898 is the mass perunit length 119872
119905is the tip mass and 120575(119909) is the Dirac delta
function
6 Shock and Vibration
The internal moment is given by the integral of the stressdistribution over the cross section of the beam shown in
119872(119909 119905) = minus119887 [int
minusℎ1199042
minusℎ119901minusℎ1199042
119879119901119910119889119910 + int
ℎ1199042
minusℎ1199042
119879119904119910119889119910
+ int
ℎ1199042+ℎ119901
ℎ1199042
119879119901119910119889119910]
(15)
where 119887 is the width of the beam ℎ119901is the thickness of
each PZT layer ℎ119904is the thickness of the substructure (see
Figure 4) and 119879 is the stress Superscripts 119901 and 119904 stand forPZT and substructure respectively The stress distributionacross the section is given by the strain-stress relations of thesubstructure and PZT layers which are given by
119879119904= 119884119904119878119904 (16)
119879119901= 119884119901119878119901minus 11988931119884119901119864 (17)
where 119878 is the strain 119884 is Youngrsquos modulus 119889 is thepiezoelectric constant and 119864 is the electric fieldThe bendingstrain can be expressed in terms of the radius of curvature as
119878 (119909 119910 119905) = minus1199101205972119908rel (119909 119905)
1205971199092 (18)
Since the piezoelectric layers are connected in parallelthey have the same voltage but the instantaneous electricfields are in opposite directions The electric field can bewritten in terms of the voltage across the PZT V(119905) and thethickness of the PZT layer ℎ
119901 by the following equation
1198641 (119905) = minus
V (119905)ℎ119901
in the top layer
1198642 (119905) =
V (119905)ℎ119901
in the bottom layer(19)
Replacing (16) to (19) in (15) we obtain
119872(119909 119905) =2119887
3[119884119904ℎ3
119904
12+ 119884119901((ℎ119901+ℎ119904
2)
3
minusℎ3
119904
8)]
sdot1205972119908rel (119909 119905)
1205971199092
+11988931119884119901119887
ℎ119901
[ℎ2
119904
4minus (ℎ119901+ℎ119904
2)
2
] V (119905)
(20)
Equation (20) can be expressed as
119872(119909 119905) = 1198841198681205972119908rel (119909 119905)
1205971199092
+ 120599V (119905) [Φ (119909) minus Φ (119909 minus 119871)] (21)
where 119871 is the length of the beam and Φ the Heavisidefunction It has been assumed that the PZT layer covers
the entire length of the beam The constant 119884119868 represents thebending stiffness of the beam and is given by
119884119868 =2119887
3[119884119904ℎ3
119904
12+ 119884119901((ℎ119901+ℎ119904
2)
3
minusℎ3
119904
8)] (22)
and the term 120599 represents the coupling between the strain andthe voltage and is given by
120599 =11988931119884119901119887
ℎ119901
[ℎ2
119904
4minus (ℎ119901+ℎ119904
2)
2
] (23)
Substituting (21) in (14) we obtain the mechanical equa-tion of motion with electrical coupling
1198841198681205974119908rel (119909 119905)
1205971199094+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
+ 120599V (119905) [120597120575 (119909)
120597119909minus120597120575 (119909 minus 119871)
120597119909]
= minus [119898 +119872119905120575 (119909 minus 119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(24)
32 Electrical Equation To derive the electrical equation letus start with the piezoelectric constitutive relation given in
119863 = 11988931119879119901+ 120576119879
33119864 (25)
where119863 is the electric displacement and 12057611987933is the permittivity
at constant stress Replacing the definition of 119879119901 from (17)yields
119863(119909 119905) = 11988931 (119884119901119878119901minus 11988931119864) + 120576
119879
33119864
= 11988931119884119901119878119901+ 119864 (120576
119879
33minus 1198892
31119884119901)
= 11988931119884119901119878119901minus 120576119878
33
V (119905)ℎ119901
(26)
with 12057611987833= 1198892
31119884119901minus 120576119879
33 The strain 119878119901 evaluated at the middle
of the PZT layer is given by
119878119901= minusℎ119901119888
1205972119908rel (119909 119905)
1205971199092 (27)
where ℎ119901119888= ℎ1199042 + ℎ
1199012 is the distance of the center of the
PZT layer to the neural axis Thus
119863 (119909 119905) = minus11988931119884119901ℎ119901119888
1205972119908rel (119909 119905)
1205971199092minus 120576119878
33
V (119905)ℎ119901
(28)
The electric charge is obtained by integrating the electricdisplacement over the area of the beamThe resulting expres-sion is
119902 (119905) = minusint
119871
0
(11988931119884119901ℎ119901119888
1205972119908rel (119909 119905)
1205971199092+ 120576119878
33
V (119905)ℎ119901
)119887119889119909 (29)
Shock and Vibration 7
Therefore the current generated by one PZT layer is
119894 (119905) =120597119902 (119905)
120597119905
= minus120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(30)
Since both layers are connected in parallel the currentgenerated by the bimorph beam is given by
119894 (119905) = 1198941 (119905) + 1198942 (119905)
= minus2120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 2int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(31)
If the beam is connected directly to a resistive load the voltageis given by V(119905) = 119877
119897119894(119905) Substituting this relation in (31) yields
V (119905) = minus2119877119897 [120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
+ int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909]
(32)
The electrical equation with mechanical coupling isobtained by rearranging (32)
V (119905)2119877119897
+120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
= minusint
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(33)
4 Voltage and Power Output of an Array ofPiezoelectric Beams Connected by Springs
Equations (24) and (33) are the electromechanical equationsfor a single bimorph beam under transverse vibrations Letus consider the case of multiple beams connected by springswith a parallel electrical connection as shown in Figure 5 Inthat case the mechanical equation of motion is given by
1198841198681205974119908rel (119909 119905)
1205971199094+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
+ 120599
119873
sum
119894=1
V119894 (119905) [
120597120575 (119909 minus (119894 minus 1) 119871)
120597119909minus120597120575 (119909 minus 119894119871)
120597119909]
= minus[119898 +
119873
sum
119894=1
119872119905119894120575 (119909 minus 119894119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(34)
(t)Rl
Figure 5 A parallel electrical connection of an array of bimorphbeams connected by springs
where V119894(119905) is the voltage of the 119894th beam Since the beams are
electrically connected in parallel they have the same voltagethus V
1(119905) = V
2(119905) = sdot sdot sdot = V
119899(119905) = V(119905) It has been assumed
that all the beams are identical except for the tip mass Itshould be noted that the springs are included as boundaryconditionsTherefore they affect the vibrationmodes but notthe equation of motion
The solution for the relative motion of the system can berepresented by a sum of its mode shapes as
119908rel (119909 119905) = sum119903
120601119903 (119909) 120578119903 (119905) (35)
where 120601119903(119909) and 120578
119903(119905) are the 119903th mode shape and modal
coordinate respectively The 119903th mode shape can be definedby segments as
120601119903 (119909)
=
1206011199031 (119909) 0 le 119909 le 119871
1206011199032 (119909 minus 119871) 119871 le 119909 le 2119871
120601119903119873 (119909 minus (119873 minus 1) 119871) (119873 minus 1) 119871 le 119909 le 119873119871
(36)
where 120601119903119894
is the 119903th mode shape of the 119894th beam 119871 is thelength of one beam and119873 is the number of beams
Replacing (35) themechanical equation ofmotion can bewritten as
[1205962
119903120578119903 (119905) + 2120585119903120596119903 120578119903 (119905) + 120578
119903 (119905)] 120572119903 +
119873
sum
119894=1
120594119903119894V (119905)
= [minus1198981205972119908119887 (119909 119905)
1205971199052minus 119888119886
120597119908119887 (119909 119905)
120597119905] 120574119903
minus1205972119908119887 (119909 119905)
1205971199052
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(37)
8 Shock and Vibration
with
120594119903119894= 120599(minus120601
119903119894 (119871) +120597120601119903119894 (119871)
120597119909)
120585119903=119888119904119868120596119903
2119884119868+
119888119886
2119898120596119903
120572119903= int
119873119871
0
1198981206012
119903(119909) 119889119909
120574119903= int
119873119871
0
120601119903 (119909) 119889119909
(38)
Assuming a harmonic basemotion119908119887(119909 119905) = 119884
0119890119895120596119905 and
a harmonic response 120578119903(119905) = 119873
119903119890119895120596119905 and V(119905) = 119881119890119895120596119905 (37)
becomes
[1205962
119899119903119873119903+ 2119895120585119903120596119899119903120596119873119903minus 1205962119873119903] 120572119903+ 119881
119873
sum
119894=1
120594119903119894
= (1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(39)
Rearranging (39) the amplitude of the temporal term119873119903
can be written as follows
119873119903
=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871) minus 119881sum
119873
119894=1120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(40)
For simplicity (40) can be expressed as
119873119903= 119886119903minus 119881
119873
sum
119894=1
119887119903119894 (41)
with
119886119903=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871)
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
119887119903119894=
120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(42)
The electrical equation needs to be derived for each beamindividually The relative motion for the 119894th beam is given by
119908rel119894 (119909 119905) = sum119903
120601119903119894 (119909) 120578119903 (119905) (43)
Replacing (43) in (30) yields
119894119894 (119905) = minus2
120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 211988931119884119901ℎ119901119888119887sum
119903
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(44)
If the beams are connected directly to a resistive load thevoltage is given by V(119905) = 119877
119897119894(119905) = 119877
119897sum119894119894119894(119905) substituting this
condition equation (44) becomes
V (119905)2119877119897
= minus120576119878
33119887119871119873
ℎ119901
120597V (119905)120597119905
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(45)
Replacing the harmonic response condition in (45) yields
119881
2119877119897
= minus119895120596120576119878
33119887119871119873
ℎ119901
119881
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
119895120596119873119903
(46)
Rearranging (46) we obtain
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596119873119903 (47)
with
120593119903119894= minus
11988931119884119901ℎ119901119888ℎ119901
120576119878
33119871119873
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120581119888= 119895120596 +
ℎ119901
2119877119897120576119878
33119887119871119873
(48)
Replacing 119873119903from (41) in (47) the following equation is
obtained
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596(119886
119903minus 119881
119873
sum
119896=1
119887119903119896) (49)
Rearranging (49) the voltage amplitude can be written asshown in
119881 =sum119903sum119894120593119903119894119895120596119886119903
120581119888+ sum119903sum119894120593119903119894119895120596sum119873
119896=1119887119903119896
(50)
The voltage FRF is defined as the ratio of the voltageoutput to the base acceleration119881minus1205962119884
0 The instantaneous
power output can be obtained from the relation 119875(119905) =V2(119905)119877
119897Therefore the FRF of the power output is simply the
square of the voltage FRF divided by 119877119897
5 Results
51 Analytical This section presents the results of the powerFRF for different number of beams and different combina-tions of tip masses and spring constants in the array Thegeometric material and electromechanical parameters of thebeams are given in Table 1 These are the parameters of theexperimental beams used in the next section
The voltage is computed by (50) and the power FRF iscomputed as 1198812[119877
119897(12059621198840)2] The value of the external load
Shock and Vibration 9
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(a) Connected
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 6 Power FRF with an increasing number of identical bimorph beams119872119905119894=16
= 03119898119871 and 119896119894=15
= 031198841198681198713
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(a) Connected
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 7 Power FRF with an increasing number of bimorph beams with different tip masses Two beams1198721199051= 03119898119871119872
1199052= 035119898119871
119896 = 031198841198681198713 three beams119872
1199051= 025119898119871119872
1199052= 030119898119871119872
1199053= 035119898119871 119896
12= 03119884119868119871
3 six beams1198721199051= 015119898119871119872
1199052= 020119898119871
1198721199053= 030119898119871119872
1199054= 035119898119871119872
1199055= 040119898119871119872
1199056= 045119898119871 119896
119894=15= 03119884119868119871
3
Table 1 Parameters of the bimorph beams
Length of the beam 119871 (mm) 482Width of the beam 119887 (mm) 10Thickness of the substructure ℎ
119904(mm) 019
Thickness of the PZT ℎ119901(mm) 024
Youngrsquos modulus of the substructure 119884119904(GPa) 404
Youngrsquos modulus of the PZT 119884119901(GPa) 55
Mass density of the substructure 120588119904(kgm3) 1519
Mass density of the PZT 120588119901(kgm3) 8140
Piezoelectric constant 11988931(pmV) minus180
Permittivity 12057611990433(nFm) 1328
Strain rate damping 119888119904119868 (kgm4s) 116 times 10minus6
Viscous air damping 119888119886(kgs) 04
resistance is defined for each case as the optimum resistorthat maximizes the power output To study the advantagesand disadvantages of the proposed system the results arecompared with those obtained by an array of disconnectedbeams (with no springs)
First the case of identical beams is investigated theresults are shown in Figure 6(a) It is obtained that in thecase of identical beams the only modes that are excitedare the modes where all beams move in phase such asmodes 1 4 and 7 in Figure 3 In this case it is also foundthat the natural frequencies do not depend on the springstiffness because all beams are in phase and the springs donot deform Furthermore the results are exactly the same asthe ones obtained with disconnected beams as illustrated inFigure 6(b)
In the case of beams with different tip masses the resultsof the FRF power do depend on the springs connecting thebeams as shown in Figure 7 If the beams are disconnectedand new beams with different tip masses are added to thearray then the frequency band of the device is increasedobtaining a similar power amplitude for all resonant frequen-cies On the other hand if the beams are connected by springand new beams are included in the array then the frequencyband is increased but the amplitude related to each resonantfrequency varies
For an array of piezoelectric beams connected by springsthe frequency band of operation and the power amplitude
10 Shock and Vibration
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
01mL 02mL01mL 03mL
01mL 04mL01mL 05mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k = 02YIL3k = 03YIL3
k = 04YIL3
(b) Varying spring constants (1198721199051 = 03119898119871 1198721199052 = 04119898119871)
Figure 8 Power FRF of two bimorph beams with different combinations of tip masses and spring constants
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
02mL 03mL 04mL02mL 02mL 04mL
03mL 03mL 04mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k =
k =k = 02YIL3 02YIL302YIL3 04YIL3
02YIL3 06YIL3
(b) Varying spring constants (1198721199051 = 02119898119871 1198721199052 = 03119898119871 1198721199053 =04119898119871)
Figure 9 Power FRF of three bimorph beams with different combinations of tip masses and spring constants
associated with each resonant frequency depend on thecombination of tipmasses and springs constantsThis ismoreevident in Figures 8 and 9 where different combinations ofsprings and tip masses are studied for the case of two andthree beams In general increasing the tip mass of the beamsdecreases someof the resonant frequencies and increasing thespring stiffness increases the frequency band
To compare the performances of an array of beamsconnected by springs and one of the disconnected beams acase of ten beams with both situations is plotted in Figure 10It is possible to see that the array with connected beamshas a slightly wider frequency band which increases withan increment of the springs constants Furthermore bycomparing the areas below both curves in the frequencyrange 50ndash200Hz the array with connected beams providesan area between 20 and 36 larger than the array withdisconnected beams Thus by connecting the beams theperformance of the array can be improved It should benoted that the previous results are obtained with arbitraryvalues for the tip masses and spring constantsTherefore thisperformance can be further improved by a proper design
52 Experimental To validate the proposed model a pre-liminary experimental campaign was performed Figure 11presents the experimental setup It consists of a signal genera-tor connected to a power amplifier and to an electrodynamicshaker which produces a controlled base vibration An arrayof three piezoelectric bimorph beams is mounted over theshaker Figure 12 shows this array The cantilever bimorphbeams have a tip mass at the free-end and can be connectedwith springs between them Three experiments were per-formed In the first case no springs were used and eachbeam was connected independently to a 1 kΩ resistive loadin this case the voltage generated by each beam is monitoredindividually In the second case no springs were used andthe beams were electrically connected in parallel to a resistiveload of 1 kΩ In the last case the beams were connected withsprings between them and they were electrically connected inparallel to a resistive load of 1 kΩ
The properties of the bimorph beams are the ones givenin Table 1 the tip masses are119872
1199051= 65119898119871119872
1199052= 35119898119871
1198721199053= 19119898119871 and the spring constants are 119896
1= 1198962=
2921198841198681198713
Shock and Vibration 11
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected02YIL3
(a) 119896119894=19 = 021198841198681198713
50 100 150 200
Frequency (Hz)
10minus4
Pow
er F
RF (W
s4m
2)
10minus3
10minus2
10minus1
100
Connected k =
Disconnected03YIL3
(b) 119896119894=19 = 031198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected04YIL3
(c) 119896119894=19 = 041198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected05YIL3
(d) 119896119894=19 = 051198841198681198713
Figure 10 Power FRF with ten bimorph beams1198721199051= 01119898119871119872
1199052= 015119898119871119872
1199053= 02119898119871119872
1199054= 025119898119871119872
1199055= 03119898119871119872
1199056= 035119898119871
1198721199057= 04119898119871119872
1199058= 045119898119871119872
1199059= 05119898119871119872
11990510= 055119898119871
Figure 11 Experimental setup
Figure 12 Experimental bimorph beams
During the three experiments the system is subjected toa harmonic base displacement with frequencies in the range20ndash70Hz The results for the first experiment are presentedin Figure 13(a) and are compared with the ones obtainedby an analytical model in Figure 13(b) Both analytical andexperimental curves are very similar The only difference isthat the experimental power increases from beams 1 to 3 andthe analytical model predicted the opposite behavior Thisunexpected behavior seems to be related to a resonance ofthe supporting structure in the three experimental curves itis possible to see an increment of the power near 60Hz whichcan be explained by a resonance of the structure
Figure 14 presents the experimental and analytical resultsfor the second and third experiments It is possible to seea good agreement between the experimental and analyticalcurves although similarly to the results in Figure 13(a) thereis an increment of the power near 60Hz that confirms astructural resonance
Comparing the areas below both experimental curves inFigure 13(b) in the frequency range 20ndash70Hz the array withconnected beams provides an area 5 larger than the arraywith disconnected beamswhich confirms that the addition ofsprings can increment the generated powerThese results canbe improved by using softer springs which have lower energylosses and by solving the structural resonance problem
12 Shock and Vibration
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 13 Power FRF for the three beams with no springs and independent electrical connections to 1 kΩ resistive loads (a) experimentalmeasurement (b) analytical results
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 14 Power FRF for the three beams with and without springs and electrical connection in parallel to a resistive load of 1 kΩ(a) experimental measurement (b) analytical results
6 Conclusions
This paper investigated the use of an array of bimorphbeams connected by springs as a broadband vibration-basedenergy harvester The equations for the voltage and poweroutput of the full system have been deduced based on theanalytical solution of the cantilevered piezoelectric energyharvester with Euler-Bernoulli beam assumptions To studythe advantages and disadvantages of the proposed system theresults are compared with those of an array of disconnectedbeams (with no springs) The analytical model is validatedwith experimental measurements of three bimorph beamswith and without springs
Both analytical and experimental results show that con-necting the array of beams with springs allows increasingthe frequency band of operation and increasing the amountof power generated Thus the maximal frequency band fora certain array of beams can be extended by connectingthem by springs and at the same time the output power isincreased The performance of this system can be optimizedfor specific applications by a proper selection of the tipmassesand spring constants
The preliminary experimental results are consistent withthe analytical ones but further research is necessary toimprove the experimental setup and to investigate differentexperimental conditions such as variations of the springsstiffness
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
References
[1] S R Anton and H A Sodano ldquoA review of power harvestingusing piezoelectric materials (2003ndash2006)rdquo Smart Materialsand Structures vol 16 no 3 article R1 2007
[2] S Chalasani and J M Conrad ldquoA survey of energy harvestingsources for embedded systemsrdquo in Proceedings of the IEEESoutheastCon pp 442ndash447 April 2008
[3] K A Cook-Chennault NThambi and AM Sastry ldquoPoweringMEMS portable devicesmdasha review of non-regenerative andregenerative power supply systems with special emphasis on
Shock and Vibration 13
piezoelectric energy harvesting systemsrdquo Smart Materials andStructures vol 17 no 4 Article ID 043001 2008
[4] A Dewan S U Ay M N Karim and H Beyenal ldquoAlternativepower sources for remote sensors a reviewrdquo Journal of PowerSources vol 245 pp 129ndash143 2014
[5] A Harb ldquoEnergy harvesting state-of-the-artrdquo RenewableEnergy vol 36 no 10 pp 2641ndash2654 2011
[6] S Saadon and O Sidek ldquoA review of vibration-based MEMSpiezoelectric energy harvestersrdquo Energy Conversion and Man-agement vol 52 no 1 pp 500ndash504 2011
[7] H A Sodano D J Inman and G Park ldquoA review of powerharvesting from vibration using piezoelectric materialsrdquo Shockand Vibration Digest vol 36 no 3 pp 197ndash205 2004
[8] R J M Vullers R van Schaijk I Doms C Van Hoofand R Mertens ldquoMicropower energy harvestingrdquo Solid-StateElectronics vol 53 no 7 pp 684ndash693 2009
[9] R Calio U B Rongala D Camboni et al ldquoPiezoelectric energyharvesting solutionsrdquo Sensors vol 14 no 3 pp 4755ndash4790 2014
[10] J Twiefel and H Westermann ldquoSurvey on broadband tech-niques for vibration energy harvestingrdquo Journal of IntelligentMaterial Systems and Structures vol 24 no 11 pp 1291ndash13022013
[11] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavengingrdquo Journal of Sound and Vibration vol 292no 3ndash5 pp 987ndash998 2006
[12] S M Shahruz ldquoLimits of performance of mechanical band-pass filters used in energy scavengingrdquo Journal of Sound andVibration vol 293 no 1-2 pp 449ndash461 2006
[13] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavenging multi-degree-of-freedom modelsrdquo Journalof Vibration and Control vol 14 no 5 pp 753ndash768 2008
[14] J-Q Liu H-B Fang Z-Y Xu et al ldquoA MEMS-based piezo-electric power generator array for vibration energy harvestingrdquoMicroelectronics Journal vol 39 no 5 pp 802ndash806 2008
[15] HC Lin PHWu I C Lien andYC Shu ldquoAnalysis of an arrayof piezo-electric energy harvesters connected in seriesrdquo SmartMaterials and Structures vol 22 no 9 Article ID 094026 2013
[16] S-J Jang E Rustighi M J Brennan Y P Lee and H-JJung ldquoDesign of a 2DOF vibrational energy harvesting devicerdquoJournal of IntelligentMaterial Systems and Structures vol 22 no5 pp 443ndash448 2011
[17] L Tang and Y Yang ldquoA multiple-degree-of-freedom piezoelec-tric energy harvesting modelrdquo Journal of Intelligent MaterialSystems and Structures vol 23 no 14 pp 1631ndash1647 2012
[18] Z Yang and J Yang ldquoConnected vibrating piezoelectricbimorph beams as a wide-band piezoelectric power harvesterrdquoJournal of IntelligentMaterial Systems and Structures vol 20 no5 pp 569ndash574 2009
[19] W Zhou G R Penamalli and L Zuo ldquoAn efficient vibrationenergy harvester with a multi-mode dynamic magnifierrdquo SmartMaterials and Structures vol 21 no 1 Article ID 015014 2012
[20] D S Nguyen E Halvorsen G U Jensen and A VoglldquoFabrication and characterization of a widebandMEMS energyharvester utilizing nonlinear springsrdquo Journal of Micromechan-ics andMicroengineering vol 20 no 12 Article ID 125009 2010
[21] S D Nguyen and E Halvorsen ldquoNonlinear springs forbandwidth-tolerant vibration energy harvestingrdquo Journal ofMicroelectromechanical Systems vol 20 no 6 pp 1225ndash12272011
[22] S BoisseauGDespesse andBA Seddik ldquoNonlinear h-shapedsprings to improve efficiency of vibration energy harvestersrdquo
Journal of Applied Mechanics vol 80 no 6 Article ID 0610132013
[23] S Priya ldquoAdvances in energy harvesting using low profilepiezoelectric transducersrdquo Journal of Electroceramics vol 19 no1 pp 165ndash182 2007
[24] S Roundy P K Wright and J Rabaey ldquoA study of lowlevel vibrations as a power source for wireless sensor nodesrdquoComputer Communications vol 26 no 11 pp 1131ndash1144 2003
[25] S Roundy and P K Wright ldquoA piezoelectric vibration basedgenerator for wireless electronicsrdquo Smart Materials and Struc-tures vol 13 no 5 pp 1131ndash1142 2004
[26] H A Sodano G Park and D J Inman ldquoEstimation of electriccharge output for piezoelectric energy harvestingrdquo Strain vol40 no 2 pp 49ndash58 2004
[27] N E Dutoit B L Wardle and S-G Kim ldquoDesign considera-tions for mems-scale piezoelectric mechanical vibration energyharvestersrdquo Integrated Ferroelectrics vol 71 no 1 pp 121ndash1602005
[28] S-N Chen G-J Wang andM-C Chien ldquoAnalytical modelingof piezoelectric vibration-induced micro power generatorrdquoMechatronics vol 16 no 7 pp 379ndash387 2006
[29] A Erturk and D J Inman ldquoIssues in mathematical modelingof piezoelectric energy harvestersrdquo Smart Materials and Struc-tures vol 17 no 6 Article ID 065016 2008
[30] F Lu H P Lee and S P Lim ldquoModeling and analysis of micropiezoelectric power generators for micro-electromechanical-systems applicationsrdquo Smart Materials and Structures vol 13no 1 pp 57ndash63 2004
[31] S Jiang X Li S Guo Y Hu J Yang andQ Jiang ldquoPerformanceof a piezoelectric bimorph for scavenging vibration energyrdquoSmartMaterials and Structures vol 14 no 4 pp 769ndash774 2005
[32] A Erturk and D J Inman ldquoA distributed parameter elec-tromechanical model for cantilevered piezoelectric energy har-vestersrdquo Journal of Vibration and Acoustics vol 130 no 4Article ID 041002 2008
[33] A Erturk and D J Inman ldquoAn experimentally validatedbimorph cantilever model for piezoelectric energy harvestingfrom base excitationsrdquo Smart Materials and Structures vol 18no 2 Article ID 025009 2009
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
2 Shock and Vibration
Mt Rl (t)
+
minus
PiezoceramicSubstructure
(a)
Mt Rl (t)
+
minus
PiezoceramicSubstructure
(b)
Figure 1 Cantilever piezoelectric energy configurations (a) unimorph (b) bimorph (parallel connection)
an operating frequency equal to their resonance frequencyand if the excitation is slightly different there is a significantdrop in the level of power generated This problem has ledto the search for broadband energy harvesters Twiefel andWestermann [10] present a survey of broadband techniquesfor vibration-based energy harvesting These techniques areclassified as follows
(i) Linear generators arrays combined modes
(ii) Nonlinear generators bimultiple stable vibroimpactresonant tuning
(iii) Advanced electronic networks switching networkssynthetic impedance
According to the researchers there is no universal solutiondepending on the application one technique may be betterthan the others For example linear generators are moreefficient for stochastic excitations whereas nonlinear tech-niques are more suitable for harmonic excitations or very lowexcitation frequencies
In the case of linear generators recent research hasbeen focused on developing systems that exhibit multipleresonances that are closely spaced The simplest solution isto use multiple cantilevers The cantilevers are built withdifferent geometries (large tip mass or width) so each beamhas a slightly different resonant frequency thus increasing thebandwidth of the full system A first design of a cantileverarray was presented by Shahruz [11 12] He concluded thatif the dimensions of the beams and masses are chosenappropriately then the system performs as a broadbandharvester However the maximal frequency band is limitedand is independent of the beams dimensions This approachhas been studied extensively since then [13ndash15]
Another method is to use structures with two or moredegrees of freedom resulting in two or more resonantfrequencies that are closely spaced [16ndash19] Z Yang andJ Yang [18] analyzed the power generated by the flexuralvibrations of two piezoelectric beams connected by springsthey showed that this structure can be used as a broadbandenergy harvester Nevertheless the model presented by theauthors is restricted to two beams and the extension tomore than two beams is not trivial Springs have also beenused in MEMS-based energy harvesters where it has been
demonstrated that the use of nonlinear springs increases thegenerated power [20ndash22]
The research community related to piezoelectric energyharvesting is very wide and includes researchers frommechanical electrical materials and civil engineering Thishas led to the development of many different models for thesame problem a piezoelectric cantilevered beam Most ofthese models use simplifications such as oversimplifying ofthe piezoelectric coupling as viscous damping [23] using ofthe static deflection of the beam in a dynamic model [24 25]no inclusion of the piezoelectric effect in the mechanicalmodel [26] or approximate solutions as the Rayleigh-Ritzapproximation [27 28] Erturk and Inman [29] providea comprehensive discussion of the different simplificationsfound in literature to model piezoelectric beams highlight-ing the common indiscretions and providing correctionsand necessary clarifications for researchers from differentengineering fields The best solution can be obtained bysolving the continuous electromechanical model analyticallyDifferent approaches for solving the continuousmodel can befound in literature [30ndash33]
In this paper we present a general approach to character-ize multiple piezoelectric beams connected with springs Theanalytical solution of the cantilevered piezoelectric energyharvester with Euler-Bernoulli beam assumptions proposedby Erturk and Inman [33] is extended to the case of an arrayof piezoelectric beams connected by springs The proposedconfiguration is attractive since it combines the properties ofbeam arrays and of systems withmultiple degrees of freedomFurthermore the results demonstrate that connecting anarray of beams with springs allows increasing the frequencyband of operation and increasing the amount of powergenerated when compared with an array of disconnectedbeams
The remainder of this work is organized as followsSection 2 presents the structure under investigation andthe methodology to compute its modes shapes and naturalfrequencies Section 3 provides the derivation of the elec-tromechanical equations of a bimorph cantilever beam InSection 4 the equations of the output voltage and power ofan array of bimorph cantilever beams connected by springsare derived Section 5 presents analytical and experimentalresults for different combinations of tip masses and springsconstants comparing the results with those of an array ofdisconnected beams (with no springs) Finally conclusionsand forthcoming work are presented in Section 6
Shock and Vibration 3
Mt1
Mt2
MtN
k1
k2
kN
(a) Full system
kiminus1
kiwi(x t)
Mtiminus1
Mti
Mti+1
X
(b) Scheme of the 119894th beam
Figure 2 Structure consisting in an array of piezoelectric beams connected by springs
2 Structure and Vibration Modes
The structure under investigation is an array of piezoelectricbeams connected by springs as illustrated in Figure 2(a) Allbeams are clamped at the left-end to a vertical wallThe right-end of each beam is connected to a concentrated tip massand to its neighbors beams by springs The inertia of theconcentrated masses is neglected All beams have identicaldimensions and the only differences are the tip masses
To derive the vibration modes of this system it isnecessary to study the boundary conditions of each beam sep-arately Looking at the 119894th beam as shown in Figure 2(b) andassuming a separation-of-variables solution the transversedisplacement of the 119894th beam can be expressed as 119908
119894(119909 119905) =
120601119894(119909)120578(119905) The function 120601
119894(119909) represents the mode shape and
depends on the boundary conditions In an Euler-Bernoullibeam the general solution for 120601
119894(119909) is
120601119894 (119909) = 120582
119894
1sin (120573119909) + 120582119894
2cos (120573119909) + 120582119894
3sinh (120573119909)
+ 120582119894
4cosh (120573119909)
(1)
The boundary conditions at the left-end are
120601119894 (0) = 0
120597120601119894 (0)
120597119909= 0
(2)
which implies that
120582119894
4= minus120582119894
2
120582119894
3= minus120582119894
1
(3)
Therefore the mode shape can be written as
120601119894 (119909) = 120582
119894
1[sin (120573119909) minus sinh (120573119909)]
+ 120582119894
2[cos (120573119909) minus cosh (120573119909)]
(4)
At the right-end the bending moment and shear forcemust satisfy the following conditions
1198841198681205972119908119894
1205971199092= 0
120597
120597119909[119884119868
1205972119908119894
1205971199092] = minus119872
119905119894
1205972119908119894
1205971199052minus 119896119894(119908119894minus 119908119894+1)
minus 119896119894minus1(119908119894minus 119908119894minus1)
(5)
where 119884 is Youngrsquos modulus 119868 is the area moment of inertiaof the cross section 119872
119905119894is the 119894th tip mass and 119896
119894is the
stiffness of the 119894th spring Replacing that 119908119894(119909 119905) = 120601
119894(119909)120578(119905)
and 1205972120578(119905)1205971199052 = minus1205962119899120578(119905) 120596
119899being the natural frequency
yields
1205972120601119894 (119871)
1205971199092= 0
1198841198681205973120601119894 (119871)
1205971199093= 1205962
119899119872119905119894120601119894 (119871) minus 119896119894 [120601119894 (119871) minus 120601119894+1 (119871)]
minus 119896119894minus1[120601119894 (119871) minus 120601119894minus1 (119871)]
(6)
4 Shock and Vibration
where 119871 is the length of the beam Replacing the definition of120601119894(119909) from (4) the following equations are obtained
[
0 0 1205721
1205722
0 0
1205723(119896119894minus1) 1205724(119896119894minus1) 1205725(119872119905119894 119896119894 119896119894minus1) 1205726(119872119905119894 119896119894 119896119894minus1) 1205723(119896119894) 1205724(119896119894)]
[[[[[[[[[[[[
[
120582119894minus1
1
120582119894minus1
2
120582119894
1
120582119894
2
120582119894+1
1
120582119894+1
2
]]]]]]]]]]]]
]
=
[[[[[[[[[[[
[
0
0
0
0
0
0
]]]]]]]]]]]
]
(7)
with
1205721= sin (120573119871) + sinh (120573119871)
1205722= cos (120573119871) + cosh (120573119871)
1205723(119896119894) = 119896119894(sinh (120573119871) minus sin (120573119871))
1205724(119896119894) = 119896119894(cosh (120573119871) minus cos (120573119871))
1205725(119872119905119894 119896119894 119896119894minus1)
= (119896119894+ 119896119894minus1minus1198721199051198941205962
119899) (sin (120573119871) minus sinh (120573119871))
+ 1198841198681205733(cos (120573119871) + cosh (120573119871))
1205726(119872119905119894 119896119894 119896119894minus1)
= (119896119894+ 119896119894minus1minus1198721199051198941205962
119899) (cos (120573119871) minus cosh (120573119871))
+ 1198841198681205733(sinh (120573119871) minus sin (120573119871))
(8)
Equation (7) can be obtained for each of the119873 beams inthe structure Particular cases are the first and last beams Inthe case of the first beam there is no previous neighbor thusthe following equation is obtained
[
1205721
1205722
0 0
1205725(1198721199051 1198961 0) 120572
6(1198721199051 1198961 0) 120572
3(1198961) 1205724(1198961)]
[[[[[[
[
1205821
1
1205821
2
1205822
1
1205822
2
]]]]]]
]
=
[[[[[
[
0
0
0
0
]]]]]
]
(9)
whereas for the last beam there is no next neighbor and theresulting equations are
[
0 0 1205721
1205722
1205723(119896119873minus1) 1205724(119896119873minus1) 1205725(119872119905119873 0 119896119873minus1) 1205726(119872119905119873 0 119896119873minus1)]
[[[[[[
[
120582119873minus1
1
120582119873minus1
2
120582119873
1
120582119873
2
]]]]]]
]
=
[[[[[
[
0
0
0
0
]]]]]
]
(10)
Coupling all equations yields a system of 2times119873 equationswhich can be written as the following matrix equation
C120582 = 0 (11)
where C is a coefficient matrix and 120582 = [12058211 1205821
2 1205822
1 1205822
2
120582119873
1 120582119873
2]119879 This equation can have nonzero solution for 120582 only
if the determinant of the coefficient matrix vanishes (ie thecoefficient matrix is singular) Setting the determinant of Cequal to zero yields the characteristic equation that is satisfiedby an infinite number of choices for 120573 denoted by 120573
119903 The
vector of unknowns 120582 is calculated by replacing 120573119903in the
coefficient matrix C and then computing its null space Foreach value of 120573
119903a different vector 120582
119903is obtained
Given 120573119903and 120582
119903 the mode shapes are given by
120601119903119894 (119909) = 120582
119894
1199031[sin (120573
119903119909) minus sinh (120573
119903119909)]
+ 120582119894
1199032[cos (120573
119903119909) minus cosh (120573
119903119909)]
(12)
The undamped natural frequency associated with the 119903thmode is given by
120596119899119903= 1205732
119903radic119884119868
119898 (13)
where119898 is the mass per unit length As an example Figure 3illustrates the vibration modes of three elastically connected
Shock and Vibration 5
120573rL = 1536
(a) Mode 1
120573rL = 1573
(b) Mode 2
120573rL = 1640
(c) Mode 3
120573rL = 4192
(d) Mode 4 (e) Mode 5 (f) Mode 6
(g) Mode 7 (h) Mode 8 (i) Mode 9
Figure 3 First nine vibration modes of three elastically connected beams with identical tip masses and spring constants (119872119905119894=13
=
03119898119871 11989612= 03119864119868119871
3)
beams In this example all beams have a mass tip of 03119898119871and all springs have a stiffness of 031198841198681198713
3 Electromechanical Model of a BimorphPiezoelectric Beam
31 Mechanical Equation Let us consider the bimorph can-tilever beam shown in Figure 4 The bimorph beam consistsin two layers of piezoelectric material (PZT) bonded to asubstructure layer Assuming an Euler-Bernoulli beam theequation of motion can be written as [33]
1205972119872(119909 119905)
1205971199092+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
= minus [119898 +119872119905120575 (119909 minus 119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(14)
b
x(t)
Mt
hp
hp
hs
PZT
PZTSubstructureY0e
i120596t
Figure 4 Scheme of the bimorph piezoelectric beam
where 119908rel(119909 119905) is the transverse deflection of the beamrelative to the base 119908
119887(119905) is the base displacement119872(119909 119905) is
the internal bending moment 119888119904is the equivalent coefficient
of strain rate damping 119888119886is the viscous air damping 119868 is the
area moment of inertia of the cross section119898 is the mass perunit length 119872
119905is the tip mass and 120575(119909) is the Dirac delta
function
6 Shock and Vibration
The internal moment is given by the integral of the stressdistribution over the cross section of the beam shown in
119872(119909 119905) = minus119887 [int
minusℎ1199042
minusℎ119901minusℎ1199042
119879119901119910119889119910 + int
ℎ1199042
minusℎ1199042
119879119904119910119889119910
+ int
ℎ1199042+ℎ119901
ℎ1199042
119879119901119910119889119910]
(15)
where 119887 is the width of the beam ℎ119901is the thickness of
each PZT layer ℎ119904is the thickness of the substructure (see
Figure 4) and 119879 is the stress Superscripts 119901 and 119904 stand forPZT and substructure respectively The stress distributionacross the section is given by the strain-stress relations of thesubstructure and PZT layers which are given by
119879119904= 119884119904119878119904 (16)
119879119901= 119884119901119878119901minus 11988931119884119901119864 (17)
where 119878 is the strain 119884 is Youngrsquos modulus 119889 is thepiezoelectric constant and 119864 is the electric fieldThe bendingstrain can be expressed in terms of the radius of curvature as
119878 (119909 119910 119905) = minus1199101205972119908rel (119909 119905)
1205971199092 (18)
Since the piezoelectric layers are connected in parallelthey have the same voltage but the instantaneous electricfields are in opposite directions The electric field can bewritten in terms of the voltage across the PZT V(119905) and thethickness of the PZT layer ℎ
119901 by the following equation
1198641 (119905) = minus
V (119905)ℎ119901
in the top layer
1198642 (119905) =
V (119905)ℎ119901
in the bottom layer(19)
Replacing (16) to (19) in (15) we obtain
119872(119909 119905) =2119887
3[119884119904ℎ3
119904
12+ 119884119901((ℎ119901+ℎ119904
2)
3
minusℎ3
119904
8)]
sdot1205972119908rel (119909 119905)
1205971199092
+11988931119884119901119887
ℎ119901
[ℎ2
119904
4minus (ℎ119901+ℎ119904
2)
2
] V (119905)
(20)
Equation (20) can be expressed as
119872(119909 119905) = 1198841198681205972119908rel (119909 119905)
1205971199092
+ 120599V (119905) [Φ (119909) minus Φ (119909 minus 119871)] (21)
where 119871 is the length of the beam and Φ the Heavisidefunction It has been assumed that the PZT layer covers
the entire length of the beam The constant 119884119868 represents thebending stiffness of the beam and is given by
119884119868 =2119887
3[119884119904ℎ3
119904
12+ 119884119901((ℎ119901+ℎ119904
2)
3
minusℎ3
119904
8)] (22)
and the term 120599 represents the coupling between the strain andthe voltage and is given by
120599 =11988931119884119901119887
ℎ119901
[ℎ2
119904
4minus (ℎ119901+ℎ119904
2)
2
] (23)
Substituting (21) in (14) we obtain the mechanical equa-tion of motion with electrical coupling
1198841198681205974119908rel (119909 119905)
1205971199094+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
+ 120599V (119905) [120597120575 (119909)
120597119909minus120597120575 (119909 minus 119871)
120597119909]
= minus [119898 +119872119905120575 (119909 minus 119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(24)
32 Electrical Equation To derive the electrical equation letus start with the piezoelectric constitutive relation given in
119863 = 11988931119879119901+ 120576119879
33119864 (25)
where119863 is the electric displacement and 12057611987933is the permittivity
at constant stress Replacing the definition of 119879119901 from (17)yields
119863(119909 119905) = 11988931 (119884119901119878119901minus 11988931119864) + 120576
119879
33119864
= 11988931119884119901119878119901+ 119864 (120576
119879
33minus 1198892
31119884119901)
= 11988931119884119901119878119901minus 120576119878
33
V (119905)ℎ119901
(26)
with 12057611987833= 1198892
31119884119901minus 120576119879
33 The strain 119878119901 evaluated at the middle
of the PZT layer is given by
119878119901= minusℎ119901119888
1205972119908rel (119909 119905)
1205971199092 (27)
where ℎ119901119888= ℎ1199042 + ℎ
1199012 is the distance of the center of the
PZT layer to the neural axis Thus
119863 (119909 119905) = minus11988931119884119901ℎ119901119888
1205972119908rel (119909 119905)
1205971199092minus 120576119878
33
V (119905)ℎ119901
(28)
The electric charge is obtained by integrating the electricdisplacement over the area of the beamThe resulting expres-sion is
119902 (119905) = minusint
119871
0
(11988931119884119901ℎ119901119888
1205972119908rel (119909 119905)
1205971199092+ 120576119878
33
V (119905)ℎ119901
)119887119889119909 (29)
Shock and Vibration 7
Therefore the current generated by one PZT layer is
119894 (119905) =120597119902 (119905)
120597119905
= minus120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(30)
Since both layers are connected in parallel the currentgenerated by the bimorph beam is given by
119894 (119905) = 1198941 (119905) + 1198942 (119905)
= minus2120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 2int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(31)
If the beam is connected directly to a resistive load the voltageis given by V(119905) = 119877
119897119894(119905) Substituting this relation in (31) yields
V (119905) = minus2119877119897 [120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
+ int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909]
(32)
The electrical equation with mechanical coupling isobtained by rearranging (32)
V (119905)2119877119897
+120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
= minusint
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(33)
4 Voltage and Power Output of an Array ofPiezoelectric Beams Connected by Springs
Equations (24) and (33) are the electromechanical equationsfor a single bimorph beam under transverse vibrations Letus consider the case of multiple beams connected by springswith a parallel electrical connection as shown in Figure 5 Inthat case the mechanical equation of motion is given by
1198841198681205974119908rel (119909 119905)
1205971199094+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
+ 120599
119873
sum
119894=1
V119894 (119905) [
120597120575 (119909 minus (119894 minus 1) 119871)
120597119909minus120597120575 (119909 minus 119894119871)
120597119909]
= minus[119898 +
119873
sum
119894=1
119872119905119894120575 (119909 minus 119894119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(34)
(t)Rl
Figure 5 A parallel electrical connection of an array of bimorphbeams connected by springs
where V119894(119905) is the voltage of the 119894th beam Since the beams are
electrically connected in parallel they have the same voltagethus V
1(119905) = V
2(119905) = sdot sdot sdot = V
119899(119905) = V(119905) It has been assumed
that all the beams are identical except for the tip mass Itshould be noted that the springs are included as boundaryconditionsTherefore they affect the vibrationmodes but notthe equation of motion
The solution for the relative motion of the system can berepresented by a sum of its mode shapes as
119908rel (119909 119905) = sum119903
120601119903 (119909) 120578119903 (119905) (35)
where 120601119903(119909) and 120578
119903(119905) are the 119903th mode shape and modal
coordinate respectively The 119903th mode shape can be definedby segments as
120601119903 (119909)
=
1206011199031 (119909) 0 le 119909 le 119871
1206011199032 (119909 minus 119871) 119871 le 119909 le 2119871
120601119903119873 (119909 minus (119873 minus 1) 119871) (119873 minus 1) 119871 le 119909 le 119873119871
(36)
where 120601119903119894
is the 119903th mode shape of the 119894th beam 119871 is thelength of one beam and119873 is the number of beams
Replacing (35) themechanical equation ofmotion can bewritten as
[1205962
119903120578119903 (119905) + 2120585119903120596119903 120578119903 (119905) + 120578
119903 (119905)] 120572119903 +
119873
sum
119894=1
120594119903119894V (119905)
= [minus1198981205972119908119887 (119909 119905)
1205971199052minus 119888119886
120597119908119887 (119909 119905)
120597119905] 120574119903
minus1205972119908119887 (119909 119905)
1205971199052
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(37)
8 Shock and Vibration
with
120594119903119894= 120599(minus120601
119903119894 (119871) +120597120601119903119894 (119871)
120597119909)
120585119903=119888119904119868120596119903
2119884119868+
119888119886
2119898120596119903
120572119903= int
119873119871
0
1198981206012
119903(119909) 119889119909
120574119903= int
119873119871
0
120601119903 (119909) 119889119909
(38)
Assuming a harmonic basemotion119908119887(119909 119905) = 119884
0119890119895120596119905 and
a harmonic response 120578119903(119905) = 119873
119903119890119895120596119905 and V(119905) = 119881119890119895120596119905 (37)
becomes
[1205962
119899119903119873119903+ 2119895120585119903120596119899119903120596119873119903minus 1205962119873119903] 120572119903+ 119881
119873
sum
119894=1
120594119903119894
= (1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(39)
Rearranging (39) the amplitude of the temporal term119873119903
can be written as follows
119873119903
=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871) minus 119881sum
119873
119894=1120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(40)
For simplicity (40) can be expressed as
119873119903= 119886119903minus 119881
119873
sum
119894=1
119887119903119894 (41)
with
119886119903=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871)
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
119887119903119894=
120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(42)
The electrical equation needs to be derived for each beamindividually The relative motion for the 119894th beam is given by
119908rel119894 (119909 119905) = sum119903
120601119903119894 (119909) 120578119903 (119905) (43)
Replacing (43) in (30) yields
119894119894 (119905) = minus2
120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 211988931119884119901ℎ119901119888119887sum
119903
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(44)
If the beams are connected directly to a resistive load thevoltage is given by V(119905) = 119877
119897119894(119905) = 119877
119897sum119894119894119894(119905) substituting this
condition equation (44) becomes
V (119905)2119877119897
= minus120576119878
33119887119871119873
ℎ119901
120597V (119905)120597119905
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(45)
Replacing the harmonic response condition in (45) yields
119881
2119877119897
= minus119895120596120576119878
33119887119871119873
ℎ119901
119881
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
119895120596119873119903
(46)
Rearranging (46) we obtain
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596119873119903 (47)
with
120593119903119894= minus
11988931119884119901ℎ119901119888ℎ119901
120576119878
33119871119873
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120581119888= 119895120596 +
ℎ119901
2119877119897120576119878
33119887119871119873
(48)
Replacing 119873119903from (41) in (47) the following equation is
obtained
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596(119886
119903minus 119881
119873
sum
119896=1
119887119903119896) (49)
Rearranging (49) the voltage amplitude can be written asshown in
119881 =sum119903sum119894120593119903119894119895120596119886119903
120581119888+ sum119903sum119894120593119903119894119895120596sum119873
119896=1119887119903119896
(50)
The voltage FRF is defined as the ratio of the voltageoutput to the base acceleration119881minus1205962119884
0 The instantaneous
power output can be obtained from the relation 119875(119905) =V2(119905)119877
119897Therefore the FRF of the power output is simply the
square of the voltage FRF divided by 119877119897
5 Results
51 Analytical This section presents the results of the powerFRF for different number of beams and different combina-tions of tip masses and spring constants in the array Thegeometric material and electromechanical parameters of thebeams are given in Table 1 These are the parameters of theexperimental beams used in the next section
The voltage is computed by (50) and the power FRF iscomputed as 1198812[119877
119897(12059621198840)2] The value of the external load
Shock and Vibration 9
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(a) Connected
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 6 Power FRF with an increasing number of identical bimorph beams119872119905119894=16
= 03119898119871 and 119896119894=15
= 031198841198681198713
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(a) Connected
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 7 Power FRF with an increasing number of bimorph beams with different tip masses Two beams1198721199051= 03119898119871119872
1199052= 035119898119871
119896 = 031198841198681198713 three beams119872
1199051= 025119898119871119872
1199052= 030119898119871119872
1199053= 035119898119871 119896
12= 03119884119868119871
3 six beams1198721199051= 015119898119871119872
1199052= 020119898119871
1198721199053= 030119898119871119872
1199054= 035119898119871119872
1199055= 040119898119871119872
1199056= 045119898119871 119896
119894=15= 03119884119868119871
3
Table 1 Parameters of the bimorph beams
Length of the beam 119871 (mm) 482Width of the beam 119887 (mm) 10Thickness of the substructure ℎ
119904(mm) 019
Thickness of the PZT ℎ119901(mm) 024
Youngrsquos modulus of the substructure 119884119904(GPa) 404
Youngrsquos modulus of the PZT 119884119901(GPa) 55
Mass density of the substructure 120588119904(kgm3) 1519
Mass density of the PZT 120588119901(kgm3) 8140
Piezoelectric constant 11988931(pmV) minus180
Permittivity 12057611990433(nFm) 1328
Strain rate damping 119888119904119868 (kgm4s) 116 times 10minus6
Viscous air damping 119888119886(kgs) 04
resistance is defined for each case as the optimum resistorthat maximizes the power output To study the advantagesand disadvantages of the proposed system the results arecompared with those obtained by an array of disconnectedbeams (with no springs)
First the case of identical beams is investigated theresults are shown in Figure 6(a) It is obtained that in thecase of identical beams the only modes that are excitedare the modes where all beams move in phase such asmodes 1 4 and 7 in Figure 3 In this case it is also foundthat the natural frequencies do not depend on the springstiffness because all beams are in phase and the springs donot deform Furthermore the results are exactly the same asthe ones obtained with disconnected beams as illustrated inFigure 6(b)
In the case of beams with different tip masses the resultsof the FRF power do depend on the springs connecting thebeams as shown in Figure 7 If the beams are disconnectedand new beams with different tip masses are added to thearray then the frequency band of the device is increasedobtaining a similar power amplitude for all resonant frequen-cies On the other hand if the beams are connected by springand new beams are included in the array then the frequencyband is increased but the amplitude related to each resonantfrequency varies
For an array of piezoelectric beams connected by springsthe frequency band of operation and the power amplitude
10 Shock and Vibration
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
01mL 02mL01mL 03mL
01mL 04mL01mL 05mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k = 02YIL3k = 03YIL3
k = 04YIL3
(b) Varying spring constants (1198721199051 = 03119898119871 1198721199052 = 04119898119871)
Figure 8 Power FRF of two bimorph beams with different combinations of tip masses and spring constants
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
02mL 03mL 04mL02mL 02mL 04mL
03mL 03mL 04mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k =
k =k = 02YIL3 02YIL302YIL3 04YIL3
02YIL3 06YIL3
(b) Varying spring constants (1198721199051 = 02119898119871 1198721199052 = 03119898119871 1198721199053 =04119898119871)
Figure 9 Power FRF of three bimorph beams with different combinations of tip masses and spring constants
associated with each resonant frequency depend on thecombination of tipmasses and springs constantsThis ismoreevident in Figures 8 and 9 where different combinations ofsprings and tip masses are studied for the case of two andthree beams In general increasing the tip mass of the beamsdecreases someof the resonant frequencies and increasing thespring stiffness increases the frequency band
To compare the performances of an array of beamsconnected by springs and one of the disconnected beams acase of ten beams with both situations is plotted in Figure 10It is possible to see that the array with connected beamshas a slightly wider frequency band which increases withan increment of the springs constants Furthermore bycomparing the areas below both curves in the frequencyrange 50ndash200Hz the array with connected beams providesan area between 20 and 36 larger than the array withdisconnected beams Thus by connecting the beams theperformance of the array can be improved It should benoted that the previous results are obtained with arbitraryvalues for the tip masses and spring constantsTherefore thisperformance can be further improved by a proper design
52 Experimental To validate the proposed model a pre-liminary experimental campaign was performed Figure 11presents the experimental setup It consists of a signal genera-tor connected to a power amplifier and to an electrodynamicshaker which produces a controlled base vibration An arrayof three piezoelectric bimorph beams is mounted over theshaker Figure 12 shows this array The cantilever bimorphbeams have a tip mass at the free-end and can be connectedwith springs between them Three experiments were per-formed In the first case no springs were used and eachbeam was connected independently to a 1 kΩ resistive loadin this case the voltage generated by each beam is monitoredindividually In the second case no springs were used andthe beams were electrically connected in parallel to a resistiveload of 1 kΩ In the last case the beams were connected withsprings between them and they were electrically connected inparallel to a resistive load of 1 kΩ
The properties of the bimorph beams are the ones givenin Table 1 the tip masses are119872
1199051= 65119898119871119872
1199052= 35119898119871
1198721199053= 19119898119871 and the spring constants are 119896
1= 1198962=
2921198841198681198713
Shock and Vibration 11
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected02YIL3
(a) 119896119894=19 = 021198841198681198713
50 100 150 200
Frequency (Hz)
10minus4
Pow
er F
RF (W
s4m
2)
10minus3
10minus2
10minus1
100
Connected k =
Disconnected03YIL3
(b) 119896119894=19 = 031198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected04YIL3
(c) 119896119894=19 = 041198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected05YIL3
(d) 119896119894=19 = 051198841198681198713
Figure 10 Power FRF with ten bimorph beams1198721199051= 01119898119871119872
1199052= 015119898119871119872
1199053= 02119898119871119872
1199054= 025119898119871119872
1199055= 03119898119871119872
1199056= 035119898119871
1198721199057= 04119898119871119872
1199058= 045119898119871119872
1199059= 05119898119871119872
11990510= 055119898119871
Figure 11 Experimental setup
Figure 12 Experimental bimorph beams
During the three experiments the system is subjected toa harmonic base displacement with frequencies in the range20ndash70Hz The results for the first experiment are presentedin Figure 13(a) and are compared with the ones obtainedby an analytical model in Figure 13(b) Both analytical andexperimental curves are very similar The only difference isthat the experimental power increases from beams 1 to 3 andthe analytical model predicted the opposite behavior Thisunexpected behavior seems to be related to a resonance ofthe supporting structure in the three experimental curves itis possible to see an increment of the power near 60Hz whichcan be explained by a resonance of the structure
Figure 14 presents the experimental and analytical resultsfor the second and third experiments It is possible to seea good agreement between the experimental and analyticalcurves although similarly to the results in Figure 13(a) thereis an increment of the power near 60Hz that confirms astructural resonance
Comparing the areas below both experimental curves inFigure 13(b) in the frequency range 20ndash70Hz the array withconnected beams provides an area 5 larger than the arraywith disconnected beamswhich confirms that the addition ofsprings can increment the generated powerThese results canbe improved by using softer springs which have lower energylosses and by solving the structural resonance problem
12 Shock and Vibration
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 13 Power FRF for the three beams with no springs and independent electrical connections to 1 kΩ resistive loads (a) experimentalmeasurement (b) analytical results
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 14 Power FRF for the three beams with and without springs and electrical connection in parallel to a resistive load of 1 kΩ(a) experimental measurement (b) analytical results
6 Conclusions
This paper investigated the use of an array of bimorphbeams connected by springs as a broadband vibration-basedenergy harvester The equations for the voltage and poweroutput of the full system have been deduced based on theanalytical solution of the cantilevered piezoelectric energyharvester with Euler-Bernoulli beam assumptions To studythe advantages and disadvantages of the proposed system theresults are compared with those of an array of disconnectedbeams (with no springs) The analytical model is validatedwith experimental measurements of three bimorph beamswith and without springs
Both analytical and experimental results show that con-necting the array of beams with springs allows increasingthe frequency band of operation and increasing the amountof power generated Thus the maximal frequency band fora certain array of beams can be extended by connectingthem by springs and at the same time the output power isincreased The performance of this system can be optimizedfor specific applications by a proper selection of the tipmassesand spring constants
The preliminary experimental results are consistent withthe analytical ones but further research is necessary toimprove the experimental setup and to investigate differentexperimental conditions such as variations of the springsstiffness
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
References
[1] S R Anton and H A Sodano ldquoA review of power harvestingusing piezoelectric materials (2003ndash2006)rdquo Smart Materialsand Structures vol 16 no 3 article R1 2007
[2] S Chalasani and J M Conrad ldquoA survey of energy harvestingsources for embedded systemsrdquo in Proceedings of the IEEESoutheastCon pp 442ndash447 April 2008
[3] K A Cook-Chennault NThambi and AM Sastry ldquoPoweringMEMS portable devicesmdasha review of non-regenerative andregenerative power supply systems with special emphasis on
Shock and Vibration 13
piezoelectric energy harvesting systemsrdquo Smart Materials andStructures vol 17 no 4 Article ID 043001 2008
[4] A Dewan S U Ay M N Karim and H Beyenal ldquoAlternativepower sources for remote sensors a reviewrdquo Journal of PowerSources vol 245 pp 129ndash143 2014
[5] A Harb ldquoEnergy harvesting state-of-the-artrdquo RenewableEnergy vol 36 no 10 pp 2641ndash2654 2011
[6] S Saadon and O Sidek ldquoA review of vibration-based MEMSpiezoelectric energy harvestersrdquo Energy Conversion and Man-agement vol 52 no 1 pp 500ndash504 2011
[7] H A Sodano D J Inman and G Park ldquoA review of powerharvesting from vibration using piezoelectric materialsrdquo Shockand Vibration Digest vol 36 no 3 pp 197ndash205 2004
[8] R J M Vullers R van Schaijk I Doms C Van Hoofand R Mertens ldquoMicropower energy harvestingrdquo Solid-StateElectronics vol 53 no 7 pp 684ndash693 2009
[9] R Calio U B Rongala D Camboni et al ldquoPiezoelectric energyharvesting solutionsrdquo Sensors vol 14 no 3 pp 4755ndash4790 2014
[10] J Twiefel and H Westermann ldquoSurvey on broadband tech-niques for vibration energy harvestingrdquo Journal of IntelligentMaterial Systems and Structures vol 24 no 11 pp 1291ndash13022013
[11] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavengingrdquo Journal of Sound and Vibration vol 292no 3ndash5 pp 987ndash998 2006
[12] S M Shahruz ldquoLimits of performance of mechanical band-pass filters used in energy scavengingrdquo Journal of Sound andVibration vol 293 no 1-2 pp 449ndash461 2006
[13] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavenging multi-degree-of-freedom modelsrdquo Journalof Vibration and Control vol 14 no 5 pp 753ndash768 2008
[14] J-Q Liu H-B Fang Z-Y Xu et al ldquoA MEMS-based piezo-electric power generator array for vibration energy harvestingrdquoMicroelectronics Journal vol 39 no 5 pp 802ndash806 2008
[15] HC Lin PHWu I C Lien andYC Shu ldquoAnalysis of an arrayof piezo-electric energy harvesters connected in seriesrdquo SmartMaterials and Structures vol 22 no 9 Article ID 094026 2013
[16] S-J Jang E Rustighi M J Brennan Y P Lee and H-JJung ldquoDesign of a 2DOF vibrational energy harvesting devicerdquoJournal of IntelligentMaterial Systems and Structures vol 22 no5 pp 443ndash448 2011
[17] L Tang and Y Yang ldquoA multiple-degree-of-freedom piezoelec-tric energy harvesting modelrdquo Journal of Intelligent MaterialSystems and Structures vol 23 no 14 pp 1631ndash1647 2012
[18] Z Yang and J Yang ldquoConnected vibrating piezoelectricbimorph beams as a wide-band piezoelectric power harvesterrdquoJournal of IntelligentMaterial Systems and Structures vol 20 no5 pp 569ndash574 2009
[19] W Zhou G R Penamalli and L Zuo ldquoAn efficient vibrationenergy harvester with a multi-mode dynamic magnifierrdquo SmartMaterials and Structures vol 21 no 1 Article ID 015014 2012
[20] D S Nguyen E Halvorsen G U Jensen and A VoglldquoFabrication and characterization of a widebandMEMS energyharvester utilizing nonlinear springsrdquo Journal of Micromechan-ics andMicroengineering vol 20 no 12 Article ID 125009 2010
[21] S D Nguyen and E Halvorsen ldquoNonlinear springs forbandwidth-tolerant vibration energy harvestingrdquo Journal ofMicroelectromechanical Systems vol 20 no 6 pp 1225ndash12272011
[22] S BoisseauGDespesse andBA Seddik ldquoNonlinear h-shapedsprings to improve efficiency of vibration energy harvestersrdquo
Journal of Applied Mechanics vol 80 no 6 Article ID 0610132013
[23] S Priya ldquoAdvances in energy harvesting using low profilepiezoelectric transducersrdquo Journal of Electroceramics vol 19 no1 pp 165ndash182 2007
[24] S Roundy P K Wright and J Rabaey ldquoA study of lowlevel vibrations as a power source for wireless sensor nodesrdquoComputer Communications vol 26 no 11 pp 1131ndash1144 2003
[25] S Roundy and P K Wright ldquoA piezoelectric vibration basedgenerator for wireless electronicsrdquo Smart Materials and Struc-tures vol 13 no 5 pp 1131ndash1142 2004
[26] H A Sodano G Park and D J Inman ldquoEstimation of electriccharge output for piezoelectric energy harvestingrdquo Strain vol40 no 2 pp 49ndash58 2004
[27] N E Dutoit B L Wardle and S-G Kim ldquoDesign considera-tions for mems-scale piezoelectric mechanical vibration energyharvestersrdquo Integrated Ferroelectrics vol 71 no 1 pp 121ndash1602005
[28] S-N Chen G-J Wang andM-C Chien ldquoAnalytical modelingof piezoelectric vibration-induced micro power generatorrdquoMechatronics vol 16 no 7 pp 379ndash387 2006
[29] A Erturk and D J Inman ldquoIssues in mathematical modelingof piezoelectric energy harvestersrdquo Smart Materials and Struc-tures vol 17 no 6 Article ID 065016 2008
[30] F Lu H P Lee and S P Lim ldquoModeling and analysis of micropiezoelectric power generators for micro-electromechanical-systems applicationsrdquo Smart Materials and Structures vol 13no 1 pp 57ndash63 2004
[31] S Jiang X Li S Guo Y Hu J Yang andQ Jiang ldquoPerformanceof a piezoelectric bimorph for scavenging vibration energyrdquoSmartMaterials and Structures vol 14 no 4 pp 769ndash774 2005
[32] A Erturk and D J Inman ldquoA distributed parameter elec-tromechanical model for cantilevered piezoelectric energy har-vestersrdquo Journal of Vibration and Acoustics vol 130 no 4Article ID 041002 2008
[33] A Erturk and D J Inman ldquoAn experimentally validatedbimorph cantilever model for piezoelectric energy harvestingfrom base excitationsrdquo Smart Materials and Structures vol 18no 2 Article ID 025009 2009
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
Shock and Vibration 3
Mt1
Mt2
MtN
k1
k2
kN
(a) Full system
kiminus1
kiwi(x t)
Mtiminus1
Mti
Mti+1
X
(b) Scheme of the 119894th beam
Figure 2 Structure consisting in an array of piezoelectric beams connected by springs
2 Structure and Vibration Modes
The structure under investigation is an array of piezoelectricbeams connected by springs as illustrated in Figure 2(a) Allbeams are clamped at the left-end to a vertical wallThe right-end of each beam is connected to a concentrated tip massand to its neighbors beams by springs The inertia of theconcentrated masses is neglected All beams have identicaldimensions and the only differences are the tip masses
To derive the vibration modes of this system it isnecessary to study the boundary conditions of each beam sep-arately Looking at the 119894th beam as shown in Figure 2(b) andassuming a separation-of-variables solution the transversedisplacement of the 119894th beam can be expressed as 119908
119894(119909 119905) =
120601119894(119909)120578(119905) The function 120601
119894(119909) represents the mode shape and
depends on the boundary conditions In an Euler-Bernoullibeam the general solution for 120601
119894(119909) is
120601119894 (119909) = 120582
119894
1sin (120573119909) + 120582119894
2cos (120573119909) + 120582119894
3sinh (120573119909)
+ 120582119894
4cosh (120573119909)
(1)
The boundary conditions at the left-end are
120601119894 (0) = 0
120597120601119894 (0)
120597119909= 0
(2)
which implies that
120582119894
4= minus120582119894
2
120582119894
3= minus120582119894
1
(3)
Therefore the mode shape can be written as
120601119894 (119909) = 120582
119894
1[sin (120573119909) minus sinh (120573119909)]
+ 120582119894
2[cos (120573119909) minus cosh (120573119909)]
(4)
At the right-end the bending moment and shear forcemust satisfy the following conditions
1198841198681205972119908119894
1205971199092= 0
120597
120597119909[119884119868
1205972119908119894
1205971199092] = minus119872
119905119894
1205972119908119894
1205971199052minus 119896119894(119908119894minus 119908119894+1)
minus 119896119894minus1(119908119894minus 119908119894minus1)
(5)
where 119884 is Youngrsquos modulus 119868 is the area moment of inertiaof the cross section 119872
119905119894is the 119894th tip mass and 119896
119894is the
stiffness of the 119894th spring Replacing that 119908119894(119909 119905) = 120601
119894(119909)120578(119905)
and 1205972120578(119905)1205971199052 = minus1205962119899120578(119905) 120596
119899being the natural frequency
yields
1205972120601119894 (119871)
1205971199092= 0
1198841198681205973120601119894 (119871)
1205971199093= 1205962
119899119872119905119894120601119894 (119871) minus 119896119894 [120601119894 (119871) minus 120601119894+1 (119871)]
minus 119896119894minus1[120601119894 (119871) minus 120601119894minus1 (119871)]
(6)
4 Shock and Vibration
where 119871 is the length of the beam Replacing the definition of120601119894(119909) from (4) the following equations are obtained
[
0 0 1205721
1205722
0 0
1205723(119896119894minus1) 1205724(119896119894minus1) 1205725(119872119905119894 119896119894 119896119894minus1) 1205726(119872119905119894 119896119894 119896119894minus1) 1205723(119896119894) 1205724(119896119894)]
[[[[[[[[[[[[
[
120582119894minus1
1
120582119894minus1
2
120582119894
1
120582119894
2
120582119894+1
1
120582119894+1
2
]]]]]]]]]]]]
]
=
[[[[[[[[[[[
[
0
0
0
0
0
0
]]]]]]]]]]]
]
(7)
with
1205721= sin (120573119871) + sinh (120573119871)
1205722= cos (120573119871) + cosh (120573119871)
1205723(119896119894) = 119896119894(sinh (120573119871) minus sin (120573119871))
1205724(119896119894) = 119896119894(cosh (120573119871) minus cos (120573119871))
1205725(119872119905119894 119896119894 119896119894minus1)
= (119896119894+ 119896119894minus1minus1198721199051198941205962
119899) (sin (120573119871) minus sinh (120573119871))
+ 1198841198681205733(cos (120573119871) + cosh (120573119871))
1205726(119872119905119894 119896119894 119896119894minus1)
= (119896119894+ 119896119894minus1minus1198721199051198941205962
119899) (cos (120573119871) minus cosh (120573119871))
+ 1198841198681205733(sinh (120573119871) minus sin (120573119871))
(8)
Equation (7) can be obtained for each of the119873 beams inthe structure Particular cases are the first and last beams Inthe case of the first beam there is no previous neighbor thusthe following equation is obtained
[
1205721
1205722
0 0
1205725(1198721199051 1198961 0) 120572
6(1198721199051 1198961 0) 120572
3(1198961) 1205724(1198961)]
[[[[[[
[
1205821
1
1205821
2
1205822
1
1205822
2
]]]]]]
]
=
[[[[[
[
0
0
0
0
]]]]]
]
(9)
whereas for the last beam there is no next neighbor and theresulting equations are
[
0 0 1205721
1205722
1205723(119896119873minus1) 1205724(119896119873minus1) 1205725(119872119905119873 0 119896119873minus1) 1205726(119872119905119873 0 119896119873minus1)]
[[[[[[
[
120582119873minus1
1
120582119873minus1
2
120582119873
1
120582119873
2
]]]]]]
]
=
[[[[[
[
0
0
0
0
]]]]]
]
(10)
Coupling all equations yields a system of 2times119873 equationswhich can be written as the following matrix equation
C120582 = 0 (11)
where C is a coefficient matrix and 120582 = [12058211 1205821
2 1205822
1 1205822
2
120582119873
1 120582119873
2]119879 This equation can have nonzero solution for 120582 only
if the determinant of the coefficient matrix vanishes (ie thecoefficient matrix is singular) Setting the determinant of Cequal to zero yields the characteristic equation that is satisfiedby an infinite number of choices for 120573 denoted by 120573
119903 The
vector of unknowns 120582 is calculated by replacing 120573119903in the
coefficient matrix C and then computing its null space Foreach value of 120573
119903a different vector 120582
119903is obtained
Given 120573119903and 120582
119903 the mode shapes are given by
120601119903119894 (119909) = 120582
119894
1199031[sin (120573
119903119909) minus sinh (120573
119903119909)]
+ 120582119894
1199032[cos (120573
119903119909) minus cosh (120573
119903119909)]
(12)
The undamped natural frequency associated with the 119903thmode is given by
120596119899119903= 1205732
119903radic119884119868
119898 (13)
where119898 is the mass per unit length As an example Figure 3illustrates the vibration modes of three elastically connected
Shock and Vibration 5
120573rL = 1536
(a) Mode 1
120573rL = 1573
(b) Mode 2
120573rL = 1640
(c) Mode 3
120573rL = 4192
(d) Mode 4 (e) Mode 5 (f) Mode 6
(g) Mode 7 (h) Mode 8 (i) Mode 9
Figure 3 First nine vibration modes of three elastically connected beams with identical tip masses and spring constants (119872119905119894=13
=
03119898119871 11989612= 03119864119868119871
3)
beams In this example all beams have a mass tip of 03119898119871and all springs have a stiffness of 031198841198681198713
3 Electromechanical Model of a BimorphPiezoelectric Beam
31 Mechanical Equation Let us consider the bimorph can-tilever beam shown in Figure 4 The bimorph beam consistsin two layers of piezoelectric material (PZT) bonded to asubstructure layer Assuming an Euler-Bernoulli beam theequation of motion can be written as [33]
1205972119872(119909 119905)
1205971199092+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
= minus [119898 +119872119905120575 (119909 minus 119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(14)
b
x(t)
Mt
hp
hp
hs
PZT
PZTSubstructureY0e
i120596t
Figure 4 Scheme of the bimorph piezoelectric beam
where 119908rel(119909 119905) is the transverse deflection of the beamrelative to the base 119908
119887(119905) is the base displacement119872(119909 119905) is
the internal bending moment 119888119904is the equivalent coefficient
of strain rate damping 119888119886is the viscous air damping 119868 is the
area moment of inertia of the cross section119898 is the mass perunit length 119872
119905is the tip mass and 120575(119909) is the Dirac delta
function
6 Shock and Vibration
The internal moment is given by the integral of the stressdistribution over the cross section of the beam shown in
119872(119909 119905) = minus119887 [int
minusℎ1199042
minusℎ119901minusℎ1199042
119879119901119910119889119910 + int
ℎ1199042
minusℎ1199042
119879119904119910119889119910
+ int
ℎ1199042+ℎ119901
ℎ1199042
119879119901119910119889119910]
(15)
where 119887 is the width of the beam ℎ119901is the thickness of
each PZT layer ℎ119904is the thickness of the substructure (see
Figure 4) and 119879 is the stress Superscripts 119901 and 119904 stand forPZT and substructure respectively The stress distributionacross the section is given by the strain-stress relations of thesubstructure and PZT layers which are given by
119879119904= 119884119904119878119904 (16)
119879119901= 119884119901119878119901minus 11988931119884119901119864 (17)
where 119878 is the strain 119884 is Youngrsquos modulus 119889 is thepiezoelectric constant and 119864 is the electric fieldThe bendingstrain can be expressed in terms of the radius of curvature as
119878 (119909 119910 119905) = minus1199101205972119908rel (119909 119905)
1205971199092 (18)
Since the piezoelectric layers are connected in parallelthey have the same voltage but the instantaneous electricfields are in opposite directions The electric field can bewritten in terms of the voltage across the PZT V(119905) and thethickness of the PZT layer ℎ
119901 by the following equation
1198641 (119905) = minus
V (119905)ℎ119901
in the top layer
1198642 (119905) =
V (119905)ℎ119901
in the bottom layer(19)
Replacing (16) to (19) in (15) we obtain
119872(119909 119905) =2119887
3[119884119904ℎ3
119904
12+ 119884119901((ℎ119901+ℎ119904
2)
3
minusℎ3
119904
8)]
sdot1205972119908rel (119909 119905)
1205971199092
+11988931119884119901119887
ℎ119901
[ℎ2
119904
4minus (ℎ119901+ℎ119904
2)
2
] V (119905)
(20)
Equation (20) can be expressed as
119872(119909 119905) = 1198841198681205972119908rel (119909 119905)
1205971199092
+ 120599V (119905) [Φ (119909) minus Φ (119909 minus 119871)] (21)
where 119871 is the length of the beam and Φ the Heavisidefunction It has been assumed that the PZT layer covers
the entire length of the beam The constant 119884119868 represents thebending stiffness of the beam and is given by
119884119868 =2119887
3[119884119904ℎ3
119904
12+ 119884119901((ℎ119901+ℎ119904
2)
3
minusℎ3
119904
8)] (22)
and the term 120599 represents the coupling between the strain andthe voltage and is given by
120599 =11988931119884119901119887
ℎ119901
[ℎ2
119904
4minus (ℎ119901+ℎ119904
2)
2
] (23)
Substituting (21) in (14) we obtain the mechanical equa-tion of motion with electrical coupling
1198841198681205974119908rel (119909 119905)
1205971199094+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
+ 120599V (119905) [120597120575 (119909)
120597119909minus120597120575 (119909 minus 119871)
120597119909]
= minus [119898 +119872119905120575 (119909 minus 119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(24)
32 Electrical Equation To derive the electrical equation letus start with the piezoelectric constitutive relation given in
119863 = 11988931119879119901+ 120576119879
33119864 (25)
where119863 is the electric displacement and 12057611987933is the permittivity
at constant stress Replacing the definition of 119879119901 from (17)yields
119863(119909 119905) = 11988931 (119884119901119878119901minus 11988931119864) + 120576
119879
33119864
= 11988931119884119901119878119901+ 119864 (120576
119879
33minus 1198892
31119884119901)
= 11988931119884119901119878119901minus 120576119878
33
V (119905)ℎ119901
(26)
with 12057611987833= 1198892
31119884119901minus 120576119879
33 The strain 119878119901 evaluated at the middle
of the PZT layer is given by
119878119901= minusℎ119901119888
1205972119908rel (119909 119905)
1205971199092 (27)
where ℎ119901119888= ℎ1199042 + ℎ
1199012 is the distance of the center of the
PZT layer to the neural axis Thus
119863 (119909 119905) = minus11988931119884119901ℎ119901119888
1205972119908rel (119909 119905)
1205971199092minus 120576119878
33
V (119905)ℎ119901
(28)
The electric charge is obtained by integrating the electricdisplacement over the area of the beamThe resulting expres-sion is
119902 (119905) = minusint
119871
0
(11988931119884119901ℎ119901119888
1205972119908rel (119909 119905)
1205971199092+ 120576119878
33
V (119905)ℎ119901
)119887119889119909 (29)
Shock and Vibration 7
Therefore the current generated by one PZT layer is
119894 (119905) =120597119902 (119905)
120597119905
= minus120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(30)
Since both layers are connected in parallel the currentgenerated by the bimorph beam is given by
119894 (119905) = 1198941 (119905) + 1198942 (119905)
= minus2120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 2int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(31)
If the beam is connected directly to a resistive load the voltageis given by V(119905) = 119877
119897119894(119905) Substituting this relation in (31) yields
V (119905) = minus2119877119897 [120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
+ int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909]
(32)
The electrical equation with mechanical coupling isobtained by rearranging (32)
V (119905)2119877119897
+120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
= minusint
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(33)
4 Voltage and Power Output of an Array ofPiezoelectric Beams Connected by Springs
Equations (24) and (33) are the electromechanical equationsfor a single bimorph beam under transverse vibrations Letus consider the case of multiple beams connected by springswith a parallel electrical connection as shown in Figure 5 Inthat case the mechanical equation of motion is given by
1198841198681205974119908rel (119909 119905)
1205971199094+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
+ 120599
119873
sum
119894=1
V119894 (119905) [
120597120575 (119909 minus (119894 minus 1) 119871)
120597119909minus120597120575 (119909 minus 119894119871)
120597119909]
= minus[119898 +
119873
sum
119894=1
119872119905119894120575 (119909 minus 119894119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(34)
(t)Rl
Figure 5 A parallel electrical connection of an array of bimorphbeams connected by springs
where V119894(119905) is the voltage of the 119894th beam Since the beams are
electrically connected in parallel they have the same voltagethus V
1(119905) = V
2(119905) = sdot sdot sdot = V
119899(119905) = V(119905) It has been assumed
that all the beams are identical except for the tip mass Itshould be noted that the springs are included as boundaryconditionsTherefore they affect the vibrationmodes but notthe equation of motion
The solution for the relative motion of the system can berepresented by a sum of its mode shapes as
119908rel (119909 119905) = sum119903
120601119903 (119909) 120578119903 (119905) (35)
where 120601119903(119909) and 120578
119903(119905) are the 119903th mode shape and modal
coordinate respectively The 119903th mode shape can be definedby segments as
120601119903 (119909)
=
1206011199031 (119909) 0 le 119909 le 119871
1206011199032 (119909 minus 119871) 119871 le 119909 le 2119871
120601119903119873 (119909 minus (119873 minus 1) 119871) (119873 minus 1) 119871 le 119909 le 119873119871
(36)
where 120601119903119894
is the 119903th mode shape of the 119894th beam 119871 is thelength of one beam and119873 is the number of beams
Replacing (35) themechanical equation ofmotion can bewritten as
[1205962
119903120578119903 (119905) + 2120585119903120596119903 120578119903 (119905) + 120578
119903 (119905)] 120572119903 +
119873
sum
119894=1
120594119903119894V (119905)
= [minus1198981205972119908119887 (119909 119905)
1205971199052minus 119888119886
120597119908119887 (119909 119905)
120597119905] 120574119903
minus1205972119908119887 (119909 119905)
1205971199052
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(37)
8 Shock and Vibration
with
120594119903119894= 120599(minus120601
119903119894 (119871) +120597120601119903119894 (119871)
120597119909)
120585119903=119888119904119868120596119903
2119884119868+
119888119886
2119898120596119903
120572119903= int
119873119871
0
1198981206012
119903(119909) 119889119909
120574119903= int
119873119871
0
120601119903 (119909) 119889119909
(38)
Assuming a harmonic basemotion119908119887(119909 119905) = 119884
0119890119895120596119905 and
a harmonic response 120578119903(119905) = 119873
119903119890119895120596119905 and V(119905) = 119881119890119895120596119905 (37)
becomes
[1205962
119899119903119873119903+ 2119895120585119903120596119899119903120596119873119903minus 1205962119873119903] 120572119903+ 119881
119873
sum
119894=1
120594119903119894
= (1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(39)
Rearranging (39) the amplitude of the temporal term119873119903
can be written as follows
119873119903
=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871) minus 119881sum
119873
119894=1120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(40)
For simplicity (40) can be expressed as
119873119903= 119886119903minus 119881
119873
sum
119894=1
119887119903119894 (41)
with
119886119903=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871)
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
119887119903119894=
120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(42)
The electrical equation needs to be derived for each beamindividually The relative motion for the 119894th beam is given by
119908rel119894 (119909 119905) = sum119903
120601119903119894 (119909) 120578119903 (119905) (43)
Replacing (43) in (30) yields
119894119894 (119905) = minus2
120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 211988931119884119901ℎ119901119888119887sum
119903
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(44)
If the beams are connected directly to a resistive load thevoltage is given by V(119905) = 119877
119897119894(119905) = 119877
119897sum119894119894119894(119905) substituting this
condition equation (44) becomes
V (119905)2119877119897
= minus120576119878
33119887119871119873
ℎ119901
120597V (119905)120597119905
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(45)
Replacing the harmonic response condition in (45) yields
119881
2119877119897
= minus119895120596120576119878
33119887119871119873
ℎ119901
119881
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
119895120596119873119903
(46)
Rearranging (46) we obtain
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596119873119903 (47)
with
120593119903119894= minus
11988931119884119901ℎ119901119888ℎ119901
120576119878
33119871119873
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120581119888= 119895120596 +
ℎ119901
2119877119897120576119878
33119887119871119873
(48)
Replacing 119873119903from (41) in (47) the following equation is
obtained
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596(119886
119903minus 119881
119873
sum
119896=1
119887119903119896) (49)
Rearranging (49) the voltage amplitude can be written asshown in
119881 =sum119903sum119894120593119903119894119895120596119886119903
120581119888+ sum119903sum119894120593119903119894119895120596sum119873
119896=1119887119903119896
(50)
The voltage FRF is defined as the ratio of the voltageoutput to the base acceleration119881minus1205962119884
0 The instantaneous
power output can be obtained from the relation 119875(119905) =V2(119905)119877
119897Therefore the FRF of the power output is simply the
square of the voltage FRF divided by 119877119897
5 Results
51 Analytical This section presents the results of the powerFRF for different number of beams and different combina-tions of tip masses and spring constants in the array Thegeometric material and electromechanical parameters of thebeams are given in Table 1 These are the parameters of theexperimental beams used in the next section
The voltage is computed by (50) and the power FRF iscomputed as 1198812[119877
119897(12059621198840)2] The value of the external load
Shock and Vibration 9
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(a) Connected
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 6 Power FRF with an increasing number of identical bimorph beams119872119905119894=16
= 03119898119871 and 119896119894=15
= 031198841198681198713
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(a) Connected
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 7 Power FRF with an increasing number of bimorph beams with different tip masses Two beams1198721199051= 03119898119871119872
1199052= 035119898119871
119896 = 031198841198681198713 three beams119872
1199051= 025119898119871119872
1199052= 030119898119871119872
1199053= 035119898119871 119896
12= 03119884119868119871
3 six beams1198721199051= 015119898119871119872
1199052= 020119898119871
1198721199053= 030119898119871119872
1199054= 035119898119871119872
1199055= 040119898119871119872
1199056= 045119898119871 119896
119894=15= 03119884119868119871
3
Table 1 Parameters of the bimorph beams
Length of the beam 119871 (mm) 482Width of the beam 119887 (mm) 10Thickness of the substructure ℎ
119904(mm) 019
Thickness of the PZT ℎ119901(mm) 024
Youngrsquos modulus of the substructure 119884119904(GPa) 404
Youngrsquos modulus of the PZT 119884119901(GPa) 55
Mass density of the substructure 120588119904(kgm3) 1519
Mass density of the PZT 120588119901(kgm3) 8140
Piezoelectric constant 11988931(pmV) minus180
Permittivity 12057611990433(nFm) 1328
Strain rate damping 119888119904119868 (kgm4s) 116 times 10minus6
Viscous air damping 119888119886(kgs) 04
resistance is defined for each case as the optimum resistorthat maximizes the power output To study the advantagesand disadvantages of the proposed system the results arecompared with those obtained by an array of disconnectedbeams (with no springs)
First the case of identical beams is investigated theresults are shown in Figure 6(a) It is obtained that in thecase of identical beams the only modes that are excitedare the modes where all beams move in phase such asmodes 1 4 and 7 in Figure 3 In this case it is also foundthat the natural frequencies do not depend on the springstiffness because all beams are in phase and the springs donot deform Furthermore the results are exactly the same asthe ones obtained with disconnected beams as illustrated inFigure 6(b)
In the case of beams with different tip masses the resultsof the FRF power do depend on the springs connecting thebeams as shown in Figure 7 If the beams are disconnectedand new beams with different tip masses are added to thearray then the frequency band of the device is increasedobtaining a similar power amplitude for all resonant frequen-cies On the other hand if the beams are connected by springand new beams are included in the array then the frequencyband is increased but the amplitude related to each resonantfrequency varies
For an array of piezoelectric beams connected by springsthe frequency band of operation and the power amplitude
10 Shock and Vibration
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
01mL 02mL01mL 03mL
01mL 04mL01mL 05mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k = 02YIL3k = 03YIL3
k = 04YIL3
(b) Varying spring constants (1198721199051 = 03119898119871 1198721199052 = 04119898119871)
Figure 8 Power FRF of two bimorph beams with different combinations of tip masses and spring constants
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
02mL 03mL 04mL02mL 02mL 04mL
03mL 03mL 04mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k =
k =k = 02YIL3 02YIL302YIL3 04YIL3
02YIL3 06YIL3
(b) Varying spring constants (1198721199051 = 02119898119871 1198721199052 = 03119898119871 1198721199053 =04119898119871)
Figure 9 Power FRF of three bimorph beams with different combinations of tip masses and spring constants
associated with each resonant frequency depend on thecombination of tipmasses and springs constantsThis ismoreevident in Figures 8 and 9 where different combinations ofsprings and tip masses are studied for the case of two andthree beams In general increasing the tip mass of the beamsdecreases someof the resonant frequencies and increasing thespring stiffness increases the frequency band
To compare the performances of an array of beamsconnected by springs and one of the disconnected beams acase of ten beams with both situations is plotted in Figure 10It is possible to see that the array with connected beamshas a slightly wider frequency band which increases withan increment of the springs constants Furthermore bycomparing the areas below both curves in the frequencyrange 50ndash200Hz the array with connected beams providesan area between 20 and 36 larger than the array withdisconnected beams Thus by connecting the beams theperformance of the array can be improved It should benoted that the previous results are obtained with arbitraryvalues for the tip masses and spring constantsTherefore thisperformance can be further improved by a proper design
52 Experimental To validate the proposed model a pre-liminary experimental campaign was performed Figure 11presents the experimental setup It consists of a signal genera-tor connected to a power amplifier and to an electrodynamicshaker which produces a controlled base vibration An arrayof three piezoelectric bimorph beams is mounted over theshaker Figure 12 shows this array The cantilever bimorphbeams have a tip mass at the free-end and can be connectedwith springs between them Three experiments were per-formed In the first case no springs were used and eachbeam was connected independently to a 1 kΩ resistive loadin this case the voltage generated by each beam is monitoredindividually In the second case no springs were used andthe beams were electrically connected in parallel to a resistiveload of 1 kΩ In the last case the beams were connected withsprings between them and they were electrically connected inparallel to a resistive load of 1 kΩ
The properties of the bimorph beams are the ones givenin Table 1 the tip masses are119872
1199051= 65119898119871119872
1199052= 35119898119871
1198721199053= 19119898119871 and the spring constants are 119896
1= 1198962=
2921198841198681198713
Shock and Vibration 11
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected02YIL3
(a) 119896119894=19 = 021198841198681198713
50 100 150 200
Frequency (Hz)
10minus4
Pow
er F
RF (W
s4m
2)
10minus3
10minus2
10minus1
100
Connected k =
Disconnected03YIL3
(b) 119896119894=19 = 031198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected04YIL3
(c) 119896119894=19 = 041198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected05YIL3
(d) 119896119894=19 = 051198841198681198713
Figure 10 Power FRF with ten bimorph beams1198721199051= 01119898119871119872
1199052= 015119898119871119872
1199053= 02119898119871119872
1199054= 025119898119871119872
1199055= 03119898119871119872
1199056= 035119898119871
1198721199057= 04119898119871119872
1199058= 045119898119871119872
1199059= 05119898119871119872
11990510= 055119898119871
Figure 11 Experimental setup
Figure 12 Experimental bimorph beams
During the three experiments the system is subjected toa harmonic base displacement with frequencies in the range20ndash70Hz The results for the first experiment are presentedin Figure 13(a) and are compared with the ones obtainedby an analytical model in Figure 13(b) Both analytical andexperimental curves are very similar The only difference isthat the experimental power increases from beams 1 to 3 andthe analytical model predicted the opposite behavior Thisunexpected behavior seems to be related to a resonance ofthe supporting structure in the three experimental curves itis possible to see an increment of the power near 60Hz whichcan be explained by a resonance of the structure
Figure 14 presents the experimental and analytical resultsfor the second and third experiments It is possible to seea good agreement between the experimental and analyticalcurves although similarly to the results in Figure 13(a) thereis an increment of the power near 60Hz that confirms astructural resonance
Comparing the areas below both experimental curves inFigure 13(b) in the frequency range 20ndash70Hz the array withconnected beams provides an area 5 larger than the arraywith disconnected beamswhich confirms that the addition ofsprings can increment the generated powerThese results canbe improved by using softer springs which have lower energylosses and by solving the structural resonance problem
12 Shock and Vibration
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 13 Power FRF for the three beams with no springs and independent electrical connections to 1 kΩ resistive loads (a) experimentalmeasurement (b) analytical results
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 14 Power FRF for the three beams with and without springs and electrical connection in parallel to a resistive load of 1 kΩ(a) experimental measurement (b) analytical results
6 Conclusions
This paper investigated the use of an array of bimorphbeams connected by springs as a broadband vibration-basedenergy harvester The equations for the voltage and poweroutput of the full system have been deduced based on theanalytical solution of the cantilevered piezoelectric energyharvester with Euler-Bernoulli beam assumptions To studythe advantages and disadvantages of the proposed system theresults are compared with those of an array of disconnectedbeams (with no springs) The analytical model is validatedwith experimental measurements of three bimorph beamswith and without springs
Both analytical and experimental results show that con-necting the array of beams with springs allows increasingthe frequency band of operation and increasing the amountof power generated Thus the maximal frequency band fora certain array of beams can be extended by connectingthem by springs and at the same time the output power isincreased The performance of this system can be optimizedfor specific applications by a proper selection of the tipmassesand spring constants
The preliminary experimental results are consistent withthe analytical ones but further research is necessary toimprove the experimental setup and to investigate differentexperimental conditions such as variations of the springsstiffness
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
References
[1] S R Anton and H A Sodano ldquoA review of power harvestingusing piezoelectric materials (2003ndash2006)rdquo Smart Materialsand Structures vol 16 no 3 article R1 2007
[2] S Chalasani and J M Conrad ldquoA survey of energy harvestingsources for embedded systemsrdquo in Proceedings of the IEEESoutheastCon pp 442ndash447 April 2008
[3] K A Cook-Chennault NThambi and AM Sastry ldquoPoweringMEMS portable devicesmdasha review of non-regenerative andregenerative power supply systems with special emphasis on
Shock and Vibration 13
piezoelectric energy harvesting systemsrdquo Smart Materials andStructures vol 17 no 4 Article ID 043001 2008
[4] A Dewan S U Ay M N Karim and H Beyenal ldquoAlternativepower sources for remote sensors a reviewrdquo Journal of PowerSources vol 245 pp 129ndash143 2014
[5] A Harb ldquoEnergy harvesting state-of-the-artrdquo RenewableEnergy vol 36 no 10 pp 2641ndash2654 2011
[6] S Saadon and O Sidek ldquoA review of vibration-based MEMSpiezoelectric energy harvestersrdquo Energy Conversion and Man-agement vol 52 no 1 pp 500ndash504 2011
[7] H A Sodano D J Inman and G Park ldquoA review of powerharvesting from vibration using piezoelectric materialsrdquo Shockand Vibration Digest vol 36 no 3 pp 197ndash205 2004
[8] R J M Vullers R van Schaijk I Doms C Van Hoofand R Mertens ldquoMicropower energy harvestingrdquo Solid-StateElectronics vol 53 no 7 pp 684ndash693 2009
[9] R Calio U B Rongala D Camboni et al ldquoPiezoelectric energyharvesting solutionsrdquo Sensors vol 14 no 3 pp 4755ndash4790 2014
[10] J Twiefel and H Westermann ldquoSurvey on broadband tech-niques for vibration energy harvestingrdquo Journal of IntelligentMaterial Systems and Structures vol 24 no 11 pp 1291ndash13022013
[11] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavengingrdquo Journal of Sound and Vibration vol 292no 3ndash5 pp 987ndash998 2006
[12] S M Shahruz ldquoLimits of performance of mechanical band-pass filters used in energy scavengingrdquo Journal of Sound andVibration vol 293 no 1-2 pp 449ndash461 2006
[13] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavenging multi-degree-of-freedom modelsrdquo Journalof Vibration and Control vol 14 no 5 pp 753ndash768 2008
[14] J-Q Liu H-B Fang Z-Y Xu et al ldquoA MEMS-based piezo-electric power generator array for vibration energy harvestingrdquoMicroelectronics Journal vol 39 no 5 pp 802ndash806 2008
[15] HC Lin PHWu I C Lien andYC Shu ldquoAnalysis of an arrayof piezo-electric energy harvesters connected in seriesrdquo SmartMaterials and Structures vol 22 no 9 Article ID 094026 2013
[16] S-J Jang E Rustighi M J Brennan Y P Lee and H-JJung ldquoDesign of a 2DOF vibrational energy harvesting devicerdquoJournal of IntelligentMaterial Systems and Structures vol 22 no5 pp 443ndash448 2011
[17] L Tang and Y Yang ldquoA multiple-degree-of-freedom piezoelec-tric energy harvesting modelrdquo Journal of Intelligent MaterialSystems and Structures vol 23 no 14 pp 1631ndash1647 2012
[18] Z Yang and J Yang ldquoConnected vibrating piezoelectricbimorph beams as a wide-band piezoelectric power harvesterrdquoJournal of IntelligentMaterial Systems and Structures vol 20 no5 pp 569ndash574 2009
[19] W Zhou G R Penamalli and L Zuo ldquoAn efficient vibrationenergy harvester with a multi-mode dynamic magnifierrdquo SmartMaterials and Structures vol 21 no 1 Article ID 015014 2012
[20] D S Nguyen E Halvorsen G U Jensen and A VoglldquoFabrication and characterization of a widebandMEMS energyharvester utilizing nonlinear springsrdquo Journal of Micromechan-ics andMicroengineering vol 20 no 12 Article ID 125009 2010
[21] S D Nguyen and E Halvorsen ldquoNonlinear springs forbandwidth-tolerant vibration energy harvestingrdquo Journal ofMicroelectromechanical Systems vol 20 no 6 pp 1225ndash12272011
[22] S BoisseauGDespesse andBA Seddik ldquoNonlinear h-shapedsprings to improve efficiency of vibration energy harvestersrdquo
Journal of Applied Mechanics vol 80 no 6 Article ID 0610132013
[23] S Priya ldquoAdvances in energy harvesting using low profilepiezoelectric transducersrdquo Journal of Electroceramics vol 19 no1 pp 165ndash182 2007
[24] S Roundy P K Wright and J Rabaey ldquoA study of lowlevel vibrations as a power source for wireless sensor nodesrdquoComputer Communications vol 26 no 11 pp 1131ndash1144 2003
[25] S Roundy and P K Wright ldquoA piezoelectric vibration basedgenerator for wireless electronicsrdquo Smart Materials and Struc-tures vol 13 no 5 pp 1131ndash1142 2004
[26] H A Sodano G Park and D J Inman ldquoEstimation of electriccharge output for piezoelectric energy harvestingrdquo Strain vol40 no 2 pp 49ndash58 2004
[27] N E Dutoit B L Wardle and S-G Kim ldquoDesign considera-tions for mems-scale piezoelectric mechanical vibration energyharvestersrdquo Integrated Ferroelectrics vol 71 no 1 pp 121ndash1602005
[28] S-N Chen G-J Wang andM-C Chien ldquoAnalytical modelingof piezoelectric vibration-induced micro power generatorrdquoMechatronics vol 16 no 7 pp 379ndash387 2006
[29] A Erturk and D J Inman ldquoIssues in mathematical modelingof piezoelectric energy harvestersrdquo Smart Materials and Struc-tures vol 17 no 6 Article ID 065016 2008
[30] F Lu H P Lee and S P Lim ldquoModeling and analysis of micropiezoelectric power generators for micro-electromechanical-systems applicationsrdquo Smart Materials and Structures vol 13no 1 pp 57ndash63 2004
[31] S Jiang X Li S Guo Y Hu J Yang andQ Jiang ldquoPerformanceof a piezoelectric bimorph for scavenging vibration energyrdquoSmartMaterials and Structures vol 14 no 4 pp 769ndash774 2005
[32] A Erturk and D J Inman ldquoA distributed parameter elec-tromechanical model for cantilevered piezoelectric energy har-vestersrdquo Journal of Vibration and Acoustics vol 130 no 4Article ID 041002 2008
[33] A Erturk and D J Inman ldquoAn experimentally validatedbimorph cantilever model for piezoelectric energy harvestingfrom base excitationsrdquo Smart Materials and Structures vol 18no 2 Article ID 025009 2009
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
4 Shock and Vibration
where 119871 is the length of the beam Replacing the definition of120601119894(119909) from (4) the following equations are obtained
[
0 0 1205721
1205722
0 0
1205723(119896119894minus1) 1205724(119896119894minus1) 1205725(119872119905119894 119896119894 119896119894minus1) 1205726(119872119905119894 119896119894 119896119894minus1) 1205723(119896119894) 1205724(119896119894)]
[[[[[[[[[[[[
[
120582119894minus1
1
120582119894minus1
2
120582119894
1
120582119894
2
120582119894+1
1
120582119894+1
2
]]]]]]]]]]]]
]
=
[[[[[[[[[[[
[
0
0
0
0
0
0
]]]]]]]]]]]
]
(7)
with
1205721= sin (120573119871) + sinh (120573119871)
1205722= cos (120573119871) + cosh (120573119871)
1205723(119896119894) = 119896119894(sinh (120573119871) minus sin (120573119871))
1205724(119896119894) = 119896119894(cosh (120573119871) minus cos (120573119871))
1205725(119872119905119894 119896119894 119896119894minus1)
= (119896119894+ 119896119894minus1minus1198721199051198941205962
119899) (sin (120573119871) minus sinh (120573119871))
+ 1198841198681205733(cos (120573119871) + cosh (120573119871))
1205726(119872119905119894 119896119894 119896119894minus1)
= (119896119894+ 119896119894minus1minus1198721199051198941205962
119899) (cos (120573119871) minus cosh (120573119871))
+ 1198841198681205733(sinh (120573119871) minus sin (120573119871))
(8)
Equation (7) can be obtained for each of the119873 beams inthe structure Particular cases are the first and last beams Inthe case of the first beam there is no previous neighbor thusthe following equation is obtained
[
1205721
1205722
0 0
1205725(1198721199051 1198961 0) 120572
6(1198721199051 1198961 0) 120572
3(1198961) 1205724(1198961)]
[[[[[[
[
1205821
1
1205821
2
1205822
1
1205822
2
]]]]]]
]
=
[[[[[
[
0
0
0
0
]]]]]
]
(9)
whereas for the last beam there is no next neighbor and theresulting equations are
[
0 0 1205721
1205722
1205723(119896119873minus1) 1205724(119896119873minus1) 1205725(119872119905119873 0 119896119873minus1) 1205726(119872119905119873 0 119896119873minus1)]
[[[[[[
[
120582119873minus1
1
120582119873minus1
2
120582119873
1
120582119873
2
]]]]]]
]
=
[[[[[
[
0
0
0
0
]]]]]
]
(10)
Coupling all equations yields a system of 2times119873 equationswhich can be written as the following matrix equation
C120582 = 0 (11)
where C is a coefficient matrix and 120582 = [12058211 1205821
2 1205822
1 1205822
2
120582119873
1 120582119873
2]119879 This equation can have nonzero solution for 120582 only
if the determinant of the coefficient matrix vanishes (ie thecoefficient matrix is singular) Setting the determinant of Cequal to zero yields the characteristic equation that is satisfiedby an infinite number of choices for 120573 denoted by 120573
119903 The
vector of unknowns 120582 is calculated by replacing 120573119903in the
coefficient matrix C and then computing its null space Foreach value of 120573
119903a different vector 120582
119903is obtained
Given 120573119903and 120582
119903 the mode shapes are given by
120601119903119894 (119909) = 120582
119894
1199031[sin (120573
119903119909) minus sinh (120573
119903119909)]
+ 120582119894
1199032[cos (120573
119903119909) minus cosh (120573
119903119909)]
(12)
The undamped natural frequency associated with the 119903thmode is given by
120596119899119903= 1205732
119903radic119884119868
119898 (13)
where119898 is the mass per unit length As an example Figure 3illustrates the vibration modes of three elastically connected
Shock and Vibration 5
120573rL = 1536
(a) Mode 1
120573rL = 1573
(b) Mode 2
120573rL = 1640
(c) Mode 3
120573rL = 4192
(d) Mode 4 (e) Mode 5 (f) Mode 6
(g) Mode 7 (h) Mode 8 (i) Mode 9
Figure 3 First nine vibration modes of three elastically connected beams with identical tip masses and spring constants (119872119905119894=13
=
03119898119871 11989612= 03119864119868119871
3)
beams In this example all beams have a mass tip of 03119898119871and all springs have a stiffness of 031198841198681198713
3 Electromechanical Model of a BimorphPiezoelectric Beam
31 Mechanical Equation Let us consider the bimorph can-tilever beam shown in Figure 4 The bimorph beam consistsin two layers of piezoelectric material (PZT) bonded to asubstructure layer Assuming an Euler-Bernoulli beam theequation of motion can be written as [33]
1205972119872(119909 119905)
1205971199092+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
= minus [119898 +119872119905120575 (119909 minus 119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(14)
b
x(t)
Mt
hp
hp
hs
PZT
PZTSubstructureY0e
i120596t
Figure 4 Scheme of the bimorph piezoelectric beam
where 119908rel(119909 119905) is the transverse deflection of the beamrelative to the base 119908
119887(119905) is the base displacement119872(119909 119905) is
the internal bending moment 119888119904is the equivalent coefficient
of strain rate damping 119888119886is the viscous air damping 119868 is the
area moment of inertia of the cross section119898 is the mass perunit length 119872
119905is the tip mass and 120575(119909) is the Dirac delta
function
6 Shock and Vibration
The internal moment is given by the integral of the stressdistribution over the cross section of the beam shown in
119872(119909 119905) = minus119887 [int
minusℎ1199042
minusℎ119901minusℎ1199042
119879119901119910119889119910 + int
ℎ1199042
minusℎ1199042
119879119904119910119889119910
+ int
ℎ1199042+ℎ119901
ℎ1199042
119879119901119910119889119910]
(15)
where 119887 is the width of the beam ℎ119901is the thickness of
each PZT layer ℎ119904is the thickness of the substructure (see
Figure 4) and 119879 is the stress Superscripts 119901 and 119904 stand forPZT and substructure respectively The stress distributionacross the section is given by the strain-stress relations of thesubstructure and PZT layers which are given by
119879119904= 119884119904119878119904 (16)
119879119901= 119884119901119878119901minus 11988931119884119901119864 (17)
where 119878 is the strain 119884 is Youngrsquos modulus 119889 is thepiezoelectric constant and 119864 is the electric fieldThe bendingstrain can be expressed in terms of the radius of curvature as
119878 (119909 119910 119905) = minus1199101205972119908rel (119909 119905)
1205971199092 (18)
Since the piezoelectric layers are connected in parallelthey have the same voltage but the instantaneous electricfields are in opposite directions The electric field can bewritten in terms of the voltage across the PZT V(119905) and thethickness of the PZT layer ℎ
119901 by the following equation
1198641 (119905) = minus
V (119905)ℎ119901
in the top layer
1198642 (119905) =
V (119905)ℎ119901
in the bottom layer(19)
Replacing (16) to (19) in (15) we obtain
119872(119909 119905) =2119887
3[119884119904ℎ3
119904
12+ 119884119901((ℎ119901+ℎ119904
2)
3
minusℎ3
119904
8)]
sdot1205972119908rel (119909 119905)
1205971199092
+11988931119884119901119887
ℎ119901
[ℎ2
119904
4minus (ℎ119901+ℎ119904
2)
2
] V (119905)
(20)
Equation (20) can be expressed as
119872(119909 119905) = 1198841198681205972119908rel (119909 119905)
1205971199092
+ 120599V (119905) [Φ (119909) minus Φ (119909 minus 119871)] (21)
where 119871 is the length of the beam and Φ the Heavisidefunction It has been assumed that the PZT layer covers
the entire length of the beam The constant 119884119868 represents thebending stiffness of the beam and is given by
119884119868 =2119887
3[119884119904ℎ3
119904
12+ 119884119901((ℎ119901+ℎ119904
2)
3
minusℎ3
119904
8)] (22)
and the term 120599 represents the coupling between the strain andthe voltage and is given by
120599 =11988931119884119901119887
ℎ119901
[ℎ2
119904
4minus (ℎ119901+ℎ119904
2)
2
] (23)
Substituting (21) in (14) we obtain the mechanical equa-tion of motion with electrical coupling
1198841198681205974119908rel (119909 119905)
1205971199094+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
+ 120599V (119905) [120597120575 (119909)
120597119909minus120597120575 (119909 minus 119871)
120597119909]
= minus [119898 +119872119905120575 (119909 minus 119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(24)
32 Electrical Equation To derive the electrical equation letus start with the piezoelectric constitutive relation given in
119863 = 11988931119879119901+ 120576119879
33119864 (25)
where119863 is the electric displacement and 12057611987933is the permittivity
at constant stress Replacing the definition of 119879119901 from (17)yields
119863(119909 119905) = 11988931 (119884119901119878119901minus 11988931119864) + 120576
119879
33119864
= 11988931119884119901119878119901+ 119864 (120576
119879
33minus 1198892
31119884119901)
= 11988931119884119901119878119901minus 120576119878
33
V (119905)ℎ119901
(26)
with 12057611987833= 1198892
31119884119901minus 120576119879
33 The strain 119878119901 evaluated at the middle
of the PZT layer is given by
119878119901= minusℎ119901119888
1205972119908rel (119909 119905)
1205971199092 (27)
where ℎ119901119888= ℎ1199042 + ℎ
1199012 is the distance of the center of the
PZT layer to the neural axis Thus
119863 (119909 119905) = minus11988931119884119901ℎ119901119888
1205972119908rel (119909 119905)
1205971199092minus 120576119878
33
V (119905)ℎ119901
(28)
The electric charge is obtained by integrating the electricdisplacement over the area of the beamThe resulting expres-sion is
119902 (119905) = minusint
119871
0
(11988931119884119901ℎ119901119888
1205972119908rel (119909 119905)
1205971199092+ 120576119878
33
V (119905)ℎ119901
)119887119889119909 (29)
Shock and Vibration 7
Therefore the current generated by one PZT layer is
119894 (119905) =120597119902 (119905)
120597119905
= minus120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(30)
Since both layers are connected in parallel the currentgenerated by the bimorph beam is given by
119894 (119905) = 1198941 (119905) + 1198942 (119905)
= minus2120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 2int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(31)
If the beam is connected directly to a resistive load the voltageis given by V(119905) = 119877
119897119894(119905) Substituting this relation in (31) yields
V (119905) = minus2119877119897 [120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
+ int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909]
(32)
The electrical equation with mechanical coupling isobtained by rearranging (32)
V (119905)2119877119897
+120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
= minusint
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(33)
4 Voltage and Power Output of an Array ofPiezoelectric Beams Connected by Springs
Equations (24) and (33) are the electromechanical equationsfor a single bimorph beam under transverse vibrations Letus consider the case of multiple beams connected by springswith a parallel electrical connection as shown in Figure 5 Inthat case the mechanical equation of motion is given by
1198841198681205974119908rel (119909 119905)
1205971199094+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
+ 120599
119873
sum
119894=1
V119894 (119905) [
120597120575 (119909 minus (119894 minus 1) 119871)
120597119909minus120597120575 (119909 minus 119894119871)
120597119909]
= minus[119898 +
119873
sum
119894=1
119872119905119894120575 (119909 minus 119894119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(34)
(t)Rl
Figure 5 A parallel electrical connection of an array of bimorphbeams connected by springs
where V119894(119905) is the voltage of the 119894th beam Since the beams are
electrically connected in parallel they have the same voltagethus V
1(119905) = V
2(119905) = sdot sdot sdot = V
119899(119905) = V(119905) It has been assumed
that all the beams are identical except for the tip mass Itshould be noted that the springs are included as boundaryconditionsTherefore they affect the vibrationmodes but notthe equation of motion
The solution for the relative motion of the system can berepresented by a sum of its mode shapes as
119908rel (119909 119905) = sum119903
120601119903 (119909) 120578119903 (119905) (35)
where 120601119903(119909) and 120578
119903(119905) are the 119903th mode shape and modal
coordinate respectively The 119903th mode shape can be definedby segments as
120601119903 (119909)
=
1206011199031 (119909) 0 le 119909 le 119871
1206011199032 (119909 minus 119871) 119871 le 119909 le 2119871
120601119903119873 (119909 minus (119873 minus 1) 119871) (119873 minus 1) 119871 le 119909 le 119873119871
(36)
where 120601119903119894
is the 119903th mode shape of the 119894th beam 119871 is thelength of one beam and119873 is the number of beams
Replacing (35) themechanical equation ofmotion can bewritten as
[1205962
119903120578119903 (119905) + 2120585119903120596119903 120578119903 (119905) + 120578
119903 (119905)] 120572119903 +
119873
sum
119894=1
120594119903119894V (119905)
= [minus1198981205972119908119887 (119909 119905)
1205971199052minus 119888119886
120597119908119887 (119909 119905)
120597119905] 120574119903
minus1205972119908119887 (119909 119905)
1205971199052
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(37)
8 Shock and Vibration
with
120594119903119894= 120599(minus120601
119903119894 (119871) +120597120601119903119894 (119871)
120597119909)
120585119903=119888119904119868120596119903
2119884119868+
119888119886
2119898120596119903
120572119903= int
119873119871
0
1198981206012
119903(119909) 119889119909
120574119903= int
119873119871
0
120601119903 (119909) 119889119909
(38)
Assuming a harmonic basemotion119908119887(119909 119905) = 119884
0119890119895120596119905 and
a harmonic response 120578119903(119905) = 119873
119903119890119895120596119905 and V(119905) = 119881119890119895120596119905 (37)
becomes
[1205962
119899119903119873119903+ 2119895120585119903120596119899119903120596119873119903minus 1205962119873119903] 120572119903+ 119881
119873
sum
119894=1
120594119903119894
= (1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(39)
Rearranging (39) the amplitude of the temporal term119873119903
can be written as follows
119873119903
=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871) minus 119881sum
119873
119894=1120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(40)
For simplicity (40) can be expressed as
119873119903= 119886119903minus 119881
119873
sum
119894=1
119887119903119894 (41)
with
119886119903=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871)
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
119887119903119894=
120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(42)
The electrical equation needs to be derived for each beamindividually The relative motion for the 119894th beam is given by
119908rel119894 (119909 119905) = sum119903
120601119903119894 (119909) 120578119903 (119905) (43)
Replacing (43) in (30) yields
119894119894 (119905) = minus2
120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 211988931119884119901ℎ119901119888119887sum
119903
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(44)
If the beams are connected directly to a resistive load thevoltage is given by V(119905) = 119877
119897119894(119905) = 119877
119897sum119894119894119894(119905) substituting this
condition equation (44) becomes
V (119905)2119877119897
= minus120576119878
33119887119871119873
ℎ119901
120597V (119905)120597119905
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(45)
Replacing the harmonic response condition in (45) yields
119881
2119877119897
= minus119895120596120576119878
33119887119871119873
ℎ119901
119881
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
119895120596119873119903
(46)
Rearranging (46) we obtain
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596119873119903 (47)
with
120593119903119894= minus
11988931119884119901ℎ119901119888ℎ119901
120576119878
33119871119873
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120581119888= 119895120596 +
ℎ119901
2119877119897120576119878
33119887119871119873
(48)
Replacing 119873119903from (41) in (47) the following equation is
obtained
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596(119886
119903minus 119881
119873
sum
119896=1
119887119903119896) (49)
Rearranging (49) the voltage amplitude can be written asshown in
119881 =sum119903sum119894120593119903119894119895120596119886119903
120581119888+ sum119903sum119894120593119903119894119895120596sum119873
119896=1119887119903119896
(50)
The voltage FRF is defined as the ratio of the voltageoutput to the base acceleration119881minus1205962119884
0 The instantaneous
power output can be obtained from the relation 119875(119905) =V2(119905)119877
119897Therefore the FRF of the power output is simply the
square of the voltage FRF divided by 119877119897
5 Results
51 Analytical This section presents the results of the powerFRF for different number of beams and different combina-tions of tip masses and spring constants in the array Thegeometric material and electromechanical parameters of thebeams are given in Table 1 These are the parameters of theexperimental beams used in the next section
The voltage is computed by (50) and the power FRF iscomputed as 1198812[119877
119897(12059621198840)2] The value of the external load
Shock and Vibration 9
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(a) Connected
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 6 Power FRF with an increasing number of identical bimorph beams119872119905119894=16
= 03119898119871 and 119896119894=15
= 031198841198681198713
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(a) Connected
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 7 Power FRF with an increasing number of bimorph beams with different tip masses Two beams1198721199051= 03119898119871119872
1199052= 035119898119871
119896 = 031198841198681198713 three beams119872
1199051= 025119898119871119872
1199052= 030119898119871119872
1199053= 035119898119871 119896
12= 03119884119868119871
3 six beams1198721199051= 015119898119871119872
1199052= 020119898119871
1198721199053= 030119898119871119872
1199054= 035119898119871119872
1199055= 040119898119871119872
1199056= 045119898119871 119896
119894=15= 03119884119868119871
3
Table 1 Parameters of the bimorph beams
Length of the beam 119871 (mm) 482Width of the beam 119887 (mm) 10Thickness of the substructure ℎ
119904(mm) 019
Thickness of the PZT ℎ119901(mm) 024
Youngrsquos modulus of the substructure 119884119904(GPa) 404
Youngrsquos modulus of the PZT 119884119901(GPa) 55
Mass density of the substructure 120588119904(kgm3) 1519
Mass density of the PZT 120588119901(kgm3) 8140
Piezoelectric constant 11988931(pmV) minus180
Permittivity 12057611990433(nFm) 1328
Strain rate damping 119888119904119868 (kgm4s) 116 times 10minus6
Viscous air damping 119888119886(kgs) 04
resistance is defined for each case as the optimum resistorthat maximizes the power output To study the advantagesand disadvantages of the proposed system the results arecompared with those obtained by an array of disconnectedbeams (with no springs)
First the case of identical beams is investigated theresults are shown in Figure 6(a) It is obtained that in thecase of identical beams the only modes that are excitedare the modes where all beams move in phase such asmodes 1 4 and 7 in Figure 3 In this case it is also foundthat the natural frequencies do not depend on the springstiffness because all beams are in phase and the springs donot deform Furthermore the results are exactly the same asthe ones obtained with disconnected beams as illustrated inFigure 6(b)
In the case of beams with different tip masses the resultsof the FRF power do depend on the springs connecting thebeams as shown in Figure 7 If the beams are disconnectedand new beams with different tip masses are added to thearray then the frequency band of the device is increasedobtaining a similar power amplitude for all resonant frequen-cies On the other hand if the beams are connected by springand new beams are included in the array then the frequencyband is increased but the amplitude related to each resonantfrequency varies
For an array of piezoelectric beams connected by springsthe frequency band of operation and the power amplitude
10 Shock and Vibration
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
01mL 02mL01mL 03mL
01mL 04mL01mL 05mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k = 02YIL3k = 03YIL3
k = 04YIL3
(b) Varying spring constants (1198721199051 = 03119898119871 1198721199052 = 04119898119871)
Figure 8 Power FRF of two bimorph beams with different combinations of tip masses and spring constants
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
02mL 03mL 04mL02mL 02mL 04mL
03mL 03mL 04mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k =
k =k = 02YIL3 02YIL302YIL3 04YIL3
02YIL3 06YIL3
(b) Varying spring constants (1198721199051 = 02119898119871 1198721199052 = 03119898119871 1198721199053 =04119898119871)
Figure 9 Power FRF of three bimorph beams with different combinations of tip masses and spring constants
associated with each resonant frequency depend on thecombination of tipmasses and springs constantsThis ismoreevident in Figures 8 and 9 where different combinations ofsprings and tip masses are studied for the case of two andthree beams In general increasing the tip mass of the beamsdecreases someof the resonant frequencies and increasing thespring stiffness increases the frequency band
To compare the performances of an array of beamsconnected by springs and one of the disconnected beams acase of ten beams with both situations is plotted in Figure 10It is possible to see that the array with connected beamshas a slightly wider frequency band which increases withan increment of the springs constants Furthermore bycomparing the areas below both curves in the frequencyrange 50ndash200Hz the array with connected beams providesan area between 20 and 36 larger than the array withdisconnected beams Thus by connecting the beams theperformance of the array can be improved It should benoted that the previous results are obtained with arbitraryvalues for the tip masses and spring constantsTherefore thisperformance can be further improved by a proper design
52 Experimental To validate the proposed model a pre-liminary experimental campaign was performed Figure 11presents the experimental setup It consists of a signal genera-tor connected to a power amplifier and to an electrodynamicshaker which produces a controlled base vibration An arrayof three piezoelectric bimorph beams is mounted over theshaker Figure 12 shows this array The cantilever bimorphbeams have a tip mass at the free-end and can be connectedwith springs between them Three experiments were per-formed In the first case no springs were used and eachbeam was connected independently to a 1 kΩ resistive loadin this case the voltage generated by each beam is monitoredindividually In the second case no springs were used andthe beams were electrically connected in parallel to a resistiveload of 1 kΩ In the last case the beams were connected withsprings between them and they were electrically connected inparallel to a resistive load of 1 kΩ
The properties of the bimorph beams are the ones givenin Table 1 the tip masses are119872
1199051= 65119898119871119872
1199052= 35119898119871
1198721199053= 19119898119871 and the spring constants are 119896
1= 1198962=
2921198841198681198713
Shock and Vibration 11
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected02YIL3
(a) 119896119894=19 = 021198841198681198713
50 100 150 200
Frequency (Hz)
10minus4
Pow
er F
RF (W
s4m
2)
10minus3
10minus2
10minus1
100
Connected k =
Disconnected03YIL3
(b) 119896119894=19 = 031198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected04YIL3
(c) 119896119894=19 = 041198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected05YIL3
(d) 119896119894=19 = 051198841198681198713
Figure 10 Power FRF with ten bimorph beams1198721199051= 01119898119871119872
1199052= 015119898119871119872
1199053= 02119898119871119872
1199054= 025119898119871119872
1199055= 03119898119871119872
1199056= 035119898119871
1198721199057= 04119898119871119872
1199058= 045119898119871119872
1199059= 05119898119871119872
11990510= 055119898119871
Figure 11 Experimental setup
Figure 12 Experimental bimorph beams
During the three experiments the system is subjected toa harmonic base displacement with frequencies in the range20ndash70Hz The results for the first experiment are presentedin Figure 13(a) and are compared with the ones obtainedby an analytical model in Figure 13(b) Both analytical andexperimental curves are very similar The only difference isthat the experimental power increases from beams 1 to 3 andthe analytical model predicted the opposite behavior Thisunexpected behavior seems to be related to a resonance ofthe supporting structure in the three experimental curves itis possible to see an increment of the power near 60Hz whichcan be explained by a resonance of the structure
Figure 14 presents the experimental and analytical resultsfor the second and third experiments It is possible to seea good agreement between the experimental and analyticalcurves although similarly to the results in Figure 13(a) thereis an increment of the power near 60Hz that confirms astructural resonance
Comparing the areas below both experimental curves inFigure 13(b) in the frequency range 20ndash70Hz the array withconnected beams provides an area 5 larger than the arraywith disconnected beamswhich confirms that the addition ofsprings can increment the generated powerThese results canbe improved by using softer springs which have lower energylosses and by solving the structural resonance problem
12 Shock and Vibration
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 13 Power FRF for the three beams with no springs and independent electrical connections to 1 kΩ resistive loads (a) experimentalmeasurement (b) analytical results
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 14 Power FRF for the three beams with and without springs and electrical connection in parallel to a resistive load of 1 kΩ(a) experimental measurement (b) analytical results
6 Conclusions
This paper investigated the use of an array of bimorphbeams connected by springs as a broadband vibration-basedenergy harvester The equations for the voltage and poweroutput of the full system have been deduced based on theanalytical solution of the cantilevered piezoelectric energyharvester with Euler-Bernoulli beam assumptions To studythe advantages and disadvantages of the proposed system theresults are compared with those of an array of disconnectedbeams (with no springs) The analytical model is validatedwith experimental measurements of three bimorph beamswith and without springs
Both analytical and experimental results show that con-necting the array of beams with springs allows increasingthe frequency band of operation and increasing the amountof power generated Thus the maximal frequency band fora certain array of beams can be extended by connectingthem by springs and at the same time the output power isincreased The performance of this system can be optimizedfor specific applications by a proper selection of the tipmassesand spring constants
The preliminary experimental results are consistent withthe analytical ones but further research is necessary toimprove the experimental setup and to investigate differentexperimental conditions such as variations of the springsstiffness
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
References
[1] S R Anton and H A Sodano ldquoA review of power harvestingusing piezoelectric materials (2003ndash2006)rdquo Smart Materialsand Structures vol 16 no 3 article R1 2007
[2] S Chalasani and J M Conrad ldquoA survey of energy harvestingsources for embedded systemsrdquo in Proceedings of the IEEESoutheastCon pp 442ndash447 April 2008
[3] K A Cook-Chennault NThambi and AM Sastry ldquoPoweringMEMS portable devicesmdasha review of non-regenerative andregenerative power supply systems with special emphasis on
Shock and Vibration 13
piezoelectric energy harvesting systemsrdquo Smart Materials andStructures vol 17 no 4 Article ID 043001 2008
[4] A Dewan S U Ay M N Karim and H Beyenal ldquoAlternativepower sources for remote sensors a reviewrdquo Journal of PowerSources vol 245 pp 129ndash143 2014
[5] A Harb ldquoEnergy harvesting state-of-the-artrdquo RenewableEnergy vol 36 no 10 pp 2641ndash2654 2011
[6] S Saadon and O Sidek ldquoA review of vibration-based MEMSpiezoelectric energy harvestersrdquo Energy Conversion and Man-agement vol 52 no 1 pp 500ndash504 2011
[7] H A Sodano D J Inman and G Park ldquoA review of powerharvesting from vibration using piezoelectric materialsrdquo Shockand Vibration Digest vol 36 no 3 pp 197ndash205 2004
[8] R J M Vullers R van Schaijk I Doms C Van Hoofand R Mertens ldquoMicropower energy harvestingrdquo Solid-StateElectronics vol 53 no 7 pp 684ndash693 2009
[9] R Calio U B Rongala D Camboni et al ldquoPiezoelectric energyharvesting solutionsrdquo Sensors vol 14 no 3 pp 4755ndash4790 2014
[10] J Twiefel and H Westermann ldquoSurvey on broadband tech-niques for vibration energy harvestingrdquo Journal of IntelligentMaterial Systems and Structures vol 24 no 11 pp 1291ndash13022013
[11] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavengingrdquo Journal of Sound and Vibration vol 292no 3ndash5 pp 987ndash998 2006
[12] S M Shahruz ldquoLimits of performance of mechanical band-pass filters used in energy scavengingrdquo Journal of Sound andVibration vol 293 no 1-2 pp 449ndash461 2006
[13] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavenging multi-degree-of-freedom modelsrdquo Journalof Vibration and Control vol 14 no 5 pp 753ndash768 2008
[14] J-Q Liu H-B Fang Z-Y Xu et al ldquoA MEMS-based piezo-electric power generator array for vibration energy harvestingrdquoMicroelectronics Journal vol 39 no 5 pp 802ndash806 2008
[15] HC Lin PHWu I C Lien andYC Shu ldquoAnalysis of an arrayof piezo-electric energy harvesters connected in seriesrdquo SmartMaterials and Structures vol 22 no 9 Article ID 094026 2013
[16] S-J Jang E Rustighi M J Brennan Y P Lee and H-JJung ldquoDesign of a 2DOF vibrational energy harvesting devicerdquoJournal of IntelligentMaterial Systems and Structures vol 22 no5 pp 443ndash448 2011
[17] L Tang and Y Yang ldquoA multiple-degree-of-freedom piezoelec-tric energy harvesting modelrdquo Journal of Intelligent MaterialSystems and Structures vol 23 no 14 pp 1631ndash1647 2012
[18] Z Yang and J Yang ldquoConnected vibrating piezoelectricbimorph beams as a wide-band piezoelectric power harvesterrdquoJournal of IntelligentMaterial Systems and Structures vol 20 no5 pp 569ndash574 2009
[19] W Zhou G R Penamalli and L Zuo ldquoAn efficient vibrationenergy harvester with a multi-mode dynamic magnifierrdquo SmartMaterials and Structures vol 21 no 1 Article ID 015014 2012
[20] D S Nguyen E Halvorsen G U Jensen and A VoglldquoFabrication and characterization of a widebandMEMS energyharvester utilizing nonlinear springsrdquo Journal of Micromechan-ics andMicroengineering vol 20 no 12 Article ID 125009 2010
[21] S D Nguyen and E Halvorsen ldquoNonlinear springs forbandwidth-tolerant vibration energy harvestingrdquo Journal ofMicroelectromechanical Systems vol 20 no 6 pp 1225ndash12272011
[22] S BoisseauGDespesse andBA Seddik ldquoNonlinear h-shapedsprings to improve efficiency of vibration energy harvestersrdquo
Journal of Applied Mechanics vol 80 no 6 Article ID 0610132013
[23] S Priya ldquoAdvances in energy harvesting using low profilepiezoelectric transducersrdquo Journal of Electroceramics vol 19 no1 pp 165ndash182 2007
[24] S Roundy P K Wright and J Rabaey ldquoA study of lowlevel vibrations as a power source for wireless sensor nodesrdquoComputer Communications vol 26 no 11 pp 1131ndash1144 2003
[25] S Roundy and P K Wright ldquoA piezoelectric vibration basedgenerator for wireless electronicsrdquo Smart Materials and Struc-tures vol 13 no 5 pp 1131ndash1142 2004
[26] H A Sodano G Park and D J Inman ldquoEstimation of electriccharge output for piezoelectric energy harvestingrdquo Strain vol40 no 2 pp 49ndash58 2004
[27] N E Dutoit B L Wardle and S-G Kim ldquoDesign considera-tions for mems-scale piezoelectric mechanical vibration energyharvestersrdquo Integrated Ferroelectrics vol 71 no 1 pp 121ndash1602005
[28] S-N Chen G-J Wang andM-C Chien ldquoAnalytical modelingof piezoelectric vibration-induced micro power generatorrdquoMechatronics vol 16 no 7 pp 379ndash387 2006
[29] A Erturk and D J Inman ldquoIssues in mathematical modelingof piezoelectric energy harvestersrdquo Smart Materials and Struc-tures vol 17 no 6 Article ID 065016 2008
[30] F Lu H P Lee and S P Lim ldquoModeling and analysis of micropiezoelectric power generators for micro-electromechanical-systems applicationsrdquo Smart Materials and Structures vol 13no 1 pp 57ndash63 2004
[31] S Jiang X Li S Guo Y Hu J Yang andQ Jiang ldquoPerformanceof a piezoelectric bimorph for scavenging vibration energyrdquoSmartMaterials and Structures vol 14 no 4 pp 769ndash774 2005
[32] A Erturk and D J Inman ldquoA distributed parameter elec-tromechanical model for cantilevered piezoelectric energy har-vestersrdquo Journal of Vibration and Acoustics vol 130 no 4Article ID 041002 2008
[33] A Erturk and D J Inman ldquoAn experimentally validatedbimorph cantilever model for piezoelectric energy harvestingfrom base excitationsrdquo Smart Materials and Structures vol 18no 2 Article ID 025009 2009
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
Shock and Vibration 5
120573rL = 1536
(a) Mode 1
120573rL = 1573
(b) Mode 2
120573rL = 1640
(c) Mode 3
120573rL = 4192
(d) Mode 4 (e) Mode 5 (f) Mode 6
(g) Mode 7 (h) Mode 8 (i) Mode 9
Figure 3 First nine vibration modes of three elastically connected beams with identical tip masses and spring constants (119872119905119894=13
=
03119898119871 11989612= 03119864119868119871
3)
beams In this example all beams have a mass tip of 03119898119871and all springs have a stiffness of 031198841198681198713
3 Electromechanical Model of a BimorphPiezoelectric Beam
31 Mechanical Equation Let us consider the bimorph can-tilever beam shown in Figure 4 The bimorph beam consistsin two layers of piezoelectric material (PZT) bonded to asubstructure layer Assuming an Euler-Bernoulli beam theequation of motion can be written as [33]
1205972119872(119909 119905)
1205971199092+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
= minus [119898 +119872119905120575 (119909 minus 119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(14)
b
x(t)
Mt
hp
hp
hs
PZT
PZTSubstructureY0e
i120596t
Figure 4 Scheme of the bimorph piezoelectric beam
where 119908rel(119909 119905) is the transverse deflection of the beamrelative to the base 119908
119887(119905) is the base displacement119872(119909 119905) is
the internal bending moment 119888119904is the equivalent coefficient
of strain rate damping 119888119886is the viscous air damping 119868 is the
area moment of inertia of the cross section119898 is the mass perunit length 119872
119905is the tip mass and 120575(119909) is the Dirac delta
function
6 Shock and Vibration
The internal moment is given by the integral of the stressdistribution over the cross section of the beam shown in
119872(119909 119905) = minus119887 [int
minusℎ1199042
minusℎ119901minusℎ1199042
119879119901119910119889119910 + int
ℎ1199042
minusℎ1199042
119879119904119910119889119910
+ int
ℎ1199042+ℎ119901
ℎ1199042
119879119901119910119889119910]
(15)
where 119887 is the width of the beam ℎ119901is the thickness of
each PZT layer ℎ119904is the thickness of the substructure (see
Figure 4) and 119879 is the stress Superscripts 119901 and 119904 stand forPZT and substructure respectively The stress distributionacross the section is given by the strain-stress relations of thesubstructure and PZT layers which are given by
119879119904= 119884119904119878119904 (16)
119879119901= 119884119901119878119901minus 11988931119884119901119864 (17)
where 119878 is the strain 119884 is Youngrsquos modulus 119889 is thepiezoelectric constant and 119864 is the electric fieldThe bendingstrain can be expressed in terms of the radius of curvature as
119878 (119909 119910 119905) = minus1199101205972119908rel (119909 119905)
1205971199092 (18)
Since the piezoelectric layers are connected in parallelthey have the same voltage but the instantaneous electricfields are in opposite directions The electric field can bewritten in terms of the voltage across the PZT V(119905) and thethickness of the PZT layer ℎ
119901 by the following equation
1198641 (119905) = minus
V (119905)ℎ119901
in the top layer
1198642 (119905) =
V (119905)ℎ119901
in the bottom layer(19)
Replacing (16) to (19) in (15) we obtain
119872(119909 119905) =2119887
3[119884119904ℎ3
119904
12+ 119884119901((ℎ119901+ℎ119904
2)
3
minusℎ3
119904
8)]
sdot1205972119908rel (119909 119905)
1205971199092
+11988931119884119901119887
ℎ119901
[ℎ2
119904
4minus (ℎ119901+ℎ119904
2)
2
] V (119905)
(20)
Equation (20) can be expressed as
119872(119909 119905) = 1198841198681205972119908rel (119909 119905)
1205971199092
+ 120599V (119905) [Φ (119909) minus Φ (119909 minus 119871)] (21)
where 119871 is the length of the beam and Φ the Heavisidefunction It has been assumed that the PZT layer covers
the entire length of the beam The constant 119884119868 represents thebending stiffness of the beam and is given by
119884119868 =2119887
3[119884119904ℎ3
119904
12+ 119884119901((ℎ119901+ℎ119904
2)
3
minusℎ3
119904
8)] (22)
and the term 120599 represents the coupling between the strain andthe voltage and is given by
120599 =11988931119884119901119887
ℎ119901
[ℎ2
119904
4minus (ℎ119901+ℎ119904
2)
2
] (23)
Substituting (21) in (14) we obtain the mechanical equa-tion of motion with electrical coupling
1198841198681205974119908rel (119909 119905)
1205971199094+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
+ 120599V (119905) [120597120575 (119909)
120597119909minus120597120575 (119909 minus 119871)
120597119909]
= minus [119898 +119872119905120575 (119909 minus 119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(24)
32 Electrical Equation To derive the electrical equation letus start with the piezoelectric constitutive relation given in
119863 = 11988931119879119901+ 120576119879
33119864 (25)
where119863 is the electric displacement and 12057611987933is the permittivity
at constant stress Replacing the definition of 119879119901 from (17)yields
119863(119909 119905) = 11988931 (119884119901119878119901minus 11988931119864) + 120576
119879
33119864
= 11988931119884119901119878119901+ 119864 (120576
119879
33minus 1198892
31119884119901)
= 11988931119884119901119878119901minus 120576119878
33
V (119905)ℎ119901
(26)
with 12057611987833= 1198892
31119884119901minus 120576119879
33 The strain 119878119901 evaluated at the middle
of the PZT layer is given by
119878119901= minusℎ119901119888
1205972119908rel (119909 119905)
1205971199092 (27)
where ℎ119901119888= ℎ1199042 + ℎ
1199012 is the distance of the center of the
PZT layer to the neural axis Thus
119863 (119909 119905) = minus11988931119884119901ℎ119901119888
1205972119908rel (119909 119905)
1205971199092minus 120576119878
33
V (119905)ℎ119901
(28)
The electric charge is obtained by integrating the electricdisplacement over the area of the beamThe resulting expres-sion is
119902 (119905) = minusint
119871
0
(11988931119884119901ℎ119901119888
1205972119908rel (119909 119905)
1205971199092+ 120576119878
33
V (119905)ℎ119901
)119887119889119909 (29)
Shock and Vibration 7
Therefore the current generated by one PZT layer is
119894 (119905) =120597119902 (119905)
120597119905
= minus120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(30)
Since both layers are connected in parallel the currentgenerated by the bimorph beam is given by
119894 (119905) = 1198941 (119905) + 1198942 (119905)
= minus2120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 2int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(31)
If the beam is connected directly to a resistive load the voltageis given by V(119905) = 119877
119897119894(119905) Substituting this relation in (31) yields
V (119905) = minus2119877119897 [120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
+ int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909]
(32)
The electrical equation with mechanical coupling isobtained by rearranging (32)
V (119905)2119877119897
+120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
= minusint
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(33)
4 Voltage and Power Output of an Array ofPiezoelectric Beams Connected by Springs
Equations (24) and (33) are the electromechanical equationsfor a single bimorph beam under transverse vibrations Letus consider the case of multiple beams connected by springswith a parallel electrical connection as shown in Figure 5 Inthat case the mechanical equation of motion is given by
1198841198681205974119908rel (119909 119905)
1205971199094+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
+ 120599
119873
sum
119894=1
V119894 (119905) [
120597120575 (119909 minus (119894 minus 1) 119871)
120597119909minus120597120575 (119909 minus 119894119871)
120597119909]
= minus[119898 +
119873
sum
119894=1
119872119905119894120575 (119909 minus 119894119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(34)
(t)Rl
Figure 5 A parallel electrical connection of an array of bimorphbeams connected by springs
where V119894(119905) is the voltage of the 119894th beam Since the beams are
electrically connected in parallel they have the same voltagethus V
1(119905) = V
2(119905) = sdot sdot sdot = V
119899(119905) = V(119905) It has been assumed
that all the beams are identical except for the tip mass Itshould be noted that the springs are included as boundaryconditionsTherefore they affect the vibrationmodes but notthe equation of motion
The solution for the relative motion of the system can berepresented by a sum of its mode shapes as
119908rel (119909 119905) = sum119903
120601119903 (119909) 120578119903 (119905) (35)
where 120601119903(119909) and 120578
119903(119905) are the 119903th mode shape and modal
coordinate respectively The 119903th mode shape can be definedby segments as
120601119903 (119909)
=
1206011199031 (119909) 0 le 119909 le 119871
1206011199032 (119909 minus 119871) 119871 le 119909 le 2119871
120601119903119873 (119909 minus (119873 minus 1) 119871) (119873 minus 1) 119871 le 119909 le 119873119871
(36)
where 120601119903119894
is the 119903th mode shape of the 119894th beam 119871 is thelength of one beam and119873 is the number of beams
Replacing (35) themechanical equation ofmotion can bewritten as
[1205962
119903120578119903 (119905) + 2120585119903120596119903 120578119903 (119905) + 120578
119903 (119905)] 120572119903 +
119873
sum
119894=1
120594119903119894V (119905)
= [minus1198981205972119908119887 (119909 119905)
1205971199052minus 119888119886
120597119908119887 (119909 119905)
120597119905] 120574119903
minus1205972119908119887 (119909 119905)
1205971199052
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(37)
8 Shock and Vibration
with
120594119903119894= 120599(minus120601
119903119894 (119871) +120597120601119903119894 (119871)
120597119909)
120585119903=119888119904119868120596119903
2119884119868+
119888119886
2119898120596119903
120572119903= int
119873119871
0
1198981206012
119903(119909) 119889119909
120574119903= int
119873119871
0
120601119903 (119909) 119889119909
(38)
Assuming a harmonic basemotion119908119887(119909 119905) = 119884
0119890119895120596119905 and
a harmonic response 120578119903(119905) = 119873
119903119890119895120596119905 and V(119905) = 119881119890119895120596119905 (37)
becomes
[1205962
119899119903119873119903+ 2119895120585119903120596119899119903120596119873119903minus 1205962119873119903] 120572119903+ 119881
119873
sum
119894=1
120594119903119894
= (1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(39)
Rearranging (39) the amplitude of the temporal term119873119903
can be written as follows
119873119903
=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871) minus 119881sum
119873
119894=1120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(40)
For simplicity (40) can be expressed as
119873119903= 119886119903minus 119881
119873
sum
119894=1
119887119903119894 (41)
with
119886119903=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871)
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
119887119903119894=
120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(42)
The electrical equation needs to be derived for each beamindividually The relative motion for the 119894th beam is given by
119908rel119894 (119909 119905) = sum119903
120601119903119894 (119909) 120578119903 (119905) (43)
Replacing (43) in (30) yields
119894119894 (119905) = minus2
120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 211988931119884119901ℎ119901119888119887sum
119903
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(44)
If the beams are connected directly to a resistive load thevoltage is given by V(119905) = 119877
119897119894(119905) = 119877
119897sum119894119894119894(119905) substituting this
condition equation (44) becomes
V (119905)2119877119897
= minus120576119878
33119887119871119873
ℎ119901
120597V (119905)120597119905
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(45)
Replacing the harmonic response condition in (45) yields
119881
2119877119897
= minus119895120596120576119878
33119887119871119873
ℎ119901
119881
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
119895120596119873119903
(46)
Rearranging (46) we obtain
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596119873119903 (47)
with
120593119903119894= minus
11988931119884119901ℎ119901119888ℎ119901
120576119878
33119871119873
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120581119888= 119895120596 +
ℎ119901
2119877119897120576119878
33119887119871119873
(48)
Replacing 119873119903from (41) in (47) the following equation is
obtained
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596(119886
119903minus 119881
119873
sum
119896=1
119887119903119896) (49)
Rearranging (49) the voltage amplitude can be written asshown in
119881 =sum119903sum119894120593119903119894119895120596119886119903
120581119888+ sum119903sum119894120593119903119894119895120596sum119873
119896=1119887119903119896
(50)
The voltage FRF is defined as the ratio of the voltageoutput to the base acceleration119881minus1205962119884
0 The instantaneous
power output can be obtained from the relation 119875(119905) =V2(119905)119877
119897Therefore the FRF of the power output is simply the
square of the voltage FRF divided by 119877119897
5 Results
51 Analytical This section presents the results of the powerFRF for different number of beams and different combina-tions of tip masses and spring constants in the array Thegeometric material and electromechanical parameters of thebeams are given in Table 1 These are the parameters of theexperimental beams used in the next section
The voltage is computed by (50) and the power FRF iscomputed as 1198812[119877
119897(12059621198840)2] The value of the external load
Shock and Vibration 9
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(a) Connected
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 6 Power FRF with an increasing number of identical bimorph beams119872119905119894=16
= 03119898119871 and 119896119894=15
= 031198841198681198713
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(a) Connected
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 7 Power FRF with an increasing number of bimorph beams with different tip masses Two beams1198721199051= 03119898119871119872
1199052= 035119898119871
119896 = 031198841198681198713 three beams119872
1199051= 025119898119871119872
1199052= 030119898119871119872
1199053= 035119898119871 119896
12= 03119884119868119871
3 six beams1198721199051= 015119898119871119872
1199052= 020119898119871
1198721199053= 030119898119871119872
1199054= 035119898119871119872
1199055= 040119898119871119872
1199056= 045119898119871 119896
119894=15= 03119884119868119871
3
Table 1 Parameters of the bimorph beams
Length of the beam 119871 (mm) 482Width of the beam 119887 (mm) 10Thickness of the substructure ℎ
119904(mm) 019
Thickness of the PZT ℎ119901(mm) 024
Youngrsquos modulus of the substructure 119884119904(GPa) 404
Youngrsquos modulus of the PZT 119884119901(GPa) 55
Mass density of the substructure 120588119904(kgm3) 1519
Mass density of the PZT 120588119901(kgm3) 8140
Piezoelectric constant 11988931(pmV) minus180
Permittivity 12057611990433(nFm) 1328
Strain rate damping 119888119904119868 (kgm4s) 116 times 10minus6
Viscous air damping 119888119886(kgs) 04
resistance is defined for each case as the optimum resistorthat maximizes the power output To study the advantagesand disadvantages of the proposed system the results arecompared with those obtained by an array of disconnectedbeams (with no springs)
First the case of identical beams is investigated theresults are shown in Figure 6(a) It is obtained that in thecase of identical beams the only modes that are excitedare the modes where all beams move in phase such asmodes 1 4 and 7 in Figure 3 In this case it is also foundthat the natural frequencies do not depend on the springstiffness because all beams are in phase and the springs donot deform Furthermore the results are exactly the same asthe ones obtained with disconnected beams as illustrated inFigure 6(b)
In the case of beams with different tip masses the resultsof the FRF power do depend on the springs connecting thebeams as shown in Figure 7 If the beams are disconnectedand new beams with different tip masses are added to thearray then the frequency band of the device is increasedobtaining a similar power amplitude for all resonant frequen-cies On the other hand if the beams are connected by springand new beams are included in the array then the frequencyband is increased but the amplitude related to each resonantfrequency varies
For an array of piezoelectric beams connected by springsthe frequency band of operation and the power amplitude
10 Shock and Vibration
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
01mL 02mL01mL 03mL
01mL 04mL01mL 05mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k = 02YIL3k = 03YIL3
k = 04YIL3
(b) Varying spring constants (1198721199051 = 03119898119871 1198721199052 = 04119898119871)
Figure 8 Power FRF of two bimorph beams with different combinations of tip masses and spring constants
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
02mL 03mL 04mL02mL 02mL 04mL
03mL 03mL 04mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k =
k =k = 02YIL3 02YIL302YIL3 04YIL3
02YIL3 06YIL3
(b) Varying spring constants (1198721199051 = 02119898119871 1198721199052 = 03119898119871 1198721199053 =04119898119871)
Figure 9 Power FRF of three bimorph beams with different combinations of tip masses and spring constants
associated with each resonant frequency depend on thecombination of tipmasses and springs constantsThis ismoreevident in Figures 8 and 9 where different combinations ofsprings and tip masses are studied for the case of two andthree beams In general increasing the tip mass of the beamsdecreases someof the resonant frequencies and increasing thespring stiffness increases the frequency band
To compare the performances of an array of beamsconnected by springs and one of the disconnected beams acase of ten beams with both situations is plotted in Figure 10It is possible to see that the array with connected beamshas a slightly wider frequency band which increases withan increment of the springs constants Furthermore bycomparing the areas below both curves in the frequencyrange 50ndash200Hz the array with connected beams providesan area between 20 and 36 larger than the array withdisconnected beams Thus by connecting the beams theperformance of the array can be improved It should benoted that the previous results are obtained with arbitraryvalues for the tip masses and spring constantsTherefore thisperformance can be further improved by a proper design
52 Experimental To validate the proposed model a pre-liminary experimental campaign was performed Figure 11presents the experimental setup It consists of a signal genera-tor connected to a power amplifier and to an electrodynamicshaker which produces a controlled base vibration An arrayof three piezoelectric bimorph beams is mounted over theshaker Figure 12 shows this array The cantilever bimorphbeams have a tip mass at the free-end and can be connectedwith springs between them Three experiments were per-formed In the first case no springs were used and eachbeam was connected independently to a 1 kΩ resistive loadin this case the voltage generated by each beam is monitoredindividually In the second case no springs were used andthe beams were electrically connected in parallel to a resistiveload of 1 kΩ In the last case the beams were connected withsprings between them and they were electrically connected inparallel to a resistive load of 1 kΩ
The properties of the bimorph beams are the ones givenin Table 1 the tip masses are119872
1199051= 65119898119871119872
1199052= 35119898119871
1198721199053= 19119898119871 and the spring constants are 119896
1= 1198962=
2921198841198681198713
Shock and Vibration 11
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected02YIL3
(a) 119896119894=19 = 021198841198681198713
50 100 150 200
Frequency (Hz)
10minus4
Pow
er F
RF (W
s4m
2)
10minus3
10minus2
10minus1
100
Connected k =
Disconnected03YIL3
(b) 119896119894=19 = 031198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected04YIL3
(c) 119896119894=19 = 041198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected05YIL3
(d) 119896119894=19 = 051198841198681198713
Figure 10 Power FRF with ten bimorph beams1198721199051= 01119898119871119872
1199052= 015119898119871119872
1199053= 02119898119871119872
1199054= 025119898119871119872
1199055= 03119898119871119872
1199056= 035119898119871
1198721199057= 04119898119871119872
1199058= 045119898119871119872
1199059= 05119898119871119872
11990510= 055119898119871
Figure 11 Experimental setup
Figure 12 Experimental bimorph beams
During the three experiments the system is subjected toa harmonic base displacement with frequencies in the range20ndash70Hz The results for the first experiment are presentedin Figure 13(a) and are compared with the ones obtainedby an analytical model in Figure 13(b) Both analytical andexperimental curves are very similar The only difference isthat the experimental power increases from beams 1 to 3 andthe analytical model predicted the opposite behavior Thisunexpected behavior seems to be related to a resonance ofthe supporting structure in the three experimental curves itis possible to see an increment of the power near 60Hz whichcan be explained by a resonance of the structure
Figure 14 presents the experimental and analytical resultsfor the second and third experiments It is possible to seea good agreement between the experimental and analyticalcurves although similarly to the results in Figure 13(a) thereis an increment of the power near 60Hz that confirms astructural resonance
Comparing the areas below both experimental curves inFigure 13(b) in the frequency range 20ndash70Hz the array withconnected beams provides an area 5 larger than the arraywith disconnected beamswhich confirms that the addition ofsprings can increment the generated powerThese results canbe improved by using softer springs which have lower energylosses and by solving the structural resonance problem
12 Shock and Vibration
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 13 Power FRF for the three beams with no springs and independent electrical connections to 1 kΩ resistive loads (a) experimentalmeasurement (b) analytical results
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 14 Power FRF for the three beams with and without springs and electrical connection in parallel to a resistive load of 1 kΩ(a) experimental measurement (b) analytical results
6 Conclusions
This paper investigated the use of an array of bimorphbeams connected by springs as a broadband vibration-basedenergy harvester The equations for the voltage and poweroutput of the full system have been deduced based on theanalytical solution of the cantilevered piezoelectric energyharvester with Euler-Bernoulli beam assumptions To studythe advantages and disadvantages of the proposed system theresults are compared with those of an array of disconnectedbeams (with no springs) The analytical model is validatedwith experimental measurements of three bimorph beamswith and without springs
Both analytical and experimental results show that con-necting the array of beams with springs allows increasingthe frequency band of operation and increasing the amountof power generated Thus the maximal frequency band fora certain array of beams can be extended by connectingthem by springs and at the same time the output power isincreased The performance of this system can be optimizedfor specific applications by a proper selection of the tipmassesand spring constants
The preliminary experimental results are consistent withthe analytical ones but further research is necessary toimprove the experimental setup and to investigate differentexperimental conditions such as variations of the springsstiffness
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
References
[1] S R Anton and H A Sodano ldquoA review of power harvestingusing piezoelectric materials (2003ndash2006)rdquo Smart Materialsand Structures vol 16 no 3 article R1 2007
[2] S Chalasani and J M Conrad ldquoA survey of energy harvestingsources for embedded systemsrdquo in Proceedings of the IEEESoutheastCon pp 442ndash447 April 2008
[3] K A Cook-Chennault NThambi and AM Sastry ldquoPoweringMEMS portable devicesmdasha review of non-regenerative andregenerative power supply systems with special emphasis on
Shock and Vibration 13
piezoelectric energy harvesting systemsrdquo Smart Materials andStructures vol 17 no 4 Article ID 043001 2008
[4] A Dewan S U Ay M N Karim and H Beyenal ldquoAlternativepower sources for remote sensors a reviewrdquo Journal of PowerSources vol 245 pp 129ndash143 2014
[5] A Harb ldquoEnergy harvesting state-of-the-artrdquo RenewableEnergy vol 36 no 10 pp 2641ndash2654 2011
[6] S Saadon and O Sidek ldquoA review of vibration-based MEMSpiezoelectric energy harvestersrdquo Energy Conversion and Man-agement vol 52 no 1 pp 500ndash504 2011
[7] H A Sodano D J Inman and G Park ldquoA review of powerharvesting from vibration using piezoelectric materialsrdquo Shockand Vibration Digest vol 36 no 3 pp 197ndash205 2004
[8] R J M Vullers R van Schaijk I Doms C Van Hoofand R Mertens ldquoMicropower energy harvestingrdquo Solid-StateElectronics vol 53 no 7 pp 684ndash693 2009
[9] R Calio U B Rongala D Camboni et al ldquoPiezoelectric energyharvesting solutionsrdquo Sensors vol 14 no 3 pp 4755ndash4790 2014
[10] J Twiefel and H Westermann ldquoSurvey on broadband tech-niques for vibration energy harvestingrdquo Journal of IntelligentMaterial Systems and Structures vol 24 no 11 pp 1291ndash13022013
[11] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavengingrdquo Journal of Sound and Vibration vol 292no 3ndash5 pp 987ndash998 2006
[12] S M Shahruz ldquoLimits of performance of mechanical band-pass filters used in energy scavengingrdquo Journal of Sound andVibration vol 293 no 1-2 pp 449ndash461 2006
[13] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavenging multi-degree-of-freedom modelsrdquo Journalof Vibration and Control vol 14 no 5 pp 753ndash768 2008
[14] J-Q Liu H-B Fang Z-Y Xu et al ldquoA MEMS-based piezo-electric power generator array for vibration energy harvestingrdquoMicroelectronics Journal vol 39 no 5 pp 802ndash806 2008
[15] HC Lin PHWu I C Lien andYC Shu ldquoAnalysis of an arrayof piezo-electric energy harvesters connected in seriesrdquo SmartMaterials and Structures vol 22 no 9 Article ID 094026 2013
[16] S-J Jang E Rustighi M J Brennan Y P Lee and H-JJung ldquoDesign of a 2DOF vibrational energy harvesting devicerdquoJournal of IntelligentMaterial Systems and Structures vol 22 no5 pp 443ndash448 2011
[17] L Tang and Y Yang ldquoA multiple-degree-of-freedom piezoelec-tric energy harvesting modelrdquo Journal of Intelligent MaterialSystems and Structures vol 23 no 14 pp 1631ndash1647 2012
[18] Z Yang and J Yang ldquoConnected vibrating piezoelectricbimorph beams as a wide-band piezoelectric power harvesterrdquoJournal of IntelligentMaterial Systems and Structures vol 20 no5 pp 569ndash574 2009
[19] W Zhou G R Penamalli and L Zuo ldquoAn efficient vibrationenergy harvester with a multi-mode dynamic magnifierrdquo SmartMaterials and Structures vol 21 no 1 Article ID 015014 2012
[20] D S Nguyen E Halvorsen G U Jensen and A VoglldquoFabrication and characterization of a widebandMEMS energyharvester utilizing nonlinear springsrdquo Journal of Micromechan-ics andMicroengineering vol 20 no 12 Article ID 125009 2010
[21] S D Nguyen and E Halvorsen ldquoNonlinear springs forbandwidth-tolerant vibration energy harvestingrdquo Journal ofMicroelectromechanical Systems vol 20 no 6 pp 1225ndash12272011
[22] S BoisseauGDespesse andBA Seddik ldquoNonlinear h-shapedsprings to improve efficiency of vibration energy harvestersrdquo
Journal of Applied Mechanics vol 80 no 6 Article ID 0610132013
[23] S Priya ldquoAdvances in energy harvesting using low profilepiezoelectric transducersrdquo Journal of Electroceramics vol 19 no1 pp 165ndash182 2007
[24] S Roundy P K Wright and J Rabaey ldquoA study of lowlevel vibrations as a power source for wireless sensor nodesrdquoComputer Communications vol 26 no 11 pp 1131ndash1144 2003
[25] S Roundy and P K Wright ldquoA piezoelectric vibration basedgenerator for wireless electronicsrdquo Smart Materials and Struc-tures vol 13 no 5 pp 1131ndash1142 2004
[26] H A Sodano G Park and D J Inman ldquoEstimation of electriccharge output for piezoelectric energy harvestingrdquo Strain vol40 no 2 pp 49ndash58 2004
[27] N E Dutoit B L Wardle and S-G Kim ldquoDesign considera-tions for mems-scale piezoelectric mechanical vibration energyharvestersrdquo Integrated Ferroelectrics vol 71 no 1 pp 121ndash1602005
[28] S-N Chen G-J Wang andM-C Chien ldquoAnalytical modelingof piezoelectric vibration-induced micro power generatorrdquoMechatronics vol 16 no 7 pp 379ndash387 2006
[29] A Erturk and D J Inman ldquoIssues in mathematical modelingof piezoelectric energy harvestersrdquo Smart Materials and Struc-tures vol 17 no 6 Article ID 065016 2008
[30] F Lu H P Lee and S P Lim ldquoModeling and analysis of micropiezoelectric power generators for micro-electromechanical-systems applicationsrdquo Smart Materials and Structures vol 13no 1 pp 57ndash63 2004
[31] S Jiang X Li S Guo Y Hu J Yang andQ Jiang ldquoPerformanceof a piezoelectric bimorph for scavenging vibration energyrdquoSmartMaterials and Structures vol 14 no 4 pp 769ndash774 2005
[32] A Erturk and D J Inman ldquoA distributed parameter elec-tromechanical model for cantilevered piezoelectric energy har-vestersrdquo Journal of Vibration and Acoustics vol 130 no 4Article ID 041002 2008
[33] A Erturk and D J Inman ldquoAn experimentally validatedbimorph cantilever model for piezoelectric energy harvestingfrom base excitationsrdquo Smart Materials and Structures vol 18no 2 Article ID 025009 2009
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
6 Shock and Vibration
The internal moment is given by the integral of the stressdistribution over the cross section of the beam shown in
119872(119909 119905) = minus119887 [int
minusℎ1199042
minusℎ119901minusℎ1199042
119879119901119910119889119910 + int
ℎ1199042
minusℎ1199042
119879119904119910119889119910
+ int
ℎ1199042+ℎ119901
ℎ1199042
119879119901119910119889119910]
(15)
where 119887 is the width of the beam ℎ119901is the thickness of
each PZT layer ℎ119904is the thickness of the substructure (see
Figure 4) and 119879 is the stress Superscripts 119901 and 119904 stand forPZT and substructure respectively The stress distributionacross the section is given by the strain-stress relations of thesubstructure and PZT layers which are given by
119879119904= 119884119904119878119904 (16)
119879119901= 119884119901119878119901minus 11988931119884119901119864 (17)
where 119878 is the strain 119884 is Youngrsquos modulus 119889 is thepiezoelectric constant and 119864 is the electric fieldThe bendingstrain can be expressed in terms of the radius of curvature as
119878 (119909 119910 119905) = minus1199101205972119908rel (119909 119905)
1205971199092 (18)
Since the piezoelectric layers are connected in parallelthey have the same voltage but the instantaneous electricfields are in opposite directions The electric field can bewritten in terms of the voltage across the PZT V(119905) and thethickness of the PZT layer ℎ
119901 by the following equation
1198641 (119905) = minus
V (119905)ℎ119901
in the top layer
1198642 (119905) =
V (119905)ℎ119901
in the bottom layer(19)
Replacing (16) to (19) in (15) we obtain
119872(119909 119905) =2119887
3[119884119904ℎ3
119904
12+ 119884119901((ℎ119901+ℎ119904
2)
3
minusℎ3
119904
8)]
sdot1205972119908rel (119909 119905)
1205971199092
+11988931119884119901119887
ℎ119901
[ℎ2
119904
4minus (ℎ119901+ℎ119904
2)
2
] V (119905)
(20)
Equation (20) can be expressed as
119872(119909 119905) = 1198841198681205972119908rel (119909 119905)
1205971199092
+ 120599V (119905) [Φ (119909) minus Φ (119909 minus 119871)] (21)
where 119871 is the length of the beam and Φ the Heavisidefunction It has been assumed that the PZT layer covers
the entire length of the beam The constant 119884119868 represents thebending stiffness of the beam and is given by
119884119868 =2119887
3[119884119904ℎ3
119904
12+ 119884119901((ℎ119901+ℎ119904
2)
3
minusℎ3
119904
8)] (22)
and the term 120599 represents the coupling between the strain andthe voltage and is given by
120599 =11988931119884119901119887
ℎ119901
[ℎ2
119904
4minus (ℎ119901+ℎ119904
2)
2
] (23)
Substituting (21) in (14) we obtain the mechanical equa-tion of motion with electrical coupling
1198841198681205974119908rel (119909 119905)
1205971199094+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
+ 120599V (119905) [120597120575 (119909)
120597119909minus120597120575 (119909 minus 119871)
120597119909]
= minus [119898 +119872119905120575 (119909 minus 119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(24)
32 Electrical Equation To derive the electrical equation letus start with the piezoelectric constitutive relation given in
119863 = 11988931119879119901+ 120576119879
33119864 (25)
where119863 is the electric displacement and 12057611987933is the permittivity
at constant stress Replacing the definition of 119879119901 from (17)yields
119863(119909 119905) = 11988931 (119884119901119878119901minus 11988931119864) + 120576
119879
33119864
= 11988931119884119901119878119901+ 119864 (120576
119879
33minus 1198892
31119884119901)
= 11988931119884119901119878119901minus 120576119878
33
V (119905)ℎ119901
(26)
with 12057611987833= 1198892
31119884119901minus 120576119879
33 The strain 119878119901 evaluated at the middle
of the PZT layer is given by
119878119901= minusℎ119901119888
1205972119908rel (119909 119905)
1205971199092 (27)
where ℎ119901119888= ℎ1199042 + ℎ
1199012 is the distance of the center of the
PZT layer to the neural axis Thus
119863 (119909 119905) = minus11988931119884119901ℎ119901119888
1205972119908rel (119909 119905)
1205971199092minus 120576119878
33
V (119905)ℎ119901
(28)
The electric charge is obtained by integrating the electricdisplacement over the area of the beamThe resulting expres-sion is
119902 (119905) = minusint
119871
0
(11988931119884119901ℎ119901119888
1205972119908rel (119909 119905)
1205971199092+ 120576119878
33
V (119905)ℎ119901
)119887119889119909 (29)
Shock and Vibration 7
Therefore the current generated by one PZT layer is
119894 (119905) =120597119902 (119905)
120597119905
= minus120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(30)
Since both layers are connected in parallel the currentgenerated by the bimorph beam is given by
119894 (119905) = 1198941 (119905) + 1198942 (119905)
= minus2120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 2int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(31)
If the beam is connected directly to a resistive load the voltageis given by V(119905) = 119877
119897119894(119905) Substituting this relation in (31) yields
V (119905) = minus2119877119897 [120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
+ int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909]
(32)
The electrical equation with mechanical coupling isobtained by rearranging (32)
V (119905)2119877119897
+120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
= minusint
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(33)
4 Voltage and Power Output of an Array ofPiezoelectric Beams Connected by Springs
Equations (24) and (33) are the electromechanical equationsfor a single bimorph beam under transverse vibrations Letus consider the case of multiple beams connected by springswith a parallel electrical connection as shown in Figure 5 Inthat case the mechanical equation of motion is given by
1198841198681205974119908rel (119909 119905)
1205971199094+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
+ 120599
119873
sum
119894=1
V119894 (119905) [
120597120575 (119909 minus (119894 minus 1) 119871)
120597119909minus120597120575 (119909 minus 119894119871)
120597119909]
= minus[119898 +
119873
sum
119894=1
119872119905119894120575 (119909 minus 119894119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(34)
(t)Rl
Figure 5 A parallel electrical connection of an array of bimorphbeams connected by springs
where V119894(119905) is the voltage of the 119894th beam Since the beams are
electrically connected in parallel they have the same voltagethus V
1(119905) = V
2(119905) = sdot sdot sdot = V
119899(119905) = V(119905) It has been assumed
that all the beams are identical except for the tip mass Itshould be noted that the springs are included as boundaryconditionsTherefore they affect the vibrationmodes but notthe equation of motion
The solution for the relative motion of the system can berepresented by a sum of its mode shapes as
119908rel (119909 119905) = sum119903
120601119903 (119909) 120578119903 (119905) (35)
where 120601119903(119909) and 120578
119903(119905) are the 119903th mode shape and modal
coordinate respectively The 119903th mode shape can be definedby segments as
120601119903 (119909)
=
1206011199031 (119909) 0 le 119909 le 119871
1206011199032 (119909 minus 119871) 119871 le 119909 le 2119871
120601119903119873 (119909 minus (119873 minus 1) 119871) (119873 minus 1) 119871 le 119909 le 119873119871
(36)
where 120601119903119894
is the 119903th mode shape of the 119894th beam 119871 is thelength of one beam and119873 is the number of beams
Replacing (35) themechanical equation ofmotion can bewritten as
[1205962
119903120578119903 (119905) + 2120585119903120596119903 120578119903 (119905) + 120578
119903 (119905)] 120572119903 +
119873
sum
119894=1
120594119903119894V (119905)
= [minus1198981205972119908119887 (119909 119905)
1205971199052minus 119888119886
120597119908119887 (119909 119905)
120597119905] 120574119903
minus1205972119908119887 (119909 119905)
1205971199052
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(37)
8 Shock and Vibration
with
120594119903119894= 120599(minus120601
119903119894 (119871) +120597120601119903119894 (119871)
120597119909)
120585119903=119888119904119868120596119903
2119884119868+
119888119886
2119898120596119903
120572119903= int
119873119871
0
1198981206012
119903(119909) 119889119909
120574119903= int
119873119871
0
120601119903 (119909) 119889119909
(38)
Assuming a harmonic basemotion119908119887(119909 119905) = 119884
0119890119895120596119905 and
a harmonic response 120578119903(119905) = 119873
119903119890119895120596119905 and V(119905) = 119881119890119895120596119905 (37)
becomes
[1205962
119899119903119873119903+ 2119895120585119903120596119899119903120596119873119903minus 1205962119873119903] 120572119903+ 119881
119873
sum
119894=1
120594119903119894
= (1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(39)
Rearranging (39) the amplitude of the temporal term119873119903
can be written as follows
119873119903
=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871) minus 119881sum
119873
119894=1120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(40)
For simplicity (40) can be expressed as
119873119903= 119886119903minus 119881
119873
sum
119894=1
119887119903119894 (41)
with
119886119903=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871)
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
119887119903119894=
120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(42)
The electrical equation needs to be derived for each beamindividually The relative motion for the 119894th beam is given by
119908rel119894 (119909 119905) = sum119903
120601119903119894 (119909) 120578119903 (119905) (43)
Replacing (43) in (30) yields
119894119894 (119905) = minus2
120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 211988931119884119901ℎ119901119888119887sum
119903
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(44)
If the beams are connected directly to a resistive load thevoltage is given by V(119905) = 119877
119897119894(119905) = 119877
119897sum119894119894119894(119905) substituting this
condition equation (44) becomes
V (119905)2119877119897
= minus120576119878
33119887119871119873
ℎ119901
120597V (119905)120597119905
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(45)
Replacing the harmonic response condition in (45) yields
119881
2119877119897
= minus119895120596120576119878
33119887119871119873
ℎ119901
119881
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
119895120596119873119903
(46)
Rearranging (46) we obtain
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596119873119903 (47)
with
120593119903119894= minus
11988931119884119901ℎ119901119888ℎ119901
120576119878
33119871119873
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120581119888= 119895120596 +
ℎ119901
2119877119897120576119878
33119887119871119873
(48)
Replacing 119873119903from (41) in (47) the following equation is
obtained
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596(119886
119903minus 119881
119873
sum
119896=1
119887119903119896) (49)
Rearranging (49) the voltage amplitude can be written asshown in
119881 =sum119903sum119894120593119903119894119895120596119886119903
120581119888+ sum119903sum119894120593119903119894119895120596sum119873
119896=1119887119903119896
(50)
The voltage FRF is defined as the ratio of the voltageoutput to the base acceleration119881minus1205962119884
0 The instantaneous
power output can be obtained from the relation 119875(119905) =V2(119905)119877
119897Therefore the FRF of the power output is simply the
square of the voltage FRF divided by 119877119897
5 Results
51 Analytical This section presents the results of the powerFRF for different number of beams and different combina-tions of tip masses and spring constants in the array Thegeometric material and electromechanical parameters of thebeams are given in Table 1 These are the parameters of theexperimental beams used in the next section
The voltage is computed by (50) and the power FRF iscomputed as 1198812[119877
119897(12059621198840)2] The value of the external load
Shock and Vibration 9
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(a) Connected
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 6 Power FRF with an increasing number of identical bimorph beams119872119905119894=16
= 03119898119871 and 119896119894=15
= 031198841198681198713
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(a) Connected
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 7 Power FRF with an increasing number of bimorph beams with different tip masses Two beams1198721199051= 03119898119871119872
1199052= 035119898119871
119896 = 031198841198681198713 three beams119872
1199051= 025119898119871119872
1199052= 030119898119871119872
1199053= 035119898119871 119896
12= 03119884119868119871
3 six beams1198721199051= 015119898119871119872
1199052= 020119898119871
1198721199053= 030119898119871119872
1199054= 035119898119871119872
1199055= 040119898119871119872
1199056= 045119898119871 119896
119894=15= 03119884119868119871
3
Table 1 Parameters of the bimorph beams
Length of the beam 119871 (mm) 482Width of the beam 119887 (mm) 10Thickness of the substructure ℎ
119904(mm) 019
Thickness of the PZT ℎ119901(mm) 024
Youngrsquos modulus of the substructure 119884119904(GPa) 404
Youngrsquos modulus of the PZT 119884119901(GPa) 55
Mass density of the substructure 120588119904(kgm3) 1519
Mass density of the PZT 120588119901(kgm3) 8140
Piezoelectric constant 11988931(pmV) minus180
Permittivity 12057611990433(nFm) 1328
Strain rate damping 119888119904119868 (kgm4s) 116 times 10minus6
Viscous air damping 119888119886(kgs) 04
resistance is defined for each case as the optimum resistorthat maximizes the power output To study the advantagesand disadvantages of the proposed system the results arecompared with those obtained by an array of disconnectedbeams (with no springs)
First the case of identical beams is investigated theresults are shown in Figure 6(a) It is obtained that in thecase of identical beams the only modes that are excitedare the modes where all beams move in phase such asmodes 1 4 and 7 in Figure 3 In this case it is also foundthat the natural frequencies do not depend on the springstiffness because all beams are in phase and the springs donot deform Furthermore the results are exactly the same asthe ones obtained with disconnected beams as illustrated inFigure 6(b)
In the case of beams with different tip masses the resultsof the FRF power do depend on the springs connecting thebeams as shown in Figure 7 If the beams are disconnectedand new beams with different tip masses are added to thearray then the frequency band of the device is increasedobtaining a similar power amplitude for all resonant frequen-cies On the other hand if the beams are connected by springand new beams are included in the array then the frequencyband is increased but the amplitude related to each resonantfrequency varies
For an array of piezoelectric beams connected by springsthe frequency band of operation and the power amplitude
10 Shock and Vibration
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
01mL 02mL01mL 03mL
01mL 04mL01mL 05mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k = 02YIL3k = 03YIL3
k = 04YIL3
(b) Varying spring constants (1198721199051 = 03119898119871 1198721199052 = 04119898119871)
Figure 8 Power FRF of two bimorph beams with different combinations of tip masses and spring constants
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
02mL 03mL 04mL02mL 02mL 04mL
03mL 03mL 04mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k =
k =k = 02YIL3 02YIL302YIL3 04YIL3
02YIL3 06YIL3
(b) Varying spring constants (1198721199051 = 02119898119871 1198721199052 = 03119898119871 1198721199053 =04119898119871)
Figure 9 Power FRF of three bimorph beams with different combinations of tip masses and spring constants
associated with each resonant frequency depend on thecombination of tipmasses and springs constantsThis ismoreevident in Figures 8 and 9 where different combinations ofsprings and tip masses are studied for the case of two andthree beams In general increasing the tip mass of the beamsdecreases someof the resonant frequencies and increasing thespring stiffness increases the frequency band
To compare the performances of an array of beamsconnected by springs and one of the disconnected beams acase of ten beams with both situations is plotted in Figure 10It is possible to see that the array with connected beamshas a slightly wider frequency band which increases withan increment of the springs constants Furthermore bycomparing the areas below both curves in the frequencyrange 50ndash200Hz the array with connected beams providesan area between 20 and 36 larger than the array withdisconnected beams Thus by connecting the beams theperformance of the array can be improved It should benoted that the previous results are obtained with arbitraryvalues for the tip masses and spring constantsTherefore thisperformance can be further improved by a proper design
52 Experimental To validate the proposed model a pre-liminary experimental campaign was performed Figure 11presents the experimental setup It consists of a signal genera-tor connected to a power amplifier and to an electrodynamicshaker which produces a controlled base vibration An arrayof three piezoelectric bimorph beams is mounted over theshaker Figure 12 shows this array The cantilever bimorphbeams have a tip mass at the free-end and can be connectedwith springs between them Three experiments were per-formed In the first case no springs were used and eachbeam was connected independently to a 1 kΩ resistive loadin this case the voltage generated by each beam is monitoredindividually In the second case no springs were used andthe beams were electrically connected in parallel to a resistiveload of 1 kΩ In the last case the beams were connected withsprings between them and they were electrically connected inparallel to a resistive load of 1 kΩ
The properties of the bimorph beams are the ones givenin Table 1 the tip masses are119872
1199051= 65119898119871119872
1199052= 35119898119871
1198721199053= 19119898119871 and the spring constants are 119896
1= 1198962=
2921198841198681198713
Shock and Vibration 11
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected02YIL3
(a) 119896119894=19 = 021198841198681198713
50 100 150 200
Frequency (Hz)
10minus4
Pow
er F
RF (W
s4m
2)
10minus3
10minus2
10minus1
100
Connected k =
Disconnected03YIL3
(b) 119896119894=19 = 031198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected04YIL3
(c) 119896119894=19 = 041198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected05YIL3
(d) 119896119894=19 = 051198841198681198713
Figure 10 Power FRF with ten bimorph beams1198721199051= 01119898119871119872
1199052= 015119898119871119872
1199053= 02119898119871119872
1199054= 025119898119871119872
1199055= 03119898119871119872
1199056= 035119898119871
1198721199057= 04119898119871119872
1199058= 045119898119871119872
1199059= 05119898119871119872
11990510= 055119898119871
Figure 11 Experimental setup
Figure 12 Experimental bimorph beams
During the three experiments the system is subjected toa harmonic base displacement with frequencies in the range20ndash70Hz The results for the first experiment are presentedin Figure 13(a) and are compared with the ones obtainedby an analytical model in Figure 13(b) Both analytical andexperimental curves are very similar The only difference isthat the experimental power increases from beams 1 to 3 andthe analytical model predicted the opposite behavior Thisunexpected behavior seems to be related to a resonance ofthe supporting structure in the three experimental curves itis possible to see an increment of the power near 60Hz whichcan be explained by a resonance of the structure
Figure 14 presents the experimental and analytical resultsfor the second and third experiments It is possible to seea good agreement between the experimental and analyticalcurves although similarly to the results in Figure 13(a) thereis an increment of the power near 60Hz that confirms astructural resonance
Comparing the areas below both experimental curves inFigure 13(b) in the frequency range 20ndash70Hz the array withconnected beams provides an area 5 larger than the arraywith disconnected beamswhich confirms that the addition ofsprings can increment the generated powerThese results canbe improved by using softer springs which have lower energylosses and by solving the structural resonance problem
12 Shock and Vibration
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 13 Power FRF for the three beams with no springs and independent electrical connections to 1 kΩ resistive loads (a) experimentalmeasurement (b) analytical results
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 14 Power FRF for the three beams with and without springs and electrical connection in parallel to a resistive load of 1 kΩ(a) experimental measurement (b) analytical results
6 Conclusions
This paper investigated the use of an array of bimorphbeams connected by springs as a broadband vibration-basedenergy harvester The equations for the voltage and poweroutput of the full system have been deduced based on theanalytical solution of the cantilevered piezoelectric energyharvester with Euler-Bernoulli beam assumptions To studythe advantages and disadvantages of the proposed system theresults are compared with those of an array of disconnectedbeams (with no springs) The analytical model is validatedwith experimental measurements of three bimorph beamswith and without springs
Both analytical and experimental results show that con-necting the array of beams with springs allows increasingthe frequency band of operation and increasing the amountof power generated Thus the maximal frequency band fora certain array of beams can be extended by connectingthem by springs and at the same time the output power isincreased The performance of this system can be optimizedfor specific applications by a proper selection of the tipmassesand spring constants
The preliminary experimental results are consistent withthe analytical ones but further research is necessary toimprove the experimental setup and to investigate differentexperimental conditions such as variations of the springsstiffness
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
References
[1] S R Anton and H A Sodano ldquoA review of power harvestingusing piezoelectric materials (2003ndash2006)rdquo Smart Materialsand Structures vol 16 no 3 article R1 2007
[2] S Chalasani and J M Conrad ldquoA survey of energy harvestingsources for embedded systemsrdquo in Proceedings of the IEEESoutheastCon pp 442ndash447 April 2008
[3] K A Cook-Chennault NThambi and AM Sastry ldquoPoweringMEMS portable devicesmdasha review of non-regenerative andregenerative power supply systems with special emphasis on
Shock and Vibration 13
piezoelectric energy harvesting systemsrdquo Smart Materials andStructures vol 17 no 4 Article ID 043001 2008
[4] A Dewan S U Ay M N Karim and H Beyenal ldquoAlternativepower sources for remote sensors a reviewrdquo Journal of PowerSources vol 245 pp 129ndash143 2014
[5] A Harb ldquoEnergy harvesting state-of-the-artrdquo RenewableEnergy vol 36 no 10 pp 2641ndash2654 2011
[6] S Saadon and O Sidek ldquoA review of vibration-based MEMSpiezoelectric energy harvestersrdquo Energy Conversion and Man-agement vol 52 no 1 pp 500ndash504 2011
[7] H A Sodano D J Inman and G Park ldquoA review of powerharvesting from vibration using piezoelectric materialsrdquo Shockand Vibration Digest vol 36 no 3 pp 197ndash205 2004
[8] R J M Vullers R van Schaijk I Doms C Van Hoofand R Mertens ldquoMicropower energy harvestingrdquo Solid-StateElectronics vol 53 no 7 pp 684ndash693 2009
[9] R Calio U B Rongala D Camboni et al ldquoPiezoelectric energyharvesting solutionsrdquo Sensors vol 14 no 3 pp 4755ndash4790 2014
[10] J Twiefel and H Westermann ldquoSurvey on broadband tech-niques for vibration energy harvestingrdquo Journal of IntelligentMaterial Systems and Structures vol 24 no 11 pp 1291ndash13022013
[11] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavengingrdquo Journal of Sound and Vibration vol 292no 3ndash5 pp 987ndash998 2006
[12] S M Shahruz ldquoLimits of performance of mechanical band-pass filters used in energy scavengingrdquo Journal of Sound andVibration vol 293 no 1-2 pp 449ndash461 2006
[13] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavenging multi-degree-of-freedom modelsrdquo Journalof Vibration and Control vol 14 no 5 pp 753ndash768 2008
[14] J-Q Liu H-B Fang Z-Y Xu et al ldquoA MEMS-based piezo-electric power generator array for vibration energy harvestingrdquoMicroelectronics Journal vol 39 no 5 pp 802ndash806 2008
[15] HC Lin PHWu I C Lien andYC Shu ldquoAnalysis of an arrayof piezo-electric energy harvesters connected in seriesrdquo SmartMaterials and Structures vol 22 no 9 Article ID 094026 2013
[16] S-J Jang E Rustighi M J Brennan Y P Lee and H-JJung ldquoDesign of a 2DOF vibrational energy harvesting devicerdquoJournal of IntelligentMaterial Systems and Structures vol 22 no5 pp 443ndash448 2011
[17] L Tang and Y Yang ldquoA multiple-degree-of-freedom piezoelec-tric energy harvesting modelrdquo Journal of Intelligent MaterialSystems and Structures vol 23 no 14 pp 1631ndash1647 2012
[18] Z Yang and J Yang ldquoConnected vibrating piezoelectricbimorph beams as a wide-band piezoelectric power harvesterrdquoJournal of IntelligentMaterial Systems and Structures vol 20 no5 pp 569ndash574 2009
[19] W Zhou G R Penamalli and L Zuo ldquoAn efficient vibrationenergy harvester with a multi-mode dynamic magnifierrdquo SmartMaterials and Structures vol 21 no 1 Article ID 015014 2012
[20] D S Nguyen E Halvorsen G U Jensen and A VoglldquoFabrication and characterization of a widebandMEMS energyharvester utilizing nonlinear springsrdquo Journal of Micromechan-ics andMicroengineering vol 20 no 12 Article ID 125009 2010
[21] S D Nguyen and E Halvorsen ldquoNonlinear springs forbandwidth-tolerant vibration energy harvestingrdquo Journal ofMicroelectromechanical Systems vol 20 no 6 pp 1225ndash12272011
[22] S BoisseauGDespesse andBA Seddik ldquoNonlinear h-shapedsprings to improve efficiency of vibration energy harvestersrdquo
Journal of Applied Mechanics vol 80 no 6 Article ID 0610132013
[23] S Priya ldquoAdvances in energy harvesting using low profilepiezoelectric transducersrdquo Journal of Electroceramics vol 19 no1 pp 165ndash182 2007
[24] S Roundy P K Wright and J Rabaey ldquoA study of lowlevel vibrations as a power source for wireless sensor nodesrdquoComputer Communications vol 26 no 11 pp 1131ndash1144 2003
[25] S Roundy and P K Wright ldquoA piezoelectric vibration basedgenerator for wireless electronicsrdquo Smart Materials and Struc-tures vol 13 no 5 pp 1131ndash1142 2004
[26] H A Sodano G Park and D J Inman ldquoEstimation of electriccharge output for piezoelectric energy harvestingrdquo Strain vol40 no 2 pp 49ndash58 2004
[27] N E Dutoit B L Wardle and S-G Kim ldquoDesign considera-tions for mems-scale piezoelectric mechanical vibration energyharvestersrdquo Integrated Ferroelectrics vol 71 no 1 pp 121ndash1602005
[28] S-N Chen G-J Wang andM-C Chien ldquoAnalytical modelingof piezoelectric vibration-induced micro power generatorrdquoMechatronics vol 16 no 7 pp 379ndash387 2006
[29] A Erturk and D J Inman ldquoIssues in mathematical modelingof piezoelectric energy harvestersrdquo Smart Materials and Struc-tures vol 17 no 6 Article ID 065016 2008
[30] F Lu H P Lee and S P Lim ldquoModeling and analysis of micropiezoelectric power generators for micro-electromechanical-systems applicationsrdquo Smart Materials and Structures vol 13no 1 pp 57ndash63 2004
[31] S Jiang X Li S Guo Y Hu J Yang andQ Jiang ldquoPerformanceof a piezoelectric bimorph for scavenging vibration energyrdquoSmartMaterials and Structures vol 14 no 4 pp 769ndash774 2005
[32] A Erturk and D J Inman ldquoA distributed parameter elec-tromechanical model for cantilevered piezoelectric energy har-vestersrdquo Journal of Vibration and Acoustics vol 130 no 4Article ID 041002 2008
[33] A Erturk and D J Inman ldquoAn experimentally validatedbimorph cantilever model for piezoelectric energy harvestingfrom base excitationsrdquo Smart Materials and Structures vol 18no 2 Article ID 025009 2009
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
Shock and Vibration 7
Therefore the current generated by one PZT layer is
119894 (119905) =120597119902 (119905)
120597119905
= minus120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(30)
Since both layers are connected in parallel the currentgenerated by the bimorph beam is given by
119894 (119905) = 1198941 (119905) + 1198942 (119905)
= minus2120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 2int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(31)
If the beam is connected directly to a resistive load the voltageis given by V(119905) = 119877
119897119894(119905) Substituting this relation in (31) yields
V (119905) = minus2119877119897 [120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
+ int
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909]
(32)
The electrical equation with mechanical coupling isobtained by rearranging (32)
V (119905)2119877119897
+120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
= minusint
119871
0
(11988931119884119901ℎ1199011198881198871205973119908rel (119909 119905)
1205971199092120597119905) 119889119909
(33)
4 Voltage and Power Output of an Array ofPiezoelectric Beams Connected by Springs
Equations (24) and (33) are the electromechanical equationsfor a single bimorph beam under transverse vibrations Letus consider the case of multiple beams connected by springswith a parallel electrical connection as shown in Figure 5 Inthat case the mechanical equation of motion is given by
1198841198681205974119908rel (119909 119905)
1205971199094+ 1198881199041198681205975119908rel (119909 119905)
1205971199094120597119905+ 119888119886
120597119908rel (119909 119905)
120597119905
+ 1198981205972119908rel (119909 119905)
1205971199052
+ 120599
119873
sum
119894=1
V119894 (119905) [
120597120575 (119909 minus (119894 minus 1) 119871)
120597119909minus120597120575 (119909 minus 119894119871)
120597119909]
= minus[119898 +
119873
sum
119894=1
119872119905119894120575 (119909 minus 119894119871)]
1205972119908119887 (119905)
1205971199052minus 119888119886
120597119908119887 (119905)
120597119905
(34)
(t)Rl
Figure 5 A parallel electrical connection of an array of bimorphbeams connected by springs
where V119894(119905) is the voltage of the 119894th beam Since the beams are
electrically connected in parallel they have the same voltagethus V
1(119905) = V
2(119905) = sdot sdot sdot = V
119899(119905) = V(119905) It has been assumed
that all the beams are identical except for the tip mass Itshould be noted that the springs are included as boundaryconditionsTherefore they affect the vibrationmodes but notthe equation of motion
The solution for the relative motion of the system can berepresented by a sum of its mode shapes as
119908rel (119909 119905) = sum119903
120601119903 (119909) 120578119903 (119905) (35)
where 120601119903(119909) and 120578
119903(119905) are the 119903th mode shape and modal
coordinate respectively The 119903th mode shape can be definedby segments as
120601119903 (119909)
=
1206011199031 (119909) 0 le 119909 le 119871
1206011199032 (119909 minus 119871) 119871 le 119909 le 2119871
120601119903119873 (119909 minus (119873 minus 1) 119871) (119873 minus 1) 119871 le 119909 le 119873119871
(36)
where 120601119903119894
is the 119903th mode shape of the 119894th beam 119871 is thelength of one beam and119873 is the number of beams
Replacing (35) themechanical equation ofmotion can bewritten as
[1205962
119903120578119903 (119905) + 2120585119903120596119903 120578119903 (119905) + 120578
119903 (119905)] 120572119903 +
119873
sum
119894=1
120594119903119894V (119905)
= [minus1198981205972119908119887 (119909 119905)
1205971199052minus 119888119886
120597119908119887 (119909 119905)
120597119905] 120574119903
minus1205972119908119887 (119909 119905)
1205971199052
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(37)
8 Shock and Vibration
with
120594119903119894= 120599(minus120601
119903119894 (119871) +120597120601119903119894 (119871)
120597119909)
120585119903=119888119904119868120596119903
2119884119868+
119888119886
2119898120596119903
120572119903= int
119873119871
0
1198981206012
119903(119909) 119889119909
120574119903= int
119873119871
0
120601119903 (119909) 119889119909
(38)
Assuming a harmonic basemotion119908119887(119909 119905) = 119884
0119890119895120596119905 and
a harmonic response 120578119903(119905) = 119873
119903119890119895120596119905 and V(119905) = 119881119890119895120596119905 (37)
becomes
[1205962
119899119903119873119903+ 2119895120585119903120596119899119903120596119873119903minus 1205962119873119903] 120572119903+ 119881
119873
sum
119894=1
120594119903119894
= (1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(39)
Rearranging (39) the amplitude of the temporal term119873119903
can be written as follows
119873119903
=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871) minus 119881sum
119873
119894=1120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(40)
For simplicity (40) can be expressed as
119873119903= 119886119903minus 119881
119873
sum
119894=1
119887119903119894 (41)
with
119886119903=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871)
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
119887119903119894=
120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(42)
The electrical equation needs to be derived for each beamindividually The relative motion for the 119894th beam is given by
119908rel119894 (119909 119905) = sum119903
120601119903119894 (119909) 120578119903 (119905) (43)
Replacing (43) in (30) yields
119894119894 (119905) = minus2
120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 211988931119884119901ℎ119901119888119887sum
119903
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(44)
If the beams are connected directly to a resistive load thevoltage is given by V(119905) = 119877
119897119894(119905) = 119877
119897sum119894119894119894(119905) substituting this
condition equation (44) becomes
V (119905)2119877119897
= minus120576119878
33119887119871119873
ℎ119901
120597V (119905)120597119905
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(45)
Replacing the harmonic response condition in (45) yields
119881
2119877119897
= minus119895120596120576119878
33119887119871119873
ℎ119901
119881
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
119895120596119873119903
(46)
Rearranging (46) we obtain
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596119873119903 (47)
with
120593119903119894= minus
11988931119884119901ℎ119901119888ℎ119901
120576119878
33119871119873
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120581119888= 119895120596 +
ℎ119901
2119877119897120576119878
33119887119871119873
(48)
Replacing 119873119903from (41) in (47) the following equation is
obtained
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596(119886
119903minus 119881
119873
sum
119896=1
119887119903119896) (49)
Rearranging (49) the voltage amplitude can be written asshown in
119881 =sum119903sum119894120593119903119894119895120596119886119903
120581119888+ sum119903sum119894120593119903119894119895120596sum119873
119896=1119887119903119896
(50)
The voltage FRF is defined as the ratio of the voltageoutput to the base acceleration119881minus1205962119884
0 The instantaneous
power output can be obtained from the relation 119875(119905) =V2(119905)119877
119897Therefore the FRF of the power output is simply the
square of the voltage FRF divided by 119877119897
5 Results
51 Analytical This section presents the results of the powerFRF for different number of beams and different combina-tions of tip masses and spring constants in the array Thegeometric material and electromechanical parameters of thebeams are given in Table 1 These are the parameters of theexperimental beams used in the next section
The voltage is computed by (50) and the power FRF iscomputed as 1198812[119877
119897(12059621198840)2] The value of the external load
Shock and Vibration 9
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(a) Connected
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 6 Power FRF with an increasing number of identical bimorph beams119872119905119894=16
= 03119898119871 and 119896119894=15
= 031198841198681198713
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(a) Connected
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 7 Power FRF with an increasing number of bimorph beams with different tip masses Two beams1198721199051= 03119898119871119872
1199052= 035119898119871
119896 = 031198841198681198713 three beams119872
1199051= 025119898119871119872
1199052= 030119898119871119872
1199053= 035119898119871 119896
12= 03119884119868119871
3 six beams1198721199051= 015119898119871119872
1199052= 020119898119871
1198721199053= 030119898119871119872
1199054= 035119898119871119872
1199055= 040119898119871119872
1199056= 045119898119871 119896
119894=15= 03119884119868119871
3
Table 1 Parameters of the bimorph beams
Length of the beam 119871 (mm) 482Width of the beam 119887 (mm) 10Thickness of the substructure ℎ
119904(mm) 019
Thickness of the PZT ℎ119901(mm) 024
Youngrsquos modulus of the substructure 119884119904(GPa) 404
Youngrsquos modulus of the PZT 119884119901(GPa) 55
Mass density of the substructure 120588119904(kgm3) 1519
Mass density of the PZT 120588119901(kgm3) 8140
Piezoelectric constant 11988931(pmV) minus180
Permittivity 12057611990433(nFm) 1328
Strain rate damping 119888119904119868 (kgm4s) 116 times 10minus6
Viscous air damping 119888119886(kgs) 04
resistance is defined for each case as the optimum resistorthat maximizes the power output To study the advantagesand disadvantages of the proposed system the results arecompared with those obtained by an array of disconnectedbeams (with no springs)
First the case of identical beams is investigated theresults are shown in Figure 6(a) It is obtained that in thecase of identical beams the only modes that are excitedare the modes where all beams move in phase such asmodes 1 4 and 7 in Figure 3 In this case it is also foundthat the natural frequencies do not depend on the springstiffness because all beams are in phase and the springs donot deform Furthermore the results are exactly the same asthe ones obtained with disconnected beams as illustrated inFigure 6(b)
In the case of beams with different tip masses the resultsof the FRF power do depend on the springs connecting thebeams as shown in Figure 7 If the beams are disconnectedand new beams with different tip masses are added to thearray then the frequency band of the device is increasedobtaining a similar power amplitude for all resonant frequen-cies On the other hand if the beams are connected by springand new beams are included in the array then the frequencyband is increased but the amplitude related to each resonantfrequency varies
For an array of piezoelectric beams connected by springsthe frequency band of operation and the power amplitude
10 Shock and Vibration
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
01mL 02mL01mL 03mL
01mL 04mL01mL 05mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k = 02YIL3k = 03YIL3
k = 04YIL3
(b) Varying spring constants (1198721199051 = 03119898119871 1198721199052 = 04119898119871)
Figure 8 Power FRF of two bimorph beams with different combinations of tip masses and spring constants
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
02mL 03mL 04mL02mL 02mL 04mL
03mL 03mL 04mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k =
k =k = 02YIL3 02YIL302YIL3 04YIL3
02YIL3 06YIL3
(b) Varying spring constants (1198721199051 = 02119898119871 1198721199052 = 03119898119871 1198721199053 =04119898119871)
Figure 9 Power FRF of three bimorph beams with different combinations of tip masses and spring constants
associated with each resonant frequency depend on thecombination of tipmasses and springs constantsThis ismoreevident in Figures 8 and 9 where different combinations ofsprings and tip masses are studied for the case of two andthree beams In general increasing the tip mass of the beamsdecreases someof the resonant frequencies and increasing thespring stiffness increases the frequency band
To compare the performances of an array of beamsconnected by springs and one of the disconnected beams acase of ten beams with both situations is plotted in Figure 10It is possible to see that the array with connected beamshas a slightly wider frequency band which increases withan increment of the springs constants Furthermore bycomparing the areas below both curves in the frequencyrange 50ndash200Hz the array with connected beams providesan area between 20 and 36 larger than the array withdisconnected beams Thus by connecting the beams theperformance of the array can be improved It should benoted that the previous results are obtained with arbitraryvalues for the tip masses and spring constantsTherefore thisperformance can be further improved by a proper design
52 Experimental To validate the proposed model a pre-liminary experimental campaign was performed Figure 11presents the experimental setup It consists of a signal genera-tor connected to a power amplifier and to an electrodynamicshaker which produces a controlled base vibration An arrayof three piezoelectric bimorph beams is mounted over theshaker Figure 12 shows this array The cantilever bimorphbeams have a tip mass at the free-end and can be connectedwith springs between them Three experiments were per-formed In the first case no springs were used and eachbeam was connected independently to a 1 kΩ resistive loadin this case the voltage generated by each beam is monitoredindividually In the second case no springs were used andthe beams were electrically connected in parallel to a resistiveload of 1 kΩ In the last case the beams were connected withsprings between them and they were electrically connected inparallel to a resistive load of 1 kΩ
The properties of the bimorph beams are the ones givenin Table 1 the tip masses are119872
1199051= 65119898119871119872
1199052= 35119898119871
1198721199053= 19119898119871 and the spring constants are 119896
1= 1198962=
2921198841198681198713
Shock and Vibration 11
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected02YIL3
(a) 119896119894=19 = 021198841198681198713
50 100 150 200
Frequency (Hz)
10minus4
Pow
er F
RF (W
s4m
2)
10minus3
10minus2
10minus1
100
Connected k =
Disconnected03YIL3
(b) 119896119894=19 = 031198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected04YIL3
(c) 119896119894=19 = 041198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected05YIL3
(d) 119896119894=19 = 051198841198681198713
Figure 10 Power FRF with ten bimorph beams1198721199051= 01119898119871119872
1199052= 015119898119871119872
1199053= 02119898119871119872
1199054= 025119898119871119872
1199055= 03119898119871119872
1199056= 035119898119871
1198721199057= 04119898119871119872
1199058= 045119898119871119872
1199059= 05119898119871119872
11990510= 055119898119871
Figure 11 Experimental setup
Figure 12 Experimental bimorph beams
During the three experiments the system is subjected toa harmonic base displacement with frequencies in the range20ndash70Hz The results for the first experiment are presentedin Figure 13(a) and are compared with the ones obtainedby an analytical model in Figure 13(b) Both analytical andexperimental curves are very similar The only difference isthat the experimental power increases from beams 1 to 3 andthe analytical model predicted the opposite behavior Thisunexpected behavior seems to be related to a resonance ofthe supporting structure in the three experimental curves itis possible to see an increment of the power near 60Hz whichcan be explained by a resonance of the structure
Figure 14 presents the experimental and analytical resultsfor the second and third experiments It is possible to seea good agreement between the experimental and analyticalcurves although similarly to the results in Figure 13(a) thereis an increment of the power near 60Hz that confirms astructural resonance
Comparing the areas below both experimental curves inFigure 13(b) in the frequency range 20ndash70Hz the array withconnected beams provides an area 5 larger than the arraywith disconnected beamswhich confirms that the addition ofsprings can increment the generated powerThese results canbe improved by using softer springs which have lower energylosses and by solving the structural resonance problem
12 Shock and Vibration
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 13 Power FRF for the three beams with no springs and independent electrical connections to 1 kΩ resistive loads (a) experimentalmeasurement (b) analytical results
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 14 Power FRF for the three beams with and without springs and electrical connection in parallel to a resistive load of 1 kΩ(a) experimental measurement (b) analytical results
6 Conclusions
This paper investigated the use of an array of bimorphbeams connected by springs as a broadband vibration-basedenergy harvester The equations for the voltage and poweroutput of the full system have been deduced based on theanalytical solution of the cantilevered piezoelectric energyharvester with Euler-Bernoulli beam assumptions To studythe advantages and disadvantages of the proposed system theresults are compared with those of an array of disconnectedbeams (with no springs) The analytical model is validatedwith experimental measurements of three bimorph beamswith and without springs
Both analytical and experimental results show that con-necting the array of beams with springs allows increasingthe frequency band of operation and increasing the amountof power generated Thus the maximal frequency band fora certain array of beams can be extended by connectingthem by springs and at the same time the output power isincreased The performance of this system can be optimizedfor specific applications by a proper selection of the tipmassesand spring constants
The preliminary experimental results are consistent withthe analytical ones but further research is necessary toimprove the experimental setup and to investigate differentexperimental conditions such as variations of the springsstiffness
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
References
[1] S R Anton and H A Sodano ldquoA review of power harvestingusing piezoelectric materials (2003ndash2006)rdquo Smart Materialsand Structures vol 16 no 3 article R1 2007
[2] S Chalasani and J M Conrad ldquoA survey of energy harvestingsources for embedded systemsrdquo in Proceedings of the IEEESoutheastCon pp 442ndash447 April 2008
[3] K A Cook-Chennault NThambi and AM Sastry ldquoPoweringMEMS portable devicesmdasha review of non-regenerative andregenerative power supply systems with special emphasis on
Shock and Vibration 13
piezoelectric energy harvesting systemsrdquo Smart Materials andStructures vol 17 no 4 Article ID 043001 2008
[4] A Dewan S U Ay M N Karim and H Beyenal ldquoAlternativepower sources for remote sensors a reviewrdquo Journal of PowerSources vol 245 pp 129ndash143 2014
[5] A Harb ldquoEnergy harvesting state-of-the-artrdquo RenewableEnergy vol 36 no 10 pp 2641ndash2654 2011
[6] S Saadon and O Sidek ldquoA review of vibration-based MEMSpiezoelectric energy harvestersrdquo Energy Conversion and Man-agement vol 52 no 1 pp 500ndash504 2011
[7] H A Sodano D J Inman and G Park ldquoA review of powerharvesting from vibration using piezoelectric materialsrdquo Shockand Vibration Digest vol 36 no 3 pp 197ndash205 2004
[8] R J M Vullers R van Schaijk I Doms C Van Hoofand R Mertens ldquoMicropower energy harvestingrdquo Solid-StateElectronics vol 53 no 7 pp 684ndash693 2009
[9] R Calio U B Rongala D Camboni et al ldquoPiezoelectric energyharvesting solutionsrdquo Sensors vol 14 no 3 pp 4755ndash4790 2014
[10] J Twiefel and H Westermann ldquoSurvey on broadband tech-niques for vibration energy harvestingrdquo Journal of IntelligentMaterial Systems and Structures vol 24 no 11 pp 1291ndash13022013
[11] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavengingrdquo Journal of Sound and Vibration vol 292no 3ndash5 pp 987ndash998 2006
[12] S M Shahruz ldquoLimits of performance of mechanical band-pass filters used in energy scavengingrdquo Journal of Sound andVibration vol 293 no 1-2 pp 449ndash461 2006
[13] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavenging multi-degree-of-freedom modelsrdquo Journalof Vibration and Control vol 14 no 5 pp 753ndash768 2008
[14] J-Q Liu H-B Fang Z-Y Xu et al ldquoA MEMS-based piezo-electric power generator array for vibration energy harvestingrdquoMicroelectronics Journal vol 39 no 5 pp 802ndash806 2008
[15] HC Lin PHWu I C Lien andYC Shu ldquoAnalysis of an arrayof piezo-electric energy harvesters connected in seriesrdquo SmartMaterials and Structures vol 22 no 9 Article ID 094026 2013
[16] S-J Jang E Rustighi M J Brennan Y P Lee and H-JJung ldquoDesign of a 2DOF vibrational energy harvesting devicerdquoJournal of IntelligentMaterial Systems and Structures vol 22 no5 pp 443ndash448 2011
[17] L Tang and Y Yang ldquoA multiple-degree-of-freedom piezoelec-tric energy harvesting modelrdquo Journal of Intelligent MaterialSystems and Structures vol 23 no 14 pp 1631ndash1647 2012
[18] Z Yang and J Yang ldquoConnected vibrating piezoelectricbimorph beams as a wide-band piezoelectric power harvesterrdquoJournal of IntelligentMaterial Systems and Structures vol 20 no5 pp 569ndash574 2009
[19] W Zhou G R Penamalli and L Zuo ldquoAn efficient vibrationenergy harvester with a multi-mode dynamic magnifierrdquo SmartMaterials and Structures vol 21 no 1 Article ID 015014 2012
[20] D S Nguyen E Halvorsen G U Jensen and A VoglldquoFabrication and characterization of a widebandMEMS energyharvester utilizing nonlinear springsrdquo Journal of Micromechan-ics andMicroengineering vol 20 no 12 Article ID 125009 2010
[21] S D Nguyen and E Halvorsen ldquoNonlinear springs forbandwidth-tolerant vibration energy harvestingrdquo Journal ofMicroelectromechanical Systems vol 20 no 6 pp 1225ndash12272011
[22] S BoisseauGDespesse andBA Seddik ldquoNonlinear h-shapedsprings to improve efficiency of vibration energy harvestersrdquo
Journal of Applied Mechanics vol 80 no 6 Article ID 0610132013
[23] S Priya ldquoAdvances in energy harvesting using low profilepiezoelectric transducersrdquo Journal of Electroceramics vol 19 no1 pp 165ndash182 2007
[24] S Roundy P K Wright and J Rabaey ldquoA study of lowlevel vibrations as a power source for wireless sensor nodesrdquoComputer Communications vol 26 no 11 pp 1131ndash1144 2003
[25] S Roundy and P K Wright ldquoA piezoelectric vibration basedgenerator for wireless electronicsrdquo Smart Materials and Struc-tures vol 13 no 5 pp 1131ndash1142 2004
[26] H A Sodano G Park and D J Inman ldquoEstimation of electriccharge output for piezoelectric energy harvestingrdquo Strain vol40 no 2 pp 49ndash58 2004
[27] N E Dutoit B L Wardle and S-G Kim ldquoDesign considera-tions for mems-scale piezoelectric mechanical vibration energyharvestersrdquo Integrated Ferroelectrics vol 71 no 1 pp 121ndash1602005
[28] S-N Chen G-J Wang andM-C Chien ldquoAnalytical modelingof piezoelectric vibration-induced micro power generatorrdquoMechatronics vol 16 no 7 pp 379ndash387 2006
[29] A Erturk and D J Inman ldquoIssues in mathematical modelingof piezoelectric energy harvestersrdquo Smart Materials and Struc-tures vol 17 no 6 Article ID 065016 2008
[30] F Lu H P Lee and S P Lim ldquoModeling and analysis of micropiezoelectric power generators for micro-electromechanical-systems applicationsrdquo Smart Materials and Structures vol 13no 1 pp 57ndash63 2004
[31] S Jiang X Li S Guo Y Hu J Yang andQ Jiang ldquoPerformanceof a piezoelectric bimorph for scavenging vibration energyrdquoSmartMaterials and Structures vol 14 no 4 pp 769ndash774 2005
[32] A Erturk and D J Inman ldquoA distributed parameter elec-tromechanical model for cantilevered piezoelectric energy har-vestersrdquo Journal of Vibration and Acoustics vol 130 no 4Article ID 041002 2008
[33] A Erturk and D J Inman ldquoAn experimentally validatedbimorph cantilever model for piezoelectric energy harvestingfrom base excitationsrdquo Smart Materials and Structures vol 18no 2 Article ID 025009 2009
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
8 Shock and Vibration
with
120594119903119894= 120599(minus120601
119903119894 (119871) +120597120601119903119894 (119871)
120597119909)
120585119903=119888119904119868120596119903
2119884119868+
119888119886
2119898120596119903
120572119903= int
119873119871
0
1198981206012
119903(119909) 119889119909
120574119903= int
119873119871
0
120601119903 (119909) 119889119909
(38)
Assuming a harmonic basemotion119908119887(119909 119905) = 119884
0119890119895120596119905 and
a harmonic response 120578119903(119905) = 119873
119903119890119895120596119905 and V(119905) = 119881119890119895120596119905 (37)
becomes
[1205962
119899119903119873119903+ 2119895120585119903120596119899119903120596119873119903minus 1205962119873119903] 120572119903+ 119881
119873
sum
119894=1
120594119903119894
= (1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840
119873
sum
119894=1
119872119905119894120601119903 (119894119871)
(39)
Rearranging (39) the amplitude of the temporal term119873119903
can be written as follows
119873119903
=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871) minus 119881sum
119873
119894=1120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(40)
For simplicity (40) can be expressed as
119873119903= 119886119903minus 119881
119873
sum
119894=1
119887119903119894 (41)
with
119886119903=
(1198981205962minus 119895120596119888119886) 1198840120574119903+ 12059621198840sum119873
119894=1119872119905119894120601119903 (119894119871)
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
119887119903119894=
120594119903119894
(1205962119903+ 2119895120585119903120596119903120596 minus 1205962) 120572
119903
(42)
The electrical equation needs to be derived for each beamindividually The relative motion for the 119894th beam is given by
119908rel119894 (119909 119905) = sum119903
120601119903119894 (119909) 120578119903 (119905) (43)
Replacing (43) in (30) yields
119894119894 (119905) = minus2
120576119878
33119887119871
ℎ119901
120597V (119905)120597119905
minus 211988931119884119901ℎ119901119888119887sum
119903
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(44)
If the beams are connected directly to a resistive load thevoltage is given by V(119905) = 119877
119897119894(119905) = 119877
119897sum119894119894119894(119905) substituting this
condition equation (44) becomes
V (119905)2119877119897
= minus120576119878
33119887119871119873
ℎ119901
120597V (119905)120597119905
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120578119903 (119905)
(45)
Replacing the harmonic response condition in (45) yields
119881
2119877119897
= minus119895120596120576119878
33119887119871119873
ℎ119901
119881
minus 11988931119884119901ℎ119901119888119887sum
119903
sum
119894
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
119895120596119873119903
(46)
Rearranging (46) we obtain
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596119873119903 (47)
with
120593119903119894= minus
11988931119884119901ℎ119901119888ℎ119901
120576119878
33119871119873
120597120601119903119894 (119909)
120597119909
10038161003816100381610038161003816100381610038161003816
119871
0
120581119888= 119895120596 +
ℎ119901
2119877119897120576119878
33119887119871119873
(48)
Replacing 119873119903from (41) in (47) the following equation is
obtained
119881120581119888= sum
119903
sum
119894
120593119903119894119895120596(119886
119903minus 119881
119873
sum
119896=1
119887119903119896) (49)
Rearranging (49) the voltage amplitude can be written asshown in
119881 =sum119903sum119894120593119903119894119895120596119886119903
120581119888+ sum119903sum119894120593119903119894119895120596sum119873
119896=1119887119903119896
(50)
The voltage FRF is defined as the ratio of the voltageoutput to the base acceleration119881minus1205962119884
0 The instantaneous
power output can be obtained from the relation 119875(119905) =V2(119905)119877
119897Therefore the FRF of the power output is simply the
square of the voltage FRF divided by 119877119897
5 Results
51 Analytical This section presents the results of the powerFRF for different number of beams and different combina-tions of tip masses and spring constants in the array Thegeometric material and electromechanical parameters of thebeams are given in Table 1 These are the parameters of theexperimental beams used in the next section
The voltage is computed by (50) and the power FRF iscomputed as 1198812[119877
119897(12059621198840)2] The value of the external load
Shock and Vibration 9
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(a) Connected
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 6 Power FRF with an increasing number of identical bimorph beams119872119905119894=16
= 03119898119871 and 119896119894=15
= 031198841198681198713
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(a) Connected
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 7 Power FRF with an increasing number of bimorph beams with different tip masses Two beams1198721199051= 03119898119871119872
1199052= 035119898119871
119896 = 031198841198681198713 three beams119872
1199051= 025119898119871119872
1199052= 030119898119871119872
1199053= 035119898119871 119896
12= 03119884119868119871
3 six beams1198721199051= 015119898119871119872
1199052= 020119898119871
1198721199053= 030119898119871119872
1199054= 035119898119871119872
1199055= 040119898119871119872
1199056= 045119898119871 119896
119894=15= 03119884119868119871
3
Table 1 Parameters of the bimorph beams
Length of the beam 119871 (mm) 482Width of the beam 119887 (mm) 10Thickness of the substructure ℎ
119904(mm) 019
Thickness of the PZT ℎ119901(mm) 024
Youngrsquos modulus of the substructure 119884119904(GPa) 404
Youngrsquos modulus of the PZT 119884119901(GPa) 55
Mass density of the substructure 120588119904(kgm3) 1519
Mass density of the PZT 120588119901(kgm3) 8140
Piezoelectric constant 11988931(pmV) minus180
Permittivity 12057611990433(nFm) 1328
Strain rate damping 119888119904119868 (kgm4s) 116 times 10minus6
Viscous air damping 119888119886(kgs) 04
resistance is defined for each case as the optimum resistorthat maximizes the power output To study the advantagesand disadvantages of the proposed system the results arecompared with those obtained by an array of disconnectedbeams (with no springs)
First the case of identical beams is investigated theresults are shown in Figure 6(a) It is obtained that in thecase of identical beams the only modes that are excitedare the modes where all beams move in phase such asmodes 1 4 and 7 in Figure 3 In this case it is also foundthat the natural frequencies do not depend on the springstiffness because all beams are in phase and the springs donot deform Furthermore the results are exactly the same asthe ones obtained with disconnected beams as illustrated inFigure 6(b)
In the case of beams with different tip masses the resultsof the FRF power do depend on the springs connecting thebeams as shown in Figure 7 If the beams are disconnectedand new beams with different tip masses are added to thearray then the frequency band of the device is increasedobtaining a similar power amplitude for all resonant frequen-cies On the other hand if the beams are connected by springand new beams are included in the array then the frequencyband is increased but the amplitude related to each resonantfrequency varies
For an array of piezoelectric beams connected by springsthe frequency band of operation and the power amplitude
10 Shock and Vibration
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
01mL 02mL01mL 03mL
01mL 04mL01mL 05mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k = 02YIL3k = 03YIL3
k = 04YIL3
(b) Varying spring constants (1198721199051 = 03119898119871 1198721199052 = 04119898119871)
Figure 8 Power FRF of two bimorph beams with different combinations of tip masses and spring constants
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
02mL 03mL 04mL02mL 02mL 04mL
03mL 03mL 04mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k =
k =k = 02YIL3 02YIL302YIL3 04YIL3
02YIL3 06YIL3
(b) Varying spring constants (1198721199051 = 02119898119871 1198721199052 = 03119898119871 1198721199053 =04119898119871)
Figure 9 Power FRF of three bimorph beams with different combinations of tip masses and spring constants
associated with each resonant frequency depend on thecombination of tipmasses and springs constantsThis ismoreevident in Figures 8 and 9 where different combinations ofsprings and tip masses are studied for the case of two andthree beams In general increasing the tip mass of the beamsdecreases someof the resonant frequencies and increasing thespring stiffness increases the frequency band
To compare the performances of an array of beamsconnected by springs and one of the disconnected beams acase of ten beams with both situations is plotted in Figure 10It is possible to see that the array with connected beamshas a slightly wider frequency band which increases withan increment of the springs constants Furthermore bycomparing the areas below both curves in the frequencyrange 50ndash200Hz the array with connected beams providesan area between 20 and 36 larger than the array withdisconnected beams Thus by connecting the beams theperformance of the array can be improved It should benoted that the previous results are obtained with arbitraryvalues for the tip masses and spring constantsTherefore thisperformance can be further improved by a proper design
52 Experimental To validate the proposed model a pre-liminary experimental campaign was performed Figure 11presents the experimental setup It consists of a signal genera-tor connected to a power amplifier and to an electrodynamicshaker which produces a controlled base vibration An arrayof three piezoelectric bimorph beams is mounted over theshaker Figure 12 shows this array The cantilever bimorphbeams have a tip mass at the free-end and can be connectedwith springs between them Three experiments were per-formed In the first case no springs were used and eachbeam was connected independently to a 1 kΩ resistive loadin this case the voltage generated by each beam is monitoredindividually In the second case no springs were used andthe beams were electrically connected in parallel to a resistiveload of 1 kΩ In the last case the beams were connected withsprings between them and they were electrically connected inparallel to a resistive load of 1 kΩ
The properties of the bimorph beams are the ones givenin Table 1 the tip masses are119872
1199051= 65119898119871119872
1199052= 35119898119871
1198721199053= 19119898119871 and the spring constants are 119896
1= 1198962=
2921198841198681198713
Shock and Vibration 11
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected02YIL3
(a) 119896119894=19 = 021198841198681198713
50 100 150 200
Frequency (Hz)
10minus4
Pow
er F
RF (W
s4m
2)
10minus3
10minus2
10minus1
100
Connected k =
Disconnected03YIL3
(b) 119896119894=19 = 031198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected04YIL3
(c) 119896119894=19 = 041198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected05YIL3
(d) 119896119894=19 = 051198841198681198713
Figure 10 Power FRF with ten bimorph beams1198721199051= 01119898119871119872
1199052= 015119898119871119872
1199053= 02119898119871119872
1199054= 025119898119871119872
1199055= 03119898119871119872
1199056= 035119898119871
1198721199057= 04119898119871119872
1199058= 045119898119871119872
1199059= 05119898119871119872
11990510= 055119898119871
Figure 11 Experimental setup
Figure 12 Experimental bimorph beams
During the three experiments the system is subjected toa harmonic base displacement with frequencies in the range20ndash70Hz The results for the first experiment are presentedin Figure 13(a) and are compared with the ones obtainedby an analytical model in Figure 13(b) Both analytical andexperimental curves are very similar The only difference isthat the experimental power increases from beams 1 to 3 andthe analytical model predicted the opposite behavior Thisunexpected behavior seems to be related to a resonance ofthe supporting structure in the three experimental curves itis possible to see an increment of the power near 60Hz whichcan be explained by a resonance of the structure
Figure 14 presents the experimental and analytical resultsfor the second and third experiments It is possible to seea good agreement between the experimental and analyticalcurves although similarly to the results in Figure 13(a) thereis an increment of the power near 60Hz that confirms astructural resonance
Comparing the areas below both experimental curves inFigure 13(b) in the frequency range 20ndash70Hz the array withconnected beams provides an area 5 larger than the arraywith disconnected beamswhich confirms that the addition ofsprings can increment the generated powerThese results canbe improved by using softer springs which have lower energylosses and by solving the structural resonance problem
12 Shock and Vibration
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 13 Power FRF for the three beams with no springs and independent electrical connections to 1 kΩ resistive loads (a) experimentalmeasurement (b) analytical results
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 14 Power FRF for the three beams with and without springs and electrical connection in parallel to a resistive load of 1 kΩ(a) experimental measurement (b) analytical results
6 Conclusions
This paper investigated the use of an array of bimorphbeams connected by springs as a broadband vibration-basedenergy harvester The equations for the voltage and poweroutput of the full system have been deduced based on theanalytical solution of the cantilevered piezoelectric energyharvester with Euler-Bernoulli beam assumptions To studythe advantages and disadvantages of the proposed system theresults are compared with those of an array of disconnectedbeams (with no springs) The analytical model is validatedwith experimental measurements of three bimorph beamswith and without springs
Both analytical and experimental results show that con-necting the array of beams with springs allows increasingthe frequency band of operation and increasing the amountof power generated Thus the maximal frequency band fora certain array of beams can be extended by connectingthem by springs and at the same time the output power isincreased The performance of this system can be optimizedfor specific applications by a proper selection of the tipmassesand spring constants
The preliminary experimental results are consistent withthe analytical ones but further research is necessary toimprove the experimental setup and to investigate differentexperimental conditions such as variations of the springsstiffness
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
References
[1] S R Anton and H A Sodano ldquoA review of power harvestingusing piezoelectric materials (2003ndash2006)rdquo Smart Materialsand Structures vol 16 no 3 article R1 2007
[2] S Chalasani and J M Conrad ldquoA survey of energy harvestingsources for embedded systemsrdquo in Proceedings of the IEEESoutheastCon pp 442ndash447 April 2008
[3] K A Cook-Chennault NThambi and AM Sastry ldquoPoweringMEMS portable devicesmdasha review of non-regenerative andregenerative power supply systems with special emphasis on
Shock and Vibration 13
piezoelectric energy harvesting systemsrdquo Smart Materials andStructures vol 17 no 4 Article ID 043001 2008
[4] A Dewan S U Ay M N Karim and H Beyenal ldquoAlternativepower sources for remote sensors a reviewrdquo Journal of PowerSources vol 245 pp 129ndash143 2014
[5] A Harb ldquoEnergy harvesting state-of-the-artrdquo RenewableEnergy vol 36 no 10 pp 2641ndash2654 2011
[6] S Saadon and O Sidek ldquoA review of vibration-based MEMSpiezoelectric energy harvestersrdquo Energy Conversion and Man-agement vol 52 no 1 pp 500ndash504 2011
[7] H A Sodano D J Inman and G Park ldquoA review of powerharvesting from vibration using piezoelectric materialsrdquo Shockand Vibration Digest vol 36 no 3 pp 197ndash205 2004
[8] R J M Vullers R van Schaijk I Doms C Van Hoofand R Mertens ldquoMicropower energy harvestingrdquo Solid-StateElectronics vol 53 no 7 pp 684ndash693 2009
[9] R Calio U B Rongala D Camboni et al ldquoPiezoelectric energyharvesting solutionsrdquo Sensors vol 14 no 3 pp 4755ndash4790 2014
[10] J Twiefel and H Westermann ldquoSurvey on broadband tech-niques for vibration energy harvestingrdquo Journal of IntelligentMaterial Systems and Structures vol 24 no 11 pp 1291ndash13022013
[11] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavengingrdquo Journal of Sound and Vibration vol 292no 3ndash5 pp 987ndash998 2006
[12] S M Shahruz ldquoLimits of performance of mechanical band-pass filters used in energy scavengingrdquo Journal of Sound andVibration vol 293 no 1-2 pp 449ndash461 2006
[13] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavenging multi-degree-of-freedom modelsrdquo Journalof Vibration and Control vol 14 no 5 pp 753ndash768 2008
[14] J-Q Liu H-B Fang Z-Y Xu et al ldquoA MEMS-based piezo-electric power generator array for vibration energy harvestingrdquoMicroelectronics Journal vol 39 no 5 pp 802ndash806 2008
[15] HC Lin PHWu I C Lien andYC Shu ldquoAnalysis of an arrayof piezo-electric energy harvesters connected in seriesrdquo SmartMaterials and Structures vol 22 no 9 Article ID 094026 2013
[16] S-J Jang E Rustighi M J Brennan Y P Lee and H-JJung ldquoDesign of a 2DOF vibrational energy harvesting devicerdquoJournal of IntelligentMaterial Systems and Structures vol 22 no5 pp 443ndash448 2011
[17] L Tang and Y Yang ldquoA multiple-degree-of-freedom piezoelec-tric energy harvesting modelrdquo Journal of Intelligent MaterialSystems and Structures vol 23 no 14 pp 1631ndash1647 2012
[18] Z Yang and J Yang ldquoConnected vibrating piezoelectricbimorph beams as a wide-band piezoelectric power harvesterrdquoJournal of IntelligentMaterial Systems and Structures vol 20 no5 pp 569ndash574 2009
[19] W Zhou G R Penamalli and L Zuo ldquoAn efficient vibrationenergy harvester with a multi-mode dynamic magnifierrdquo SmartMaterials and Structures vol 21 no 1 Article ID 015014 2012
[20] D S Nguyen E Halvorsen G U Jensen and A VoglldquoFabrication and characterization of a widebandMEMS energyharvester utilizing nonlinear springsrdquo Journal of Micromechan-ics andMicroengineering vol 20 no 12 Article ID 125009 2010
[21] S D Nguyen and E Halvorsen ldquoNonlinear springs forbandwidth-tolerant vibration energy harvestingrdquo Journal ofMicroelectromechanical Systems vol 20 no 6 pp 1225ndash12272011
[22] S BoisseauGDespesse andBA Seddik ldquoNonlinear h-shapedsprings to improve efficiency of vibration energy harvestersrdquo
Journal of Applied Mechanics vol 80 no 6 Article ID 0610132013
[23] S Priya ldquoAdvances in energy harvesting using low profilepiezoelectric transducersrdquo Journal of Electroceramics vol 19 no1 pp 165ndash182 2007
[24] S Roundy P K Wright and J Rabaey ldquoA study of lowlevel vibrations as a power source for wireless sensor nodesrdquoComputer Communications vol 26 no 11 pp 1131ndash1144 2003
[25] S Roundy and P K Wright ldquoA piezoelectric vibration basedgenerator for wireless electronicsrdquo Smart Materials and Struc-tures vol 13 no 5 pp 1131ndash1142 2004
[26] H A Sodano G Park and D J Inman ldquoEstimation of electriccharge output for piezoelectric energy harvestingrdquo Strain vol40 no 2 pp 49ndash58 2004
[27] N E Dutoit B L Wardle and S-G Kim ldquoDesign considera-tions for mems-scale piezoelectric mechanical vibration energyharvestersrdquo Integrated Ferroelectrics vol 71 no 1 pp 121ndash1602005
[28] S-N Chen G-J Wang andM-C Chien ldquoAnalytical modelingof piezoelectric vibration-induced micro power generatorrdquoMechatronics vol 16 no 7 pp 379ndash387 2006
[29] A Erturk and D J Inman ldquoIssues in mathematical modelingof piezoelectric energy harvestersrdquo Smart Materials and Struc-tures vol 17 no 6 Article ID 065016 2008
[30] F Lu H P Lee and S P Lim ldquoModeling and analysis of micropiezoelectric power generators for micro-electromechanical-systems applicationsrdquo Smart Materials and Structures vol 13no 1 pp 57ndash63 2004
[31] S Jiang X Li S Guo Y Hu J Yang andQ Jiang ldquoPerformanceof a piezoelectric bimorph for scavenging vibration energyrdquoSmartMaterials and Structures vol 14 no 4 pp 769ndash774 2005
[32] A Erturk and D J Inman ldquoA distributed parameter elec-tromechanical model for cantilevered piezoelectric energy har-vestersrdquo Journal of Vibration and Acoustics vol 130 no 4Article ID 041002 2008
[33] A Erturk and D J Inman ldquoAn experimentally validatedbimorph cantilever model for piezoelectric energy harvestingfrom base excitationsrdquo Smart Materials and Structures vol 18no 2 Article ID 025009 2009
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
Shock and Vibration 9
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(a) Connected
One beamTwo beams
Four beamsSix beams
50 100 150 200
Frequency (Hz)
10minus2
100
10minus4
10minus6
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 6 Power FRF with an increasing number of identical bimorph beams119872119905119894=16
= 03119898119871 and 119896119894=15
= 031198841198681198713
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(a) Connected
Two beamsFour beams
Six beams
50 100 150 200
Frequency (Hz)
10minus4
10minus3
10minus2
10minus1
100
Pow
er F
RF (W
s4m
2)
(b) Disconnected
Figure 7 Power FRF with an increasing number of bimorph beams with different tip masses Two beams1198721199051= 03119898119871119872
1199052= 035119898119871
119896 = 031198841198681198713 three beams119872
1199051= 025119898119871119872
1199052= 030119898119871119872
1199053= 035119898119871 119896
12= 03119884119868119871
3 six beams1198721199051= 015119898119871119872
1199052= 020119898119871
1198721199053= 030119898119871119872
1199054= 035119898119871119872
1199055= 040119898119871119872
1199056= 045119898119871 119896
119894=15= 03119884119868119871
3
Table 1 Parameters of the bimorph beams
Length of the beam 119871 (mm) 482Width of the beam 119887 (mm) 10Thickness of the substructure ℎ
119904(mm) 019
Thickness of the PZT ℎ119901(mm) 024
Youngrsquos modulus of the substructure 119884119904(GPa) 404
Youngrsquos modulus of the PZT 119884119901(GPa) 55
Mass density of the substructure 120588119904(kgm3) 1519
Mass density of the PZT 120588119901(kgm3) 8140
Piezoelectric constant 11988931(pmV) minus180
Permittivity 12057611990433(nFm) 1328
Strain rate damping 119888119904119868 (kgm4s) 116 times 10minus6
Viscous air damping 119888119886(kgs) 04
resistance is defined for each case as the optimum resistorthat maximizes the power output To study the advantagesand disadvantages of the proposed system the results arecompared with those obtained by an array of disconnectedbeams (with no springs)
First the case of identical beams is investigated theresults are shown in Figure 6(a) It is obtained that in thecase of identical beams the only modes that are excitedare the modes where all beams move in phase such asmodes 1 4 and 7 in Figure 3 In this case it is also foundthat the natural frequencies do not depend on the springstiffness because all beams are in phase and the springs donot deform Furthermore the results are exactly the same asthe ones obtained with disconnected beams as illustrated inFigure 6(b)
In the case of beams with different tip masses the resultsof the FRF power do depend on the springs connecting thebeams as shown in Figure 7 If the beams are disconnectedand new beams with different tip masses are added to thearray then the frequency band of the device is increasedobtaining a similar power amplitude for all resonant frequen-cies On the other hand if the beams are connected by springand new beams are included in the array then the frequencyband is increased but the amplitude related to each resonantfrequency varies
For an array of piezoelectric beams connected by springsthe frequency band of operation and the power amplitude
10 Shock and Vibration
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
01mL 02mL01mL 03mL
01mL 04mL01mL 05mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k = 02YIL3k = 03YIL3
k = 04YIL3
(b) Varying spring constants (1198721199051 = 03119898119871 1198721199052 = 04119898119871)
Figure 8 Power FRF of two bimorph beams with different combinations of tip masses and spring constants
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
02mL 03mL 04mL02mL 02mL 04mL
03mL 03mL 04mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k =
k =k = 02YIL3 02YIL302YIL3 04YIL3
02YIL3 06YIL3
(b) Varying spring constants (1198721199051 = 02119898119871 1198721199052 = 03119898119871 1198721199053 =04119898119871)
Figure 9 Power FRF of three bimorph beams with different combinations of tip masses and spring constants
associated with each resonant frequency depend on thecombination of tipmasses and springs constantsThis ismoreevident in Figures 8 and 9 where different combinations ofsprings and tip masses are studied for the case of two andthree beams In general increasing the tip mass of the beamsdecreases someof the resonant frequencies and increasing thespring stiffness increases the frequency band
To compare the performances of an array of beamsconnected by springs and one of the disconnected beams acase of ten beams with both situations is plotted in Figure 10It is possible to see that the array with connected beamshas a slightly wider frequency band which increases withan increment of the springs constants Furthermore bycomparing the areas below both curves in the frequencyrange 50ndash200Hz the array with connected beams providesan area between 20 and 36 larger than the array withdisconnected beams Thus by connecting the beams theperformance of the array can be improved It should benoted that the previous results are obtained with arbitraryvalues for the tip masses and spring constantsTherefore thisperformance can be further improved by a proper design
52 Experimental To validate the proposed model a pre-liminary experimental campaign was performed Figure 11presents the experimental setup It consists of a signal genera-tor connected to a power amplifier and to an electrodynamicshaker which produces a controlled base vibration An arrayof three piezoelectric bimorph beams is mounted over theshaker Figure 12 shows this array The cantilever bimorphbeams have a tip mass at the free-end and can be connectedwith springs between them Three experiments were per-formed In the first case no springs were used and eachbeam was connected independently to a 1 kΩ resistive loadin this case the voltage generated by each beam is monitoredindividually In the second case no springs were used andthe beams were electrically connected in parallel to a resistiveload of 1 kΩ In the last case the beams were connected withsprings between them and they were electrically connected inparallel to a resistive load of 1 kΩ
The properties of the bimorph beams are the ones givenin Table 1 the tip masses are119872
1199051= 65119898119871119872
1199052= 35119898119871
1198721199053= 19119898119871 and the spring constants are 119896
1= 1198962=
2921198841198681198713
Shock and Vibration 11
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected02YIL3
(a) 119896119894=19 = 021198841198681198713
50 100 150 200
Frequency (Hz)
10minus4
Pow
er F
RF (W
s4m
2)
10minus3
10minus2
10minus1
100
Connected k =
Disconnected03YIL3
(b) 119896119894=19 = 031198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected04YIL3
(c) 119896119894=19 = 041198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected05YIL3
(d) 119896119894=19 = 051198841198681198713
Figure 10 Power FRF with ten bimorph beams1198721199051= 01119898119871119872
1199052= 015119898119871119872
1199053= 02119898119871119872
1199054= 025119898119871119872
1199055= 03119898119871119872
1199056= 035119898119871
1198721199057= 04119898119871119872
1199058= 045119898119871119872
1199059= 05119898119871119872
11990510= 055119898119871
Figure 11 Experimental setup
Figure 12 Experimental bimorph beams
During the three experiments the system is subjected toa harmonic base displacement with frequencies in the range20ndash70Hz The results for the first experiment are presentedin Figure 13(a) and are compared with the ones obtainedby an analytical model in Figure 13(b) Both analytical andexperimental curves are very similar The only difference isthat the experimental power increases from beams 1 to 3 andthe analytical model predicted the opposite behavior Thisunexpected behavior seems to be related to a resonance ofthe supporting structure in the three experimental curves itis possible to see an increment of the power near 60Hz whichcan be explained by a resonance of the structure
Figure 14 presents the experimental and analytical resultsfor the second and third experiments It is possible to seea good agreement between the experimental and analyticalcurves although similarly to the results in Figure 13(a) thereis an increment of the power near 60Hz that confirms astructural resonance
Comparing the areas below both experimental curves inFigure 13(b) in the frequency range 20ndash70Hz the array withconnected beams provides an area 5 larger than the arraywith disconnected beamswhich confirms that the addition ofsprings can increment the generated powerThese results canbe improved by using softer springs which have lower energylosses and by solving the structural resonance problem
12 Shock and Vibration
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 13 Power FRF for the three beams with no springs and independent electrical connections to 1 kΩ resistive loads (a) experimentalmeasurement (b) analytical results
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 14 Power FRF for the three beams with and without springs and electrical connection in parallel to a resistive load of 1 kΩ(a) experimental measurement (b) analytical results
6 Conclusions
This paper investigated the use of an array of bimorphbeams connected by springs as a broadband vibration-basedenergy harvester The equations for the voltage and poweroutput of the full system have been deduced based on theanalytical solution of the cantilevered piezoelectric energyharvester with Euler-Bernoulli beam assumptions To studythe advantages and disadvantages of the proposed system theresults are compared with those of an array of disconnectedbeams (with no springs) The analytical model is validatedwith experimental measurements of three bimorph beamswith and without springs
Both analytical and experimental results show that con-necting the array of beams with springs allows increasingthe frequency band of operation and increasing the amountof power generated Thus the maximal frequency band fora certain array of beams can be extended by connectingthem by springs and at the same time the output power isincreased The performance of this system can be optimizedfor specific applications by a proper selection of the tipmassesand spring constants
The preliminary experimental results are consistent withthe analytical ones but further research is necessary toimprove the experimental setup and to investigate differentexperimental conditions such as variations of the springsstiffness
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
References
[1] S R Anton and H A Sodano ldquoA review of power harvestingusing piezoelectric materials (2003ndash2006)rdquo Smart Materialsand Structures vol 16 no 3 article R1 2007
[2] S Chalasani and J M Conrad ldquoA survey of energy harvestingsources for embedded systemsrdquo in Proceedings of the IEEESoutheastCon pp 442ndash447 April 2008
[3] K A Cook-Chennault NThambi and AM Sastry ldquoPoweringMEMS portable devicesmdasha review of non-regenerative andregenerative power supply systems with special emphasis on
Shock and Vibration 13
piezoelectric energy harvesting systemsrdquo Smart Materials andStructures vol 17 no 4 Article ID 043001 2008
[4] A Dewan S U Ay M N Karim and H Beyenal ldquoAlternativepower sources for remote sensors a reviewrdquo Journal of PowerSources vol 245 pp 129ndash143 2014
[5] A Harb ldquoEnergy harvesting state-of-the-artrdquo RenewableEnergy vol 36 no 10 pp 2641ndash2654 2011
[6] S Saadon and O Sidek ldquoA review of vibration-based MEMSpiezoelectric energy harvestersrdquo Energy Conversion and Man-agement vol 52 no 1 pp 500ndash504 2011
[7] H A Sodano D J Inman and G Park ldquoA review of powerharvesting from vibration using piezoelectric materialsrdquo Shockand Vibration Digest vol 36 no 3 pp 197ndash205 2004
[8] R J M Vullers R van Schaijk I Doms C Van Hoofand R Mertens ldquoMicropower energy harvestingrdquo Solid-StateElectronics vol 53 no 7 pp 684ndash693 2009
[9] R Calio U B Rongala D Camboni et al ldquoPiezoelectric energyharvesting solutionsrdquo Sensors vol 14 no 3 pp 4755ndash4790 2014
[10] J Twiefel and H Westermann ldquoSurvey on broadband tech-niques for vibration energy harvestingrdquo Journal of IntelligentMaterial Systems and Structures vol 24 no 11 pp 1291ndash13022013
[11] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavengingrdquo Journal of Sound and Vibration vol 292no 3ndash5 pp 987ndash998 2006
[12] S M Shahruz ldquoLimits of performance of mechanical band-pass filters used in energy scavengingrdquo Journal of Sound andVibration vol 293 no 1-2 pp 449ndash461 2006
[13] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavenging multi-degree-of-freedom modelsrdquo Journalof Vibration and Control vol 14 no 5 pp 753ndash768 2008
[14] J-Q Liu H-B Fang Z-Y Xu et al ldquoA MEMS-based piezo-electric power generator array for vibration energy harvestingrdquoMicroelectronics Journal vol 39 no 5 pp 802ndash806 2008
[15] HC Lin PHWu I C Lien andYC Shu ldquoAnalysis of an arrayof piezo-electric energy harvesters connected in seriesrdquo SmartMaterials and Structures vol 22 no 9 Article ID 094026 2013
[16] S-J Jang E Rustighi M J Brennan Y P Lee and H-JJung ldquoDesign of a 2DOF vibrational energy harvesting devicerdquoJournal of IntelligentMaterial Systems and Structures vol 22 no5 pp 443ndash448 2011
[17] L Tang and Y Yang ldquoA multiple-degree-of-freedom piezoelec-tric energy harvesting modelrdquo Journal of Intelligent MaterialSystems and Structures vol 23 no 14 pp 1631ndash1647 2012
[18] Z Yang and J Yang ldquoConnected vibrating piezoelectricbimorph beams as a wide-band piezoelectric power harvesterrdquoJournal of IntelligentMaterial Systems and Structures vol 20 no5 pp 569ndash574 2009
[19] W Zhou G R Penamalli and L Zuo ldquoAn efficient vibrationenergy harvester with a multi-mode dynamic magnifierrdquo SmartMaterials and Structures vol 21 no 1 Article ID 015014 2012
[20] D S Nguyen E Halvorsen G U Jensen and A VoglldquoFabrication and characterization of a widebandMEMS energyharvester utilizing nonlinear springsrdquo Journal of Micromechan-ics andMicroengineering vol 20 no 12 Article ID 125009 2010
[21] S D Nguyen and E Halvorsen ldquoNonlinear springs forbandwidth-tolerant vibration energy harvestingrdquo Journal ofMicroelectromechanical Systems vol 20 no 6 pp 1225ndash12272011
[22] S BoisseauGDespesse andBA Seddik ldquoNonlinear h-shapedsprings to improve efficiency of vibration energy harvestersrdquo
Journal of Applied Mechanics vol 80 no 6 Article ID 0610132013
[23] S Priya ldquoAdvances in energy harvesting using low profilepiezoelectric transducersrdquo Journal of Electroceramics vol 19 no1 pp 165ndash182 2007
[24] S Roundy P K Wright and J Rabaey ldquoA study of lowlevel vibrations as a power source for wireless sensor nodesrdquoComputer Communications vol 26 no 11 pp 1131ndash1144 2003
[25] S Roundy and P K Wright ldquoA piezoelectric vibration basedgenerator for wireless electronicsrdquo Smart Materials and Struc-tures vol 13 no 5 pp 1131ndash1142 2004
[26] H A Sodano G Park and D J Inman ldquoEstimation of electriccharge output for piezoelectric energy harvestingrdquo Strain vol40 no 2 pp 49ndash58 2004
[27] N E Dutoit B L Wardle and S-G Kim ldquoDesign considera-tions for mems-scale piezoelectric mechanical vibration energyharvestersrdquo Integrated Ferroelectrics vol 71 no 1 pp 121ndash1602005
[28] S-N Chen G-J Wang andM-C Chien ldquoAnalytical modelingof piezoelectric vibration-induced micro power generatorrdquoMechatronics vol 16 no 7 pp 379ndash387 2006
[29] A Erturk and D J Inman ldquoIssues in mathematical modelingof piezoelectric energy harvestersrdquo Smart Materials and Struc-tures vol 17 no 6 Article ID 065016 2008
[30] F Lu H P Lee and S P Lim ldquoModeling and analysis of micropiezoelectric power generators for micro-electromechanical-systems applicationsrdquo Smart Materials and Structures vol 13no 1 pp 57ndash63 2004
[31] S Jiang X Li S Guo Y Hu J Yang andQ Jiang ldquoPerformanceof a piezoelectric bimorph for scavenging vibration energyrdquoSmartMaterials and Structures vol 14 no 4 pp 769ndash774 2005
[32] A Erturk and D J Inman ldquoA distributed parameter elec-tromechanical model for cantilevered piezoelectric energy har-vestersrdquo Journal of Vibration and Acoustics vol 130 no 4Article ID 041002 2008
[33] A Erturk and D J Inman ldquoAn experimentally validatedbimorph cantilever model for piezoelectric energy harvestingfrom base excitationsrdquo Smart Materials and Structures vol 18no 2 Article ID 025009 2009
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
10 Shock and Vibration
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
01mL 02mL01mL 03mL
01mL 04mL01mL 05mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k = 02YIL3k = 03YIL3
k = 04YIL3
(b) Varying spring constants (1198721199051 = 03119898119871 1198721199052 = 04119898119871)
Figure 8 Power FRF of two bimorph beams with different combinations of tip masses and spring constants
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
02mL 03mL 04mL02mL 02mL 04mL
03mL 03mL 04mL
(a) Varying tip masses (119896 = 031198841198681198713)
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
k =
k =k = 02YIL3 02YIL302YIL3 04YIL3
02YIL3 06YIL3
(b) Varying spring constants (1198721199051 = 02119898119871 1198721199052 = 03119898119871 1198721199053 =04119898119871)
Figure 9 Power FRF of three bimorph beams with different combinations of tip masses and spring constants
associated with each resonant frequency depend on thecombination of tipmasses and springs constantsThis ismoreevident in Figures 8 and 9 where different combinations ofsprings and tip masses are studied for the case of two andthree beams In general increasing the tip mass of the beamsdecreases someof the resonant frequencies and increasing thespring stiffness increases the frequency band
To compare the performances of an array of beamsconnected by springs and one of the disconnected beams acase of ten beams with both situations is plotted in Figure 10It is possible to see that the array with connected beamshas a slightly wider frequency band which increases withan increment of the springs constants Furthermore bycomparing the areas below both curves in the frequencyrange 50ndash200Hz the array with connected beams providesan area between 20 and 36 larger than the array withdisconnected beams Thus by connecting the beams theperformance of the array can be improved It should benoted that the previous results are obtained with arbitraryvalues for the tip masses and spring constantsTherefore thisperformance can be further improved by a proper design
52 Experimental To validate the proposed model a pre-liminary experimental campaign was performed Figure 11presents the experimental setup It consists of a signal genera-tor connected to a power amplifier and to an electrodynamicshaker which produces a controlled base vibration An arrayof three piezoelectric bimorph beams is mounted over theshaker Figure 12 shows this array The cantilever bimorphbeams have a tip mass at the free-end and can be connectedwith springs between them Three experiments were per-formed In the first case no springs were used and eachbeam was connected independently to a 1 kΩ resistive loadin this case the voltage generated by each beam is monitoredindividually In the second case no springs were used andthe beams were electrically connected in parallel to a resistiveload of 1 kΩ In the last case the beams were connected withsprings between them and they were electrically connected inparallel to a resistive load of 1 kΩ
The properties of the bimorph beams are the ones givenin Table 1 the tip masses are119872
1199051= 65119898119871119872
1199052= 35119898119871
1198721199053= 19119898119871 and the spring constants are 119896
1= 1198962=
2921198841198681198713
Shock and Vibration 11
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected02YIL3
(a) 119896119894=19 = 021198841198681198713
50 100 150 200
Frequency (Hz)
10minus4
Pow
er F
RF (W
s4m
2)
10minus3
10minus2
10minus1
100
Connected k =
Disconnected03YIL3
(b) 119896119894=19 = 031198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected04YIL3
(c) 119896119894=19 = 041198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected05YIL3
(d) 119896119894=19 = 051198841198681198713
Figure 10 Power FRF with ten bimorph beams1198721199051= 01119898119871119872
1199052= 015119898119871119872
1199053= 02119898119871119872
1199054= 025119898119871119872
1199055= 03119898119871119872
1199056= 035119898119871
1198721199057= 04119898119871119872
1199058= 045119898119871119872
1199059= 05119898119871119872
11990510= 055119898119871
Figure 11 Experimental setup
Figure 12 Experimental bimorph beams
During the three experiments the system is subjected toa harmonic base displacement with frequencies in the range20ndash70Hz The results for the first experiment are presentedin Figure 13(a) and are compared with the ones obtainedby an analytical model in Figure 13(b) Both analytical andexperimental curves are very similar The only difference isthat the experimental power increases from beams 1 to 3 andthe analytical model predicted the opposite behavior Thisunexpected behavior seems to be related to a resonance ofthe supporting structure in the three experimental curves itis possible to see an increment of the power near 60Hz whichcan be explained by a resonance of the structure
Figure 14 presents the experimental and analytical resultsfor the second and third experiments It is possible to seea good agreement between the experimental and analyticalcurves although similarly to the results in Figure 13(a) thereis an increment of the power near 60Hz that confirms astructural resonance
Comparing the areas below both experimental curves inFigure 13(b) in the frequency range 20ndash70Hz the array withconnected beams provides an area 5 larger than the arraywith disconnected beamswhich confirms that the addition ofsprings can increment the generated powerThese results canbe improved by using softer springs which have lower energylosses and by solving the structural resonance problem
12 Shock and Vibration
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 13 Power FRF for the three beams with no springs and independent electrical connections to 1 kΩ resistive loads (a) experimentalmeasurement (b) analytical results
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 14 Power FRF for the three beams with and without springs and electrical connection in parallel to a resistive load of 1 kΩ(a) experimental measurement (b) analytical results
6 Conclusions
This paper investigated the use of an array of bimorphbeams connected by springs as a broadband vibration-basedenergy harvester The equations for the voltage and poweroutput of the full system have been deduced based on theanalytical solution of the cantilevered piezoelectric energyharvester with Euler-Bernoulli beam assumptions To studythe advantages and disadvantages of the proposed system theresults are compared with those of an array of disconnectedbeams (with no springs) The analytical model is validatedwith experimental measurements of three bimorph beamswith and without springs
Both analytical and experimental results show that con-necting the array of beams with springs allows increasingthe frequency band of operation and increasing the amountof power generated Thus the maximal frequency band fora certain array of beams can be extended by connectingthem by springs and at the same time the output power isincreased The performance of this system can be optimizedfor specific applications by a proper selection of the tipmassesand spring constants
The preliminary experimental results are consistent withthe analytical ones but further research is necessary toimprove the experimental setup and to investigate differentexperimental conditions such as variations of the springsstiffness
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
References
[1] S R Anton and H A Sodano ldquoA review of power harvestingusing piezoelectric materials (2003ndash2006)rdquo Smart Materialsand Structures vol 16 no 3 article R1 2007
[2] S Chalasani and J M Conrad ldquoA survey of energy harvestingsources for embedded systemsrdquo in Proceedings of the IEEESoutheastCon pp 442ndash447 April 2008
[3] K A Cook-Chennault NThambi and AM Sastry ldquoPoweringMEMS portable devicesmdasha review of non-regenerative andregenerative power supply systems with special emphasis on
Shock and Vibration 13
piezoelectric energy harvesting systemsrdquo Smart Materials andStructures vol 17 no 4 Article ID 043001 2008
[4] A Dewan S U Ay M N Karim and H Beyenal ldquoAlternativepower sources for remote sensors a reviewrdquo Journal of PowerSources vol 245 pp 129ndash143 2014
[5] A Harb ldquoEnergy harvesting state-of-the-artrdquo RenewableEnergy vol 36 no 10 pp 2641ndash2654 2011
[6] S Saadon and O Sidek ldquoA review of vibration-based MEMSpiezoelectric energy harvestersrdquo Energy Conversion and Man-agement vol 52 no 1 pp 500ndash504 2011
[7] H A Sodano D J Inman and G Park ldquoA review of powerharvesting from vibration using piezoelectric materialsrdquo Shockand Vibration Digest vol 36 no 3 pp 197ndash205 2004
[8] R J M Vullers R van Schaijk I Doms C Van Hoofand R Mertens ldquoMicropower energy harvestingrdquo Solid-StateElectronics vol 53 no 7 pp 684ndash693 2009
[9] R Calio U B Rongala D Camboni et al ldquoPiezoelectric energyharvesting solutionsrdquo Sensors vol 14 no 3 pp 4755ndash4790 2014
[10] J Twiefel and H Westermann ldquoSurvey on broadband tech-niques for vibration energy harvestingrdquo Journal of IntelligentMaterial Systems and Structures vol 24 no 11 pp 1291ndash13022013
[11] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavengingrdquo Journal of Sound and Vibration vol 292no 3ndash5 pp 987ndash998 2006
[12] S M Shahruz ldquoLimits of performance of mechanical band-pass filters used in energy scavengingrdquo Journal of Sound andVibration vol 293 no 1-2 pp 449ndash461 2006
[13] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavenging multi-degree-of-freedom modelsrdquo Journalof Vibration and Control vol 14 no 5 pp 753ndash768 2008
[14] J-Q Liu H-B Fang Z-Y Xu et al ldquoA MEMS-based piezo-electric power generator array for vibration energy harvestingrdquoMicroelectronics Journal vol 39 no 5 pp 802ndash806 2008
[15] HC Lin PHWu I C Lien andYC Shu ldquoAnalysis of an arrayof piezo-electric energy harvesters connected in seriesrdquo SmartMaterials and Structures vol 22 no 9 Article ID 094026 2013
[16] S-J Jang E Rustighi M J Brennan Y P Lee and H-JJung ldquoDesign of a 2DOF vibrational energy harvesting devicerdquoJournal of IntelligentMaterial Systems and Structures vol 22 no5 pp 443ndash448 2011
[17] L Tang and Y Yang ldquoA multiple-degree-of-freedom piezoelec-tric energy harvesting modelrdquo Journal of Intelligent MaterialSystems and Structures vol 23 no 14 pp 1631ndash1647 2012
[18] Z Yang and J Yang ldquoConnected vibrating piezoelectricbimorph beams as a wide-band piezoelectric power harvesterrdquoJournal of IntelligentMaterial Systems and Structures vol 20 no5 pp 569ndash574 2009
[19] W Zhou G R Penamalli and L Zuo ldquoAn efficient vibrationenergy harvester with a multi-mode dynamic magnifierrdquo SmartMaterials and Structures vol 21 no 1 Article ID 015014 2012
[20] D S Nguyen E Halvorsen G U Jensen and A VoglldquoFabrication and characterization of a widebandMEMS energyharvester utilizing nonlinear springsrdquo Journal of Micromechan-ics andMicroengineering vol 20 no 12 Article ID 125009 2010
[21] S D Nguyen and E Halvorsen ldquoNonlinear springs forbandwidth-tolerant vibration energy harvestingrdquo Journal ofMicroelectromechanical Systems vol 20 no 6 pp 1225ndash12272011
[22] S BoisseauGDespesse andBA Seddik ldquoNonlinear h-shapedsprings to improve efficiency of vibration energy harvestersrdquo
Journal of Applied Mechanics vol 80 no 6 Article ID 0610132013
[23] S Priya ldquoAdvances in energy harvesting using low profilepiezoelectric transducersrdquo Journal of Electroceramics vol 19 no1 pp 165ndash182 2007
[24] S Roundy P K Wright and J Rabaey ldquoA study of lowlevel vibrations as a power source for wireless sensor nodesrdquoComputer Communications vol 26 no 11 pp 1131ndash1144 2003
[25] S Roundy and P K Wright ldquoA piezoelectric vibration basedgenerator for wireless electronicsrdquo Smart Materials and Struc-tures vol 13 no 5 pp 1131ndash1142 2004
[26] H A Sodano G Park and D J Inman ldquoEstimation of electriccharge output for piezoelectric energy harvestingrdquo Strain vol40 no 2 pp 49ndash58 2004
[27] N E Dutoit B L Wardle and S-G Kim ldquoDesign considera-tions for mems-scale piezoelectric mechanical vibration energyharvestersrdquo Integrated Ferroelectrics vol 71 no 1 pp 121ndash1602005
[28] S-N Chen G-J Wang andM-C Chien ldquoAnalytical modelingof piezoelectric vibration-induced micro power generatorrdquoMechatronics vol 16 no 7 pp 379ndash387 2006
[29] A Erturk and D J Inman ldquoIssues in mathematical modelingof piezoelectric energy harvestersrdquo Smart Materials and Struc-tures vol 17 no 6 Article ID 065016 2008
[30] F Lu H P Lee and S P Lim ldquoModeling and analysis of micropiezoelectric power generators for micro-electromechanical-systems applicationsrdquo Smart Materials and Structures vol 13no 1 pp 57ndash63 2004
[31] S Jiang X Li S Guo Y Hu J Yang andQ Jiang ldquoPerformanceof a piezoelectric bimorph for scavenging vibration energyrdquoSmartMaterials and Structures vol 14 no 4 pp 769ndash774 2005
[32] A Erturk and D J Inman ldquoA distributed parameter elec-tromechanical model for cantilevered piezoelectric energy har-vestersrdquo Journal of Vibration and Acoustics vol 130 no 4Article ID 041002 2008
[33] A Erturk and D J Inman ldquoAn experimentally validatedbimorph cantilever model for piezoelectric energy harvestingfrom base excitationsrdquo Smart Materials and Structures vol 18no 2 Article ID 025009 2009
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
Shock and Vibration 11
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected02YIL3
(a) 119896119894=19 = 021198841198681198713
50 100 150 200
Frequency (Hz)
10minus4
Pow
er F
RF (W
s4m
2)
10minus3
10minus2
10minus1
100
Connected k =
Disconnected03YIL3
(b) 119896119894=19 = 031198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected04YIL3
(c) 119896119894=19 = 041198841198681198713
50 100 150 200
Frequency (Hz)
Pow
er F
RF (W
s4m
2)
10minus4
10minus3
10minus2
10minus1
100
Connected k =
Disconnected05YIL3
(d) 119896119894=19 = 051198841198681198713
Figure 10 Power FRF with ten bimorph beams1198721199051= 01119898119871119872
1199052= 015119898119871119872
1199053= 02119898119871119872
1199054= 025119898119871119872
1199055= 03119898119871119872
1199056= 035119898119871
1198721199057= 04119898119871119872
1199058= 045119898119871119872
1199059= 05119898119871119872
11990510= 055119898119871
Figure 11 Experimental setup
Figure 12 Experimental bimorph beams
During the three experiments the system is subjected toa harmonic base displacement with frequencies in the range20ndash70Hz The results for the first experiment are presentedin Figure 13(a) and are compared with the ones obtainedby an analytical model in Figure 13(b) Both analytical andexperimental curves are very similar The only difference isthat the experimental power increases from beams 1 to 3 andthe analytical model predicted the opposite behavior Thisunexpected behavior seems to be related to a resonance ofthe supporting structure in the three experimental curves itis possible to see an increment of the power near 60Hz whichcan be explained by a resonance of the structure
Figure 14 presents the experimental and analytical resultsfor the second and third experiments It is possible to seea good agreement between the experimental and analyticalcurves although similarly to the results in Figure 13(a) thereis an increment of the power near 60Hz that confirms astructural resonance
Comparing the areas below both experimental curves inFigure 13(b) in the frequency range 20ndash70Hz the array withconnected beams provides an area 5 larger than the arraywith disconnected beamswhich confirms that the addition ofsprings can increment the generated powerThese results canbe improved by using softer springs which have lower energylosses and by solving the structural resonance problem
12 Shock and Vibration
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 13 Power FRF for the three beams with no springs and independent electrical connections to 1 kΩ resistive loads (a) experimentalmeasurement (b) analytical results
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 14 Power FRF for the three beams with and without springs and electrical connection in parallel to a resistive load of 1 kΩ(a) experimental measurement (b) analytical results
6 Conclusions
This paper investigated the use of an array of bimorphbeams connected by springs as a broadband vibration-basedenergy harvester The equations for the voltage and poweroutput of the full system have been deduced based on theanalytical solution of the cantilevered piezoelectric energyharvester with Euler-Bernoulli beam assumptions To studythe advantages and disadvantages of the proposed system theresults are compared with those of an array of disconnectedbeams (with no springs) The analytical model is validatedwith experimental measurements of three bimorph beamswith and without springs
Both analytical and experimental results show that con-necting the array of beams with springs allows increasingthe frequency band of operation and increasing the amountof power generated Thus the maximal frequency band fora certain array of beams can be extended by connectingthem by springs and at the same time the output power isincreased The performance of this system can be optimizedfor specific applications by a proper selection of the tipmassesand spring constants
The preliminary experimental results are consistent withthe analytical ones but further research is necessary toimprove the experimental setup and to investigate differentexperimental conditions such as variations of the springsstiffness
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
References
[1] S R Anton and H A Sodano ldquoA review of power harvestingusing piezoelectric materials (2003ndash2006)rdquo Smart Materialsand Structures vol 16 no 3 article R1 2007
[2] S Chalasani and J M Conrad ldquoA survey of energy harvestingsources for embedded systemsrdquo in Proceedings of the IEEESoutheastCon pp 442ndash447 April 2008
[3] K A Cook-Chennault NThambi and AM Sastry ldquoPoweringMEMS portable devicesmdasha review of non-regenerative andregenerative power supply systems with special emphasis on
Shock and Vibration 13
piezoelectric energy harvesting systemsrdquo Smart Materials andStructures vol 17 no 4 Article ID 043001 2008
[4] A Dewan S U Ay M N Karim and H Beyenal ldquoAlternativepower sources for remote sensors a reviewrdquo Journal of PowerSources vol 245 pp 129ndash143 2014
[5] A Harb ldquoEnergy harvesting state-of-the-artrdquo RenewableEnergy vol 36 no 10 pp 2641ndash2654 2011
[6] S Saadon and O Sidek ldquoA review of vibration-based MEMSpiezoelectric energy harvestersrdquo Energy Conversion and Man-agement vol 52 no 1 pp 500ndash504 2011
[7] H A Sodano D J Inman and G Park ldquoA review of powerharvesting from vibration using piezoelectric materialsrdquo Shockand Vibration Digest vol 36 no 3 pp 197ndash205 2004
[8] R J M Vullers R van Schaijk I Doms C Van Hoofand R Mertens ldquoMicropower energy harvestingrdquo Solid-StateElectronics vol 53 no 7 pp 684ndash693 2009
[9] R Calio U B Rongala D Camboni et al ldquoPiezoelectric energyharvesting solutionsrdquo Sensors vol 14 no 3 pp 4755ndash4790 2014
[10] J Twiefel and H Westermann ldquoSurvey on broadband tech-niques for vibration energy harvestingrdquo Journal of IntelligentMaterial Systems and Structures vol 24 no 11 pp 1291ndash13022013
[11] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavengingrdquo Journal of Sound and Vibration vol 292no 3ndash5 pp 987ndash998 2006
[12] S M Shahruz ldquoLimits of performance of mechanical band-pass filters used in energy scavengingrdquo Journal of Sound andVibration vol 293 no 1-2 pp 449ndash461 2006
[13] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavenging multi-degree-of-freedom modelsrdquo Journalof Vibration and Control vol 14 no 5 pp 753ndash768 2008
[14] J-Q Liu H-B Fang Z-Y Xu et al ldquoA MEMS-based piezo-electric power generator array for vibration energy harvestingrdquoMicroelectronics Journal vol 39 no 5 pp 802ndash806 2008
[15] HC Lin PHWu I C Lien andYC Shu ldquoAnalysis of an arrayof piezo-electric energy harvesters connected in seriesrdquo SmartMaterials and Structures vol 22 no 9 Article ID 094026 2013
[16] S-J Jang E Rustighi M J Brennan Y P Lee and H-JJung ldquoDesign of a 2DOF vibrational energy harvesting devicerdquoJournal of IntelligentMaterial Systems and Structures vol 22 no5 pp 443ndash448 2011
[17] L Tang and Y Yang ldquoA multiple-degree-of-freedom piezoelec-tric energy harvesting modelrdquo Journal of Intelligent MaterialSystems and Structures vol 23 no 14 pp 1631ndash1647 2012
[18] Z Yang and J Yang ldquoConnected vibrating piezoelectricbimorph beams as a wide-band piezoelectric power harvesterrdquoJournal of IntelligentMaterial Systems and Structures vol 20 no5 pp 569ndash574 2009
[19] W Zhou G R Penamalli and L Zuo ldquoAn efficient vibrationenergy harvester with a multi-mode dynamic magnifierrdquo SmartMaterials and Structures vol 21 no 1 Article ID 015014 2012
[20] D S Nguyen E Halvorsen G U Jensen and A VoglldquoFabrication and characterization of a widebandMEMS energyharvester utilizing nonlinear springsrdquo Journal of Micromechan-ics andMicroengineering vol 20 no 12 Article ID 125009 2010
[21] S D Nguyen and E Halvorsen ldquoNonlinear springs forbandwidth-tolerant vibration energy harvestingrdquo Journal ofMicroelectromechanical Systems vol 20 no 6 pp 1225ndash12272011
[22] S BoisseauGDespesse andBA Seddik ldquoNonlinear h-shapedsprings to improve efficiency of vibration energy harvestersrdquo
Journal of Applied Mechanics vol 80 no 6 Article ID 0610132013
[23] S Priya ldquoAdvances in energy harvesting using low profilepiezoelectric transducersrdquo Journal of Electroceramics vol 19 no1 pp 165ndash182 2007
[24] S Roundy P K Wright and J Rabaey ldquoA study of lowlevel vibrations as a power source for wireless sensor nodesrdquoComputer Communications vol 26 no 11 pp 1131ndash1144 2003
[25] S Roundy and P K Wright ldquoA piezoelectric vibration basedgenerator for wireless electronicsrdquo Smart Materials and Struc-tures vol 13 no 5 pp 1131ndash1142 2004
[26] H A Sodano G Park and D J Inman ldquoEstimation of electriccharge output for piezoelectric energy harvestingrdquo Strain vol40 no 2 pp 49ndash58 2004
[27] N E Dutoit B L Wardle and S-G Kim ldquoDesign considera-tions for mems-scale piezoelectric mechanical vibration energyharvestersrdquo Integrated Ferroelectrics vol 71 no 1 pp 121ndash1602005
[28] S-N Chen G-J Wang andM-C Chien ldquoAnalytical modelingof piezoelectric vibration-induced micro power generatorrdquoMechatronics vol 16 no 7 pp 379ndash387 2006
[29] A Erturk and D J Inman ldquoIssues in mathematical modelingof piezoelectric energy harvestersrdquo Smart Materials and Struc-tures vol 17 no 6 Article ID 065016 2008
[30] F Lu H P Lee and S P Lim ldquoModeling and analysis of micropiezoelectric power generators for micro-electromechanical-systems applicationsrdquo Smart Materials and Structures vol 13no 1 pp 57ndash63 2004
[31] S Jiang X Li S Guo Y Hu J Yang andQ Jiang ldquoPerformanceof a piezoelectric bimorph for scavenging vibration energyrdquoSmartMaterials and Structures vol 14 no 4 pp 769ndash774 2005
[32] A Erturk and D J Inman ldquoA distributed parameter elec-tromechanical model for cantilevered piezoelectric energy har-vestersrdquo Journal of Vibration and Acoustics vol 130 no 4Article ID 041002 2008
[33] A Erturk and D J Inman ldquoAn experimentally validatedbimorph cantilever model for piezoelectric energy harvestingfrom base excitationsrdquo Smart Materials and Structures vol 18no 2 Article ID 025009 2009
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
12 Shock and Vibration
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
Beam 1Beam 2
Beam 3
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 13 Power FRF for the three beams with no springs and independent electrical connections to 1 kΩ resistive loads (a) experimentalmeasurement (b) analytical results
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(a) Experimental
20 25 30 35 40 45 50 55 60 65 70
Frequency (Hz)
ConnectedDisconnected
Pow
er F
RF (W
s4m
2)
10minus7
10minus6
10minus9
10minus8
10minus5
10minus4
(b) Analytical
Figure 14 Power FRF for the three beams with and without springs and electrical connection in parallel to a resistive load of 1 kΩ(a) experimental measurement (b) analytical results
6 Conclusions
This paper investigated the use of an array of bimorphbeams connected by springs as a broadband vibration-basedenergy harvester The equations for the voltage and poweroutput of the full system have been deduced based on theanalytical solution of the cantilevered piezoelectric energyharvester with Euler-Bernoulli beam assumptions To studythe advantages and disadvantages of the proposed system theresults are compared with those of an array of disconnectedbeams (with no springs) The analytical model is validatedwith experimental measurements of three bimorph beamswith and without springs
Both analytical and experimental results show that con-necting the array of beams with springs allows increasingthe frequency band of operation and increasing the amountof power generated Thus the maximal frequency band fora certain array of beams can be extended by connectingthem by springs and at the same time the output power isincreased The performance of this system can be optimizedfor specific applications by a proper selection of the tipmassesand spring constants
The preliminary experimental results are consistent withthe analytical ones but further research is necessary toimprove the experimental setup and to investigate differentexperimental conditions such as variations of the springsstiffness
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
References
[1] S R Anton and H A Sodano ldquoA review of power harvestingusing piezoelectric materials (2003ndash2006)rdquo Smart Materialsand Structures vol 16 no 3 article R1 2007
[2] S Chalasani and J M Conrad ldquoA survey of energy harvestingsources for embedded systemsrdquo in Proceedings of the IEEESoutheastCon pp 442ndash447 April 2008
[3] K A Cook-Chennault NThambi and AM Sastry ldquoPoweringMEMS portable devicesmdasha review of non-regenerative andregenerative power supply systems with special emphasis on
Shock and Vibration 13
piezoelectric energy harvesting systemsrdquo Smart Materials andStructures vol 17 no 4 Article ID 043001 2008
[4] A Dewan S U Ay M N Karim and H Beyenal ldquoAlternativepower sources for remote sensors a reviewrdquo Journal of PowerSources vol 245 pp 129ndash143 2014
[5] A Harb ldquoEnergy harvesting state-of-the-artrdquo RenewableEnergy vol 36 no 10 pp 2641ndash2654 2011
[6] S Saadon and O Sidek ldquoA review of vibration-based MEMSpiezoelectric energy harvestersrdquo Energy Conversion and Man-agement vol 52 no 1 pp 500ndash504 2011
[7] H A Sodano D J Inman and G Park ldquoA review of powerharvesting from vibration using piezoelectric materialsrdquo Shockand Vibration Digest vol 36 no 3 pp 197ndash205 2004
[8] R J M Vullers R van Schaijk I Doms C Van Hoofand R Mertens ldquoMicropower energy harvestingrdquo Solid-StateElectronics vol 53 no 7 pp 684ndash693 2009
[9] R Calio U B Rongala D Camboni et al ldquoPiezoelectric energyharvesting solutionsrdquo Sensors vol 14 no 3 pp 4755ndash4790 2014
[10] J Twiefel and H Westermann ldquoSurvey on broadband tech-niques for vibration energy harvestingrdquo Journal of IntelligentMaterial Systems and Structures vol 24 no 11 pp 1291ndash13022013
[11] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavengingrdquo Journal of Sound and Vibration vol 292no 3ndash5 pp 987ndash998 2006
[12] S M Shahruz ldquoLimits of performance of mechanical band-pass filters used in energy scavengingrdquo Journal of Sound andVibration vol 293 no 1-2 pp 449ndash461 2006
[13] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavenging multi-degree-of-freedom modelsrdquo Journalof Vibration and Control vol 14 no 5 pp 753ndash768 2008
[14] J-Q Liu H-B Fang Z-Y Xu et al ldquoA MEMS-based piezo-electric power generator array for vibration energy harvestingrdquoMicroelectronics Journal vol 39 no 5 pp 802ndash806 2008
[15] HC Lin PHWu I C Lien andYC Shu ldquoAnalysis of an arrayof piezo-electric energy harvesters connected in seriesrdquo SmartMaterials and Structures vol 22 no 9 Article ID 094026 2013
[16] S-J Jang E Rustighi M J Brennan Y P Lee and H-JJung ldquoDesign of a 2DOF vibrational energy harvesting devicerdquoJournal of IntelligentMaterial Systems and Structures vol 22 no5 pp 443ndash448 2011
[17] L Tang and Y Yang ldquoA multiple-degree-of-freedom piezoelec-tric energy harvesting modelrdquo Journal of Intelligent MaterialSystems and Structures vol 23 no 14 pp 1631ndash1647 2012
[18] Z Yang and J Yang ldquoConnected vibrating piezoelectricbimorph beams as a wide-band piezoelectric power harvesterrdquoJournal of IntelligentMaterial Systems and Structures vol 20 no5 pp 569ndash574 2009
[19] W Zhou G R Penamalli and L Zuo ldquoAn efficient vibrationenergy harvester with a multi-mode dynamic magnifierrdquo SmartMaterials and Structures vol 21 no 1 Article ID 015014 2012
[20] D S Nguyen E Halvorsen G U Jensen and A VoglldquoFabrication and characterization of a widebandMEMS energyharvester utilizing nonlinear springsrdquo Journal of Micromechan-ics andMicroengineering vol 20 no 12 Article ID 125009 2010
[21] S D Nguyen and E Halvorsen ldquoNonlinear springs forbandwidth-tolerant vibration energy harvestingrdquo Journal ofMicroelectromechanical Systems vol 20 no 6 pp 1225ndash12272011
[22] S BoisseauGDespesse andBA Seddik ldquoNonlinear h-shapedsprings to improve efficiency of vibration energy harvestersrdquo
Journal of Applied Mechanics vol 80 no 6 Article ID 0610132013
[23] S Priya ldquoAdvances in energy harvesting using low profilepiezoelectric transducersrdquo Journal of Electroceramics vol 19 no1 pp 165ndash182 2007
[24] S Roundy P K Wright and J Rabaey ldquoA study of lowlevel vibrations as a power source for wireless sensor nodesrdquoComputer Communications vol 26 no 11 pp 1131ndash1144 2003
[25] S Roundy and P K Wright ldquoA piezoelectric vibration basedgenerator for wireless electronicsrdquo Smart Materials and Struc-tures vol 13 no 5 pp 1131ndash1142 2004
[26] H A Sodano G Park and D J Inman ldquoEstimation of electriccharge output for piezoelectric energy harvestingrdquo Strain vol40 no 2 pp 49ndash58 2004
[27] N E Dutoit B L Wardle and S-G Kim ldquoDesign considera-tions for mems-scale piezoelectric mechanical vibration energyharvestersrdquo Integrated Ferroelectrics vol 71 no 1 pp 121ndash1602005
[28] S-N Chen G-J Wang andM-C Chien ldquoAnalytical modelingof piezoelectric vibration-induced micro power generatorrdquoMechatronics vol 16 no 7 pp 379ndash387 2006
[29] A Erturk and D J Inman ldquoIssues in mathematical modelingof piezoelectric energy harvestersrdquo Smart Materials and Struc-tures vol 17 no 6 Article ID 065016 2008
[30] F Lu H P Lee and S P Lim ldquoModeling and analysis of micropiezoelectric power generators for micro-electromechanical-systems applicationsrdquo Smart Materials and Structures vol 13no 1 pp 57ndash63 2004
[31] S Jiang X Li S Guo Y Hu J Yang andQ Jiang ldquoPerformanceof a piezoelectric bimorph for scavenging vibration energyrdquoSmartMaterials and Structures vol 14 no 4 pp 769ndash774 2005
[32] A Erturk and D J Inman ldquoA distributed parameter elec-tromechanical model for cantilevered piezoelectric energy har-vestersrdquo Journal of Vibration and Acoustics vol 130 no 4Article ID 041002 2008
[33] A Erturk and D J Inman ldquoAn experimentally validatedbimorph cantilever model for piezoelectric energy harvestingfrom base excitationsrdquo Smart Materials and Structures vol 18no 2 Article ID 025009 2009
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
Shock and Vibration 13
piezoelectric energy harvesting systemsrdquo Smart Materials andStructures vol 17 no 4 Article ID 043001 2008
[4] A Dewan S U Ay M N Karim and H Beyenal ldquoAlternativepower sources for remote sensors a reviewrdquo Journal of PowerSources vol 245 pp 129ndash143 2014
[5] A Harb ldquoEnergy harvesting state-of-the-artrdquo RenewableEnergy vol 36 no 10 pp 2641ndash2654 2011
[6] S Saadon and O Sidek ldquoA review of vibration-based MEMSpiezoelectric energy harvestersrdquo Energy Conversion and Man-agement vol 52 no 1 pp 500ndash504 2011
[7] H A Sodano D J Inman and G Park ldquoA review of powerharvesting from vibration using piezoelectric materialsrdquo Shockand Vibration Digest vol 36 no 3 pp 197ndash205 2004
[8] R J M Vullers R van Schaijk I Doms C Van Hoofand R Mertens ldquoMicropower energy harvestingrdquo Solid-StateElectronics vol 53 no 7 pp 684ndash693 2009
[9] R Calio U B Rongala D Camboni et al ldquoPiezoelectric energyharvesting solutionsrdquo Sensors vol 14 no 3 pp 4755ndash4790 2014
[10] J Twiefel and H Westermann ldquoSurvey on broadband tech-niques for vibration energy harvestingrdquo Journal of IntelligentMaterial Systems and Structures vol 24 no 11 pp 1291ndash13022013
[11] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavengingrdquo Journal of Sound and Vibration vol 292no 3ndash5 pp 987ndash998 2006
[12] S M Shahruz ldquoLimits of performance of mechanical band-pass filters used in energy scavengingrdquo Journal of Sound andVibration vol 293 no 1-2 pp 449ndash461 2006
[13] S M Shahruz ldquoDesign of mechanical band-pass filters forenergy scavenging multi-degree-of-freedom modelsrdquo Journalof Vibration and Control vol 14 no 5 pp 753ndash768 2008
[14] J-Q Liu H-B Fang Z-Y Xu et al ldquoA MEMS-based piezo-electric power generator array for vibration energy harvestingrdquoMicroelectronics Journal vol 39 no 5 pp 802ndash806 2008
[15] HC Lin PHWu I C Lien andYC Shu ldquoAnalysis of an arrayof piezo-electric energy harvesters connected in seriesrdquo SmartMaterials and Structures vol 22 no 9 Article ID 094026 2013
[16] S-J Jang E Rustighi M J Brennan Y P Lee and H-JJung ldquoDesign of a 2DOF vibrational energy harvesting devicerdquoJournal of IntelligentMaterial Systems and Structures vol 22 no5 pp 443ndash448 2011
[17] L Tang and Y Yang ldquoA multiple-degree-of-freedom piezoelec-tric energy harvesting modelrdquo Journal of Intelligent MaterialSystems and Structures vol 23 no 14 pp 1631ndash1647 2012
[18] Z Yang and J Yang ldquoConnected vibrating piezoelectricbimorph beams as a wide-band piezoelectric power harvesterrdquoJournal of IntelligentMaterial Systems and Structures vol 20 no5 pp 569ndash574 2009
[19] W Zhou G R Penamalli and L Zuo ldquoAn efficient vibrationenergy harvester with a multi-mode dynamic magnifierrdquo SmartMaterials and Structures vol 21 no 1 Article ID 015014 2012
[20] D S Nguyen E Halvorsen G U Jensen and A VoglldquoFabrication and characterization of a widebandMEMS energyharvester utilizing nonlinear springsrdquo Journal of Micromechan-ics andMicroengineering vol 20 no 12 Article ID 125009 2010
[21] S D Nguyen and E Halvorsen ldquoNonlinear springs forbandwidth-tolerant vibration energy harvestingrdquo Journal ofMicroelectromechanical Systems vol 20 no 6 pp 1225ndash12272011
[22] S BoisseauGDespesse andBA Seddik ldquoNonlinear h-shapedsprings to improve efficiency of vibration energy harvestersrdquo
Journal of Applied Mechanics vol 80 no 6 Article ID 0610132013
[23] S Priya ldquoAdvances in energy harvesting using low profilepiezoelectric transducersrdquo Journal of Electroceramics vol 19 no1 pp 165ndash182 2007
[24] S Roundy P K Wright and J Rabaey ldquoA study of lowlevel vibrations as a power source for wireless sensor nodesrdquoComputer Communications vol 26 no 11 pp 1131ndash1144 2003
[25] S Roundy and P K Wright ldquoA piezoelectric vibration basedgenerator for wireless electronicsrdquo Smart Materials and Struc-tures vol 13 no 5 pp 1131ndash1142 2004
[26] H A Sodano G Park and D J Inman ldquoEstimation of electriccharge output for piezoelectric energy harvestingrdquo Strain vol40 no 2 pp 49ndash58 2004
[27] N E Dutoit B L Wardle and S-G Kim ldquoDesign considera-tions for mems-scale piezoelectric mechanical vibration energyharvestersrdquo Integrated Ferroelectrics vol 71 no 1 pp 121ndash1602005
[28] S-N Chen G-J Wang andM-C Chien ldquoAnalytical modelingof piezoelectric vibration-induced micro power generatorrdquoMechatronics vol 16 no 7 pp 379ndash387 2006
[29] A Erturk and D J Inman ldquoIssues in mathematical modelingof piezoelectric energy harvestersrdquo Smart Materials and Struc-tures vol 17 no 6 Article ID 065016 2008
[30] F Lu H P Lee and S P Lim ldquoModeling and analysis of micropiezoelectric power generators for micro-electromechanical-systems applicationsrdquo Smart Materials and Structures vol 13no 1 pp 57ndash63 2004
[31] S Jiang X Li S Guo Y Hu J Yang andQ Jiang ldquoPerformanceof a piezoelectric bimorph for scavenging vibration energyrdquoSmartMaterials and Structures vol 14 no 4 pp 769ndash774 2005
[32] A Erturk and D J Inman ldquoA distributed parameter elec-tromechanical model for cantilevered piezoelectric energy har-vestersrdquo Journal of Vibration and Acoustics vol 130 no 4Article ID 041002 2008
[33] A Erturk and D J Inman ldquoAn experimentally validatedbimorph cantilever model for piezoelectric energy harvestingfrom base excitationsrdquo Smart Materials and Structures vol 18no 2 Article ID 025009 2009
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of