Available online at www.sciencedirect.

com

ScienceDirect
International Journal of Pavement Research and Technology 11 (2018) 381–387
www.elsevier.com/locate/IJPRT

Synergistic performance of piezoelectric transducers and

asphalt pavement
Hongduo Zhao a, Luyao Qin b,⇑, Jianming Ling a
a
Key Laboratory of Road and Traffic Engineering of Ministry of Education, Tongji University, Shanghai 201804, China
b
Shanghai Municipal Engineering Design General Institute (Group) CO., LTD, Shanghai 200092, China

Received 1 March 2017; received in revised form 31 August 2017; accepted 11 September 2017
Available online 15 November 2017

Abstract

The deformation of piezoelectric transducer and asphalt pavement differs because of their different stiffness. As the piezoelectric trans-
ducer is embedded in asphalt pavement, the differential deformation may cause a non-uniform stress and degrade their performance. In
this paper, the synergistic performance of piezoelectric transducer and asphalt pavement is studied through finite element analysis.
Results show that the energy conversion efficiency of the transducers decreases with increasing burial depth and modulus of surface
course. Under the vehicle loading, the piezoelectric transducer and asphalt pavement are easily separated by shear stress. For the asphalt
pavement, the change in the stress of pavement structure and surface deflection is very small, which means that the differential deforma-
tion has no effect on fatigue cracking. Furthermore, an energy harvesting system composed of transducers array and asphalt plate is
made and evaluated in a laboratory. Results show that the transducers within asphalt pavement have acceptable efficiency and high
durability.
Ó 2017 Chinese Society of Pavement Engineering. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Synergistic performance; Piezoelectric transducer; Asphalt pavement; Energy harvesting; Cymbal; Differential deformation

1. Introduction
Asphalt pavement during its life’s cycle is frequented by
millions of vehicles that receive strain, stress and kinetic
energy from the work of load and gravity of the vehicle.
Based on piezoelectricity theory, these energies can be con-
verted into electric energy by applying a load to the embed-
ding piezoelectric transducer.
Most of typical piezoelectric transducers are used to
harvest ambient energy from asphalt pavement, such as
the Multilayer [1,2], PZT (Lead zirconate titanate) piles
⇑ Corresponding author.
E-mail addresses: hdzhao@tongji.edu.cn (H. Zhao), qinluyao@smedi.
com (L. Qin), jmling@tongji.edu.cn (J. Ling).
Peer review under responsibility of Chinese Society of Pavement
Engineering.
[3], Cymbal [4] and Bridge [5]. As the main piezoelectric
material for the transducer, the stiffness of PZT is more
than 100 GPa. Therefore, the stiffness of most piezoelectric
transducers is more than asphalt pavement. As the piezo-
electric transducer embedded in asphalt pavement, the dif-
ferential deformation may degrade their performance and
even cause the cracking. Although the Cymbal and Bridge
have the moderate stiffness close to asphalt pavement [6],
there is little research aiming at the change in performance
of piezoelectric transducer and asphalt pavement.
The purpose of this paper is discussing the synergistic
performance of piezoelectric transducer and asphalt pave-
ment. The relation between energy conversion efficiency
of the transducer and structure parameters of pavement
is presented. The failure mode of transducer and asphalt
pavement is also analyzed. On this basis, the performance
of an optimized energy harvesting system is evaluated.

https://doi.org/10.1016/j.ijprt.2017.09.008
1996-6814/Ó 2017 Chinese Society of Pavement Engineering. Production and hosting by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
382 H. Zhao et al. / International Journal of Pavement Research and Technology 11 (2018) 381–387
2. Bridge piezoelectric transducer
Considering the reasonable efficiency and moderate stiff-
ness, Bridge transducer is taken as the study object in this
paper. Arch [7] is one of the modified Bridge transducers,
whose structure is shown in Fig. 1. The metal caps are arch
designed geometries to transform the vertical stress into
horizontal. Besides, the air cavity between the metal cap
and ceramic also decreases the elastic modulus in vertical
direction.
When the uniform static load is applied to the metal
caps in the vertical direction, the metal cap can convert
and amplify the vertical stress to a much larger axial stress
of the PZT disk. At the same time, a polarization P3
appears on the vertical surface (assumed as the 3rd axial
direction) of PZT. Then the piezoelectric equation is
described as Eq. (1) [8].
X
6
P3 ¼ d 3i T i ¼ d 31 T x þ d 33 T y
i¼1
where P3 is the piezoelectric polarization at direction 3,
whose magnitude is equal to the charge density at the
applied surface; d is piezoelectric strain constant tensor;
Tx is the horizontal stress of PZT disk; Ty is the vertical
stress of PZT disk.
Then the electric potential (voltage) V3 and electric
energy stored UE are calculated by Eqs. (2) and (3):
Z
P3 tp
V3 ¼ dtp ¼ T ðd 31 T x þ d 33 T y Þ ð2Þ
eT33 e33
1 1 et e0 A
U E ¼ V 23 C ¼ V 23 r
2 2 tp
where eT33 is the relative dielectric constant of PZT; e0 is the
dielectric constant of vacuum; C is the electric capacity of
PZT ceramic.
In order to avoid the influence of natural factors on the
transducers in the pavement environment, the transducer
should be encapsulated. The encapsulation mode is shown
in Fig. 2. The transducer is encapsulated in a thin-walled
metal cylinder and surrounded by filler material. It is worth
noting that the stiffness of filler material should be closed to
asphalt pavement so that the differential deformation is
less.
Fig. 1. Structure of Arch piezoelectric transducer.
Fig. 2. Encapsulation mode of piezoelectric transducer.
In addition, several arch transducers are designed to
array for construction. The specific way is that transducers
are bonded together to form the array. According to the
effective area of vehicle load, the array with 20 mm of
ð1Þ
length and width is suggested and shown in Fig. 3. Its
height is the same as in the encapsulation mode. In order
to control the differential deformation, the stiffness of
bonding layer should be close to that of asphalt pavement.
3. Finite element analysis
3.1. Model and parameters
In this finite element model, elastic material properties
are used for asphalt pavement and piezoelectric transducer.
The constraint is set as ‘‘tie” and the interface is assumed to
ð3Þ continuity condition. The asphalt pavement with 1.875 m
of width and 2.60 m of depth is a semi-rigid structure,
whose parameters are shown in Table 1.
In this paper, the contact pressure between tire and
pavement is assumed to equal the tire pressure (25 kN for
one tire), and the shape of contact area is assumed to be
round. Under the tire load, Arch array is embedded about
3 cm deep into the surface course. The material properties
of Arch array are shown in Table 2. On the account of high
piezoelectric strain constant tensor and relative dielectric
constant, the PZT-5H is chosen as PZT material of piezo-
electric transducer [9]. Based on the above conditions, the
two-dimensional finite element model of Arch array and
asphalt pavement is built and shown in Fig. 4.
3.2. Results of transducers
In the model, the six Arch transducers are numbered
one to six from left to right. For comparing the perfor-
mance degradation, an Arch, which is not embedded in
the pavement, is numbered zero. The electric potential of
each Arch and the maximum stress of PZT disk are shown
in Table 3.
From the Table 3, it can be found that the electric
potential and maximum stress is decreased when the trans-
ducer is embedded in the pavement. According to Eqs. (2)
 H. Zhao et al. / International Journal of Pavement Research and Technology 11 (2018) 381–387 383

Fig. 3. Structure of Arch transducer array.

Table 1
Parameters of semi-rigid asphalt pavement structure.
Structure layers Thickness Density Modulus Poisson’s ratio
(cm) (kg/m3) (MPa)
Surface 15 2400 1200 0.25
Base 30 2000 1500 0.25
Subbase 15 1800 200 0.30
Subgrade 200 1800 20 0.35

Table 2
Material properties of Arch transducers array.
Component Material Modulus Poisson’s ratio
(MPa)
Metal cap and cylinder Stainless steel 207000 0.30
Filler material Sand 20 0.35
Bonding layer Cement 2000 0.30 Fig. 4. Finite element mode of transducer and asphalt pavement.

and (3), the energy conversion efficiency is proportional to
electric potential and stress. Thus, the less electric potential
means that the energy conversion efficiency is decreased.
The reason is that the vertical stress of transducer is
decreased. And one fundamental reason is that the stiffness
of bonding layer is more than that of the transducer and
asphalt pavement.
Considering the major influence factors to energy con-
version efficiency, the sensitivity of burial depth, thickness
and modulus of surface course are analyzed. The sensitivity
analysis results are shown in Figs. 5–7. The results show
that the energy conversion efficiency of the transducers
decreases with increasing burial depth and modulus of sur-
face course, the decrements are about 13 V/cm and 12
V/200 MPa. However, the thickness of surface course has
little influence on the energy conversion efficiency.
On the other hand, the less stress of PZT disk means
that it is hard to be broken. Taking the number 2
384 H. Zhao et al. / International Journal of Pavement Research and Technology 11 (2018) 381–387

Table 3
Results of electric potential and maximum stress.
Number 0 1 2 3 4 5 6
Electric voltage (V) 556.3 204.6 287.9 198.9 197.0 287.9 205.2
Maximum tensile stress (MPa) 33.2 14.2 23.8 13.8 13.8 23.8 14.2
Maximum shearing stress (MPa) 6.48 2.25 3.11 2.17 2.16 3.11 2.25

Fig. 5. Burial depth sensitivity analysis. Fig. 7. Stress distribution of PZT disk.

Fig. 8. Shear stress in the contact interface.

Fig. 6. Surface course sensitivity analysis.

strength of modified asphalt is about 0.3 MPa [10]. From
the figure, it can found that change trend of three transduc-
transducer for example, the stress distribution of PZT disk
ers is almost same. The maximum shear stress of number 1
in horizontal direction is shown in Fig. 7. It shows that the
Arch transducer is more than others because its position is
maximum tensile stress and shearing stress are both at the
the end of the force effect direction. In both ends of contact
inner corner of the contact area between cap and PZT disk.
interface, the shear stress is more than the bonding
It means the PZT disk is easier to crack by the concen-
strength. This dangerous area accounts for 50% of the total
trated stress.
area, which means that the piezoelectric transducer and
In addition, the horizontal force caused by the moving
asphalt pavement are easily separated by shear stress.
vehicle cannot be ignored. Considering the most unfavor-
able condition, the drag coefficient between surface and
wheel is 0.95, and the horizontal stress is assumed to uni- 3.3. Results of asphalt pavement
form distribution. After calculation, the results of shear
stress in the contact interface between Arch array and In the asphalt pavement structure design, the principal
asphalt pavement are shown in Fig. 8. The bonding stress of structure layer bottom and surface deflection is
the main index. In this paper, they are also used to evaluate
 H. Zhao et al. / International Journal of Pavement Research and Technology 11 (2018) 381–387
Table 4
Principal stress of structure layer bottom with and without transducers.
Surface Base Subbase
bottom bottom bottom
Pavement with transducers 0.050 MPa 0.091 MPa 0.005 MPa
Pavement without 0.051 MPa 0.092 MPa 0.005 MPa
transducers
the state of fatigue cracking and the whole rigidity of the
pavement. The principal stress of structure layer bottom
with and without transducers is shown in Table 4. It can
be found that the principal stress of structure layer bottom
with and without transducers is almost the same. The
embedding transducers have obvious influence on the area
within 2 cm and the maximum different stress is only 0.06
MPa.
Using the pavement cross section as the horizontal ordi-
nate, the surface deflection is shown in Fig. 9. It shows that
the surface deflection increases when the transducers are
embedded in the pavement. The maximum difference of
surface deflection is 0.028 mm, which is right above the
transducers. However, it is small enough to be ignored.
Overall, the change in the stress of pavement structure
and surface deflection is very small. It means that the stiff-
ness of Arch array is close to that of asphalt pavement and
the differential deformation has no effect on fatigue
cracking.
4. Test and evaluation
In order to evaluate the synergistic performance of
piezoelectric and asphalt pavement, an energy harvesting
system composed of transducers array and asphalt plate
is made and tested in a laboratory. The AC-10 asphalt mix-
ture is used for bonding layer and the asphalt plate. In this
way, the difference in stiffness is resolved and the bonding
strength is improved. The Arch array, composed of four
Fig. 9. Surface deflection of asphalt pavement.
385
transducers, is embedded in the asphalt plate with a depth
of 2 cm. The encapsulation mode of Arch and the asphalt
plate is shown in Figs. 10 and 11. Their size and material
are the same as the finite element model.
The performance of piezoelectric transducer and asphalt
pavement was tested and evaluated by MMLS3 (1/3 Model
Mobile Load Simulator). MMLS3 is a pavement testing
frame with four groups of tire imitating the tire motion
of motor vehicles. In the test, its wheel velocity and applied
load of the tires are 2.5 m/s and 0.75 MPa, respectively. In
addition, the Tektronix oscilloscope was used to collect the
electric signal and voltage.
In the test and evaluation, two transducers did not gen-
erate the electric signal because their wire was broken. The
other two transducers keep a stable power output during
the test. However, their energy conversion efficiency is
reduced about 15% compared with those that were not
embedded in the plate. In the repeated wheel load, the max-
imum electric voltage is 202 V and the change trend of elec-
tric voltage is shown in Fig. 12. The figure shows that the
electric voltage increased in the beginning and then
decreased. The reason is that the compaction of asphalt
Fig. 10. Arch transducer.
386 H. Zhao et al. / International Journal of Pavement Research and Technology 11 (2018) 381–387
Fig. 11. Asphalt plate.
Fig. 12. Change trend of electric voltage.
Fig. 13. Rutting damage.
is not enough in the fabrication. In the beginning, the com-
paction increases with the increasing wheel loading times,
meanwhile the vertical stress of transducer increases. The
compaction maximizes the electric voltage output stability
for some time which then decreases gradually because of
the performance degradation. Results indicate that the
transducer within asphalt pavement has acceptable effi-
ciency and high durability. At the same time, some struc-
ture parameters have proved to be reasonable.
On the other hand, a rut depth of 6 mm appeared in the
middle of asphalt plate, which is shown in Fig. 13. From
the figure the obvious rutting damage but no crack can
be seen. The main reason is also the low degree of com-
paction. If the fabrication technique is improved, the syn-
ergistic performance of transducers and asphalt would be
improved.
5. Conclusions
In this paper, the synergistic performance of piezoelec-
tric transducer and asphalt pavement is studied through
finite element analysis and laboratory test. Some significant
conclusions are made as follows:
(1) The energy conversion efficiency of the transducer
decreases when it is embedded in asphalt pavement.
The decrement is proportional to the burial depth
and modulus of surface course.
(2) The stress concentration at the inner corner of the
contact area between cap and PZT disk is easily bro-
ken. Moreover, the shear stress is more than the
bonding strength in both ends of contact interface,
which may lead to the separation of transducer and
asphalt pavement.
(3) The change in the stress of structure layer bottom and
surface deflection is very small, which means that the
differential deformation has no effect on fatigue
cracking of asphalt pavement.
(4) The Arch transducers within asphalt pavement have
acceptable efficiency and high durability. The struc-
ture parameters of Arch transducer and encapsula-
tion mode have proved to be reasonable. However,
the bonding layer and fabrication technique should
be improved for further application.
References
[1] A. Heinzmann, E. Henning, B. Kolle, et al., Properties of PZT
multilayer actuators, in: 8th International Conference on New
Actuators, Bremen, Germany, 2002.
[2] K. Uchino, Ferroelectric Devices, CRC Press, New York, 2010.
[3] H. Zhao, Y. Tao, Y. Niu, et al., Harvesting Energy from asphalt
pavement by piezoelectric generator, J. Wuhan Univ. Technol. Mater.
29 (5) (2014).
[4] A. Dogan, ‘‘Flextensional ‘Moonie and cymbal’ actuators.” a thesis in
materials program submitted in partial fulfillment of the requirements
for the degree of Doctor of Philosophy, 1994, pp. 63-66.
[5] Y. Tao, H. Zhao, J. Ling, Analysis of performance of Bridge
piezoelectric transducer within asphalt pavement. Shanghai High-
ways, Shanghai, China, 2012.
[6] H. Zhao, J. Ling, J. Yu, A comparative analysis of piezoelectric
transducer for harvesting energy from asphalt pavement, J. Ceram.
Soc. Japan 120 (2012) 317–323.
 H. Zhao et al. / International Journal of Pavement Research and Technology 11 (2018) 381–387 387

[7] H. Zhao, L. Qin, J. Ling, Test and analysis of bridge transducers for
harvesting energy from asphalt pavement, Int. J. Transp. Sci.
Technol. 4 (1) (2015) 17–28.
[8] A.A. Vives, Piezoelectric Transducers and Applications Second
Edition, Springer, New York, U.S.A, 2008.
[9] D. Song, M. Xiao, Piezoelectric Effects and Application, Popular
science press, Beijing, China, 1987.
[10] Y. Wang, Y. Zhou, A. Yao, et al., Test of shear and pull-off between
asphalt and concrete on bridge deck pavement structure, J. Chang’an
Univ. (Natl. Sci. Ed.) 29 (6) (2009) 15–18.