Energy Conversion and Management 163 (2018) 196–207

Contents lists available at ScienceDirect

Energy Conversion and Management

journal homepage: www.elsevier.com/locate/enconman

Fabrication and performance of a power generation device based on stacked T

piezoelectric energy-harvesting units for pavements
⁎
Chaohui Wanga, , Shuai Wanga, Qiang Joshua Lib, Xingju Wangc, Zhiwei Gaoa, Lu Zhangd
a
School of Highway, Chang’an University, Xi’an 710064, China
b
School of Civil and Environmental Engineering, Oklahoma State University, Stillwater, OK 74074, USA
c
School of Traffic and Transportation, Shijiazhuang Tiedao University, Shijiazhuang 050043, China
d
Hebei Branch of China State Construction Engineering Corporation, Shijiazhuang 050011, China

A R T I C L E I N F O A B S T R A C T

Keywords: Although piezoelectric power generation for pavements is technically feasible, harvesting energy in a stable and
Energy-harvesting efficient manner still presents several challenges. This paper designs and assesses stacked piezoelectric energy-
Piezoelectric harvesting power generation devices for pavements, in which energy output and road coupling are considered.
Power generation device Following corresponding technical specifications, the dimensions of the proposed power generation device are
Stacked units
determined, and the unit is assembled for subsequent testing in terms of its electrical energy outputs. The energy
Electrical output
outputs are analyzed in various series and parallel configurations of the units, different loads, and frequency
Pavement engineering
conditions. Results indicate that power generation device with the dimension of 100 mm × 100 mm achieves
optimal power generation outputs. The electrical output is positively correlated to the number of parallel pie-
zoelectric energy-harvesting units, the magnitude of load, and the frequency of traffic action. The maximum
output at the 0.2 MPa load magnitude and 10 Hz load frequency is 0.88 mW, and the corresponding optimum
load is 20 kΩ. The maximum output at the 0.7 MPa and 15 Hz load level can reach up to 11.67 mW, and the
corresponding optimum load is 10 kΩ. Subsequently, Mechanical Testing and Simulation (MTS) is performed
with 40,000 load cycles, and the overall deformation of the power generation device is minimal and the elec-
trical outputs of the piezoelectric energy-harvesting units remain stable. Finally, future research plan is dis-
cussed.

1. Introduction piezoelectric materials and pavement materials involves the combina-

tion of piezoelectric and pavement materials and the preparation of
Environmental micro-energy collection and utilization has gradu- piezoelectric composite materials for pavements through a certain
ally become an important research direction in various industries due to process used directly in pavement paving. However, limited by material
environmental pollution, energy shortage, and other issues. Existing properties and restrictions of the preparation process, the studies about
piezoelectric energy harvesting techniques could convert mechanical piezoelectric composite materials only produced a weak electro-me-
vibration energy into electrical energy through the piezoelectric effect chanical response. Tan et al. [8] prepared 0–3 piezoelectric ceramic/
of piezoelectric materials [1–6]. If such techniques could be applied to asphalt composite piezoelectric material by hot embossing and
road pavement engineering to develop piezoelectric power pavements, achieved a maximum output voltage of 7.2 V under dynamic load.
then the concept of green transition from the abundant mechanical Wang et al. [9] prepared two forms of piezoelectric asphalt concrete
vibration of roads to electrical energy can be realized. At present, the with d31 and d33 by using tourmaline, piezoelectric ceramic, and gra-
research on piezoelectric power technologies for pavements have phite. The output voltage reached up to 2.4 V under the combined
proved that piezoelectric power pavements are feasible in technical treatment of three measures, namely, insulation treatment, piezo-
aspects. The two technical approaches are the technology of power electric material fibrosis, and polarization treatment. Many problems
pavements based on the integration of piezoelectric materials and pa- arise in the preparation of high-efficiency piezoelectric composites for
vement materials and the technology of power pavements based on pavements. For instance, the technical difficulties and interference
embedded piezoelectric energy-harvesting units [7]. factors involved are numerous, and the energy output effect of piezo-
The technology of power pavements based on the integration of electric composites is poor. Meanwhile, road performance is difficult to

⁎
Corresponding author.
E-mail address: wchh0205@chd.edu.cn (C. Wang).

https://doi.org/10.1016/j.enconman.2018.02.045
Received 21 November 2017; Received in revised form 2 February 2018; Accepted 9 February 2018
Available online 23 February 2018
0196-8904/ © 2018 Elsevier Ltd. All rights reserved.
C. Wang et al. Energy Conversion and Management 163 (2018) 196–207

ensure. Therefore, studies should explore the research and development output is not ideal. Moreover, most studies conducted preliminary ex-
of piezoelectric materials, the method of polarization and optimization plorations with software simulations or simple indoor tests on plants.
of the preparation process, and the improvements of road performance Hence, substantive research into integrated power generation devices
to improve the comprehensive application performance of road piezo- for improving energy output is still limited.
electric composites. In summary, many countries have studied different types of piezo-
In power pavements based on embedded piezoelectric energy-har- electric power pavement technologies from different levels and dif-
vesting units, the power pavement is embedded with piezoelectric en- ferent angles. The mainstream ideas are biased toward integrated
ergy-harvesting units for energy conversion. Compared with the tech- power generation devices based on piezoelectric energy-harvesting
nology of power pavements based on the integration of piezoelectric units. However, the design and implementation methods of the devices
materials and pavement materials, power pavements embedded with disclosed in existing studies are few, and the research and development
piezoelectric energy-harvesting units exhibit high energy output level, of high-efficiency power generation devices for pavements based on
strong controllability, and relatively simple implementation. Hence, the piezoelectric energy-harvesting units is still in the initial stage.
latter type is likely to realize pavement piezoelectric power generation To realize the application of piezoelectric power generation pave-
in the future. At present, many countries have carried out relevant re- ments, we design a power generation device for pavements that in-
search. Zhao and Najini et al. [10,11] used finite element software to tegrates energy output and road coupling. The structure and size of the
compare the energy output efficiency of multi-layer cantilevers and power generation device are optimized based on traffic load char-
cymbals in the road environment and concluded that the energy output acteristics. The complete production and assembly process are estab-
of a multi-layer transducer is relatively superior. Xiang et al. [12] stu- lished. The energy output of the piezoelectric power generation device
died the theoretical parameters of piezoelectric energy-harvesting units is studied comprehensively by means of a test platform. Then, the en-
embedded in pavement and found that driving speed, basic conditions, ergy output status of different power generation devices is analyzed in
and system damping exert a significant effect on the output voltage and different series and parallel configurations, different loads, and dif-
output power of piezoelectric energy-harvesting units. Kim et al. [13] ferent frequency conditions. Finally, the working durability of the
found that load case, loading frequency, and layers of energy-har- proposed power generation device is evaluated.
vesting units are the key points to increase output voltage. Moure et al.
[14] optimized the structure parameters of cymbal harvesters through a 2. Design of piezoelectric power generation device for pavements
laboratory test and found that a single cymbal harvester can collect up
to 16 µW power under a single driving load. The technology of power A piezoelectric power generation device is essentially a carrier de-
pavements based on embedded piezoelectric energy-harvesting units is vice integrated with piezoelectric harvesting units. the energy output of
clearly still in the exploratory stage. Although the technology has the device is directly affected by the power generation performance of
prospective application in energy acquisition, problems such as the the piezoelectric energy-harvesting units, and the device must be laid in
imperfect optimization of the matching of piezoelectric energy-har- the pavement structure to create a piezoelectric effect under driving
vesting units and road traffic environment and the non-ideal energy load rolling. Therefore, based on power generation theory of energy-
output require immediate solutions. harvesting units, the structure of the proposed device is designed based
Given the limited energy output of a single energy-harvesting unit, a on electro-mechanical conversion principles of the energy-harvesting
substantial number of piezoelectric elements must be paved in a pa- units and the structural characteristics of roads. The size of proposed
vement structure in a specific array for the pavement to achieve a device is optimized for driving load characteristics on pavements.
substantial piezoelectric energy output. If piezoelectric energy-har-
vesting units were buried one by one, construction operations will be- 2.1. Power generation theory of energy-harvesting units
come complex, power generation performance will weaken, and
structural damage problems will arise. Therefore, the development of a Since piezoelectric energy-harvesting units are consisted of energy
power generation device that is based on integrated piezoelectric en- conversion components of power generation devices, the type of pie-
ergy-harvesting units will serve as the mainstream technology for the zoelectric material has a significant impact on energy conversion effi-
future realization of piezoelectric power pavements. The Innowattech ciency. Piezoelectric Ceramic Transducer (PZT) is the most widely used
Company of Israel [15,16] pioneered the application of piezoelectric material for energy-harvesting units, and its power - electric conversion
technology in power pavement systems and implanted a large number modes are mainly d31 (the direction of electric field is perpendicular to
of piezoelectric crystals in ordinary asphalt pavements; for 1 km of the direction of axial stress) and d33 (the direction of the electric field
road, about 100–200 kW·h of electricity can be produced, but so far, and the direction of the stress are the same). Cook-Chennault et al. [4]
this application has not been widely promoted. The University of Texas found that the piezoelectric coefficient d33 of PZT is much larger than
in the United States [17] developed a power generation device com- d31, indicating that d33 mode has higher energy conversion efficiency.
prising three layers of PZT prism in a series connection. The output The third type of piezoelectric equations is used to analyze the power
power was measured to be 3.5 mW under a 10 Hz simulated driving generation performance of PZT. The piezoelectric equation with d33
load. Jiang [18,19] combined three stacked piezoelectric energy-har- mode is shown below [23].
vesting units and built a simple power generation device for analog D
S3 (t ) = s33 T3 (t ) + g t D3 (t ) (1)
testing; the results showed that the maximum output power is up to 85
mW with the matching load size of 350 kΩ. Roshani [20] carried out a T
E3 (t ) = −g33 T3 (t ) + β33 D3 (t ) (2)
uniaxial compression test on an assembled power generation device
with piezoelectric energy-harvesting units clamped by two pairs of where S is strain tensor, T is stress tensor, D is electric displacement
copper; the results indicated that the number and arrangement of pie- tensor, E is electric strength tensor, sD is open elastic compliance con-
zoelectric energy-harvesting units can cause changes in output power. stant matrix, βT is free dielectric isolation rate matrix, g is voltage
Wang [21,22] investigated the ways to achieve piezoelectric power constant matrix, gt is transposed matrix of g. If the PZT is subjected to
pavements and designed the basic structure of a piezoelectric power vertical load (F = Fmax sin(ωt ) ), then D3, T3, E3, S3 and F are sinusoidal
generation device; this work provides a reference for the research and functions of time t and have the same frequency.
application of piezoelectric power pavements. Nevertheless, many Stacked piezoelectric energy-harvesting unit is one of the d33 mode
problems still need to be addressed in the development of integrated piezoelectric structures, which can effectively improve the voltage
power generation devices. For example, the optimization of the output, and also enhance the energy conversion efficiency under low
matching of road traffic environment is not perfect, and the energy frequency state [24]. Therefore, stacked piezoelectric energy-

197
C. Wang et al. Energy Conversion and Management 163 (2018) 196–207

harvesting unit is more suitable for the road environment of high stress, when paving the asphalt pavement, we choose a carrier substrate with
low frequency vibration. The output voltage and power of the unit with high-temperature insulation function and large structural strength. The
parallel connection can be characterized by the following equation shell and bolt gap are treated for waterproofing and high temperature
[23]. stability given the complex construction and working environment in-
d33 Fm ωnhR volved in the practical application of power generation devices.
Uop =
T 2 2 2
h2 + ω2A2 (ε33 )nR (3) 2.3. Device dimension
2 2 2 2 2
d33 Fm ω n h R
Pop = T 2 2 2 The efficient use of load and the optimal output of piezoelectric
2[h2 + ω2A2 (ε33 )nR] (4)
energy can be achieved by matching the size and position of the power
The output voltage and power of the unit with series connection can generation device with road load characteristics. Therefore, considering
be characterized by the following equation. the lateral distribution characteristics of the wheel path, tire ground
d33 Fm ωnhR characteristics, and driving rolling conditions, we optimize the size of
Uos = the piezoelectric power generation device and establish an optimal
T 2 2
nh2 + ω2A2 (ε33 )R (5) design scheme of piezoelectric power generation device under different
2 2 2 2 2 traffic conditions.
d33 Fm ω n h R
Pos = T 2 2
2[n2h2 + ω2A2 (ε33 )R] (6) (1) Dimensional optimum design of power generation device based on
T the lateral distribution of wheel path
where d33 is the piezoelectric coefficient, ε33 is the dielectric constant,
Cp is the capacitance of the unit, h is the thickness of the unit, A is the
area of the unit, n is the number of units with series and parallel con- The distribution of vehicle wheel path has obvious agglomeration
nection, Fm is the frequency, R is the load resistance. because of the constraint guidance of traffic signs. According to statis-
The output power of the stacked piezoelectric energy-harvesting tics, about 27% of the width of the road bears nearly 60% of vehicle
unit with series and parallel connection is related not only to the d33, h, load [25]. Considering the engineering economy of piezoelectric power
T
A, and ε33 , but to the n, Fm, ω /2π, R. Therefore, the energy output effects pavement applications, we lay the power generation device in the range
of the devices will be tested with different series and parallel config- of the main track width. The initial limit of the transverse maximum
urations of the energy-harvesting units, different loads, and different size of a single piezoelectric power device should not exceed 500 mm.
frequency conditions.
(2) Dimensional optimum design of power generation device based on
2.2. Structural design tire grounding characteristics

The energy output efficiency of integrated piezoelectric power In contrast to the circular uniform load in the traditional pavement
generation devices is affected by road traffic characteristics in addition mechanics model, the tires of actual driving load tend to be rectangular
to the influence of piezoelectric energy units. Therefore, we compre- [26]. The lateral ground width of a tire is basically the same as the tread
hensively consider the electro-mechanical conversion of piezoelectric width of the tire, and the tread width of tires is 0.7–0.85 times the
energy-harvesting units and road structural characteristics and consider section width [27,28]. According to the model of the common types of
the following technical requirements of a piezoelectric power genera- car tires, the horizontal ground width of a car tire is calculated to be in
tion device. (1) The electro-mechanical conversion of a piezoelectric the range of 110–220 mm, and that of a truck tire is 130–260 mm. The
power unit should be facilitated. (2) The coupling of the device with longitudinal tire length increases as load increases, and it decreases as
road structure should be at a high level, and the device must show good tire pressure increases [28–30]. With increasing load under normal tire
durability. (3) The piezoelectric harvesting unit and its transmission pressure, the longitudinal tire grounding length of 205-type cars in-
lines should be effectively fixed and protected. (4) The construction and creases from 104 mm to 142 mm, with the tire grounding length being
maintenance of the device should be convenient, and the device itself 124 mm. The tire length of 11.00–20-type heavy vehicles is usually
must be reusable. Considering the technical requirements above, we between 200 and 300 mm. Given the use of most traffic loads, the
design a piezoelectric power generation device that is based on in- vertical size of a power generation device for urban application should
tegrated stacked piezoelectric harvesting units. The device is mainly be between 100 and 120 mm, and the size for heavy traffic conditions
composed of the rigid bearing shell, the connecting component, the should be less than or between 200 and 300 mm. However, the size of
energy-harvesting unit and its carrier substrate, and other components. the power generation device should enable it to afford most of the
The structure is shown in Fig. 1. traffic load. Thus, the size of the device should be at the lower limit.
Considering potential thermal depolarization of the harvesting unit In summary, the balance of all traffic loads can be ensured by setting

Fig. 1. Schematic structural diagram of power generation device.

198
C. Wang et al. Energy Conversion and Management 163 (2018) 196–207

the vertical and horizontal dimensions of the power generation device 3.2. Optimization of internal configurations
to 100 mm × 100 mm for an urban traffic environment. The installation
of units within the device, construction convenience, production costs, Different configurations of the components of the power generation
and other factors must be considered. The size of the device should be device can provide a different stress environment for the energy-har-
set to 100 mm × 100 mm and 150 mm × 150 mm for a heavy traffic vesting units. Thus, the energy output of the device is different. In this
environment. A 100 mm × 100 mm-type power generation device can regard, the electro-mechanical responses of the device under different
be used in both urban and heavy traffic environments. Therefore, the configurations must be analyzed to determine the optimal configuration
100 mm × 100 mm-type power generation device is used for the pro- of the components of the device. The electro-mechanical responses of
duction and performance research. different power generation devices are tested under different cover
plates, different connecting components, and different specifications of
3. Fabrication and optimization of the device the energy-harvesting units. The results should provide a reference for
the optimization of the internal configuration of the device. To facilitate
The effective application of the electro-mechanical conversion per- the recording of observations, we arrange four energy-harvesting units
formance of a power generation device needs to be based on the precise in the 100 mm × 100 mm-type device in a numbered Z shape. We de-
coordination of the components. According to the functional require- note the units as #1, #2, #3, and #4.
ments and material characteristics of different structural components,
the corresponding process is used to achieve standardized processing, 3.2.1. Different top plates
and the assembly method is optimized. The steel plate and hard aluminum alloy plate can meet the re-
quirements of the power generation device in terms of the stiffness and
3.1. Device fabrication process flatness of the top plate. The steel plate with the same thickness as the
aluminum alloy plate is more resistant to load deformation than the
The device fabrication process can be divided into three steps: shell latter. Hence, it is more conducive to the uniform transmission of loads.
preparation, fabrication of connecting components and substrate, and However, the quality of the steel plate is almost three times that of
device assembly, etc. aluminum. Hence, the pre-pressure of the energy-harvesting units in-
creases, and their electro-mechanical responses are retained. Therefore,
3.1.1. Shell preparation to clarify the influence of different plate types on the electro-mechan-
The shell structure of the power generation device includes a top ical responses of the device, we test the responses of the device with
plate, bottom plate, and sidewall. The top plate and bottom plate are different plate types are with a rutting instrument (Fig. 4). The test
hard materials that can be processed quickly and exhibit high rigidity, results are shown in Table 1. The eight-layer energy-harvesting units
high strength, high flatness, and good durability. The technical re- are used to test the energy output effect of the device.
quirements of the sidewall material are relatively low and should be the The electro-mechanical responses of the device with a steel plate
same as those of the materials of the top plate and bottom plate. The and aluminum plate are at almost the same level. The difference be-
shell preparation process of the device is shown in Fig. 2. tween the output voltages of the same energy-harvesting unit using
different plates is extremely small. The maximum difference in voltage
3.1.2. Fabrication of connecting components and substrate is only 0.4 V, which can be ignored in comparison with the output
The connecting components are located between the top plate and voltage (Table 1). Therefore, a steel plate and aluminum plate of the
the energy-harvesting unit. These components show good rigidity, same thickness have the same effect on the electro-mechanical re-
strength, and fatigue resistance. The steel material is selected with the sponses of the device. However, the aluminum plate is more suitable for
same diameter as the energy-harvesting unit and then processed into realizing the weight reduction of the device. Hence, the aluminum plate
equally tall columns to uniformly adjust the height by the Computer is used as the top plate of the power generation device.
Number Control (CNC). At the same time, the output voltages of the energy-harvesting units
The fiber insulation board has the function of heat insulation. The are uneven, and the output voltages of the #2 and #3 units are rela-
pile-type limit hole of the energy-harvesting unit is used to facilitate the tively low. This condition may be attributed to the degree of error be-
assembly of the energy-harvesting unit, and the diameter is 30.2 mm. tween the power generation devices after assembly or to the gap be-
The substrate is leveled by CNC so that the height of the hole is the tween the electro-mechanical characteristics of the units themselves.
same.
3.2.2. Different connecting components
3.1.3. Device assembly With its energy transfer function, the energy-harvesting unit must
The height of the energy-harvesting units and the connecting com- transfer the traffic load with the top plate of the device. The trans-
ponents are calibrated for the second time during the assembly of the mission characteristics of the intermediate connecting components play
power generation device to reduce the possible error at the 0.01 mm a key role in load transmission. The other parts of the device are kept
level between the energy-harvesting units and the connecting compo- unchanged. The electro-mechanical responses of the device are mea-
nents. The assembly process of the device is shown in Fig. 3. sured with ball-type and truncated cone-type connecting components.

Fig. 2. Shell preparation.

199
C. Wang et al. Energy Conversion and Management 163 (2018) 196–207

Fig. 3. Assembly process of the device.

3.2.3. Energy-harvesting units of different specifications

The electro-mechanical characteristics of the device with energy-
harvesting units of different specifications influence the energy output
effect of the device. The aluminum plate is used as the top plate, and the
electro-mechanical responses of the 6, 8, and 10 layers of the energy-
harvesting unit are tested. The output voltage of each energy-harvesting
unit is recorded, as shown in Fig. 6.
The output voltage at the same position decreases with the increase
in the number of layers of the energy-harvesting unit (Fig. 6). The
average output voltage of the energy-harvesting units of different spe-
cifications is as follows: U6 = 7.3 V; U8 = 6.35 V; U10 = 2.3 V. The en-
ergy-harvesting unit has an internal electrode, but the internal elec-
Fig. 4. Electro-mechanical response test of the device. trode penetrates into the piezoelectric ceramic layer after high-
temperature sintering. At this point, the energy-harvesting unit is
Table 1 equivalent to a monolith, and each layer produces the same micro-de-
Test data for electro-mechanical responses of the device of different plate types. formation. Hence, a thick energy-harvesting unit requires a high load
incentive to produce the same degree of voltage output. Then, the ca-
Energy-harvesting units #1 #2 #3 #4
pacitance values of the energy-harvesting units of different specifica-
Output voltage (V) Aluminum plate 6.8 4.8 4.0 6.0 tions are measured as follows: C6 = 137 nF; C8 = 193 nF. They are
Steel plate 6.8 4.6 4.2 5.8 substituted into the power calculation formula:
Δ (Difference) 0 0.2 −0.2 0.2
Q = CU (7)

where Q is the quantity of electric charge, C is the capacitance, U is the

voltage. The results are as follows: Q6 = 137 nF × 7.3 V = 1000
nC < Q8 = 193 nF × 6.35 V = 1226 nC. The eight-layer energy-har-
vesting unit exhibits excellent voltage output effect.
In summary, the optimal configuration of the components of the
proposed device is determined to be aluminum top plate + ball-type
connecting component + eight-layer energy-harvesting unit according
to the electro-mechanical responses of the device with different plates,
different connecting components, and energy-harvesting units of dif-
ferent specifications.

Fig. 5. No-load voltage of the energy-harvesting unit with different connecting compo-
nents.

The output voltages of the energy-harvesting units are recorded, as

shown in Fig. 5. The electro-mechanical responses of the device with a
ball-type connecting component are slightly better than those of the
device with a truncated cone-type component. The reason is that the
force of the ball-type connecting component is more likely to be con-
centrated when the device is crushed by rubber wheels. The force of the
truncated cone type is relatively dispersed because of the large force
area. Therefore, the output voltages of the capturing units are low. The
ball-type connecting component is adopted as the internal connecting
component of the device. Fig. 6. No-load voltage of the energy-harvesting unit with different specifications.

200
C. Wang et al. Energy Conversion and Management 163 (2018) 196–207

4. Power generation performance

Road traffic environments vary in different application environ-

ments, and the corresponding energy outputs also differ. Therefore, the
energy output characteristics of devices are studied with the MTS si-
mulation test system. The energy output effects of different devices are
tested with different series and parallel configurations of the energy-
harvesting units, different loads, and different frequency conditions.

4.1. Testing parameters

The road compressive stress and the frequency conditions produced

by the same device are different under different driving loads. For the Fig. 7. MTS simulation test.
compressive stress of the road, the vertical compressive stress value
corresponding to the depth of the asphalt pavement is about
0.2–0.7 MPa under static standard axle load [31], and the tire pressure
of a car is generally 0.2–0.3 MPa. Thus, the simulated test stress level is
set to 0.2–0.7 MPa to comprehensively evaluate the energy output ef-
fects of the device in different traffic environments.
For the frequency of driving load, there are two kinds of vehicle
load acting on the device: those approaching the device and those de-
parting the device. When the load is gradually approaching, the com-
pressive stress on the device is also gradually increased. When the load
is applied directly above the device, the compressive stress reaches the
maximum. When the load is departing the device, the compressive
stress gradually dissipates to zero. This process can be considered as a
half-wave sinusoidal load. The device is applied for only a short time
and the rolling frequency of the device can be determined by the fol-
lowing formula.
f = l/v (8)
Fig. 8. No-load voltage of energy-harvesting units with different loads.
where f is the driving rolling frequency, v is the driving speed, l is the
vehicle wheelbase.
The front and rear wheelbase of 2.5 m and travel speed of 80 km/h trucks but also to light traffic conditions dominated by cars.
are taken as typical driving characteristics, the frequency of the front The different output voltages of the energy-harvesting units are
and rear wheels is calculated to be about 10 Hz, and the pavement analyzed under different stress conditions. The output voltages of units
structure exhibits a certain vibration frequency of approximately 15 Hz #2, #3, and #4 are basically the same. By contrast, the output voltage
under vehicle load [32]. Therefore, the frequencies are set to 5, 10, and of unit #1 is significantly higher than that of the other units, and it
15 Hz to facilitate the simulation load of the device. increases gradually with the increase of stress level. This characteristic
The MTS Landmark electro-hydraulic loading servo system is used is attributable to two factors. First, the electro-mechanical character-
as the loading equipment to fit the road compression stress produced by istics of unit #1 are better than those of the other units. Second, the
traffic load. The area and the stress level of the device are converted overall height of unit #1 is slightly greater than that of the other units
into corresponding force values, and the loads are loaded in a half-wave after assembly. Hence, the force of unit #1 is greater than that of the
sine manner in the simulation test. other units, and this condition leads to a high output voltage. When the
positions of units #1 and #2 are reversed, unit #1 still shows a higher
output voltage compared with the other units, but the gap is reduced to
4.2. Electro-mechanical responses of energy-harvesting units
some extent. Meanwhile, unit #2 exhibits a certain degree of im-
provement. Therefore, the differences in output voltages for the units
We seek to determine the electro-mechanical responses of the en-
can be attributed to the two reasons.
ergy-harvesting units in the device when traffic loads and road struc-
tural characteristics are different. Therefore, different load traffic en-
vironments are simulated by changing the loading stress level. The 4.3. Power generation performance with different series and parallel
different stress levels of the device are converted to fixed force values configurations
according to the area of the device. Finally, the output voltage of each
energy-harvesting unit is tested by the MTS Landmark (MTS810) at the According to the principle of piezoelectric ceramics, the electric
loading frequency of 10 Hz (Fig. 7). The results of test are recorded and energy produced by piezoelectric ceramics is an energy form of a single
plotted in Fig. 8. pulse with little energy and exhibits the characteristics of general
All the energy-harvesting units in the power generation device show micro-electric energy with high voltage, low current, and small output
good electro-mechanical responses (Fig. 8). The output voltage of each energy level. Therefore, to evaluate the power generation capacity of
energy-harvesting unit increases linearly with the increase of load. In the device accurately, we accurately test the output voltage of the en-
particular, the maximum output voltage of energy-harvesting unit #1 is ergy-harvesting units and their power under different operating con-
24 V under the standard axial load of 0.7 MPa. The minimum output ditions. The energy output status of the device is subsequently eval-
voltage of units #2 and #3 is 15.2 V. The minimum output voltage of all uated.
energy-harvesting units can reach 4.2 V at a small stress level of
0.2 MPa. These values can meet the technical requirements of energy- 4.3.1. Output voltage
harvesting in terms of output voltage magnitude. Thus, the device can The energy-harvesting units in the device utilize different series and
be applied not only to heavy traffic conditions dominated by large parallel configurations according to different requirements of the

201
C. Wang et al. Energy Conversion and Management 163 (2018) 196–207

different parallel connections of the energy-harvesting units are

adopted. In the early stage, the amplitude is obvious. It then becomes
relatively gentle and finally tends to be stable and infinitely close to the
no-load voltage of energy-harvesting units. We find that when load
resistance is high, the output voltage reaches the maximum point as it
approaches circuit breaking.
Unlike the case shown in Fig. 8, the output voltage of the energy-
harvesting units in series is slightly lower than the sum of the output
voltages of the energy-harvesting units. Moreover, the output voltage of
all energy-harvesting units in parallel is slightly lower than the max-
imum value of their output voltages. The result is explained as follows.
First, a certain phase difference exists between the output voltages of
the energy-harvesting units, and the whole law of output voltage is
disturbed. Second, the force not unloaded by the MTS on the device
increases, thereby causing the output voltage to decrease. Despite the
Fig. 9. No-load voltage of energy-harvesting units with different series–parallel connec- above phenomena, the output voltages of the energy-harvesting units
tions. show that the two factors do not influence the effects of the ser-
ies–parallel connections of the energy-harvesting units.
electrical output. Theoretically, when the phases of energy-harvesting
units are the same, the output voltage of the energy-harvesting units
with a series configuration is the sum of the output voltages of all en- 4.3.2. Output power
ergy-harvesting units. The output voltage of the harvesting units with a Theoretically, the optimal output power of the device is propor-
parallel configuration is the maximum output voltage of each energy- tional to the number of parallel connections of the energy-harvesting
harvesting unit. units, and the optimal load is inversely proportional to the number of
As the output current of piezoelectric ceramic materials is low, a parallel connections of the energy-harvesting units. To further clarify
parallel connection is usually used in the device to output high power at the influence of the number of parallel connections on the output
minimal load resistance. Therefore, the output voltage is tested by power, we test the relationship between the output power and the load
changing the series–parallel connection of each energy-harvesting unit resistance of the different parallel modes of the energy-harvesting units
with the condition of 0.7 MPa–10 Hz. The experimental data are plotted when the load condition of 0.7 MPa–10 Hz remains unchanged. The
in Fig. 9. The relation between the output voltage and the load re- results are plotted in Fig. 11.
sistance of different energy-harvesting units in parallel connection As shown in Fig. 11, the output power of the energy-harvesting units
mode is shown in Fig. 10. increases with the increase of the number of parallel connections. The
As shown in Fig. 9, when the series connection is adopted, the peak point of the output power curve gradually shifts left with the in-
output voltage of the energy-harvesting units increases with the in- crease of the number of parallel connections, that is, the corresponding
crease in the number of energy-harvesting units. When the parallel optimal load resistance is gradually reduced.
connection is adopted, the output voltage of the energy-harvesting units The matching resistance of load size can be effectively reduced by
is not affected by the number of energy-harvesting units, and it is the parallel connections of the energy-harvesting units. Moreover, the
roughly stable at the same output level. Given the same number of more parallel the energy-harvesting units are, the smaller the optimum
energy-harvesting units, the output voltage in the series connection load is. The optimum load of the four energy-harvesting units in parallel
mode has multiple relationships with the output voltage in the parallel connection is 20 kΩ, which is 0.67 times that of the three energy-har-
connection mode. For example, if the four energy-harvesting units are vesting units in parallel connection, 0.4 times that of the two energy-
all connected in series, the output voltage is 67.8 V, which is 4.03 times harvesting units in parallel connection, and 0.17 times that of one en-
that in parallel connection (16.8 V). This result is consistent with the ergy-harvesting unit. When the output power is constant, a small op-
theoretical expectation. timum load equates to a good application of energy harvesting.
As shown in Fig. 10, the output voltage of the energy-harvesting Therefore, the four power harvesting units should be connected in
unit increases with the increase of the load resistance when the parallel to the device in the actual application.

Fig. 10. Relationship between output voltage and load resistance of energy-harvesting Fig. 11. Relationship between output power and load resistance of energy-harvesting
units with different parallel connections under 0.7 MPa–10 Hz. units with 0.7 MPa–10 Hz.

202
C. Wang et al. Energy Conversion and Management 163 (2018) 196–207

(a) Energy-harvesting unit #1 (b) Four energy-harvesting units in parallel

Fig. 12. Relationship curves between output voltage and load resistance with different loads. (a) Energy-harvesting unit #1. (b) Four energy-harvesting units in parallel.

4.4. Power generation performance with different loads output power of the four energy-harvesting units in parallel is 8.19 mW.
Thus, we can deduce that the power output of the energy-harvesting
When the device is applied to the power generation pavement, the units in parallel is considerable. This result shows that the power
traffic load varies, and the load characteristics of its output energy generation device can be potentially applied to urban traffic sections.
acquisition and storage circuit are different. To facilitate the research Further analysis shows that the output power of the energy-harvesting
on the energy output effect of the device and the design of the latter units in parallel is about four times greater than that of a single energy-
energy acquisition circuit, we must clarify the voltage output effect and harvesting unit.
power variation law with different load sizes and load resistance con- At the same time, the output power increases gradually with the
ditions. increase of load resistance, and the maximum power output is obtained
at a resistance value. Then, the output power decreases gradually with
4.4.1. Output voltage the increase of the load resistance. The whole change process verifies
When the output voltage is tested, the different resistances are the existence of an optimum load resistance to maximize the power
connected to the output end of each energy-harvesting unit, and the output of energy-harvesting units. Therefore, energy-harvesting units
oscilloscope is connected in parallel to both ends of the resistor. The should work at the optimal load in actual applications to obtain a re-
output voltage of the device is tested with different load conditions at latively good power output. In contrast to the illustration in Fig. 13(a)
the loading frequency of 10 Hz (Fig. 12). and (b), the optimum load for a single energy-harvesting unit is about
As shown in Fig. 12, the output voltage of the energy-harvesting 120 kΩ while the optimum load for the four energy-harvesting units in
units increases with the increase of the load resistance with different parallel is approximately 20 kΩ. Therefore, the parallel mode of the
loads. In the early stage, the amplitude is obvious. It later becomes energy-harvesting units can obviously reduce the best matching re-
relatively gentle and finally tends to be stable and infinitely close to the sistance. Such condition is convenient for practical applications.
no-load voltage of the energy-harvesting units. When the load re-
sistance is high, the output voltage reaches the maximum value as it 4.5. Power generation performance with different frequency conditions
approaches circuit breaking. At the same time, with the increase of load
level, the change trend of voltage is steep with the change of resistance The composition of road traffic load is complicated. Road level,
in the early stage. This condition indicates that the unit can output high vehicle type, lane width, number of lanes, and external environment
voltage with small load resistance. This case is conducive to the im- greatly impact the speed of traffic. The fluctuation of road vibration
provement of output power. frequency caused by these factors also falls in a certain range. Thus, the
output voltage characteristics and power of the device with different
4.4.2. Output power frequency conditions should be tested to clarify the correlation between
When the output power is tested, the device is also connected by output characteristics and vibration frequency.
external loads, and the output power is calculated indirectly according
to the terminal voltage recorded by the oscilloscope and the load re- 4.5.1. Output voltage
sistance. According to the curve of the output voltage, the load of a Existing research results show that the maximum output voltage can
single energy-harvesting unit, and the load of the four energy-har- be achieved at the resonant frequency of the energy-harvesting unit, but
vesting units in parallel under different loading conditions with a 10 Hz the frequency is usually 1000 Hz. By contrast, the road environment
loading frequency, the output power of the device is calculated with mostly involves low frequency or even ultra-low frequency vibration
different loads (Fig. 13). conditions. Hence, the output voltages of the device under 0.3 and
As shown in Fig. 13, the output power of a single energy-harvesting 0.5 MPa stress levels at 5, 10, and 15 Hz are tested. The test results are
unit and that of four energy-harvesting units in parallel increases with plotted in Fig. 14.
the increase of the load level. At the load level of 0.2 MPa, the max- As shown in Fig. 14, the output voltage of the energy-harvesting
imum output power of a single energy-harvesting unit is 0.15 mW, and unit rises with the increase of loading frequency in the frequency of
the maximum output power of the four energy-harvesting units in road traffic load. Under the frequency vibration of 5 Hz, the output
parallel is 0.88 mW. At the load level of 0.7 MPa, the maximum output voltage of the energy-harvesting unit is low. It then improves when the
power of a single energy-harvesting unit is 1.83 mW, and the maximum frequency vibration is 15 Hz. The ideal electric energy can thus be

203
C. Wang et al. Energy Conversion and Management 163 (2018) 196–207

(a) Energy-harvesting unit #1 (b) Four energy-harvesting units in parallel

Fig. 13. Relationship curves between output power and load resistance with different loads. (a) Energy-harvesting unit #1. (b) Four energy-harvesting units in parallel.

obtained when the speed is high.

Compared with the increase of load frequency, the load magnitude
of the device has greater impact on the output voltage of the energy-
harvesting units. When the load frequency is increased from 5 Hz to
15 Hz, the output voltage of the energy-harvesting units increases by
2.7 V maximally. When the stress level is increased from 0.3 MPa to
0.5 MPa, the average maximum output voltage increases by 6.6 V, and
the output voltage is almost doubled at 0.3 MPa when the mean value is
0.5 MPa. This result shows 15 Hz as the limit of speed. Moreover,
0.5 MPa is still far from the high road stress level. Therefore, the stress
level of the device at the time of laying should be improved. It can be
laid in a traffic environment or a pavement structure layer having a
high stress level to improve the energy effect of the device.
The same test method described above is used to perform the output
voltage test of the energy-harvesting units. The results of the test are Fig. 15. Relationship curves between output voltage and load resistance with different
shown in Fig. 15. frequency conditions (four energy-harvesting units in parallel).

As shown in Fig. 15, the output voltage of the energy-harvesting

units increases with the increase of load resistance under different pattern with the increase of the number of energy-harvesting units in
frequency conditions. The output voltage increases sharply in the early parallel (Fig. 10) and the increase of frequency (Fig. 15). This result
stage and becomes stable and close to the no-load voltage of the energy- indirectly indicates that high frequency and a large number of energy-
harvesting units. When the load resistance is high, the output voltage harvesting units in parallel are favorable for the output power of the
reaches the maximum value as it approaches circuit breaking. Com- device.
paring Figs. 10 and 15 show that the output voltages exhibit the same

Fig. 14. Average values of no-load voltage of energy-harvesting units with different frequencies: (a) 0.3 MPa. (b) 0.5 MPa.

204
C. Wang et al. Energy Conversion and Management 163 (2018) 196–207

Fig. 16. Relationship curves between output power and load resistance with different
frequency conditions (four energy-harvesting units in parallel).

4.5.2. Output power Fig. 17. Force value of dynamic load.

The output power of the four energy-harvesting units in parallel is
measured at frequencies of 5, 10, and 15 when the loading force of
4.6. Durability test
0.7 MPa is kept unchanged. The results of the test are recorded and
shown in Fig. 16.
There are two specific problems during implementation. First, the
As shown in Fig. 16, the output power of the four energy-harvesting
structure of the device could be damaged under overweight load, and
units in parallel increases with the increase of load frequency, and the
the excessive deformation of the device could cause damage to the road
peak value of the output power shifts to the left. This condition implies
structure in the process of construction or utilization. Second, the pie-
that the optimal load size gradually decreases. Hence, the increase of
zoelectric properties of the energy-harvesting units could be attenuated
frequency can effectively reduce the matching load resistance. The
under a large amount of load in the working period of several years. To
power output of the four energy-harvesting units in parallel reaches
carry out further research on these two problems, we apply the MTS test
11.67 mW at 0.7 MPa–15 Hz, and the corresponding best load is 10 KΩ,
system to simulate the case with a long time period and large load. The
which is 3.1 times of the output power and 0.25 times of the load re-
dynamic load is set to 0.7 MPa stress level (converted to 7 kN according
sistance at 0.7 MPa–5 Hz. The result shows that the device is suitable for
to the area), and the number of cycles is 40,000 (Fig. 17). The degree of
environments involving high speed driving.
deformation of the device during the test is shown in Fig. 18.
In summary, we conduct a correlation analysis of the power gen-
The whole deformation of the device is about 0.15 mm with the load
eration performance of the device. The maximum output power of the
level of 0.7 MPa. The deformation is less than that between the upper
device with different load conditions is shown in Table 2.
and lower layers of the pavement with the same load condition
The power of the 100 mm × 100 mm-type device increases with the
(0.5 mm), and thus, it will not damage the pavement structure (Fig. 18).
increase of load and frequency. The maximum output power is 0.88
At the same time, the whole deformation of the device is always stable
mW, and the optimum load is 20 kΩ with 0.2 MPa stress. Therefore, the
within 40,000 loading cycles. Hence, the components of the device are
power generation device in city traffic environment can achieve col-
not damaged by fatigue, and the device achieves great working stability
lection and application levels. When the load level increases to
and durability.
0.7 MPa–15 Hz, the maximum output power of the power generation
To determine the attenuation of the piezoelectric properties of the
device is 11.67 mW, and the corresponding load decreases to 10 kΩ. At
energy-harvesting units under long-term load, we study the electrical
this point, the power generation device achieves maximum power
capacities of the energy-harvesting units before and after the test. Their
generation capacity. However, there are variations with the theoretical
values are 193.6 and 192.9 nF, and the difference between them is only
value. The device is a carrier device that uniformly transmits the load of
0.7 nF, which is almost negligible. Therefore, we consider that the at-
the top plate to the internal energy-harvesting unit, And the mechanical
tenuation of piezoelectric properties does not occur in the energy-har-
loss during mechanical transmission leads to a reduction in the stress
vesting units. In this case, we can deduce that the energy-harvesting
transmitted partially, resulting in a difference between the power
units exhibit excellent durability and can achieve a stable output of
generation performance of the energy-harvesting unit and its theore-
tical performance. Therefore, how to reduce the mechanical loss that
make the device achieve the optimal power generation effect will be
our team's next plan.

Table 2
Maximum output power of 100 mm × 100 mm-type device with different load cases (four
energy-harvesting units in parallel).

Load case 0.2 MPa 0.5 MPa 0.6 MPa 0.7 MPa 0.7 MPa 0.7 MPa

10 Hz 5 Hz 15 Hz

Maximum output 0.88 4.61 6.27 8.19 3.72 11.67

power (mW)
Optimal load (kΩ) 20 20 20 20 40 10
Fig. 18. Deformation of the device.

205
C. Wang et al. Energy Conversion and Management 163 (2018) 196–207

Table 3
Comparisons of typical piezoelectric power generation technology.

Device (Organization) Advantage Disadvantage

Innowattech Its piezoelectric power generation technology was first applied to power There is no published design and implementation methodology for this
generation systems of pavement and the single impulse compression technology, and piezoelectric products made by the piezoelectric ceramic
energy generation system was designed. It is reported that its power supplier Channel Technologies face the challenge of road strength and
generation is about 100–200 kWh/km [14,15], which has a better power durability [33]
generation effect
Virginia Tech Piezoelectric power generation device was designed independently funded The output power of harvester does not meet the supplier's expectations
by the US DOT, which can produce 400–700 V with the 4000 veh/d heavy [33], and the technique did not consider the mechanical advantage that can
traffic, and an energy-harvesting unit can generate 0.2–0.35 mA [34] change the force pattern of the energy-harvesting unit
Genziko A resonant or persistently vibrating energy-harvesting technology was Functional piezoelectric materials that convert vibrational energy or stress
designed [35]. That is, the device continues to output electric energy after into electrical energy are not disclosed, and the actual pavement
the initial wheel impact, and the device has better power generation environment usually has inconsistent or different loads or frequencies,
performance which severely constrain the road's practicability [12]
The Research Device The device for pavements is a self-designed product. The design process At present, the device is in the laboratory test phase and has not been
takes into account the electro-mechanical conversion of the energy- applied in a large scale yet. And like other typical devices, the energy
harvesting units and the structural characteristics of roads, the load conversion efficiency, long-term stability of the device in complex
characteristics (the lateral distribution characteristics of the wheel path, pavement environment still need to be further improved. And they are
tire ground characteristics, and driving rolling conditions) and other exactly the next contents
factors. The performance research of the device is based on the real road
environment with different load and frequencies. And the device has
application feasibility

electric energy under long-term frequent load. frequency conditions is defined. The durability of the device is also
tested, and future research plan is discussed. The main conclusions are
5. Comparison analysis summarized as below.

Thus far the energy output effect of the device is analyzed system- (1) Based on power generation theory of energy-harvesting units, the
atically under different loads and frequencies conditions. The result structure of the power generation device for pavements is designed
shows that the device offers prospective application in research and based on piezoelectric energy-harvesting units and with con-
practice. However, it has not been applied in road paving. Therefore, sideration of the electro-mechanical conversion principles of the
the technical characteristics of typical piezoelectric power generation energy-harvesting units and the structural characteristics of roads.
devices are compared and analyzed (Table 3) in order to develop future The size of proposed device is optimized for driving load char-
research plan. acteristics on pavements.
As shown in Table 3, the typical piezoelectric power generation (2) The manufacturing process of the unitized device is detailed, and
devices or technologies have shown good power generation effect, but the fabrication and assembly of the device are carried out. Based on
some of the energy-harvesting units are not efficiently matched with the the electro-mechanical responses of different combinations of the
road environment, and some of the tests were completed in the la- device, the optimal internal configuration of the device is de-
boratory only and have not been tested for road feasibility. And they termined to be aluminum top plate + ball-type connecting com-
severely restricted road practicality of the device. ponent + eight-layer energy-harvesting unit.
Although the research device has not yet been put into practice, the (3) The energy-harvesting units within the device show good electro-
original design process has fully considered the electro-mechanical mechanical responses. The output voltage in the series connection
conversion of the energy-harvesting units and the structural char- mode has multiple relationships with the output voltage in the
acteristics of roads, the load characteristics (the lateral distribution parallel connection mode in the same number of energy-harvesting
characteristics of the wheel path, tire ground characteristics, and units. The output power of the energy-harvesting units increases
driving rolling conditions) and other factors. In addition, the perfor- with the number of parallel connections. Four energy-harvesting
mance of the power generation device is evaluated based on the real units should be used in parallel in actual applications.
road environment with different load and frequencies. (4) The output voltage and maximum output power of the energy-
Future research plans include performing energy conversion more harvesting units are positively correlated to the load size and fre-
efficiently, long-term stability, maintenance of the device in complex quency. The corresponding optimum load resistance decreases with
pavement environment. It is also desirable to pave pilot road to test the the increase of loading frequency and number of parallel units. The
energy output effect of the power generation pavement, and monitor maximum output power of the 100 mm × 100 mm-type device can
the power generation performance, safety performance and durability reach up to 11.67 mW when the four energy-harvesting units are
of power generation pavement with the wireless communication tech- connected in parallel. The corresponding optimal load is 10 KΩ. In
nology. Lastly, the smart power generation pavement should be im- this regard, the device show good collection and utilization pro-
plemented through smart grid control technology, especially in urban spect.
roads, highways, toll station service areas and other occasions. (5) The working durability of the device is evaluated by means of MTS.
The device is tested under 40,000 load cycles at the dynamic
loading of 0.7 MPa. No change is observed in the capacitance of the
6. Conclusions energy-harvesting units before and after the test, thereby indicating
the excellent durability of the device.
We design a piezoelectric power generation device for pavements (6) Feasibility of the application device is studied, and future research
with consideration of energy output and pavement coupling. The fab- plans and directions are discussed.
rication and assembly of the device are carried out, and the electrical
output characteristics of the device are studied systematically. The
energy output effect of the device with energy-harvesting units of dif-
ferent series and parallel configurations, different loads, and different

206
C. Wang et al. Energy Conversion and Management 163 (2018) 196–207

Acknowledgement 2016;112:246–53.
[15] www.innowattech.co.il.
[16] Lee JY, Lee MK, Oh JG, et al. Study on the energy conversion from the dynamic load
This paper reports the research activities mainly requested and of vehicles on the road using piezoelectric materials. Mater Sci Forum 2010:57–60.
sponsored by the International Cooperation and Exchange Plan of [17] Papagiannakis AT, Dessouky S, Montoya A, et al. Energy harvesting from roadways.
Science and Technology in Shaanxi Province of China (Program No. Proc Comput Sci 2016;83:758–65.
[18] Jiang XT, Li YC, Li JC, et al. Piezoelectric energy harvesting from traffic-induced
2017KW-033) and the Fundamental Research Funds for the Central pavement vibrations. J Renew Sustain Energy 2014;6(4):977–86.
Universities (Program No. 310821162013). That sponsorship and in- [19] Jiang XT, Li YC, Wang J, et al. Electromechanical modeling and experimental
terest are gratefully acknowledged. analysis of a compression-based piezoelectric vibration energy harvesting unit. Int J
Smart Nano Mater 2014;3(5):152–68.
[20] Roshani H, Dessouky S, Montoya A, et al. Energy harvesting from asphalt pavement
References roadways vehicle-induced stresses: a feasibility study. Appl Energy
2016;182:210–8.
[21] Wang CH, Chen S, Li YW, et al. Design of piezoelectric elements' protection mea-
[1] Kim HS, Kim JH, Kim J. A review of piezoelectric energy harvesting based on vi-
sures and energy output of intelligent power pavement. China J Highw Transp
bration. Int J Precis Eng Manuf 2011;12(6):1129–41.
2016;29(5):41–9.
[2] Fan KQ, Chao FB, Zhang JG, et al. Design and experimental verification of a bi-
[22] Wang CH. An energy harvesting device for improving the power effect used on
directional nonlinear piezoelectric energy harvesting unit. Energy Convers Manage
pavement. CN patent no. 201621011448.4; 30 August 2016.
2014;86(10):561–7.
[23] Dang Y, Dong W, Bai F. Performance of piezoelectric stack for energy harvesting
[3] Guan M, Liao WH. Design and analysis of a piezoelectric energy harvesting unit for
from low frequency vibration. J Power Supply 2011;9(4):56–62.
rotational motion system. Energy Convers Manage 2016;111:239–44.
[24] Platt SR, Farritor S, Haider H. On low-frequency electric power generation with PZT
[4] Cook-Chennault KA, Thambi N, Sastry AM. Powering MEMS portable devices - a
ceramics. IEEE/ASME Trans Mechatron 2005;10(2):240–52.
review of non-regenerative and regenerative power supply systems with emphasis
[25] Liu LP, Sun LJ. Research on wheelpath lateral distribution for freeway asphalt
on piezoelectric energy harvesting systems. Smart Mater Struct 2008;17(4):1240–6.
pavements. J Tongji Univ Nat Sci 2005;33(11):1449–52.
[5] Saadon S, Sidek O. A review of vibration-based MEMS piezoelectric energy har-
[26] Hu XD, Sun LJ. Measuring tire ground pressure distribution of heavy vehicle. J
vesting units. Energy Convers Manage 2011;52(1):500–4.
Tongji Univ Nat Sci 2005;33(11):1443–8.
[6] Lee J, Choi B. Development of a piezoelectric energy harvesting system for im-
[27] Anghelache G, Moisescu R, Sorohan S, et al. Measuring system for investigation of
plementing wireless sensors on the tires. Energy Convers Manage 2014;78:32–8.
tri-axial stress distribution across the tyre–road contact patch. Measurement
[7] Wang CH, Zhao JX. Development and prospect of intelligent piezoelectric power
2011;44(3):559–68.
generation pavement technology. Road Mach Constr Mech 2017;34(5):19–24.
[28] Taghavifar H, Mardani A. Potential of functional image processing technique for the
[8] Tan YQ, Zhong Y, Lv JF, et al. Preparation and properties of PZT/asphalt-based
measurements of contact area and contact pressure of a radial ply tire in a soil bin
piezoelectric composites used on pavement. J Build Mater 2013;16(6):975–80.
testing facility. Measurement 2013;46(10):4038–44.
[9] Wang CH, Wang HL, Li YW, et al. Study on technology of power pavement based on
[29] Liang C, Wang G, An DF, et al. Tread wear and footprint geometrical characters of
integration of piezoelectric material and pavement material. J Highw Transp Res
truck bus radial tires. Chin J Mech Eng 2013;26(3):506–11.
Dev 2016;33(11):14–9.
[30] Tielking JT, Abraham MA. Measurement of truck tire footprint pressures; 1994.
[10] Zhao H, Ling J, Yu J. A comparative analysis of piezoelectric transducers for har-
[31] Huang RB, Niu YL, Zhao HD, et al. Technical approach and research prospect of
vesting energy from asphalt pavement. J Ceram Soc Jpn 2012;120(1404):317–23.
piezoelectric energy harvest from highway. China J Highw Transp 2012;25(6):1–8.
[11] Najini H, Senthil AM. Investigation on the selection of piezoelectric materials for
[32] Xiong H, Wang L, Wang D, et al. Piezoelectric energy harvesting from traffic in-
the design of an energy harvesting unit system to generate energy from traffic. Int J
duced deformation of pavements. Int J Pavement Res Technol 2012;5(5):333–7.
Eng Appl Sci 2016;3(2):43–9.
[33] Hill D, Agarwal A, Tong N, et al. Assessment of piezoelectric materials for roadway
[12] Xiang HJ, Wang JJ, Shi ZF, et al. Theoretical analysis of piezoelectric energy har-
energy harvesting cost of energy and demonstration roadmap - final project report.
vesting from traffic induced deformation of pavements. Smart Mater Struct
California Energy Commission; 2014.
2013;22(9):095024.
[34] Xiong, et al. New Technologies for Development of Renewable Energy in the Public
[13] Kim S, Shen J, Ahad M, et al. Piezoelectric energy harvesting system assessment for
Right - of -Way. DTFH61-10-C-00016. FHWA 9th Quarterly Report. Virginia Tech;
highway sustainability. In: 52nd ASC annual international conference proceedings,
2012.
Utah: the Associated Schools of Construction; 2016.
[35] Genziko RPG product brochure. Technical data; 2012.
[14] Moure A, Izquierdo M, Rueda S. Feasible integration in asphalt of piezoelectric
cymbals for vibration energy harvesting. Energy Convers Manage

207