Modeling a new energy harvesting pavement system with experimental verification

Applied Energy xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Applied Energy journal homepage: www.elsevier.com/locate/apenergy

Modeling a new energy harvesting pavement system with experimental verification ⁎

Lukai Guo , Qing Lu Department of Civil and Environmental Engineering, University of South Florida, 4202 E. Fowler Avenue, Tampa, FL 33620-4202, United States

H I G H L I G H T S energy harvesting pavement system is built using conductive asphalt mixtures and piezoelectric materials. • AA new electromechanical model is used to analyze and optimize the new system. • Thethree-degree-of-freedom of this new system is verified by laboratory tests. • The feasibility • maximum electrical power from the new system can reach to 300 mW.

A R T I C L E I N F O

A B S T R A C T

Keywords: Piezoelectric Energy harvesting pavement Multi degree of freedom electromechanical model Laboratory test

A novel design of an energy harvesting pavement system (EHPS) is introduced in this paper. The basic concept behind this design is to transform asphalt layers into a piezoelectric energy harvester to collect dissipated vehicle kinetic energy in a large-scale system. This EHPS design consists of two conductive asphalt layers and one piezoelectric material layer. To verify the feasibility of the design, this ongoing study theoretically analyzed the EHPS via a three-degree-of-freedom electromechanical model and practically tested a prototype in the laboratory. As a result, voltage outputs measured in the laboratory from the prototype design matched those estimated from the electromechanical model. Through testing the effects of several components in the EHPS on electricity generation, this study confirms that using more flexible conductive asphalt mixtures and arranging more piezoelectric elements with a higher piezoelectric stress constant can increase electrical outputs from the EHPS. Regarding specific external vibration conditions, a high frequency of external vibration can lead to a dramatic effect of each piezoelectric element’s capacitance on increasing electrical outputs, but also can reduce the benefit from adding more piezoelectric elements to produce higher electrical outputs. After optimizing this EHPS prototype by adding more piezoelectric elements with higher piezoelectric stress constant and improving the flexibility of conductive asphalt mixtures, the maximum electric power from the proposed EHPS can be increased from approximately 1.2 mW to 300 mW under a high frequency (30 Hz) external vibration. The levelized cost of electricity of this EHPS can be $19.15/kWh on a high-volume roadway within a 15-year service life.

1. Introduction One potentially important component for energy harvesting in the transportation sector is pavement. On the six million-km roadways in the U.S., a huge amount of vehicle kinetic energy is wasted every day due to rolling resistance and vehicle vibration [1–3]. Given that such wasted energy is collectible and convertible into electrical energy through different techniques [4,5], an advanced pavement system may be turned into a new “energy farm” to reduce society’s need for coal consumption for electricity production. Photovoltaic (PV) panels, pipe systems, thermoelectrical generators,

⁎

piezoelectric transducers, and pyroelectric materials recently have been studied and developed for collecting wasted energies from pavement [4]. Most of these, however, either are impractical or have limited efficiency. For example, the performance of PV panels paved on the surface of roadways can be challenged by the abrasion and contamination on its transparent layer [6] and also can be highly dependent on sunlight [7]; the pipe system that stores thermal energy can significantly affect pavement internal structures [8–10]; the efficiency of thermoelectrical generator based on the Seebeck effect is relatively low, especially under the constraint of insufficient temperature difference between asphalt layers in a traditional asphalt pavement [11]; and

Corresponding author. E-mail addresses: [email protected] (L. Guo), [email protected] (Q. Lu).

http://dx.doi.org/10.1016/j.apenergy.2017.09.045 Received 23 March 2017; Received in revised form 22 August 2017; Accepted 10 September 2017 0306-2619/ © 2017 Elsevier Ltd. All rights reserved.

Please cite this article as: Guo, L., Applied Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.09.045

Applied Energy xxx (xxxx) xxx–xxx

L. Guo, Q. Lu

rectifiers with diodes were connected with energy harvesters to avoid neutralization of the opposite voltages generated from the piezoelectric disks. To ensure sufficient damage resistance and electrical outputs of the piezoelectric disks under heavy traffic loads in the field, laboratory tests were conducted prior to the field installation [7,23]. They collected around 116 mW instant power output with 3.1 mW average power output from each energy harvester. However, the results from their field tests at the Troutville weigh station also showed significant degradation of power generation from the energy harvesters over one year after the installation [7,23]. As can be seen, the main concepts in the previous studies were limited by embedding prefabricated piezoelectric transducers into pavement to generate electric power, which can power only sensors and some low-power infrastructure utilities. To maintain a traditional layer structure of asphalt pavement while exploring efficient ways of energy harvesting from pavement, this study introduces a new system into asphalt pavement to collect kinetic energy from traffic. Compared to the piezoelectric energy harvesting methods designed in previous studies, which embed pre-fabricated piezoelectric harvesters into pavement, the approach proposed in this study may transform an entire pavement layer into a piezoelectric energy harvester to collect more dissipated kinetic energy on a larger scale. A prototype of a new energy harvesting pavement system (EHPS) was first designed in this study. A multi-degree-of-freedom electromechanical model (MDOF-EM) was then built for the EHPS. To verify the feasibility of the EHPS design and the accuracy of the MDOF-EM, a series of laboratory tests were conducted. After ensuring the accuracy of the MDOF-EM, the effect of each component of the EHPS on the electrical output is discussed based on the MDOF-EM to optimize the proposed EHPS.

the low electrical output from pyroelectric materials inside pavement layers remains an issue [12]. Among all energy harvesting techniques, the application of piezoelectric transducers on producing electricity from pavement seems promising and has been developed in the last 10 years. In 2010, a study was conducted at the University of California Pavement Research Center (UCPRC) to analyze the electric power generated from a cymbal transducer embedded in pavement using the finite element method (FEM). That study estimated that a 32 mm diameter cymbal transducer can produce 1.2 mW electrical energy under a 20 Hz vehicle load [13]. In 2012, Yao et al. modified a trapezoidal bridge transducer to an arc bridge transducer as an energy harvester in the pavement. Based on FEM, they found that the latter generated 232 V electric voltage under a pressure of 0.4 MPa, which had 78 V more from a trapezoidal bridge transducer [14]. Li verified this finding but stated that an arc bridge piezoelectric transducer may have a shorter service life due to higher stress [15]. In the same year, Kim et al. installed piezoelectric cantilevers as speed bumps above ground and compared them with the piezoelectric cantilevers embedded underground in the field. They captured a significant drop in electrical output from the piezoelectric cantilevers after moving them from underground to above ground (7.6 mW versus 63.9 mW) and also observed that if vehicle speed exceeded 20 km/h, more piezoelectric cantilevers were able to generate more electric power [16]. In 2013, several studies on using piezoelectric transducers to produce electricity from pavement were performed. Sun et al. modified the size of a piezoelectric transducer to 280 × 280 × 20 mm and suggested embedding it in pavement at a depth of 40 mm. Based on FEM, they estimated that the electrical output from their design could reach to 1.8 mW [17]. Daniel et al. simulated finite element models with different external factors (traffic load and electric load) and internal factors (geometry and material) and evaluated that 1.2 mW electric power under a dynamic load with 50 N amplitude and 2 Hz frequency can be produced from their metal-protected cymbal design [18]. In the same year, Xiang et al. developed a pavement dynamic deformation model using Fourier transform and Cauchy’s residue theorem to better analyze the effect of pavement structure on the electric power output from a piezoelectric transducer inside pavement. Their results demonstrated that vehicle speed and pavement damping property can influence the electric power output from the piezoelectric transducer [19]. In 2015, Zhang et al. updated the system model by replacing a beam model with a Kirchhoff plate model for pavement. They found that the wheel load can only trigger the transducers within a distance of 4 m, and the instantaneous electric power from each transducer can reach 47.3 mW under a four-wheel load [20]. In 2016, Roshani et al. built an energy harvester using piezoelectric disks sandwiched by two copper plates. They set two polystyrene sheets to fix the piezoelectric disks and glued them onto the copper plates using electrically-conductive epoxy containing silver adhesive. Based on a power level of around 10 mW from their energy harvester, their study confirmed the feasibility of using such energy harvester to activate multiple sensors and power LED traffic lights in a roadway infrastructure system [21]. In 2017, regarding the brittle property of piezoelectric ceramic materials, Jung et al. replaced lead zirconate titanate (PZT) ceramic with polyvinylidene fluoride (PVDF) film to develop a flexible piezoelectric polymer-based device for harvesting energy in a roadway [21]. An instantaneous power output of up to 200 mW was generated from their energy harvester module under a traffic load of 2450 N at a speed of 8 km/h [22]. For field tests, Virginia Tech demonstrated six prefabricated energy harvesters in the pavement at the I-81 Troutville weigh station [7,23]. To prevent potential damage to the brittle piezoelectric materials, the stress on piezoelectric disks was analyzed first by FEM simulation and controlled by adjusting the shape and spacing of the disks. Multiple

2. Design of an EHPS in asphalt pavement The concept for developing an EHPS is basically trying to “transform pavement layers into a massive piezoelectric transducer,” as shown in Fig. 1. This new system consists of two conductive asphalt layers and one piezoelectric material layer. The two conductive asphalt layers are made of conductive asphalt mixtures with conductive additives, such as steel wool, graphite, and carbon black. The piezoelectric layer integrates piezoelectric materials with asphalt mixtures and other insulation materials. When the piezoelectric layer is vibrated, each piezoelectric cell is polarized because of the deviations of positive and negative charge centers (piezoelectric effect). Then, the piezoelectric layer will charge the upper and lower conductive layers. Since the piezoelectric layer itself is insulated, electric charges will be stored in the upper and lower layers without leakage. The EHPS assembles conductive asphalt mixtures, regular asphalt mixtures, and piezoelectric materials into one large-scale energy harvester for the first time. Three basic advantages of this EHPS design are as follows:

• The cover area of the EHPS can be adjusted to a large extent to • •

collect more dissipated kinetic energy than traditional piezoelectric transducers. Embedding conductive asphalt layers instead of inserting metal panels into the pavement structure may have less impact on the original pavement performance. Designs of the piezoelectric layer and the conductive asphalt layer can be customized by choosing different sizes (from large to powder size) and shapes (e.g., pile, ball, and roof) of piezoelectric elements and selecting proper aggregates and conductive fillers (e.g., steel fiber, graphite, and carbon), respectively.

The most important rule behind this EHPS design is that the upper and bottom conductive asphalt layers should never be connected; if connected, the electric power produced from this EHPS will be consumed under a short circuit condition. Since the piezoelectric elements 2

Applied Energy xxx (xxxx) xxx–xxx

L. Guo, Q. Lu

Fig. 1. Concept for transforming piezoelectric transducer into EHPS.

segment of conductive asphalt layer are parallel-connected. Then, those segments, which can be regarded as several huge piezoelectric transducers can be connected in a series to produce higher voltage for less energy consumption during transmitting. Given that the material properties of all materials in the EHPS are close to those of traditional pavement materials, the above modifications will not dramatically affect the structural performance of the EHPS. Based on the structure of this EHPS prototype, in addition to generating a considerable amount of electric power, the following additional functions may be potentially utilized in real applications:

and asphalt mixtures in the piezoelectric layer are all insulation, this EHPS design will perform well in an ideal condition. However, this rule can be challenged by the potential environment impact from rainwater. If rainwater flows into the voids inside the piezoelectric layer, it may become conductive. To overcome this issue, two countermeasures may be taken:

• Use very dense asphalt mixtures to limit voids inside the piezo-

•

electric layer. This strategy should work better if the piezoelectric layer is thicker. Meanwhile, the area using conductive materials in the conductive layer can be shrunk by paving regular asphalt mixtures on the layer edge. This adjustment can avoid the connection between the two conductive layers by the rainwater emerging through a drainage component. Pave or spray a thin waterproof layer inside the piezoelectric layer. Since the piezoelectric disk is well waterproofed, the waterproof layer is needed only to block the connection of voids inside the asphalt mixtures. The location of the waterproof layer depends on the conductivity of the waterproof material itself—if the waterproof material is conductive, it can be sprayed on the interface between the piezoelectric layer and the conductive layer; if the waterproof material is insulative, it can be inserted only in the middle of the piezoelectric layer without cutting the piezoelectric elements and their connection with the conductive layer.

• The EHPS may provide a new way of deicing pavement in the • •

Another challenge of the EHPS is to avoid the offset of the positive charges and negative charges generated within such a large area of the EHPS. On one hand, the vibration of each individual vehicle can generate alternating voltage from the piezoelectric materials, and any negative charges collected on one conductive layer can be quickly neutralized by positive charges. As a typical solution, one diode can be added on each output wire, as only negative or positive charges can be transmitted through an output wire to the external power storage equipment (e.g., capacitor). On the other hand, since there are several vibration sources (vehicles) within an adjacent area of the EHPS, positive charges and negative charges can be produced and can offset each other simultaneously. To better mitigate the interaction between vehicles which may weaken the performance of the EHPS, regular asphalt mixtures can be paved to cut the EHPS into several segments in the longitudinal direction. The specific length of each EHPS segment can be set based on the local traffic condition (e.g., traffic average speed, traffic volume, average axle distance of vehicles). An additional advantage of dividing the EHPS into several segments is to better integrate the electric power from the EHPS. All piezoelectric elements under one

winter—if the conductive layers are connected each other, such a short circuit can consume the electric power to form thermal energy and directly heat the pavement. Since the electricity generated from the EHPS can be collected as a signal to measure the speed and weight of a vehicle, the EHPS can also be regarded as a large-scale sensor system. The EHPS can be self-healing—micro cracks developing in the conductive layer can affect the conductivity of conductive layer and then be detected by a slight change in the electrical signal. Once the conductive layers are connected each other, the cracked component can be heated to heal micro cracks by melting asphalt.

3. Electromechanical model of EHPS prototype To theoretically analyze the effects of material properties of conductive asphalt mixtures and piezoelectric elements on electricity generation from the EHPS, it was simplified into a three-degree-offreedom system—a harmonic excitation (representing traffic loads) was applied on the first conductive layer and then was transmitted through the piezoelectric layer to the second conductive layer, as shown in Fig. 2. Mechanical property parameters of the conductive asphalt concrete layers include masses (ma and mb), damping coefficients (ca and cb), and spring constants (ka and kb). The spring constants of the asphalt mixtures (ka and kb) can be estimated using Eq. (1) with the resilient modulus of asphalt mixtures (Mr), the cross section areas of asphalt mixtures (Aa and Ab), and the thicknesses of asphalt mixtures (Ta and Tb). The damping coefficients of asphalt mixtures can be estimated using Eq. (2) with the damping ratio ζ, spring constants ka and kb, and the masses of asphalt mixtures ma and mb. 3

Applied Energy xxx (xxxx) xxx–xxx

L. Guo, Q. Lu

Fig. 2. Three-degree-of-freedom model of EHPS prototype.

ki =

Mr Ai Ti

(i = a,b)

ci = 2ζ ki mi

(i = a,b)

(2)

Q = Ap e33

Tp

(8)

R

Ap e33 Tp

(9)

is based only on the material properties of the piezo-

electric element,

Ap e33 Tp

can be regarded as an electromechanic coupling

coefficient P. Then, the governing equations of this three-degree-offreedom electromechanical model can be written as:

⎧ ma x¨a + ka (x a−x p−xb) + ca (x ȧ −x ṗ −xḃ ) = Q0Cos (ωt ) ⎪ mp x¨p + nkp (x p−xb) + nPVp = Q0Cos (ωt )−ma x¨a ⎪ nkp (x p−xb) + nPVp = mb x¨b + cb xḃ + kb xb ⎨ ⎪ Vp ⎪ − nPx ṗ + nCp Vṗ + R = 0 ⎩

(10)

The above multi-degree-of-freedom Eq. (10) can be transformed to the following matrix equations, Eqs. (11) and (12):

(4)

Vp

Vp

Vp d d Vp + = n Q (t ) R dt dt Since

xp

Ap e33

(7)

nCp

By substituting Eq. (4) into Eq. (3), the induction force Fe can be expressed by Eq. (5):

Fe = −

D3 = d33 σ3 + ε33 E3

∫A D3 dA) =

(3)

TP

(6)

d ( dt

Since the kinetic energy is dissipated mainly from the work done by induction force inside the piezoelectric elements, the induction force Fe is used to calculate the amount of energy conversion from mechanical energy into electrical energy, as expressed in Eq. (3) with the displacement of the piezoelectric element (xp), the electric potential difference between the opposite surfaces of piezoelectric element (Vp), and the charge generated on the surface of piezoelectric element (Q). Meanwhile, the amount of charge Q under external loads can be estimated using Eq. (4) from the piezoelectric stress constant (e33), the cross section area of piezoelectric element (Ap), and the vertical strain of piezoelectric element (xp/Tp). The piezoelectric stress constant e33 represents a coefficient for electric displacement divided by vertical strain.

Fe x p = −Vp Q

εr ε0 AP Tp

Cp =

(1)

⎡1 ⎢0 ⎢0 ⎢0 ⎢0 ⎢ ⎢0 ⎢0 ⎢ ⎣0

(5)

The electronic model of the EHPS circuit is displayed in Fig. 3, where Cp is the capacitance of each piezoelectric element as seen in Eq. (6), Ri is the insulation leakage resistance of the whole EHPS, and Rl is the external load resistance. Based on the electric displacement field D3 in each piezoelectric element (Eq. (7)) and the Gauss law (Eq. (8)), Eq. (9) was developed to calculate the electric potential Vp generated in this overall electric circuit. In Eqs. (6)–(9), εr denotes the relative static permittivity of the piezoelectric element, ε0 represents the electric constant, Tp denotes the thickness of the piezoelectric layer, d33 is the charge or strain constant, σ3 represents the stress on the surface of the piezoelectric element, ε33 denotes the permittivity of the piezoelectric element, and R is the total resistance in the EHPS circuit.

0 1 0 0 0 0 0 0

0 0 1 0 0 0 0 0

0 0 0 1 0 0 0 0

0 0 0 0 ma ma 0 0

xa 0 ⎤ ⎡ xp ⎤ ⎢ ⎥ 0 ⎥ ⎢ xb ⎥ 0⎥ ⎢ ⎥ V 0⎥ d ⎢ p⎥ 0 ⎥ dt ⎢ x ȧ ⎥ ⎥ x ṗ ⎥ 0⎥ ⎢ ⎢ ⎥ 0 ⎥ ⎢ xḃ ⎥ ⎥ 0 ⎦ ⎢Vṗ ⎥ ⎢ ⎦ ⎣ ⎥ −1 0 0 0 0 0 0 0 0 0 0 0 − ka ca 0 − nkp nP 0 nkp + kb − nP 0

0 0 0 0 0 mp 0 0

⎡0 0 ⎢0 0 ⎢0 0 ⎢0 0 ⎢ + ⎢ ka − ka ⎢ 0 nkp ⎢ 0 − nk p ⎢ ⎢0 0 0 ⎣ ⎡0 ⎤ ⎢0 ⎥ ⎢0 ⎥ ⎢0 ⎥ = ⎢Q ⎥ cos (ωt ) ⎢ 0⎥ ⎢ Q0 ⎥ ⎢0 ⎥ ⎢ ⎦ ⎣0 ⎥

⇒M Fig. 3. Schematic symbol and electronic model of EHPS.

4

0 0 0 0 0 0 mb 0

1 R

0

0 −1 0 0 − ca 0 0

0 0 −1 0 − ca 0 cb

− nP 0

0 ⎤ ⎡ xa ⎤ 0 ⎥ ⎢ xp ⎥ 0 ⎥ ⎢ xb ⎥ − 1 ⎥ ⎢Vp ⎥ 0 ⎥⎢ ẋ ⎥ ⎥ ⎢ a⎥ 0 ⎥⎢ ẋ ⎥ p 0 ⎥⎢ ẋ ⎥ ⎥⎢ b⎥ nCp ⎥ ⎢Vṗ ⎥ ⎦⎣ ⎦

dX + DX = Qcos (ωt ) = Q {exp(iωt ) + exp(−iωt )}/2 dt

(11) (12)

Applied Energy xxx (xxxx) xxx–xxx

L. Guo, Q. Lu

2009.

In this case, since the external dynamic force is a sinusoidal form, matrix X can be assumed as X = Y exp (iwt ) to find a solution of dX M dt + DX = Qexp(iωt ) : and assumed as [iωM + D] Y = Q

Given the various results of resistivity measured in previous studies, this study used proper contents of steel wool and graphite to increase the specimen’s electrical conductivity to 1 2 log(Ω m) in laboratory tests. It is worth noting that steel fiber or wool might be severely electrically-eroded under the electric current generated from the piezoelectric layer, which should be considered when selecting materials in practice. Since all external forces from vehicles will be distributed to piezoelectric materials through the upper conductive layer, the textures of the conductive layer also play an important role in improving the electrical output from the EHPS, especially if the size of the piezoelectric element is small.

dX M dt

+ DX = Qexp(−iωt ) : X= exp (−iwt ) to find a solution of [−iωM + D] Y ∗ = Q , where Y ([Y1, Y2, Y3, Y4]) and Y ∗ ([ Y 1∗,Y ∗2,Y ∗3,Y ∗4 ]) are the real and imaginary parts of amplitudes, respectively. The full solution can be expressed as X = {Y exp(iωt ) + Y ∗exp(−iωt )}/2 . As a result, the amplitude of each mass vibration can be expressed as: Y∗

Ya =

Y1 Y1∗ , YP =

Y2 Y2∗ , Yb =

Y3 Y3∗

(13)

The amplitude of voltage generated by the piezoelectric elements can be expressed as:

V=

Y4 Y 4∗

(14) 4.1.2. Options of piezoelectric element designs For the piezoelectric material selection, currently the most popular is PZT ceramic [28]. Its electricity generation ability caused by deformation is efficient and feasible, which was confirmed by Sodano and Inman in 2005 by comparing PZT with two other types of piezoelectric devices [29]. However, PZT has an extremely brittle nature, which limits its strain capacity, especially under high-frequency cyclic loading [30]. Compared to PZT, some polymer films consist of piezoelectric polymer materials such as PVDF and poly(3,4-ethylenedioxy-thiophene)/poly(4-styrenesulfonate) (PEDOT/PSS), maintain good piezoelectric properties, and simultaneously, improve their flexibility to prevent fatigue crack damage from high-frequency vibrations [30]. Instead of modifying PZT to other piezoelectric materials, building a composite through mixing 40% PZT fibers with 60% epoxy also can improve such material’s flexibility [31] while retaining its piezoelectric performance. Similar to filling with epoxy, Churchill et al. further collected usable electric power through mixing PZT fibers with resin [32]. This type of composite is termed “macro fiber composite” (MFC). The details of current piezoelectric material properties are summarized and compared by Shen et al. in 2007 and displayed in Table 2, who stated that PZT has significantly higher power density than PVDF and MFC materials, but its size is limited by its fracture strength [33]. This study selected the PZT material for use in the piezoelectric layer considering its commercial availability and higher power density. Since all piezoelectric components should be directly attached to the upper and lower conductive layers, the size of each piezoelectric component should match the piezoelectric layer thickness. To overcome the shortcomings of the brittleness property of PZT, dielectric materials (e.g., neat asphalt binder and aggregates) were added into the piezoelectric layer to share traffic loads with piezoelectric components and ensure the good structural performance of the EHPS.

The power generated by the piezoelectric elements can be expressed as:

Pp =

Y4 Y 4∗ 2R

(15)

4. Laboratory tests on EHPS prototype 4.1. Material options and their properties in EHPS On the practical side, current available material selections inevitably limit the feasibility of the EHPS design. This section discusses the available options of the conductive asphalt mixtures and piezoelectric materials used in the EHPS. After comparing the resistivity of different conductive asphalt mixtures and the piezoelectric properties of different piezoelectric materials, proper materials were selected to prepare EHPS prototype specimens for laboratory testing. 4.1.1. Options of conductive asphalt concrete designs To collect electric charges produced by the piezoelectric materials in the piezoelectric layer, two layers with conductive materials should sandwich the piezoelectric layer to generate the negative and positive electrodes as capacitor panels. To maintain the performance of asphalt concrete, the structural properties of these two conductive layers should be close to those of common asphalt concrete layers; otherwise, the critical stress generated in the two layers’ interface may cause the layers to flake and peel. Thus, this study focuses on modifying the electrical conductivity of pavement layers by adding conductive materials into asphalt concrete rather than directly setting conductive layers made by metals or other rigid materials. In most previous studies, the initial intention of adding conductive materials into asphalt concrete was for deicing in the 1950s. Through controlling additive contents and asphalt binder contents, several studies measured the electrical resistivity of different modified asphalt samples. The main outputs of resistivity from those studies are summarized in Table 1. Since the electrical conductivity of asphalt concrete mixed with conductive additives cannot be precisely controlled, the level of resistivity in log (Ω m) was selected as a proper indicator for material selection. Based on the data in Table 1, current available options of conductive materials based on modified asphalt are still uncertain for the following two reasons:

4.2. Experimental design of energy harvesting layers in laboratory After selecting the proper materials to build the EHPS prototype in the laboratory, experimental tests were run on specimens of the multilayer system to assess and verify their mechanical and electrical properties. During simulation of traffic loads on specimens by a Material Testing System (MTS), electrical outputs from the specimens were detected and recorded using an analog-to-digital converter (ADC) board connected to a computer. This section describes details of the laboratory test design, including preparation of specimens and design of the testing system.

• Mixing steel fiber and wool, graphite, or carbon fiber with asphalt •

may significantly reduce the electrical resistivity of asphalt concrete from 1 × 1011 Ω m to 10 Ω m. However, it is very difficult to drop it to 1 × 10−2 unless under special conditions, such as a sand-bitumen ratio at 0.77 with 8.76% of steel fibers measured by Garcia et al. The results from previous studies on the same materials are somehow inconsistent; some studies did not capture the rapid drop of resistivity that was observed in other studies, such as the studies on adding steel fiber by Liu and Wu in 2011 and by Huang et al. in

4.2.1. Preparation of conductive asphalt mixture Asphalt mixtures with relatively high conductivity (up to 0.5 log [Ω m]) were successfully produced in an early study [26]. Based on the mixing method in that study, conductive asphalt mixtures were prepared in this study. To ensure sufficient space for conductive fills, this study created a stone mastic asphalt (SMA) mixture of 19 mm nominal maximum size, SMA-19. To increase the conductivity of asphalt mixtures, graphite was added to the asphalt binder at a volume ratio of 1:3 5

Applied Energy xxx (xxxx) xxx–xxx

L. Guo, Q. Lu

Table 1 Electrically-conductive asphalt concrete materials with additives in previous studies. Additive type

High resistivity stage

Low resistivity stage

Additive content (%) in transition

Reference

Additive content (%)

Resistivity level log (Ω m)

Additive content (%)

Resistivity level log (Ω m)

Steel fiber Steel wool

0–10 0–8

9 9

20–25 10–15

Graphite Steel wool 5.83% Steel wool + graphite 6.54% Steel wool + graphite

0 0–5 5–12

11.5 11.5–10 9–8

N/A

Graphite Carbon fiber Carbon black 7.3% Graphite + 3.6% Carbon fiber 9.3% Graphite + 3.6% Carbon fiber 7.3% Carbon black + 3.6% Carbon fiber 9.3% Carbon black + 3.6% Carbon fiber Steel fiber carbon fiber Graphite

5 5

10–20 8–10

[24]

28 6–9 16–28

5.5 1–2 1

N/A 5–9 12–16

[25]

N/A

7–28

1

N/A

5–8 0–3 5–10 N/A

12 12 12 N/A

13–30 5–7.5 15–30 N/A

3–0.5 4–2 6–4 N/A

8–13 3–5 10–15 N/A

N/A

N/A

N/A

N/A

N/A

ρ: 1.00E+1 Ω m

N/A

N/A

N/A

N/A

N/A

ρ: 1.00E+1.2 Ω m

N/A

N/A

N/A

N/A

N/A

ρ: 1.00E+1 Ω m

0 0–3 0–10

10 10–8 10–9

2 5 18–28

2 2–1 3–2

N/A 3–5 10–18

Table 2 Properties of different piezoelectric materials [33].

Maximum temperature (°C) Young's modulus (GPa) Yield strength (MPa) d31 (×10−12 m/V) k31 Dielectric constant

Note

They also got 82.5 × 10−2 (Ω m) by sand-bitumen ratio at 0.77 with totally 8.76% of steel fibers

[26]

ρ: 1.00E+1.2 Ω m

[27]

Table 3 Components of conductive asphalt mixture.

PZT

PVDF

MFC

Sieve size

Percent passing of aggregates

230 62 20 320 0.44 3800

80 2–4 30 23 0.12 12–13

180 16 > 30 170 – –

#200 #100 #50 #30 #16 #8 #4 9.5 mm 12.5 mm 19.0 mm 25.0 mm

3.6 8.7 10.2 11.9 14.4 16 17.8 22.3 58.6 91.9 100

(graphite to asphalt binder) [26]. To further enhance the conductivity of asphalt mixtures, coarse steel wool #3 of a length of around 3 mm was used as another binder additive. To increase mixing uniformity, 4% steel wool (by volume of asphalt binder) was first mixed with aggregates, then mixed with an asphalt binder at an ambient temperature of 20 °C. The specific compositions of SMA-19 conductive asphalt mixture are listed in Table 3. Based on suggestions from a study by Wu and Yu [11], the following steps were taken to prepare the conductive asphalt mixtures:

Aggregate Asphalt Graphite Steel wool

Type

Weight (g)

Weight (%)

N/A PG 76–22 Powder Grade #3

3950.6 315.2 210.7 99.0

86.34 6.89 4.61 2.16

4.2.2. Preparation of proposed energy harvesting structure After preparing the conductive asphalt mixtures, copper wires, and plaster of Paris (POP) and gathering the PZT disks (Navy type II, obtained from APC international, Ltd.), the prototype of the energy harvesting structure proposed in this study was built in the laboratory (see Fig. 4(b) and (c)). The prototypes of the conductive layers and the piezoelectric layer were separately fabricated in the laboratory following the steps described below.

1. Heat asphalt binder until fluidity reaches 177 °C in a forced draft oven. 2. Add graphite into the heated asphalt binder. 3. Use an electric drill with a mixing head to mix the binder clockwise and counterclockwise for about 3 min each at a 2400 rpm rotational speed. 4. Heat the modified asphalt binder again to 177 °C in the forced draft oven, and repeat Step 3 three times. 5. Blend the aggregates, limestone powder, steel wool, and modified asphalt at the ambient temperature. 6. Put the conductive asphalt mixture in the forced draft oven at 177 °C, and repeat Step 5 three times until the mixture is well blended.

• Four main steps were followed to build the prototype of conductive

layers: 1. Put the first part of the heated conductive asphalt mixture in a Marshall compactor mold. 2. Put a spiral-shaped copper wire on the conductive asphalt mixture. Sufficient length of the copper wire at both ends is preset for electricity output use.

6

Applied Energy xxx (xxxx) xxx–xxx

L. Guo, Q. Lu

Fig. 4. Measuring electrical output from specimens loaded in MTS.

Fig. 5. Electrical output from SMA specimens without POP fills.

6. Remove the tape from the PZT disks. 7. Put the POP mixture back into the oven at 65 °C for overnight drying. 8. Insert the PZT disks back into the dried POP mixture.

3. Pour the rest of the heated conductive asphalt mixture on the copper wire in the mold. 4. Compact the conductive asphalt mixture using a Marshall compactor. Eight steps were followed to build the prototype of the energyharvesting layer: 1. Wrap the PZT disks with tape to make sure insulation materials do not contaminate their surface. 2. Mix POP powder with water at a volume ratio of 2:1 in the mold, which is covered with plastic paper for easy removal of the POP mixture from the mold after it cures. 3. Embed the PZT disks into the POP mixture in the mold. The heights of the POP mixture and the PZT disks must be consistent. 4. Put the mold in an oven at 65 °C for 45 min. 5. Demold the specimen and remove the embedded PZT disks from the POP mixture.

•

4.2.3. Design of testing system An MTS Series 647 Hydraulic Wedge Grip was selected to simulate traffic loads by setting the appropriate loading frequency and magnitude. A sinusoidal load of 133 N amplitude and 1 Hz frequency was first applied to the specimens, with the number of load cycles set to 25. Since both the top and bottom anvils of the MTS were made of metal, papers and plastic films were used to isolate specimens from anvils to prevent leakage of generated electricity. After squeezing electricity from the specimen, an ADC board was set to receive and convert electrical voltage signals into a computer, which was able to save the electrical signal information for further data analysis, as shown in 7

Applied Energy xxx (xxxx) xxx–xxx

L. Guo, Q. Lu

Fig. 4(a).

Table 4 Parameters of material properties used in control variables. Basic parameters of material properties

Value

Mass of conductive asphalt mixture specimen A, ma, (kg) Mass of conductive asphalt mixture specimen B, mb, (kg) Mass of each PZT disk, mp, (kg) Thickness of conductive asphalt mixture specimen A, Ta, (m) Thickness of conductive asphalt mixture specimen B, Tb, (m) Thickness of each PZT disk, Tp, (m) Diameter of each PZT disk, dp, (m) Diameter of each conductive asphalt mixture specimen, d, (m) Resilient modulus of each conductive asphalt mixture specimen, Mr, (Pa) Young’s Modulus of PZT material, Y33, (N/m2) Damping ratio of each conductive asphalt mixture specimen, ζ Piezoelectric stress constant, e33, (C/m2) Dielectric constant, ε33 Resistance, R, (Ω)

1.241 1.393 8.22 × 10−3 0.086 0.097 0.0106 0.0114 0.100 2.70 × 109 5.4 × 1010 0.06 15.5 1900 2 × 107

Estimated parameters of material properties

Value

Spring constant of conductive asphalt mixture specimen A, ka, (N/ m) Damping coefficient of conductive asphalt mixture specimen A, ca, (N s/m) Spring constant of conductive asphalt mixture specimen B, kb, (N/ m) Damping coefficient of conductive asphalt mixture specimen B, cb, (N s/m) Spring constant of PZT material, kp, (N/m) Electromechanic coupling coefficient, P, (C/m) Capacitance of each PZT disk, Cp, (F)

2.466 × 108

4.2.4. Laboratory test results After squeezing the SMA specimen with the POP fill, as shown in Fig. 4(b), a steady electrical output was not captured at first from the SMA specimen. To improve the contact condition between PZT disks and the SMA mixtures, the POP was removed as shown in Fig. 4(c). The results are shown in Fig. 5. As can be seen, without POP, the electrical output from the proposed energy harvesting layer increased from 4.5 V to 7.6 V when the amplitude of the sinusoidal load was increased from 88 N to 133 N at 1 Hz. Since PZT disks remain intact under such high loads, the brittle PZT materials should withstand this level of impact when embedded into the piezoelectric layer of the EHPS. To verify the electromechanical model developed in this study, the theoretical results are shown in Fig. 5. The parameters used to obtain these theoretical results are those listed in Table 4. Specifically, the properties of the asphalt mixture specimens were either measured in the laboratory or estimated from the literature [24,34,35]. The properties of the PZT disks are provided by the sources from APC international, Ltd. [36] and Morgan Technical Ceramics [37]. As can be seen from Fig. 5, the theoretical results of the output voltage are slightly higher than the experimental results. Considering potential errors in the selected values of several parameters in Table 4 (e.g., MR, R, e33), the percent error (around 25%) from using the electromechanical model is deemed acceptable.

2099.25 2.186 × 108 2094.03 5.2 × 108 0.15 1.6 × 10−10

5. Electromechanical model results and discussions Numerous variables can affect electricity generation from the EHPS. In this study, they are categorized into four groups: properties of the Fig. 6. Effects of properties of conductive asphalt mixtures (MR and ζ) on output voltage and output power from EHPS under two external vibration frequencies, (a) 1 Hz and (b) 30 Hz.

8

Applied Energy xxx (xxxx) xxx–xxx

L. Guo, Q. Lu

Fig. 7. Effects of the electromechanical properties of piezoelectric elements (Cp and e33) on output voltage and output power from EHPS under two external vibration frequencies, (a) 1 Hz and (b) 30 Hz.

conductive asphalt mixtures (MR and ζ), electromechanical properties of the PZT disk (Cp and e33), design of the PZT disk (ωnp , n), and external dynamic load conditions (Q and ω). Each group is discussed separately in this section to determine design strategies that can optimize each component of the EHPS. Two variables from one group are discussed, and the other variables from the remaining three groups are regarded as control variables. The parameters used in the control variables are shown in Table 4. Two important indicators of electrical outputs, electrical voltage and electric power, are presented in Figs. 6–9.

exponentially if e33 is sufficiently large. Comparing Fig. 7(a) and (b), it can be seen that at a higher frequency of external vibration, the change of electrical outputs caused by modifying the capacitance of PZT disk is more dramatic. 5.3. Effects of ωnp and n on electrical outputs from EHPS As the most important common factor affecting the vibration amplitude, the natural frequency of PZT disk, ωnp , can play a key role in producing high electrical outputs from the EHPS. Assuming each PZT disk has a spring constant of 5.2 × 108 N/m and a mass of 8.22 × 10−3 kg, the natural frequency of each PZT disk is estimated to be 2.5 × 105. Apparently, this level of natural frequency does not match the frequency of traffic loads. Fortunately, based on Fig. 8(a), the limitation of decreasing natural frequency of a PZT disk to produce higher electrical output from the EHPS can be compensated by adding more PZT disks inside the EHPS under very low frequency (1 Hz) vibration. However, based on Fig. 8(b), under relatively high frequency (30 Hz) vibration, the benefit from adding more PZT disks to produce higher electrical outputs disappears. The output power from the EHPS holds relatively steady around 30 mW.

5.1. Effects of MR and ζ on electrical outputs from EHPS The two material property parameters of conductive asphalt mixtures, resilient modulus (MR) and damping ratio (ζ), reflect the elastic and viscous properties of asphalt mixtures, respectively. Fig. 6 illustrates that a stiffer asphalt mixture inside the EHPS leads to less electrical output (either voltage or power) from the EHPS. Meanwhile, the damping ratio of asphalt mixture does not affect the electrical output significantly at either 1 Hz or 30 Hz loading condition, although higher damping means more energy loss. 5.2. Effects of Cp and e33 on electrical outputs from EHPS

5.4. Effects of Q and ω on electrical outputs from EHPS The piezoelectric stress constant, e33, may vary significantly among different PZTs (e.g., 9.0 for PZT-2, 15.8 for PZT-5A, 23.3 for PZT-5H), and the capacitance of each PZT disk can be greatly improved by stacking. These two key electromechanical properties of piezoelectric materials can directly affect the electrical outputs from the EHPS, as illustrated in Fig. 7: increasing piezoelectric stress constant e33 improves the electrical outputs linearly from the EHPS and increasing capacitance Cp decreases the electrical outputs from the EHPS

After plotting the electrical outputs from the EHPS under two external vibration frequencies, Figs. 6–8 clearly display the interaction between the frequency of external vibration and several material properties on producing electricity from the EHPS. If all material properties of components in the EHPS are kept constant, the effects of external force on electricity output are shown in Fig. 9. As can be seen, both the amplitude and the frequency of the external force are 9

Applied Energy xxx (xxxx) xxx–xxx

L. Guo, Q. Lu

Fig. 8. Effects of design of piezoelectric elements (ωnp , n) on output voltage and output power from EHPS under two external vibration frequencies, (a) 1 Hz and (b) 30 Hz.

Fig. 9. Effects of external dynamic load conditions (Q and ω) on output voltage and output power from EHPS.

can generate a maximum electric power of more than 300 mW, as shown in Fig. 11. For the output voltage, both Figs. 10 and 11 show that the output voltage increases and then trends to constant by adding more external load resistance. For the output power, both Figs. 10 and 11 indicate that the power produced from the EHPS first increases to the maximum as the external load resistance reaches the optimum resistance and then gradually decreases to be very low after the external resistance exceeds the optimum value.

proportional to the amount of electricity generation from the EHPS.

5.5. Maximum outputs from simple EHPS and optimized EHPS To compare the maximum electric power generated from an initial EHPS and an optimized EHPS, the electrical outputs (voltage and power) from the two EHPSs under different external loads and different vibrational frequencies were plotted in Figs. 10 and 11. Fig. 10 shows that a simple EHPS (parameters listed in Table 4) can produce a maximum electric power of around 1.2 mW. This EHPS was optimized by adding more PZT disks (from 3 to 58), reducing the resilient modulus of asphalt mixtures (from 2.70 × 103 MPa to 1.00 × 103 MPa), and improving the piezoelectric stress constant e33 (from 15.5 to 23.3), which

6. Cost-effectiveness analysis of EHPS To weigh the benefits and cost from the proposed EHPS, a costeffectiveness analysis was performed based on the specific duration of 10

Applied Energy xxx (xxxx) xxx–xxx

L. Guo, Q. Lu

Fig. 10. Output voltage and output power from a simple EHPS connected by external load resistances under different levels of vibrational frequencies.

(1207 + 75) × π × (0.05)2 + the EHPS is approximately 19 × 0.099 + 31.6 × 0.211 = 18.6 ("$") . Assuming the average highway speed of vehicles is 65 mph (29 m/ s), its corresponding load frequency can be 24 Hz (value of velocity divided by 1.2) [39]. Based on the electromechanical model developed in this study, the electric power from a unit of EHPS can be around 22.2 mW under the frequency of 24 Hz, and approximately 0.001 J (2.78 × 10−10 kWh) of energy output can be generated from one tire rolling over the EHPS at a speed of 65 mph. The LCOE of the EHPS can be expressed as:

its service life under specific traffic conditions. A levelized cost of electricity (LCOE) was used as an indicator, expressed as follows:

LCOE = =

sum of costs over lifetime per unit sum of electrical energy produced over lifetime per unit Cp + Ci + Cc Wp × N × 365 × Y

(16)

where Cp = cost of PZT elements per unit ($) Ci = cost of installation per unit ($) Cc = additional cost from conductive asphalt mixture per unit ($) Wp = energy output from one unit of EHPS per vehicle (kWh) N = number of vehicles per day Y = service life (year)

LCOEEHPS =

18.6 2.78 × 10−10 × 2 × N × 365 × Y

(17)

In Eq. (17), the traffic volume on the roadway network can be as high as 3 × 105 per day in some states—for example, 319,000 in Florida [40]. The expected service life of the EHPS can range from 5 to 15 years [23,38]. Thus, the LCOE of the EHPS paved on a high-volume roadway section can vary from $19.15/kWh to $57.46/kWh depending on its real service life. Although the LCOE of the EHPS is higher than that from fossil-powered power plants ($0.15/kWh–$0.30/kWh), the EHPS can be built as an off-grid power system in the roadway and provide a more convenient way to supply clean energy to infrastructure or even electric vehicles in a modern transportation system. Meanwhile, the EHPS has plenty of room for improvement to decrease its LCOE via improving its power output, decreasing its cost, or extending its service life.

Since most flexible pavements should be rehabilitated or reconstructed by mill and overlay after 15 years when their performance deteriorates below a threshold value, the EHPS can be added into pavement layers during rehabilitation or reconstruction. Since the piezoelectric materials and conductive materials are newly-introduced into a regular pavement system, the sum of costs considered in LCOE includes only the costs from those two types of materials without any cost from regular pavement construction materials. Moreover, since the electrical output from each experimental specimen had be calculated and verified in this study, the specimen of 0.1 m diameter was used as a unit of EHPS to be more conveniently evaluated by LCOE. According to a study by Urquiza and Fernandez [38], the cost of piezoelectric elements per square meter should be around $1207, and the cost of their installation $75/m2. For the cost of conductive materials, the market prices of steel wool and graphite are approximately $19/kg and $31.6/kg, respectively. Thus, the sum of costs per unit of

7. Discussion and conclusions This paper introduces the basic concept of an ongoing study that aims to transform asphalt layers into a large-scale piezoelectric energy harvester, in which conductive asphalt mixtures, regular asphalt mixtures, and piezoelectric materials are layered to collect more dissipated Fig. 11. Output voltage and output power from optimized EHPS connected by external load resistances under different levels of vibrational frequencies.

11

Applied Energy xxx (xxxx) xxx–xxx

L. Guo, Q. Lu

kinetic energy from vehicle vibrations. As the first step to accomplish such a purpose, this study developed a laboratory prototype of a new energy harvesting pavement system (EHPS) design and built a multidegree-of-freedom electromechanical model (MDOF-EM) to analyze the effect of each component of the EHPS on the electrical output. The laboratory test results show that, with a good contact condition, the electrical output from specimens can be increased from 4.5 V to 7.6 V when the external load amplitude is increased from 88 N to 133 N at a 1 Hz frequency. These outputs demonstrate the feasibility of the EHPS design. Furthermore, the consistency between the voltage outputs from the EHPS prototype measured in the laboratory and those estimated from the MDOF-EM verify the accuracy of the electromechanical model developed in this study. Based on the electromechanical model and the electrical outputs, including both voltage outputs and power outputs, from the EHPS with different material properties, piezoelectric layer designs and external dynamic load conditions were calculated and analyzed, and the following findings were obtained:

[5] Guo L, Lu Q. Potentials of piezoelectric and thermoelectric technologies for harvesting energy from pavements. Renew Sustain Energy Rev 2017;72:761–73. [6] Kang WW, Correia AJ. A pilot study for investigation of novel methods to harvest solar energy from asphalt pavements. South Korea Goyang City: Korea Institute of Construction Technology (KICT); 2010. [7] Xiong H, Wang L. Piezoelectric energy harvester for public roadway: on-site installation and evaluation. Appl Energy 2016;174:101–7. [8] Van B, Houben L, Scarpas A, Molenaar A. Using pavement as solar collector: effect on pavement temperature and structural response. Transp Res Rec 2001;1778:140–8. [9] Zhou Z, Wang X, Zhang X, Chen G, Zuo J, Pullen S. Effectiveness of pavement-solar energy system—an experimental study. Appl Energy 2015;138:1–10. [10] Jesus VB, Munoz PP, Fresno DC, Hernandez JR. Asphalt solar collectors: a literature review. Appl Energy 2013;102:962–70. [11] Wu G, Yu X. Thermal energy harvesting system to harvest thermal energy across pavement structure. Int J Pavement Res Technol 2012:311–6. [12] Bhattacharjee S, Batra AK, Cain J. Carbon nano fiber reinforced cement composite for energy harvesting road. Green Streets Highways 2010:258–71. [13] Zhao H, Yu J, Ling J. Finite element analysis of cymbal piezoelectric transducers for harvesting energy from asphalt pavement. J Ceram Soc Jpn 2010;118:909–15. [14] Yao L, Zhao H, Dong Z, Sun Y, Gao Y. Laboratory testing of piezoelectric bridge transducers for asphalt pavement energy harvesting. Key Eng Mater 2012;492:172–5. [15] Li C. Road performance of common piezoelectric transducer for asphalt pavement energy harvesting. Appl Mech Mater 2015;744:1491–4. [16] Kim C, Kim K, Jeon J, Jeong Y, Cho J, Paik J, Kang I, Lee M, Choi B, Cho Y, Park S, Nahm S, Lee Y. Development and evaluation of the road energy harvester using piezoelectric cantilevers. J Kor Inst Electr Electron Mater Eng 2012;25:511–5. [17] Sun C, Shang G, Zhang Y. Designing piezoelectric harvesting unit from road vibration. Adv Mater Res 2013;712:1368–71. [18] Daniels A, Giuliano A, Zhu M, Tiwari A. Modeling, validation, and design analysis of a piezoelectric cymbal transducer for non-resonant energy harvesting. IEEE international conference on green computing and communications and IEEE internet of things and IEEE cyber, physical and social computing, Beijing, China; 2013. [19] Xiang HJ, Wang JJ, Shi ZF, Zhang ZW. Theoretical analysis of piezoelectrical energy harvesting from traffic induced deformation of pavements. Smart Mater Struct 2013;22(9):1–9. [20] Zhang ZW, Xiang HJ, Shi ZF. Modeling on piezoelectrical energy harvesting from pavements under traffic loads. J Intell Mater Syst Struct 2015:1–12. [21] Roshani H, Dessouky S, Montoya A, Papagiannakis AT. Energy harvesting from asphalt pavement roadways vehicle-induced stresses: a feasibility study. Appl Energy 2016;182:210–8. [22] Jung I, Shin Y, Kim S, Choi J, Kang C. Flexible piezoelectric polymer-based energy harvesting system for roadway applications. Appl Energy 2017;197:222–9. [23] Xiong H. Piezoelectrical energy harvesting for public roadways PhD dissertation Virginia Polytechnic Institute and State University; 2014. [24] Liu X, Wu S. Study on the graphite and carbon fiber modified asphalt concrete. Constr Build Mater 2011;25:1807–11. [25] Garcia A, Schlangen E, Ven MV, Liu Q. Electrical conductivity of asphalt mortar containing conductive fibers and fillers. Constr Build Mater 2009;23:3175–81. [26] Wu S, Mo L, Shui Z, Chen Z. Investigation of the conductivity of asphalt concrete containing conductive fillers. Carbon 2005;43:1358–63. [27] Huang B, Chen X, Shu X. Effects of electrically conductive additives on laboratorymeasured properties of asphalt mixtures. J Mater Civ Eng 2009;21:612–7. [28] Anton SR, Sodano HA. A review of power harvesting using piezoelectric materials (2003–2006). Smart Mater Struct 2007;16:R1–21. [29] Sodano HA, Inman DJ. Comparison of piezoelectric energy harvesting devices for recharging batteries. J Intell Mater Syst Struct 2005;16:799–807. [30] Lee CS, Joo J, Han S, Koh SK. Multifunctional transducer using poly(vinylidene fluride) active layer and highly conducting poly(3,4-ethylenedioxythiophene) electrode: actuator and generator. Appl Phys Lett 2004;85:10. [31] Mohammadi F, Khan A, Cass RB. Power generation from piezoelectric lead zirconate titanate fiber composite. Proc Mater Res Sympos 2003:746. [32] Churchill DL, Hamel MJ, Townsend CP, Arms SW. Strain energy harvesting for wireless sensor networks. Smart Struct Mater, Proc SPIE 2003:5055. [33] Shen D, Choe SY, Kim DJ. Analysis of piezoelectric materials for energy harvesting devices under high-g vibrations. Jpn J Appl Phys 2007;46:6755–60. [34] Shafabakhsh G, Naderpour H, Motamedi M. Dynamic analysis and determination of maximum tensile strain of bottom asphalt concrete for different vehicle velocities. Eng J 2015;19:112. [35] Karacasu M, Okur V, Er A. A study on the rheological properties of recycled rubbermodified asphalt mixtures. Sci World J 2015. Article ID 258586. [36] APC. Physical and piezoelectric properties of APC materials. < https://www. americanpiezo.com/apc-materials/piezoelectric-properties.html > [accessed December 2015]. [37] Berlincourt D, Krueger HHA. Properties of Morgan electro ceramic ceramics. Morgan electro ceramics tech. publication TP-226, Cleveland, OH, USA; 2002. [38] Urquiza D, Fernandez JF. Feasible integration in asphalt of piezoelectric cymbals for vibration energy harvesting. Energy Convers Manage 2016;112:246–53. [39] Sun L, Duan Y. Dynamic response of top-down cracked asphalt concrete pavement under a half-sinusoidal impact load. Acta Mech 2013;224:1865–77. [40] Florida Department of Transportation. Geographic information system. < http:// www.dot.state.fl.us/planning/statistics/gis/ > [accessed May 20, 2017].

• Increasing the stiffness of conductive asphalt mixture layers inside

•

•

•

the EHPS can reduce the electrical output from the EHPS. Meanwhile, decreasing the viscosity of the conductive asphalt mixture cannot sufficiently improve the efficiency of the EHPS at either 1 Hz or 30 Hz load frequency, although more damping means more energy loss. Using piezoelectric elements with a higher piezoelectric stress constant can improve the electrical outputs linearly from the EHPS. On another side, increasing capacitance may exponentially decrease the electrical outputs from the EHPS if the piezoelectric stress constant is sufficiently large. In other words, the piezoelectric stress constant and the capacitance of each piezoelectric element interact with electrical outputs from the EHPS. It is difficult to practically decrease the natural frequency of a piezoelectric element to match traffic load frequency to produce a higher electrical output from the EHPS. Fortunately, adding more piezoelectric elements inside the EHPS can be an alternative option for the EHPS to produce higher electrical output, especially under very low frequency (1 Hz) vibration. Both the amplitude and frequency of external force are proportional to the amount of electrical outputs from the EHPS. Moreover, a higher frequency (30 Hz) of external vibration can lead to a more dramatic effect of a piezoelectric element’s capacitance on electrical outputs and reduce the benefit of adding more piezoelectric elements to produce higher electrical outputs.

The electromechanical model in this study shows that the maximum electric power generated from a simple EHPS prototype can be around 1.2 mW. After optimizing this EHPS by adding more piezoelectric elements with a higher piezoelectric stress constant and replacing more flexible conductive asphalt mixtures, the maximum electric power from the EHPS prototype can be increased to 300 mW under a high frequency (30 Hz) external vibration. The levelized cost of electricity of this EHPS can reach $19.15/kWh on a high-volume roadway within a 15-year service life. References [1] Lu XP, Segal L. Vehicular energy losses associated with the traversal of an uneven road. Veh Syst Dynam: Int J Veh Mech Mobility 1986;15:342–52. [2] Gunselmann W. Technologies for increased energy efficiency in railway systems. Power electronics and applications on 2005 European conference, Horsaalzentrum der TU Dresden, Germany; 2005. [3] Holmberg K, Anderson P, Erdemir A. Global energy consumption due to friction in passenger cars. Tribol Int 2012;47:221–34. [4] Symeoni A. A review on energy harvesting from roads. Thesis, Sweden: KTH Royal Institute of Technology; 2012.

12