An investigation of energy harvesting from renewable sources with ...

Home

Search

Collections

Journals

About

Contact us

My IOPscience

An investigation of energy harvesting from renewable sources with PVDF and PZT

This article has been downloaded from IOPscience. Please scroll down to see the full text article. 2011 Smart Mater. Struct. 20 055019 (http://iopscience.iop.org/0964-1726/20/5/055019) View the table of contents for this issue, or go to the journal homepage for more

Download details: IP Address: 193.63.48.50 The article was downloaded on 26/04/2011 at 08:37

Please note that terms and conditions apply.

IOP PUBLISHING

SMART MATERIALS AND STRUCTURES

Smart Mater. Struct. 20 (2011) 055019 (6pp)

doi:10.1088/0964-1726/20/5/055019

An investigation of energy harvesting from renewable sources with PVDF and PZT D Vatansever, R L Hadimani, T Shah and E Siores Institute for Materials Research and Innovation, University of Bolton, Bolton BL35AB, UK E-mail: [email protected]

Received 7 October 2010, in final form 24 March 2011 Published 19 April 2011 Online at stacks.iop.org/SMS/20/055019 Abstract Piezoelectric materials have been in use for many years; however, with an increasing concern about global warming, piezoelectricity has gained significant importance in research and development for extracting energy from the environment. In this work the voltage responses of ceramic based piezoelectric fibre composite structures (PFCs) and polymer based piezoelectric strips, PVDF (polyvinylidene fluoride), were evaluated when subjected to various wind speeds and water droplets in order to investigate the possibility of energy generation from these two natural renewable energy sources for utilization in low power electronic devices. The effects of material dimensions, drop mass, releasing height of the drops and wind speed on the voltage output were studied and the power was calculated. This work showed that piezoelectric polymer materials can generate higher voltage/power than ceramic based piezoelectric materials and it was proved that producing energy from renewable sources such as rain drops and wind is possible by using piezoelectric polymer materials. (Some figures in this article are in colour only in the electronic version)

zirconate titanate) fibre composites to wind and rain drops with the aim of using piezoelectric materials in outdoor applications and producing energy from these renewable sources.

1. Introduction Both ceramic and polymer based piezoelectric materials have found a wide range of applications in many areas. With an increasing concern about global warming, the direction of research has changed slightly from just sensors and actuators to an interest in generating and storing electricity from renewable sources. Numerous energy harvesting techniques have gained significant importance and intensive research is being carried out on extracting and storing energy from the environment [1–5]. Previous works showed that in most conditions ceramic based piezoelectric structures (PFCs) produced higher voltages than polymer based piezoelectric structures [6–9]. However, PFCs are either rigid or brittle which limits the range of their applications. The use of polymer based piezoelectric materials provides some advantages over the other piezoelectric materials, since polymeric materials are lead free, inexpensive and easy to process. Moreover, due to its light weight and flexibility, polyvinylidene fluoride (PVDF) has an advantage for many applications where flexibility is critical, such as backpack straps [10]. In this work, we have comparatively studied the voltage responses of both piezoelectric PVDF structures and PZT (lead 0964-1726/11/055019+06$33.00

2. Direct piezoelectric effect of materials for energy harvesting The piezoelectric voltage constant in the 31-mode is higher for polymer piezoelectric materials because of their bending ability, which is 216 × 10−3 V m N−1 for PVDF films while it is only 10 × 10−3 V m N−1 for PZTs [11]. PZT has been preeminent due to its piezoelectricity among other piezoelectric materials, with a piezoelectric coefficient (d33 ) of 220 pC N−1 [12], while PVDF has a much lower piezoelectric coefficient of d33 ≈ 35 pC N−1 [13–15]. The piezoelectric voltage coefficient (g ) relates the electric field generated by an applied mechanical stress and can be measured by the following equation [16]:

gi j = V 2wt/3 Fi l

(1)

where ‘V ’ is the maximum voltage generated, ‘w’ is the width, ‘t ’ is the thickness, ‘l ’ is the length of the piezoelectric material and ‘ F ’ is the force applied. The piezoelectric voltage constant is an important parameter for assessing the material suitability 1

© 2011 IOP Publishing Ltd Printed in the UK & the USA

Smart Mater. Struct. 20 (2011) 055019

D Vatansever et al

Table 1. Characteristics of the investigated piezoelectric polymer and ceramic fibre composite samples. Material name

Length (mm)

Width (mm)

Thickness (mm)

Capacitance (nF)

Minimum radius of curvature (mm)

Supplier

LDT1-28K LDT4-28K PZT-single layer PZT-bimorph

41 171 140 140

16 22 15 15

0.20 0.20 0.32 0.75

1.38 11 4.96 8.41

3.00 5.00 19.00 Rigid

MEAS MEAS ACI ACI

for sensing applications (sensor) and is expressed in voltmeter/newtons (V m N−1 ). It has been reported that the human body is a source of energy which can be harvested in the form of electrical energy [17]. The author estimated the amount of power which can be generated by human body motion by developing a power harvesting device that can be implanted into a shoe [6]. Three different devices, a unimorph strip made from piezoceramic composite material, a stave made from a multilayer laminate of PVDF and a shoe-mounted rotary magnetic generator, were employed and comparatively studied. Although PVDF was the best suited material for the application due to its flexibility, its calculated power output was much lower than the other piezoelectric materials used. Energy scavenging from the movement of the ribs during respiration in mongrel dogs has also been investigated using PVDF films which generated 18 V [18]. Harvestable energy can be calculated with respect to the relationship between voltage generated and current. The relationship can be expressed as

P = VI

Figure 1. Rain drop experimental setup: PVDF sample with half of the length located on the support and half of the length left free to oscillate.

The samples listed in table 1 were subjected to both wind and water drop experiments in order to explore the possibility of using piezoelectric materials in outdoor energy scavenging applications. With the availability of the new piezoelectric polymer fibres [19] it is now possible to produce textile structures which can be used for energy scavenging applications. These flexible piezoelectric structures may be designed and integrated into various forms of textile structures for outdoor applications such as clothing, awnings, tents, sails, etc.

(2)

where ‘ P ’ is the power (watts, W), ‘V ’ is the voltage generated by the piezoelectric material (volts, V) and ‘ I ’ is the current (amperes, A) measured with pure resistance.

3. Experimental details 3.1. The materials investigated

3.2. Rain drop experiment setup

For energy generation purposes from the environment, a variety of materials have been selected and objectively compared. The single layer piezoelectric fibre composite material consisting of unidirectionally aligned brittle piezoelectric PZT fibres with a diameter of 250 μm embedded in an epoxy matrix, sandwiched between two copper clad polyamide laminates, and a PZT-bimorph material consisting of two 250 μm single layer PFCs adhered to either side of a rigid metal centre shim were obtained from Advanced Cerametrics Inc. (ACI), USA. Two laminated piezoelectric PVDF specimens were also used, where two polyester laminates are attached to either side of a 28 μm thick piezoelectric film element, namely LDT1-28K and LDT4-28K, respectively. The PVDF specimens were obtained from Measurement Specialities Inc. (MEAS), USA. The names, dimensions and characteristics of the piezoelectric polymer and ceramic specimens are given in table 1. The minimum radius of curvature shows the degree of flexibility of the samples. The smaller minimum radius of curvature values represent better flexibility.

In the first experiment various piezoelectric specimens were subjected to the impact of water drops with approximate masses of 50 and 7.5 mg from different heights. The samples were located on a rigid support. The voltages generated by the PVDF strips and PZT composite structures were measured and recorded by an oscilloscope (figure 1). The piezoelectric samples were located on a rigid support on one end and the other end was not supported to allow easy oscillations. Various lengths of the strip materials were left free to allow oscillations in order to generate voltage. Samples were also located on the edge of the support and the whole length of the samples was free to oscillate for voltage generation. The sample was held rigidly at one end by double sided adhesive tape at the bottom and a single sided adhesive tape on the top, as seen in figure 1. The droplets were released at low frequency such that the droplets reached the sample only after the previous droplet had flown off the sample. The location of the droplet was adjusted on the sample by moving the base on 2


D Vatansever et al

4.5 m s−1 , 6 m s−1 , 8 m s−1 and 10 m s−1 , in order to record the peak voltage generated by fluctuation of the samples.

4. Results and discussion The piezoelectric responses of the polymer and ceramic based piezoelectric samples at different excitation parameters were studied and the results are evaluated in this section. For both the rain drop and wind tunnel experiments, each test was repeated ten times for each sample with an approximate error of less than 5%. Figures 3(a) and (b) show generated peak voltages with respect to different releasing heights of water drops when half of the sample’s length was located on the rigid support. Figures 3(a) and (b) show that the short PVDF sample, LDT1-28K, exhibited the highest peak voltage when it was subjected to impact caused by water drops in this experiment. An increase in the drop releasing height yields higher voltage generation which is in agreement with the results of increasing impact force reported previously [20, 21]. The long PVDF sample, LDT4-28K, showed a lower voltage generation especially when subjected to 7.5 mg water drops which were not sufficiently heavy to induce oscillations of the sample. When the same sample was subjected to a larger drop (50 mg) its peak voltage generation was higher than ceramic based samples. A previous study [7] reported that 8.4 mW energy was generated from a semi-flexible PZT laminated and pre-stressed spring metal strip in a shoe; however the energy generated by a flexible multilaminar PVDF bimorph stave was only 1.3 mW. Furthermore, a more recent work showed that ceramic fibre based piezoelectric material generated much higher voltage (40 V) compared to polymer based piezoelectric structures under vibration and impact [8]. However, in our study ceramic materials generated a relatively smaller peak voltage compared to polymer based piezoelectric materials. This indicates that the rigidity of the ceramic material hinders the oscillations caused by water drops. Thus the mass of the

Figure 2. Wind experiment setup: a long PVDF sample is fixed to the wall and the aerometer is 20 cm in front of the wind tunnel.

which the sample was located. It was ensured that the droplet reached the sample on the ‘dot’ marked on the sample. Impact was induced by releasing water from different heights (20, 50 and 100 cm) on top of the materials in order to obtain the effect of different parameters on the voltage generation ability of the piezoelectric samples. The voltage generated with respect to the weight of drops and height was recorded by using a digital oscilloscope. 3.3. Wind experiment setup The second experiment was to subject the piezoelectric samples to wind at various wind speeds simulated by a custom built wind tunnel. Wind speed was measured using an aerometer placed 20 cm in front of the tunnel. To create turbulence a valve was located between the wind source and the samples and the turbulence was created by opening and closing the mechanism of the valve (figure 2). The generated voltage with respect to wind speed was measured by using an oscilloscope. Measurements were taken for each sample at least ten times at wind speeds of 3 m s−1 ,

Figure 3. (a) Generated peak voltage with respect to various heights when 7.5 mg water drops were released, on PVDF strips and PZT composites with half the length of the samples located on the support. (The inset shows the voltage response versus time for a single height of release.) (b) Generated peak voltage with respect to various heights when 50 mg water drops were released, on PVDF strips and PZT composites with half the length of the samples located on the support. (The inset shows the voltage response versus time for a single height of release.)

3


D Vatansever et al

Figure 4. (a) Generated peak voltage with respect to various heights when 7.5 mg water drops were released, on PVDF strips and PZT composites with the full length of the samples located on the edge of the support. (The inset shows the voltage response versus time for a single height of release.) (b) Generated peak voltage with respect to various heights when 50 mg water drops were released, on PVDF strips and PZT composites with the full length of the samples located on the edge of the support. (The inset shows the voltage response versus time for a single height of release.)

water drops was not sufficient to activate the ceramic materials piezoelectrically. The highest peak voltage generation was observed using the flexible piezoelectric polymer material with the dimensions of 16 mm × 4 mm × 0.2 mm. This is due to its smaller length—the forward strain is faster in shorter films which contributes to the generation of substantially higher peak voltage. Figures 4(a) and (b) present the generated peak voltages with respect to different releasing heights of water drops. When the location of the samples was changed to allow the whole length of the samples to oscillate freely, the long PVDF sample generated lower voltage while the voltage generated by the shorter one was not affected. For polymer based samples, it can be argued that the longer the length the slower the fluctuation and, hence, a relatively lower voltage is generated. The reason is that when long and flexible samples are located on the support from the edge, their own weight stretches the material and applies some stress onto it; therefore the tested samples become pre-stressed. This may be the reason why the longer PVDF sample (LDT428K) shows a larger increase in the voltage output with an increase in the drop size and releasing height (figure 3(b)) as compared to the results reported for the same sample when it is supported from its edge (figure 4(b)). For the PZT based structures, the effect of various parameters—drop size, releasing height and position of the specimen from the support edge—did not have any significant influence on the voltage generation. However, the single layer PZT sample showed slightly higher voltage generation than the PZT-bimorph, as it is the rigidity of the composite that hinders the oscillations and the bimorph structure is more rigid than the single layer structure. The piezoelectric voltage coefficient of the long PVDF strips was calculated from the rain drop experimental data obtained when the sample was located on the rigid support and the water droplets of mass 50 mg were used. Since the

Figure 5. Generated peak voltage with respect to various free oscillating sample lengths of PVDF strips.

water drops exert bending stress in the sample, the stress is in the longitudinal direction of the strip and perpendicular to the polarization direction, hence the piezoelectric voltage coefficient was calculated in the 31-mode (equation (1)) and found to be 205 × 10−3 V m N−1 , which is close to the value previously reported by Sessler [22, 23] and in the Piezo Sensors Technical Manual (216 × 10−3 V m N−1 ) [11]. The small difference in the values may be caused by the protective lamination of the samples used in this study. Taking the rain drop test results into consideration, to determine the optimum length for the highest voltage generation the same test was repeated on the longer PVDF sample by varying the free length of the sample from the support. 2, 4, 6, 8, 10, and 12 cm of the sample length was left free to oscillate and the rest was located on the rigid support then subjected to falling rain drops. Generated peak voltage with respect to various free oscillating sample lengths of PVDF 4


D Vatansever et al

by attaching a flexible aerofoil to the end of a piezoelectric material and subjecting it to the wind created by a wind tunnel by Bryant et al [24–27]. Fluttering of the aerofoil caused the flexing of the piezoelectric material and as a result generated 2 mW energy at a wind speed of 2 m s−1 . In our work, current and voltage drop at the various loads (from 0.99 k to 3.14 M) and wind speeds were measured for the piezoelectric samples and there was no additional structure to accelerate the oscillation of the piezoelectric samples other than the wind turbine created in the wind tunnel. Harvestable power was calculated using equation (2). The energy generated by the shorter polymer sample was in nanowatts; however, the other samples produced microwatts. Higher current readings were recorded at higher wind speeds for all piezo samples. At the highest wind speed of 10 m s−1 the power generated by the longer PVDF piezoelectric specimen was 93.6 μW, while 6.5 μW and 3.6 μW were recorded for the PZT-single layer and PZT-bimorph samples, respectively. This indicates that the piezo-activation of more flexible structures requires less force than the stiffer structures. The measured power density of 16.2 μW cm−3 for the long PVDF sample (LDT4-28K), at a wind speed of 5 m s−1 , was more than three times higher than that reported by Priya (5.1 μW cm−3 ) for a small wind mill at the same wind speed [28, 29]. Furthermore, at 10 m s−1 wind speed, the power density of the long PVDF sample reached a value of 157.9 μW cm−3 , whilst it was only 9.67 μW cm−3 for the PZT-single layer and 2.28 μW cm−3 for the PZT-bimorph. These results show that under low impact conditions the flexible polymer based piezoelectric structures may be the preferred alternatives in energy scavenging applications.

Figure 6. Voltage responses of PVDF and PZT composite films with respect to various wind speeds.

strips is presented in figure 5. The voltage output increases with an increase in the free sample length up to 4 cm then starts to decline as the free length is increased beyond this value. The recorded voltage outputs showed that a 4 cm sample length is the optimum for the highest voltage generation, which is equivalent to the length of the shorter PVDF sample. Figure 6 shows the generated peak voltage output of PVDF and PZT composite samples over a range of wind speeds. The results presented show that initially the voltage generation of the longer PVDF film has a linear increase with increase in the wind speed and then tends to saturate at higher wind speed values. It was also observed that the turbulence created in the wind tunnel was able to sustain the oscillations of the sample, which contributed to continuous voltage generation by the samples. The longer PVDF sample (LDT4-28K) generated a maximum voltage of 61.6 V at 10 m s−1 (36 km h−1 ) wind speed while the shorter PVDF specimen (LDT1-28K) initially generated smaller voltage at the lower wind speed values and the voltage generated at 10 m s−1 was somewhat smaller (56.5 V) than that produced by LDT4-28K. However, enhancement in the voltage output of the shorter polymer sample was almost exponential at wind speeds higher than 2 m s−1 . For LDT1-28K the voltage output more than quadrupled with each doubling in the wind speed value. This suggests that it may be possible to achieve higher voltage outputs from the shorter PVDF sample with further increase in the wind speed values beyond 10 m s−1 . However, due to the limitations of our experiment setup it was not possible to achieve wind speeds higher than 10 m s−1 . Figure 6 also shows the effect of wind speed on voltage generation for the PZT composite structures. These results show that at lower wind speeds both of the ceramic based piezoelectric specimens and the shorter PVDF sample generated a similar voltage, but the final voltage generated by the PZT samples was much lower than the PVDF based structures. Furthermore, it was also observed that the bimorph PZT structure generated a lower voltage than the single layer PZT structure. At the highest wind speed, the PZT-single layer composite sample generated nearly 45 V while the peak voltage generation of the bimorph PZT was about 40 V. Energy harvesting from wind using piezoelectric materials has been studied by other researchers with different setups. Power generation from piezoelectric materials has been tested

5. Conclusions In this study the peak voltages generated by various piezoelectric materials by varying different parameters such as wind speed, water droplet weight and releasing height were investigated. Water drops released from greater heights and with larger mass caused higher voltage generation. The voltage generated by the shorter PVDF sample was higher for the water drop experiment because the reaction of the shorter film to the forward strain is faster, which is critical for higher voltage/power generation. The voltage generated by the longer film was higher in the wind tunnel experiment due to the larger surface area of the film on which the wind exerts a larger force. It may be concluded that the voltage generated in various energy scavenging methods depends on the geometry and type of the films used. The results of this study show that under certain conditions polymer based piezoelectric materials produce higher peak voltage/power than ceramic piezoelectric materials when subjected to light impact and moderate wind speeds. This work has also demonstrated that the production of energy from renewable sources for utilization in low power electronic devices in outdoor applications is possible by using flexible polymer based piezoelectric structures. 5


D Vatansever et al

Acknowledgments

[12] Hellwege K H and Hellwege A M (ed) 1996 Ferroelectrics and Related Substances: Oxides (Landolt Bornstein New Series) (Berlin: Springer) [13] Sencadas V, Filho R G and Lanceros-Mendez S 2006 J. Non-Cryst. Solids 352 2226–9 [14] De-Qing Z 2008 Chin. Phys. Lett. 25 4410 [15] Jain A, Kumar J S, Mahapatra D R and Kumar H H 2010 Proc. SPIE 7647 76472C [16] APC International Ltd. 2002 Piezoelectric Ceramics: Principles and Applications [17] Starner T 1996 IBM Syst. J. 35 1–12 [18] Hausler E and Stein E 1984 Ferroelectrics 60 277–82 [19] Hadimani R L, Siores E and Vatansever D 2010 Piezoelectric Polymer Element and Production Method and Apparatus Patent application no. GB1015399.7 Submitted, September [20] Imeson A C, Vis R and de Water E 1981 Catena 8 83–96 [21] Weishaupt U 1987 Int. J. Impact Eng. 5 663–70 [22] Sessler G M (ed) 1980 Electrets, Topics in Applied Physics (Berlin: Springer) [23] Sessler G M 1981 J. Acoust. Soc. Am. 70 1596–608 [24] Bryant M and Garcia E 2009 Proc. SPIE 7288 728812 [25] Bryant M and Garcia E 2009 Proc. SPIE 7493 74931W [26] Bryant M, Fang A and Garcia E 2010 Proc. SPIE 7643 764317 [27] Bryant M, Fang A and Garcia E 2011 J. Vib. Acoust. 133 1 [28] Priya S 2005 Appl. Phys. Lett. 87 184101 [29] Li S and Lipson H 2009 Proc. ASME 2009 Conf. on Smart Materials, Adaptive Intelligent Systems SMASIS 2009 Sept. 20–24 (Oxnard, CA)

This research was supported by the Northwest Regional Development Agency through the Knowledge Centre for Material Chemistry (KCMC).

References [1] Umeda M, Nakamura K and Ueha S 1997 Japan. J. Appl. Phys. 36 3146–51 [2] Sodano H A, Inman D J and Park G 2004 Shock Vib. Dig. 36 197–205 [3] Sodano H A, Inman D J and Park G 2005 J. Intell. Mater. Syst. Struct. 16 799–807 [4] Mateu L and Moll F 2005 J. Intell. Mater. Syst. Struct. 16 835–45 [5] Mateu L and Moll F 2006 Sensors Actuators A 132 302–10 [6] Kymissis J, Kendall C, Paradiso J A and Gershenfeld N 1998 ISWC ’98: Proc. 2nd IEEE Int. Symp. on Wearable Computers IEEE Computer Society, p 132 [7] Shenck N S and Paradiso J A 2005 IEEE Micro 21 30–42 [8] Patel I, Siores E and Shah T 2010 Sensors Actuators A 159 213–8 [9] Swallow L M, Luo J K, Siores E, Patel I and Dodds D 2008 Smart Mater. Struct. 17 025017 [10] Granstrom J, Feenstra J, Sodano H A and Farinholt K 2007 Smart Mater. Struct. 16 1810–20 [11] Measurement Specialties, Inc. (MEAS) MSI Piezo Sensors Technical Manual 1999

6