See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/343782014

Experimental investigation of energy harvesting eel in the wake of bluff body

under ocean waves

Article in Proceedings of the Institution of Mechanical Engineers Part M Journal of Engineering for the Maritime Environment · August 2020
DOI: 10.1177/1475090220949334

CITATIONS READS

9 222

7 authors, including:

Usman Latif Ehtisham Ali

Duke University National University of Sciences and Technology
14 PUBLICATIONS 113 CITATIONS 1 PUBLICATION 9 CITATIONS

SEE PROFILE SEE PROFILE

Emad Uddin Zaib Ali

National University of Sciences and Technology National University of Sciences and Technology
83 PUBLICATIONS 572 CITATIONS 36 PUBLICATIONS 224 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Collaborative Research: Experiment, Theory, and Simulation of Aeroelastic Limit Cycle Oscillations for Energy Harvesting Applications. View project

Effect of WPC treatment on valve train performance in a real prodcution engine head View project

All content following this page was uploaded by Usman Latif on 04 December 2021.

The user has requested enhancement of the downloaded file.

Original Article

Proc IMechE Part M:

J Engineering for the Maritime Environment
1–12
Experimental investigation of energy Ó IMechE 2020
Article reuse guidelines:
harvesting eel in the wake of bluff body sagepub.com/journals-permissions
DOI: 10.1177/1475090220949334

under ocean waves journals.sagepub.com/home/pim

Usman Latif1, Ehtisham Ali1, Emad Uddin1 , Zaib Ali1,

Muhammad Sajid1 , Samiur Rehman Shah1 and
Muhammad Yamin Younis2

Abstract
Investigation of the energy harvesting from deep water waves by using flexible piezoelectric eel in a controlled environ-
ment is studied. Energy harvesting potential is examined as a function of streamwise distance from the fixed cylinder and
spanwise gap along with the cylinder at different wave conditions. Output voltage and eel flapping behavior are depen-
dent on cylinder vortices caused by local wavelength and wave amplitude. Maximum energy is harvested when the eel is
placed near to the surface caused by high flapping amplitude and frequency. Similarly, at greater depth low flapping ampli-
tude is observed resulting in small output voltage. Maximum output voltages are found at the shorter wavelength and at
a streamwise distance of gx = 1.25 (where gx is the ratio of spacing ‘‘S’’ between cylinder and eel to the diameter of cylin-
der ‘‘D’’) for all spanwise gaps along with the cylinder and minimum voltages are calculated at a longer wavelength and
streamwise distance gx = 0.75. An increase of 65% in energy harvesting is observed by switching longer wavelengths (l)
to a shorter one and changing the piezo-eel spanwise gap from deep to the shallow depth. Whereas, an increase of
31.5% was found by keeping wavelength constant and changing the spanwise gap of eel. Furthermore, it is observed that
energy harvesting from the wake of a bluff body in the wavy motion of water is sensitive to the wavelength and wave
height.

Keywords
Wave energy harvesting, piezoelectric eel, bluff body, fluid-structure interaction

Date received: 8 March 2020; accepted: 1 July 2020

Introduction when fluid flows around a body.7 Several types of

underutilized energy exist in the ambient environment
Harvesting unexploited energy from the ambient fluid such as solar, mechanical vibrations, acoustic, thermal,
flow is available in abundance in the nature can be and fluid flow, and can be extracted using different
transformed it into beneficial energy. In recent times, transduction mechanisms like electrostatic,8 electro-
the focus of research has been shifted to the develop- magnetic,9 and piezoelectric.10 Among these, vibration/
ment of renewable and sustainable energy sources pressure driven energy harvester is quite popular and
instead of conventional fossil fuels.1 This results in the widely studied due to its easiness of application, high
rapid research in electrical systems in the form of devel-
opment of microelectromechanical systems (MEMS),
1
wireless sensors for monitoring structure health, indus- Department of Mechanical Engineering, School of Mechanical and
Manufacturing Engineering, National University of Sciences and
trial systems, human health, and so on.2–4 The self-
Technology, Islamabad, Pakistan
powering technologies can operate without batteries as 2
Department of Mechanical Engineering, Mirpur University of Science
these operate under the influence of flow induced vibra- and Technology, Mirpur, AJK, Pakistan
tions (FIV) and thus can easily be handled in remote
locations without the tedious process of maintenance Corresponding author:
Emad Uddin, Department of Mechanical Engineering, School of
resulting in reduction of operational cost.3–6 FIV is a Mechanical and Manufacturing Engineering, National University of
very common engineering phenomenon that is insti- Sciences and Technology, 318 A, Islamabad 46000, Pakistan.
gated by vortex shedding or aero-dynamic instability Email: emaduddin@smme.nust.edu.pk
2 Proc IMechE Part M: J Engineering for the Maritime Environment 00(0)

energy density, efficiency, and miniaturization perspec- flowing water by using a circular cylinder to create
tive.11 The piezoelectric transduction showed ample Karman vortex street.
advantages over its counterparts and gained significant Derakhshandeh and Alam25 have studied the wakes
attention for harvesting energy from vibrations of different geometry bluff bodies (circular, rectangular,
imposed by flowing fluid like vortex induced vibration square, D shape, and triangular) at different Reynolds.
(VIV), buffeting, galloping, and fluttering.12 Around Wakes of a circular cylinder are simple as compared
20% of increment in energy generation per year has with non-circular cylinders. Rezaei et al.26 presented
been observed for wind power globally.13 Wind and the numerical model for energy harvester using square-
solar power cover the large portion of renewable energy shaped tip mass under the influence of hard excitations
demand whereas there is very minor energy contribu- and base acceleration and showed efficiency improve-
tion from ocean waves in comparison to available ment in unimorph cantilever beam’s performance. They
potential.14 further addressed the limitations of narrowband by pla-
In the recent past, to harvest ocean wave energy cing similar kind of harvester in the wake of a bluff
remained under focus due to its distinctive characteris- body and observed wideband energy harvesting with
tics of concentrated, high power density, and widely significant outcomes.27 Hamlehdar et al.10 had pre-
available renewable energy but it is still lagging behind sented a review of energy harvesting using a piezoelec-
due to its high cost and maintenance. Extensive tric device. Research has been carried out to harvest
research has done to harvest wave energy using oscillat- energy from piezoelectric devices deformed in wind
ing water columns, oscillating body systems, and over- flow due to vortex-induced vibration, hydrodynamic
topping converter.15 Siegel et al.16 harvested ocean force-induced vibration, and galloping mechanism.
wave energy by using an orbital circular motion of the Different techniques and methods have been discov-
wave. Singh et al.17 have designed a system to harvest ered to harvest ocean wave energy using cantilever
ocean wave energy by using oscillating water column vibrator and vortex-induced vibrations. Tanaka et al.20
technique in deep ocean waves. With the advancement have investigated the experimental study of energy har-
in piezoelectric material for energy generation and due vesting using an array of the piezoelectric flexible
to its low cost and simplicity, research gained attention device when placed transversally and vertically to the
to harvest energy from waves using a piezoelectric orbital motion of the ocean wave. Mutsuda et al.28
material. Allen et al.18 presented the concept of using a investigated the output behavior of the piezoelectric
piezoelectric flag in the wake of a bluff body, the flat flexible device for the different wave height, wave
plate was used as a bluff body, and observed the effect period, wave steepness, device depth, and different
of the ratio of the length of the membrane to the dia- device thickness. Output voltage increases with increas-
meter of bluff body (L/D) on energy output along with ing wave amplitude, wave height, deformation rate,
the feasibility of placing a piezoelectric flag in the wake. and greater stiffness. The output voltage was calculated
Techet et al.19 deployed multiple flags behind a bluff for device depth location and more voltage generated
body and observed dynamical behavior and their influ- with the device completely submerged as compared
ence on one another by observing out of phase motion with the device half-submerged in water. Vinh and
so he suggested a minimum distance must be main- Kim29 replaced the piezoelectric film with ionic poly-
tained between flags. mer film and found the best result were due to its flexi-
Tanaka et al.20 presented an experimental study bility under low-frequency ocean waves.
regarding energy harvesting from the wave by using Van Den Bremer et al.30 have studied the trajectories
piezoelectric eels installed in an array in horizontal and of water particles for surface gravity waves visually in
vertical. The array of eels was exposed to direct interac- the experimental flume and their effect with depth
tion with waves at different wave heights and water below the free surface. It was observed that near the
depths. Shi et al.21 worked on polyethylene terephtha- surface, particle orbital motion was circular but at
late membrane with a low aspect ratio of 1.67 in the greater depth magnitude of orbit decrease and horse-
wake of a square cylinder and analyzed the effect of shoe motion of the particle was observed. Particle
Reynolds number 3200 \ Re \ 12000. Furthermore, image velocimetry was performed to study wakes
they proposed the three modes of flapping dynamics around a horizontal circular cylinder in the uniform
with a set of Reynolds number range including, cantile- flow.31,32 Liu et al.33 had performed numerical simula-
ver beam like flapping, the mixture of a cantilever and tion for waves and circular cylinder interaction. Yang
limited traveling wave-like flapping, and quasi-periodic and Rockwell34 and Ozgoren and Rockwell35 had stud-
flapping. Rezaei et al.22 studied the effect of softening ied the interaction of deep water waves with the fixed
and magnetic restoring forces on the performance of an vertical cylinder and operated particle image velocime-
acoustic-based energy harvesting system. Mutsuda try to investigate the wake behavior spanwise to the
et al.23 proposed a hybrid system for energy harvesting cylinder for different wave conditions. Patterns of vor-
from the ocean by combining floating wind power tur- tices are different along the span of the cylinder due to
bines with piezoelectric sheets to obtain energy from the variation of local wave amplitude. Wu and Lin36
tides and waves. Pan et al.24 performed simulation fol- demonstrated the experimental study of water waves
lowed by experimentation to harvest energy from interaction with the vertical circular cylinder. Although
Latif et al. 3

Table 1. Wave parameters. harvesting that can be employed anywhere in the ocea-
nic environment with ease.
Wave period Wavelength Wave height Steepness ratio
(s) (mm) (mm) H/l

0.71 790 60 0.075 Experimental setup

0.65 660 65 0.098
0.6 560 70 0.125 Experiments were conducted in a wave flume having
0.57 508 75 0.147 transparent test section of 2000 3 400 3 500 mm (l 3 w
3 h). Waves were generated using a flap-type wave
maker fixed at one end of the tunnel. Vertical mesh
type wave absorber was fixed at the other end to damp
extensive research was carried out to harvest energy in out the incoming waves and to reduce the wave reflec-
the ocean caused by uniform fluid flow in the wakes of tion. Schematic diagrams of the complete experimental
the bluff body but there are no reports of harvesting setup are described in Figures 1 and 2. A flexible
energy in the wake of a bluff body under waves to date. PVDF eel is used (DT2-052K with lead attachments,
This paper presents the analysis of energy harvesting Measurement Specialties, Inc.) for energy harvesting.
from different modes of the vortex in the wake of the The dimensions of the piezoelectric eel are 62 mm
cylinder under wave loading. Energy generation has length (L), 12 mm width (W), and thickness 64 mm (t).
been investigated by flapping of piezoelectric flexible eel The piezoelectric eel having the properties Young’s
in the wake of a fixed circular cylinder under ocean Modulus E = 1.38 GPa, Poisson’s ratio n = 0.46, den-
waves. The focus of the study is to analyze the eel flap- sity r = 1.75 3 103 kg/m3 and impedance Z = 1 MO is
ping behavior by altering streamwise distance gx and attached to a thin steel rod. The solid frame structure
spanwise gap gy from the vertical cylinder at different was used to hold the piezoelectric eel with a fixed lead-
wave conditions and its impact on energy harvesting. ing edge and free trailing edge capable of adjusting the
The current study shows an increase of 61% in energy position behind the fixed vertical cylinder. A circular
harvesting from the vortices generated by the bluff cylinder with a 25 mm diameter (D) was used as a bluff
body placed upstream. Thus making it feasible for prac- body and eel is mounted behind it.
tical application and it has added advantage of easy Waves with different parameters were generated as
implementation. The streamwise distance gx in the wake described in Table 1. For all wave conditions, the
of a cylinder was varied at different spanwise gaps to streamwise gap gx (ratio of the distance from the center
observe the appropriate position for optimal synchroni- of the cylinder to the cylinder diameter ‘‘S/D’’) is varied
zation of wake flow with eel. A range of wavelengths from 0.75 to 1.75 and the spanwise gap from gy = H/2
were experimentally studied with different periods and to 3H/2 where H is the wave height, as represented by
wave heights as described in Table 1, where the pro- the top view and side view in Figure 2(a) and (b),
posed system attains efficient performance. The current respectively. Undulating behavior of eel was monitored
study addresses main issues and key challenges from by using the high speed camera (Sony Cyber-shot
former studies and illustrates a new concept of energy DSC-RX100 IV) mounted under the transparent bed

Figure 1. Schematic of the experimental setup.

4 Proc IMechE Part M: J Engineering for the Maritime Environment 00(0)

Figure 2. (a) Schematic representation of piezoelectric eel and cylinder as a function of streamwise distance, dotted line showing
initial position of eel, whereas solid line represents its flapping with amplitude ‘‘A’’. (b) Schematic representation of piezoelectric eel
as a function of the spanwise gap along with the cylinder.

of the test section and recorded the videos at the rate of the next subsections. Furthermore, energy harvesting
50 fps due to viscous fluid. The test section was illumi- can also be represented in terms of power output using
nated with LED lights, Data acquisition was done by formula Pout = V2rms/Ropt instead of Vrms by just add-
using DAQ card (National Instruments, USB-6009) ing the optimal resistance in the circuit. If one knows
with LabViewÒ software, and sample rate was set to about the impedance of the system then optimal resis-
50 Hz. Matlab image processing technique was used to tance can easily be found out using maximum power
process the recorded videos and a flapping envelope is transfer theorem40 as it is 1 MO for the current study.
shown as a superposition of the captured images.
Case-1 (gy = H/2)
Results and discussion Piezoelectric eel spanwise gap was fixed at gy = H/2
Strain in the piezoelectric membrane is generated by (near the water surface) and dynamics were investigated
placing it in the wake of a cylindrical bluff body to get by changing streamwise distance gx at different wave
the benefit of external force to vibrate applied by conditions. At fixed gy, cylinder vortices and piezoelec-
shedded vortices. In addition to this, the piezoelectric tric eel deformation mainly depend upon wavelength l,
eel is also forced from the orbital motion of water par- wave height H, and streamwise distance from the bluff
ticles as it is also under the influence of wave loading. body gx, as shown in Figure 3. Maximum and mini-
Flow kinetic energy is utilized to generate time- mum peak to peak flapping displacement is represented
dependent strain in the eel which then is converted to by a superimposed images of the eel motion at different
electrical energy due to piezoelectric effect. The waves streamwise distances gx and wave conditions in Figure
having different characteristics like wavelength, period, 4(a’) and (b’), with their corresponding time scale his-
height, and vortices generated by the bluff body induce tory of flapping for 10 s as shown in Figure 4(a) and
time-varying forces on the eel causing it to undergo (b). At longer wavelength for a range of streamwise
deformation, which causes the development of an elec- gap, corresponding wave height is small, resulting in a
trical charge. The periodic loading for energy genera- small scale non-phase-locked vortex formation which
tion is achieved by the alternate vortex shedding from undergoes small deformation in the piezo eel. At a
two sides of the cylinder, which is the function of the shorter wavelength for a range of streamwise gap, wave
piezoelectric coupling coefficient and strain change height increases which led to the formation of strong
rate.37 The energy harvested by the piezoelectric eel is a vorticity concentration and thus increases strain energy
function of eel thickness, flapping frequency, the capa- in piezoelectric eel. This behavior is due to the circular
citance of the piezoelectric layer, and oscillating ampli- orbital motion of water particles largely effected by
tude.19 Flapping frequency and amplitude are wave height as the orbital diameter is the height of the
highlighted as the key parameters on which the wave.41
extracted energy depends.38,39 It is expected that the Figure 3 shows the output voltage, peak to peak
flapping frequencies and oscillating amplitudes with amplitude, and flapping frequency response at fixed
varying wave parameters will enhance energy harvest- span-wise gap gy while altering wavelength, wave
ing. Both of these factors are associated with the output height, and streamwise gap. A general trend of increas-
as P = yt 3 f3 where yt corresponds to the displacement ing voltage for decreasing wavelength can be observed
of eel and f is its oscillation frequency.4 Hence, flapping for all streamwise gaps gx. Flapping frequency and
behavior, that is, amplitude (A/L) and frequency (FF) peak to peak amplitude shows reverse behavior for an
are analyzed for generated voltages against combined increasing streamwise gap gx which describes the poor
loading of waves and vortex shedding and presented in coupling of piezoelectric eel with vortices. A vortex
Latif et al. 5

Figure 3. (a) Root mean square voltage, (b) flapping amplitude A/L, and (c) flapping frequency, as a function of wavelength and
streamwise distance gx for the fixed spanwise gap (gy = H/2).

formed adjacent to the surface of the cylinder which decreases the diameter of particle orbital motion,41
grows to a specific distance and finally dissipates their which results in low peak to peak amplitude A/L as
energy while gap increases at fixed depth.34 Hence, the compared to case-1. Figure 5 shows the surface plot of
streamwise gape plays a key role in the performance of voltages Vrms, flapping amplitude, and frequency.
energy harvester. Point ‘‘mx’’ (l = 508 mm, gx = 1.25) Furthermore, in Figure 5(a) point ‘‘mx’’ (l = 508 mm,
in Figure 3(a) represents the maximum output voltages gx = 1.25) indicates a maximum root mean square vol-
(Vrms = 1.43 v) followed by a large flapping amplitude tage Vrms = 1.28 v caused by flapping amplitude A/
(A/L = 0.48) and flapping frequency (FF = 0.06). Eel L = 0.4 and flapping frequency FF = 0.06. Figure 6(a)
envelope (trace of flapping of eel) is shown in Figure and (a’) show their corresponding peak to peak ampli-
4(a’) and its evolution of tail position time scale tude real-time response and eel envelope, respectively.
response is shown in Figure 4(a). Higher flapping At shorter wavelength and close to the cylinder
amplitude and frequency cause a higher strain rate gx \ 1.5, cylinder wakes get dominant which results in
resulting in high voltage generation. Similarly, point large flapping amplitude, but maximum peak to peak
‘‘mn’’ (l = 790 mm, gx = 1.75) in Figure 3(a) shows the amplitude is less than the previous case due to a
minimum output voltages (Vrms = 0.56 v) generated decrease in particle orbital diameter at greater depth.
caused by the low peak to peak amplitude (A/L = 0.11) Along the span of the cylinder, patterns of velocity and
and flapping frequency (FF = 0.05). Figure 4(b’) shows vorticity decrease due to a decrease in local wavelength
flapping envelop and corresponding time scale ampli- beneath the free surface.34
tude is shown in Figure 4(b). Whereas point ‘‘mn’’ (l = 790 mm, gx = 0.75) in
Figure 5(a) indicates the minimum root mean square
voltage (Vrms = 0.6 v) due to minimum peak to peak
Case-2 (gy = H) amplitude A/L = 0.14 and flapping frequency
Piezoelectric eel spanwise gap increased from gy = H/2 FF = 0.04 as represented by the envelope of superim-
to H and eel flapping behavior was analyzed under the posed images in Figure 6(b’) and evolution of tail posi-
same wave conditions and streamwise distance gx as tion as a function of time in Figure 6(b). Longer
used for case-1. An increase in the submergence depth wavelength with smaller wave amplitude produced less
6 Proc IMechE Part M: J Engineering for the Maritime Environment 00(0)

Figure 4. (a) Flapping behavior of tail position and (a’) corresponding envelop of the eel at gx = 1.25 and l = 508 mm.
(b) Flapping behavior of tail position and (b’) corresponding envelop of the eel at gx = 1.75 and l = 790 mm.

Figure 5. (a) Root mean square voltage, (b) peak to peak amplitude, and (c) flapping frequency, as a function of wavelength and
streamwise distance gx for the fixed spanwise gap (gy = H).
Latif et al. 7

Figure 6. (a) Flapping behavior of tail position and (a’) corresponding envelop of the eel at gx = 1.25 and l = 508 mm.
(b) Flapping behavior of tail position and (b’) corresponding envelop of the eel at gx = 0.75 and l = 790 mm.

strain in the eel. As can be observed in the surface plot, horseshoe-shaped induced by return flow.30 Point ‘‘mx’’
longer the wavelength the lesser the deformation pro- (wavelength l = 508 mm, gx = 1.25) in Figure 7(a)
duced in the piezoelectric eel. As a result, a good vol- shows the maximum output voltage Vrms = 0.98 v
tage region can be observed until wavelength caused by their corresponding peak to peak amplitude
l = 660 mm for all range of streamwise distance gx but A/L = 0.27 as represented in Figure 7(b) and flapping
further increasing the wavelength causes a drop in vol- frequency FF = 0.06 in Figure 7(c). Figure 8 illustrates
tage region up to l = 790 mm. the amplitude response and eel flapping envelop at
point ‘‘mx’’ respectively. At gx = 1.25, a good voltage
region is observed for all wavelengths. In larger depth
Case-3 (gy = 3H/2) beneath the free surface, optimal flapping amplitude
Piezoelectric eel spanwise gap was further increased to reached and surpassed, due to the relationship between
maximum gy of 3H/2 and eel flapping dynamics were wave velocity, fluid forces, and fluid damping.20
observed under the same wave conditions and stream- Figure 7(c) represents the flapping frequency which
wise gap. Case-3 is dissimilar to previous cases due to became the focus of interest. The high flapping region
the fragile wave effect at a deeper level. The energy of can be observed at the shorter wavelength at all stream-
the wave decayed as we gradually increased the depth. wise gaps unlike the previous cases. This is because
Water particles become unaffected as depth approaches water particle orbital diameter decreases progressively
to l/2.41 From Figure 7(a) it can be observed that the with depth, as a result, smaller orbit caused small vor-
maximum voltage generated caused by eel deformation tices in the wake region. Point ‘‘mn’’ (l = 790 mm, gx =
as represented in Figure 7(b) is less than the previous 0.75) in Figure 7(a) indicates the minimum voltage gen-
cases due to weak vortex formation at greater depth. eration Vrms = 0.5 v at wavelength, illustrated by the
At larger depth, the magnitude of the particle orbits flapping envelope in Figure 8(b’) and evolution of tail
decrease and orbital circular motion is converted into position in Figure 8(b). At higher submergence depth,
8 Proc IMechE Part M: J Engineering for the Maritime Environment 00(0)

Figure 7. (a) Root mean square voltage, (b) peak to peak amplitude (A/L), and (c) flapping frequency, as a function of wavelength
and streamwise gap gx for the fixed spanwise gap (gy = 3H/2).

low amplitude wave results in low flapping amplitude and l3–4. Further increase in gap provides the optimal
A/L = 0.08 of eel and minimum flapping frequency point in the wake for both the flags where the system is
FF = 0.05 as shown in Figure 7(b) and (c), respectively. fully synchronized with the oncoming vortices from the
upstream body which results in high strain rate and
thus energy harvesting. Further increase in streamwise
Effect of the streamwise gap on the system’s optimal gap showed a significant change in output voltages
performance which shows that gap for the rear flag is not appropri-
Figure 9 shows the response of the proposed energy ate for proper coupling with wake flow of the upstream
harvesting system, piezoelectric eel behind a cylindrical body and it is interacting destructively resultantly a
bluff body, at different streamwise gaps gx. The stream- drop in output voltages is recorded. So the optimal
wise gap plays a key role in achieving systems optimal streamwise gap for front and the rear flag is found to
response at which maximum energy output is obtained be gx = 1.25. After this system starts to lose its connec-
for piezoelectric eel, that is, gx = 1.25. At low spacing tivity with wakes and a gradual drop in energy output
gx = 0.75, the system showed poor performance, hence is observed with further increase in streamwise gap.
yielded low output voltages as shown in Figure 10.
Further increase in the gap doesn’t show any gain
except for shorter wavelength l. It’s interesting to note
Overall annotation
that at some points the system showed anomalous Piezoelectric eel dynamics were affected by changing
behavior. The piezoelectric eel showed a substantial the spanwise gap below the free surface from gy = H/2
gain in output voltages for l1–4 (where l is the wave- to 3H/2 and streamwise distance from gx = 0.75 to 1.75
length, 1 represents to case-1 where gy = H/2 and 4 for a range of wavelength l = 508 mm to 790 mm.
represent to fourth wavelength i.e. l = 508 mm) at Maximum voltage, Vrms = 1.43 v, was achieved in case-
gx = 1.0. When the gap is increased from 1.0 to 1.25, 1 at gx = 1.25 and l = 508 mm where the eel was placed
the system showed an affirmative response and near to the surface. As the concentration of velocity
sharp gain in output specifically for the case of l2–4 and vorticity in the wake of the circular cylinder highly
Latif et al. 9

Figure 8. (a) Flapping behavior of tail position and (a’) corresponding envelop of the eel at gx = 1.25 and l = 508 mm. (b) Flapping
behavior of tail position and (b’) corresponding envelop of the eel at gx = 0.75 and l = 790 mm.

dependent on local wavelength and wave amplitude.34 gx = 1.25. Similarly, the comparison of the output vol-
Furthermore, the wave-particle orbit diameter is large tage at fixed streamwise distance gx = 0.75 is shown in
near the surface and decreases gradually with depth.30 Figure 10(b).
Although, the three-dimensional patterns of vortex for- A general trend of high voltage region was observed
mations are complex in nature along the span of the between streamwise distance gx = 1 ; 1.5 and wave-
cylinder due to the reason that wave-particle trajec- length l 4 660 mm for all spanwise gaps as shown in
tories are circular rather than unidirectional and their Figure 9. While at longer wavelength l . 660 mm, a
orbital axes are orthogonal to the cylinder axis. very low voltage region was observed for all streamwise
Similarly, minimum voltage (Vrms = 0.5 v) was and spanwise distance as shown in Figures 3, 5, and 7.
obtained at maximum spanwise gap gy = 3H/2 due to a Further analysis of Figure 9 illustrates that the gx =
decrease in orbital diameter and small wave affect. 1.25 is an optimal streamwise distance from the cylin-
Vortex formation is closely allied with depth along the der at which maximum energy harvesting can be
span of the vertical cylinder.35 Figures 3, 5, and 7 pro- achieved by taking advantage of wake flow and beyond
vide the comparison of voltage Vrms, peak amplitude this wake flow does not connect with piezoelectric eel
(A/L), and flapping frequency (FF) for all cases. The resulting in poor or no energy output. Close to the
voltage region for cases 1 and 2 is similar in both cases cylinder eel flapped in the wakes of the cylinder and by
following with similar flapping amplitude region and displacing it away from the cylinder to gx . 1.25,
frequency region. At the shorter wavelength of wakes dissipated their energy, and eel does not get ben-
l = 508 mm, high voltage was observed for all gx and efitted from oncoming vortices. Consequently, eel
maximum at gx = 1.25. In contrast, a low voltage flapped only due to the orbital circular motion of water
region was observed in case 3 for all gx except at gx = particles. Apart from this, the conflicting behavior of
1.25 where a measurable voltage region was obtained flapping frequency for case 3 is due to the small orbital
for all wavelengths. Figure 10(a) provides the compari- diameter of water particles at greater depth.
son of output voltage as a function of wavelength for The study of oscillating piezo eel excited by fluid
all span-wise gap gy at a fixed streamwise distance of movement is usually represented in terms of free and
10 Proc IMechE Part M: J Engineering for the Maritime Environment 00(0)

Figure 9. Output voltages as a function of the streamwise gap

(gx) for different wavelengths and spanwise gaps (gy).

forced vibration. Allen et al.18 stated that the if anyone

considers a basic picture of the movement of the piezo-
eel in the transverse direction, y(x, t), as a mechanical
system of mass m exposed to an exterior forcing field,
then maximum flow energy transfer to the piezoelectric
eel resulting in higher energy harvesting. Eel’s oscilla-
tion is the result of a combination of both free and
forced vibration, Figure 10. Output voltage as a function of wavelength for all
spanwise gaps at (a) gx = 1.25 and (b) gx = 0.75.
 
m y (t, x) + H( y , y, x, t) = f1 + f2 ð1Þ
Where, streamwise gap. For all cases, output voltage and eel
f1 is the free vibration caused by the self-induced behavior were studied and represented. It was observed
effects of the unsteady wake that the wake of the bluff body was highly dependent
f2 is the forced vibration and sum of fa (caused by on wave height and local wavelength beneath the free
the vortex shedding/interaction) and denotes the effect surface. Near the free surface voltage obtained was
of the exterior forcing field that has its source distant maximum caused by the high peak to peak amplitude
from the body being excited. This type of external force and flapping frequency due to larger particle orbital
is classified as extraneously induced excitation (EIE) diameter and strong vortex interaction with piezoelec-
and is a feature of gusting as the buffeting and vortex tric eel. It was observed that the highest gain in energy
induced vibration, both belongs to the family of forced harvesting result was obtained at streamwise distance
vibration.42 gx = 1.25 and at a shorter wavelength for all cases.
fb (contribution of wave force and can be predicted Similarly, the minimum voltage was calculated at a lon-
by using Morison equation). ger wavelength and at a higher value of gy. As the sub-
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi mergence depth increases, the output voltage decreases,
pD2  1
f b = cd r u + ci rDu u2 + w2 and the minimum voltage was calculated at the maxi-
4 2 mum spanwise gap. At greater depth, orbital diameter
So, it’s clear from equation (1) that the higher the force, decreases, and the concentration of vorticity reduces
the higher would be displacement in terms of deforma- due to a weak wave effect. It was also observed that the
tion of the piezoelectric eel, which excites or boosts the forcing components or vibration can be increased to
piezo effect, consequently, the higher would be energy obtain maximum energy output by deploying bluff on
harvesting. the upstream side. Because by doing this, we get force
from vortex shedding in addition to the orbital motion
of wavy particles to generate maximum strain in piezo-
Conclusion electric eel. By using the piezoelectric eel, an energy har-
In this research work, energy harvesting was studied by vesting system studied in this paper, which can be used
flapping of flexible piezoelectric eel caused by vortex- to power low energy-consuming electronic devices and
induced vibration (VIV) behind a fixed vertical circular sensors in the ocean buoys in deep water waves. It has
cylinder under different wave conditions. Three cases added benefit of working in any wind/wave orientation
were performed for a different spanwise gap along with as previously done studies were limited to place piezo-
the cylinder for all range of wavelengths and electric eel at an angle to incoming wave without a bluff
Latif et al. 11

body. So, the proposed system has an advantage over 10. Hamlehdar M, Kasaeian A and Safaei MR. Energy har-
previous studies. vesting from fluids flow using piezoelectrics: a critical
review. Renew Energy 2019; 143: 1826–1838.
11. Wang J, Geng L, Ding L, et al. The state-of-the-art
Future work review on energy harvesting from flow-induced vibra-
tions. Appl Energy 2020; 267: 114902.
For future work, we propose an on-field implementa- 12. Abdelkefi A. Aeroelastic energy harvesting: a review. Int
tion of the proposed system. Several offshore struc- J Eng Sci 2016; 100: 112–135.
tures, riser pipes, and buoys are installed and are in 13. Solaun K and Cerdá E. Impacts of climate change on
operation worldwide, so no need to use a specialized wind energy power – Four wind farms in Spain. Renew
bluff body with the special arrangement and it can be Energy 2020; 145: 1306–1316.
14. REN 21 Renewables Now. Renewables global status
implemented with the stated structures.
report 2019, vol. 105, 2019, https://www.ren21.net/
reports/global-status-report/; https://www.ren21.net/wp-
Declaration of conflicting interests content/uploads/2019/05/gsr_2019_full_report_en.pdf
15. Falcão AF de O. Wave energy utilization: a review of
The author(s) declared no potential conflicts of interest
the technologies. Renew Sustain Energy Rev 2010; 14(3):
with respect to the research, authorship, and/or publi- 899–918.
cation of this article. 16. Siegel SG, Fagley C and Nowlin S. Experimental wave
termination in a 2D wave tunnel using a cycloidal wave
Funding energy converter. Appl Ocean Res 2012; 38: 92–99.
17. Singh U, Abdussamie N and Hore J. Hydrodynamic per-
The author(s) received no financial support for the formance of a floating offshore OWC wave energy con-
research, authorship, and/or publication of this article. verter: an experimental study. Renew Sustain Energy Rev
2020; 117: 109501.
ORCID iDs 18. Allen JJ and Smits AJ. Energy harvesting eel. J Fluids
Struct 2001; 15(3–4): 629–640.
Emad Uddin https://orcid.org/0000-0002-5177-4554 19. Techet AH, Allen JJ and Smits AJ. Piezoelectric eels for
Muhammad Sajid https://orcid.org/0000-0001-6281- energy harvesting in the Ocean. Proc Int Offshore Polar
3136 Eng Conf 2002; 12: 713–718.
20. Tanaka Y, Oko T, Mutsuda H, et al. An experimental
study of wave power generation using a flexible piezo-
References
electric device. J Ocean Wind Energy 2015; 2: 28–36.
1. Karim M, Munir A, Karim M, et al. Energy revolution 21. Shi S, New TH and Liu Y. Flapping dynamics of a low
for our common future: an evaluation of the emerging aspect-ratio energy-harvesting membrane immersed in
international renewable energy law. Energies 2018; 11(7): a square cylinder wake. Exp Therm Fluid Sci 2013; 46:
1769. 151–161.
2. Sultana A, Alam MM, Middya TR, et al. A pyroelectric 22. Rezaei M, Talebitooti R and Friswell MI. Efficient
generator as a self-powered temperature sensor for sus- acoustic energy harvesting by deploying magnetic restor-
tainable thermal energy harvesting from waste heat and ing force. Smart Mater Struct 2019; 28: 105037.
human body heat. Appl Energy 2018; 221: 299–307. 23. Mutsuda H, Watanabe R, Azuma S, et al. OMAE2013-
3. Ottman GK, Hofmann HF, Bhatt AC, et al. Adaptive 10078. Ocean power generator using flexible piezoelectric
piezoelectric energy harvesting circuit for wireless remote device. In: Proceedings of the ASME 2013 32nd interna-
power supply. IEEE Trans Power Electron 2002; 17(5): tional conference on ocean, offshore and arctic engineering,
669–676. OMAE2013, Nantes, France, 9–14 June 2013, pp.1–9.
4. Mitcheson PD, Yeatman EM, Rao GK, et al. Energy New York: ASME International.
harvesting from human and machine motion for wireless 24. Pan F, Xu Z, Jin L, et al. Designed simulation and experi-
electronic devices. Proc IEEE 2008; 96: 1457–1486. ment of a piezoelectric energy harvesting system based on
5. Sudevalayam S and Kulkarni P. Energy harvesting sen- vortex-induced vibration. IEEE Trans Ind Appl 2017; 53:
sor nodes: survey and implications. IEEE Commun Surv 3890–3897.
Tutorials 2011; 13: 443–461. 25. Derakhshandeh JF and Alam MM. A review of bluff
6. Priya S. Advances in energy harvesting using low profile body wakes. Ocean Eng 2019; 182: 475–488.
piezoelectric transducers. J Electroceramics 2007; 19: 26. Rezaei M, Talebitooti R and Rahmanian S. Efficient energy
165–182. harvesting from nonlinear vibrations of PZT beam under
7. Blevins RD. Flow-induced vibration. New York: Van simultaneous resonances. Energy 2019; 182: 369–380.
Nostrand Reinhold, 1990. 27. Rezaei M and Talebitooti R. Wideband PZT energy har-
8. Lallart M, Pruvost S and Guyomar D. Electrostatic vesting from the wake of a bluff body in varying flow
energy harvesting enhancement using variable equivalent speeds. Int J Mech Sci 2019; 163: 105135.
permittivity. Phys Lett Sect A Gen At Solid State Phys 28. Mutsuda H, Tanaka Y, Doi Y, et al. Application of a
2011; 375(45): 3921–3924. flexible device coating with piezoelectric paint for har-
9. Beeby SP, Torah RN, Tudor MJ, et al. A micro electro- vesting wave energy. Ocean Eng 2019; 172: 170–182.
magnetic generator for vibration energy harvesting. J 29. Vinh ND and Kim HM. Ocean-based electricity generat-
Micromech Microeng 2007; 17(7): 1257–1265. ing system utilizing the electrochemical conversion of
12 Proc IMechE Part M: J Engineering for the Maritime Environment 00(0)

wave energy by ionic polymer-metal composites. Electro- 36. Wu H and Lin Y. An experimental investigation of the
chem Commun 2017; 75: 64–68. characteristics of a cylinder interacting with waves and
30. Van Den Bremer TS, Whittaker C, Calvert R, et al. flow in the dominant-load regime. Appl Ocean Res 2019;
Experimental study of particle trajectories below deep- 89: 154–175.
water surface gravity wave groups. J Fluid Mech 2019; 37. Doaré O and Michelin S. Piezoelectric coupling in
879: 168–186. energy-harvesting fluttering flexible plates: linear stability
31. Oshkai P and Rockwell D. Free surface wave interaction analysis and conversion efficiency. J Fluids Struct 2011;
with a horizontal cylinder. J Fluids Struct 1999; 13: 935– 27: 1357–1375.
954. 38. Michelin S and Doaré D. Energy harvesting efficiency of
32. Tumen Ozdil NF and Akilli H. Flow comparison around piezoelectric flags in axial flows. J Fluid Mech 2013; 714:
horizontal single and tandem cylinders at different 489–504.
immersion elevations. Ocean Eng 2019; 189: 106352. 39. Shoele K and Mittal R. Energy harvesting by flow-
33. Liu Z, Wan D and Hu C. Numerical investigation of reg- induced flutter in a simple model of an inverted piezo-
ular waves interaction with two fixed cylinders in tandem electric flag. J Fluid Mech 2016; 790: 582–606.
arrangement. Proc Int Conf Offshore Mech Arct Eng – 40. Poole C and Darwazeh I. Introduction. In: Microwave
OMAE 2018; 7A: 1–9. active circuit analysis and design. Amsterdam, The Neth-
34. Yang Y and Rockwell D. Interaction of a deep-water erlands: Elsevier, 2016.
wave with a vertical cylinder: flow structure and loading. 41. Thurman HV. Essentials of oceanography. 7th ed. Upper
J Fluid Mech 2004; 520: 267–295. Saddle River, NJ: Prentice Hall, 2002.
35. Ozgoren M and Rockwell D. Interaction of a deep-water 42. Armandei M, Fernandes AC and Rostami AB. Hydroe-
wave with a vertical cylinder at low Keulegan-Carpenter lastic buffeting assessment over a vertically hinged flat
number: transition from phase-locked modes of vortex plate. Exp Tech 2016; 40: 833–839.
formation. Phys Fluids 2004; 16: 2700–2703.

View publication stats