986 JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 36, NO.

4, FEBRUARY 15, 2018

Carbon Cantilever Beam Health Inspection Using a

Polymer Fiber Bragg Grating Array
Antreas Theodosiou , Student Member, IEEE, Michael Komodromos , Senior Member, IEEE,
and Kyriacos Kalli , Member, IEEE

Abstract—We demonstrate a quasi-distributed sensor for can- lower (∼60 dB/km) in the near infra-red (NIR), thus improv-
tilever health inspection measurements using a fiber Bragg grating ing the potential for POF-sensing applications. CYTOP has the
(FBG) array inscribed in a polymer optical fiber. The FBGs were same chemical backbone as PMMA but the C-H bonds, respon-
characterized and calibrated for axial strain, temperature, and
relative humidity prior to their mounting on a carbon cantilever sible for absorption at 1550 nm are replaced with C-F, shifting
beam, the tail rotor of a helicopter. By using the zero-crossing the vibrational resonance to lower frequencies and reducing the
demodulation algorithm, we recovered the time-dependent, wave- NIR wavelength loss [12]–[15]. The excellent optical proper-
length response from each Bragg grating sensor and the vibration ties of CYTOP make it incompatible with traditional UV laser
response of the beam was extracted. We used the response of the FBG inscription methods, as there is no significant photosensi-
beam to study how the addition of masses at different positions on
the beam influences the vibrational behavior and mimics the lo- tivity in the UV wavelength region. Additionally, commercially
cation of “damage” through the time-dependent results. We show available CYTOP polymer fibres are multimode in nature, which
that health inspection measurements are feasible with polymer- results in multiple peak gratings; complicating sensor interroga-
based fiber Bragg gratings, offering accurate and rapid detection tion [16], [17]. To overcome both of these potential difficulties,
of damage points on a structural beam. we use a direct-write, plane-by-plane femtosecond (fs) laser in-
Index Terms—Femtosecond laser, fibre Bragg gratings, polymer scription method for the fabrication of FBG sensors in CYTOP
fibre, structural monitoring, vibration measurements. optical fibre [18]–[21]. Fs-laser inscription can readily induce
I. INTRODUCTION refractive index changes in all transparent materials, such as
perfluorinated fibres. Here we inscribe FBGs in a multimode
OLYMER optical fibres (POFs) have received significant
P interest during the last two decades, and this is primarily
because of their low Young’s modulus, which is approximately
gradient index POF producing a controlled inscription size for
each grating plane. In order to reduce the number of fibre modes
coupling to the grating, we limit its spatial extent to the central
25 times less than conventional glass fibres. When coupled with part of the core, in the region where the gradient index profile
fibre Bragg grating (FBG) sensors the increased elasticity of- peaks. In this way we are able to excite the strongest lower
fers a high tuning range, improved strain and pressure sensi- order modes thereby generating single peak POF-FBG spectra.
tivity; whereas thermal measurements are generally limited to The grating sensitivity is calibrated for axial strain, temperature
80-100 °C [1]–[7]. Several industrial applications have been and relative humidity. We develop a 2-m long FBG array as
proposed for POF-FBGs [8]–[11], however, the optical attenua- a quasi-distributed sensor and apply it to the health monitor-
tion loss of these fibres is a major drawback that limits their use. ing of a carbon cantilever beam, the tail rotor of a helicopter.
The most commonly used polymer in fibre form, polymethyl The key advantages for the quasi-distributed sensor system is
methacrylate (PMMA), has a transmission loss exceeding 100 the intrinsic absolute wavelength encoding for data recovery,
dB/m at 1550 nm, and this restricts the operational fibre length the relatively low cost, an inherent simplicity of use and a fast
to a few centimetres. This problem can be addressed through response time for rapid data recovery.
the use of a perfluorinated POF based on the cyclic transparent The use of composite materials is on the increase in several
optical polymer (CYTOP), for which the loss is significantly industries, principally for aviation, aerospace and marine appli-
cations. Carbon fibre composites are typically used where high
Manuscript received July 28, 2017; revised September 29, 2017 and October strength and light weight are paramount. However, composite
21, 2017; accepted October 24, 2017. Date of publication October 31, 2017; rotor blades are susceptible to the typical damage characteristics
date of current version February 24, 2018. (Corresponding author: Kyriacos of any composite material such as delamination effects, matrix
Kalli.)
A. Theodosiou is with the Nanophotonics Research Laboratory, Cyprus Uni- cracking and fibre breakage. This has an impact on various prop-
versity of Technology, Limassol 3036, Cyprus, and also with Frederick Univer- erties of the composite rotor blade, affecting the rigidity, degree
sity, Nicosia 1036, Cyprus (e-mail: theodosiou.antreas@gmail.com). of deflection and frequency of vibration. Considering this re-
M. Komodromos is with Frederick University, Nicosia 1036, Cyprus (e-mail:
eng.km@frederick.ac.cy). quirement we studied the behaviour of a helicopter rear rotor
K. Kalli is with the Nanophotonics Research Laboratory, Cyprus University blade, an essential stabilising component for helicopter flight.
of Technology, Limassol 3036, Cyprus (e-mail: kyriacos.kalli@cut.ac.cy). In addition to rotating, the tail rotor blades also change pitch
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org. angle; counteracting airframe torque with anti-torque tail rotor
Digital Object Identifier 10.1109/JLT.2017.2768414 thrust. The tail rotor can fail if the rotor stops turning or the pitch

0733-8724 © 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications standards/publications/rights/index.html for more information.

Authorized licensed use limited to: UNIV ESTADUAL PAULISTA JULIO DE MESQUITA FILHO. Downloaded on August 14,2024 at 21:35:28 UTC from IEEE Xplore. Restrictions apply.
 THEODOSIOU et al.: CARBON CANTILEVER BEAM HEALTH INSPECTION USING A POLYMER FIBER BRAGG GRATING ARRAY 987

change mechanism stops functioning. It is vital to inspect the

health condition of key components such as air-blades, before,
after, and during the flight, as failure occurs on a short time
scale. Current inspection and monitoring methods for hollow
aluminium air blades [22]–[25], are unsuitable for use with car-
bon fibre composites. There is no way to reasonably estimate and
monitor the rotor health condition when the helicopter is flying.
The potential of fibre optic sensors for this type of application
is clear, given their small volume, light weight and immunity
to electromagnetic interference, and having no significant in-
fluence on the structure under test. Similar health monitoring
inspection can be applied to applications such as spacecraft,
Fig. 1. Reflection spectrum of a POF-FBG inscribed in the gradient-index
wind turbines, and civil structures [26]–[28]. We utilized the multimode CYTOP fibre, for a plane pitch of 2.315 μm.
specialized sensor array to study the behaviour and structural
health of the carbon cantilever, by monitoring the wavelength
response with respect to the time. In order to mimic signs of
rotor damage we attached small weights to different locations
on the cantilever beam, from which the dynamic mode shapes
of the beam were extracted. Furthermore, we adjusted the de-
gree of ‘damage’ by using different masses and distinguished
between them using the time-dependent wavelength response of
the gratings [29]. Our goal here is to observe the deviation from
the normal response, the degree of deflection. All the spectra
were recovered using a FBG demodulator and the wavelength
shifts of the FBGs with respect to time were processed using
the Zero-Crossing Algorithm (ZCA) [30]. The paper is struc-
tured as follows. Section II details the inscription process and
sensor calibration. Section III briefly discusses the operating
Fig. 2. POF-FBG wavelength shift for 1.3% elongation of its total length.
principles and the basic ZCA equations for data recovery. Fi-
nally, Section IV describes the experimental details and results
for health monitoring of the cantilever carbon beam. band light source (Thorlabs ASE730), coupled to a commercial
spectrometer was used to recover an optimised FBG reflection
spectrum. Fig. 1 shows clear multi-mode suppression and one
II. FIBRE BRAGG GRATING INSCRIPTION AND strong peak. The grating was located at 1560 nm with full width
CHARACTERISATION half maximum (FWHM) bandwidth of 1.67 nm.
The inscription setup combines a femtosecond laser system The CYTOP FBG was calibrated for temperature, axial strain
(HighQ laser femtoREGEN) (operating at 517 nm and produc- and relative humidity variations. The grating was fixed between
ing pulses of 220-fs duration) and an air-bearing translation two manual stages using epoxy glue and axial strain steps of
stage system (Aerotech) for accurate and controlled motion 700 με up to 12600 με were applied to the grating. The total
during the inscription process. The laser beam was focused wavelength shift for 12600 με or ∼1.3% strain of the grating
and directed into the fibre from above; through a long working was ∼18 nm, as shown in Fig. 2, giving strain sensitivity of
distance microscope objective mounted on a third translation 1.44 pm/με.
stage. The commercially available fibre had a core diameter of In order to measure the temperature sensitivity of the grating,
62.5 μm, with a cladding layer of 20 μm and an additional 490 a single mode glass fibre (SMF28) was connected to the polymer
μm polyester and polycarbonate jacket. The pulse energy at the sample using UV glue and placed in a climate chamber (the glue
exit of the laser was ∼100 nJ per pulse and a pulse picker set the joint was external to the chamber). The relative humidity of the
repetition rate to 5 kHz. Each grating was inscribed using the chamber was kept at a constant 23%, whereas the temperature
direct-write, plane-by-plane inscription method. This method was linearly increased from 35 °C up to 60 °C for 80 minutes
affords flexibility in creating two-dimensional refractive index and “snapshots” of the spectrum were recorded every 1 minute.
planes that have controllable width and depth. The gratings were The temperature response of the FBG (see Fig. 3) indicates a
fabricated with 300 planes, at the centre of the fibre core and sensitivity of 17.62 pm/°C.
the width of each plane (the plane is written transversely across We also considered the contribution of humidity, as this can
the fibre axis) was 15 μm to produce a clean, single-peak FBG have an impact on measurements over extended time periods.
spectrum. The laser delivered approximately 100 pulses per mi- The wavelength shift of POF-FBGs due to the water absorp-
cron during the inscription. The inscribed gratings were 4th tion is mainly based on two mechanisms. The first relates to
order and had a period of ∼2.3 μm, with small differences in the grating pitch change due to the swelling of the polymer
the period to arrive at discrete reflecting wavelengths. A broad- through water intake, and the second relates to the change in

Authorized licensed use limited to: UNIV ESTADUAL PAULISTA JULIO DE MESQUITA FILHO. Downloaded on August 14,2024 at 21:35:28 UTC from IEEE Xplore. Restrictions apply.
 988 JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 36, NO. 4, FEBRUARY 15, 2018

Fig. 5. The transformation of the maximum peak to zero crossing due to the
Fig. 3. Temperature response of POF-FBG inscribed in CYTOP fibre for 35 Hilbert transformation.
to 60 °C according to the climate chamber temperature sensor.

where Rref (l) is the autocorrelation of the reference spectrum.

We compute the Discrete Fourier Transform (DFT) of the au-
tocorrelated signal. The DFT signal is multiplied by a linear
filter that utilizes a 90 degrees phase shift, whilst keeping the
real value spectral components unchanged in order to utilise the
Hilbert Transformation (HT) [33].
Specifically, R (k) denotes the DFT of the signal multiplied
by the HT multiplier operator, H(k), that introduces the phase
shift in the frequency domain as shown below,
R (k) = H (k) .R (k) = −j.sgn (k) .R (k) , (2)
⎧
⎪ N
⎪
⎪ j, k = 1, 2, . . . , −1
⎪
⎨ 2
N
Fig. 4. Humidity response of the POF-FBG inscribed in CYTOP fibre with a H (k) = −j.sgn (k) = 0, k = 0, ,
⎪
⎪ 2
62.5-μm core diameter and 490-μm outer polyester and polycarbonate jacket. ⎪
⎪
⎩ −j, k = N +1, N +2, . . . , N −1
2 2
refractive index due to the increased water content. While the (3)
perfluorinated polymers such as the CYTOP are not hydrophilic,
the polycarbonate-based outer fibre jacket shows high water ab- From equation (2) we obtain the spectrum with positive fre-
sorption. Thus, swelling of the polycarbonate-based outer fibre quencies, whereas all negative frequencies are rejected. The HT
jacket is expected to affect the resonance wavelength of the of the auto-correlation was computed by applying the inverse-
grating [5]. We placed the sensor in the climate chamber where DFT as shown below, with k = 0, 1 . . . , M − 1 and M denotes
the chamber temperature was kept at a constant 30 °C, whereas the number of DFT points,
M −1
the relative humidity was changed from 50 to 90% with interval 1   j 2π k n
steps of 10%, with a stabilization period of 90 minutes for each R̃ref (n) = R ref (k) .e M , (4)
M
step. The corresponding humidity sensitivity was measured to k =0
be 14.7 pm/%RH, Fig. 4. 
R̃ref (l) is the Hilbert transform of Rref (l) and Rref (k) denote
the phase shifted signal of the Rref (l) as shown in equation (2).
III. ZCA DEMODULATION ALGORITHM The same procedure is repeated for the perturbed spectrum by
The ZCA was used to obtain the time-dependent wave- calculating the cross correlation product between the perturbed
length response of the POF-FBGs [30], offering improved and reference spectrum. The HT of the resulting cross corre-
accuracy compared to other common algorithms such as the lated signal transforms the maximum peak to a zero crossing,
cross-correlation algorithm and the centroid algorithm for mul- Fig. 5. The zero-crossing is calculated by using the equation
tiple peak spectra [31], [32]. Briefly the algorithm begins with of the straight line, y = ax + b and setting y = 0. The final
a set of samples of a reference FBG reflection spectrum. The wavelength shift is found by first subtracting the zero-crossing
reference spectrum is read from the sensor before any external coordinates of the reference (pref ), and the perturbed spec-
perturbation is applied to it. Subsequently, we find the autocor- trums (pnew ), and then multiplying with the sampling rate of the
relation of the reference, system,
2N
 −1 Δλ = (pnew − pref ) .dλ, (5)
Rref (l) = Yref (λl ) .Yref (λl + n ) , (1)
n =0 where dλ is the sampling wavelength rate.

Authorized licensed use limited to: UNIV ESTADUAL PAULISTA JULIO DE MESQUITA FILHO. Downloaded on August 14,2024 at 21:35:28 UTC from IEEE Xplore. Restrictions apply.
 THEODOSIOU et al.: CARBON CANTILEVER BEAM HEALTH INSPECTION USING A POLYMER FIBER BRAGG GRATING ARRAY 989

Fig. 6. Reflection spectrum of a multiple-FBG array recovered using a com-

mercial FBG demodulator spectrometer.

Fig. 7. Experimental setup for the cantilever health monitoring system. A

multiple FBG sensor array was mounted on a helicopter rear rotor blade. The
physical location of the FBGs on the beam is indicated with the numbers 1–4.

IV. CANTILEVER BEAM AND DAMAGE ADJUSTMENT

In order to measure the vibration response of the cantilever
beam we used a FBG array consisting of 4 gratings that were
physically separated by 8 cm. The fibre sample had total length
of 2 m. The FBGs were spectrally separated with small vari-
ations in their periods and hence operating wavelengths of
1548 nm, 1561 nm, 1570 nm, and 1580 nm, and the optimised
illumination of the grating array is shown in Fig. 6. We note
that the FWHM of the FBGs does not affect the interrogation
results as we track the peak wavelength shift using the ZCA
principles. The carbon fibre blade from the rear rotor of a heli- Fig. 8. Wavelength response of the polymer FBGs with respect to time for
gratings located according to Fig. 7, (a) at the base of the beam position 1,
copter, of dimensions 36 cm × 6 cm, was fixed to the edge of (b) position 2, (c) position 3, (d) position 4.
an optical table, with the FBG array mounted along the beam
length, Fig. 7. During the experiment, the beam was gently
pushed down and released in order to excite the first vibration to their physical position, extracting the vibrational beam shape
mode of the structure. The demodulator (Ibsen IMON 512 HS) directly.
captured data at frequencies up to 35 kHz for an integration We added weights along the length of the beam, close to the
time of 10 μs, and the time response is shown in Fig. 8. We FBGs position, to mimic “damage”, and compared their respec-
observe that the grating located at the fixed base of the beam tive vibrational displacements with and without the additional
[position 1, Fig. 8(a)] shows no measurable fluctuation, as ex- masses; all measurements were repeated four times. The nor-
pected, whereas amplitude excursions increase as we move to malised vibrational shapes are shown in Fig. 9(a)–(e) for differ-
the end of the beam [position 2–4, Fig. 8(b)–(d), respectively], ent time intervals while the reference vibrational shape shown
and we measure the wavelength shift of each grating according in Fig. 9(a) with no weight added to the beam. The minimum

Authorized licensed use limited to: UNIV ESTADUAL PAULISTA JULIO DE MESQUITA FILHO. Downloaded on August 14,2024 at 21:35:28 UTC from IEEE Xplore. Restrictions apply.
 990 JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 36, NO. 4, FEBRUARY 15, 2018

Fig. 10. Time-dependent FBG wavelength response for (a) no mass, (b) 17-g
mass, and (c) 34-g mass added on the beam at location 2 for the grating operating
at 1561 nm.

are shifted and this leads to a change in the beam’s vibration

response; now the contribution of the other FBGs becomes im-
portant in identifying the damage position [35], [36]. For exam-
Fig. 9. Response of the cantilever with the addition of weights to the beam for
ple, in our case the FBG in position 2 experiences compression
(a) No added weight, (b) weight at position 1, (c) at position 2, (d) at position when that in position 4 experiences expansion.
3, (e) at position 4. Another important factor for the health condition during the
inspection procedure of the beams is the level of damage. For
fluctuation was noted when weights were added to position 1 this reason, we attempted to ascertain the level of “damage”,
[see Fig. 9(b)] and the maximum when weights were added to repeating our measurements using different masses of, 0 g, 17
position 4 [see Fig. 9(e)]. Comparing Fig. 9(a)–(e) we observe g, and 34 g. The weights were fixed securely to position 2 of the
that significant information regarding the damage position is beam and we tracked the grating peak that was closest to that
given by the last point of the vibrational shape, which corre- point (operating at 1561 nm). From the time-dependent wave-
sponds to the position of the 4th FBG; in this particular case the length results shown in Fig. 10(a)–(c) for the three weights, we
maximum sensitivity is at the edge of the beam. It is well known observe a gradual shift off the fluctuation normal. The max-
that as damage increases the resonance frequencies of the beam imum peak-to-peak fluctuation for the first measured period

Authorized licensed use limited to: UNIV ESTADUAL PAULISTA JULIO DE MESQUITA FILHO. Downloaded on August 14,2024 at 21:35:28 UTC from IEEE Xplore. Restrictions apply.
 THEODOSIOU et al.: CARBON CANTILEVER BEAM HEALTH INSPECTION USING A POLYMER FIBER BRAGG GRATING ARRAY 991

for real-time monitoring during flight. However, the current

paper aims to show that low-loss CYTOP POFs is feasible to
use for such applications and not to present a refined health
monitoring method as the current technology is new and will
require further refinement to ensure the long-term accuracy and
robustness of the method.
Finally, we consider that the gradual development of perflu-
orinated low-loss POFs combined with FBG technology offers
a realistic solution for the use of polymers fibres in real-world
industrial applications, as realised by recent developments [34].

ACKNOWLEDGMENT
Fig. 11. The peak-to-peak fluctuation of the FBG time-dependent response, The authors would like to thank Ultra 555T LAE Ltd., Cyprus,
as the mass of the weights increasing. for their support.

REFERENCES
corresponded to a wavelength shift of 1 nm for 0 g, ∼0.4 nm for
17 g and ∼0.15 nm for 34 g. In Fig. 11 we observe the declining [1] H. Y. Liu, G. D. Peng, and P. L. Chu, “Thermal tuning of polymer optical
fiber Bragg gratings,” IEEE Photon. Technol. Lett., vol. 13, no. 8, pp. 824–
peak-to-peak fluctuation of the POF-FBG when the mass of the 826, Aug. 2001.
weight is increased. It is important to note that, all the mea- [2] P. L. Chu, G. D. Peng, and Z. Xiong, “Highly-tunable polymer optical
surements refer to the use of the same FBG, with the weights fiber gratings,” in Proc. 8th Int. Plastic Opt. Fibers Conf., vol. 99, 1999,
pp. 242–245.
added to exactly the same position in each case. We consider [3] Z. Xiong, G. D. Peng, B. Wu, and P. L. Chu, “Highly tunable Bragg
that useful outcomes were extracted regarding the prediction of gratings in single-mode polymer optical fibers,” IEEE Photon. Technol.
health condition of the beam. The position of the damage could Lett., vol. 11, no. 3, pp. 352–354, Mar. 1999.
[4] H. Dobb, D. J. Webb, K. Kalli, A. Argyros, M. C. J. Large, and M. A.
be detected by the fluctuation of mode shape, whereas the level van Eijkelenborg, “Continuous wave ultraviolet light-induced fiber Bragg
of the damage could be defined by the maximum peak-to-peak gratings in few- and single-mode microstructured polymer optical fibers,”
response of the FBG. Clearly a careful calibration is required Opt. Lett., vol. 30, no. 24, pp. 3296–3298, 2005.
[5] G. Woyessa, A. Fasano, C. Markos, H. K. Rasmussen, and O. Bang,
for each device under test, however, it is often most impor- “Low loss polycarbonate polymer optical fiber for high temperature FBG
tant to discern that the response of the component has changed humidity sensing,” IEEE Photon. Technol. Lett., vol. 29, no. 7, pp. 575–
from its original behaviour, to be replaced accordingly prior to 578, Apr. 2017.
[6] G. Woyessa et al., “Zeonex-PMMA microstructured polymer optical
component failure. FBGs for simultaneous humidity and temperature sensing,” Opt. Lett.,
vol. 42, no. 6, pp. 1161–1164, 2017.
[7] A. Pospori, C. A. F. Marques, D. Sáez-Rodrı́guez, K. Nielsen, O. Bang,
V. CONCLUSION and D. J. Webb, “Thermal and chemical treatment of polymer optical fiber
Bragg grating sensors for enhanced mechanical sensitivity,” Opt. Fiber
We have inscribed a FBG array in a multimode, gradient index Technol., vol. 36, pp. 68–74, 2017.
CYTOP polymer fibre using the plane-by-plane fs-laser inscrip- [8] C. A. F. Marques, A. Pospori, D. Saez-Rodriquez, K. Nielsen, O. Bang, and
tion method. The physical characteristics of the gratings were D. J. Webb, “Aviation fuel gauging sensor utilizing multiple diaphragm
sensors incorporating polymer optical fiber Bragg gratings,” IEEE Sens.
adapted to minimize the coupling between the higher order fibre J., vol. 16, no. 15, pp. 6122–6129, Aug. 2016.
modes and the gratings, thereby producing ‘single–peak’ FBGs [9] P. Stajanca, O. Cetinkaya, M. Schukar, P. Mergo, D. J. Webb, and K.
in multimode fibre. The POF-FBGs were carefully calibrated Krebber, “Molecular alignment relaxation in polymer optical fibers for
sensing applications,” Opt. Fiber Technol., vol. 28, pp. 11–17, 2016.
for strain, temperature and relative humidity and the responses [10] C. A. F. Marques, G. D. Peng, and D. J. Webb, “High performance
were found to be 1.44 pm/με, 17.62 pm/°C and 14.7 pm/%RH, liquid level monitoring system based on polymer fiber Bragg gratings
respectively. Subsequently, the POF-FBG array was mounted on embedded in silicone rubber diaphragms,” Proc. SPIE, vol. 9507, 2015,
Art. no. 95070N.
a carbon fibre helicopter blade mounted as a cantilever beam. [11] D. Gallego and H. Lamela, “Microstructured polymer optical fiber sen-
By applying small vertical force on the beam we recovered the sors for optoacoustic endoscopy,” Proc. SPIE, vol. 10064, 2017, Art.
time-dependent wavelength response for each particular POF- no. 1006412.
[12] A. Theodosiou, A. Lacraz, and K. Kalli, “Femtosecond laser inscription
FBG of the array using the zero-crossing algorithm. of multiplexed FBG sensors in CYTOP polymer optical fibres,” in Proc.
Based on these responses, we extract the vibrational profile Photon. Fiber Technol., Bragg Gratings, Photosensitivity Poling Glass
of the structure and the influence of the position of the damage Waveguides, 2016, pp. 815–816.
[13] A. Lacraz, A. Theodosiou, and K. Kalli, “Low loss polymer fiber Bragg
site by adding small weights to the beam. The results showed gratings sensors for effective optical sensing of strain and temperature,”
higher fluctuation when the weights placed close to the base in Proc. IEEE Photon. Conf., 2016, pp. 815–816.
of the cantilever, and the lower variation when placed close [14] A. Lacraz, A. Theodosiou, and K. Kalli, “Femtosecond laser inscribed
Bragg grating arrays in long lengths of polymer optical fibres; a route to
to the edge of the beam compared with the response obtained practical sensing with POF,” Electron. Lett., vol. 52, no. 19, pp. 1626–
with no additional weights. Moreover, with different masses 1627, 2016.
we adjusted the “damage” to the beam and we observed the [15] A. Theodosiou, M. Polis, A. Lacraz, K. Kalli, M. Komodromos, and
A. Stassis, “Comparative study of multimode CYTOP graded index and
correlation between the magnitude of the fluctuation and the single-mode silica fibre Bragg grating array for the mode shape capturing
added mass. The proposed method has the potential be employed of a free-free metal beam,” Proc. SPIE, vol. 9886, 2016, Art. no. 98860O.

Authorized licensed use limited to: UNIV ESTADUAL PAULISTA JULIO DE MESQUITA FILHO. Downloaded on August 14,2024 at 21:35:28 UTC from IEEE Xplore. Restrictions apply.
 992 JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 36, NO. 4, FEBRUARY 15, 2018

[16] T. Mizunami, T. V. Djambova, T. Niiho, and S. Gupta, “Bragg grating [29] A. Theodosiou, M. Komodromos, and K. Kalli, “Health monitoring of
in multimode and few-mode optical fibers,” J. Lightw. Technol., vol. 18, carbon cantilever using femtosecond laser inscribed FBG array in gradient-
no. 2, pp. 230–235, Feb. 2000. index CYTOP polymer fibre,” in Proc. 25th Opt. Fibre Sens. Conf., Jeju,
[17] C. Lu and Y. Cui, “Fiber Bragg grating spectra in multimode optical Apr. 2017, pp. 1–4.
fibers,” J. Lightw. Technol., vol. 24, no. 1, pp. 598–604, Jan. 2006. [30] A. Theodosiou, K. Kalli, and M. Komodromos, “Accurate and fast demod-
[18] M. Koerdt et al., “Fabrication and characterization of Bragg gratings in ulation algorithm for multi-peak FBG reflection spectra using a combina-
perfluorinated polymer optical fibres and their embedding in composites,” tion of cross-correlation and Hilbert transformation,” J. Lightw. Technol.,
Mechatronics, vol. 34, pp. 137–146, 2016. vol. 35, no. 18, pp. 3956–3962, Sep. 2017.
[19] A. Lacraz, M. Polis, A. Theodosiou, C. Koutsides, and K. Kalli, “Fem- [31] C. Huang, W. Jing, K. Liu, Y. Zhang, and G. D. Peng, “Demodulation
tosecond laser inscribed Bragg gratings in low loss CYTOP polymer of fiber Bragg grating sensor using cross-correlation algorithm,” IEEE
optical fiber,” IEEE Photon. Technol. Lett., vol. 27, no. 7, pp. 693–696, Photon. Technol. Lett., vol. 19, no. 9, pp. 707–709, May 2007.
Apr. 2015. [32] D. Ganziy, O. Jespersen, G. Woyessa, B. Rose, and O. Bang, “Dynamic
[20] A. Lacraz, M. Polis, A. Theodosiou, C. Koutsides, and K. Kalli, “Bragg gate algorithm for multimode fiber Bragg grating sensor system,” Appl.
grating inscription in CYTOP polymer optical fibre using a femtosecond Opt., vol. 54, no. 18, pp. 5657–5661, 2015.
laser,” Proc. SPIE, vol. 9507, 2015, Art. no. 95070K. [33] Y. W. Liu, “Hilbert Transform and applications,” in Fourier Transform
[21] A. Theodosiou et al., “Modified fs-laser inscribed FBG array for rapid Application, S. Salih, Ed. Rijeka, Croatia: InTech, 2012, ISBN: 978-953-
mode shape capture of free-free vibrating beams,” IEEE Photon. Technol. 51-0518-3.
Lett., vol. 28, no. 14, pp. 1509–1512, Jul. 2016. [34] M. Aressy, Y. B. Braouli, A. Lacraz, A. Theodosiou, and K. Kalli, “Char-
[22] J. Korumpas and C. Peterson, Atmosphere Detector for Helicopter Blades, acterisation of a geosynthetic strip instrumented with FBG’s sensors in
Patent US3 795 147A, Mar. 5, 1974. low loss polymer optical fibre,” in Proc. Workshop Spec. Opt. Fibers,
[23] G. P. Engels and M. C. Thomas, Helicopter Blade Crack Detection System, Limassol, Cyprus, Oct. 2017, pp. 202–203.
Patent US5 205 710A, Apr. 27, 1993. [35] A. K. Pandey, M. Biswas, and M. M. Samman, “Damage detection from
[24] R. A. Parr, Blade Inspection System and Method, Patent US3 667 862A, changes in curvature mode shapes,” J. Sound Vib., vol. 145, no. 2, pp. 321–
Jun. 6, 1972. 332, 1991.
[25] C. D. Sam and J. J. Mahoney, Helicopter Blade Crack Indicator, Patent [36] J. R. Kim, Y. S. Ryu, H. M. Cho, and N. Stubbs, “Damage identification
US3 985 318A, Oct. 12, 1976. in beam-type structures: Frequency-based method vs mode-shape-based
[26] W. Ecke, I. Latka, R. Willsch, A. Reutlinger, and R. Graue, “Fibre optic method,” Eng. Struct., vol. 25, pp. 57–67, 2003.
sensor network for spacecraft health monitoring,” Meas. Sci. Technol.,
vol. 12, no. 7, pp. 974–980, 2001.
[27] S. Song, Y. Hou, M. Guo, L. Wang, X. Tong, and J. Wu, “An investigation
on the aggregate-shape embedded piezoelectric sensor for civil infras-
tructure health monitoring,” Constr. Build. Mater., vol. 131, pp. 57–65,
2017.
[28] P. J. Schubel, R. J. Crossley, E. K. G. Boateng, and J. R. Hutchinson,
“Review of structural health and cure monitoring techniques for large
wind turbine blades,” Renewable Energy, vol. 51, pp. 113–123, 2013. Authors’ biographies not available at the time of publication.

Authorized licensed use limited to: UNIV ESTADUAL PAULISTA JULIO DE MESQUITA FILHO. Downloaded on August 14,2024 at 21:35:28 UTC from IEEE Xplore. Restrictions apply.