1616 IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 22, NO. 21, NOVEMBER 1, 2010
Strain and Temperature Discrimination Using Modal
Interferometry in Bragg Fibers
Orlando Frazão, Luís M. N. Amaral, José M. Baptista, Philippe Roy, Raphael Jamier, and Sébastien Février
Abstract—The strain and temperature sensing characteristics of
a modal interferometer based on two Bragg fibers have been inves-
used present a different external cladding shape. It appears that
the sensitivity to the sensing parameters are different for the two
Bragg fibers, which makes it possible to fabricate several sensing
configurations based on the combination of these two Bragg fibers
for strain and temperature discrimination.
Index Terms—Bragg fiber, interferometer, modal interferom-
etry, strain and temperature measurement.
posed of a low index core surrounded by concentric dielectric
layers of high and low index. Many studies have been con-
ducted on these bandgap fibers due to their unusual proper-
ties (low bending loss, dispersion properties, spectral filtering,
etc.). For instance, such a fiber can propagate a single large
Gaussian mode and, therefore, be useful for power delivery ,
 or high-power lasers . Recently, the first polarization-
maintaining photonic bandgap Bragg fiber has been demon-
strated . Furthermore, the chromatic dispersion curve can be
tailored. Thus, these fibres can also be used in various applica-
tions such as supercontinuum generation  or dispersion com-
pensation . Regarding optical fiber sensing, such a bandgap
was demonstrated in a band rejection filter sensor configuration
 or a surface-plasmon resonance sensor for refractive index
measurement  or even as a temperature strain sensor .
In this letter, one proposes the utilization of two different
pieces of Bragg fibers to perform modal sensing interferometry
fibers characteristics, the linearly polarized fundamental mode
LP01 and the high order mode LP02 can be excited simultane-
ously creating a modal interferometer in the 1550-nm region.
The two different pieces of Bragg fibers were spliced in series
and the return interferometric signals were discriminated using
RAGG fibers, whose theoretical model was proposed in
1978 by Yeh et al. , are photonic bandgap fibers com-
Manuscript received June 24, 2010; revised July 27, 2010; accepted
September 03, 2010. Date of publication September 13, 2010; date of current
version October 15, 2010. This work was performed in the framework of
European Program COST, Action 299.
O. Frazão, L. M. N. Amaral, and J. M. Baptista are with INESC Porto–Insti-
tuto de Engenharia de Sistemas e Computadores do Porto, 4169-007 Porto, Por-
tugal, and also with Departamento de Física da Faculdade de Ciências da Uni-
versidade do Porto, 4169-007 Porto, Portugal (e-mail: firstname.lastname@example.org).
P. Roy,R. Jamier, andS. Février are with Xlim, Photonics Department, UMR
6172 CNRS University of Limoges, 87060 Limoges, France.
Color versions of one or more of the figures in this letter are available online
Digital Object Identifier 10.1109/LPT.2010.2076357
(b) of the two Bragg fibers.
coherence addressing and the interrogation was performed ap-
plying a heterodyne technique.
II. EXPERIMENTAL RESULTS
Bragg fibers, which are anti-resonant reflecting optical
waveguides (ARROW), allow the propagation of light in a
core with a refractive index smaller than the microstructured
periodic cladding one. Light, launched in the fiber core, is
coherently reflected onto the cladding and then is trapped in
the fiber core for certain wave vectors. Strictly, we do not
have modal propagation, but if the number of layers is not too
small the propagation approaches a modal-type one. In these
experiments, both fibers used have been drawn from the same
preform to an external diameter of 125
between the two fibers is the pure silica external cladding
shape. A first fiber drawing was made to obtain the first fiber.
Then the external shape of the preform has been mechanically
modified so as to have an octagonal shape. A second fiber
drawing was then carried out leading to the second fiber. The
cross-section images of the both fibers are shown in the inset of
Fig. 2. The both fibers are characterized by a depressed-index
core surrounded by three bilayers of high and low index. The
index difference between the core and the pure silica low-index
layers is close to
[see Fig. 1(a)]. The core diameters
of the fibers are similar,
35.7 m. Fig. 1(b) superimposes the
attenuation spectra of the both fibers measured by the cut-back
technique. Three low-loss bandgap regions can be observed.
In the vicinity of
nm corresponding to the center
of the first bandgap, the loss is
setup of the sensing interferometer is shown in Fig. 2. The lead
fibers are SMF 28. The sensing head consists of two pieces
m. The difference
0.2 dB m. The experimental
1041-1135/$26.00 © 2010 IEEE
FRAZÃO et al.: STRAIN AND TEMPERATURE DISCRIMINATION USING MODAL INTERFEROMETRY1617
Fig. 2. Experimental setup showing the sensing head and the receiving inter-
ferometer. Inset: cross-section of the both Bragg fibers.
of the both Bragg fibers with different lengths (30 mm for the
fiber with the circular external cladding shape and 350 mm for
the octagonal one), with a small length of connecting standard
single-mode fiber (SMF 28) put in between them. The loss
splice is approximately 1.5 dB for two splices between the SMF
28. The total sensing head loss is 3 dB. By nature, a photonic
bandgap fiber is a multimode waveguide. The single-mode
behavior is reached by making a modal discrimination based on
the confinement loss of each mode. Thus, a Bragg fiber exhibits
single-mode propagation only if the losses of the high order
modes are higher than those of the fundamental mode. Due
to the short lengths of Bragg fibers used in these experiments,
a multimode propagation is expected. Calculations using the
full-vector transfer matrix method have confirmed that the
both pieces of Bragg fibers propagate the two quasi-modes
noted LP01 and LP02 . Therefore, each piece of Bragg
fiber constitutes a modal interferometer based on these two
quasi-modes, which means that the sensing head has two-in
series modal interferometers with different path imbalances
due to the different fiber lengths. These path imbalances are
much larger than the EDFA source coherence length, meaning
that in the time domain at the end of the return SMF 28 there is
no noticeable interference. This situation changes if a second
interferometer is introduced in the reception region to imple-
ment the white-light readout technique. As shown in Fig. 2, the
light returned from the sensing head is injected into a receiver
Michelson interferometer with an open air path in one of the
arms, which is used to match the optical path difference of
one of the modal interferometers in the sensing head. In this
way, coherence addressing of that interferometer is performed
and an interferometric signal is generated at the input of the
photodetector. Applying a saw tooth waveform with proper
amplitude to a piezoelectric transducer (PZT) fiber stretcher
in one arm of the receiving interferometer, an electric carrier
can be obtained through electrical filtering of the detected
signal, whose phase is a replica of the phase of the tandem
modal interferometer. This phase is then measured using a
lock-in amplifier. To read the signal from the second modal
Fig. 3. Channeled spectrum of the interferometric sensing head.
Fig. 4. Phase variation of the interferometric sensing system when strain
(a) and temperature (b) are applied to the sensing head.
interferometer, the Michelson interferometer is tuned to it by
arising from a Bragg fiber modal interferometer, the channeled
spectrum at the input of the processing interferometer was
obtained when only the piece of the Bragg fiber with the
octagonal external cladding shape was present in the sensing
head. Fig. 3 shows this measured spectrum, which is typical
of a two beam interferometer, indicating that the structure
proposed works as a modal interferometer involving essentially
the two quasimodes LP01 and LP02 of the Bragg fiber. After,
the sensing head integrating the two pieces of Bragg fibers
was characterized for strain and temperature. In both cases, the
phase data from each modal interferometer was obtained tuning
the path imbalance of the processing interferometer to that of
the modal interferometer. The results obtained are plotted in
Fig. 4. It is important to underline that the readout phase signal
is the phase difference between the light that propagates in the
two modes excited by the input standard single-mode fiber.
Therefore, the accumulated phase difference shall essentially
be due to different modal elasto-optic coefficients (for the case
of strain) and different modal thermo-optic coefficients (for
the case of temperature). In the case of strain, the data shown
1618 IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 22, NO. 21, NOVEMBER 1, 2010 Download full-text
Fig. 5. Sensor output as determined by (2) for applied temperature and strain.
in Fig. 4(a) indicate sensitivity magnitudes of 1.6 mrad
(for the Bragg fiber with the circular external cladding shape)
and 2.1 mrad
(for the Bragg fiber with the octagonal one).
For temperature, the coefficients are 22 mrad
circular Bragg fiber) and 11 mrad
The strain coefficients are first-order independent of the fiber
length, but the temperature coefficients scale approximately
with the length of the modal interferometers. Therefore, by
normalizing, the temperature coefficients are 7.3 mrad cm C
(for the circular Bragg fiber) and 0.3 mrad cm C (for the
octagonal one). The difference in the strain and temperature
coefficients of the two Bragg fibers, more particularly for the
case of temperature where the difference is by a factor of 24,
are certainly due to different frozen internal stresses applied
to the fibers during their fabrication process. This is a topic
where further investigation is on target. This difference in the
sensitivity coefficients opens the possibility to perform with
this sensing head simultaneous measurement of strain and
temperature. Indeed, with the coefficients derived from Fig. 4
it can be written
C (for the
C (for the octagonal one).
This relation permits from the measured phase variation of
the two modal interferometers to derive the associated varia-
tions of temperature and strain. The performance of this simul-
taneous measurement configuration was experimentally deter-
mined by undertaking strain variations in a range of 1000
at a fixed temperature (22 C) and the other way around, i.e.,
temperature variations in a range of 50 C for a specific ap-
plied strain (500
), as shown in Fig. 5. From these results,
root-mean-square deviations of
terms of performance comparison, this configuration presents a
matrix condition number of 21.3, which is a better result when
comparing with the configurations presented in .
0.9 C and8 were deter-
Itwas demonstrated thatitis possible tobuildupmodal inter-
to strain and temperature variations. Two pieces of Bragg fibers
other with an octagonal one. They showed distinct differential
phase sensitivity to those measured. A sensing head with two
modal interferometers in series was implemented, addressed in
coherence and with heterodyne interrogation. The results ob-
form the functionality of simultaneous measurement of strain
and temperature. Two important features of this work are the
use of Braggfibers as interferometers and thedifferent achieved
sensitivities with different Bragg fiber cladding geometries.
The authors acknowledge S. L. Semjonov, M. E. Likhachev,
M. M. Bubnov, and E. M. Dianov from the Fiber Optics Re-
search Center, Moscow, and V. F. Khopin, M. Y. Salganskii,
and A. N. Guryanov from the Institute of Chemistry of High
Purity Substances, and N. Novgorod for the fabrication of the
 P. Yeh, A. Yariv, and E. Marom, “Theory of Bragg fiber,” J. Opt. Soc.
Amer., vol. 68, no. 9, pp. 1196–1201, Sep. 1978.
 S. Février, P. Viale, F. Gerome, P. Leproux, P. Roy, J.-M. Blondy, B.
Dussardier, and G. Monnom, “Very large effective area singlemode
photonic bandgap fiber,” Electron. Lett., vol. 39, no. 2, pp. 1240–1242,
 S. Février, R. Jamier, J.-M. Blondy, S. L. Semjonov, M. E. Likhachev,
M. M. Bubnov, E. M. Dianov, V. F. Khopin, M. Y. Salganskii, and
A. N. Guryanov, “Low-loss singlemode large mode area all-silica pho-
tonic bandgap fiber,” Opt. Express, vol. 14, no. 2, pp. 562–569, Jan.
 S. Février, D. D. Gaponov, P. Roy, M. E. Likhachev, S. L. Semjonov,
M. M. Bubnov, E. M. Dianov, M. Y. Yashkov, V. F. Khopin, M. Y.
Salganskii, and A. N. Guryanov, “High-power photonic-bandgap fiber
laser,” Opt. Lett., vol. 33, no. 9, pp. 989–991, May 2008.
 M. E. Likhachev, A. D. Pryamikov, D. A. Gaponov, M. M. Bubnov, M.
Y. Salganskii, V. F. Khopin, A. N. Guryanov, and S. Février, “Polar-
ization-maintaining photonic bandgap Bragg fiber,” Opt. Lett., vol. 34,
no. 9, pp. 1366–1368, May 2009.
 R. Jamier, S. Février, N. Ducros, M. E. Likhachev, and M. Y. Salgan-
skii, “Tight control of the spectral broadening obtained by nonlinear
conversion in a photonic bandgap fiber,” in Proc. CLEO/IQEC, Balti-
more, MD, May 31– Jun. 5, 2009, Paper JWA53.
Blondy, M. E. Likhachev, M. M. Bubnov, S. L. Semjonov, and E. M.
Dianov, “Highly dispersive large mode area photonic bandgap fiber,”
Opt. Lett., vol. 32, pp. 1208–1210, May 2007.
jection fiber filter and fiber sensor based on a Bragg fiber of transversal
resonant structure,” Opt. Express, vol. 16, no. 21, pp. 16489–16495,
 L. Ma, T. Katagiri, and Y. Matsuura, “Surface-plasmon resonance
sensor using silica-core Bragg fiber,” Opt. Lett., vol. 34, no. 7, pp.
1069–1071, Apr. 2009.
 O. Frazão, J. M. Baptista, J. L. Santos, P. Roy, R. Jamier, and S.
Février, “Bragg fibre for sensing applications,” Proc. SPIE, vol. 7714,
pp. 7714–7732, May 1–5, 2010.
 K. Kalli, A. G. Simpson, K. Zhou, L. Zhang, and I. Bennion, “Tai-
loring the temperature and strain coefficients of Type I and Type IA
dualgrating sensors—The impactofhydrogenation conditions,”Meas.
Sci. Technol., vol. 17, no. 5, pp. 949–954, May 2006.