Highly sensitive long-period fiber-grating strain
sensor with low temperature sensitivity
Yi-Ping Wang
Department of Electrical Engineering, Hong Kong Polytechnic University, Kowloon, Hong Kong, China,
and State Key Lab of Advanced Optical Communication Systems and Networks, Shanghai Jiao Tong University,
Shanghai 200240, China
Limin Xiao, D. N. Wang, and Wei Jin
Department of Electrical Engineering, The Hong Kong Polytechnic University, Kowloon, Hong Kong, China
Received July 13, 2006; revised August 24, 2006; accepted August 31, 2006;
posted September 11, 2006 (Doc. ID 73008); published November 9, 2006
A long-period fiber-grating sensor with a high strain sensitivity of −7.6 pm/
and a low temperature sen-
sitivity of
3.91 pm/ °C is fabricated by use of focused CO
2
laser beam to carve periodic grooves on a large-
mode-area photonic crystal fiber. Such a strain sensor can effectively reduce the cross-sensitivity between
strain and temperature, and the temperature-induced strain error obtained is only
0.5
/°Cwithout using
temperature compensation.
© 2006 Optical Society of America
OCIS codes: 060.2370, 050.2770, 060.2340
.
During application of long-period fiber-grating
(LPFG) strain sensors, one of the main difficulties is
the cross sensitivity between the strain and the
temperature.
1,2
The common methods for cross-
sensitivity reduction are using temperature compen-
sation and simultaneous strain and temperature
measurement.
1–3
By use of CO
2
laser radiation, a
LPFG with a strain sensitivity of −0.45 pm/
and a
temperature sensitivity of 58 pm/ °C was written in
Corning SMF-28 fiber,
3
and another LPFG with a
strain sensitivity of −0.19 pm/
and a temperature
sensitivity of 10.9 pm/ °C was inscribed in a photonic
crystal fiber (PCF).
4
In this Letter, we present a
LPFG sensor with a high strain sensitivity and a low
temperature sensitivity. Such a LPFG is fabricated
by use of focused CO
2
laser beam to carve periodic
grooves on a large-mode-area (LMA) PCF and can ef-
fectively reduce the cross sensitivity between strain
and temperature without using temperature compen-
sation.
As shown in Fig. 1, a CO
2
laser (SYNRAD 48-1)
with a maximum output power of 10 W was em-
ployed. A broadband light source and an optical spec-
trum analyzer (HP 70004A) were used to observe the
transmission spectrum evolution of LPFG. A LMA-10
PCF was situated at the focal plane of the CO
2
laser
beam, with one of the fiber ends fixed. A small weight
of ⬃5 g at the free end of the fiber was employed to
avoid the weight-induced macrobend of the fiber and
to provide a tensile strain in the fiber. The focused
CO
2
laser beam can scan the fiber via a computer-
controlled 2D optical scanner. Figure 2(a) illustrates
the cross section of the PCF employed. The PCF has
a center-to-center distance between the air holes of
⌳=6.1
m and an average air-hole diameter of d
=1.8
m. The holes are arranged in a hexagonal pat-
tern, which has a diameter of 65
m. The core diam-
eter D
co
=⌳共2−d / ⌳兲=10.4
m, and the outer diam-
eter of the PCF is 125
m.
As shown in Fig. 1, the focused CO
2
laser beam
scans repeatedly for M times along the X direction at
the location corresponding to the first grating period
of the fiber. Then the laser beam is shifted by a grat-
ing pitch along the Y direction and scans repeatedly
again M times to generate the next grating period.
This scanning and shifting process is carried out pe-
riodically N times (N is the number of grating peri-
ods) along the fiber’s axis until the final grating pe-
riod is produced. The above process may be repeated
for K cycles from the first grating period to the final
grating period until the high-quality LPFG is pro-
duced. The repeated scanning of the focused CO
2
la-
ser beam creates a local high temperature in the fi-
ber, which leads to the gasification of SiO
2
on the
surface of the fiber. As a result, periodic grooves with
a depth of ⬃15
m and a width of ⬃50
m are carved
on the fiber, as shown in Fig. 2(b). Such grooves can
induce periodic refractive index modulations along
the fiber axis due to the photoelastic effect, thus cre-
ating a LPFG. The groove’s depth, which indicates
the efficiency of CO
2
laser heating and the refractive
index modulation, depends on the fabrication param-
eters. In our experiments, the diameter of the focused
CO
2
laser beam spot is ⬃35
m, the line speed of la-
ser beam scanning on the fiber is 2.326 mm/s, the
pulse repetition rate is 10 kHz, the pulse width is
4.1
s, and average output power of the CO
2
laser is
⬃0.5 W.
As shown in Fig. 3, a high-quality LPFG, i.e.,
LPFG
1
, with asymmetric periodic grooves, a low in-
Fig. 1. Experimental setup for LPFG fabrication.
3414 OPTICS LETTERS / Vol. 31, No. 23 / December 1, 2006
0146-9592/06/233414-3/$15.00 © 2006 Optical Society of America
sertion loss of ⬃2 dB, and two attenuation dips (Dip
11
and Dip
12
) is produced after nine scanning cycles,
where the resonant wavelengths and the peak trans-
mission attenuations of Dip
11
and Dip
12
are
11
=1546.87 nm, A
11
=−31.136 dB,
12
=1351.86 nm, and
A
12
=−29.232 dB, respectively. To observe the differ-
ence between the strain characteristics of the LPFGs
with and without periodic grooves, another LPFG
without periodic grooves, i.e., LPFG
2
, was written in
the same type PCF by decreasing the power of the
CO
2
laser and the number of scanning times. The in-
sertion loss of 1.5 dB and two attenuation dips (Dip
21
and Dip
22
) were observed in the transmission spec-
trum of LPFG
2
, and the resonant wavelengths and
the peak transmission attenuations of Dip
21
and
Dip
22
are
21
=1548.38 nm, A
21
=−27.142 dB,
22
=1352.85 nm, and A
22
=−26.084 dB, respectively. The
grating pitch of both LPFG
1
and LPFG
2
is 410
m.
No grooves were observed in the LPFGs fabricated
by Zhu et al.
5
in the same type of PCF and by Rao
et al.
3
in standard single-mode fiber. During our
LPFG fabrication, however, periodic grooves were
carved intentionally on the fiber, as shown in Fig.
2(b). Such periodic grooves have essentially no contri-
bution to the LPFG’s insertion loss, which is similar
to the case of the corrugated LPFG fabricated by hy-
drofluoric acid etching.
6
This is because these grooves
are totally confined within the outer cladding and
have no influence on the light transmission in the
core of the fiber. The insertion loss of the LPFG writ-
ten in the PCF is mainly due to the
CO
2
-laser-induced collapse of air holes and the non-
periodicity and disorder of the refractive index modu-
lation. During Zhu’s
5
fabrication, the CO
2
laser beam
was focused to a larger light spot with a diameter of
⬃180
m, and the fiber is moved by adjusting a
transmission stage, which leads to the difficulty of fo-
cusing the laser beam at the desired location of the
fiber. As a result, the insertion loss, i.e., ⬃2 dB, of our
LPFGs is much smaller than that, i.e., ⬃16 dB, of the
LPFG written by Zhu et al.
5
in the same type of PCF.
When an optical fiber with an asymmetric struc-
ture, e.g., periodic grooves on one side of the fiber, is
stretched longitudinally, small lateral bends, i.e., pe-
riodic microbends, will be induced in the grooved sec-
tion of the fiber.
7
Thus periodic microbends can be ob-
served when the CO
2
-laser-carved LPFG with
asymmetric grooves is stretched, as shown in Fig. 4.
The refractive index modulation in the CO
2
-laser-
carved LPFG that is stretched can be expressed as
⌬n = ⌬n
residual
+ ⌬n
groove
+ ⌬n
stretch
, 共1兲
where ⌬n
residual
is the initial refractive index pertur-
bation induced by the residual stress relaxation re-
sulting from the high local temperature, which is
similar to the case of the CO
2
-laser-induced LPFGs
without periodic grooves
3,5
; ⌬n
groove
is the initial re-
fractive index perturbation induced by the periodic
grooves on the fiber, which is similar to ⌬n
corrugated
in
the corrugated LPFGs fabricated by hydrofluoric acid
etching
6
; ⌬n
stretch
is the refractive index perturbation
induced by the stretching force and can be expressed
as
⌬n
stretch
= ⌬n
strain
+ ⌬n
microbend
, 共2兲
where ⌬n
strain
is the refractive index perturbation in-
duced by the difference between the stretch-induced
tensile strains in the grooved and ungrooved regions
via the photoelastic effect
6,7
; ⌬n
microbend
is the refrac-
tive index perturbation induced by the stretch-
induced microbends in the CO
2
-laser-carved LPFGs
with asymmetric grooves.
7,8
Such stretch-induced mi-
crobends effectively enhance refractive index modu-
lation in the CO
2
-laser-carved LPFGs, which is simi-
lar to the case of the microbend-induced LPFG
reported.
8
The amplitude of microbends depends
strongly on the stretching force applied and the
groove parameters such as the depth and width of
Fig. 2. (a) Scanning electron micrograph of the cross sec-
tion of the PCF employed. (b) CCD photograph of the LPFG
with periodic grooves, obtained via a fiber fusion splicer
from ERICSSON.
Fig. 3. (Color online) Transmission spectra evolution of
the LPFG with the number, K, of scanning cycles increased
from 1 to 9, where N = 40 and M =5.
Fig. 4. Schematic of the CO
2
-laser-carved LPFG (a) before
and (b) after a stretching force, F, is applied, where ⌳, D,
and W are the grating pitch, the depth, and width of the
grooves, respectively.
December 1, 2006 / Vol. 31, No. 23 / OPTICS LETTERS 3415
the groove, the grating pitch, and the critical period-
icity of the fiber.
7
Therefore the refractive index
modulation efficiency in the CO
2
-laser-carved LPFG
is higher than that in the CO
2
-laser-induced LPFG
without periodic grooves.
With the stretching force applied, the curvature of
the stretch-induced microbends and the stretch-
induced tensile strain difference between the grooved
and the ungrooved regions increase, which leads to
the increase of ⌬n
strain
and ⌬n
microbend
. As shown in
Fig. 5(a), with the increase of the tensile strain, the
resonant wavelengths of LPFG
1
with periodic
grooves shift rapidly toward the shorter wavelength
and a good linearity is observed, whereas that of
LPFG
2
without periodic grooves shifts slowly toward
the shorter wavelength. The strain sensitivities of
the resonant wavelengths for Dip
11
, Dip
12
, Dip
21
, and
Dip
22
are −7.60, −6.71, −0.31, and −0.30 pm/
, re-
spectively. It is obvious that the strain sensitivity of
resonant wavelength, e.g., Dip
11
and Dip
21
, of the
LPFGs written by CO
2
laser in the same fiber (the
LMA PCF) is increased by 25 times by means of carv-
ing periodic grooves on one side of the fiber. The
strain sensitivity of our LPFG is much higher than
that of other CO
2
-laser-induced LPFGs without peri-
odic grooves.
3,4
The resonant wavelengths of the cor-
rugated LPFG with symmetric grooves is insensitive
to the tensile strain,
6
which is because no periodic
microbend is induced in the LPFG with symmetric
grooves. Therefore asymmetric grooves, instead of
symmetric grooves, should be created in order to in-
crease the strain sensitivity of the LPFGs fabricated,
and the strain sensitivity of the LPFG depends
strongly on the depth and width of the grooves, as
can be seen in Ref. 7.
As shown in Fig. 5(b), both resonant wavelength
and peak transmission attenuation of LPFG
1
and
LPFG
2
hardly change when the temperature in-
creases from 20°C to 100°C. The temperature sensi-
tivities of resonant wavelengths for Dip
11
, Dip
12
,
Dip
21
, and Dip
22
are only 3.91, 4.62, 3.84, and
4.63 pm/ °C, respectively. Thus, compared with the
temperature sensitivity of the LPFG written in stan-
dard single-mode fiber,
3
e.g., 58 pm/ °C, the resonant
wavelength of our LPFGs written by CO
2
laser in the
large-mode-area PCF is insensitive to the tempera-
ture, and the periodic grooves have no influence on
temperature sensitivity. The reduction of the tem-
perature sensitivity of the LPFGs obtained is due to
the air-hole structure of the pure silica PCF. Accord-
ing to the strain and temperature sensitivity of the
resonant wavelength for Dip
11
, the temperature-
induced strain measurement error is only 0.5
/°C
without using temperature compensation. Therefore
our LPFG strain sensors can effectively reduce the
cross sensitivity between strain and temperature.
In conclusion, an LPFG strain sensor with a high
strain sensitivity of −7.6 pm/
was achieved, and
the temperature-induced strain measurement error
is only 0.5
/ °C in the case of no temperature com-
pensation. Such an LPFG is fabricated by use of fo-
cused CO
2
laser beam to carve periodic grooves on a
large-mode-area PCF, and the strain sensitivity of
the LPFG obtained is increased by 25 times com-
pared with that of the LPFG without periodic
grooves.
This work was supported by a research grant of the
Hong Kong Polytechnic University in a Postdoctoral
Research Fellowship scheme (G-YX5l) and the Na-
tional Science Foundation of China (60507013). D. N.
Wang’s e-mail address is eednwang@polyu.edu.hk,
and Yi-Ping Wang’s e-mail address is
ypwang@china.com.
References
1. V. Bhatia, D. K. Campbell, D. Sherr, T. G. DAlberto, N.
A. Zabaronick, G. A. TenEyck, K. A. Murphy, and R. O.
Claus, Opt. Eng. 36, 1872 (1997).
2. Y. P. Wang, Y. J. Rao, Z. L. Ran, T. Zhu, and A. Z. Hu,
IEEE Photon. Technol. Lett. 15, 251 (2003).
3. Y. J. Rao, Y. P. Wang, Z. L. Ran, and T. Zhu, J.
Lightwave Technol. 21, 1320 (2003).
4. Y. Zhu, P. Shum, H. W. Bay, M. Yan, X. Yu, and J. Hu,
Opt. Lett. 30, 367 (2005).
5. Y. Zhu, P. Shum, J. H. Chong, M. K. Rao, and C. Lu,
Opt. Lett. 28, 2467 (2003).
6. C. Y. Lin, L. A. Wang, and G. W. Chern, J. Lightwave
Technol. 19, 1159 (2001).
7. M. Vaziri and C. L. Chen, Appl. Opt. 32, 6399 (1993).
8. C. H. Lin, Q. Li, A. A. Au, Y. Jiang, E. B. Wu, and H. P.
Lee, J. Lightwave Technol. 22, 1818 (2004).
Fig. 5. (Color online) Resonant wavelength of LPFG
1
and LPFG
2
via (a) the tensile strain and (b) the temperature.
⽧,Dip
11
of LPFG
1
; 䊏,Dip
12
of LPFG
1
; 䉱,Dip
21
of LPFG
2
; ⫻,Dip
22
of LPFG
2
.
3416 OPTICS LETTERS / Vol. 31, No. 23 / December 1, 2006
