Real time and simultaneous measurement of displacement and temperature using fiber loop with polymer coating and fiber Bragg grating Lei Chen, Weigang Zhang, Yongji Liu, Li Wang, Jonathan Sieg, Biao Wang, Quan Zhou, Liyu Zhang, and Tieyi Yan Citation: Review of Scientific Instruments 85, 075002 (2014); doi: 10.1063/1.4889885 View online: http://dx.doi.org/10.1063/1.4889885 View Table of Contents: http://scitation.aip.org/content/aip/journal/rsi/85/7?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Multiplex and simultaneous measurement of displacement and temperature using tapered fiber and fiber Bragg grating Rev. Sci. Instrum. 83, 053109 (2012); 10.1063/1.4718360 Simultaneous measurement of temperature and strain by combining a fiber Bragg grating and the pigtail fiber covered with epoxy resin Rev. Sci. Instrum. 82, 064904 (2011); 10.1063/1.3600902 A multiplexed fiber Bragg grating sensor for simultaneous salinity and temperature measurement J. Appl. Phys. 103, 053107 (2008); 10.1063/1.2890156 Bragg grating performance in Er–Sn-doped germanosilicate fiber for simultaneous measurement of wide range temperature (to 500°C) and strain Rev. Sci. Instrum. 74, 4858 (2003); 10.1063/1.1619553 Fiber optic sensor for dual measurement of temperature and strain using a combined fluorescence lifetime decay and fiber Bragg grating technique Rev. Sci. Instrum. 72, 3186 (2001); 10.1063/1.1372171
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 152.11.242.100 On: Thu, 25 Dec 2014 02:09:25
REVIEW OF SCIENTIFIC INSTRUMENTS 85, 075002 (2014)
Real time and simultaneous measurement of displacement and temperature using fiber loop with polymer coating and fiber Bragg grating Lei Chen, Weigang Zhang,a) Yongji Liu, Li Wang, Jonathan Sieg, Biao Wang, Quan Zhou, Liyu Zhang, and Tieyi Yan Key Laboratory of Optical Information Science and Technology, Ministry of Education, Institute of Modern Optics, Nankai University, Tianjin 300071, China
(Received 5 February 2014; accepted 29 June 2014; published online 15 July 2014) An all-fiber sensor scheme for real time and simultaneous displacement and temperature measurement is presented and demonstrated. The sensor head is formed by cascading a fiber loop with polymer coating with a fiber Bragg grating. The compatibility of the two components is fully utilized. A sensor resolution of 0.14314 V/μm in displacement and 0.00795 nm/◦ C in temperature are experimentally achieved within a displacement range of 0–50 μm and a temperature range of 20 ◦ C–75 ◦ C, respectively. The fiber loop with the protection of polymer coating is mechanically reliable, which means the sensor head also suits measuring dynamic displacement. A 500 Hz mechanical micro-vibration is successfully measured by the proposed sensor experimentally. In the last part, we perform a test making the sensor reach its maximum deformation and find the surviving sensor still possesses the same responsiveness as before. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4889885] I. INTRODUCTION
Optical fiber loops have attracted a lot of interest in recent years due to their unique physical and optical properties. An optical fiber loop has special bending loss,1 whispering gallery modes2 and its polarization characteristic3 is pronounced. Depending on these properties, novel optical devices based on fiber loops have been proposed, such as the edge filter,4 the optical frequency wavemeter,5 and so on. Additionally, optical sensors based on fiber loops have also been frequently studied because their pronounced bending characteristic can provide an efficient and controlled interaction of the guided light with external sensors, such as the displacement sensor,6 refractive index sensor,7 temperature sensor,8 and so on. In general, two main methods for displacement measurement in optic fibers may be identified for wavelength demodulation and intensity demodulation. For the former method, one or more dips are formed by coupling or cohering the light shown in the spectrum, which can be affected by the displacement. By recording the shift of the dips in the spectrum, the displacement can be read. For the latter one, the fiber loss is dependent on displacement. When the transmission light suffers from the loss, the intensity will decrease, resulting in a decrease of the voltage. Consequently, the displacement can be read from the voltage. Additionally, the displacement sensor based on intensity demodulation is a good candidate for detecting vibration. Recently, some kinds of fiber grating or cavity such as fiber Bragg gratings (FBGs),9–11 tilted-FBGs,12 long period fiber gratings (LPGs),13, 14 and F-P cavities14 were proposed to be the sensor head. The FBG based sensor head is used a) Author to whom correspondence should be addressed. Electronic mail:
[email protected]
0034-6748/2014/85(7)/075002/6/$30.00
very commonly in the measurement of displacement. In displacement sensor head experiments, most of the FBGs are glued to a cantilevered beam. Since the uniform FBGs suffer from an asymmetrical force, the original uniform FBG becomes a chirped FBG; by measuring the reflected bandwidth or the intensity, the displacement can be read. However, relatively complicated force transfer components need to be employed on these sensors, causing them to lack consistent results; moreover, the largest sensitivity is only 0.6234 nm/mm and 165.48 mV/mm.9 Therefore, they are not practical for micro-displacement measurement. The sensor head based on LPG is well developed in the field of detecting displacement because LPGs offer high cladding and bending loss. Reference 13 reported that the resolution of their sensor head is about 11.5 μm in a range of 240 μm by employing the bending loss of a LPG; Ref. 14 proposed a sensor head consisting of a LPG cascading an air cavity, and a high sensitivity of 0.2155 nm/μm was found in the measurement range of 0–140 μm. Moreover, the coherence of orthorhombic polarization mode can also be utilized in the measurement of displacement, especially micro-displacement. In Ref. 15, several dips are formed from the orthorhombic polarization mode. When the sensor head is displaced, the wavelength of the dips will shift to a new wavelength due to a slight change in the coherent length. Consequently, by measuring the differences in their spectra, the displacement can be read. A sensitivity of 1.1 nm/μm with a smallest measurable displacement of 80 μm can be reached. More recently, the combination of an adiabat taper and a FBG is used to measure microdisplacement based on the bend loss of an adiabat taper; the function of FBG in that structure is multiplex measurement of different points rather than its previous function. According to their measured results, the largest sensitivities reach 0.11 dB/μm for displacement in a range of 0–400 μm, and about 0.0097 nm/ ◦ C in a temperature range of 20 ◦ C –70 ◦ C.16
85, 075002-1
© 2014 AIP Publishing LLC
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 152.11.242.100 On: Thu, 25 Dec 2014 02:09:25
075002-2
Chen et al.
Rev. Sci. Instrum. 85, 075002 (2014)
Additionally, increasing the bending loss of a fiber is another method to measure micro-displacement. In Ref. 6, a bared high bend loss semi-fiber loop coated absorption layer is used as a displacement sensor. Since the output intensity has a high sensitivity to the bend radii, it can indirectly measure the displacement. However, the drawback of this principle is the unwanted phenomenon that the bend loss is not a monotonous function due to the reflected light from the interface of air and silica. The function of the additional absorption layer serves to remove the reflected light and make the loss intensity linear with the bend radii. A high resolution of 40 nm with a high sensitivity of 0.268 dB/μm can be reached in a range of 15 μm at a centre wavelength of 1550 nm. Finally, whispering gallery modes (WGMs) can also be used to detect displacement. In Ref. 17, a bare fiber loop is explained as a cavity of whispering gallery. The spectrum of their sensor head is composed of several comb-shaped dips. The radius of a fiber loop affects the mode field radius, and the dips will shift due to the contribution from the displacement. Another laser with a centre wavelength of 1555 nm is added whose narrow band spectrum acts as a narrow band filter and a photodetector is positioned to detect the output intensity. The original dip in the 1555 nm will shift depending on displacement, and the vibration information can also be measured. Though these sensor heads are compact, easy to fabricate, and low cost, these sensor heads are hardly used in practice for the following three reasons. First, the reported structures2, 6–13, 17 are not mechanically reliable without the protection of a polymer coating, i.e., it is hard for the bare part(s) of the sensors to undergo long time measurement or an abrupt change of environment without mechanical damage. Second, the reported structures undergo temperature, ambient refractive index, or humidity crosstalk due to the material thermal expansion coefficients and the evanescent field.14 Third, the sensors, particularly interferometers,15 require chemical processing, and thus are not very reproducible. Additionally, the sensors based on the aforementioned wavelength demodulation method cannot measure dynamic physic parameter. We also find that neither of commercial devices offers dual sensing options.18 In this paper, we present a combined fiber loop with polymer protection coating (FLPPC) and FBG probe in which the FLPPC that proves to be nearly independent of ambient temperature19 ; this allows a clean separation between temperature and displacement sensing, which is essential in the presence of a steep gradient. Using the proposed probe, we provide the results of micro-displacement and temperature measurements during displacement in a range of 0–50 μm (sensitivities 0.14313 V/μm) and temperatures in a range of
20 ◦ C–75 ◦ C (sensitivities 0.00795 nm/◦ C). Furthermore, the proposed sensor head is compact (about 4 cm) and low cost (unless 0.2 US dollar). Moreover, the proposed structure can be easily developed into a star-shaped topological structure sensor for multipoint sensing by cascading different FBGs and detecting the reflected light from them. Here, we set a limit of displacement measurement range of the sensor in order to make the range ability of the bend radius within two of the inflection points like the phenomenon described in Ref. 19, which is commonly used in engineering to avoid non-monotonically changing of dependent variable as the independent variable. In this regard, because of maintaining the polymer coating, the crosstalk between displacement and ambient temperature/refractive index can be totally canceled out, which can be confirmed by the following performed experiments. Moreover, the “healthy” structure also can serve for resisting the mechanical damage in the long time measurement or an abrupt change of environment, which can be confirmed from the last part. Additionally, we can notice that the method of packaging can effectively keep the reproducibility of each sensor head. When one optical fiber loop cascaded with a FBG works as a sensor head, the transmission intensity is only affected by the bending radius of the optical fiber loop and the peak wavelength which is at the FBG’s centre wavelength, is only affected by temperature, i.e., the fiber loop with polymer coating is highly sensitive to displacement response but independent of temperature response; on the contrary, the fixed FBG is sensitive to temperature response but independent of displacement response; no crosstalk exists between the two parameters. Thus, two parameters are easily obtained. II. EXPERIMENTAL SETUP
The proposed sensor is structured with two slides, a section of standard single mode fiber, and a FBG. We first engrave a half loop cavity in one slide by using a CO2 laser (Han’s Laser, CO2–H3O). Then the section of commercial single mode fiber connected to the FBG (center wavelength is 1550 nm, and 3 dB wavelength band is 0.3 nm) is coiled up into the cavity to fabricate and keep the structure—cascading a fiber loop with polymer protection coating packaged in the circle cavity and a FBG. After that, another slide covers it and secures it tightly. The sensor head making process is shown in Fig. 1. Fig. 2(a) shows the experimental setup of the displacement sensing system. Both ends of the sensor head are fixed on the translation stages, while the sensor keeps vertical with another translation stage. Two of the translation stages are
FIG. 1. Fabricating of the sensor head.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 152.11.242.100 On: Thu, 25 Dec 2014 02:09:25
075002-3
Chen et al.
Rev. Sci. Instrum. 85, 075002 (2014)
FIG. 2. (a) Schematic diagram of the displacement sensing system; (b) schematic diagram of the real time displacement sensing system.
fixed during the experiment to keep the sensor head unmovable, while the third translation stage is movable in the vertical direction, and the displacement of it can be read from the micrometer caliper on it. A beam of light from a broadband source (BBS) with a wavelength range of 1525–1568 nm is coupled into the sensor head. When the light with a wavelength range of 1525 –1568 nm propagates through the fiber loop and reaches the FBG, the reflected light from the FBG only possess a 3 dB wavelength band of 0.3 nm centered 1550 nm, while the transmission light possess the whole wavelength except for the part reflected by the FBG (see inset of Fig. 1). We compose a photo-detector (Thorlabs, DET01CFC) and an oscilloscope (OSC: YOKOGAWA, DL9140), and an optical spectrum analyzer (OSA: YOKOGAWA, AQ6370C) by a 1 × 2 coupler of which the splitting ratio is 1:1 to achieve real time and simultaneous measurement of displacement and temperature. The first part transmission light will be received by a photo-detector (or O-E converter) and shown on the OSC with a resolution of 0.01 V, and the second part transmission light will be detected by an OSA with a spectral resolution of 0.01 nm. To confirm the sensor head respond displacement in real time, the vibration sensing experiment is performed. The vibration source, a commercial speaker, is placed on a Lab jack and driven by a computer, as shown in Fig. 2(b). We vertically elevate the Lab jack and carefully adjust the voltage to quasilinear range (as shown in Fig. 3). When the sensor is subject to the speaker vibration, the changing electronic signals are shown on the OSC. The principle of our proposed sensor head is based on the bending loss sensitive to the equivalent bend radius of the optical fiber loop. The equivalent radius of the optical fiber loop would decrease when the displacement is applied on the optical fiber loop’s top. Fig. 3 shows the relationship between the bending loss and the bending radius. The bending loss dependence on wavelength and bend radius is shown in the inset of Fig. 3. From the inset, we can see that the bending loss varies with the bending radius periodically according to the
theory of Ref. 1. If a monochromatic light is coupled into the sensor, the change in output voltage from the photo-detector is proportional to one of the curves varying with different the bending radius. For the sake of increasing the sensitivity of displacement and detecting temperature simultaneously, the light source is selected as BBS in the experiment rather than a traditional monochromatic laser. Therefore, the change in output voltage at different wavelengths should be summed up to calculate total change in output voltage V(R) and its result is shown in Eq. (1), αBC (λ, R), (1) V (R) = S(λ)D(λ) λ
here, λ represents wavelength, S(λ) is the input laser intensity, D(λ) is the detector responsivity, and R is bend radii, respectively. Based on the BBS wavelength range (1525–1568 nm) in the experiment, the result of output voltage by summing up Eq. (1) is shown in Fig. 3. It is obvious that in the
FIG. 3. Simulation of the change of output voltage dependence on bend radius with a wavelength range from 1525 nm to 1568 nm. (Inset) simulation of the bending loss dependence on radius and wavelength.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 152.11.242.100 On: Thu, 25 Dec 2014 02:09:25
075002-4
Chen et al.
vicinities of 5.2 mm, 6.2 mm, and 7.2 mm, a quasi-linear relationship between output voltage and radius is obtained.7, 8 When the top of the fiber loop with the described radius are subjected to a displacement, the equivalent bending radius of the fiber loop have a little change. That slight change in equivalent bending radius will contribute to output voltage decrease quasi-linearly. As a result, these sections are well-suited for displacement sensing. Owning to the coefficient of thermal expansion of polymer coating smaller than silica, the volume of the bare fiber will be bound by the polymer protection coating which will cancel out the influence of ambient temperature on output voltage. From the analysis above we know the bending loss of fiber loop will affect the output voltage and that structure is independent on temperature. However, the centre wavelength of the FBG’s transmission spectrum is independent on the bending loss, while it is sensitive to the ambient temperature. Consequently, the ambient temperature of the sensor head can be obtained independent of displacement. Therefore, the proposed sensor possesses the ability that the ambient temperature and the displacement can be effectively measured independently. A radius of 6.6 mm is selected for the loop curve in the experiment for the following two reasons: First, from Fig. 3, we can see that if the initial fiber loop radius is larger than 7.5 mm, the output voltage becomes insensitive to the decreasing radius. In this regard, the initial radii should be less than 7.5 mm. Second, technically, a cavity in the slide is very fragile if its radius is less than 5.5 mm. Considering the above two limitations, the radius selections are narrowed down to two sections. We select the smaller one as initial radii of the optical fiber loop in order to pursue a larger sensitivity of displacement. III. EXPERIMENT AND RESULTS
Fig. 4 shows the transmission spectra when the whole configuration is settled at a room temperature and measure at different displacement. At the output, the transmission loss peak wavelength is 1550.08 nm, which corresponds to the resonant peak of the FBG and reflects the ambient temperature
FIG. 4. Transmission spectra of the sensor when the displacements of the sensor head are applied ranging from 0 μm to 50 μm at room temperature.
Rev. Sci. Instrum. 85, 075002 (2014)
FIG. 5. Transmission loss peaks wavelength and output voltages when displacement of the sensor head changes from 0 μm to 50 μm at room temperature.
of the sensor; the intensities of the whole spectra are varied from different displacements, which can be converted to voltage and reflect the displacement. The transmission loss peak of the sensor head is easily distinguished in the spectrum. When displacements change by moving the movable translation stages such as Fig. 2(a) depicts, the output intensities of whole spectrum will change and lead to the varying of the output voltage. By monitoring the output voltage and the transmission loss peak wavelength, the displacements and ambient temperature of the sensor head will be obtained. We apply displacement to the sensor head and keep temperature stable, and monitor the output voltage and the peak wavelength via the OSC and OSA. The results are shown in Fig. 5. As we expected, the output voltage decreases as the equivalent radius decreases and it can be fitted as y = −0.14314x + 11.42524; while the peak wavelength is stable (or more accurately, the changing of transmission loss peak wavelength in the range of system resolution—0.01 nm) and it can be fitted as y = 1150.08. The sensitivity of the sensor head for displacement is 0.14314 V/μm ranging from 0 to 50 μm. The error bars of the experimental data in x- and y- axes are times 5 and 10 in order to make it visible. Fig. 6 shows the transmission spectra when the whole configuration is fixed at 0 μm displacement and settled at a temperature ranging from 20 ◦ C to 75 ◦ C. The inset is the magnification of the peaks at different temperatures. The output voltage is monitored at the same time by OSC. Both the wavelength and output voltage are shown in Fig. 7. From Fig. 7, we can see that the output voltage in different temperatures keeps a constant value (or more accurately, the changing of the voltage is under the system resolution—0.01 V) and it can be fitted as y = 12.18; while the transmission loss peak wavelength varies quasi-linearly as the temperature and it can be fitted as y = 0.00795x −0.18381. The sensitivity of the sensor head for temperature is 0.00795 nm/ ◦ C ranging from 20 ◦ C to 75 ◦ C. Notice that the error bar in the x-axis is times 10 for the sake of making the error bar visible. To do more research and confirm the real time characteristics of the sensor head respond, a vibration experiment
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 152.11.242.100 On: Thu, 25 Dec 2014 02:09:25
075002-5
Chen et al.
FIG. 6. Transmission spectra of the sensor head when temperature applied range from 20 ◦ C to 75 ◦ C.
is performed, and the processing is described in Fig. 2(b). A 5 V, 500.00 Hz sinusoidal voltage waveform is applied experimentally. Fig. 8(a) shows a representative measurement result at time domain and its corresponding frequency domain. The inset of Fig. 8(a) is the magnification of the red circle and shows the details of the measured signal. The output voltage of the sensor varied alike a sinusoidal waveform with very uniform amplitude. And the demodulated frequency agrees well with the electric signal, with the SNR of about 65 dB, as shown in Fig. 8(b). It is worth mentioning that the corresponding harmonics of the induced vibration frequency is also detected, which may be due to the slightly non-axial alignment between the speaker and the sensor.20 Nevertheless, the tone of the main frequency overcomes the second, third, and fourth harmonics with a difference of 43, 38, and 53 dB, respectively, which allows clear identification of the induced vibration frequency. However, what need alludes is the temperature measurement will be limited if the influence of the spectrum contributed by the amplitude of a vibration source is comparable with a loss amplitude of a FBG.
FIG. 7. Transmission loss peak wavelength and output voltage of the sensor when whole configuration is fixed at 0 μm and settled it at a temperature ranging from 20 ◦ C to 75 ◦ C.
Rev. Sci. Instrum. 85, 075002 (2014)
FIG. 8. Responses of the sensor head to speaker driven by a 5 V, 500 Hz sinusoidal wave in (a) time and (b) corresponding frequency domain.
Obviously, when a fiber loop with polymer coating packaged in the circle cavity is cascaded with a FBG, simultaneous measurement of displacement and temperature can be achieved by detecting the output voltage and the transmission loss peak wavelength of the FBG. Moreover, since existing the protection polymer coating, the proposed senor is reliable mechanically which can be used in long time and real time measurement. Additionally, the crosstalk with temperature or ambient refractive index due to the material thermal expansion coefficients and evanescent field can be totally cancelled out (In fact, the experiments about put the single fiber loop with polymer coating in the temperature chamber ranging from 20 ◦ C to 75 ◦ C and drip a liquid with 1.8 index of refractive are performed. Although the outcomes are not plotted here, we still can image easily that it does not have any effects on output voltage and wavelength due to the protection of polymer coating). The disadvantage of this sensor is output voltage non-monotonically dependent on decrease of equivalent bend radius, but it can be easily solved by set a limit—50 μm—on the range of displacement measurement. In the last part, we make the sensor reach its maximum deformation in order to check its ability survive in the harsh environment. Both ends of the sensor head are fixed on the translation stages, while the sensor maintains a vertical position with another translation stage. Two of the translation stages are fixed during the experiment to keep the sensor unmovable; another is movable in the vertical direction and inward to induce bending in the sensor. The steel beam on the
FIG. 9. (a) Original shape of the sensor, (b) Steel beam forces the sensor to maximum bending point, (c) Steel beam moves outward to allow sensor to resume its shape.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 152.11.242.100 On: Thu, 25 Dec 2014 02:09:25
075002-6
Chen et al.
vertical translation stages moves until the slide bends to a maximum without breaking and then moves outward to allow the sensor to resume its shape. These processes are pictured in Fig. 9, which shows the sensor possesses the ability to withstand a maximum deformation and still can work when the environment improves. IV. CONCLUSION
In conclusion, we present a simple structure for simultaneous displacement and temperature measurement, by utilizing the transmission intensity of an optical fiber loop’s decrease as the equivalent bending radius decreases and independent of ambient temperature; red shifting response of a FBG’s peak wavelength due to the increasing of ambient temperature; and the compatibility of the two components. After applying suitable packaging, the proposed sensor head is able to measure displacement and temperature with sensitivities reaching up to 0.14314 V/μm for displacement in the range of 0–50 μm and 0.00795 nm/◦ C for temperature between 20 ◦ C and 75 ◦ C. Additionally, such a sensing method can be used for real time and simultaneous measurement of displacement and temperature in the building industry, such as bridges, dam buildings, and so on. Moreover, because of the polymer coating protection, the structure is reliable even under harsh condition in mechanic, and even certain chemical parameters such as refractive-index, humidity, and so on. ACKNOWLEDGMENTS
This work was supported by the National Natural Science Foundation under Grant Nos. 11274181, 10974100,
Rev. Sci. Instrum. 85, 075002 (2014)
10674075, 11104149, the Doctoral Scientific Fund Project of the Ministry of Education under Grant No. 20120031110033 and by the Tianjin Key Program of Application Foundations and Future Technology Research Project under Grant No. 10JCZDJC24300, China. 1 Q.
Wang, G. Farrell, and T. Freir, Opt. Exp. 13, 4476 (2005). H. Nam and S.-Z. Yin, IEEE Photon. Technol. Lett. 17, 2391 (2005). 3 Q. Wang, G. Rajan, P.-F. Wang, and G. Farrell, Opt. Exp. 15, 4909 (2007). 4 P.-F. Wang, G. Farrell, and Y. Semenova, Appl. Opt. 48, 3055 (2009). 5 Q. Wang, G. Farrell, T. Freir, G. Rajan, and P.-F. Wang, Opt. Lett. 31, 1785 (2006). 6 P.-F. Wang, G. Brambilla, Y. Semenova, Q. Wu, and G. Farrell, J. Opt. 13, 075402 (2011). 7 P.-F. Wang, Y. Semenova, Q. Wu, G. Farrell, Y.-Q. Ti, and J. Zheng, Appl. Opt. 48, 6044 (2009). 8 P.-F. Wang, Y. Semenova, Q. Wu, J. Zheng, and G. Farrell, Appl. Opt. 49, 1744 (2010). 9 J. H. Ng, X.-Q Zhou, X.-F Yang, and J.-Z. Hao, Opt. Commun. 273, 398 (2007). 10 X.-Y. Dong, X.-F. Yang, C.-L. Zhao, L. Ding, P. Shum, and N. Q. Ngo, Smart Mater. Struct. 14, N7 (2005). 11 C.-Y. Shen and C. Zhong, Sens. Actuat. A 170, 51 (2011). 12 Q. Jiang and D. Hu, IEEE Sens. J. 11, 1776 (2011). 13 J. M. Baptista, S. F. Santos, G. Rego, O. Frazáo, and J. L. Santos, Opt. Commun. 260, 8 (2006). 14 L. Qi, C.-L. Zhao, Y.-P. Wang, J. Kang, Z.-X. Zhang, and S. Z. Jin, Opt. Exp. 21, 3193 (2013). 15 L.-L. Xu, Y. Li, and B.-J. Li, Opt. Lett. 38, 64 (2013). 16 C.-K. Ji, C.-L. Zhao, J. Kang, X.-Y. Dong, and S.-Z. Jin, Rev. Sci. Instrum. 83, 053109 (2012). 17 P.-F. Wang, Y. Semenova, A. Sun, Q. Wu, and G. Farrell, Proc. SPIE 7726, 772623 (2010). 18 D. Tosi, E. Gino Macchi, G. Braschi, A. Cigada, M. Gallati, S. Rossi, S. Poeggel, G. Leen, and E. Lewis, Appl. Opt. 53, 2136 (2014). 19 R.-P. Hu and X.-G. Huang, IEEE Sens. J. 9, 1952 (2009). 20 Y. Huang, T. Guo, C. Lu, and H. Y. Tam, IEEE Photon. Technol. Lett. 22, 1235 (2010). 2 S.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 152.11.242.100 On: Thu, 25 Dec 2014 02:09:25
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 152.11.242.100 On: Thu, 25 Dec 2014 02:09:25
REVIEW OF SCIENTIFIC INSTRUMENTS 85, 075002 (2014)
Real time and simultaneous measurement of displacement and temperature using fiber loop with polymer coating and fiber Bragg grating Lei Chen, Weigang Zhang,a) Yongji Liu, Li Wang, Jonathan Sieg, Biao Wang, Quan Zhou, Liyu Zhang, and Tieyi Yan Key Laboratory of Optical Information Science and Technology, Ministry of Education, Institute of Modern Optics, Nankai University, Tianjin 300071, China
(Received 5 February 2014; accepted 29 June 2014; published online 15 July 2014) An all-fiber sensor scheme for real time and simultaneous displacement and temperature measurement is presented and demonstrated. The sensor head is formed by cascading a fiber loop with polymer coating with a fiber Bragg grating. The compatibility of the two components is fully utilized. A sensor resolution of 0.14314 V/μm in displacement and 0.00795 nm/◦ C in temperature are experimentally achieved within a displacement range of 0–50 μm and a temperature range of 20 ◦ C–75 ◦ C, respectively. The fiber loop with the protection of polymer coating is mechanically reliable, which means the sensor head also suits measuring dynamic displacement. A 500 Hz mechanical micro-vibration is successfully measured by the proposed sensor experimentally. In the last part, we perform a test making the sensor reach its maximum deformation and find the surviving sensor still possesses the same responsiveness as before. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4889885] I. INTRODUCTION
Optical fiber loops have attracted a lot of interest in recent years due to their unique physical and optical properties. An optical fiber loop has special bending loss,1 whispering gallery modes2 and its polarization characteristic3 is pronounced. Depending on these properties, novel optical devices based on fiber loops have been proposed, such as the edge filter,4 the optical frequency wavemeter,5 and so on. Additionally, optical sensors based on fiber loops have also been frequently studied because their pronounced bending characteristic can provide an efficient and controlled interaction of the guided light with external sensors, such as the displacement sensor,6 refractive index sensor,7 temperature sensor,8 and so on. In general, two main methods for displacement measurement in optic fibers may be identified for wavelength demodulation and intensity demodulation. For the former method, one or more dips are formed by coupling or cohering the light shown in the spectrum, which can be affected by the displacement. By recording the shift of the dips in the spectrum, the displacement can be read. For the latter one, the fiber loss is dependent on displacement. When the transmission light suffers from the loss, the intensity will decrease, resulting in a decrease of the voltage. Consequently, the displacement can be read from the voltage. Additionally, the displacement sensor based on intensity demodulation is a good candidate for detecting vibration. Recently, some kinds of fiber grating or cavity such as fiber Bragg gratings (FBGs),9–11 tilted-FBGs,12 long period fiber gratings (LPGs),13, 14 and F-P cavities14 were proposed to be the sensor head. The FBG based sensor head is used a) Author to whom correspondence should be addressed. Electronic mail:
[email protected]
0034-6748/2014/85(7)/075002/6/$30.00
very commonly in the measurement of displacement. In displacement sensor head experiments, most of the FBGs are glued to a cantilevered beam. Since the uniform FBGs suffer from an asymmetrical force, the original uniform FBG becomes a chirped FBG; by measuring the reflected bandwidth or the intensity, the displacement can be read. However, relatively complicated force transfer components need to be employed on these sensors, causing them to lack consistent results; moreover, the largest sensitivity is only 0.6234 nm/mm and 165.48 mV/mm.9 Therefore, they are not practical for micro-displacement measurement. The sensor head based on LPG is well developed in the field of detecting displacement because LPGs offer high cladding and bending loss. Reference 13 reported that the resolution of their sensor head is about 11.5 μm in a range of 240 μm by employing the bending loss of a LPG; Ref. 14 proposed a sensor head consisting of a LPG cascading an air cavity, and a high sensitivity of 0.2155 nm/μm was found in the measurement range of 0–140 μm. Moreover, the coherence of orthorhombic polarization mode can also be utilized in the measurement of displacement, especially micro-displacement. In Ref. 15, several dips are formed from the orthorhombic polarization mode. When the sensor head is displaced, the wavelength of the dips will shift to a new wavelength due to a slight change in the coherent length. Consequently, by measuring the differences in their spectra, the displacement can be read. A sensitivity of 1.1 nm/μm with a smallest measurable displacement of 80 μm can be reached. More recently, the combination of an adiabat taper and a FBG is used to measure microdisplacement based on the bend loss of an adiabat taper; the function of FBG in that structure is multiplex measurement of different points rather than its previous function. According to their measured results, the largest sensitivities reach 0.11 dB/μm for displacement in a range of 0–400 μm, and about 0.0097 nm/ ◦ C in a temperature range of 20 ◦ C –70 ◦ C.16
85, 075002-1
© 2014 AIP Publishing LLC
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 152.11.242.100 On: Thu, 25 Dec 2014 02:09:25
075002-2
Chen et al.
Rev. Sci. Instrum. 85, 075002 (2014)
Additionally, increasing the bending loss of a fiber is another method to measure micro-displacement. In Ref. 6, a bared high bend loss semi-fiber loop coated absorption layer is used as a displacement sensor. Since the output intensity has a high sensitivity to the bend radii, it can indirectly measure the displacement. However, the drawback of this principle is the unwanted phenomenon that the bend loss is not a monotonous function due to the reflected light from the interface of air and silica. The function of the additional absorption layer serves to remove the reflected light and make the loss intensity linear with the bend radii. A high resolution of 40 nm with a high sensitivity of 0.268 dB/μm can be reached in a range of 15 μm at a centre wavelength of 1550 nm. Finally, whispering gallery modes (WGMs) can also be used to detect displacement. In Ref. 17, a bare fiber loop is explained as a cavity of whispering gallery. The spectrum of their sensor head is composed of several comb-shaped dips. The radius of a fiber loop affects the mode field radius, and the dips will shift due to the contribution from the displacement. Another laser with a centre wavelength of 1555 nm is added whose narrow band spectrum acts as a narrow band filter and a photodetector is positioned to detect the output intensity. The original dip in the 1555 nm will shift depending on displacement, and the vibration information can also be measured. Though these sensor heads are compact, easy to fabricate, and low cost, these sensor heads are hardly used in practice for the following three reasons. First, the reported structures2, 6–13, 17 are not mechanically reliable without the protection of a polymer coating, i.e., it is hard for the bare part(s) of the sensors to undergo long time measurement or an abrupt change of environment without mechanical damage. Second, the reported structures undergo temperature, ambient refractive index, or humidity crosstalk due to the material thermal expansion coefficients and the evanescent field.14 Third, the sensors, particularly interferometers,15 require chemical processing, and thus are not very reproducible. Additionally, the sensors based on the aforementioned wavelength demodulation method cannot measure dynamic physic parameter. We also find that neither of commercial devices offers dual sensing options.18 In this paper, we present a combined fiber loop with polymer protection coating (FLPPC) and FBG probe in which the FLPPC that proves to be nearly independent of ambient temperature19 ; this allows a clean separation between temperature and displacement sensing, which is essential in the presence of a steep gradient. Using the proposed probe, we provide the results of micro-displacement and temperature measurements during displacement in a range of 0–50 μm (sensitivities 0.14313 V/μm) and temperatures in a range of
20 ◦ C–75 ◦ C (sensitivities 0.00795 nm/◦ C). Furthermore, the proposed sensor head is compact (about 4 cm) and low cost (unless 0.2 US dollar). Moreover, the proposed structure can be easily developed into a star-shaped topological structure sensor for multipoint sensing by cascading different FBGs and detecting the reflected light from them. Here, we set a limit of displacement measurement range of the sensor in order to make the range ability of the bend radius within two of the inflection points like the phenomenon described in Ref. 19, which is commonly used in engineering to avoid non-monotonically changing of dependent variable as the independent variable. In this regard, because of maintaining the polymer coating, the crosstalk between displacement and ambient temperature/refractive index can be totally canceled out, which can be confirmed by the following performed experiments. Moreover, the “healthy” structure also can serve for resisting the mechanical damage in the long time measurement or an abrupt change of environment, which can be confirmed from the last part. Additionally, we can notice that the method of packaging can effectively keep the reproducibility of each sensor head. When one optical fiber loop cascaded with a FBG works as a sensor head, the transmission intensity is only affected by the bending radius of the optical fiber loop and the peak wavelength which is at the FBG’s centre wavelength, is only affected by temperature, i.e., the fiber loop with polymer coating is highly sensitive to displacement response but independent of temperature response; on the contrary, the fixed FBG is sensitive to temperature response but independent of displacement response; no crosstalk exists between the two parameters. Thus, two parameters are easily obtained. II. EXPERIMENTAL SETUP
The proposed sensor is structured with two slides, a section of standard single mode fiber, and a FBG. We first engrave a half loop cavity in one slide by using a CO2 laser (Han’s Laser, CO2–H3O). Then the section of commercial single mode fiber connected to the FBG (center wavelength is 1550 nm, and 3 dB wavelength band is 0.3 nm) is coiled up into the cavity to fabricate and keep the structure—cascading a fiber loop with polymer protection coating packaged in the circle cavity and a FBG. After that, another slide covers it and secures it tightly. The sensor head making process is shown in Fig. 1. Fig. 2(a) shows the experimental setup of the displacement sensing system. Both ends of the sensor head are fixed on the translation stages, while the sensor keeps vertical with another translation stage. Two of the translation stages are
FIG. 1. Fabricating of the sensor head.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 152.11.242.100 On: Thu, 25 Dec 2014 02:09:25
075002-3
Chen et al.
Rev. Sci. Instrum. 85, 075002 (2014)
FIG. 2. (a) Schematic diagram of the displacement sensing system; (b) schematic diagram of the real time displacement sensing system.
fixed during the experiment to keep the sensor head unmovable, while the third translation stage is movable in the vertical direction, and the displacement of it can be read from the micrometer caliper on it. A beam of light from a broadband source (BBS) with a wavelength range of 1525–1568 nm is coupled into the sensor head. When the light with a wavelength range of 1525 –1568 nm propagates through the fiber loop and reaches the FBG, the reflected light from the FBG only possess a 3 dB wavelength band of 0.3 nm centered 1550 nm, while the transmission light possess the whole wavelength except for the part reflected by the FBG (see inset of Fig. 1). We compose a photo-detector (Thorlabs, DET01CFC) and an oscilloscope (OSC: YOKOGAWA, DL9140), and an optical spectrum analyzer (OSA: YOKOGAWA, AQ6370C) by a 1 × 2 coupler of which the splitting ratio is 1:1 to achieve real time and simultaneous measurement of displacement and temperature. The first part transmission light will be received by a photo-detector (or O-E converter) and shown on the OSC with a resolution of 0.01 V, and the second part transmission light will be detected by an OSA with a spectral resolution of 0.01 nm. To confirm the sensor head respond displacement in real time, the vibration sensing experiment is performed. The vibration source, a commercial speaker, is placed on a Lab jack and driven by a computer, as shown in Fig. 2(b). We vertically elevate the Lab jack and carefully adjust the voltage to quasilinear range (as shown in Fig. 3). When the sensor is subject to the speaker vibration, the changing electronic signals are shown on the OSC. The principle of our proposed sensor head is based on the bending loss sensitive to the equivalent bend radius of the optical fiber loop. The equivalent radius of the optical fiber loop would decrease when the displacement is applied on the optical fiber loop’s top. Fig. 3 shows the relationship between the bending loss and the bending radius. The bending loss dependence on wavelength and bend radius is shown in the inset of Fig. 3. From the inset, we can see that the bending loss varies with the bending radius periodically according to the
theory of Ref. 1. If a monochromatic light is coupled into the sensor, the change in output voltage from the photo-detector is proportional to one of the curves varying with different the bending radius. For the sake of increasing the sensitivity of displacement and detecting temperature simultaneously, the light source is selected as BBS in the experiment rather than a traditional monochromatic laser. Therefore, the change in output voltage at different wavelengths should be summed up to calculate total change in output voltage V(R) and its result is shown in Eq. (1), αBC (λ, R), (1) V (R) = S(λ)D(λ) λ
here, λ represents wavelength, S(λ) is the input laser intensity, D(λ) is the detector responsivity, and R is bend radii, respectively. Based on the BBS wavelength range (1525–1568 nm) in the experiment, the result of output voltage by summing up Eq. (1) is shown in Fig. 3. It is obvious that in the
FIG. 3. Simulation of the change of output voltage dependence on bend radius with a wavelength range from 1525 nm to 1568 nm. (Inset) simulation of the bending loss dependence on radius and wavelength.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 152.11.242.100 On: Thu, 25 Dec 2014 02:09:25
075002-4
Chen et al.
vicinities of 5.2 mm, 6.2 mm, and 7.2 mm, a quasi-linear relationship between output voltage and radius is obtained.7, 8 When the top of the fiber loop with the described radius are subjected to a displacement, the equivalent bending radius of the fiber loop have a little change. That slight change in equivalent bending radius will contribute to output voltage decrease quasi-linearly. As a result, these sections are well-suited for displacement sensing. Owning to the coefficient of thermal expansion of polymer coating smaller than silica, the volume of the bare fiber will be bound by the polymer protection coating which will cancel out the influence of ambient temperature on output voltage. From the analysis above we know the bending loss of fiber loop will affect the output voltage and that structure is independent on temperature. However, the centre wavelength of the FBG’s transmission spectrum is independent on the bending loss, while it is sensitive to the ambient temperature. Consequently, the ambient temperature of the sensor head can be obtained independent of displacement. Therefore, the proposed sensor possesses the ability that the ambient temperature and the displacement can be effectively measured independently. A radius of 6.6 mm is selected for the loop curve in the experiment for the following two reasons: First, from Fig. 3, we can see that if the initial fiber loop radius is larger than 7.5 mm, the output voltage becomes insensitive to the decreasing radius. In this regard, the initial radii should be less than 7.5 mm. Second, technically, a cavity in the slide is very fragile if its radius is less than 5.5 mm. Considering the above two limitations, the radius selections are narrowed down to two sections. We select the smaller one as initial radii of the optical fiber loop in order to pursue a larger sensitivity of displacement. III. EXPERIMENT AND RESULTS
Fig. 4 shows the transmission spectra when the whole configuration is settled at a room temperature and measure at different displacement. At the output, the transmission loss peak wavelength is 1550.08 nm, which corresponds to the resonant peak of the FBG and reflects the ambient temperature
FIG. 4. Transmission spectra of the sensor when the displacements of the sensor head are applied ranging from 0 μm to 50 μm at room temperature.
Rev. Sci. Instrum. 85, 075002 (2014)
FIG. 5. Transmission loss peaks wavelength and output voltages when displacement of the sensor head changes from 0 μm to 50 μm at room temperature.
of the sensor; the intensities of the whole spectra are varied from different displacements, which can be converted to voltage and reflect the displacement. The transmission loss peak of the sensor head is easily distinguished in the spectrum. When displacements change by moving the movable translation stages such as Fig. 2(a) depicts, the output intensities of whole spectrum will change and lead to the varying of the output voltage. By monitoring the output voltage and the transmission loss peak wavelength, the displacements and ambient temperature of the sensor head will be obtained. We apply displacement to the sensor head and keep temperature stable, and monitor the output voltage and the peak wavelength via the OSC and OSA. The results are shown in Fig. 5. As we expected, the output voltage decreases as the equivalent radius decreases and it can be fitted as y = −0.14314x + 11.42524; while the peak wavelength is stable (or more accurately, the changing of transmission loss peak wavelength in the range of system resolution—0.01 nm) and it can be fitted as y = 1150.08. The sensitivity of the sensor head for displacement is 0.14314 V/μm ranging from 0 to 50 μm. The error bars of the experimental data in x- and y- axes are times 5 and 10 in order to make it visible. Fig. 6 shows the transmission spectra when the whole configuration is fixed at 0 μm displacement and settled at a temperature ranging from 20 ◦ C to 75 ◦ C. The inset is the magnification of the peaks at different temperatures. The output voltage is monitored at the same time by OSC. Both the wavelength and output voltage are shown in Fig. 7. From Fig. 7, we can see that the output voltage in different temperatures keeps a constant value (or more accurately, the changing of the voltage is under the system resolution—0.01 V) and it can be fitted as y = 12.18; while the transmission loss peak wavelength varies quasi-linearly as the temperature and it can be fitted as y = 0.00795x −0.18381. The sensitivity of the sensor head for temperature is 0.00795 nm/ ◦ C ranging from 20 ◦ C to 75 ◦ C. Notice that the error bar in the x-axis is times 10 for the sake of making the error bar visible. To do more research and confirm the real time characteristics of the sensor head respond, a vibration experiment
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 152.11.242.100 On: Thu, 25 Dec 2014 02:09:25
075002-5
Chen et al.
FIG. 6. Transmission spectra of the sensor head when temperature applied range from 20 ◦ C to 75 ◦ C.
is performed, and the processing is described in Fig. 2(b). A 5 V, 500.00 Hz sinusoidal voltage waveform is applied experimentally. Fig. 8(a) shows a representative measurement result at time domain and its corresponding frequency domain. The inset of Fig. 8(a) is the magnification of the red circle and shows the details of the measured signal. The output voltage of the sensor varied alike a sinusoidal waveform with very uniform amplitude. And the demodulated frequency agrees well with the electric signal, with the SNR of about 65 dB, as shown in Fig. 8(b). It is worth mentioning that the corresponding harmonics of the induced vibration frequency is also detected, which may be due to the slightly non-axial alignment between the speaker and the sensor.20 Nevertheless, the tone of the main frequency overcomes the second, third, and fourth harmonics with a difference of 43, 38, and 53 dB, respectively, which allows clear identification of the induced vibration frequency. However, what need alludes is the temperature measurement will be limited if the influence of the spectrum contributed by the amplitude of a vibration source is comparable with a loss amplitude of a FBG.
FIG. 7. Transmission loss peak wavelength and output voltage of the sensor when whole configuration is fixed at 0 μm and settled it at a temperature ranging from 20 ◦ C to 75 ◦ C.
Rev. Sci. Instrum. 85, 075002 (2014)
FIG. 8. Responses of the sensor head to speaker driven by a 5 V, 500 Hz sinusoidal wave in (a) time and (b) corresponding frequency domain.
Obviously, when a fiber loop with polymer coating packaged in the circle cavity is cascaded with a FBG, simultaneous measurement of displacement and temperature can be achieved by detecting the output voltage and the transmission loss peak wavelength of the FBG. Moreover, since existing the protection polymer coating, the proposed senor is reliable mechanically which can be used in long time and real time measurement. Additionally, the crosstalk with temperature or ambient refractive index due to the material thermal expansion coefficients and evanescent field can be totally cancelled out (In fact, the experiments about put the single fiber loop with polymer coating in the temperature chamber ranging from 20 ◦ C to 75 ◦ C and drip a liquid with 1.8 index of refractive are performed. Although the outcomes are not plotted here, we still can image easily that it does not have any effects on output voltage and wavelength due to the protection of polymer coating). The disadvantage of this sensor is output voltage non-monotonically dependent on decrease of equivalent bend radius, but it can be easily solved by set a limit—50 μm—on the range of displacement measurement. In the last part, we make the sensor reach its maximum deformation in order to check its ability survive in the harsh environment. Both ends of the sensor head are fixed on the translation stages, while the sensor maintains a vertical position with another translation stage. Two of the translation stages are fixed during the experiment to keep the sensor unmovable; another is movable in the vertical direction and inward to induce bending in the sensor. The steel beam on the
FIG. 9. (a) Original shape of the sensor, (b) Steel beam forces the sensor to maximum bending point, (c) Steel beam moves outward to allow sensor to resume its shape.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 152.11.242.100 On: Thu, 25 Dec 2014 02:09:25
075002-6
Chen et al.
vertical translation stages moves until the slide bends to a maximum without breaking and then moves outward to allow the sensor to resume its shape. These processes are pictured in Fig. 9, which shows the sensor possesses the ability to withstand a maximum deformation and still can work when the environment improves. IV. CONCLUSION
In conclusion, we present a simple structure for simultaneous displacement and temperature measurement, by utilizing the transmission intensity of an optical fiber loop’s decrease as the equivalent bending radius decreases and independent of ambient temperature; red shifting response of a FBG’s peak wavelength due to the increasing of ambient temperature; and the compatibility of the two components. After applying suitable packaging, the proposed sensor head is able to measure displacement and temperature with sensitivities reaching up to 0.14314 V/μm for displacement in the range of 0–50 μm and 0.00795 nm/◦ C for temperature between 20 ◦ C and 75 ◦ C. Additionally, such a sensing method can be used for real time and simultaneous measurement of displacement and temperature in the building industry, such as bridges, dam buildings, and so on. Moreover, because of the polymer coating protection, the structure is reliable even under harsh condition in mechanic, and even certain chemical parameters such as refractive-index, humidity, and so on. ACKNOWLEDGMENTS
This work was supported by the National Natural Science Foundation under Grant Nos. 11274181, 10974100,
Rev. Sci. Instrum. 85, 075002 (2014)
10674075, 11104149, the Doctoral Scientific Fund Project of the Ministry of Education under Grant No. 20120031110033 and by the Tianjin Key Program of Application Foundations and Future Technology Research Project under Grant No. 10JCZDJC24300, China. 1 Q.
Wang, G. Farrell, and T. Freir, Opt. Exp. 13, 4476 (2005). H. Nam and S.-Z. Yin, IEEE Photon. Technol. Lett. 17, 2391 (2005). 3 Q. Wang, G. Rajan, P.-F. Wang, and G. Farrell, Opt. Exp. 15, 4909 (2007). 4 P.-F. Wang, G. Farrell, and Y. Semenova, Appl. Opt. 48, 3055 (2009). 5 Q. Wang, G. Farrell, T. Freir, G. Rajan, and P.-F. Wang, Opt. Lett. 31, 1785 (2006). 6 P.-F. Wang, G. Brambilla, Y. Semenova, Q. Wu, and G. Farrell, J. Opt. 13, 075402 (2011). 7 P.-F. Wang, Y. Semenova, Q. Wu, G. Farrell, Y.-Q. Ti, and J. Zheng, Appl. Opt. 48, 6044 (2009). 8 P.-F. Wang, Y. Semenova, Q. Wu, J. Zheng, and G. Farrell, Appl. Opt. 49, 1744 (2010). 9 J. H. Ng, X.-Q Zhou, X.-F Yang, and J.-Z. Hao, Opt. Commun. 273, 398 (2007). 10 X.-Y. Dong, X.-F. Yang, C.-L. Zhao, L. Ding, P. Shum, and N. Q. Ngo, Smart Mater. Struct. 14, N7 (2005). 11 C.-Y. Shen and C. Zhong, Sens. Actuat. A 170, 51 (2011). 12 Q. Jiang and D. Hu, IEEE Sens. J. 11, 1776 (2011). 13 J. M. Baptista, S. F. Santos, G. Rego, O. Frazáo, and J. L. Santos, Opt. Commun. 260, 8 (2006). 14 L. Qi, C.-L. Zhao, Y.-P. Wang, J. Kang, Z.-X. Zhang, and S. Z. Jin, Opt. Exp. 21, 3193 (2013). 15 L.-L. Xu, Y. Li, and B.-J. Li, Opt. Lett. 38, 64 (2013). 16 C.-K. Ji, C.-L. Zhao, J. Kang, X.-Y. Dong, and S.-Z. Jin, Rev. Sci. Instrum. 83, 053109 (2012). 17 P.-F. Wang, Y. Semenova, A. Sun, Q. Wu, and G. Farrell, Proc. SPIE 7726, 772623 (2010). 18 D. Tosi, E. Gino Macchi, G. Braschi, A. Cigada, M. Gallati, S. Rossi, S. Poeggel, G. Leen, and E. Lewis, Appl. Opt. 53, 2136 (2014). 19 R.-P. Hu and X.-G. Huang, IEEE Sens. J. 9, 1952 (2009). 20 Y. Huang, T. Guo, C. Lu, and H. Y. Tam, IEEE Photon. Technol. Lett. 22, 1235 (2010). 2 S.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 152.11.242.100 On: Thu, 25 Dec 2014 02:09:25
