Abstract
A micro-newton strain force and temperature synchronous fiber sensor with a high Q-factor is proposed. The sensor is based on a commercial quartz microbubble (QMB, the diameter is less than 80 µm) that is attached to the end surface of the suspending taper integrated in the hollow core fiber. The multi-beam interference and long-active-length make the sensor show both high sensitivity (0.150 nm/mN) and Q-factor (1470 based on the 3dB-bandwidth). The actual detection limit of the strain force reaches about 50 µN. The UV-cured polymer between the QMB and taper improves the temperature sensitivity. The strain force and temperature can be demodulated synchronously by using band-pass filtering and sensing matrix. The sensor can have actual application in micro-newton strain force detection as its low cost and flexible structure.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Optical fiber structures such as Mach-Zehnder interferometers (MZIs) [1,2], Fabry-Perot interferometers (FPIs) [3,4], fiber brag gratings (FBGs) [5,6] have been widely used for strain force sensing. Among these kinds of structures, sensors based on FPIs are most popular because of their advantages such as small size, good durability, and ease to fabrication. Air cavities based on different shapes such as cylinder [7,8], oblate spheroid [9,10], cuboid [11] and expanded bubble [12–14] have been fabricated between single mode fibers (SMFs) to form fiber in-line FPIs. However, their strain force sensitivities are generally less than 0.04 nm/mN due to the limitation of the interference principle [9,11].
To further improve the sensitivity, a kind of long-active-length FP strain force sensor based on the suspending taper in the fiber has been reported [15–18]. The interference length Li is separate from the cavity active-length La, and the strain force sensitivity is proportional to the ratio of La-to-Li. When the active-length is increased to 1360 µm by using capillary-taper structure, the strain force sensitivity reaches as high as 0.841 nm/mN [18]. However, as the FP interference length of these structures is only less than 10µm, the free spectrum range (FSR) is more than 200 nm. Meanwhile, as the large reflection loss at the taper end face, the interference quality degenerates. As a result, the Q-factor of the interference valley based on 3dB-bandwidth is generally less than 70 for this kind of sensor. In fact, the Q-factor and sensitivity both determinate the detection limit (DL) [19], which is the key parameter to evaluate the actual performance of the sensor. The sensitivities of the above reported sensors are improved by sacrificing the Q-factor values, which influences the actual DL. There needs to be an effective way to improve the sensitivity and Q-factor at the same time. For common fiber in-line FPIs, multi-beam interferences based on the cascading cavities are often employed to increase the Q-factor values. For example, Deng et al. proposed a high-temperature sensor by use of cascaded FPI based on four in-fiber mirrors. The Q-factor of their sensor is more than 1800, which effectively improves the DL value [20]. Therefore, the mothed of multi-beam interference can be introduced to the strain force sensor with a long-active-length.
Besides, the strain force sensor is usually used in the environment with a temperature variation, where the strain force and temperature are needed to be detected at the same time. A temperature sensor such as FBG can be combined with the long-active-length strain force sensor [21]. However, the reported FBGs and long-active-length strain force sensors show low temperature sensitivities (generally less than 30 pm/°C) [21] due to the low thermo-optical coefficient (TOC) and thermal expansion coefficient (TEC) of quartz material.
In this paper, we proposed a long-active-length fiber strain force sensor with both high sensitivity and high Q-factor. A commercial quartz microbubble (QMB, the diameter is less than 80 µm) is attached to the end surface of the suspending taper by the UV polymer. The suspending taper is fabricated by arc discharge fiber tapering, and its length can be more than 3000 µm. The taper is then plugged into a section of HCF that is spliced to the SMF in advance. The joint between the taper and HCF is sealed by arc discharge in a Fiber Fusion Splicer. The end face of SMF and two side faces of the QMB form three reflection mirrors. Owing to the multi-beam interference, the Q-factor of the interference valley based on the 3dB-bandwidth is improved to be 1470. Meanwhile, owing to the long active length, the strain force sensitivity reaches 0.150 nm/mN. As a result, the actual DL of the strain force is about 50 µN. Besides, two different interference fringes can be extracted individually by using band-pass filtering. They are both sensitive to strain force and temperature. The two parameters can be demodulated synchronously by using sensing matrix with the maximum sensitivities of 0.667 nm/mN and −0.251 nm/°C.
2. Theory
The schematic diagram of the sensor is shown in Figs. 1(a) and 1(b). The length between the SMF end face and left-face of the QMB is L1. The diameter of the QMB is L2. The length of HCF is L. The side wall thickness of the QMB is about 200nm, which can be ignored. There is a section of UV-cured polymer whose length is L3 between the QMB and taper.
The previously reported strain sensors are usually based on two-beam FPI. We assume that the first reflection face is the flat end-face of the SMF and the second reflection face is the interface between air and quartz. The length of the interference length is assumed to be L1. According to the FP interference theory [22], the reflection spectrum can be expressed as
For our structure, as the refractive index (RI) of the UV polymer is close to the quartz material, the reflections at the interfaces QMB-UV polymer and UV polymer-taper can be ignored. The end-face of SMF and two side faces of the QMB form three reflection mirrors. The sensor in this paper is mainly based on three-beam FPI. According to the theory of three-beam FPI [22], the reflection spectrum can be expressed as
There are generally three types of interference fringes superimposed in the reflection spectrum of the common three-beam FPI [23], which corresponds to three FP cavities formed by two of three reflection mirrors. As the reflectivity of the two side faces of the QMB is much smaller than that of the SMF due to the transmission and reflection losses, the interference fringe formed by the cavity between the two side faces of the QMB can be ignored. Therefore, there are mainly two types of interference fringes superimposed in the reflection spectrum corresponds to the cavity length L1 and L1 + L2. The central wavelength values of the mth-order interference valley for the first-type interference fringe and m´th-order interference valley for the second-type interference fringe can be expressed as
As the two types of interference fringes have different FSR values, they can be extracted individually by using band-pass filtering.When the sensor is stretched, the central wavelength shift sensitivity of the first-type interference fringe can be expressed as
In the same way, the central wavelength shift sensitivity of the second-type interference fringe can be expressed as
When the temperature changes, the UV polymer between the QMB and taper will expand or shrink, causing L1 and L1 + L2 changing. The temperature sensitivity for the first-type interference fringe can be expressed as
where, α=∂L3/(L3∂T) is the TEC of the UV polymer. In the same way, the temperature sensitivity for the second-type interference fringe can be expressed as3. Fabrication
The fabrication process includes three steps as shown in Fig. 2. In the first step, the structure consisting of a section of SMF and a short section of HCF (SMF-HCF) was used to catch a QMB as shown in Fig. 2(a). The HCF is made by drawing the quartz capillary by Rui-feng chromatographic device company. A few QMBs purchased from 3M Corporation with a type of GB-K1 were put on a glass slide. SMF-HCF was controlled by a five-dimensional displacement platform that consists of three translation axes and two rotary axes driven by five manual spiral micrometers. The opening of the SMF-HCF was moved to reach the QMB with a proper size by the adjusting three translation axes, and then the SMF-HCF was controlled to be perpendicular to the upper surface of the glass slide by adjusting two rotary axes. The QMB was caught in the HCF and it would attach on the inner wall of the HCF due to Van Der Waals force. In the second step, a fiber taper was prepared beforehand by using the two-step arc discharge fiber tapering method [14], and then some UV polymer (NOA81) was dipped on the end surface of the fiber taper. The fiber taper was placed on a displacement platform and it was adjusted to make the end surface of fiber taper touch QMB as shown in Fig. 2(b). Once the location of the fiber taper and QMB was adjusted well, the structure was exposed in UV light (365nm) until the UV polymer was absolutely solidified. In the last step, the taper was then plugged into a section of HCF that was spliced to the SMF in advance as Fig. 2(c). This step was operated in a Fiber Fusion Splicer, and the joint between the taper and HCF is sealed by arc discharge. It’s worth noting that the QMB will be destroyed by high temperature if it is too close from the discharging point. Therefore, the length of the fiber taper should be more than 3mm. The image of the final fabricated fiber sensor is shown in Figs. 2(d) and 2(e). The diameter of the QMB is about 75 µm and the diameter of the fiber taper is about 60 µm. The fiber taper is strong enough to support the QMB. The thickness of the polymer is about 10 µm. The inner and outer diameter of HCF is 80 µm and 125 µm respectively. The length of HCF is about 4500 µm and the length of cavity 1 is about 14 µm.
The measured reflection spectrum of the sensor is shown in Fig. 3(a), and the fringe pattern is similar with the simulation result shown in Fig. 1(d). The Q-factor value of the interference valley near the wavelength 1430 nm based on the 3dB-bandwidth is as high as 622 owing to the superposition of interference fringes. Figure 3 (b) is the fast Fourier transform (FFT) result of the spectrum. There are two main peaks in the frequency spectrum, which corresponds to the interference length L1 and L1 + L2. Their spatial frequencies are 0.015/nm and 0.080/nm. The two types of interference fringes are extracted individually by using band-pass filtering. For the second dominant peak (Peak II), there is a small peak near its left, which corresponds to the interference accrued in the QMB. It will influence the filtering result if the bandwidth for the band-pass filtering is larger than 0.01/nm. Therefore, the bandwidth is set as 0.009/nm. The spectrum results are shown in Figs. 3(c) and 3(d). The fringe I with larger FSR results from the interference cavity with the length of L1, and the fringe II with smaller FSR results from the interference cavity with the length of L1 + L2. The two kinds of fringes can also be achieved by using Inverse Fourier Transform (IFT) method based on the frequency values of the two dominant peaks, which can avoid the influence of the bandwidth selection on band-pass filtering.
4. Experimental results and discussions
Strain experiments were carried out with the sensor. Strain force was applied by loading objects on the sensor. The value of strain force is increased by increasing the mass of the objects that is measured by an Electronic Scale. At each strain force, the reflection spectrum was recorded, and two types of interference fringes (I and II) were extracted from the original spectrum by using band-pass filtering. The original spectrum and the extracted two types of interference fringes at different strain force conditions are shown in Figs. 4(a)–4(c). Three spectra all redshift with the increase of the strain force. The interference valley near wavelength 1430 nm was chosen to show its wavelength shift as the variation of the strain force for three spectra. The linear fitted results are shown in Fig. 4(d). The sensitivities of fringe I and II are 0.667 nm/mN and 0.092 nm/mN, which are basically consistent with the calculated theory values (0.683 nm/mN and 0.107 nm/mN). The little difference between the calculated and experimental sensitivity values may result from the error of the measurement of the interference length. The sensitivity of the superimposed spectrum is 0.124 nm/mN, which is between the sensitivities of fringe I and II agreeing with the theoretical analysis. The three spectra all show good linearities with values of 99.9%, 99.2% and 98.5%, respectively.
Temperature experiment was also carried out by heating the sensor in the water bath. The sensor was heated from 30°C to 70°C with a step of 5℃. The original spectrum and the extracted fringes I and II all redshift with the increase of the temperature as shown in Figs. 5(a)–5(c). The linear fitted results between the wavelength shift and temperature are shown in Fig. 5(d). The temperature sensitivity of the fringe I is −0.251 nm/℃ with a linearity of 99.5%, and the temperature sensitivity of the fringe II is −0.038 nm/℃ with a linearity of 98.3%. The calculated theory values for fringes I and II are −0.266 nm/°C and −0.042nm/°C. The calculated and experimental sensitivity values also coincide with each other. The sensitivity of the superimposed spectrum is −0.049 nm/°C with a linearity of 98.2%. Owing to the long active length and large TEC of the UV polymer, the sensor shows both high sensitivities in strain force and temperature sensing. As fringes I and II can be extracted from the single sensor at the same time, sensitive synchronous detection of strain force and temperature can be achieved by using the follow sensing matrix based on the fringes I and II:
Here, S1-F =0.667 nm/mN, S1-T =−0.251 nm/°C, S2-F =0.092 nm/mN, and S2-T =−0.038 nm/°C. The solutions of the inequalities can be shown in the coordinate system by using the linear programming. The result is shown in Fig. 6. The variations of strain force and temperature can only be detected when they are beyond the area enclosed by the red lines. Smaller area shown by using the linear programming results in smaller value of DL. From Fig. 6 we can also see that, the DL of the strain force is influenced by that of the temperature. If the environmental temperature fluctuates in a large range, the DL of the strain force will degenerate. If the values of R and temperature fluctuation is known, the DL value can be obtained from Fig. 6.
As the Q-factor value of the interference valley in the original superimposed spectrum near the wavelength 1430nm is 622, the sensor is expected to show better wavelength demodulation resolution and smaller wavelength shift noise. Therefore, it has additional advantages in small strain force detection. If the synchronous detection of strain force and temperature is not required, the QMB can be fixed on the end face of the taper by using CO2 laser heating [25]. As the UV polymer is not used, the influence of the temperature crosstalk on strain force detection can be ignored due to the small TEC (10−7/°C) of quartz material. To show the advantage of the sensor in small strain force detection, we fabricated another sensor (L1 = 8 µm, L = 3500 mm) to further improve the strain force sensitivity and the Q-factor. The measured reflection spectrum is shown in Fig. 7(a). As the transmission and reflection losses are reduced as the decrease of L1, the fringe contrast of the spectrum near the wavelength 1470 nm reaches about 19 dB. As a result, the Q-factor based on the 3dB-bandwidth reaches as high as 1633. The strain force response of the sensor was also tested, the sensitivity results are shown in Fig. 7(b). The strain force sensitivities of fringe I and II are 0.991 nm/mN and 0.089 nm/mN. The strain force sensitivity of the chosen valley with high Q-factor in the actual reflection spectrum is 0.150 nm/mN.
To test the actual DL in small strain force detection, an extremely light object with the mass of 0. 02g (corresponding to the strain force of 196 µN) was loaded on the sensor and then removed for three cycles in constant room temperature. In each step, the reflection spectrum of the sensor was recorded repeatedly for 20 times at an interval of 2 s. The central wavelength value of the chosen valley with high Q-factor is measured by using Lorentz fitting of the valley. The wavelength shift values in the whole test process are shown in Fig. 8(a). From the result we can see, the sensor can easily show the small strain force change of 196 µN. The measured wavelength errors at fixed strain force for three times are shown in Fig. 8(b). The wavelength variation is about 0.008 nm owing to the large Q-factor value [19]. By putting this value and sensitivity 0.150 nm/mN into DL = R/S, DL can be measured to be 53 µN.
The Q-factor and sensitivity of the proposed sensor is compared with other reported strain force sensors as shown in Table 1. Our sensor shows large Q-factor and high sensitivity at the same time. Besides, synchronous measurement of strain force and temperature can be achieved by using the single sensor, which further increases the competitiveness of the sensor in actual application.
5. Conclusion
A suspending taper and QMB was integrated in the HCF and a long-active-length FP strain force sensor is achieved. The suspending taper structure can effectively improve the strain fore sensitivity. Meanwhile, the three reflection faces increase the Q-factor value of the interference fringe, and the DL of about 50 µN is realized for the proposed strain force sensor. Though the QMB can be fixed on the end-face of the taper by using CO2 laser heating, UV-cured polymer can be used between the QMB and taper to achieve sensitive temperature detection at the same time. The strain force and temperature can be demodulated synchronously by using band-pass filtering and sensing matrix.
Funding
National Natural Science Foundation of China (11874010, 11874133); Natural Science Foundation of Shandong Province (ZR2021MF111).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. H. Zhang, M. Zhang, J. Kang, X. Zhang, and J. Yang, “High Sensitivity Fiber-Optic Strain Sensor Based on Modified Microfiber-Assisted Open-Cavity Mach-Zehnder Interferometer,” J. Lightwave Technol 39(13), 4556–4563 (2021). [CrossRef]
2. S. Wang, Y. Ma, W. Chen, S. Wang, Y. Yi, X. Li, C. Lu, S. Zhang, T. Geng, W. Sun, and L. Yuan, “Ultrasensitive Strain Sensor Based on Mach-Zehnder Interferometer With Bent Structures,” J. Lightwave Technol 39(21), 6958–6967 (2021). [CrossRef]
3. R. M. André, S. Pevec, M. Becker, J. Dellith, M. Rothhardt, M. B. Marques, D. Donlagic, H. Bartelt, and O. Frazão, “Focused ion beam post-processing of optical fiber Fabry-Perot cavities for sensing applications,” Opt. Express 22(11), 13102–13108 (2014). [CrossRef]
4. E. Cibula and D. Donlagic, “In-line short cavity Fabry-Perot strain sensor for quasi distributed measurement utilizing standard OTDR,” Opt. Express 15(14), 8719–8730 (2007). [CrossRef]
5. R. Oliveira, L. Bilro, T. H. R. Marques, C. M. B. Cordeiro, and R. Nogueira, “Strain Sensitivity Enhancement of a Sensing Head Based on ZEONEX Polymer FBG in Series With Silica Fiber,” J. Lightwave Technol 36(22), 5106–5112 (2018). [CrossRef]
6. R. Zhang, Z. Zhu, and G. Wu, “Static pure strain sensing using dual–comb spectroscopy with FBG sensors,” Opt. Express 27(23), 34269–34283 (2019). [CrossRef]
7. G. K. B. Costa, P. M. P. Gouvêa, L. M. B. Soares, J. M. B. Pereira, F. Favero, A. M. B. Braga, P. Palffy-Muhoray, A. C. Bruno, and I. C. S. Carvalho, “In-fiber Fabry-Perot interferometer for strain and magnetic field sensing,” Opt. Express. 24(13), 14690–14696 (2016). [CrossRef]
8. Y. Ouyang, H. Guo, X. Ouyang, Y. Zhao, Z. Zheng, C. Zhou, and A. Zhou, “An In-Fiber Dual Air-Cavity Fabry–Perot Interferometer for Simultaneous Measurement of Strain and Directional Bend,” IEEE Sensors J 17(11), 3362–3366 (2017). [CrossRef]
9. F. C. Favero, L. Araujo, G. Bouwmans, V. Finazzi, J. Villatoro, and V. Pruneri, “Spheroidal Fabry-Perot microcavities in optical fibers for high-sensitivity sensing,” Opt. Express 20(7), 7112–7118 (2012). [CrossRef]
10. P. Zhang, L. Zhang, Z. Wang, X. Zhang, and Z. Shang, “Sapphire derived fiber based Fabry-Perot interferometer with an etched micro air cavity for strain measurement at high temperatures,” Opt. Express 27(19), 27112–27123 (2019). [CrossRef]
11. Y. Liu, S. Qu, W. Qu, and R. Que, “A Fabry–Perot cuboid cavity across the fibre for high-sensitivity strain force sensing,” J. Opt. 16(10), 105401 (2014). [CrossRef]
12. S. Liu, Y. Wang, C. Liao, G. Wang, Z. Li, Q. Wang, J. Zhou, K. Yang, X. Zhong, J. Zhao, and J. Tang, “High-sensitivity strain sensor based on in-fiber improved Fabry–Perot interferometer,” Opt. Express 39(7), 2121–2124 (2014). [CrossRef]
13. C. Yin, Z. Cao, Z. Zhang, T. Shui, R. Wang, J. Wang, L. Lu, S. Zhen, and B. Yu, “Temperature Independent Ultrasensitive Fabry–Perot All-Fiber Strain Sensor Based on a Bubble-Expanded Microcavity,” IEEE Photonics J. 6(4), 6802009 (2014). [CrossRef]
14. Z. Yong, X. Feng, M. Chen, R. Tong, and P. Yun, “Optical Fiber Axial Contact Force Sensor Based on Bubble-Expanded Fabry–Pérot Interferometer,” Sens. Actuators, A 270, 318–326 (2018). [CrossRef]
15. S. Pevec and D. Donlagic, “All-fiber, long-active-length Fabry-Perot strain sensor,” Opt. Express 19(16), 15641–15651 (2011). [CrossRef]
16. K. Zhou, M. Ai, Z. Qian, X. Gao, Z. Hu, Q. Li, L. Yuan, Z. Wang, Y. Huang, T. Tu, J. Cui, C. Li, and G. Guo, “High-sensitivity strain sensor with an in-fiber air-bubble Fabry-Perot interferometer,” Applied Physics Lett 113(18), 181901 (2018). [CrossRef]
17. T. Yang, Z. Ran, X. He, Z. Li, Z. Xie, Y. Wang, Y. Rao, X. Qiao, Z. He, P. He, Y. Yang, and F. Min, “Temperature-Compensated Multifunctional All-Fiber Sensors for Precise Strain/High-Pressure Measurement,” J. Lightwave Technol 37(18), 4634–4642 (2019). [CrossRef]
18. Y. Liu, C. Lang, X. Wei, and S. Qu, “Strain force sensor with ultra-high sensitivity based on fiber inline Fabry-Perot micro-cavity plugged by cantilever taper,” Opt. Express. 25(7), 7797–7806 (2017). [CrossRef]
19. I. M. White and X. Fan, “On the performance quantification of resonant refractive index sensors,” Opt. Express 16(2), 1020–1028 (2008). [CrossRef]
20. J. Deng, D. N. Wang, and H. Zhang, “Femtosecond Laser Inscribed Multiple In-Fiber Reflection Mirrors for High-Temperature Sensing,” J. Lightwave Technol 37(21), 5537–5541 (2019). [CrossRef]
21. X. Zhao, Y. Zhang, W. Zhang, Z. Li, L. Kong, L. Yu, J. Ge, and T. Yan, “Ultra-high sensitivity and temperature-compensated Fabry–Perot strain sensor based on tapered FBG,” Opt & Laser Technol 124, 105997 (2020). [CrossRef]
22. Z. L. Ran, Y. J. Rao, W. J. Liu, X. Liao, and K. S. Chiang, “Laser-micromachined Fabry-Perot optical fiber tip sensor for high-resolution temperature-independent measurement of refractive index,” Opt. Express 16(3), 2252–2263 (2008). [CrossRef]
23. W. Zhang, H. Li, L. Zhu, M. Dong, and F. Meng, “Dual-Parameter Optical Fiber Probe Based on a Three-Beam Fabry-Perot Interferometer,” IEEE Sensors J. 21(4), 4635–4643 (2021). [CrossRef]
24. O. Arrizabalaga, G. Durana, J. Zubia, and J. Villatoro, “Accurate microthermometer based on off center polymer caps onto optical fiber tips,” Sens. Actuators B 272, 612–617 (2018). [CrossRef]
25. K. Tian, J. B. Yu, X. Wang, H. Y. Zhao, D. J. Liu, E. Lewis, G. Farrell, and P. Wang, “Miniature Fabry-Perot interferometer based on a movable microsphere reflector,” Opt. Lett 45(3), 787–790 (2020). [CrossRef]