Open Access
Add to BibSonomyAbstract
In the present study, a novel packaged long-period fiber grating (PLPFG) strain sensor is first presented. The MEMS process was utilized to fabricate the packaged optical fiber strain sensor. The sensor structure consisted of etched optical fiber sandwiched between two layers of thick photoresist SU-8 3050 and then packaged with poly (dimethylsiloxane) (PDMS) polymer material to construct the PLPFG strain sensor. The PDMS packaging material was used to prevent the glue effect, wherein glue flows into the LPFG structure and reduces coupling strength, in the surface bonding process. Because the fiber grating was packaged with PDMS material, it was effectively protected and made robust. The resonance attenuation dip of PLPFG grows when it is loading. This study explored the size effect of the grating period and fiber diameter of PLPFG via tensile testing. The experimental results found that the best strain sensitivity of the PLPFG strain sensor was −0.0342 dB/με, and that an R2 value of 0.963 was reached.
© 2014 Optical Society of America
1. Introduction
Because long-period fiber grating (LPFG) has many special features, including high sensitivity, light weight, and a small diameter, in addition to not being subject to the interference of electromagnetic waves, it has been commonly applied in the monitoring and measurement of many physical quantities such as strain [1], pressure [2], and torsion [3], as well as temperature [4–6] in highly corrosive and electromagnetic environments. Among these measurements, the measurement of strain is a most indispensable part of the industrial engineering field. The novel packaged long-period fiber grating (PLPFG) strain sensor proposed in the present study consists of small cladding fiber with a periodic polymer grating structure inside packaging material. The sensing principle of a PLPFG strain sensor is such that, when the fiber grating is acted upon by a strain, the periodic grating structure will generate a periodic deformation to cause a periodic change in the refractive index, there by forming a long-period fiber grating. When light is being transmitted, the optical attenuation dip loss and the change of resonant wavelength will be generated according to the phase matching condition, thus causing a change in the optical spectrum, and the change in the strain value and the change in the resonant attenuation dip loss will have a linear relationship.
In 1992, M. Vaziri et al. [7] proposed an approach using epoxy resin as protective layers coupled with a chemical etching method to produce an LPFG strain gauge; however, the fiber structure formed by this chemical etching method will be uneven, such that its grating period will not be effectively controlled. Furthermore, the strength of the resulting sensor is questionable. In 2001, L.A. Wang et al. [8] used metal film as protective layers coupled with hydrofluoric acid etching to produce a corrugated LPFG, and used it to measure physical quantities including strain, bending, and torsion. The strain sensitivity of such LPFG can reach −0.037 dB/με, but its measurement range is too small and not robust (i.e., it is weak without packaging), so its application will be limited. In 2006, Y.P. Wang et al. [9] used a CO2 laser method to produce an LPFG sensor on fiber, with one characteristic of such a sensor being that it can significantly reduce the interactions caused by temperature. Therefore, it is an athermal strain sensor. When conducting strain loading, its strain sensitivity can reach −7.6 pm/με, while its temperature sensitivity can be reduced to 3.91 pm/°C. In 2007 and 2009, T. Zhu et al. [10, 11] presented an ultra-long-period fiber grating with a periodic groove structure (G-ULPFG) fabricated via high-frequency CO2 laser using an edge-written method. The experimental results show that G-ULPFG could be used as a highly sensitive optical refractometer with temperature self-compensation. However, the G-ULPFG is not robust owing to the groove structure on the fiber and is also not suitable for mass production. In 2008, C.L. Zhao et al. [12] used a laser processing method to produce an LPFG structure on photonic crystal fiber, with a strain resolution reaching 0.3 με. In 2008, P. Caldas et al. [13] used an arc discharging method to produce an LPFG, and constructed a precision stage to conduct linear isotropic stretching with tensile strain and stretching of up to 2000 με; the tested strain sensitivity reached 0.172 pm/με. In 2012, Z. Xin et al. [14] used a CO2 laser method to produce a structure of LPFG on fiber, and constructed a fixed platform to exert a force by way of hanging weights on a weighing pan to achieve the effect of force measurement, and the force sensitivity reached 0.75 dB/N.
Other major methods for producing LPFG include the amplitude mask method [15, 16], the mechanical pressure method [17], and the photoresist method [18]. The gratings produced by the aforementioned methods are contactable and open-type, so if a strain sensor is adhered onto the object to be measured, the adhesive will penetrate around the grating, affecting the refractive index of the fiber, and thus directly affecting its measurement accuracy. Therefore, this study proposed an optical fiber strain sensor based on PLPFG. The proposed PLPFG is an etched optical fiber that is sandwiched between two thick photoresists with a periodic structure and packaged within PDMS polymer. When surface bonding the PLPFG strain sensor onto an object to be measured, the PDMS can isolate the adhesive and prevent it from penetrating into optical fiber grating. So, this method not only can solve the problem of difficult to adhere/surface-bonding LPFG onto an object to be measured, but can also enhance the strength and stability of the grating. Moreover, the manufacturing cost and the strain sensitivity of the PLPFG strain sensor proposed by this study are better than those of the sensors suggested by the aforementioned studies, and the proposed PLPFG strain sensor also has convenience.
2. Operating principle of the PLPFG strain sensor
When the grating period length of a fiber grating is within 100 μm~1 mm, the grating is defined as an LPFG. The LPFG consists of the periodic refractive index change. When light is being transmitted in LPFG, the periodic refractive index grating structure will generate the resonant attenuation dip in the spectrum based on the coupled mode theory [18]. In this study, the periodic refractive index variations of the proposed PLPFG are caused by the external strain loading and the periodic composite soft (PDMS) and hard (SU-8) polymer structures of the PLPFG. When external loading is applied on the PLPFG, the strain will rise according to the periodic soft (PDMS) and hard (SU-8) areas. The resulting periodic refractive index variance in the fiber forms the spectrum character of the LPFG. Therefore, the strain loading can tune the resonant dip attenuation loss of the PLPFG. When this PLPFG suffers strain, the PLPFG is reflected as loss variation of the resonant dip. Hence, we can measure the strain variation by monitoring the resonant dip loss.
The resonant wavelength phase-matching condition of LPFG can be calculated as follows:
where Λ is the grating period, is the effective refractive index of the core mode, and is the effective refractive index of the cladding mode. And when light is being transmitted in LPFG, the transmission loss can be controlled according to the couple mode theory [18, 19] for LPFG, and the transmission loss is calculated as:where L indicates the length of the LPFG, and is the AC component of the coupling coefficient between the core and the cladding. The transmission loss of a PLPFG can be deduced from the AC component of the coupling coefficient between the core and the cladding. Transmission loss is a function of , which is proportional to the amplitude of changes in the refractive index because of variation in the strain field.From the above formula, it can be seen that the transmission loss of an LPFG is a cosine-squared function, and that the transmission loss and the resonant wavelength are related to the coupling coefficient and grating length. Therefore, the loss can be changed by external loading. From Eq. (2), we can measure the strain by monitoring the transmission loss of the PLPFG.
3. Process of the PLPFG
3.1 PLPFG strain sensor
The photolithography MEMS process was adopted as the process for producing the PLPFG strain sensor, and the material used in the process included SU-8 3050 negative photoresist, etched single mode optical fiber, and PDMS. Before starting the process, the 4-inch wafer sputtered with a copper film of approximately 200 nm thickness, and the single mode optical fiber must be etched to 48 μm with buffer silicon oxide etching (BOE) solution at 40 °C; the sputtered copper film on the wafer is used as a sacrificial layer to make the sensor separate from the wafer, while the etching of the optical fiber is done under constant temperature because that makes it easier to obtain an accurate etching diameter. The production process started with using a spin coater to evenly spin-coat photoresist onto the surface of the wafer by making use of the centrifugal force generated by the rotation of the spin coater. Then the coated wafer was placed on a heating plate with a temperature of 95 °C to carry out the soft bake (SB) operation, the main purpose of which is to allow the vaporization of organic solvents in the photoresist. After the soft baking, a double side exposing machine was used to carry out the exposure operation. In this process, a 365 nm wavelength UV light was used, and the penetration and shadowing of light were controlled by a mask; the photoresist areas exposed to the UV light generated bonding and hardened. After completion of the exposure operation, the wafer was moved onto a hot baking plate to carry out the post exposure baking (PEB) operation in order to eliminate standing waves generated in the exposure operation, and then the hot baking plate was turned off to allow the temperature to slowly cool to room temperature. This operation enhanced the strength of the photoresist structure. Finally, after completion of all the aforementioned operations, the wafer was immersed in a developing solution and rotated by a spinner to remove areas not exposed to UV light. Upon completion of this operation, the designed bottom periodic structure was obtained.
Through the above steps, the patterned SU-8 3050 photoresist structure (37.5 μm) was first spin-coated onto the surface of the wafer, and then the optical fiber etched to 48 μm was pasted onto the patterned SU-8 structure. The steps were then repeated once again using the photoresist SU-8 3050 and the spin coater with a corresponding number of revolutions to form a structure of approximately 125 μm thickness to cover the etched optical fiber. Finally, the prepared PDMS was spin-coated by the spin coater onto the surface of the wafer to form a thickness of about 162.5 μm, and then the wafer was placed into a vacuum oven to cure at a temperature of 100 °C for one hour. With this curing, the processing of the wafer itself was completed.
The completed fiber grating on the wafer was then immersed in a ferric chloride solution for the releasing process, whereby the photoresist layer was separated from the wafer because the thin copper film sacrificial layer was etched away by the ferric chloride solution. Figure 1 presents a diagram of the process for fabricating the PLPFG strain sensor, and Fig. 2 presents a schematic diagram of the dimensions of each part of the PLPFG strain sensor.
4. Experimental setup
4.1 PLPFG tensile test
The main purpose of the tensile test was to measure the changes of resonant attenuation dip loss of the PLPFG. The PLPFG was loaded with axial stresses by the precision stage. The experimental setup is shown in Fig. 3. Experimental equipment included the broadband light source (superluminescent diode, SLD), optical spectrum analyzer (OSA), computer, load cells, and precision stage. Two ends of the PLPFG sensor were respectively connected to the broadband light source and the optical spectrum analyzer. Optical fibers at two ends of the sensor were respectively fixed on the load cell and the precision stage. The strain gauge used to verify the experimental results in this study was a coil strain gauge, its operating principle being to make use of any deformation of the strain gauge to cause changes in resistance which are then amplified and balanced by a Wheatstone bridge. Finally, a data acquisition system is used to acquire and convert the resulting data to strain values. As an external tensile strain is applied, the strain increases accordingly in different sections of the PLPFG. Thus, a periodic refractive index variance in the fiber core is obtained. As a result, the spectra of the PLPFG deform because of the strain, and the attenuation loss of the PLPFG can be changed with the strain in the sensing applications.
5. Results and discussion
5.1 Appearance of the PLPFG device
In this study, a novel PLPFG was fabricated via the LIGA-like MEMS process. The PLPFG strain sensor consisted of etched optical fiber sandwiched between two layers of thick photoresist SU-8 3050 and then packaged within PDMS polymer material. Figure 4 presents images of the etched optical fiber sandwiched by and embedded in the photoresist and PDMS polymer, respectively. Therefore, the image of the etched optical fiber in Fig. 4(a) is unfocused and unclear. Moreover, the packaging material (PDMS) was used to prevent the glue effect in the surface bonding process, that is, to provide effective protection and make the PLPFG robust. Figure 4(a) shows a photograph taken through an optical microscope (OM) of the PLPFG with a 30 mm gauge length, 670 μm period, and 72 μm diameter for the etched region. The width and thickness of the sandwiching photoresist were 1.5 mm and 150 μm, respectively. Figure 4(b) is an SEM image of the PLPFG. In Fig. 4 the periodic rectangular structure shown is the SU-8 photoresist grating.
5.2 PLPFG tensile character corresponds to different optical fiber diameters and grating periods
In this study, the spectra of PLPFG with different fiber diameters and periods under strain loading were investigated. We observed and compared the spectra of PLPFG with different fiber diameters under loading. The tensile test spectra of PLPFG with a period of 670 μm and diameters of 72 μm, 66 μm, 60 μm, and 54 μm, respectively, are shown in Fig. 5. The spectra of PLPFG deform under increasing loading. Before loading, there is a weak resonance dip in the transmission spectrum. The resonance dip of PLPFG intensifies with loading, and the wavelength of PLPFG slightly drifts toward shorter wavelengths during a tensile test. Moreover, a red shift is induced as the fiber diameter is reduced. The resonant wavelengths of PLPFG (all with a period of 670 μm, and with respective fiber diameters of 72 μm, 66 μm, 60 μm, and 54 μm) were 1553.5 nm, 1562.5 nm, 1568.5 nm, and 1577.5 nm, respectively. As the fiber diameter becomes smaller, the wavelength of PLPFG will drift toward longer wavelengths.
Fig. 5 Spectra of PLPFGs with a period of 670 μm and respective diameters of 72 μm, 66 μm, 60 μm, and 54 μm under various tensile loadings.
In order to compare the effects of the designed parameters (different fiber diameters and grating periods) on the PLPFG wavelength, strain calibration experiments were conducted on PLPFG with grating periods of 640~670 μm, and diameters of 72 μm, 66 μm, 60 μm, and 54 μm. The corresponding spectra are shown in Fig. 6. The tests showed that the wavelengths of PLPFG are larger when the size of the period increases and when the fiber diameter decreases. With loading, the resonant dip increases. The maximum resonant dip of the PLPFG (period: 670 μm, diameter: 72 μm) was −28.312 dB at a wavelengths of 1553.5 nm under loading, as shown in Fig. 6.
Fig. 6 Spectra of PLPFGs with periods of 640~670 μm and diameters of 72 μm, 66 μm, 60 μm, and 54 μm.
When PLPFGs with various periods are under loading, the transmission dip varies. Figure 7 shows the development of transmission dips by observing the spectra of the PLPFGs (670 μm) with different loading increments. The curves of the transmission/loading relationship are squared-harmonic functions. The dips of the PLPFGs signify growth up to the maximum transmission loss with loading. This can be explained using (2), which indicates that the transmission loss of the resonant dip has a cosine-squared dependence. Therefore, the relationship between tensile loading and transmission loss forms a quadratic polynomial curve. The diagram in Fig. 7 shows how the resonant attenuation dip of different PLPFGs varied under different strain loadings. The best sensitivity for the tested PLPFG was −0.0342 dB/με (with an R2 value of 0.963) for the PLPFG with a period of 670 μm and diameter of 54 μm. Moreover, with a decrease in the diameter of the optical fiber, the slope of the transmission loss-loading curve increases. In other words, the sensitivity of the PLPFG can be improved by reducing the diameter of the PLPFG. The diameter of the thin cladding fiber for PLPFG with high sensitivity (large slope) was 54 μm for achieving the tunable transmission dip loss with high loading sensitivity. In this way, a highly sensitive PLPFG strain sensor can be achieved.
Fig. 7 Resonant attenuation dip loss values of the PLPFGs (all with a period of 670 μm and with respective diameters of 72 μm, 66 μm, 60 μm, and 54 μm) changed with the various strain loadings.
Figure 8 shows the fiber diameter and wavelength analysis diagram of the different PLPFGs. As we reduce the fiber diameter of the optical fiber, the resonant wavelength linearly shifts towards the blue wavelength at an average rate of around 1.33 nm/μm, and the average linearity is 0.9755.
Fig. 8 Relation between the fiber diameter and wavelength of the PLPFGs with different periods (period 640~670 μm).
Figure 9 shows the grating period and wavelength correspondence analysis diagram of the PLPFGs with different diameters. From the figure it can be confirmed that in the LPFG wavelength formula (1), the wavelength will become smaller as the grating period increases. The ratios of the resonant wavelengths to periods are roughly constant, with values of −1.47, −1.50, −1.39, and −1.28 nm/μm, respectively. The average slope is −1.41 nm/μm. Moreover, the linearity of the fitting curves of the PLPFG with diameters of 72 μm, 66 μm, 60 μm, and 54 μm are 0.998, 0.990, 0.953, and 0.946, respectively. This phenomenon indicates that the resonant wavelengths are proportional to the periods. The average coefficient of determination (R-squared) was greater than 0.9717, suggesting high correlation. Therefore, we can obtain a designed resonant wavelength PLPFG by fine tuning the period and fiber diameter.
Fig. 9 Relation between the grating period and wavelength of the PLPFGs with various diameters (72 μm, 66 μm, 60 μm, and 54 μm).
5. Conclusion
A proposed PLPFG fabricated by the MEMS process with a PDMS packaging is reported in this paper. When PLPFG is subjected to strain loading, the transmission spectra are changed with resonance attenuation loss. PLPFG exhibit tunable resonance attenuation loss with strain loading. Through tensile testing and comparison with coil strain gauge calibration, it is confirmed that the PLPFG strain sensor researched and developed in this study can indeed be used as an effective sensor. In the experiment, the generated strain sensitivity of the PLPFG resonant attenuation dip loss reached −0.0342 dB/με, and the R2 value reached 0.963. The sensor’s sensing effect is thus even better than those of commercially available strain sensors. With the PDMS packaging material, the PLPFG is rendered more effective and robust for use as a strain sensor. The sensor can be adhered onto an object to be measured without affecting the measurement results, a feature which enhances the practicality of the PLPFG strain sensor.
Future research will be conducted to develop a multi-function sensor capable of measuring a variety of physical quantities based on packaged fiber grating. Moreover, the resonant dip loss of the proposed PLPFG can be tuned by the loadings for loss tunable filter applications.
Acknowledgments
This work is funded by the National Science Council, Taiwan (grant number NSC-100-2628-E-151-002-MY3).
References and links
1. K. J. Han, Y. W. Lee, J. Kwon, S. Roh, J. Jung, and B. Lee, “Simultaneous measurement of strain and temperature incorporating a long-period fiber grating inscribed on a polarization-maintaining fiber,” IEEE Photon. Technol. Lett. 16(9), 2114–2116 (2004). [CrossRef]
2. W. J. Bock, J. Chen, P. Mikulic, and T. Eftimov, “A Novel Fiber-Optic Tapered Long-Period Grating Sensor for Pressure Monitoring,” IEEE Trans. Instrum. Meas. 56(4), 1176–1180 (2007). [CrossRef]
3. D. E. Ceballos-Herrera, I. Torres-Gomez, A. Martinez-Rios, L. Garcia, and J. J. Sanchez-Mondragon, “Torsion Sensing Characteristics of Mechanically Induced Long-Period Holey Fiber Gratings,” IEEE Sens. J. 10(7), 1200–1205 (2010). [CrossRef]
4. R. I. Mata-Chávez, J. M. Estudillo-Ayala, R. Rojas-Laguna, E. Vargas-Rodriguez, I. Torres-Fomez, D. Monzon-Hernandez, M. Trejo-Duran, and J. R. Perez-Chimal, “Study of temperature sensing in a novel fattened electric arc induced LPFG,” Proc. SPIE 7430, 743014 (2009). [CrossRef]
5. H. J. Kim, O. J. Kown, S. B. Lee, and Y. G. Han, “Measurement of temperature and refractive index based on surface long-period gratings deposited onto a D-shaped photonic crystal fiber,” Appl. Phys. B 102(1), 81–85 (2011). [CrossRef]
6. I. K. Hwang, S. H. Yun, and B. Y. Kim, “Long-period fiber gratings based on periodic microbends,” Opt. Lett. 24(18), 1263–1265 (1999). [CrossRef] [PubMed]
7. M. Vaziri and C. L. Chen, “Etched fibers as strain gauges,” J. Lightwave Technol. 10(6), 836–841 (1992). [CrossRef]
8. C. Y. Lin, L. A. Wang, and G. W. Chern, “Corrugated long-period fiber gratings as strain, torsion, and bending sensors,” J. Lightwave Technol. 19(8), 1159–1168 (2001). [CrossRef]
9. Y. P. Wang, L. Xiao, D. N. Wang, and W. Jin, “Highly sensitive long-period fiber-grating strain sensor with low temperature sensitivity,” Opt. Lett. 31(23), 3414–3416 (2006). [CrossRef] [PubMed]
10. T. Zhu, Y. Song, Y. J. Rao, and Y. Zhu, “Highly Sensitive Optical Refractometer Based on Edge-Written Ultra-Long-Period Fiber Grating Formed by Periodic Grooves,” IEEE Sens. J. 9(6), 678–681 (2009). [CrossRef]
11. T. Zhu, Y. J. Rao, J. L. Wang, and Y. Song, “A Highly Sensitive Fiber-Optic Refractive Index Sensor Based on an Edge-Written Long-Period Fiber Grating,” IEEE Photon. Technol. Lett. 19(24), 1946–1948 (2007). [CrossRef]
12. C. L. Zhao, L. Xiao, J. Ju, M. S. Demokan, and W. Jin, “Strain and Temperature Characteristics of a Long-Period Grating Written in a Photonic Crystal Fiber and Its Application as a Temperature-Insensitive Strain Sensor,” J. Lightwave Technol. 26(2), 220–227 (2008). [CrossRef]
13. P. Caldas, G. Rego, J. L. Santos, and O. V. Ivanov, “Characterization of response of a dual resonance of an arc-induced long-period grating to various physical parameters,” in IEEE Sensors Conference (2008), pp. 329–332. [CrossRef]
14. Z. Xin, C. Shen, J. Chu, and C. Zhong, “Long-Period Fiber Grating Based Cantilever Strain Sensor,” in Symposium on Photonics and Optoelectronics (SOPO) (2012), pp. 1–3.
15. T. Mizunami, Y. Sho, K. Yamamoto, and Y. Ishida, “Long-period fiber-gratings produced by exposure with a low-pressure mercury lamp and their sensing characteristics,” Opt. Commun. 282(24), 4699–4705 (2009). [CrossRef]
16. L. Zhang, Y. Liu, L. Everall, J. A. R. Williams, and I. Bennion, “Design and realization of long-period grating devices in conventional and high birefringence fibers and their novel applications as fiber-optic load sensors,” IEEE J. Sel. Top. Quantum Electron. 5(5), 1373–1378 (1999). [CrossRef]
17. S. Savin, M. J. F. Digonnet, G. S. Kino, and H. J. Shaw, “Tunable mechanically induced long-period fiber gratings,” Opt. Lett. 25(10), 710–712 (2000). [CrossRef] [PubMed]
18. C. C. Chiang, T. C. Cheng, H. J. Chang, and L. Tsai, “Sandwiched long-period fiber grating filter based on periodic SU8-thick photoresist technique,” Opt. Lett. 34(23), 3677–3679 (2009). [CrossRef] [PubMed]
19. T. Erdogan, “Fiber grating spectra,” J. Lightwave Technol. 15(8), 1277–1294 (1997). [CrossRef]
