Open Access
Add to BibSonomy
Abstract
We present a hybrid Brillouin/Rayleigh sensor for multiparameter sensing in optical fibers. The system makes use of a single laser pulse to excite both Rayleigh and Brillouin backscattering in the same optical fiber. In the detection path, the backscattered signals are separated based on their different wavelengths. The system is capable of determining simultaneously the Brillouin frequency shift (BFS) of the fiber, as well as the frequency contents of any vibration acting on the same fiber as recovered by phase sensitive OTDR (ϕ-OTDR) measurements. The reported experiments show the possibility to perform simultaneous temperature and vibration measurements, as well as to perform dynamic strain measurements combining the information provided by slope-assisted Brillouin scattering measurements, with those provided by amplitude-based ϕ-OTDR measurements.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Distributed optical fiber sensors (DOFS) represent a very effective technology for the continuous monitoring of structures. Conventionally, DOFS make use of some form of light scattering inside a silica glass optical fiber [1]. Rayleigh scattering is caused by random fluctuations in the refractive index of the fiber and can be used for both static (strain and temperature) and dynamic (acoustic/vibration) sensing. Static sensing techniques based on Rayleigh scattering usually involve the acquisition of the backscattered light for a range of laser wavelengths. These methods are not always suited for monitoring applications, either because the temperature/strain range is too narrow [2], or because the measurement range is too short [3]. Rayleigh scattering can be exploited to perform vibration (dynamic) sensing as well. In such a case, the amplitude (or the phase) of the backscattered light excited by a coherent laser pulse is acquired at a fixed wavelength. Mechanical vibrations or acoustic perturbances are usually identified by comparing consecutive phase sensitive OTDR (ϕ-OTDR) measurements, as they induce variations in the amplitude and phase of the backscattered light at the corresponding position. A different mechanism of light scattering is the Brillouin scattering, which involves the interaction between optical and acoustic fields in the optical fiber: any changes in the tensional state or temperature of the fiber alter its elastic properties, and in particular, the frequency of the acoustic wave generated either thermally (in case of spontaneous scattering) or by electrostriction (in case of stimulated scattering). This frequency coincides with the Doppler shift experienced by the incident light. Thus, the strain or the temperature of the fiber is recovered by measuring the frequency shift between the incident and scattered light, which is known as the Brillouin Frequency shift (BFS). Brillouin methods are capable of high spatial resolution (down to the mm range) [4,5], and very long sensing lengths (up to one hundred km and more) [6]. However, they cannot capture weak vibrations with nɛ-level strain, mainly because of the relatively poor sensitivity of the BFS to strain (∼ 50 kHz/µɛ).
Hybrid sensors, exploiting (at least) two scattering mechanisms for multiparameter sensing, may provide more valuable information and a comprehensive identification of fault events. For example, in the context of pipeline leakage detection, both temperature changes and vibration can detect the leak event [7], while using the two effects simultaneously may improve the detection rate and reduce false alarms. In the geotechnical field, both static ground deformation and vibration can be used as precursors of slope failure events [8,9]. A few works have been reported, exploiting Rayleigh and Brillouin scattering simultaneously on the same fiber. For example, Ref. [10] exploits the two scattering mechanisms to separate strain effects from temperature. In that case, the Rayleigh scattering was used for static measurements in the frequency domain, while vibration sensing was not demonstrated. In Ref. [11], combined temperature and vibration sensing has been reported, using ϕ-OTDR for vibration measurements and Brillouin-OTDR for temperature sensing. Notably, this method has the advantage of requiring access to a single fiber end for both vibration and static measurements. However, the setup is quite complicated as it adopts a specially synthetized optical pulse sequence to avoid the onset of stimulated Brillouin scattering. We finally note that, a temperature/vibration sensor based on the combined use of Rayleigh and Raman scattering has been demonstrated in Ref. [12]; however, in that case the weak efficiency of Raman scattering in single-mode fibers limits the accuracy (and/or the measurement time) of temperature measurements.
In this paper, we demonstrate an integrated system for the simultaneous measurement of temperature, strain, and vibration along the same fiber. The system is only slightly different from a conventional Brillouin Optical Time-Domain Analysis (BOTDA) setup; therefore, it can be used to add vibration sensing capabilities to almost any BOTDA system.
2. Experimental setup
The experimental setup employed for hybrid Brillouin/Rayleigh scattering measurements is shown in Fig. 1. The scheme is very similar to a common BOTDA setup: The laser light generated by an external cavity laser (ECL) with 10-kHz linewidth, 1550-nm wavelength, and an optical power of 18 dBm, is split by a 50/50 optical coupler. In the lower branch the light is pulsed through an electro-optic modulator (EOM), then amplified, polarization-switched, and finally sent to one end of the fiber under test (FUT). In the upper branch, the light is sent to another EOM for suppressed-carrier, dual sideband modulation. The two sidebands are shifted from the carrier by a frequency close to the BFS (∼ 10.7 GHz) of the FUT. The optical switch (OS) in Fig. 1 is usually in a close state (ON position), therefore the probe light is injected into the other end of the FUT. Therefore, the lower frequency (Stokes) sideband interacts with the pulse along the FUT, creating an intense acoustic wave which backscatters the pulse itself. The backscattered light is sent to a narrowband (∼ 5 GHz) fiber Bragg grating (FBG), reflecting the sole Stokes component, while letting the Rayleigh scattering to pass through. The reflected Stokes light is pre-amplified up to -3 dBm, before being detected by an optical photodetector and digitized. Differently from a conventional BOTDA setup, the light at the transmission port of the FBG is detected too. The latter includes the light backscattered by the FUT due to Rayleigh scattering, as well as the anti-Stokes component of the probe light. In order to remove the anti-Stokes component, an optical circulator and a tunable FBG are employed, with the FBG tuned in order to reflect the Rayleigh scattering only. The two photodetectors shown in Fig. 1 are identical and characterized by a conversion gain of 1300 V/W. Data acquisition is performed at a sampling frequency of 250 MS/s (one sampled point every 40 cm along the FUT). We observe that, while the two FBGs in Fig. 1 guarantee that the modulation sidebands impressed into the probe light will not reach PD2, the residual optical carrier leaking through the EOM will be detected by PD2 together with the Rayleigh scattering from the FUT, adding noise in the ϕ-OTDR measurements. As we will see later, this noise can be eliminated by setting the optical switch in its open state (OFF position). However, in that case the measurement of Rayleigh and Brillouin scattering are not carried out simultaneously.
Fig. 1. Experimental setup. EDFA: Erbium-Doped Fiber Amplifier. PS: polarization switch; EOM: electro-optic modulator; FBG: fiber Bragg grating; FUT: fiber under test; OS: optical switch; PD: photodetector.
As in any conventional BOTDA sensor, the (static) temperature or strain conditions of the FUT are retrieved from the BFS profile, which in turn is determined by scanning the radiofrequency applied to the EOM in the probe branch. The experiments reported here have been carried out using a linearly frequency-swept probe light, in order to get rid of the settling time of the microwave source [13]. In brief, the method consists in the application of a frequency modulated continuous wave (FMCW) signal to the EOM in the probe branch. Differently from the optical chirp chain method reported in Ref. [14], the sweep rate is so low, that the frequency of the probe light is basically constant during its interaction with the pulse pump; therefore, the resulting signal is very similar to a conventional BOTDA working with a stepped probe frequency. In our test, the RF was swept from 10,650 MHz to 10,850 MHz in 1 s, while acquiring the Brillouin traces with an averaging factor of 1024 and a pulse repetition rate of 100 kHz. Each measurement required the acquisition of the Brillouin traces for two orthogonal states of polarization of the pump light, i.e., for the two states of the polarization switch shown in Fig. 1; thus, the RF frequency was swept twice for each measurement, resulting in an overall acquisition time of 2 s. The pulse width was set to 20 ns, corresponding to a spatial resolution of 2 m. A 200-m spool of single-mode fiber, AcoustiSens by OFS, has been employed for the tests. In this specialty fiber, a ultraweak Bragg grating is UV-inscribed during the draw process, in order to increase the Rayleigh backscatter by ∼ 13 dB at the working wavelength, while maintaining a low attenuation (< 0.7 dB/km at 1550-nm) [15]. As discussed in Ref. [16], an increase of the Rayleigh backscattering by a certain factor, leads to an identical increase in the mean dynamic SNR. Before the tests, the fiber was calibrated in terms of BFS dependence from temperature and strain, revealing a sensitivity of 0.99 MHz/°C and 47 kHz/µɛ, respectively. A length of ∼5m of the FUT was wound around a pipe. During the experiment, the pipe (and therefore the fiber wound on it) was heated using a space heater. Simultaneously, the same piece of fiber was put in vibration by means of a magnetostrictive actuator bolted to the pipe and driven by a sine wave current at 150 Hz. We report in Fig. 2(a) the acquired BFS profiles, as determined by processing the acquired BOTDA data by conventional fitting. The inset shows a zoomed view in the heated zone, which permits to appreciate the 2-m spatial resolution of the BOTDA measurements. Figure 2(b) shows the temperature over time at the heated position. We note that the heater was switched off after 400 s, and that the temperature follows the expected behavior. The estimated temperature resolution is 0.1°C, which should be compared to the resolution of 0.5°C (at a spatial resolution of 5m and with an acquisition time of 1 minute) reported in Ref. [12].
Fig. 2. (a) BFS profiles and (b) corresponding temperature evolution over the heated piece of fiber (z≈192m). In (a), the inset shows a zoomed view of the BFS profiles around the heated zone.
In Fig. 3(a) we show the spectrogram of the ϕ-OTDR signal acquired at the same position, computed using a window size of 2.5 s, an overlap of 50%, and a number of averages equal to 256 (resulting in an acquisition rate of ∼391 Hz). While the 150-Hz vibration is clearly seen in the reported spectrogram, we also note that some noise is visible in the measurement. As discussed earlier, part of this noise should be attributed to the residual probe light at the laser frequency, leaking from the EOM and reaching PD2 together with the Rayleigh scattering. Note that the EOMs employed in our setup had a nominal extinction ratio of 25 dB. In order to analyze the role played by the EOM leakage, we have performed a new ϕ-OTDR measurement with the optical switch set in its open position (OFF state). In this case, no probe light can reach PD2 and therefore the measurement is free from any interference due to EOM leakage. We show in Fig. 3(b) the results of this ϕ-OTDR measurement. Apparently, the vibration is better resolved in this case. In fact, by calculating the signal-to-noise ratio (SNR) as the ratio between the 150-Hz component and the average of the fast-Fourier Transform (FFT) magnitude at the vibrating position, we estimate an SNR of 14.6 dB for the measurement shown in Fig. 3(a), and 19.8 dB for that shown in Fig. 3(b). While removing the probe light is therefore beneficial for ϕ-OTDR (vibration) measurements, Rayleigh and Brillouin measurements cannot be performed simultaneously in this case. On the other hand, a high-extinction ratio EOM could be used in the probe generation branch, in order to reduce the impact of the residual carrier on the SNR of the ϕ-OTDR measurements.
Another experiment was carried out by attaching a piece of the fiber on a 50 cm × 30 cm × 0.3 cm ($L \times W \times H$) aluminum plate. The fiber was glued four times across the plate, in order to reach a 2-m length of bonded fiber. The plate was fixed at its four corners, and excited at its first resonance frequency (50 Hz) by use of a magnetic shaker (TIRAvib 50009) attached to the center of the plate. For this test, the BOTDA setup was operated in the slope-assisted configuration, i.e., with the probe frequency fixed on the rising slope of the Brillouin Gain Spectrum (BGS) [17,18]. A preliminary acquisition of the BGS allowed us to derive a transduction coefficient for the slope-assisted measurements of $1.65\,mV/MHz$ at $f_{RF} = 10720MHz$. The measurements were carried out by acquiring, for one minute, the Brillouin and Rayleigh scattering signals simultaneously (the optical switch was permanently set in its close state). The number of averages was set to 512 in both cases, resulting in an acquisition rate of ∼ 195 Hz. For BOTDA measurements, the first half of each acquisition set (256 averages) was taken for one state-of-polarization of the pump, while the other 256 averages were taken for the orthogonal state. We show in Fig. 4 the results of the measurements expressed in terms of strain amplitude in case of Brillouin scattering, or voltage amplitude in case of Rayleigh scattering. In both cases, the signal amplitude was evaluated by taking the FFT of the acquired trace at the vibrating position, and by retaining the component at the excitation frequency (50 Hz). Data are reported for different excitation energies, varied by acting on the amplitude of the voltage signal applied to the TIRAvib controller, from 100 mVpk-pk to 1.4 Vpk-pk. The same plots also report the SNR of the measurements, evaluated as the ratio between the 50-Hz component of the acquired signal, and the average of the magnitude spectrum at the vibrating position.
Fig. 4. Vibration amplitude of the plate and corresponding SNR as a function of the excitation voltage for (a) slope-assisted BOTDA measurements or (b) ϕ-OTDR measurements.
As expected, the strain amplitude detected by Brillouin measurements gradually increases with the excitation voltage, from 0.94 µɛ to 15.42 µɛ. For excitation voltages higher than 0.6V this increase appears linear as expected, while for weaker excitations the response deviates from linearity, probably due to the influence of noise. The SNR also increases with the excitation voltage. On the contrary, the vibration amplitude detected by ϕ-OTDR (as well as the SNR) varies non-monotonically, as a consequence of the random relation between strain and intensity in Rayleigh backscattering measurements [18]. We also note that the SNR of the ϕ-OTDR measurements is generally better than BOTDA measurements, especially for lower strain amplitudes. This is still more evident if we consider not only the vibrating position, but the whole fiber. As an example, we compare in Fig. 5 the signals acquired over the entire fiber length at the maximum excitation energy (1.4 Vpk-pk). We see that the ϕ-OTDR measurement is much cleaner than the BOTDA measurement, despite the slightly lower SNR at the vibrating position (SNR = 17.3 dB for BOTDA, SNR = 16.7 dB for ϕ-OTDR).
Fig. 5. Vibration amplitude of the plate for an excitation voltage of 1.4 Vpk-pk, as detected by (a) slope-assisted BOTDA measurements or (b) ϕ-OTDR measurements.
Thus, while non quantitative, the Rayleigh scattering measurement detects the vibration more clearly. We believe that this is a consequence of the much higher strain sensitivity of the ϕ-OTDR signal, compared to the SA-BOTDA signal. For example, for a typical Brillouin gain spectrum bandwidth of 35 MHz, the strain sensitivity of the Brillouin gain is ∼$0.25\% /{\mathrm{\mu}}\varepsilon $. On the contrary, the strain sensitivity of the ϕ-OTDR signal can be as high as $10\% /n\varepsilon $ [1]. This evidence suggests the possibility of combining the information provided by the Rayleigh and Brillouin scattering measurements. In practice, one may use the Brillouin measurement in order to estimate the amount of applied strain (which cannot be retrieved from the amplitude of the ϕ-OTDR measurement [1]) while the Rayleigh measurement can be used to determine the vibration spectrum along the entire fiber with more fidelity. Of course, the “fusion” of the two measurements is possible as long as the vibration is large enough to be detected by the SA-BOTDA measurement. For nɛ-level strain amplitudes, the modulation induced by the vibration on the SA-BOTDA signal may be so weak, that it may be not possible to isolate the signal from noise. In those cases, schemes based on the measurement of the phase of the Rayleigh scattering should be employed for proper quantification of the applied strain [19,20]. Furthermore, the errors induced by power variations should be also considered when adopting the SA-BOTDA scheme for quantitative strain measurements: When large power variations are expected, more sophisticated schemes such as the double slope-assisted BOTDA method should be adopted [21].
3. Conclusions
A multiparameter scheme for BOTDA and ϕ-OTDR measurements has been demonstrated, which is only slightly more complex than a conventional BOTDA scheme. The combined use of Brillouin and Rayleigh scattering, as excited by the same laser pulse over the same fiber, permits to recover temperature/strain and vibration simultaneously. We have also shown that one can combine slope-assisted BOTDA measurements and ϕ-OTDR measurements, in order to get a more valuable information than that provided by a single measurement.
Funding
POR FESR CAMPANIA 2014/2020 (B43D1800029007).
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. A. H. Hartog, An Introduction to Distributed Optical Fibre Sensors, Fiber Optic Sensors (Taylor & Francis Group, 2018).
2. Y. Koyamada, M. Imahama, K. Kubota, and K. Hogari, “Fiber-Optic Distributed Strain and Temperature Sensing With Very High Measurand Resolution Over Long Range Using Coherent OTDR,” J. Lightwave Technol. 27(9), 1142–1146 (2009). [CrossRef]
3. B. J. Soller, D. K. Gifford, M. S. Wolfe, and M. E. Froggatt, “High resolution optical frequency domain reflectometry for characterization of components and assemblies,” Opt. Express 13(2), 666–674 (2005). [CrossRef]
4. R. Bernini and A. Minardo, L. Zeni, “Distributed sensing at cm-scale spatial resolution by BOFDA: measurements and signal processing,” IEEE Photonics J. 4(1), 48–56 (2012). [CrossRef]
5. A. Denisov, M. A. Soto, and L. Thévenaz, “Going beyond 1000000 resolved points in a Brillouin distributed fiber sensor: theoretical analysis and experimental demonstration,” Light Sci. Appl. 5(5), e16074 (2016). [CrossRef]
6. M. A. Soto, G. Bolognini, and F. Di Pasquale, “Long-range simplex-coded BOTDA sensor over 120km distance employing optical preamplification,” Opt. Lett. 36(2), 232–234 (2011). [CrossRef]
7. N. F. Adnan, M. F. Ghazali, M. M. Amin, and M. M. A. Hamat, “Leak detection in gas pipeline by acoustic and signal processing - A review,” IOP Conf. Ser.: Mater. Sci. Eng.1000120132015. [CrossRef]
8. L. Olivares, E. Damiano, R. Greco, L. Zeni, L. Picarelli, A. Minardo, A. Guida, and R. Bernini, “An Instrumented Flume to Investigate the Mechanics of Rainfall-Induced Landslides in Unsaturated Granular Soils,” Geotech. Test. J. 32(2), 108–118 (2009). [CrossRef]
9. T. Shiotani and M. Ohtsu, “Prediction of Slope Failure Based on AE Activity,” in Acoustic Emission: Standards and Technology Update, ed. S. Vahaviolos, (ASTM International, West Conshohocken, PA, 1999), 156–156-17 [CrossRef]
10. D.P. Zhou, W. Li, L. Chen, and X. Bao, “Distributed Temperature and Strain Discrimination with Stimulated Brillouin Scattering and Rayleigh Backscatter in an Optical Fiber,” Sensors 13(2), 1836–1845 (2013). [CrossRef]
11. J. Zhang, T. Zhu, H. Zhou, S. Huang, M. Liu, and W. Huang, “High spatial resolution distributed fiber System for multi-parameter sensing based on modulated pulses,” Opt. Express 24(24), 27482–27493 (2016). [CrossRef]
12. Y. Muanenda, C. J. Oton, S. Faralli, T. Nannipieri, A. Signorini, and F. Di Pasquale, “Hybrid distributed acoustic and temperature sensor using a commercial off-the-shelf DFB laser and direct detection,” Opt. Lett. 41(3), 587–590 (2016). [CrossRef]
13. A. Minardo, E. Catalano, and L. Zeni, “Cost-effective method for fast Brillouin optical time-domain analysis,” Opt. Express 24(22), 25424–25431 (2016). [CrossRef]
14. D. Zhou, Y. Dong, B. Wang, C. Pang, D. Ba, H. Zhang, Z. Lu, H. Li, and X. Bao, “Single-shot BOTDA based on an optical chirp chain probe wave for distributed ultrafast measurement,” Light Sci. Appl. 7, 32 (2018). [CrossRef]
15. P. S. Westbrook, K. S. Feder, T. Kremp, E. M. Monberg, H. Wu, B. Zhu, L. Huang, D. A. Simoff, S. Shenk, V. A. Handerek, M. Karimi, A. Nkansah, and A. Yau, “Enhanced Optical Fiber for Distributed Acoustic Sensing beyond the Limits of Rayleigh Backscattering,” iScience 23(6), 101137 (2020). [CrossRef]
16. Gabai and Eyal, “On the sensitivity of distributed acoustic sensing,” Opt. Lett. 41(24), 5648–5651 (2016). [CrossRef]
17. R. Bernini, A. Minardo, and L. Zeni, “Dynamic strain measurement in optical fibers by stimulated Brillouin scattering,” Opt. Lett. 34(17), 2613–2615 (2009). [CrossRef]
18. A. Minardo, A. Coscetta, S. Pirozzi, R. Bernini, and L. Zeni, “Experimental modal analysis of an aluminum rectangular plate by use the slope-assisted BOTDA method,” Smart Mater. Struct. 22(12), 125035 (2013). [CrossRef]
19. A. Masoudi, M. Belal, and T. Newson, “A distributed optical fibre dynamic strain sensor based on phase-OTDR,” Meas. Sci. Technol. 24(8), 085204 (2013). [CrossRef]
20. M. Sagues, E. Piñeiro, E. Cerri, A. Minardo, A. Eyal, and A. Loayssa, “Two-wavelength phase-sensitive OTDR sensor using perfect periodic correlation codes for measurement range enhancement, noise reduction and fading compensation,” Opt. Express 29(4), 6021–6035 (2021). [CrossRef]
21. A. Motil, O. Danon, Y. Peled, and M. Tur, “Pump-Power-Independent Double Slope-Assisted Distributed and Fast Brillouin Fiber-Optic Sensor,” IEEE Photon. Technol. Lett. 26(8), 797–800 (2014). [CrossRef]
