Ultrasonic imaging of seismic physical models using a phase-shifted fiber Bragg grating Jingjing Guo,1 Shigui Xue,2 Qun Zhao,2 and Changxi Yang1,* 1
State Key Laboratory of Precision Measurement Technology and Instruments, Department of Precision Instruments, Tsinghua University, Beijing 100084, China 2 Sinopec Geophysical Research Institute, Nanjing 211103, China * [email protected]
Abstract: We report what is to our knowledge the first ultrasonic imaging of seismic physical models by using a phase-shifted fiber Bragg grating (PS-FBG). Seismic models, which consist of multiple layer structures, were immersed in water. Piezoelectric (PZT) transducer was used to generate ultrasonic waves and a PS-FBG as a receiver. Two-dimensional (2D) ultrasonic images were reconstructed by scanning the PS-FBG with a highprecision position scanning device. In order to suppress the low-frequency drift of the Bragg wavelength during scanning, a tight wavelength tracking method was employed to lock the laser to the PS-FBG resonance in its reflection bandgap. The ultrasonic images captured by the PS-FBG have been compared with the images obtained by the geophysical imaging system, Sinopec, demonstrating the feasibility of our PS-FBG based imaging system in seismic modeling studies. ©2014 Optical Society of America OCIS codes: (060.2370) Fiber optics sensors; (060.3735) Fiber Bragg gratings; (110.2970) Image detection systems.
References and links 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.
M. L. Buddensiek, C. M. Krawczyk, N. Kukowski, and O. Oncken, “Performance of piezoelectric transducers in terms of amplitude and waveform,” Geophysics 74(2), T33–T45 (2009). M. Urosevic, G. Bhat, and M. H. Grochau, “Targeting nickel sulfide deposits from 3D seismic reflection data at Kambalda, Australia,” Geophysics 77(5), WC123–WC132 (2012). B. Evans, J. McDonald, and W. French, “Seismic physical modelling of reservoirs, its past, present and future,” ASEG Extended Abstracts 2007: 19th Geophysical Conference: pp. 1–6. J. K. Coper, D. C. Lawton, and G. F. Margrave, “The wedge model revisited: A physical modeling experiment,” Geophysics 75(2), T15–T21 (2010). G. L. Fradelizio, A. Levander, and C. A. Zelt, “Three-dimensional seismic-reflection imaging of a shallow buried paleochannel,” Geophysics 73(5), B85–B98 (2008). D. Sherlock, J. McKenna, and B. Evans, “Time-lapse 3-D seismic physical modelling,” Explor. Geophys. 31(2), 310–314 (2000). A. Wandler, B. Evans, and C. Link, “AVO as a fluid indicator: A physical modeling study,” Geophysics 72(1), C9–C17 (2007). F. Mahmoudian, G. F. Margrave, P. F. Daley, J. Wong, and E. Gallant, “Determining elastic constants of an orthorhombic material by physical seismic modeling,” Crewes Research Report 21 (2010). R. R. Stewart, N. Dyaur, B. Omoboya, J. J. S. de Figueiredo, M. Willis, and S. Sil, “Physical modeling of anisotropic domains: Ultrasonic imaging of laser-etched fractures in glass,” Geophysics 78(1), D11–D19 (2013). H. Lamela, D. Gallego, and A. Oraevsky, “Optoacoustic imaging using fiber-optic interferometric sensors,” Opt. Lett. 34(23), 3695–3697 (2009). A. Rosenthal, D. Razansky, and V. Ntziachristos, “High-sensitivity compact ultrasonic detector based on a piphase-shifted fiber Bragg grating,” Opt. Lett. 36(10), 1833–1835 (2011). Q. Wu and Y. Okabe, “High-sensitivity ultrasonic phase-shifted fiber Bragg grating balanced sensing system,” Opt. Express 20(27), 28353–28362 (2012). N. Takahashi, A. Hirose, and S. Takahashi, “Underwater Acoustic Sensor with Fiber Bragg Grating,” Opt. Rev. 4(6), 691–694 (1997). P. Fomitchov, “Response of a fiber Bragg grating ultrasonic sensor,” Opt. Eng. 42(4), 956–963 (2003). I. M. Perez, H. Cui, and E. Udd, “Acoustic emission detection using fiber Bragg gratings,” Proc. SPIE 4328, 209–215 (2001). H. Tsuda, E. Sato, T. Nakajima, H. Nakamura, T. Arakawa, H. Shiono, M. Minato, H. Kurabayashi, and A. Sato, “Acoustic emission measurement using a strain-insensitive fiber Bragg grating sensor under varying load conditions,” Opt. Lett. 34(19), 2942–2944 (2009).
#214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19573
17. A. Minardo, A. Cusano, R. Bernini, L. Zeni, and M. Giordano, “Response of fiber Bragg gratings to longitudinal ultrasonic waves,” IEEE Trans. Ultrason. Ferroelectr. Freq. Control 52(2), 304–312 (2005). 18. T. Liu and M. Han, “Analysis of π-Phase-Shifted Fiber Bragg Gratings for Ultrasonic Detection,” IEEE Sens. J. 12(7), 2368–2373 (2012). 19. Q. Wu, Y. Okabe, and J. Sun, “Investigation of dynamic properties of erbium fiber laser for ultrasonic sensing,” Opt. Express 22(7), 8405–8419 (2014). 20. M. A. Breazeale and J. Ford, “Ultrasonic Studies of the Nonlinear Behavior of Solids,” J. Appl. Phys. 36(11), 3486–3490 (1965). 21. D. M. Egle and D. E. Bray, “Measurement of acoustoelastic and third-order elastic constants for rail steel,” J. Acoust. Soc. Am. 60(3), 741–744 (1976).
1. Introduction Since the 1920s, seismic physical model, as a scaled analog to geological structures in reservoirs, has been a successful tool for research in seismic wave propagation and wave theoretical predictions [1, 2]. Experiments in labs are much less expensive as well as more repeatable, and controllable than in the field. A seismic modeling system was constructed employing a meter square water tank, where scaled models of geologic structures were suspended [3, 4]. 2D or 3D surveys were conducted by moving ultrasonic source and receiver over the models [5, 6]. Wandler et al. constructed physical models to distinguish fluid types from mixtures of water, brine, oil and carbon dioxide by characterizing the seismic reflection [7]. Mahmoudian et al. utilized a physical model to experimentally determine the elastic constants of an orthorhombic material [8]. Stewart et al. explored the anisotropic cracked or fractured zones by imaging laser-etched models, which was intended to assist in assessment of fracture zones in actual reservoirs [9]. Seismic physical models provide an effective way to bridge the theory and field-scale experiments and enable us to measure changes of the acoustic response in absence of a rock matrix in a nearly ideal setting. In physical modeling studies, piezoelectric (PZT) transducers usually serve as source and receiver, which are highly sensitive but have narrow bandwidth owing to their strongly resonant effects. The detectors based on thin piezoelectric polymer films, can achieve a broadband response using appropriate matching materials [10]. However, their sensitivity decreases with decreasing size and the corresponding electrical capacitance. In addition, PZT sensors, due to their electrical nature, are very sensitive to ambient electromagnetic disturbances [11, 12]. Optical means based on FBGs for acoustic detection has been studied as an alternative to PZT sensors by many researchers [13, 14]. Perez et al. utilized a broadband amplified spontaneous emission (ASE) light source and matched FBGs to detect acoustic emission (AE) signals [15]. To avoid the high intensity noise induced by ASE source, Tsuda et al. adjusted a narrow linewidth laser to the 3 dB position of the peak in the reflection spectrum of a normal FBG to achieve a high signal-to-noise ratio (SNR) [16]. According to the analyses in [17] and [18], the grating length and the spectrum slope determine the frequency response and intensity sensitivity, respectively. Consequently, FBGs with short grating and sharp slope are needed for high-frequency weak signals. A special type of FBGs with a π-phase shift in its center has aroused extensive attentions for ultrasonic detection. The π-phase shift in center of the grating decreases the effective length of sensor and exhibits an extremely sharp resonance for high sensitivity. Rosenthal et al. achieved a 10MHz bandwidth of ultrasonic detection by tuning the wavelength of a continuous-wave (CW) laser to the maximum slope of a pi-phase-shifted FBG and an inverse relation between sensitivity and effective sensing length was obtained [11]. To further improve the PS-FBG based ultrasonic sensor for weak AE signals, Wu et al. demonstrated a PS-FBG balanced system to effectively suppress the laser intensity noise and common mode noise, and a high sensitivity of 9 nε/Hz1/2 was achieved [12]. Though PS-FBG exhibits great potential for ultrasonic sensing, the environmental perturbations caused by temperature fluctuations and large quasi-static strains, easily shift the Bragg wavelength of PS-FBG out of its dynamic range. Against this problem, Wu et al. built an erbium fiber laser with an inbuilt PS-FBG for ultrasonic sensing [19]. When temperature or quasi-static strain shifts the wavelength of PSFBG, the lasing longitudinal mode will hop to a neighboring position, giving a solution to the instability of PS-FBG based sensing system. Unfortunately, due to the uncontrollable relative
#214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19574
positions of the lasing and PS-FBG wavelength, 22 dB instability of the detected signal was found in their measurements. In this study, we demonstrate the first ultrasonic imaging system based on a PS-FBG probe for seismic modeling studies. The imaging system is robust to environmental perturbations by locking the laser to the PS-FBG resonance wavelength. The PS-FBG probe, capable of mobility and ultrasonic detection was mounted on a high-precision position scanning device. By scanning the sensor probe, Two-dimensional (2D) images can be reconstructed according to time-lapse changes in ultrasonic reflection. 2. Sensor design and system setup To realize ultrasonic imaging, continuous scanning measurement of ultrasonic reflections in tested models is needed. To give mobility and reliability to the PS-FBG sensor, we contrived a sensor probe in Fig. 1. The PS-FBG with grating length of 5 mm was manufactured by the Fujikura Company. The reflectivity of the grating was designed to above 90%. The fiber grating, recoated with polyimide coating, is glued on a steel pole of 0.15 cm diameter and 5 cm length with an epoxy resin adhesive. A certain thickness of epoxy resin adhesive forms an outer sheath layer to protect the PS-FBG sensor. The Ultrasonic wave, conducted through the steel pole and epoxy resin layer, can be effectively coupled to sensing area of the PS-FBG.
Fig. 1. Structure of the mobile PS-FBG probe. Ultrasonic wave is coupled to PS-FBG through the steel pole and the epoxy resin coverage layer.
Fig. 2. Close-up of PS-FBG reflection dip measured by sweeping the TLS. The reflection bandwidth is 0.012 nm and the linear slope is 86 nm−1. Inset: wavelength range from 1547 nm to 1551.5 nm by linear range.
Figure 2 shows the reflection spectrum of PS-FBG used in the sensing system measured by sweeping a tunable laser source (TLS) from 1547 nm to 1551.5 nm. With a π-phase shift in center of the grating, an extremely narrow spectral notch is formed for highly sensitive
#214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19575
ultrasonic detection. The Bandwidth of PS-FBG is 0.012 nm and the linear slope is 86 nm−1, which is about 172 times over a normal FBG of 0.5 nm−1 (typical). Figure 3 shows the schematic diagram of the proposed imaging system. A tunable laser source (Santec, 710) with 100 kHz linewidth and 0.1 pm tunable resolution is used to interrogate the fiber sensor. An optical circulator guides the light reflected from the grating to a photoelectric detector (New Focus, 2053) with a bandwidth of 10 MHz at 0 dB gain. A signal conditioning unit is also designed for pre-amplification and denoising. The demodulated signals are eventually collected by a DAQ (Data Acquisition) card and captured by computer for image processing. Because ultrasound is not effectively transmitted through air, Plexiglas models are placed at the bottom of a 2 × 1.5 × 1 m3 (l × w × h) water tank. The PS-FBG and PZT actuator are just located on the surface of water, which is above the physical model around 7.5 cm. A pulse generator (PG) is used to drive the PZT actuator, which generates pulse ultrasonic signals with duration of 2.5 μs and is kept at a fixed position while measuring. The location arm with PS-FBG probe is programmed to move along the water surface with a spatial resolution of 0.01 cm, during which both of the ultrasonic waves and the probe’s coordinates should be real-time recorded.
Wavelength tracking
Single-mode fiber
TLS
Circulator 1 2
PG Location control arm
3 Servo
PD
Signal conditioning X Y DAQ
PS-FBG probe
Water
PZT actuator
Computer Model under test
Fig. 3. Schematic diagram of the proposed ultrasonic imaging system. The model is put into a water tank. The PZT actuator and PS-FBG are just located on the water surface, which is above the physical model 7.5 cm. TLS: tunable laser source, PD: photodiode, DAQ: Data Acquisition, PG: Pulse Generator. The PZT actuator and PS-FBG were controlled by a position scanning device with a spatial resolution of 0.01 cm.
3. Experimental results and discussion 3.1 System stability In the measurement, the wavelength of TLS was tuned to the linear range of the grating. The basic principle of this demodulation technique is based on edge filter detection. Ultrasonic waves can be directly detected by monitoring the reflected power of PS-FBG and changes in output power can be expressed as: ΔP = Δλ GP
(1)
Δλ represents the wavelength shift of PS-FBG, G is the grating slope at the laser wavelength, and P is the input power. It can be seen that the intensity sensitivity depends on the grating slope. However, due to environmental perturbations caused by temperature fluctuations and water sloshing during measurements, we found the slope rate can change significantly with time judging from the reflected power. To solve this problem, a wavelength tracking method
#214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19576
was employed to lock the laser to the 3 dB position on linear slope of the grating resonance (marked by red line in Fig. 2). We choose the 3 dB point as the locking position as it can ensure a large dynamic range in the bidirectional perturbations. Within the linear range, the demodulated signal, which could be used as an error signal of wavelength detuning between the laser and PS-FBG was fed back for fine tuning of the laser wavelength through an integral and proportional servo shown in Fig. 3. A typical control loop bandwidth of 1 kHz was used, which was fast enough to efficiently avoid the low-frequency wavelength drifts of PS-FBG, and enabled a tight laser wavelength tracking even when the sensor was subject to strong environmental perturbations. To illustrate this, we recorded the reflected power of PS-FBG in observation for about 30 minutes without ultrasound induced. Figure 4 shows a comparison of normalized reflected intensities by PS-FBG over time with/without wavelength tracking. At the beginning, the laser wavelength was tuned to about 3 dB position of the PS-FBG resonance at a slope rate of about 86 nm−1. In the case of PS-FBG without wavelength tracking, the Bragg wavelength slowly fluctuated back and forth as time went on, and the laser was eventually drifted out of the resonance wavelength range. Contrastively, in the case of wavelength tracking, the laser was finely locked to the half maximum of grating spectrum by the feedback servo, which ensured the practicability and long-term stability of the imaging system.
Fig. 4. Normalized intensity changing over time in observation for 30 minutes without ultrasound induced. Red curve: PS-FBG without wavelength tracking. Blue curve: with wavelength tracking. In the case of wavelength tracking, the half maximum point of the PSFBG resonance was chosen as the locking position to ensure a large dynamic range in the bidirectional perturbations.
3.2 Characterization of PS-FBG probe A continuous sinusoidal wave of 200 kHz was input into the PZT actuator to determine the signal-to-noise ratio (SNR) of the proposed system and the peak-to-peak voltage of the wave was 20 V. The laser power was set to 0 dBm and a high-pass digital filter with cut-off frequency of 10 kHz was employed to remove unnecessary noise. Figure 5(a) shows the temporal responses of the PS-FBG probe. As shown, the probe can clearly detect the ultrasonic waves and the peak-to-peak voltage is about 0.35 V. Figure 5(b) shows the spectral response of the probe, calculated by taking the Fourier transform of the temporal response and a high SNR of 45 dB can be observed in the figure. In addition, the signal appears frequency peak at 400 kHz and the amplitude are about 38 dB lower than at 200 kHz. This can be induced by ultrasonic harmonics due to the nonlinear effect of material [20, 21]. To further research on the properties of the sensitivity (SNR) and bandwidth of the sensor probe, the dependence of sensitivity on frequencies was also tested using a wideband PZT actuator, which has a generally flat frequency response from 100 to 1000 kHz. The input frequency of PZT actuator was set to change with a 25 kHz step. Figure 6 shows the response of PS-FBG probe on frequencies. The probe has the best sensitivity at about 200 kHz. From
#214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19577
100 to 1000 kHz, the maximum fluctuation of the probe sensitivity is about 8.8 dB and the sensitivity keeps above 37 dB all over the range, which verifies the broadband properties of the probe. The wideband probe is quite suitable for ultrasonic imaging of seismic models, where the typical seismic signal covers a wide frequency band from ~100 kHz to 1 MHz.
Fig. 5. Temporal (a) and spectral (b) responses of the PS-FBG probe, while injecting a 200 kHz continuous sinusoidal ultrasonic wave.
Fig. 6. Dependence of PS-FBG sensitivity (SNR) on frequencies from 100 to 1000 kHz.
3.3 Geophysical imaging Geophysical imaging experiments were carried out in accordance with the setup in Fig. 3. The tested model in Fig. 7 was a 4-interface Plexiglas structure, made up by a rectangular block and a wedge block. The thickness and length of the rectangular block was about 4 cm and 60 cm respectively. The wedge block with a taper angle of 3.8 was about 0.2 cm thickness at the tip and 3 cm at the bottom edge. Between the two blocks, there was a 2 cm gap filled with water. In experiment, the PZT actuator was driven by a square-wave pulse signal with duration of 2.5 μs and the scanning device with a spatial resolution of 0.01 cm, was programmed to move along the water surface. The pulse generator is triggered by the
#214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19578
Fig. 7. Side view of the physical model made up of a rectangular and a wedge Plexiglas block. The wedge block is separated by 2 cm from the rectangular one in water.
Fig. 8. Ultrasonic images of the physical model made up of a rectangular Plexiglas block and a wedge block. (a) Image reconstructed from signals of PS-FBG probe. (b) Image obtained by PZT sensor. Each channel (CHAN) corresponds to a scanning step of 0.2 cm.
scanning device and in synch with the DAQ card. Figure 8(a) shows the ultrasonic image reconstructed using our PS-FBG imaging system. For comparison, Fig. 8(b) shows the image obtained using the geophysical imaging system with PZT (resonant frequency at 200 kHz) as the sensing element, provided by Sinopec Geophysical Research Institute. Both sensors captured clear 2D images of the model shown in Fig. 7. Relatively, PS-FBG possesses stronger signals, but shows more noise than PZT sensor due to its broadband properties. Each layer of the model is clearly separated and detected. Different propagation velocities of ultrasound wave in water and Plexiglas result in slanted image of the rectangular block reconstructed by time-of-flight approach. Since the front edge of the wedge block was pretty thin (about 0.2 cm) that obvious diffraction could be seen around the tip. The clear images show that the performance of the PS-FBG ultrasonic imaging system is excellent even under large environmental disturbances during fast scanning of the PS-FBG probe in water. Further test was also done by exchanging the wedge block in Fig. 7 with an elliptic cylinder Plexiglas block shown in Fig. 9. The diameters of its long and short axis were 12 cm and 3 cm respectively. The repeated ultrasonic imaging result is shown in Fig. 10. As expected, the images are in good agreement with shape and structure of the model. In comparison, PS-FBG exhibits longer diffraction waves at edges of the elliptic model and thinner reflecting layers than the PZT sensor, indicating a high sensitivity and spatial resolution. The ultrasonic waves focused toward the center, since the ellipsoidal shape block served as a focusing lens. It could be the reason that the image of the rectangular structure was interrupted just under the elliptic block. #214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19579
Fig. 9. Photograph of the elliptic cylinder Plexiglas model. The inset shows the tested structure. The elliptic block is separated by 2 cm from the rectangular one in water.
Fig. 10. Ultrasonic images of the physical model made up of a rectangular Plexiglas block and an ellipse block. (a) Image reconstructed from signals of PS-FBG probe. (b) Image obtained by PZT sensor. Each channel (CHAN) corresponds to a scanning step of 0.2 cm.
4. Conclusion In this paper, an ultrasonic imaging system based on a PS-FBG probe has been developed for seismic physical modeling studies. PS-FBG exhibits an extremely narrow spectral notch in its reflection bandgap. Shifts of the Bragg wavelength caused by ultrasonic waves were measured by tuning a TLS to the linear range of the notch. To stabilize the imaging system, a tight wavelength tracking method has been employed to suppress the drift of the grating wavelength. By scanning the sensor probe, two-dimensional images of Plexiglas models have been clearly reconstructed by time-of-flight approach. The PS-FBG ultrasonic imaging system has presented excellent performance even under harsh situations during the fast scanning of the PS-FBG probe in water. The proposed system shows a great potential for practical ultrasonic imaging applications, not limited to seismic physical modeling studies.
#214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19580
State Key Laboratory of Precision Measurement Technology and Instruments, Department of Precision Instruments, Tsinghua University, Beijing 100084, China 2 Sinopec Geophysical Research Institute, Nanjing 211103, China * [email protected]
Abstract: We report what is to our knowledge the first ultrasonic imaging of seismic physical models by using a phase-shifted fiber Bragg grating (PS-FBG). Seismic models, which consist of multiple layer structures, were immersed in water. Piezoelectric (PZT) transducer was used to generate ultrasonic waves and a PS-FBG as a receiver. Two-dimensional (2D) ultrasonic images were reconstructed by scanning the PS-FBG with a highprecision position scanning device. In order to suppress the low-frequency drift of the Bragg wavelength during scanning, a tight wavelength tracking method was employed to lock the laser to the PS-FBG resonance in its reflection bandgap. The ultrasonic images captured by the PS-FBG have been compared with the images obtained by the geophysical imaging system, Sinopec, demonstrating the feasibility of our PS-FBG based imaging system in seismic modeling studies. ©2014 Optical Society of America OCIS codes: (060.2370) Fiber optics sensors; (060.3735) Fiber Bragg gratings; (110.2970) Image detection systems.
References and links 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.
M. L. Buddensiek, C. M. Krawczyk, N. Kukowski, and O. Oncken, “Performance of piezoelectric transducers in terms of amplitude and waveform,” Geophysics 74(2), T33–T45 (2009). M. Urosevic, G. Bhat, and M. H. Grochau, “Targeting nickel sulfide deposits from 3D seismic reflection data at Kambalda, Australia,” Geophysics 77(5), WC123–WC132 (2012). B. Evans, J. McDonald, and W. French, “Seismic physical modelling of reservoirs, its past, present and future,” ASEG Extended Abstracts 2007: 19th Geophysical Conference: pp. 1–6. J. K. Coper, D. C. Lawton, and G. F. Margrave, “The wedge model revisited: A physical modeling experiment,” Geophysics 75(2), T15–T21 (2010). G. L. Fradelizio, A. Levander, and C. A. Zelt, “Three-dimensional seismic-reflection imaging of a shallow buried paleochannel,” Geophysics 73(5), B85–B98 (2008). D. Sherlock, J. McKenna, and B. Evans, “Time-lapse 3-D seismic physical modelling,” Explor. Geophys. 31(2), 310–314 (2000). A. Wandler, B. Evans, and C. Link, “AVO as a fluid indicator: A physical modeling study,” Geophysics 72(1), C9–C17 (2007). F. Mahmoudian, G. F. Margrave, P. F. Daley, J. Wong, and E. Gallant, “Determining elastic constants of an orthorhombic material by physical seismic modeling,” Crewes Research Report 21 (2010). R. R. Stewart, N. Dyaur, B. Omoboya, J. J. S. de Figueiredo, M. Willis, and S. Sil, “Physical modeling of anisotropic domains: Ultrasonic imaging of laser-etched fractures in glass,” Geophysics 78(1), D11–D19 (2013). H. Lamela, D. Gallego, and A. Oraevsky, “Optoacoustic imaging using fiber-optic interferometric sensors,” Opt. Lett. 34(23), 3695–3697 (2009). A. Rosenthal, D. Razansky, and V. Ntziachristos, “High-sensitivity compact ultrasonic detector based on a piphase-shifted fiber Bragg grating,” Opt. Lett. 36(10), 1833–1835 (2011). Q. Wu and Y. Okabe, “High-sensitivity ultrasonic phase-shifted fiber Bragg grating balanced sensing system,” Opt. Express 20(27), 28353–28362 (2012). N. Takahashi, A. Hirose, and S. Takahashi, “Underwater Acoustic Sensor with Fiber Bragg Grating,” Opt. Rev. 4(6), 691–694 (1997). P. Fomitchov, “Response of a fiber Bragg grating ultrasonic sensor,” Opt. Eng. 42(4), 956–963 (2003). I. M. Perez, H. Cui, and E. Udd, “Acoustic emission detection using fiber Bragg gratings,” Proc. SPIE 4328, 209–215 (2001). H. Tsuda, E. Sato, T. Nakajima, H. Nakamura, T. Arakawa, H. Shiono, M. Minato, H. Kurabayashi, and A. Sato, “Acoustic emission measurement using a strain-insensitive fiber Bragg grating sensor under varying load conditions,” Opt. Lett. 34(19), 2942–2944 (2009).
#214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19573
17. A. Minardo, A. Cusano, R. Bernini, L. Zeni, and M. Giordano, “Response of fiber Bragg gratings to longitudinal ultrasonic waves,” IEEE Trans. Ultrason. Ferroelectr. Freq. Control 52(2), 304–312 (2005). 18. T. Liu and M. Han, “Analysis of π-Phase-Shifted Fiber Bragg Gratings for Ultrasonic Detection,” IEEE Sens. J. 12(7), 2368–2373 (2012). 19. Q. Wu, Y. Okabe, and J. Sun, “Investigation of dynamic properties of erbium fiber laser for ultrasonic sensing,” Opt. Express 22(7), 8405–8419 (2014). 20. M. A. Breazeale and J. Ford, “Ultrasonic Studies of the Nonlinear Behavior of Solids,” J. Appl. Phys. 36(11), 3486–3490 (1965). 21. D. M. Egle and D. E. Bray, “Measurement of acoustoelastic and third-order elastic constants for rail steel,” J. Acoust. Soc. Am. 60(3), 741–744 (1976).
1. Introduction Since the 1920s, seismic physical model, as a scaled analog to geological structures in reservoirs, has been a successful tool for research in seismic wave propagation and wave theoretical predictions [1, 2]. Experiments in labs are much less expensive as well as more repeatable, and controllable than in the field. A seismic modeling system was constructed employing a meter square water tank, where scaled models of geologic structures were suspended [3, 4]. 2D or 3D surveys were conducted by moving ultrasonic source and receiver over the models [5, 6]. Wandler et al. constructed physical models to distinguish fluid types from mixtures of water, brine, oil and carbon dioxide by characterizing the seismic reflection [7]. Mahmoudian et al. utilized a physical model to experimentally determine the elastic constants of an orthorhombic material [8]. Stewart et al. explored the anisotropic cracked or fractured zones by imaging laser-etched models, which was intended to assist in assessment of fracture zones in actual reservoirs [9]. Seismic physical models provide an effective way to bridge the theory and field-scale experiments and enable us to measure changes of the acoustic response in absence of a rock matrix in a nearly ideal setting. In physical modeling studies, piezoelectric (PZT) transducers usually serve as source and receiver, which are highly sensitive but have narrow bandwidth owing to their strongly resonant effects. The detectors based on thin piezoelectric polymer films, can achieve a broadband response using appropriate matching materials [10]. However, their sensitivity decreases with decreasing size and the corresponding electrical capacitance. In addition, PZT sensors, due to their electrical nature, are very sensitive to ambient electromagnetic disturbances [11, 12]. Optical means based on FBGs for acoustic detection has been studied as an alternative to PZT sensors by many researchers [13, 14]. Perez et al. utilized a broadband amplified spontaneous emission (ASE) light source and matched FBGs to detect acoustic emission (AE) signals [15]. To avoid the high intensity noise induced by ASE source, Tsuda et al. adjusted a narrow linewidth laser to the 3 dB position of the peak in the reflection spectrum of a normal FBG to achieve a high signal-to-noise ratio (SNR) [16]. According to the analyses in [17] and [18], the grating length and the spectrum slope determine the frequency response and intensity sensitivity, respectively. Consequently, FBGs with short grating and sharp slope are needed for high-frequency weak signals. A special type of FBGs with a π-phase shift in its center has aroused extensive attentions for ultrasonic detection. The π-phase shift in center of the grating decreases the effective length of sensor and exhibits an extremely sharp resonance for high sensitivity. Rosenthal et al. achieved a 10MHz bandwidth of ultrasonic detection by tuning the wavelength of a continuous-wave (CW) laser to the maximum slope of a pi-phase-shifted FBG and an inverse relation between sensitivity and effective sensing length was obtained [11]. To further improve the PS-FBG based ultrasonic sensor for weak AE signals, Wu et al. demonstrated a PS-FBG balanced system to effectively suppress the laser intensity noise and common mode noise, and a high sensitivity of 9 nε/Hz1/2 was achieved [12]. Though PS-FBG exhibits great potential for ultrasonic sensing, the environmental perturbations caused by temperature fluctuations and large quasi-static strains, easily shift the Bragg wavelength of PS-FBG out of its dynamic range. Against this problem, Wu et al. built an erbium fiber laser with an inbuilt PS-FBG for ultrasonic sensing [19]. When temperature or quasi-static strain shifts the wavelength of PSFBG, the lasing longitudinal mode will hop to a neighboring position, giving a solution to the instability of PS-FBG based sensing system. Unfortunately, due to the uncontrollable relative
#214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19574
positions of the lasing and PS-FBG wavelength, 22 dB instability of the detected signal was found in their measurements. In this study, we demonstrate the first ultrasonic imaging system based on a PS-FBG probe for seismic modeling studies. The imaging system is robust to environmental perturbations by locking the laser to the PS-FBG resonance wavelength. The PS-FBG probe, capable of mobility and ultrasonic detection was mounted on a high-precision position scanning device. By scanning the sensor probe, Two-dimensional (2D) images can be reconstructed according to time-lapse changes in ultrasonic reflection. 2. Sensor design and system setup To realize ultrasonic imaging, continuous scanning measurement of ultrasonic reflections in tested models is needed. To give mobility and reliability to the PS-FBG sensor, we contrived a sensor probe in Fig. 1. The PS-FBG with grating length of 5 mm was manufactured by the Fujikura Company. The reflectivity of the grating was designed to above 90%. The fiber grating, recoated with polyimide coating, is glued on a steel pole of 0.15 cm diameter and 5 cm length with an epoxy resin adhesive. A certain thickness of epoxy resin adhesive forms an outer sheath layer to protect the PS-FBG sensor. The Ultrasonic wave, conducted through the steel pole and epoxy resin layer, can be effectively coupled to sensing area of the PS-FBG.
Fig. 1. Structure of the mobile PS-FBG probe. Ultrasonic wave is coupled to PS-FBG through the steel pole and the epoxy resin coverage layer.
Fig. 2. Close-up of PS-FBG reflection dip measured by sweeping the TLS. The reflection bandwidth is 0.012 nm and the linear slope is 86 nm−1. Inset: wavelength range from 1547 nm to 1551.5 nm by linear range.
Figure 2 shows the reflection spectrum of PS-FBG used in the sensing system measured by sweeping a tunable laser source (TLS) from 1547 nm to 1551.5 nm. With a π-phase shift in center of the grating, an extremely narrow spectral notch is formed for highly sensitive
#214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19575
ultrasonic detection. The Bandwidth of PS-FBG is 0.012 nm and the linear slope is 86 nm−1, which is about 172 times over a normal FBG of 0.5 nm−1 (typical). Figure 3 shows the schematic diagram of the proposed imaging system. A tunable laser source (Santec, 710) with 100 kHz linewidth and 0.1 pm tunable resolution is used to interrogate the fiber sensor. An optical circulator guides the light reflected from the grating to a photoelectric detector (New Focus, 2053) with a bandwidth of 10 MHz at 0 dB gain. A signal conditioning unit is also designed for pre-amplification and denoising. The demodulated signals are eventually collected by a DAQ (Data Acquisition) card and captured by computer for image processing. Because ultrasound is not effectively transmitted through air, Plexiglas models are placed at the bottom of a 2 × 1.5 × 1 m3 (l × w × h) water tank. The PS-FBG and PZT actuator are just located on the surface of water, which is above the physical model around 7.5 cm. A pulse generator (PG) is used to drive the PZT actuator, which generates pulse ultrasonic signals with duration of 2.5 μs and is kept at a fixed position while measuring. The location arm with PS-FBG probe is programmed to move along the water surface with a spatial resolution of 0.01 cm, during which both of the ultrasonic waves and the probe’s coordinates should be real-time recorded.
Wavelength tracking
Single-mode fiber
TLS
Circulator 1 2
PG Location control arm
3 Servo
PD
Signal conditioning X Y DAQ
PS-FBG probe
Water
PZT actuator
Computer Model under test
Fig. 3. Schematic diagram of the proposed ultrasonic imaging system. The model is put into a water tank. The PZT actuator and PS-FBG are just located on the water surface, which is above the physical model 7.5 cm. TLS: tunable laser source, PD: photodiode, DAQ: Data Acquisition, PG: Pulse Generator. The PZT actuator and PS-FBG were controlled by a position scanning device with a spatial resolution of 0.01 cm.
3. Experimental results and discussion 3.1 System stability In the measurement, the wavelength of TLS was tuned to the linear range of the grating. The basic principle of this demodulation technique is based on edge filter detection. Ultrasonic waves can be directly detected by monitoring the reflected power of PS-FBG and changes in output power can be expressed as: ΔP = Δλ GP
(1)
Δλ represents the wavelength shift of PS-FBG, G is the grating slope at the laser wavelength, and P is the input power. It can be seen that the intensity sensitivity depends on the grating slope. However, due to environmental perturbations caused by temperature fluctuations and water sloshing during measurements, we found the slope rate can change significantly with time judging from the reflected power. To solve this problem, a wavelength tracking method
#214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19576
was employed to lock the laser to the 3 dB position on linear slope of the grating resonance (marked by red line in Fig. 2). We choose the 3 dB point as the locking position as it can ensure a large dynamic range in the bidirectional perturbations. Within the linear range, the demodulated signal, which could be used as an error signal of wavelength detuning between the laser and PS-FBG was fed back for fine tuning of the laser wavelength through an integral and proportional servo shown in Fig. 3. A typical control loop bandwidth of 1 kHz was used, which was fast enough to efficiently avoid the low-frequency wavelength drifts of PS-FBG, and enabled a tight laser wavelength tracking even when the sensor was subject to strong environmental perturbations. To illustrate this, we recorded the reflected power of PS-FBG in observation for about 30 minutes without ultrasound induced. Figure 4 shows a comparison of normalized reflected intensities by PS-FBG over time with/without wavelength tracking. At the beginning, the laser wavelength was tuned to about 3 dB position of the PS-FBG resonance at a slope rate of about 86 nm−1. In the case of PS-FBG without wavelength tracking, the Bragg wavelength slowly fluctuated back and forth as time went on, and the laser was eventually drifted out of the resonance wavelength range. Contrastively, in the case of wavelength tracking, the laser was finely locked to the half maximum of grating spectrum by the feedback servo, which ensured the practicability and long-term stability of the imaging system.
Fig. 4. Normalized intensity changing over time in observation for 30 minutes without ultrasound induced. Red curve: PS-FBG without wavelength tracking. Blue curve: with wavelength tracking. In the case of wavelength tracking, the half maximum point of the PSFBG resonance was chosen as the locking position to ensure a large dynamic range in the bidirectional perturbations.
3.2 Characterization of PS-FBG probe A continuous sinusoidal wave of 200 kHz was input into the PZT actuator to determine the signal-to-noise ratio (SNR) of the proposed system and the peak-to-peak voltage of the wave was 20 V. The laser power was set to 0 dBm and a high-pass digital filter with cut-off frequency of 10 kHz was employed to remove unnecessary noise. Figure 5(a) shows the temporal responses of the PS-FBG probe. As shown, the probe can clearly detect the ultrasonic waves and the peak-to-peak voltage is about 0.35 V. Figure 5(b) shows the spectral response of the probe, calculated by taking the Fourier transform of the temporal response and a high SNR of 45 dB can be observed in the figure. In addition, the signal appears frequency peak at 400 kHz and the amplitude are about 38 dB lower than at 200 kHz. This can be induced by ultrasonic harmonics due to the nonlinear effect of material [20, 21]. To further research on the properties of the sensitivity (SNR) and bandwidth of the sensor probe, the dependence of sensitivity on frequencies was also tested using a wideband PZT actuator, which has a generally flat frequency response from 100 to 1000 kHz. The input frequency of PZT actuator was set to change with a 25 kHz step. Figure 6 shows the response of PS-FBG probe on frequencies. The probe has the best sensitivity at about 200 kHz. From
#214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19577
100 to 1000 kHz, the maximum fluctuation of the probe sensitivity is about 8.8 dB and the sensitivity keeps above 37 dB all over the range, which verifies the broadband properties of the probe. The wideband probe is quite suitable for ultrasonic imaging of seismic models, where the typical seismic signal covers a wide frequency band from ~100 kHz to 1 MHz.
Fig. 5. Temporal (a) and spectral (b) responses of the PS-FBG probe, while injecting a 200 kHz continuous sinusoidal ultrasonic wave.
Fig. 6. Dependence of PS-FBG sensitivity (SNR) on frequencies from 100 to 1000 kHz.
3.3 Geophysical imaging Geophysical imaging experiments were carried out in accordance with the setup in Fig. 3. The tested model in Fig. 7 was a 4-interface Plexiglas structure, made up by a rectangular block and a wedge block. The thickness and length of the rectangular block was about 4 cm and 60 cm respectively. The wedge block with a taper angle of 3.8 was about 0.2 cm thickness at the tip and 3 cm at the bottom edge. Between the two blocks, there was a 2 cm gap filled with water. In experiment, the PZT actuator was driven by a square-wave pulse signal with duration of 2.5 μs and the scanning device with a spatial resolution of 0.01 cm, was programmed to move along the water surface. The pulse generator is triggered by the
#214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19578
Fig. 7. Side view of the physical model made up of a rectangular and a wedge Plexiglas block. The wedge block is separated by 2 cm from the rectangular one in water.
Fig. 8. Ultrasonic images of the physical model made up of a rectangular Plexiglas block and a wedge block. (a) Image reconstructed from signals of PS-FBG probe. (b) Image obtained by PZT sensor. Each channel (CHAN) corresponds to a scanning step of 0.2 cm.
scanning device and in synch with the DAQ card. Figure 8(a) shows the ultrasonic image reconstructed using our PS-FBG imaging system. For comparison, Fig. 8(b) shows the image obtained using the geophysical imaging system with PZT (resonant frequency at 200 kHz) as the sensing element, provided by Sinopec Geophysical Research Institute. Both sensors captured clear 2D images of the model shown in Fig. 7. Relatively, PS-FBG possesses stronger signals, but shows more noise than PZT sensor due to its broadband properties. Each layer of the model is clearly separated and detected. Different propagation velocities of ultrasound wave in water and Plexiglas result in slanted image of the rectangular block reconstructed by time-of-flight approach. Since the front edge of the wedge block was pretty thin (about 0.2 cm) that obvious diffraction could be seen around the tip. The clear images show that the performance of the PS-FBG ultrasonic imaging system is excellent even under large environmental disturbances during fast scanning of the PS-FBG probe in water. Further test was also done by exchanging the wedge block in Fig. 7 with an elliptic cylinder Plexiglas block shown in Fig. 9. The diameters of its long and short axis were 12 cm and 3 cm respectively. The repeated ultrasonic imaging result is shown in Fig. 10. As expected, the images are in good agreement with shape and structure of the model. In comparison, PS-FBG exhibits longer diffraction waves at edges of the elliptic model and thinner reflecting layers than the PZT sensor, indicating a high sensitivity and spatial resolution. The ultrasonic waves focused toward the center, since the ellipsoidal shape block served as a focusing lens. It could be the reason that the image of the rectangular structure was interrupted just under the elliptic block. #214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19579
Fig. 9. Photograph of the elliptic cylinder Plexiglas model. The inset shows the tested structure. The elliptic block is separated by 2 cm from the rectangular one in water.
Fig. 10. Ultrasonic images of the physical model made up of a rectangular Plexiglas block and an ellipse block. (a) Image reconstructed from signals of PS-FBG probe. (b) Image obtained by PZT sensor. Each channel (CHAN) corresponds to a scanning step of 0.2 cm.
4. Conclusion In this paper, an ultrasonic imaging system based on a PS-FBG probe has been developed for seismic physical modeling studies. PS-FBG exhibits an extremely narrow spectral notch in its reflection bandgap. Shifts of the Bragg wavelength caused by ultrasonic waves were measured by tuning a TLS to the linear range of the notch. To stabilize the imaging system, a tight wavelength tracking method has been employed to suppress the drift of the grating wavelength. By scanning the sensor probe, two-dimensional images of Plexiglas models have been clearly reconstructed by time-of-flight approach. The PS-FBG ultrasonic imaging system has presented excellent performance even under harsh situations during the fast scanning of the PS-FBG probe in water. The proposed system shows a great potential for practical ultrasonic imaging applications, not limited to seismic physical modeling studies.
#214763 - $15.00 USD Received 30 Jun 2014; revised 28 Jul 2014; accepted 28 Jul 2014; published 6 Aug 2014 (C) 2014 OSA 11 August 2014 | Vol. 22, No. 16 | DOI:10.1364/OE.22.019573 | OPTICS EXPRESS 19580
