Técnicas avanzadas para la interrogación en análisis óptico Brillouin en el dominio del tiempo

Abstract

Large infrastructures are key elements in our everyday lifestyle. Bridges, dikes, tunnels, highspeed railways, dams, pipeline transport for oil and gas or high power lines should all guarantee a correct and continuous operation, and thus, solutions to preserve and control the integrity of these installations are needed. In this context, fiber-optics sensors can be an interesting solution since they can be embedded within the structure with minimal impact and provide information of the stress, strain, temperature or deformation suffered by the structure. The main advantages of this technology rely in the good properties of the optical fiber: low attenuation, immunity to electromagnetic noise and deflagrations, high transmission speed, small size, light weight, flexibility, etc. For a wide range of applications, the only feasible solution for monitoring is based on fiber-optic sensors, particularly in some harsh environments. Among optical fiber sensors, distributed sensors are very adequate for monitoring big infrastructures as they allow to monitor thousands of measurement points with only one optical fiber cable and a single interrogator. Distributed fiber sensors based on stimulated Brillouin scattering (SBS) are the central object of this thesis dissertation. This is a very versatile technology since it can be used in different scenarios such as long-range monitoring distances with hundreds of kilometers with sub-meter spatial resolution or shorter-distances, around few kilometres, with centimeters spatial resolution. Although BOTDA sensors have been thoroughly studied, there still exist some limitations in these sensors. The quality of BOTDA measurements is driven by several parameters: the frequency step taken to sweep the pump-probe frequency offset, the number of averages in each time trace to reduce the noise, the Brillouin bandwidth, the sensing range (limited by the optical attenuation of the fiber) or the pulse width. All these parameters, together with the probe power (limited by pump depletion) and pump power (limited by modulation instability), have an impact on the SNR of the measurements, and ultimately the frequency uncertainty in the determination of the BFS. Finding methods to increase the performance of the sensors (improve the system SNR, reduce the measurement time, reduce the complexity or the cost of the sensor, etc.) turns critical to convert them in a real, practical and cost effective solution for many applications. This thesis dissertation concentrates on solving some of the limitations described by proposing novel techniques. The first experimental technique is a completely passive system, with no moving parts, to eliminate polarization noise in a BOTDA sensor. Polarization noise arises in BOTDA due to the strong polarization sensitivity of SBS, which leads to a decrease of the measurement SNR. This technique is based on the use of passive depolarization of the pump pulse together with balanced detection among orthogonally polarized Stokes and anti-Stokes bands of the probe signal. This approach is considerably simpler and more reliable than a polarization scrambler. The second technique proposed during this thesis dissertation is a method for distributed temperature/strain measurements in BOTDA based on the use of the nonlinear phase-shift (BPS) induced by SBS. The phase shift feature of SBS is rarely used for the development of sensors BOTDA. However, this feature shows significant interest for sensing. Employing a Sagnac interferometer (SI), the position-resolved BPS along the fiber can be obtained, benefiting from the sensitivity to nonreciprocal phase-shifts of the SI scheme. This proposal simplifies the existing methods to retrieve the BPS distribution along an optical fiber since phase modulation, filtering, and high-bandwidth detectors are not required. Moreover, the thesis presents a comparative study, theoretical and experimental, of the temperature/strain determination error when utilizing a linear fit over the BPS instead of a quadratic-fit over the Brillouin gain spectrum (BGS) in BOTDA systems as a function of the different experimental parameters. In the literature, the performance of standard (gain/loss) BOTDA systems had been thoroughly analyzed and evaluated in terms of the best achievable error in the temperature/strain determination. However, to the best of our knowledge, a similar analysis had never been performed for the phase profile case. Finally, a simple scheme allowing to perform distributed BPS measurements with very high spatial resolution over long optical fibers is proposed. This is achieved by inserting a SI in a BOCDA configuration (SI-BOCDA). Over its time-domain equivalent configuration (SIBOTDA), this approach reduces its main source of noise, and over the most usual schemes used for distributed BPS measurements, presents the key advantage of not requiring highbandwidth detection or complex modulation, while reaching unprecedented values of spatial resolution and number of resolved points for this type of measurements. The presented technique is, as far as we know, the first demonstration of a phase-measuring time-gated BOCDA.