Time-expanded phase sensitive optical time-domain reflectometry

Abstract

Distributed optical fiber sensors (DOFS) have proven a very useful, versatile and cost-effective tool for monitoring processes that require a sensor network with very high numbers of points (typically beyond 1,000 sensing points). Conventionally, the DOFS are employed for surveillance purposes in tunnels, dams, power lines or submarine cables. Today, some of the most promising distributed sensing technologies are based on Rayleigh scattering, including phase-sensitive optical time domain reflectometry (ϕOTDR) and optical frequency domain reflectometry (OFDR). On the one hand, the ϕOTDR systems are characterized by an elevated sensing range and a high sampling rate that reaches the acoustic scale, although their typical spatial resolution is limited to a few meters. On the other hand, OFDR technology usually allows recovering the spectral response of a fiber under test with high spatial resolution, even reaching the millimeter range. However, such resolution is achieved at the cost of reducing its measurement range and sampling rate. In this Thesis, a novel high-resolution interrogation system based on the ϕOTDR technique has been proposed. The main objective is to develop a distributed sensing technique capable of dynamically interrogating fibers under test with centimeter spatial resolution. The introduction of dual-frequency comb technology (DFC) has been critical to meet such objective. The spectral compression process associated with the use of DFCs has made it possible to recover fibers responses with a high spatial resolution (thanks to using a signal with a wide optical bandwidth) with reduced electrical detection bandwidths, up to five times smaller than the required in a conventional scheme. In the time domain, this decoupling between the optical and electrical detection bandwidths results in a time-expansion process of the recovered backscattered traces. For this reason, the new interrogation technique is known as time-expanded phase-sensitive optical time-domain reflectometry (TE-ϕOTDR). Another advantage associated with the employment of DFC is the possibility of implementing efficient coding strategies with automatic decoding (i.e., without having to introduce decoding strategies based on digital methods or reference measurements). All this results in an efficient, flexible, distributed sensing technique with a simple detection system that has the potential to offer high-resolution dynamic measurements in real-time. Experimentally, different proofs of concept have been carried out to validate the proposed technique. In particular, the effects of the temporal expansion on the backscattered traces have been studied. Hence, increasing the duration of the recovered scattered signals leads to an improvement of the signal-to-noise ratio (SNR) of the mentioned scattered signals. On the other hand, it has been worked on different spectral phase coding techniques that have allowed for increasing the power level of the optical waveforms used to probe the fiber under test. This way, it has been possible to compensate for the reduction of the SNR of the backscatter traces when increasing the spatial resolution of the acoustic sampling rate. Besides, a structural health monitoring of a flexible wing specimen designed for an unmanned aircraft was performed, verifying the applicability of this technique in fully distributed measurements. Furthermore, a series of proposals have been made to improve the performance and simplicity of the interrogator system. For example, different interrogation schemes have been developed to reduce the complexity and increase the stability of the experimental setup, in view of a future portable interrogator. However, the most interesting proposal is the introduction of a DFC scheme whose optical frequency combs have a quasi-integer ratio in their repetition rates. This configuration has allowed us to bring distributed acoustic sensing to distances close to one kilometer with centimeter spatial resolution. Finally, a series of conclusions are included in order to summarize all the objectives achieved during this Thesis, new applications and future lines of work.