The Spanish-Swiss research team envisions applications of their improved fiber optic sensing method both in large-scale, infrastructure projects and in biomedical monitoring. [Image: The Optical Society]
The need to repair aging infrastructure, and to monitor existing bridges, dams and other large structures, has cast new light on distributed fiber optic sensors that can detect strain and temperature changes. One common category of fiber optic sensors, however—those based on the nonlinear-optical phenomenon called stimulated Brillouin scattering (SBS)—has some fundamental limitations on spatial range and resolution that have been tough to overcome.
Now, researchers in Spain and Switzerland have broken through those barriers, revealing an approach that can detect changes in temperature or strain at a million points across a 10-km span—centimeter-scale spatial resolution—while still keeping measurement times reasonable (Opt. Lett., doi: 10.1364/OL.42.001903). The method’s resolution gains, in the view of the team, could find applications both in infrastructure monitoring over long distances and in more intimate biomedical settings.
Signal distortions
SBS fiber sensors work by sending a pulsed laser signal, the pump, through a length of fiber against a counterpropagating continuous-wave (CW) probe laser. (In practice, to prevent certain systematic errors, these systems commonly use two CW probe waves, separated by a modulation frequency related to the fiber’s material characteristics—a so-called double-sideband scheme.) The pump pulse interacts nonlinearly with the fiber to create stimulated Brillouin scattering, inelastic Stokes and anti-Stokes scattering that shifts the frequency distribution of the probe light. This so-called Brillouin frequency shift depends on material properties in the fiber that vary with strain and temperature; thus, analysis of the Brillouin shift can be used to sense changes in those parameters along the fiber’s length.
While SBS-based fiber optic sensing has already found a place in a variety of infrastructure settings, it has a number of problems. One of those is limited range. Recent analyses have shown that the increases in the probe power required for spans of many kilometers can distort the pump pulse, severely complicating the prospect of an accurate reading of the Brillouin frequency shift (and, thus, of the strain or temperature experienced by the fiber) over those broad spans.
Another problem is limited spatial resolution. Because SBS relies on a nonlinear light-matter interaction that generates acoustic waves, there’s a small but significant time lag that puts a floor on spatial resolution in time-domain techniques. Alternative techniques in the frequency and correlation domains can do much better, but also take much longer—on the order of an hour or more to measure a million points along a length of fiber.
A question of scanning
The Spanish-Swiss research team, including scientists from the University of Alcala, Spain, and EPFL, Switzerland, appear to have found a way around these quandaries. They did so by digging into the details of how the signal is actually scanned to reveal the Brillouin frequency shift tied to changes in strain or temperature.
In most time-domain SBS-based fiber sensing schemes, the frequency shift is determined by symmetrically scanning the frequency offset of the two sideband probe beams against the fixed pump frequency. It turns out, however, that this scanning method is a primary source of distortion at high probe powers. That’s because of difficult-to-quantify, asymmetric energy transfers between the two probe sidebands and the pump pulse—an effect that increases with increasing probe power.
The Alcala-EPFL team found that by changing the scanning method—such that the sideband probe beams are held at a fixed frequency difference (tied to the Stokes and anti-Stokes frequencies of the fiber), and the input pump beam is instead swept across the relevant frequencies—the distortion can be significantly reduced. That means that much higher probe-beam power levels, and thus longer spans for the fiber sensing system, become feasible. Moreover, by cleaning up distortions in the pump pulse, the same change also should enable higher spatial resolution along the fiber.
In the new technique, rather than being scanned relative to a fixed-frequency pump pulse, the two sideband probe beams themselves are held constant at the Stokes and anti-Stokes frequencies, canceling out potential distortions in the pump pulse. The optical frequency of the pump pulse is then swept across an appropriate range to scan for the Brillouin frequency response. [Image: A. Dominguez-Lopez et al., Opt. Lett., doi: 10.1364/OL.42.001903]
Single-centimeter resolution
The researchers tested out these ideas in a differential-pulse-width pair, Brillouin optical time-domain analysis (DPP-BOTDA) experiment involving a 10-km length of single-mode fiber. They found that they could resolve the Brillouin frequency shift at a million points along the fiber, for a 1-cm resolution, and were able to detect a 3-cm “hot spot” at the far end of the fiber. And, because the system remained in the time domain, it could achieve these feats in less than 20 minutes, rather than the hours required for frequency- and correlation-domain approaches.
The team believes that in addition to applications in infrastructure, the technique could find use in other domains. “Because we have such a large density of sensing points,” noted lead author Alejandro Dominguez-Lopez of the University of Alcala in a press release, the sensor “could also be used for monitoring in applications such as avionics and aerospace, where it’s important to know what is happening in every inch of a plane wing.” The researchers also think that the system’s higher resolution might benefit some biomedical applications, such detecting temperature deviations present in breast cancer.