Very Sensible Fiber
by by Dominic Siriani
…perhaps the last bad title pun you’ll have to suffer for FiO/LS 2013. Today was the Symposium on Advanced Distributed Optical Fiber Sensor Systems. In a previous post, I wrote a little bit about these fiber sensors. In really brief review, they work on the principle that fibers scatter light (Rayleigh, Raman, Brillouin…), this scattering is sensitive to perturbations (stresses, heating…), and so changes in the scattered return signal translate to sensing an environmental change. So today I got a taste of a little bit more detail about the challenges and advances in the field.
First off, maybe it’s important to discuss why these sensors are significant. After all, there are already sensors to detect changes in stresses or temperature. Well, the keyword here is “distributed.” Fiber sensors can detect changes is the environment over long ranges,
i.e., tens to hundreds of meters. And they can do so using the same types of fibers and lasers that are all over to support our enormous telecommunications network. This means that instead of using a large number of point sensors, one can employ a single distributed fiber sensor, saving a lot on cost and complexity.
Naturally, there’s a trade space associated with distributed fiber sensors, otherwise things would just be too easy. In this case, a primary trade is between resolution and range. For example, in an optical time domain reflectometry (OTDR) setup, the resolution is set by the width of the optical pulse launched down the fiber; the shorter the pulse width, the higher the resolution. However, one also has to contend with losses in the fiber, which reduces the signal strength as it gets further down the fiber and thus degrades the signal-to-noise ratio. One might consider increasing the pulse power, but then limits due to optical nonlinearities arise. Or one could try increasing the pulse width, but at the expense of resolution.
So how does one get around these issues? That was a major theme of today’s symposium, and I’ll run through just a few solutions. For one, a hybrid method can be employed, that is, a method that utilizes more than one of the scattering/sensing mechanisms. Since the use of Rayleigh, Raman, Brillouin, and fiber Bragg grating scattering all have different advantages and trades, a combination of them can result in an accurate, high resolution, large range sensing system. Another approach is to actually tailor the launched pulses, called pulse coding, to simultaneously avoid nonlinearities and still be able to solve an inverse problem that yields high resolution and large range.
Alright, the question might be why do we care? In a word, applications! Fortunately, there are several. One of the most noticeable application areas is for structural sensing. If you have an aircraft wing, a support structure, a ventilation duct, pipelines, or anything else whose extensive structural integrity needs to be excellent yet is susceptible to compromise, these fiber sensors can help. Since any stresses in the fiber can be sensed and pinpointed, attaching them to a structure of interest allows one to identify where weaknesses arise. Another application is intrusion detection. For example, one can detect vibrations that are caused by environmental disturbances. In fact, one talk today illustrated that by using an advanced technique utilizing a multicore fiber, one can even detect how the shape and bends of a fiber are distorted, so you can tell precisely how things are changed! And if you’re interested in more than just physical disturbances, it’s possible to detect very high currents in a polarization-sensitive distributed fiber sensor by employing the magnetic-field induced change in fiber birefringence.
So, as you can see, distributed fiber sensors can do a whole lot and there are many innovative ideas to make them better. It seems that this is a rapidly advancing field. In fact, a panel discussion at the conclusion of the symposium began by noting that a market estimate predicts a 300% growth in distributed fiber sensors over 2013-2017! I guess we can see a lot more of these novel sensors in the coming years.
Disclaimer: Opinions, interpretations, conclusions, and recommendations are those of the author and are not necessarily endorsed by the United States Government and MIT Lincoln Lab
Posted: 10 Oct 2013 by by Dominic Siriani