Brillouin Distributed Optical Fiber Sensor Based on a Closed-Loop Configuration

A Brillouin optical time-domain analysis (BOTDA) method based on a closed-loop control system is proposed to track fast variations of the Brillouin frequency shift along the sensing fiber. Whilst the method eliminates the gain spectral scanning, the exact distributed Brillouin frequency profile is retrieved directly from the output of a closed-loop controller with no need of post-processing. Moreover, as the operating frequency is being continuously updated to follow the Brillouin frequency change, an unlimited temperature or strain measurement range can be achieved. Both theoretical analysis and experimental results validate that the closed-loop controlled BOTDA acts as a low-pass filter that considerably rejects the noise from photodetector, with an efficiency which fundamentally outperforms basic averaging. By optimizing the closed-loop parameters, the measurement time is reduced from a few minutes to a couple of seconds compared with standard BOTDA, i.e., two orders of magnitude improvement in terms of measurement speed, while keeping the same accuracy and measurement conditions. If the sampling time interval that is limited by our instrument can be further reduced, the method offers the potentiality of km-range sensing with sub-second measurement time, with an unmatched favorable trade-off between measurand accuracy and closed-loop delay.

Brillouin Echoes for Advanced Distributed Sensing in Optical Fibres

Stella Foaleng Mafang

Brillouin scattering is particularly efficient and attractive for the implementation of strain and temperature distributed sensing in optical fibres. Recently a trend has been observed that modern advanced applications require a substantial step towards better spatial resolution, while preserving temperature/strain precision over a long range. For this purpose the state of the art does not satisfy all these requirements. In this thesis we present a radically new approach named Brillouin Echoes distributed sensing (BEDS) that allows covering these requirements. In the first part, we propose an updated configuration of the classical existing Brillouin sensor for time domain analysis allowing drastic noise reduction. Then we investigate the limitations (due to non-linear effects) of the classical Brillouin sensor in terms of long distance range measurements. The identified nonlinear effects are pump depletion due to SBS itself, self-phase modulation (SPM), modulation instability (MI), which occurs only in fibres presenting an anomalous dispersion at the pump wavelength and Raman scattering (RS). We propose the modeling of the pump depletion effect to obtain analytical expressions that are useful for the proper design of a BOTDA sensor and for the determination of a very small depletion. The model confirmed by experimental measurements is informative on the conditions maximizing the depletion effect; therefore a standard configuration can be defined to test the value of the depletion in the set-up. Furthermore, we demonstrate that SPM-induced spectral broadening can have a significant effect on the measured effective gain linewidth. Modeling and experiments have undoubtedly demonstrated that the effective gain linewidth can easily experience a two-fold increase in standard conditions when the pulse intensity profile is Gaussian. We showed that the problem can be practically circumvented by using a clean rectangular pulse with very sharp rising and falling edges. The theoretical and experimental analysis of the undesirable effects of MI and forward RS in distributed BOTDA sensors systems gives a simplified expression to predict the critical power for a given distance range. MI turns out to be the dominant nonlinear limitation since it shows the lowest critical power, but it is less critical since it can be avoided to a wide extent by using the fibre in the normal dispersion spectral region such as a DSF in the C-band. On the other hand Raman scattering can be avoided only by limiting the optical pump power and therefore is the ultimate nonlinear limitation in a distributed sensing system. Under similar conditions RS shows a critical power ∼5 times larger than MI. In the second part, we present the new approach Brillouin echo distributed sensing (BEDS) which has proved to be a powerful solution to realize sub-metric spatial resolutions in Brillouin distributed measurements. We have demonstrated both theoretically and experimentally that an optimized configuration is reached when the optical wave is π-phase shifted. The experimental tests have shown a spatial resolution down to 5 cm, with a clear margin for further improvement down to a real centimetric spatial resolution. This optimized configuration produce the best contrast independently of the pulse intensity, with a factor 2 of improvement compared to other techniques based on the same approach (dark pulse, bright pulse). This extends the dynamic range by 3 dB, which corresponds in standard loss conditions to a 5 km extension of the sensing range. An analytical developed model has proved to be an excellent tool not only for optimizing the pumping scheme but also in post-processing the measured data. Finally the potentialities of BEDS technology provide solutions in real contexts. Using the BEDS technology in landslide monitoring at laboratory scale, for the first time it became possible to observe the failure propagation in laboratory scale with an accurate precision. Furthermore, using BEDS we have proposed and demonstrated the possibility of mapping geometrical structure fluctuations along a photonic crystal fibre (PCF). Both long- and short-scale longitudinal fluctuations in the Brillouin frequency shift have been identify and quantify. Observation of Brillouin linewidth broadening in PCF fibre through distributed measurement of the Brillouin gain spectrum using BEDS has allowed fundamental understanding of SBS in PCF fibre and in their design in view of applications to optical-strain/temperature sensing.

EPFL2011

Generation and application of dynamic gratings in optical fibers using stimulated Brillouin scattering

Nikolay Primerov, Luc Thévenaz

Since the invention of optical fiber cable data transmission using light has become ubiquitous in the past years. Nonlinear optical effects, which are hardly noticed in our daily life, have become crucial and highly influential for the design and performance of advance high-capacity systems. Future optical systems will definitely require signal processing operations to be performed on data signals “on the fly” solely in the optical domain. Brillouin scattering is currently one of the most prominent nonlinear effects in optical fibers, as it has found many applications in various fields of photonics, such as slow light, new high coherent sources, filters, optical fiber sensing, etc. In this thesis we investigate a novel phenomenon based on Stimulated Brillouin Scattering, which is called dynamic Brillouin grating (DBG). Investigating the process of DBG generation in optical fibers we propose several new approaches to realize new tools for signal processing of the optical signal by using phonon-photon interactions. First an analytical model is proposed to describe localized DBG generation in optical fibers using three coupled equations for Stimulated Brillouin Scattering. Based on this model we can show how the grating is generated and which shape it has in time when two pump pulses are used with optical frequency difference equal to the Brillouin shift of the fiber. Moreover optical phase conjugation is also theoretically predicted for the optical wave reflected from DBG. Several applications of Dynamic Brillouin grating are identified and investigated. An optical signal delay line or buffer is proposed and experimentally demonstrated using DBG principle. Storage of an input optical pulse is achieved by changing the position of the localized DBG inside the fiber. Here DBG acts as a Bragg reflector for the optical pulse and change in DBG position defines the time-of-flight of the input and reflected pulse. Single pulse delays can be dynamically controlled with a buffer capacity ranging from several picoseconds to more than one microsecond. Such delays show that DBG based signal delay lines can easily outperform slow light based delay lines in terms of delay-bandwidth product. Optical delay of pulse trains is also achieved, but with maximum limited amount of delay governed by the acoustic wave decay constant, which in standard silica fibers is around 10-12 ns. Moreover microwave photonic signals can be also delayed using DBG demonstrating true time delay. In this configuration, the reflection bandwidth of the DBG defines the maximum operational radio frequency. Relation of the DBG bandwidth as a function of its length was measured and it was determined that DBG exhibits similar properties as weak uniform fiber Bragg grating (FBG). Another application of the DBG gratings is demonstrated by exploiting this phenomenon to realize microwave photonic filters. For the first time, two-, three- and four-tap microwave photonic filter configurations are realized based on DBG. Free spectral range and IV tcartsbA stop band central frequencies of the proposed microwave filters can be easily tuned by adjusting the properties and positions of DBGs. Dynamic gratings generated by amplitude modulated pump waves are used to realize optical differentiation, integration and time reversal of input optical signals. Performed experiments demonstrate that a DBG based signal processor performs operations on the complex amplitude of the input optical wave. The DBG based integrator offers high flexibility in terms of operational wavelength and variable integration time, which can be controlled by changing the DBG length. For the first time, new operations of true time reversal of several bit sequences are performed using dynamic Brillouin grating. To overcome the problem of the decay of the localized acoustic grating, a novel method based on phase modulation by pseudo-random bit sequence (PRBS) is proposed, which allows generation of localized and stationary DBG at random positions inside the fiber. Such a method opens up new possibilities for signal control, and for optical fiber sensing. Advantages and limitations of the PRBS technique in terms of signal-to-noise ratio are discussed. It is experimentally verified that PRBS generation of DBG is highly attractive for the operations on optical signals where time averaging is possible, such as in the fields of microwave photonics and distributed fiber sensing. On the other hand, it was experimentally verified that realization of the optical buffer for continuous digital data using the PRBS technique is difficult due to the substantial amount of instantaneous noise. A separate chapter of this thesis is devoted to the application of DBG to the field of distributed fiber sensing. A DBG based distributed temperature and stress sensor in polarization maintaining fibers has been realized, showing 1 cm spatial resolution, which is - to the best of our knowledge - first time, demonstrated using a pulsed-based Brillouin sensor. The demonstration is performed in a 20 m fiber, but the sensing range of the proposed DBG based sensor can be expanded up to several hundred meters. In addition, it is shown that the demonstrated spatial resolution is not the ultimate limit, and a finer spatial resolution can be achieved by simply reducing the probe pulse duration. Another Brillouin distributed temperature and stress sensor is proposed based on the DBG principle. Here DBG is generated by modulating the phases of a continuous Brillouin pump and probe waves by common PRBS with unipolar encoding and modulation speeds faster than phonon lifetime. By choosing the proper parameters of PRBS, only one localized and permanent dynamic Brillouin grating is placed inside the sensing fiber. Distributed measurements of temperature or strain are accomplished by scanning the position of the DBG inside the fiber and determining the Brillouin resonance condition at each position. Using this technique, a random access sensor with 1 cm spatial resolution is demonstrated over a 40 m standard single mode fiber. In the proposed scheme the spatial resolution can be easily changed by changing the bit duration of the PRBS, while the measurement range is controlled by the length of the PRBS sequence.

EPFL2013

Brillouin Distributed Fibre Sensing: State of the Art and Perspectives

Luc Thévenaz

Developed societies require more and more information for their safety and for their economic development. Recent disasters due to landslides, fires in tunnels and collapses of bridges are a source of serious concern on the part of the public which seeks for more safety and for an efficient prevention of these frequent recurrences of dangers. Sensing in adverse environment and extreme conditions requires dedicated techniques. Quite recent technologies may offer novel valuable solutions and give rise to a strenuous research effort. Optical technologies are an essential actor owing to their tremendous capability to transmit and process a high density of information. Optical fibre is a key component for these technologies and its potential for optical processing and for collecting information as sensing probe is still widely unexplored. The development and the popularity of telecommunications have entirely screened the fact that optical fibres may be efficiently used for sensing purposes. In this paper we present applications of a novel technique using optical fibres to monitor large structures for safety purpose. The fibre is used as sensing element and can provide distributed measurements of quantities like temperature or strain. In other words a value of temperature and/or strain can be obtained for any point along the fibre. This is made possible by using a nonlinear optical effect in the fibre called Stimulated Brillouin Scattering (SBS). Optical fibre sensors based on stimulated Brillouin scattering have now clearly demonstrated their excellent capability for long-range distributed strain and temperature measurements. The Brillouin interaction causes the coupling between optical and acoustic waves when a resonance condition is fulfilled. It turns out that this resonance condition is strain and temperature-dependent, so that determining the resonance frequency directly provides a measure of temperature or strain. The resonance frequency is an intrinsic property of the material that may be observed in any silica fibre. This is very attractive since the bare fibre itself acts as sensing element without any special processing or preparation on the fibre. Standard optical cables may thus be used, resulting in a low-cost sensing element that may be left in the structure. Since the optical effect only depends on the fibre material, it is absolutely stable in time and independent of the instrument. Different measurements performed over a long-term period are thus fully comparable. The latest developments in this class of sensors will be shown, such as the possibility to measure with a spatial resolution of 10 cm and below, while preserving the full accuracy on the determination of temperature and strain. Illustrative examples of site measurements will be presented and future prospects will be discussed.

Materials Research Society of Singapore2009