Distributed optical fibre sensor for sensing stress state

Abstract

A distributed optical fiber sensor of dynamic stress state comprises: an optical assembly configured to generate a series of optical pulses; an optical fiber of optical length L; an optical system configured to: inject through the first end at least the series of optical pulses; receive at the level of the end at least one series of output optical pulses, arising from the input pulses after propagation and retro-propagation in the fiber; generate at least one continuous reference beam or reference optical pulses on the basis of the optical assembly or of output optical pulses; produce a series of interference zones corresponding to the interference between the reference beam or a reference pulse and a signal optical pulse arising from an output optical pulse; a holographic detector comprising: a liquid-crystal light valve, the valve disposed so that it at least partially covers the interference zones, and producing holograms on the basis of the interference zones; at least one optical detector configured to detect output optical signals diffracted by the holograms.

Claims

1. A distributed optical fiber sensor of dynamic stress state, said sensor comprising: an optical assembly comprising: at least one laser (SL.sub.1) emitting at a wavelength λ, said optical assembly being configured to generate a series of optical pulses (I.sub.pi); an optical fiber (FO) exhibiting a first end (E.sub.x1) and a second end (E.sub.x2) and of optical length L; an optical system configured to: Inject, via a circulator, through said first end (E.sub.x1) at least said series of optical pulses (I.sub.pi); receive at the level of said end (E.sub.x1) at least one series of output optical pulses, an output optical pulse (I.sub.psi) being a carrier of the intensity of reflection of said ends of the fiber and of the backscattered intensity along said fiber; create, by a coupler, two optical pathways carrying output optical pulses; delay, by means for delaying, said output optical pulses on one of said pathways, to create a signal pathway carrying signal optical pulses and a reference pathway carrying reference optical pulses (I.sub.pri), said delaying means introducing an additional length to be traversed ΔL, with ΔL/2 the length of a sensitive zone defined between a position A.sub.i and a position B.sub.i at the level of said fiber and referenced from said first end; and produce a series of interference zones corresponding to the interference between said reference pulse (I.sub.pri) and a signal optical pulse (I.sub.psiS) arising from an output optical pulse (I.sub.psi); a second acousto-optical modulator (MAO.sub.2) making it possible to select gates of a duration equal to 2ΔT/c with c the speed of light in vacuum, so as to allow only backscattered waves originating from a sensitive zone of said fiber at one and the same time to interfere, the pulses being separated by a duration t.sub.R, such that t.sub.R>2L/c, the duration of said pulses t.sub.p being t.sub.p>ΔL/c and t.sub.R>t.sub.off with t.sub.off, the response time of the liquid crystals; and a holographic detector (HD) comprising: a liquid-crystal light valve (LCLV) comprising a liquid crystal layer disposed between two substrates, one of the substrates comprising a photoconductor material for said emission wavelength (λ), said valve (LCLV) being disposed so that it at least partially covers said interference zones, said valve being configured to produce holograms on the basis of said interference zones; at least one optical detector (PD) configured to detect output optical signals diffracted by said holograms; and a digital processing unit situated at the optical detector output to analyze the various probed active zones.

2. The distributed optical fiber sensor of claim 1, wherein the optical assembly comprising at least one laser emitting at a wavelength λ comprises a first acousto-optical modulator (MAO.sub.1) for generating optical pulses.

3. The distributed optical fiber sensor of claim 1, wherein said fiber is single-mode.

4. The distributed optical fiber sensor of claim 1, wherein said fiber is multimode.

5. The distributed optical fiber sensor of claim 1, wherein the emission wavelength of the optical assembly is equal to 1.5 μm.

6. A distributed optical fiber sensor of dynamic stress state, said sensor comprising: an optical assembly comprising: at least one laser (SL.sub.1) emitting at a wavelength k, said optical assembly being configured to generate a series of optical pulses (I.sub.pi); an optical fiber (FO) exhibiting a first end (E.sub.x1) and a second end (E.sub.x2) and of optical length L; an optical system configured to: inject, via a circulator, through said first end (E.sub.x1) at least said series of optical pulses (I.sub.pi); receive at the level of said end (E.sub.x1) at least one series of output optical pulses, an output optical pulse (I.sub.psi) being a carrier of the intensity of reflection of said ends of the fiber and of the backscattered intensity along said fiber; create, by a coupler, two optical pathways carrying output optical pulses; delay, by means for delaying, said output optical pulses on one of said pathways, to create a signal pathway carrying signal optical pulses and a reference pathway carrying reference optical pulses (I.sub.pri), said delaying means introducing an additional length to be traversed ΔL, with ΔL/2 the length of a sensitive zone defined between a position A.sub.i and a position B.sub.i at the level of said fiber and referenced from said first end; and produce a series of interference zones corresponding to the interference between said reference pulse (I.sub.pri) and a signal optical pulse (I.sub.psiS) arising from an output optical pulse (I.sub.psi); a second acousto-optical modulator (MAO.sub.2) making it possible to select gates of a duration equal to t.sub.R−(t.sub.p+ΔT/c), with c the speed of light in vacuum, said gates making it possible to filter the intensities of reflection of said ends of the fiber and to preserve a part of the backscattered intensity along said fiber, the pulses being separated by a duration t.sub.R, such that t.sub.R>2L/c, the duration of said pulses t.sub.p being t.sub.p>ΔT/c and t.sub.R<t.sub.off with t.sub.off, the response time of the liquid crystals; said gates making it possible to filter the intensities of reflection of said ends of the fiber and to preserve a part of the backscattered intensity along said fiber; and a holographic detector (HD) comprising: a liquid-crystal light valve (LCLV) comprising a liquid crystal layer disposed between two substrates, one of the substrates comprising a photoconductor material for said emission wavelength (λ), said valve (LCLV) being disposed so that it at least partially covers said interference zones, said valve being configured to produce holograms on the basis of said interference zones; at least one optical detector (PD) configured to detect output optical signals diffracted by said holograms; and a digital processing unit situated at the optical detector output to analyze the various probed active zones.

7. The distributed optical fiber sensor of claim 6, wherein the optical assembly comprising at least one laser emitting at a wavelength λ comprises a first acousto-optical modulator (MAO.sub.1) for generating optical pulses.

8. The distributed optical fiber sensor of claim 6, wherein said fiber is single-mode.

9. The distributed optical fiber sensor of claim 6, wherein said fiber is multimode.

10. The distributed optical fiber sensor of claim 6, wherein the emission wavelength of the optical assembly is equal to 1.5 μm.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The invention will be better understood and other advantages will become apparent on reading the nonlimiting description which follows and by virtue of the appended figures among which:

(2) FIG. 1 illustrates the recombining of the reference and signal signals on a liquid-crystal light valve thus forming an intensity array, in a distributed fiber sensor of the invention;

(3) FIG. 2 illustrates a first exemplary embodiment of a distributed fiber sensor according to the invention;

(4) FIGS. 3a and 3b relate to a second exemplary embodiment of a distributed fiber sensor according to the invention;

(5) FIGS. 4a to 4c relate to a third exemplary embodiment of a distributed fiber sensor according to the invention.

DETAILED DESCRIPTION

(6) The distributed optical fiber (fiber which is uniform over its entire length) sensor of the present invention makes it possible to exploit the principle of phase demodulation with a distributed measurement and exhibits the following main advantages by reason of the adaptive interferometer that it integrates and which are in particular: insensitivity to slow phase variations of the waves which interfere (therefore low-frequency noise filtering); the capacity to demodulate a complex wavefront (for example arising from a speckle originating from a multimode fiber, thereby making it possible to use a multimode fiber as sensor, which, with respect to a single-mode fiber, and for the same sensitive zone length, gives a sensitivity gain related to the numbers of modes excited).

(7) The sensor of the present invention comprises at least one coherent optical source, and means for generating two optical waves: a reference wave E.sub.R at the frequency ω.sub.R and a signal wave E.sub.s at the frequency ω.sub.s, which is injected into the optical fiber and analyzed at the fiber output after propagation and retro-propagation in said fiber.

(8) Each mode of the signal wave experiences the phase disturbances integrated along the optical fiber. The reference wave and the signal wave are recombined on a liquid-crystal light valve thus forming an intensity array as illustrated in FIG. 1, which depicts the photoconductor PC and the liquid-crystal cell CL, a voltage V.sub.0 being applied between two electrodes El.sub.1 and El.sub.2.

(9) The light valve behaving as an optical Kerr effect medium, the intensity array is transferred to a phase hologram, the duration of recording being the response time of the liquid crystal. Consequently, the hologram accommodates all phase disturbances that are slow relative to its recording time, being re-inscribed continuously as a function of the slow modifications of the interference pattern. The liquid crystals have a response time of the order of a some hundred milliseconds for thicknesses of the order of some ten micrometers. Having regard to these characteristic dimensions, the diffraction takes place in a Raman-Nath regime. It follows from this that the reference and signal waves will diffract on the phase hologram, inducing several diffracted orders E.sub.1, E.sub.2, E.sub.−1.

(10) To illustrate this phenomenon, the Applicants have considered the order diffracted in the direction of the reference wave. The resulting wave after the light valve consists of the transmitted part of the reference wave and of the diffracted part of the signal wave. These two waves have the same wavefront for the phase variations whose characteristic time is greater than the response time of the liquid crystals. The beating of these two optical signals on a photodiode (which converts the phase modulations into intensity modulation) therefore allows the demodulation of the phase disturbance while circumventing the slow disturbances. Moreover, the multimode character makes it possible to effect an average over the whole set of propagation modes and to increase, ultimately, the signal-to-noise ratio. The analytical calculation presented hereinbelow makes it possible to demonstrate this principle.

(11) The Applicants have undertaken the analytical calculation of the gain in sensitivity and have evaluated the power diffracted in the direction of the reference.

(12) To do this, they have studied the phenomenon of self-diffraction in a liquid-crystal light valve (LCLV) between a reference wave E.sub.R and a wave arising from a multimode optical fiber E.sub.S. The signal wave E.sub.S decomposes into the sum of the modes guided by the optical fiber. These are all phase-modulated and the Applicants have more particularly concerned themselves with the phase-amplitude conversion. The results presented hereinafter make the assumption that the modes are mutually orthogonal and that they are all polarized along the director axis of the liquid crystals.

(13) The reference wave may be written:

(14) $\begin{array}{cc}{E}_R=A_R{e}^{j{\stackrel{->}{k}}_R.Math.\stackrel{->}{r}}+c.c.& \left(1.1\right)\end{array}$

(15) With c.c: complex conjugate corresponding to the same term as the first term in the sum by replacing j by −j

(16) The signal wave may be written:

(17) $\begin{array}{cc}{E}_S=\underset{m=1}{\overset{M}{.Math.}}{E}_m+c.c.=\underset{m=1}{\overset{M}{.Math.}}{A}_m{e}^{j\left({\stackrel{->}{\beta }}_m.Math.\stackrel{->}{r}+{\varphi }_m+\Delta {\varphi }_m\sin \left(\Omega t\right)\right)}+c.c.& \left(1.2\right)\end{array}$
Where M is the number of mode, m is the index of the mode considered, Φ.sub.m is a relative phase shift between the modes and ΔΦ.sub.m is the amplitude of the phase modulation at the frequency Ω.

(18) Under steady conditions, the refractive index in the valve is sensitive only to slow variations relative to its response time τ. The latter takes the form:
n=n.sub.0+n.sub.2I.sub.LF  (1.3)
where I.sub.LF is the low-frequency contribution of the intensity array between E.sub.R and E.sub.S:
I.sub.LF=|E.sub.R+E.sub.S|.sub.LF.sup.2  (1.4)

(19) Consequently, it is necessary to determine the LF contribution of the signal wave. The Jacobi-Anger identity is accordingly used. The expression for the optical field for each mode m may be written:

(20) $\begin{array}{cc}{E}_m=A_m{e}^{j\left({\stackrel{->}{\beta }}_m.Math.\stackrel{->}{r}+{\varphi }_m\right)}\left[J_0\left({\mathrm{\Delta \varphi }}_m\right)+\underset{k\ne 0}{\overset{K}{.Math.}}{J}_k\left({\mathrm{\Delta \varphi }}_m\right)e^{\mathrm{jk}\Omega t}\right]& \left(1.5\right)\end{array}$
where the functions J.sub.k(x) are the Bessel functions of the 1.sup.st kind of order k.

(21) In conclusion, by feeding equation (3.5) into equation (3.4), it is possible to show that the low-frequency intensity of the intensity array becomes:

(22) $\begin{array}{cc}{I}_{\mathrm{LF}}={.Math.E_R.Math.}^2+\underset{m=1}{\overset{M}{.Math.}}{.Math.E_m.Math.}^2+2\underset{m=1}{\overset{M}{.Math.}}{A}_m{A}_R{J}_0\left({\mathrm{\Delta \varphi }}_m\right)\cos \left(\left({\stackrel{->}{\beta }}_m-{\stackrel{->}{k}}_R\right).Math.\stackrel{->}{r}+{\varphi }_m\right)& \left(1.6\right)\end{array}$

(23) Having regard to the characteristic dimensions of the LCLV liquid-crystal valve, the diffraction operates in the Raman-Nath regime. The optical field at output may then be written as the product of the incident optical field with the transmission coefficient of the valve:
E.sub.out=TE.sub.in=e.sup.jnk.sup.0.sup.dE.sub.in  (1.7)
with:
E.sub.in=E.sub.S+E.sub.R  (1.8)

(24) By putting:

(25) $\begin{array}{cc}\{\begin{array}{c}\gamma =k_{0}d\left[n_0+n_2{.Math.E_R.Math.}^2+n_2\underset{m=1}{\overset{M}{.Math.}}{.Math.E_m.Math.}^2\right]\\ {\mathrm{��}}_m=2k_0{\mathrm{dn}}_2{J}_0\left({\mathrm{\Delta \varphi }}_m\right)A_m{A}_R\end{array}& \left(1.9\right)\end{array}$

(26) The transmission coefficient can then be cast into the form:

(27) $\begin{array}{cc}T=e^{j\gamma }\underset{m=1}{\overset{M}{.Math.}}{e}^{j{\mathrm{��}}_m\cos \left(\left({\stackrel{->}{\beta }}_m-{\stackrel{->}{k}}_R\right).Math.\stackrel{->}{r}+{\varphi }_m\right)}& \left(1.10\right)\end{array}$

(28) By using the Jacobi-Anger identity, it is possible to express the field diffracted in the direction of the reference wave in the form:

(29) $\begin{array}{cc}{E}_d=\left\{A_R\underset{m=1}{\overset{M}{.Math.}}{J}_0\left({\mathrm{��}}_m\right)+i\underset{m=1}{\overset{M}{.Math.}}{J}_1\left({\mathrm{��}}_m\right)\left[\underset{k=1k\ne m}{\overset{K}{.Math.}}{J}_0\left({\mathrm{��}}_k\right)\right]e^{j{\mathrm{\Delta \varphi }}_m\sin \left(\Omega t\right)}\right\}{e}^{j\left({\stackrel{->}{k}}_R.Math.\stackrel{->}{r}+\gamma \right)}& \left(1.11\right)\end{array}$

(30) In order to establish a first trend in the behavior of such a device, it is possible to make the reasonable assumption that |X.sub.m|<<1.

(31) This implies that

(32) $J_0\left({\mathrm{��}}_m\right)\approx 1\mathrm{and}{J}_1\left({\mathrm{��}}_m\right)\approx \frac{{\mathrm{��}}_m}{2}.$

(33) Equation (1.11) then reduces to:

(34) $\begin{array}{cc}{E}_d=\left\{A_R+i\underset{m=1}{\overset{M}{.Math.}}\frac{{\mathrm{��}}_m}{2}{e}^{j{\mathrm{\Delta \varphi }}_m\sin \left(\Omega t\right)}\right\}{e}^{j\left({\stackrel{->}{k}}_R.Math.\stackrel{->}{r}+\gamma \right)}& \left(1.12\right)\end{array}$

(35) The intensity detected in the direction of the reference wave is thus equal to:

(36) 0$\begin{array}{cc}{.Math.E_d.Math.}^2=I_R-2I_R{k}_{0}d.Math.n_2.Math.\underset{m=1}{\overset{M}{.Math.}}{J}_0\left({\mathrm{\Delta \varphi }}_m\right)I_m\sin \left[{\mathrm{\Delta \varphi }}_m\sin \left(\Omega t\right)\right]& \left(1.13\right)\end{array}$

(37) Moreover, sin[Δϕ.sub.m sin(Ωt)]≈2J.sub.1(Δϕ.sub.m) sin(Ωt) i.e.:

(38) $\begin{array}{cc}{.Math.E_d.Math.}^2=I_R-4I_R{k}_{0}d.Math.n_2.Math.\left[\underset{m=1}{\overset{M}{.Math.}}{J}_0\left({\mathrm{\Delta \varphi }}_m\right)J_1\left({\mathrm{\Delta \varphi }}_m\right)I_m\right]\sin \left(\Omega t\right)& \left(1.14\right)\end{array}$

(39) For weak phase modulations, that is to say Δϕ.sub.m<<1, the above expression can be reduced to:

(40) $\begin{array}{cc}{.Math.E_d.Math.}^2=I_R-2I_R{k}_{0}d.Math.n_2.Math.\left[\underset{m=1}{\overset{M}{.Math.}}{\mathrm{\Delta \varphi }}_m{I}_m\right]\sin \left(\Omega t\right)& \left(1.15\right)\end{array}$

(41) If moreover, it is considered that each mode transports the same intensity I.sub.0, equation (1.15) becomes:

(42) $\begin{array}{cc}{.Math.E_d.Math.}^2=I_R-2I_R{k}_{0}d.Math.n_2.Math.I_0\left[\underset{m=1}{\overset{M}{.Math.}}{\mathrm{\Delta \varphi }}_m\right]\sin \left(\Omega t\right)& \left(1.16\right)\end{array}$

(43) Consequently, the contribution of each mode can be measured in a coherent manner. The power of the signal detected through the dynamic hologram is proportional to Σ.sub.m=1.sup.MΔϕ.sub.m.

(44) The Applicants have also estimated the detection sensitivity: To do this, they have assumed that the limiting noise of the system is an optical noise. The phase shift associated with a disturbance of the optical fiber is a statistical process.

(45) In the case of a conventional quadrature interferometer, the phase modulation is converted linearly into optical power modulation in the form:
P.sub.SMF(Δϕ)=αΔϕ  (1.17)

(46) The associated variance then takes the form:
σ.sub.SMF(Δϕ)=ασ.sub.Δϕ  (1.18)

(47) In the case of an adaptive interferometer with a multimode fiber, the modulated power may be written in accordance with (1.16):

(48) $\begin{array}{cc}{P}_{\mathrm{MMF}}\left(\mathrm{\Delta \varphi }\right)=\frac{\alpha }{M}\underset{m=1}{\overset{M}{.Math.}}{\mathrm{\Delta \varphi }}_m& \left(1.17\right)\end{array}$

(49) The factor 1/M signifies that the intensity is divided spatially over the set of modes.

(50) Consequently, the variance of the signal detected with a fiber having M modes may be written:

(51) $\begin{array}{cc}{\sigma }_{\mathrm{SMF}}\left(\mathrm{\Delta \varphi }\right)=\frac{\alpha }{\sqrt{M}}{\sigma }_{\mathrm{\Delta \varphi }}& \left(1.18\right)\end{array}$

(52) It is possible to conclude that the variance of the signal detected with a multimode fiber with respect to that obtained with a single-mode fiber is reduced by a factor √{square root over (M)}. Consequently, the signal-to-noise ratio (SNR) for a multimode fiber increases with √{square root over (M)}.

(53) The SNR ratio can be subsequently increased by differential detection, for example with two balanced photodiodes, on the diffracted waves E.sub.1 and E.sub.0 (illustrated in FIG. 1) thus making it possible to reduce the continuous component of the measured intensity.

First Exemplary Embodiment of Distributed Optical Fiber Sensor Making it Possible to Locate a Disturbance

(54) According to this first exemplary configuration, the distributed optical fiber sensor comprises a laser source SL.sub.1, an acousto-optical modulator MAO.sub.1 generating optical pulses I.sub.pi emitted every t.sub.R and of pulse duration t.sub.p, and an optical fiber FO of length L. A series of luminous pulses I.sub.pi of duration t.sub.p are thus injected into said optical fiber via a first end E.sub.x1, propagate along said optical fiber, are reflected at the level of the second end E.sub.x2, and then backscattered along said fiber, they correspond to the output optical pulses I.sub.psi which are utilized and carry information, as is illustrated in FIG. 2. According to this figure, various so-called sensitive zones are probed, and represented between the positions A.sub.1 and B.sub.1, and then between the positions A.sub.i and B.sub.i at the level of the optical fiber.

(55) Thus, a pulse introduced into the optical fiber FO of length L, via a circulator C gives rise to a back-scattered wave for the entire duration of the return journey of the pulse in the fiber, i.e. for a duration of 2×L/c. FIG. 2 depicts the luminous intensity of an output pulse I.sub.psi generated by the backscattering R.sub.d, the reflection R.sub.e by the first end and by the reflection R.sub.s by the second end.

(56) A second acousto-optical modulator MAO.sub.2 is provided at the circulator C output, as well as a coupler CPL so as to divide the output pulses on two pathways. Means Dm making it possible to introduce a delay of length ΔL corresponding to the return journey time of the light in the sensitive zone of length ΔL/2 are inserted on one of the two pathways. This delay makes it possible to produce on the LCLV valve an interference between a wave and itself shifted in time.

(57) This temporal shift corresponds to a distance shift of length ΔL/2 in the sensor. The wave which passes down the delayed pathway originates from the position A.sub.i in the fiber, the wave which passes down the undelayed pathway originates from the position B.sub.i in the fiber, which is situated ΔL/2 further on in the sensor.

(58) The interference of the backscatterings originating from the positions A.sub.i and from B.sub.i gives the phase difference between the backscattering coming from A.sub.i and the backscattering coming from B.sub.i. The phase involved is indeed that experienced by the wave arising from the position B.sub.i over the length ΔL/2. The bigger the sensitive zone length, the more decorrelated the interference patterns corresponding to the N sensitive zones in the sensor.

(59) The superposition of the N interference patterns does not make it possible to inscribe a grating in the LCLV. Indeed, the relative phase of these interference patterns being random, their superposition decreases the contrast and scrambles the fringes. This is why an acousto-optical modulator MAO.sub.2 is inserted before the separation of the backscattering into two pathways, ensured by a coupler CPL. It makes it possible to open a gate of duration 2×ΔL so as to allow only the waves originating from one sensitive zone at a time to interfere.

(60) FIG. 2 thus also depicts the interrogation via the opening of two gates G.sub.1 and G.sub.i shifted in time and relating to various emission optical pulses I.sub.pi.

(61) In this case, one of the two pathways serves as reference with reference output optical pulses I.sub.pri and the other pathway serves as signal pathway carrying signal optical pulses I.sub.psiS, all arising from the output optical pulses I.sub.psi, the optical waves of the two pathways interfering at the level of the LCLV light valve.

(62) This architecture makes it possible to locate along the fiber the phase disturbance induced by the physical quantity to be measured with a spatial resolution ΔL/2. It is measured by frequency analysis of the electrical signal delivered by the photodiode PD.

(63) The maximum rate of interrogation of two different sensitive zones of the sensor is limited by the response time “off” of the liquid crystals: i.e. t.sub.off, the return time of the liquid crystals in the light valve, thereby implying that the rate of two successive pulses, which is defined by the parameter t.sub.R, must be greater than the parameter t.sub.off.

(64) One ought to wait for the liquid crystals involved in inscribing the interference pattern of the first sensitive zone to be available again.

Second Exemplary Embodiment of Distributed Optical Fiber Sensor Making it Possible to Locate a Disturbance

(65) The second exemplary distributed fiber sensor of the invention comprises an architecture the aim of which is to allow the reading of the disturbance on all the sensitive zones of the sensor in a short time. The issue here is to be able to reconstruct a spatial array of virtual sensors (see definition hereinabove) so as to be able to construct for example an acoustic antenna with “electronically” reconfigurable spacing, this presenting a decisive advantage with respect to the solutions with dispersed sensors.

(66) Accordingly it is necessary for a “mean” grating, called a “static grating”, to be inscribed in the LCLV. This grating is obtained by interference of the reflection of an input pulse I.sub.pi with itself delayed by ΔL on the fiber extremity (fiber end) connector. The phase is therefore accumulated along the entire fiber. The slow phase variations along the entire fiber will modify the interference pattern making it possible to inscribe this mean grating. It is then possible to use this mean grating as diffraction grating for another light wave. This third-party wave is the backscattering of the pulse in the fiber.

(67) A luminous pulse of duration t.sub.p is therefore injected into the fiber and is back-scattered in the latter. It is also reflected at the end of the fiber. It is obtained for example by a laser source SL.sub.1 followed by an acousto-optical modulator MAO.sub.1. This pulse gives rise to a back-scattered wave for the entire duration of the return journey of the pulse in the fiber, i.e. for 2×L/c. The back-scattered and reflected signal is separated into two pathways, via a circulator C and by virtue of a coupler CPL.

(68) As in the first exemplary sensor, a delay of length ΔL corresponding to the return journey time of the light in the sensitive zone of length ΔL/2 is inserted into one of the two pathways by means Dm. This delay makes it possible to produce on the LCLV an interference between a wave and itself shifted in time. This temporal shift corresponds to a distance shift of length ΔL/2 in the sensor. This temporal shift fixes the minimum duration of the input pulse t.sub.p>ΔL/c. This configuration is illustrated in FIGS. 3a and 3b.

(69) Every t.sub.r, a new pulse is dispatched so that the mean grating does not wane with t.sub.r>2×L/c and t.sub.r<t.sub.off where t.sub.off is the return time for the liquid crystals to regain their initial state.

(70) The maximum length of the sensor is therefore related to t.sub.off by:
L.sub.max<c×t.sub.off/2.

(71) A photodiode PD is placed on a diffraction order. The diffraction of the back-scattered wave on the mean grating is detected on the photodiode for the entire duration of the backscattering. So as not to saturate the detector, an acousto-optical modulator MAO.sub.2 is placed before the photodiode and makes it possible to cut.sub.off the waves reflected by the ends of the fiber corresponding to the input and output connectors of the fiber. In contradistinction to the aforementioned first exemplary distributed fiber sensor, the electrical signal delivered by the photodiode contains the information regarding phase shift over the entire length of the sensitive fiber with a resolution ΔL/2 at a given instant.

(72) It is the pulse repetition frequency analyzed by a processing unit UN which allows the frequency analysis of the signal with as limit condition on the sampling f.sub.ac<f.sub.rep/2, f.sub.ac being the highest disturbance signal frequency that it is desired to detect in accordance with the Nyquist-Shannon sampling theorem which indicates that when sampling at the frequency Fe, only the frequencies below Fe/2 can be transmitted without information loss.

(73) FIG. 3b portrays the temporal superposition of an output intensity I.sub.psi arising from the position A.sub.i, that arising from the position B.sub.i in the fiber, which are denoted I.sub.Ai and I.sub.Bi, the inscribing of the static grating, the generation of temporal analysis gates and the backscattered intensity analyzed and which are illustrated more precisely by: (1): the curve I.sub.Ai (t) (2): the curve I.sub.Bi (t) (3): the curve I.sub.grating LCLV (4): the curve Gate(t), making it possible to filter the reflections of the ends of the fiber (5): the curve I.sub.PD (t), intensity relating to the backscattering by virtue of the use of a temporal gate at the level of the component MAO.sub.2.

(74) Third exemplary embodiment of distributed optical fiber sensor making it possible to locate a disturbance based on a Brillouin dynamic grating:

(75) This exemplary embodiment comprises a distributed architecture based on a dynamic Brillouin grating as movable reflector and an optical interrogation wave comprising a series of optical pulses.

(76) The Brillouin grating, generated by the interaction between two optical pulses, makes it possible to define a sensitive optical fiber portion. In this case, one is not concerned with the frequency aspect of the stimulated Brillouin interaction but solely with the reflection coefficient of the dynamic grating. A probe wave then makes it possible to probe the optical fiber.

(77) The proposed architecture is shown diagrammatically in FIGS. 4a and 4b.

(78) FIG. 4a illustrates the part of the means that are necessary for the step of writing the Brillouin grating R.sub.Bi. Two pulses arising from one and the same laser SL.sub.2 of optical frequency ω.sub.p but which are shifted temporally and spectrally inscribe a Bragg grating by Stimulated Brillouin Scattering R.sub.Bi. Thus on the basis of a single laser source SL.sub.2, there are provided after division of the laser beam, two acousto-optical modulators MAG.sub.3 and MAO.sub.4 and means of frequency shifting Df, which may be electro-optical, so as to generate two series of optical pulses at the optical frequencies ω.sub.p and ω.sub.s (the wave at the frequency ω.sub.p is represented by a thin line, the wave at the frequency ω.sub.s being represented by a thick line). Typically, the maximum length of the optical fiber L is limited to half the coherence length L.sub.coh of the laser used.

(79) The Stokes wave is shifted toward the low frequencies of ω.sub.B, the Brillouin frequency, corresponding to the Doppler effect: reflection of the pump on a movable grating. This grating, equivalent to a Bragg grating (due to the electrostriction effect between the pump wave and the Stokes wave in silica), propagates in the same direction as the pump at the speed of sound c.sub.ac in the fiber ω.sub.B=2nc.sub.ac/λ.

(80) The duration of the pulses determines the length of the grating. The grating is successively inscribed at various positions Z.sub.r in the fiber.

(81) Its position is controlled (that is to say the zone in the fiber where the reflection of the pulse at ωs crosses the pulse at ωp) by the time interval Δt between the two pulses. In practice, it is proposed to use a reflecting treatment M at the end of the fiber so as to obtain the reflection of the wave at ωs making it possible to stimulate the Brillouin scattering.

(82) Thus, on the first pathway, the first acousto-optical modulator MOA.sub.3 is used to obtain the pump pulses. On the second pathway, the frequency of the laser is shifted by a value corresponding to w.sub.B(ω.sub.s=ω.sub.p−ω.sub.s) i.e. about 10 GHz in the optical fibers, and then another acousto-optical modulator MAO.sub.4 is used to obtain the Stokes pulses.

(83) FIG. 4b illustrates the part of the means that are necessary for the step of reading the phase which consists in using this Brillouin grating R.sub.Bi as a Bragg mirror.

(84) A third “probe” wave arising from the optical assembly comprising the laser SL.sub.1 is injected into the fiber at a frequency ω.sub.s and makes a return journey between the input of the fiber and the Brillouin grating R.sub.Bi. It accumulates a phase shift on this return journey.

(85) This phase shift is the signal of interest. A hologram is caused on the LCLV valve between the return from the probe and a reference originating from the same laser. An optical assembly is used comprising a laser SL.sub.1 which emits a laser beam, divided so as to generate a laser beam F.sub.r, the other part of said beam being introduced into an acousto-optical modulator to generate a series of pulses I.sub.pi at the optical frequency ω.sub.s so as to carry out the reading step. Typically, the lifetime of the grating thus inscribed is defined by the lifetime of the acoustic phonons in the material of the optical fiber which may be in a conventional manner, silica, i.e. about 10 ns. It is therefore necessary to read the phase in the 10 ns following writing. The pulses I.sub.pi introduced into the optical fiber generate, at fiber output, pulses I.sub.psiS after reflection at the level of the Brillouin grating R.sub.Bi, interfering with the reference beam on the LCLV light valve.

(86) FIG. 4c illustrates the alternation of writing steps and of reading steps, corresponding to variations of inscription of Brillouin gratings R.sub.Bi at various sites in the optical fiber. The position of the Brillouin grating R.sub.Bi in the fiber is adjusted through the temporal shift between two writing pulses at the frequencies ω.sub.p and ω.sub.s.

(87) At each writing-reading cycle, the spectrum obtained at cycle N is subtracted from the spectrum obtained at cycle N+1 so as to have access to the information which occurs in the fiber portion corresponding to the interrogation (writing/reading) by the pulses of cycle N.

(88) The period of the writing-reading process is: T=2L/c=1/f.sub.rep

(89) One obtains the spectrum Si (t, L−Z.sub.r) of the phase disturbance signal at ti=i*T, for the Brillouin reflector R.sub.Bi at the position L−Z.sub.ri. More precisely, Si(t, L−Z.sub.r) is the spectrum of the acoustic signal which modulates the phase on a return journey between the input of the fiber and the Brillouin reflector R.sub.Bi inscribed, therefore over the length 2×L−Z.sub.ri.