GB2303445A

GB2303445A - Pressure sensor with wound optical fibre

Info

Publication number: GB2303445A
Application number: GB9514918A
Authority: GB
Inventors: Denis Mazerolle; Marie-Noolle Charasse; Pesant Jean-Pierre Le; Marc Turpin; Michel Lagier
Original assignee: Thomson CSF SA
Current assignee: Thales SA
Priority date: 1994-02-01
Filing date: 1995-07-20
Publication date: 1997-02-19
Anticipated expiration: 2015-07-20
Also published as: GB2303445B; FR2715730A1; GB9514918D0; FR2715730B1

Description

A PRESSURES SENSOR WITH WOUND OPTICAL FIBRE The present invention relates to a pressure sensor with wound optical fibre, and applies in particular to hydrophones, for example, for towed antennas or streamers or fixed arrays.

The characteristics desired for this type of hydrophone are: small volume, low cost, high sensitivity to pressure, low sensitivity to temperature, operation over a wide static pressure range (for example from 0 to 10 MPa) and multiplexability.

Several types of fibre-optic sensors based on interferopolarimetry are already known.

The first type uses a conventional single-mode optical fibre. The elasto-optical effect in optical fibres has been recognized as an attractive process for producing high-performance hydrophones. When an isotropic pressure is applied to the fibre, the length of the fibre varies, as do its refractive indices, and this induces a variation in the phase of the wave propagating through it. This phase shift is measured in an interferometric assembly, generally with a Mach-Zender interferometer which compares the phase shift created in a reference fibre that is insensitive to the pressure variation or is isolated therefrom, with that created in the fibre subjected to the pressure variation.

Although the effect to be measured is intrinsically small, high sensitivities are possible, in particular if use is made of a large length of fibre and if it is conditioned so as to multiply by a transduction ratio the pressure to which the fibre is actually subjected.

However, such pressure sensors generally exhibit high sensitivity to temperature. Indeed, the refractive index of silica (which constitutes the fibre) varies by approximately 10-5 per degree Celsius and produces, during temperature variations, phase shifts which are very much greater than the phase shifts created by the pressure variations.

This is particularly problematic in the case of hydrophones used in antennas towed at sea, where temperature variations due to marine currents are frequent and can be large.

A method for overcoming the effects of the temperature variations consists in embedding the fibre in a sufficiently large volume of material having low thermal conductivity, in order to slow and average out these effects. This method leads, however, to a significant increase in the volume of the sensor, which is not desirable.

The second type of sensor uses an intrinsically birefringent single-mode optical fibre. By definition, a birefringent single-mode optical fibre makes it possible to propagate two eigenmodes with orthogonal linear polarization, one along a so-called slow axis and the other along a so-called fast axis.

The differential propagation delay between the two modes is measured. A pressure variation is thus directly measured by interference between the mode along the fast axis and the mode along the slow axis, without it being necessary to resort to a reference fibre.

Most known birefringent fibres are obtained by breaking the axial symmetry of the fibre, by deformation of the optical core/cladding structure (for example fibres of oval section) or by adding stress zones (for example fibres known by the term "bow-tie" or by the term "panda') These fibres have a very small sensitivity SP to pressure (of the order of 5 yrad.Pa~l per metre of fibre) and a high sensitivity ST to temperature (of the order of 50 Ccrad,K'1 per metre of fibre), i.e. a ratio SP/ST of the order of 0.1 K.Pa~l, which does not make it possible to obtain a very high-performance sensor.

A side-hole birefringent single-mode optical fibre (known by the term SEF) is also known, which has an improved ratio SP/ST of the order of 15 K,Pa'l. Instead of a solid structure, this fibre has two holes on each side of the core which guides the light. Although its pressure sensitivity SP (of the order of 70 rad.Pa1 per metre of fibre) is indeed better than that of solidstructure birefringent fibres, production of a compact sensor having a pressure sensitivity of the order of 10 Ctrad.Pa'l requires a great length of fibre, of the order of 100 metres. In order to obtain the desired compactness, it is necessary to wind the fibre over a small diameter, generally less than 30 millimetres, and it is found that the sensor provides insufficient interferometric contrast.

The present invention aims to overcome these drawbacks. It proposes production of a fibre-optic pressure sensor which is compact, which has good pressure sensitivity, low temperature sensitivity and the best possible interferometric contrast. tk proyided According to a first aspect of the invention there is' a pressure sensor with wound optical fibre, comprising a shell including the optical fibre, this shell is externally subjected to a pressure to be measured and internally contains a medium which is more compliant than the shell, so that the pressure to be measured generates an anisotropic stress tensor in the optical fibre, the stress difference in a cross-section of the fibre serving to determine the pressure to be measured.

The shell is preferably formed by a tubular structure including the optical fibre, this structure having its ends closed by-flanges, the medium which is more compliant than the shell being contained inside the tubular structure.

The tubular structure may be formed by the bare or jacketed optical fibre wound with adjoining turns. The optical fibre may also be wound with spaced turns.

In order to reduce the fragility of the sensor, the wound fibre may be embedded in encapsulation.

In order to increase the transduction ratio of the sensor, provision is made to place the tubular structure in a bellows bearing on the flanges so as to increase the surface area of the flanges.

In order to increase the static pressure strength of the sensor while having a high transduction ratio, it is possible to cover the tubular structure externally with at least one layer of a flexible material. This layer may also be placed inside the tubular structure.

The two variants can also be combined.

A device for mechanically filtering the pressure to be measured may also be provided, which connects the inside of the shell to the outside. This type of sensor withstands higher static pressures and fast pressure variations.

This type of sensor functions with non-birefringent or intrinsically birefringent optical fibres.

A fibre with oval cross-section is particularly beneficial because it is spontaneously oriented in suitable fashion during winding. An advantageous oval fibre is one which includes optical core and cladding having oval cross-section, in which the major dimensions coincide, these major dimensions being perpendicular to that of the cross-section of the fibre.

According to a second aspect of the invention, there is provided a method of manufacture of a fibre with oval cross-section.

other characteristics and advantages of the invention will emerge more clearly in the following description and in the attached figures, in which: Figures la,lb represent longitudinal and transverse sections through a pressure sensor in which the first aspect of the invention is embodied; Figure Ic represents a view of the cross-section of a fibre used in a pressure sensor in which the first aspect of the invention is embodied; Figures 2a,2b represent longitudinal and transverse sections through a first pressure sensor in which the first aspect of the invention is embodied; Figures 3a,3b represent longitudinal and transverse sections through a second pressure sensor in which the first aspect of the invention is embodied; Figures 4a,4b represent longitudinal and transverse sections through a third pressure sensor in which the first aspect of the invention is embodied; ; Figures 5a,5b represent longitudinal and transverse sections through a fourt Invention pressure sensor in which the first aspect of the / is erttbodied; - Figure 6a represents a longitudinal section through a pressure sensor in which the first aspect of the invention is embodied with mechanical pressure filtering; - Figure 6b represents an electrical circuit diagram equivalent to the sensor of Figure 6a; - Figure 6c represents a curve of the transduction ratio 7' of the pressure sensor of Figure 6a as a function of angular frequency or - Figure 7a represents a longitudinal section through a pressure sensor in which the first aspect of the invention is embodied using a fibre with oval cross-section;; - Figure 7b represents the cross-section of an oval fibre such as that used in Figure 7a; - Figure 8 represents a longitudinal section through a pressure sensor in which the first aspect of the invention is embodied using a variant of an optical fibre with oval cross-section; - Figures 9a,9b,9c represent steps in the method of manufacture of an optical fibre with oval crosssection according to that represented in Figure 8.

Figures la and ib respectively represent, in longitudinal and transverse section, a fibre-optic invention pressure sensor in which the first aspect of the/ is embodied; This sensor comprises a shell 1 including the wound optical fibre 2. In these figures, the shell 1 is formed by a tubular structure 3, the two ends of which are closed by flanges 4. This shell 1 internally contains a medium 5 which is more compliant than the shell 1.

The compliance of a system is defined as the ratio of its volume variation to the pressure variation generating this volume variation.

The term compliance is used in the case of a system and the term compressibility is used in the case of a simple material.

In these figures, the tubular structure 3 is formed by the optical fibre 2 wound with adjoining turns.

The fibre 2 is, for example, a conventional non-birefrin gent single-mode fibre. The winding has, in this example, circular section but it is envisagable to give it another section in order, in particular, to reduce the bulk along one direction. The fibre 2 is represented bare, without a protective jacket, but it might have had one as in Figures 2a,2b.

The shell 1 is externally subjected to an isotropic pressure p which is applied over its entire outer surface. This pressure p generates an anisotropic stress tensor (TX, Ty, Tz) in the optical fibre 2. This pressure p is represented by the arrows in the figures.

The axes x,y,z are represented in the figures. The axis x corresponds to the winding axis of the fibre 2, the axis y is normal to the axis x, these two axes define the longitudinal section plane (Figure la). The axis z is normal to the axis y, these two axes define the transverse section plane (Figure lib).

The stress tensor is anisotropic, in particular in the x,y plane. Figure ic represents a cross-section of the fibre 2 in the x,y plane.

In this figure, the normal stresses are represented and referenced Tx along x and Ty along y. The stress Tx along x is greater than the stress Ty along y.

Because of the elasto-optical effect in the material of the fibre, the anisotropy is manifested by an induced birefringence proportional to the stress difference Ty Tx.

A pressure variation is thus manifested by a variation in the birefringence of the fibre, and this variation is measurable directly by interference between the mode along the fast axis and the mode along the slow axis.

The stress difference Ty - Tx may, in absolute value, be very much greater than the surrounding pressure p and the pressure sensor functions as an amplifier.

In Figures la, lb, the tubular structure is formed by the fibre 2 wound with adjoining turns. The fibre 2 is represented bare, that is to say without a protective jacket. It is generally made of silica.

The flanges 4 hermetically close the tubular structure 3. The medium contained inside the shell 1 may be partially vacuum or be produced based on a compressible fluid such as a gas. The flanges 4 will be produced preferably from a material which is more rigid than the material of the fibre.

The stresses Tx, Ty can be calculated to first approximation on the basis of the model of an infinitely long tube. It is assumed that the tubular structure is cylindrical. According to the normal conventions, the stresses are counted negatively for compressive forces and positively for tensile forces.

Tx t -p.b2/(b2 ~ a2) p is the pressure outside the shell a is the internal radius of the tubular structure b is the external radius of the tubular structure.

Ty - p on condition that b2 > > b2 - a2 Ty - Tx - - Tx Ty - Tx 7 . p r.p , equal tob2/(b2-a2), is the transduction ratio.

It is directly proportional to the pressure sensitivity of the sensor. This ratio may reach high values, for example several tens.

Instead of using a tubular structure 3 formed by an optical fibre wound with adjoining turns, use may be made of a tubular structure 3 comprising an optical fibre 2 wound with spaced or adjoining turns, and this fibre can be embedded in encapsulation 6. This encapsulation 6 is made of a material which is less rigid than the material of the optical fibre 2, which is generally silica. The material of the encapsulation 6 is preferably also less rigid than that of the flanges 4. The material of the encapsulation 6 may, for example, be an epoxy resin or a polyurethane with suitably chosen rigidity.

In Figures 2a,2b, the optical fibre 2 represented comprises a protective jacket 7, but it might also have been bare.

This protective jacket 7 is preferably made of a material having rigidity intermediate between that of the material of the encapsulation 6 and that of the material of the optical fibre 2.

The smaller the thickness of the encapsulation 6, the higher the amplification effect.

In order to increase the transduction ratio, it is possible, as represented in Figures 3a and 3b, to place the tubular structure 3 in a bellows 8 which bears on the flanges 4. This type of bellows 8 is highly deformable along the axis x, and hardly or scarcely deformable along the axes y and z. The shell 1 then comprises the tubular structure, the flanges and the bellows.

A space 9 is arranged between the tubular structure 3 and the bellows 8. In the example represented in Figures 3a and 3b, the tubular structure 3 is formed by the bare optical fibre 2, wound with spaced turns, and embedded in encapsulation 6.

The surface area of the flanges 4 has been increased relative to the configurations in Figures la,lb and 2a,2b. In the preceding configurations, the surface area of the flanges 4 corresponded to the cross-sectional area of the tubular structure 3.

The space 9 between the bellows 8 and the tubular structure 3 contains a medium having a higher compliance than that of the shell 1. This compliance may be close to that of the medium contained in the tubular structure 3.

This medium may be at least partially vacuum or a gas, for example.

The transduction ratio of the sensor then becomes: r = c2/(b2 - a2) c is the mean radius of the bellows 8 b and a, respectively, are the external and internal radii of the tubular structure.

A transduction ratio of the same order of magnitude can be obtained with a configuration as represented in Figures 4a and 4b. At least one layer 10 externally covers the tubular structure 3. It is made based on a material having a rigidity less than or equal to that of the material which it covers. Here, it covers the encapsulation 6, but it might also cover the fibre 2 when the latter is wound with adjoining turns. This material can be made, for example, based on silicone or other elastomer. The flanges 4 also bear on the layer 10. This layer 10 protects the tubular structure 3 without making it more rigid.

The flanges 4 have a surface area greater than the cross-sectional area of the tubular structure 3. The shell 1 then comprises the tubular structure, the flanges 4 and the layer 10.

A pressure sensor produced according to this configuration has the advantage of being more resistant to static pressure than those represented in the figures already described.

However, this configuration, just as that represented in Figures 3a and 3b, leads to large bulk.

A less voluminous sensor is represented in Figures 5a and Sb. In this variant, the layer 10 is held inside the tubular structure 3, instead of covering the exterior of the tubular structure 3. It is in contact with the internal surface of the tubular structure 3. In the configuration of Figures 5a and 5b, the tubular structure is formed by the optical fibre 2 wound with adjoining turns and there is no encapsulation. The optical fibre comprises a protective jacket 7.

This configuration has the advantage of being even less fragile to static pressure. The resistance to crushing has been improved relative to the variant of Figures 4a and 4b.

It is, of course, possible to conceive of a configuration in which the tubular structure 3 would be lined both internally and externally with at least one layer 10. It is also possible to conceive of a configuration with at least one of the layers 10 and the bellows. The choice of the material or materials, of the number of layers and of their position makes it possible to obtain the desired transduction ratio and sensitivity to various parameter such as pressure, temperature, etc.

When the pressure sensor according to the invention is used as a hydrophone, it may be subjected to a high static pressure of the order of 10 MPa. Because of the amplifying nature of the sensor, this high static pressure runs the risk of generating excessive stresses in the encapsulation 6 and of causing the tubular structure 3 to be crushed. In order to solve this problem, a solution consists in compensating for the static pressure difference between the inside of the shell 1 and the outside. This solution reduces to performing mechanical filtering. Only frequency components of the pressure which are less than a cut-off frequency are transmitted into the shell 1.

Thus, when the sensor is subjected to a high static pressure or a slow pressure variation, that is to say a variation whose frequency spectrum lies below the cut-off frequency, the pressure is transmitted in large part into the shell, which makes it possible to decrease some of the stresses to which the shell would be subjected if it were not compensated.

On the other hand, in the case of a fast pressure variation, that is to say a variation whose frequency spectrum lies above the cut-off frequency, only a small part of the pressure is transmitted into the shell, and the stresses to which it is subjected remain comparable to those to which it would be subjected if it were not compensated.

Figure 6a represents, in longitudinal section, a pressure sensor comparable to that of Figures 2a and 2b, and provided with compensation. The shell 1 contains a medium which is more compliant than the shell 1. This medium may be a mixture of the fluid in which the pressure sensor is immersed, called the "base fluid', and inclusions of compressible material, such as a foam, based on bubbles. In the case of hydrophones, the base fluid is often oil and is generally almost incompressible. The inside of the shell 1 is connected to the outside via a capillary tube 30. In this example, the shell 1 is comparable to that of Figures 2a and 2b, but the optical fibre is bare instead of having a protective jacket. This capillary tube 30 preferably passes through one of the flanges 4.

Let: - Ce be the compliance of the shell 1 when closed (without the capillary tube 30); - Cf be the compliance of the medium inside the shell 1; The compliances Ce and Cf are electrically equivalent to capacitances.

- R be the equivalent electrical resistance of the capillary tube 30; - pi be the pressure inside the shell 1; - pe be the pressure outside the shell 1.

The pressures pi and pe are electrically equivalent to voltages.

Figure 6b represents an electrical circuit diagram equivalent to the sensor of Figure 6a. The pressure pi is equivalent to the voltage across the terminals of the compliance Cf. The pressure pe is equivalent to the voltage across the terminals of the series pair formed by the compliance Cf and by the combination of the compliance Ce and the resistance R in parallel.

It may be shown that the equivalent resistance R of the capillary tube 30 is: R = 8ir77l/d2 with: - n the dynamic viscosity of the base fluid - 1 the length of the capillary tube - d the internal diameter of the capillary tube.

The transduction ratio 7' of the pressure sensor of Figure 6a is given by the relationship: r'=r.j(pe - pi)/pel 7 is the transduction ratio of the uncompensated sensor, that is to say a sensor whose shell is closed (without the capillary tube 30).

Figure 6c represents the curve of variation in T' as a function of angular frequency . This curve is deduced from the electrical circuit diagram of Figure 6b.

' remaine low when X remains less than an angular frequency oe called the cut-off angular frequency. This corresponds to a low-frequency regime. When Ca, increases and is greater than oc, T' tends to: 'HP = TCf/(Cf+Ce) This corresponds to a high-frequency regime.

T'HF may be very close to T when CfooCe, that is to say when the medium of the inside of the shell is much more compliant than the shell without the capillary tube.

The cut-off angular frequency oc has the value: oc P l/R.(Cf+Ce) This cut-off angular frequency can be adjusted to the desired value by suitably choosing the length 1 and the internal diameter d of the capillary tube.

The act of producing compensation makes it possible to overcome two drawbacks of the sensor according to the invention but which does not have compensation. The compensated sensor is less fragile to static pressure. In addition, the birefringence drifts due to slow pressure variations are limited.

Figure 7b represents, in longitudinal section, an example of a pressure sensor according to the invention with compensation. The optical fibre 2 used is a fibre of polarization-maintaining single-mode type, that is to say an intrinsically birefringent fibre. During winding, it is possible to orientate one of its optical axes along the winding direction, that is to say along the axis x.

In Figure 7a, the optical fibre 2 employed is a fibre with oval cross-section. It is particularly beneficial.

Another type of polarization-maintaining fibre might have been used, for example an SHF fibre. Because of the oval shape of the fibre, it is spontaneously wound with the major dimension of its cross-section parallel to the axis x. Figure 7b represents a cross-sectional view of the fibre used. The slow axis is referenced 12 and the fast axis is referenced 13. Its slow optical axis 12 is directed along this major dimension. Its slow optical axis 12 is thus always substantially parallel to the axis x. The fast axis 13 is directed along the minor dimension of the cross-section.

This fibre is formed by an oval guide core 14 surrounded by an oval guide cladding 15, which is embedded in an oval matrix 16 which is generally made of silica. The matrix 16 may be protected by a protective jacket 17. The core and the cladding are arranged Co- axially and the major dimensions of their cross-sections coincide with that of the fibre.

The optical fibre 2 is wound with spaced turns over a diameter of approximately 20 millimetres. The optical fibre 2 is embedded in encapsulation 6 and the whole forms the tubular structure 3. The encapsulation is of epoxy resin.

The tubular structure 3 is closed by flanges 4 made of plexiglas. The flanges 4 are adhesively bonded to the tubular structure 3.

The tubular structure 3 is immersed in a base fluid 18, for example oil. The inside S of the tubular structure contains this same base fluid. The inside of the tubular structure 3 also comprises a layer 19 of closed-cell foam or syntactic foam of the type constituted by a polyurethane matrix filled with plastic-walled gaseous microspheres.

A mechanical filtering device is provided. It comprises a capillary tube 30 which connects the outside of the tubular structure 3 to the inside.

If the wall of the tubular structure 3 is thin, smaller than one millimetre for example, such a pressure sensor may provide a high transduction factor 7, for example between 10 and 50.

A pressure sensitivity exceeding 20 mrad.Pa'l was obtained with an interferometric contrast close to the theoretical maximum for a length of fibre of approximately 20 metres. This corresponds to a pressure sensitivity SP per metre of fibre of approximately 1mrad.Pa1.

The ratio SP/ST is approximately 20 K.Pa'l. This ratio may be further increased by using a less birefringent fibre. Indeed, the temperature sensitivity ST of a birefringent fibre is generally proportional to its beat length.

An advantageous variant of the sensor in which the invention is embodied uses an optical fibre as represented by its cross-section in Figure 9c. Figure 8 represents such a sensor in longitudinal section. It differs from that of Figure 7a only as regards the optical fibre 2, which is comparable to that of Figure 9c.

This optical fibre still has an oval crosssection. It comprises an oval optical core 23 surrounded by an oval optical cladding 24, itself embedded in an oval matrix 20 optionally protected by a protective jacket 21. The major dimension of the cross-section of the core 23 coincides with the major dimension of the cross-section of the optical cladding 24. The major dimension of the cross-section of the fibre is perpendicular to that of the core 23 and of the optical cladding 24.

The fast axis of the fibre, which corresponds to the minor dimension of the cross-section of the core and of the optical cladding, now corresponds to the major dimension of the cross-section of the fibre. When the fibre is wound, the fast optical axis of the fibre is directed along the winding-axis x.

A pressure sensor using such an optical fibre has at least two advantages. The optical fibre thus wound has a smaller birefringence than that of the fibre used in Figure 7a. The ratio SP/ST has thus been increased without introducing inhomogeneity linked with the orientation of the fibre in the winding. The oval shape allows accurate positioning of the fibre during winding.

In the event that it is desired to measure positive static pressures, the birefringence increases with pressure. With the fibre of Figure 7a it decreases.

This makes it possible to avoid, for a certain pressure range, a "blind zone" in which the birefringence is very small or practically zero, which makes the measurement inaccurate because of the very low interferometric contrast.

The present invention also relates to a method of manufacture of an optical fibre such as that of Figure 9c. The starting point is a preform as represented in Figure 9a. This preform is cylindrical. It comprises an optical core 23 surrounded by an optical cladding 24 contained in a matrix 20. The core, the cladding and the matrix are circular and concentric. The core may, for example, be made of silica doped with germanium, the cladding of silica doped with fluorine or with phosphorus or with boron or with a combination of these substances, and the matrix may, for example, be made of pure silica.

This preform is machined so as to cut two flats 22 in the matrix 20. They are diametrically opposite on either side of the optical cladding 24. This preform is subjected to a drawing process which is known in the technique of manufacturing optical fibres. During drawing, the preform is given a partially circular shape at the flats 22, so as to obtain elongation of the optical cladding 24 and of the core 23 in a direction perpendicular to the flats 22.

After drawing and optional coating with a protective jacket 21, the fibre obtained is in accordance with that of Figure 9c.

Claims

CLAYS

1. A pressure sensor with wound optical fibre, comprising a shell including the optical fibre, which sensor is intended to measure a pressure external to the shell' wherein the shell internally contains a medium which is more compliant than the shell, so that the pressure to be measured generates an anisotropic stress tensor (Tx, Ty, Tz) in the optical fibre, the stress difference (Ty-Tx) in a cross-section of the fibre serving to determine the pressure to be measured.

2. A pressure sensor according to Claim 1, wherein the shell comprises at least one tubular structure including the optical fibre, and flanges which close off the two ends of the tubular structure, the medium which is more compliant than the shell being contained in the tubular structure.

3. A pressure sensor according to claim 1 or claim 2, wherein the optical fibre is wound with adjoining turns.

4. A pressure sensor according to claim 1 or claim 2, wherein the optical fibre is wound with spaced turns.

5. A pressure sensor according to any one of claims 1 to 4, wherein the wound optical fibre is embedded in encapsulation.

6. A pressure sensor according to claim 5, wherein the material of the encapsulation is less rigid than the material of the fibre.

7. A pressure sensor according to any one of claims 2 to 6, wherein the tubular structure is surrounded by a bellows which bears on the flanges, a space being arranged between the bellows and the tubular structure.

8. A pressure sensor according to claim 7, wherein the space contains a medium more compliant than the shell.

9. A pressure sensor according to/aos; of Claims 2 to 8, wherein at least one layer externally covers the tubular structure , this layer being made based on a material having a rigidity less than or equal to the rigidity of the material which it covers.

10. A pressure sensor according to/aoge of Claims 2 to 9, wherein at least one layer internally covers the tubular structure, this layer being made based on a material having a rigidity less than or equal to the rigidity of the material which it covers.

any

11. A pressure sensor according to/one of Claims 1 to 10, 6 rein it comprises means for mechanically filtering the pressure to be measured.

12. A pressure sensor according to Claim 11, wherein these means comprise a capillary tube which connects the inside of the shell to the outside.

13. A pressure sensor according to one of Claims 11 or 12, immersed in a base fluid, wherein the internal medium is formed by a mixture of the base fluid and a material, such as a foam, containing bubbles.

a

14. A pressure sensor according to/oge of Claims 1 to 13, wherein the optical fibre is an intrinsically non-birefringent fibre.

any

15. A pressure sensor according to/one of Claims 1 to 13, wherein the optical fibre is an intrinsically birefringent fibre.

16. A pressure sensor according to Claim 15, wherein the optical fibre has an oval crosssection, this fibre having an optical core and an optical cladding, which have oval cross-sections, the major dimension of which is perpendicular to that of the crosssection of the optical fibre.

17. A method of manufacture of the optical fibre used in the pressure sensor of Claim 16, cormprising the steps of machining two diametrically opposite flats in a preform of circular cross-section, concentrically comprising an optical core surrounded by an optical cladding surrounded by a matrix, all of circular cross-section, then in drawing the preform while giving a partially circular shape to the flats until the optical fibre is obtained.

18. A method according to Claim 17,wherein it involves the step of covering the fibre with a protective jacket (21).

19. A pressure sensor with wound optical fibre, substantially as described hereinbefore with reference to the accompanying drawings and as shown in Figure la, or Figure lb, or Figure 2a, or Figure 2b, or Figure 3a, or Figure 3b, or Figure 4a, or Figure 4b, or Figure 5a, or Figure 5b, or Figure 6a, or Figure 7a.

20. A method of manufacture of an optical fibre for-use in a pressure sensor according to claim 19, substantially as described hereinbefore with reference to the accocpanying drawings and as represented diagrammatically by figures 9a, 9b and 9c.