CA1207549A - Fiber coupler transducer - Google Patents

Abstract

Abstract of the Disclosure A transducer includes a fiber optic coupler 10 having a variable coupling efficiency. The fiber optic coupler includes a first fiber and a second fiber and an interaction region in which light propagating in one of the fibers is transferred to the other fiber by evanescent field coupling such that introducing light into one of the fibers may cause light to be output from both fibers.
Detectors detect the light output by the fibers and a ratio meter produces an output proportional to the ratio of light intensity received by one of the detectors to the light intensity received by the other detector, the ratio being indicative of the coupling efficiency of the fiber optic coupler. In a first embodiment the coupling efficiency varies in response to the relative position of the fibers, thereby providing a transducer for measuring displacement. In a second embodiment the coupling efficiency is termperature-dependent, thereby providing a transducer for measuring temperature.

Description

~()754~

PIBER COUPLER TRANSDUCER
Background of the Invention This invention relates to transducers used for the measurement of minute di~placements, for example, in the range of several microns or less with precise accuracy.
The invention also relates to the indirect measurement of other physical parameters, such as ~emperature and pressure, through the use of a secondary transducer whi~h provides a displacement proportional to the temperature or pressure to be measured. The invention relates to the measurement of su~h physical parameters with extremely high resolution over a relatively small range, or over a correlated qroup of dynamic ranges.
The most successful, priorl high resolution displacement transducer known to applicant is a linear variable differential transducer which comprises a transformer having a primary coil and two secondary coils symmetrically spaced on a cylindrical form. A free moving, rod-shaped magnetic core inside the coil assembly provides a path for the magnetic flux linking the coils.
~hen the primary coil is energized by an external source, voltages are induced in the two &econdary coils.
Typically, these secondary coils are connected as an opposed ~eries circuit so that the two induced vol~ages are of opposite polarity~ ~hus, the net output of the transducer is the difference between these two voltages, which is ~ero when tne core is at the center or null p~sition in the transducer. When the core is moved from the null position, the induced voltage in the coil toward which the core is moved increases, while the induced voltage in the opposite coil decreases. Moving the core from the null position produoe a differential voltage output which varies linearly with changes in oore postion. The sign of the output voltage changes as the core is moved from one ~ide of null to the other.
,. ~

120'75~9

-2-Such transducer6 provide isolation between the electrical circuit and the element whose displacement i5 being measured and of~er wide temperature range operation, long mechanical lifetime, and high sensitivity with low drift and linearity over a broad dynamic range. For example, the sens-itivity of such devices provides a minium measurable displacement as low as 0.02 microns. Other ~uch devices may have a linear range as high as 10 inches.
While such transducers have low susceptibility to hostile chemical and thermal environments, their electro magnetic nature makes them relatively susceptible to electrical and magnetic interference, so that their use in a hostile electromagnetic environment may be prohibited.
Furthermore, the dynamic range of a particular transducer, once the transducer has been manufactured, is predetermined as a trade-off against high resolution, so that a particular transducer may not be used for a variety of applications.
Finally, while the resolution of such transducers is relatively high, certain applications may require m2asurement of displacements ~maller than 0.02 microns and, to date, electromagnetic devices capable of such resolution have not been constructed.
The most commonly used temperature sensors of the prior art which provide an electrical signal output are thermocouples. Such devices are constructed by joining two pieces of different metals, often in the shape of thin wires, at a junction located in the environment whose temperatures are to be measured. Temperature changes at this jun~tion modify the electrical potential of each of the metals; and, since the metals react differently, a ~otential difference is generated between the two metals at the junction. Such thermocouples provide a high degree of temperature sensitivity, providing as much as 60 millivolts of signal increase per degree Celsius of temperature change. These thermocouples can be directly ~207549 connected to analog or digital gauges to provide an accurate, repeatable measurement of tempera~ure at the thermocouple junction. --- Calibration tables exist for all common metal pairs, such as chromel-constantan or iron-constantan.
Additionally, such thermocouples opera~e over broad temperature ranges from as low as -150 degrees Celsius to as high as 2,500 degrees Celsius with relatively small error (sometimes as low as 0.5% over the entire temperature range). Because thermocouples have low thermal mass, they reach equilibrium with their environment very rapidly, especially when they are ~onstructed to be extremely thin. Response times as low as 1 milli econd are known. While thermocouples can be protected to withstand hostile chemical environments, they are susceptible to induced errors from electromagnetic interferenceO In addition, unless well protected by shielaing which would necessarily increase the thermal mass and the response time of the thermocouple, these devices may be sensitive to some corrosive environments.
Summary of the Invention The present invention utilizes the sensitiYity of a fiber optic coupler to mechanical displacement of its coupler elements to provide an extremely high resolution, non-elec~romagnetic displacement transducer.
More specifically, applicant has found that the coupling coefficient of single mode fiber optic couplers is predictably variable in response to mutual displacement of the fiber optic elements which make up the coupler, particularly when these elements are moved in a direction perpendicular to the ~iber optic longitudinal axis.
Furthermore, be~au6e in single ~ode fiber-~ptic couplers, plural orders of coupling are achievable when the core members are closely spaced (a condition known as ~overcoupling"), a family of dynamic ranges are possible for a displacement transducer based upon-~uch ~ouplers.

~2(~7549 Each of the ranges within the family may have a relatively high resolution, so that a single coupler ~ay be used for measurements over different ranges, with different sensitivities, but with each range having a higher S resolution than would be possible i$ the entire range of the family were covered by a single dynamic range.
In a specific configuration for utili~ing a fiber optic ~oupler for displacement measurement, a single mode fiber optic coupler using evanescent field coupling includes a pair of fiber optic elements mounted in a face-to-~ace configuration within a pair of relatively moveable bases or ~ubstrates. One of these substrates is rigidly mounted in a measurement vise, and the second substrate is attached to an element whose movement is to be measured.
As the second substrate and an associated fiber optic element are displaced relative to the stationary substrate, th~ coupling efficiency of the fiber optic coupler is altered with a high degree of sensitivity to minute ~ubstrate movements.
By supplying light to the coupler input and utilizing photo-detectors on the primary and ~econdary coupler outputs 9 the ratio of light at these outputs may be measured. This ratio may be directly transformed into a measurement of displacement of the moveable fiber optic substrate relative the stationary substrate. Futhermore, by attaching the moveable substrate to an object whose displacement is to be measured, a direct reading of displacement with a high degree of accuracy and repeatability is possible, even in a hostile electromagnetic environment.
The displacement transducer of the present invention ~ay also be used to measure--the frequency~and amplitude of mechanical vibrations~ Moreover, using a secondary transducer, such as a material having known thermal expansion characteristics, the displacement transducer may ber used indirectly to measure temperature. Likewise, it ~20'754~

has been found that, by using other secondary transducers, ~uch as materials having known compressibilities, the displacement transducer of this invention may be used as a pressure sensor. Other physical properties may likewise be measured, so long as secondary transducers can be found to provide a displacement which varies in accordance with such properties.
The present invention utilizes the temperature sensitivity of index matching oil to provide a fiber optic coupler as the basis for a high resolution, non-electromagnetic temperature transducer.
More specifically, applicant has found that the coupling efficiency of a sihgle mode fiber optic coupler is predictably variable in response to ~hanges in the refractive index of index matching oil placed between the optical fibers which make up the coupler.
In a specific configuration for utilizing a fiber optic coupler for temperature measurement, the substrates are carefully superimposed to provide a predetermined coupling efficiency at a ~elected temperature and are then rigidly interconnected in this ~elected placement. Oil, having a refractive index which varies in a predictable manner with temperature, fills a space between the bases or substrates. Any one of plural disclosed configurations may be utilized for maintaining a predetermined thickness of such oil between the faces of the coupler substrates.
By supplying light to the coupler input and utilizing photo detectors at the primary and secondary coupler outputs, the ratio of light at these outputs may be measured. This ratio may be directly transformed into a measurement of the fiber optic coupler and its surroundings.
Since the fibers exiting the fiber optic coupler from opposite ends would normally require that the coupler, and in turn the area whose temperatures are to be measured, must be placed between th~ 1$9ht source and the detectors, lZ()7S4~

techniques have been developed for providing measurement of the coupling efficiency exclusively from one æide of the coupler. One of theæe techniques involves the looping of the output ends of the optical fibers over the ~urf~ce - 5 of the coupler and rigidly bonding these output endæ to the coupler surface to protect them from damage. A second technique involves the reflection of the coupled branch power at the output end and the use of a secondary coupler to normalize out any laser output fluctuations, which tO would otherwise be interpreted as temperature variations. This second method requires an optical isolator at the laser light source to prevent laser output fluctuations which would result from feedback through the primary optical coupler.
lS In order to ~eep the thermal mass of the coupler low and to thus provide a low response time of the coupler to temperature changes in the measured environment, the bases or substrates used to support the optical fibers are ground to the smallest practical dimensions.
These and other advantages of the present invention are best understood throuqh reference to the drawings, in which:
FIGURE 1 i~ a cross-sectional view of the fiber optic coupler used in the present invention;
FIGURE 2 is a cross-sectional view of the coupler of Figure 1, taken along line 2-2;
FIGURE 3 is a cross-sectional view of the coupler of Figure 1 taken along line 3-3;
FIGURE 4 is a perspective view of one element of the coupler of Figure 1, ~eparated from the other element thereof to show the configuration of the confronting face of æaid element;
FIGURE 5 is an elevation view of a test assembly including the fiber optic coupler shown in Figure 1, for measuring displacement;

~207S49 FIGURE 6 i~ graph showing the theoretical coupling efficiency of a coupler plotted as ~ function of the horizontal displacement of the optical fibers thereof;
FIGURE 7 is a plot similar to the plot of Figure 8, but showing a second thesretical fiber optic coupling ~urve for a fiber optic coupler having a core ~pacing smaller than that ~hown in ~igure 6 to generate a family of dynamic ranges for the displacement transducer;
FIGURE B is a ~chematic diagram of an electrical circuit used for measuring the coupling efficiency of the the eoupler within the test assembly of Figure 5 to provide a measurement of displacement;
FIGURE 9 is a schematic block diagram of an alternative electrical system for use in place of the electrical ~ircuit of Figure 8 for high resolution measurements of displacement;
FIGURE 10 is a perspective view showing a first structure for rigidly interconnecting the bases of ~ubstrates of the coupler of ~igure 1 for use in temperature measurement;
FIGURE ll is a perspective ~iew showing one b~se or substrate of the coupler of Figure 1 incorporating a dam for precisely ~eparating the coupler substrates in a temperature measuring coupler;
FIGURE 12 is a chart showing the relationship between temperature in degrees Celsius and coupling efficiency for a particular coupler used for temperature measurement FIGU~E 13 is perspective view ~howing a first fiber optic arran~ement for permitting the input and output fibers to exit from the coupler exclusively on one side and thus permitting the coupler to measure temperatures at ~he distal end for the fiber optic ~onnections; and PIGURE 14 is a schematic representation showing the use of an optical isolator and a secondary fiber optic coupler to permit temperature measurements using a fiber optic coupler having fiber elements exitin~ exclusively 12~7549 from one end of the primary coupler and thus providing temperature me~surement at the di~tal end of the fiber optic elements.
Detailed Descri~tion of the Preferred Embodiment . _ Referring ini~ially to ~igures 1 through 4, a fiber optic coupler 10 is shown which provides the transducing element of the displacement transducer 11 of the present invention.
The coupler 10 includes two strands, 12a and 12b, of a single mode fiber optic material mounted in longitudinal arcuate grooves 13a and 13b, respectively, formed in optically flat confronting surfaces 14a and 14b, respectively, of rectangular bases or ~ub~trates 16a and 16b, respectively.
Each of the ~trands 12a and 12b comprises a commercially available fiber of quartz glass which is doped to have a central core and an outer cladding. The applicant has found that the present invention works effectively with single mode fibers, which typically have a core diameter in the order of 10 microns or less and a cladding diameter on the order of 125 microns.
The arcuate grooves 13a and 13b have a radius of curvature which is very large compared to the diameters of the strands 12a and 12b. Thus the fiber optic strands 12a and 12b, when mounted in the grooves 13a and 13b, respectively, gradually converqe toward the center and diverge toward the edges of the substrates 16a and 16b.
At the centers of the substrates 16a and 16b the depth of the grooves 13a and 13b which mount strands 12a and 12b, respectively is less than the diameter of the strands 12a and 12b. At the edges of the substrates 16a and 16b, the -depth of the grooves 13a and 13b is preferably at least as great as the diameter of the strands 12a and 12b. Fiber optic material is removed from each of the strands 12a and 12b to form oval shaped, planar surfaces 18a and 18b, which are co-planar with the confronting ~urfaces 17a and :1207549 _9_ 17b, respectively, of the substrntes 16a and 16b, respectively. In the embodiment ~hown, the ~oupler halves 1Oa ~nd 1Ob are identical and are assembled by placing the confronting surfaces 14a and 14b of the substrates 16a and 16b together EO that the surfaces 18a and 1Bb of the strands 12a and 12b are in facing relationship. An index matching substance (not shown), ~uch as index matching oil, is provided between the confron~ing surfaces 14a and 14b. The index matching ~ubstance has a refractive index approximately equal to the refracive index of the cladding and also function to prevent the optically flat ~urfaces 14a and 14b from becoming permanently locked together.
An interaction region 32 is formed at the junction of the trands 12a and 12b~ In the interaction region 32, light is transferred between the strand~ 12a and 12b by evanescent field coupling. When the spacing between the - cores of the ~trands 12a and 12b i8 within a critical zone, ea~h strand receives a ~ignificant portion of the evanescent field energy from the other strand; and optimum coupling is achieved without ~ignificant energy loss. For a ~ingle mode fiber having a step index gradient, the critical zone can be quite narrow. In a single mode fiber of the type shown in Figures 1 through 4, for example, the required center-to-center spacing between the strands 12a and 12b at the center of coupler is typically less than a few (for example, 2-3) core diameters.
Preferably the strands t2a and 12b are symmetrical through the interaction region 32 in the plane of the facing surfaces 18a and 18b so that the facing surfaces 18a and 18b are coextensive if superimposed.
The coupler 10 includes four ports labeled A, B, C and D in Figure 1. If it is assumed that inPut light of a suitable wavelength, for example, 1.15 microns, is applied to port A, this light passes through the coupler 10 and is output at either one or both of port B and port D, depending on the amount of power that is coupled between ~207549 the ~trands 12a and 12b. In this reg~rd, the term ~normalized coupled power~ is defined a8 the power ~atio of the coupled power to the total output power. In the above example, the normalized coupled p~wer would be equal to the ratio of the power at port D to the ~um of the power output at ports B and D. This ratio is also referred to as the ~coupling efficiencyH and when ~o used is typically expressed as a percent. Thus, when the term ~normalized coupled power" is used herein, it should be understood that the corresponding coupling efficiency is equal to the normalized coupled power times 100.
The coupler 10 operates on evanescent field coupling principles in which guided modes o~ the strands 12a and 12b interact through their evanescent fields to cause light to be transferred between the strands 12a and 12b.
As previously indicated, the transfer of light between the ~trands 12a and 12b occurs at the interaction region 32.
The amount of light transferred i~ dependent upon the proximity and orientation of the cores of the strands 12a and 12b as well as the effective length of the interaction region 32. If the length of the interaction region 32 and the separation of the ~trands 12a and 12b within this region 32 are properly ~elected, light within the coupler 10 will make only one transfer between the strands t2a and 12b as it travels through the interaction regon 32. ~f the coupler 10 is properly constructed, it is possible under such conditions to couple 100% of the light input at port A to port D. If the length of the interaction region 32 is further increased, or the separation between the

3~ strands 12a and 12b further reduced, a phenomenon referred to herein as overcoupling will occur in which the light will transfer back to the ~trand from which it originated~ Several orders of overcoupling are possible. Thus, as the interaction length is still further increased or the separation is ~till further reduced, the light will make several transf~r~ibetween the 120'7S$9 strands 12a and 12b. Thus, the light may make ~ultiple transfers back and forth between the two strand~ t2a and 12b as it travels through -the region 32, the number of ~uch transfers being dependent on the length of interaction region 32 and the spacing of the ~trands 12a and 12b within this region.
Figure 5 illustrates the incorporation of the fiber optic coupler 10 into a displacement transducer configuration. In this configuration, the sub~trate 16b 0 i8 rigidly mounted within a frame or yoke 20 having a ~tepped, ~-shaped channel 22. The lower portion 24 of the channel 22 i~ narrower than the upper portion 26 and is sized to rigidly mount the lower substrate ~6b, with the bottom of substrate 16b resting on the b~ttom of the ~hannel 22.
The step transition between the upper portion 26 and the lower portion 24 is below the confronting faces 14a and 14b of the coupler 10 so that the upper substrate 16a may be translated in a direction perpendicular to the channel 22 between the sidewalls forming the upper channel portion 26. The coupler 10 is oriented 80 that the strands 12a and 12b are in a direction parallel to the channel 22 to permit such translation to laterally offset the facing ~urfaces 18a and 18b (Figure 1).
A cylindrically-shaped retainer 28 is slidably mounted to project from one sidewall of the upper portion 26 of the channel 22. This retainer 28 is spring biased to bear against one ~ide of the substrate 1~a. A differential micrometer 30 is mounted on the opposite sidewall of the upper channel portion 26. The differential micrometer 30 bears against the opposite side of substrate 16a so that the ~ubstrat~ 16a and 16b are held between the mierometer 30 and the spring biased retainer 28.
By turning the micrometer 30, the position o~ the substrate 16a may be adjusted relative to the ~ubstrate 16b to provide a home or starting position for the displacement measurement.

:~207S49 The upper substrate 16a mount~ a projecting flange 33 which abuts a pushrod 34 mounted for reciprocal movement within a guide 36. The pushrod 34 may include an enlarged distal end 38 for mounting to an element (not shown) whose displacement is to be measured.
As the pushrod 34 is displaced to the left, as viewed in Figure 5, the pushrod 34 bears upon the flange 33 and aisplaces the upper substrate 16a to the left, relative to the substrate 16b, overcoming the bias of the retainer 28 and displacing the substrate 16a away from the micrometer 30. This displacement, in turn, laterally displaces the ~trands 12a and 12b ~Figure 1).
The curve 40 of Figure 6 graphically illustrates the effect of laterally off6etting the fiber facing surfaces 18a and 18b IFigure 1) for a coupler 10 having a minimum core spacing when the surfaces 18a and 18b are superimposed which provides a 50% coupling efficiency. As shown in Figure 6, when the facing surfaces 18a and 18b of the fibers 12a and 12b are laterally offset in either direction from the superimposed configuration to increase the spacing between the strands 12a and 12b, the coupled power gradually decreases to zero. A portion of this gradual increase is substantially linear and is denoted as the ~dynamic range~ 42 in Figure 6. It will be understood, therefore, that if a coupler 10 having the characteristics shown in Figure 6 is utilized in the system of Figure 5, the micrometer 30 will ~irst be used to adjust the starting location of the upper substrate 16a to a location within the dynamic range 42. For example, the starting position may be at one end of the dynamic range 42 so that pressure on the pushrod 34 (Figure 5) can --adjust the position of the upper substrate 16a in a direction whîch utilizes the full dynamic range 42.
Alternatively, if measurement of displacement in either direction is desired, the micrometer 30 may be used to - adjust the starting position so that it is in the middle ~2075~g of the dynamic range ~2. It will be recognized that,as the pushrod 34 adjusts the position of ~he upper substrate 16a so that the coupling efficiency varies between approximately 20% and 40%, as shown in Figure 6, the ratio of light output at ports B and D (Figure 1) will vary linearly in relationship with the di~placement of the substrate 16a. Thus, thi~ ratio provides a direct linear measurement of the displacement of substrate t6a.
Referring now to ~igure 7, the coupling efficiency is plotted versus displacement form zero with surfaces 18a and 1Bb superimposed for a coupler having a minimum edge to edge core spacing which provides overcoupling. Thus, when the upper substrate 16a (Figure 1~ is laterally offset from the lower substrate 16b by 9 microns, 100%
coupling occurs. That is, light entering the coupler at port A will exit at port D, with virtually no light exiting at port B. When the trand~ 12a and 12b are moved closer to one another by displacing the upper substrate 16a toward a directly superimposed position above the substrate 16b, the light which was originally coupled from port A and strand 12a into strand 12b is recoupled into s~rand 12a. Therefore, at a displacement of 5 microns, the coupling efficiency is zero, all of the light having then been coupled back from strand 12b to ~trand 12a.
As the strands 12a and 12b are moved even closer toward a superimposed position, so that their offset is two ~icrons, 100% coupling efficiency is again achieved.
That is, the light is transferred from port A and strand 12a to strand 12b, then back to strand 12a, and finally completely back to strand 12b to exit at port D.
By placing the fiber cores suficiently close to one another~ as by grinding and polishing the cladding of the strands 12a and 12b sufficiently to place the cores very near the surfaces 18a and 18b, respectively, a larger order of overcoupling may be achieved. The overcoupling which graphically is illustra~ed in Figure 7 will provide 120~54g three separate dynamic ranges 46~ 48 and 50, each being sub6t~ntially linear, on either side of the null or superimpo6ed position. Each of the dynamic ~6, 48, 50 ranges may be used, ~electively, for displacement measurement6 and each dynamic range 46, 48, and 50 will provide a different transfer function or transducer constant.
The dynamic range 46 providing the greatest change in coupling efficiency ver~us horizontal dicplacement will provide the most accurate displacement measurement~, the highest measurement resolution, but the lowest dynamic range. Conversely, the dynamic range 50 having the lowest slope will provide the widest linear displacement excursion of the three ranges 46, 4B, 50, but the lowest resolution. Thus, by utilizing the micrometer 30 (Figure 5) to adjust the coupler 10, a dynamic range 46, 48, 50 may be selected which provides the sensitivity and range required for a particular measure~ent.
It should be recognized that, in this particular system, the three dynamic ranges 46, 48, 50 provide a linear displacement excursion of about 2, 3, and 4 microns, respectively, with a resolution of approximately 0.005 microns in each range 46, 48, S0.
Referring now to Figure 8, an electrical system 51 is ~hown or measuring the ratio of ligh~ output of ports B
and D (Figure 1) to provide a measurement of coupling efficiency and to thus generate an electrical signal which is proportional to displacement. In the arrangement shown in Figure 8, a laser diode 52 provides a continuous wave ~ignal to the fiber optic coupler 10 and specifically to port A (Figure 1). The outputs from ports B and D are connected, respectively, to a pair of detectors 54 and 56, each of which provides an output electrical signal proportional to the light intensity at ports B and D, respectively. The electrical signals output from the detectors 54, 56 are supplied to a ratio meter 58, which ~20~S4g provides a direct measurement of the ratio between the output light at port D and the output light at both ports B and D. ~his output is supplied as an analog output 60 for operating other equipment (not shown) and as a digital - 5 display 62 for providing a direct display of displacement. It will be understood that the analog output 60 and digital display 62 may include scaling factors or constants selected to provide direct displacement measurement in appropriate measurement units. The ratio meter 58, along with the analog and digital outputs 60 and 62, respectively, is a readily available electronic device well known in the prior art.
Figure 9 provides a more accurate electronic system 53 for measuring the output signals from detectors 54 and 56, and thus mechanical displacement. In this example, a laser diode 5~ supplies light to port A of the fiber optic coupler 10; and the outputs at ports B and D are monitored by detectors 54 and 56 as in Figure 8. However~ the laser diode 52 does not supply continuous light as in Figure 8, but rather, provides a pulsed light output, which is induced by a pulse generator 66 ~ontrolling a diode power supply 68. In a preferred exemplary embodiment, the pulse generator 66 may provide square wave pulses at a frequency of 2 kHz so that the output from the laser diode 52 is modulated at a 2 k~z rate.
The output from the pulse generator 66 is supplied, in addition, as a reference on line 69 to a pair of lock-in amplifiers 70 and 72, which are connected to the output of the detectors 54 and 56, respectively. The reference input 68 to the amplifiers 70, 72 provides a timing signal for enabling these amplifiers 70, 72. The amplifiers 70, 72 are well known in the prior art and each provides a ~ignal proportional to the average peak value of the output from the detectors 54, 56, respe~tively. Thus, by using the reference signal on line 69, the lock-in amplifiers 70, 72 will measure the peak output level from 120'7549 the detectors 54, 56 and will average thi~ peak level over a predetermined number o~ cycle~ of the pulse generator 66. This number of cycles i8 adjustable on each of the amplifiers 70, 72.
The output ~rom the amplifiers 70, 72 is supplied to a ratio meter 58 as described in reference to Figure 8 to provide an analog output 60 and digital display 62 corresponding to the displacement of the upper substrate 16a (Figure 1~.
By selecting the integration time o~ the lock-in amplifier~ 70, 72, that i8, the number of cycles of pulse generator 66 which are averaged by each of the amplifiers 70, 72, the band width of the measurement may be - adjusted. ~or example, an integration time of 1 second corre ponds to a measurement bandwidth of 2kHz centered about ~he pulse generator 76 frequency. Tbis reduced bandwidth substantially reduces the noise signal contribution in the output to the ratio meter 58, and thereby increases the accuracy of the me~surements while reducing the minimum time period between successive measurements to, in this particular example, one second.
That is, measurements are available only at a one-second repitition rate. Thus, by properly adjusting both amplifiers 70, 72 to an integration time which provides only the required measurement repitition rate, the noise bandwidth may be substantially reduced, and the signal-to-noise ratio significantly increased, increasing the accuracy of the measurement. An operator may thus select a proper tradeoff between measurement accuracy and repetition rate to fit the re~uirements of a particular measurement.
From the preceding description, it can be seen that the fiber optic coupler provides a non-electromagnetic displacement ~easurement which is extremely accurate.
Further, this single coupler is capable of providing more than one dynamic measurement range, if the coupler is manufactured to provide over-coupling.

~ZO'7549 Referring to Figure 10, in order to form a temperature sensor 63 similar to the coupler 10 described above for displacement measurements, it is ~eces~ary for substrates 16a and 16b to be aligned or tuned to provide a coupling efficiency which i~ approximately midway along a relatively linear portion of a curve of coupling efficiency versus core spacing. This will increase the linearity of the temperature measurement. Once the coupler 10 has been tuned to this preferred fiber spacing, the technique shown in Figure 10 may be utilized. As shown in that figure, the upper substrate 16a and lower substrate 16b may be welded along their outside edges using a low melting temperature glass 64 to prevent -shifting of the substrates 16a and 16b. Since the sizes Of the substrates 16a and 16b should be reduced to the greatest extent possible to lower the thermal mass of the temperature sensor 63, the sub~trates 16a and 16b may be clamped together in the preferred orientation and the edges ground prior to application of the welding glass ; 20 64. Prior to this welding operation, the substrates 16a and 16b should be ~eparated from one another by a layer of index matching oil.
In the embodiment shown in Figure 11, a thicker layer of index matching oil of uniform cross-section may be applied between the substrates 16a and 16b by providing, on the substrate 16b, a thin dam 68 which surrounds the face of the substrate 16b which abuts against the sub6trate 16a. This dam 68 will hold the index matching ; oil near the coupling region and will additionally provide 3U a bonding area for attaching the two substrates 16a and 16b together. The dam 68 should be deposited on the substrate 16b after the substrates 16a and 16b have been ground to the smallest size possible. Fabrication of the dam 68 can be easily achieved using standard vacuum deposition techniques which provide accurate, low-thickness films as thin as 10-angstroms. A preferable .

~20754 -la-thickne66 for the dam 68 i8 approxi~ately 1 ~icron. Once the dam 68 has been depo~ited, molecular bonding may be used, enhanced ~y`ultra~onic wave , to attach the dam 68, nnd thus the substrate 16b to the face of the sub~tr~te 16a after temperature sensor 63 has been accurately tuned. Alternatively, a thin bonding agent ~ay be used between the dam 68 and the substrate 16a.
The curve 70 of Figure 1~ graphically illustrates the effect of temperature on the index of refraction of a typical index matching oil which is placed between the sub6trate 16a and 16b. Curve 70 also fihows the relationship between temperature in degrees Celsius and the index of refraction and, in turn, the relationship between temperature and coupling efficiency.
As can be ~een from the exemplary curve 70, a typical index matching oil can provide relatively linear changes in coupling efficiency between temperatures of approximately 20 degrees and 43 degrees Celsius. By ~electing oil~ having different characteristic refractive indices with temperature, temperatures within different ranges ~ay be measured; and different resolution measurementæ of temperature may be achieved.
The temperatùre ~ensor 63 may be used in the configuration shown in Figure 1, witb the input light being supplied to port A of the coupler 10 and the output light at ports B and D being measured to determine coupling efficiency for the temperature sensor 63 if the environment whose temperatures are to be measured can conveniently be placed between the light source 52 and the photo detectors 54, 56. Since this is often not the case, the arrangement of Figure 13 may be utilized to permit the light source 52 and photo detectors ~4, 56 all to be mounted exclusively on one side of the temperature sensor 63 and to thus allow the temperature senssr 63 to be inserted into a thermal environment to be measured at the di~tal end of the optical fibers :12a and 12b~ In the ~207549 arran~ement shown in Figure 13, the optical fibers 12a and 12b exiting the temperature sensor 63 at ports B and D may be looped over the upper substrate 16a and bonded to the upper surface of the substrate 16a by any suitable bonding agent. In this configuration, it is advi-sable to melt low temperature glass 72 onto the end of the temperature sensor 63 to encapsulate and protect the looped portion of the fibers 12a and 12b.
An alternative arrangement for providing the temperature sensor 63 at the distal end of the optical fi~er interconnec-tions is shown in Figure 14. In this configuration, the fibers 12a and 12b are cut flush with the end of the temperature sensor 63 at the ports B and D (Figure 1).
A mirror 74 is placed against the end of the fiber 12a at port D to reflect all of the coupled branch power toward the temperature sensor 63. The temperature-dependent optical signal may then be detected at the second input side fiber 12b at port C, which thus becomes an output channel. The signal magnitude, Po, is proportional to PI (E)(l-E) where E is the temperature-dependent coupling efficiency of the fiber coupler 10.
- When using the configuration shown in Figure 14, it is necessary to protect the laser diode 74 from power reflected by the mirror 74 and coupled back to the fiber strand 12a.
This protection is accomplished by an optical isolator which prevents such reflected power from introducing laser output f]uctuations. Such optical isolation is accomplished, for example, by using a quarter wave plate 48 combined with a polarizer 78 in a well-known manner.
3a The mirror 74 may be simply deposited on the polished end surface of the temperature sensor 73 where the output fiber 12b has been cut. A metallic type of mirror is preferable to a dielectric, multi-layered type mirror whose reflection coefficient is also temperature sensitive. A
layer of glass 80 may be vacuum deposited on the top of the ~;~07~;4g ~20~

mirror 74 to pro~ide extra mechanical Frotection.
In this arrangement, a second optical coupler 82 is utilized along the input fiber 12a to normalize out any laser output or launching efficiency variations. Output from the coupled strand 84 of this coupler 82 is used as a reference for measuring the ratio of coupling ef~iciency.
It has been found that by increasing the thickness of the layer of index matching oil between the substrates 16a and 16b (Figure l),-the sensitivity of the temperature sensor 63 and temperature changes may be increased. This sensitivity can be further improved by polishing the substrates 16a, 16b closer to the cores of the fibers 12a, 12b and increasing the li~uid layer thickness. Reducing the size of the substrates 16a, 16b, for example, to approximately five by five by ten millimeters yields a fairly low thermal mass and thus reduced response times.
Adequate interconnection of the substrates l~a and 16b, as by welding, is important so that the physical size of the oil layer between the substrates 16a, 16b does not change the temperature. Furthermore, the substrate l~a should be constructed from material, such as quartz, which has a thermal expansion coefficient similar to that of the optical ~ibers 12a and 12b to avoid damage to the optical fibers 12a and 12b with temperature changes.
The temperature sensor 63 provides repeatable, accurate temperature measurements typically in a range from approximately 0 degrees C to 50 degrees C, depending upon the index matching li~uid used. A sensitivity as high as 400 millivolts per degree Celsius is achievable, and resolutions providing minimum detectable temperature changes of 0~001 degree Celsius may be realized. The temperature sensor 63 is insensitive to hostile electro-magnetic enviroments and may thus be utilized in locations where other temperature measuring devices are not practical.

~20'7S49 The electrical systems 51 and 53 of Figures 8 and 9, respectively, may be used in connection with the fiber optic devices of Figure 10, 13, and 14, which have temperature-dependent coupling efficiencies, to generate an electrical signal which is proportional to temperature.
The analog output 60 and digital display 62 may include appropriate scaling constants to provide a direct temperature readout in selected units, such as Celsius or Fahrenheit.

Claims (17)

The embodiments of the invention in which an exclusive property or privilege is claimed are defined as follows:

1. A transducer for measuring displacement, characterized by:
a first optical fiber having a first end, and a second optical fiber;
a light source for introducing light into said first optical fiber through said first end;
said first and second optical fibers juxtaposed to form a fiber optic coupler such that at least a portion of the light conducted by said first optical fiber is transferred to said second optical fiber, each of said fibers having a core surrounded by cladding;
a device for mounting said first and second juxtaposed fibers for relative sliding movement, without deforming said claddings, in accordance with the displacement to be measured; and a device for sensing said relative sliding movement, said sensing device including a detector for detecting light from one of said fibers.

2. A transducer as claimed in Claim 1 further characterized by said sensing device including first and second detectors for detecting light from said first and second fibers respectively and a ratiometer for providing an output in accordance with the ratio of light intensities detected from said fibers.

3. A transducer as claimed in Claim 2 further characterized by apparatus displaying the output of said ratiometer and scaling the output into the desired dimensional units.

4. A transducer according to Claim 2 further characterized in that the fibers are single mode optical fibers.

5. A transducer according to Claims 2 or 4, further charac-terized by:
a first lock-in amplifier connected between the ratio-meter and the first detector to provide to the ratiometer a signal proportional to the average peak value of the light intensity incident upon the first detector;
a second lock-in amplifier connected between the ratio-meter and the second detector to provide to the ratiometer a signal proportional to the average peak value of the light intensity incident upon the second detector; and a pulse generator connected to the first lock-in ampli-fier and to the second lock-in amplifier to control the time during which the first and second lock-in amplifiers average the output of the first and second detectors to control the bandwidth of the signal output of the ratiometer.

6. A transducer according to Claim 1, further characterized by:
a first substrate having a groove therein, the first fiber being mounted in the groove;
a second substrate having a groove therein, the second fiber being mounted in the groove;

a frame having a channel rigidly mounting the second substrate;
a retainer within the channel adjacent the first sub-substrate;
a flange extending from the first substrate; and a pushrod adjacent the flange for displacing the first substrate toward the retainer relative to the second sub-strate to laterally displace the first fiber relative to the second fiber, thereby varying the coupling efficiency of the coupler in response to a displacement of the pushrod.

7. A transducer according to Claim 6, further characterized by a micrometer connected to the frame for adjusting the initial position of the first substrate relative to the second substrate to select a dynamic range of the coupling efficiency.

8. A transducer according to Claim 6 further characterized by said first and second optical fibers being ground to remove a portion of the cladding of said fibers and to provide a planar surface of cladding coplanar said substrates.

9. A transducer as defined in Claim 1 further characterized in that the transfer of light between said first and second fibers occurs by evanescent coupling.

10. A transducer as claimed in Claim 1 further charac-terized in that said first and second fibers are juxtaposed to provide overcoupling and thereby provide plural displacement ranges for said transducer.

11. A transducer as claimed in Claim 9 further characterized in that said first and second fibers are juxtaposed to provide overcoupling and thereby provide plural displacement ranges for said transducer.

12. A transducer as claimed in Claims 10 or 11 further characterized in that the coupling between said fibers varies substantially linearly with said displacement for a plurality of displacement ranges, and that in at least one of said displace-ment ranges said optical coupling increases as the separation between said fibers increases.

13. A method of measuring the displacement of an object, characterized by:
illuminating a first end of a first optical fiber;
juxtaposing said first optical fiber with a second optical fiber so that at least a portion of the light con-ducted by said first optical fiber is transferred to said second optical fiber, each of said fibers having a core surrounded by a cladding;
relatively sliding said optical fibers, without deforming said claddings, in accordance with the displacement of the object; and measuring the light transferred between said two optical fibers to determine the displacement of the object.

14. A method of measuring the displacement of an object, as defined in Claim 13 characterized in that said fibers each have a portion of the cladding removed to form a flat, oval surface, and wherein said removing step comprises sliding said flat, oval surfaces on said optical fibers to cause relative movement between said fibers.

15. A method of measuring the displacement of an object, as defined in Claim 13 characterized in that said measuring step comprises:

measuring the light output at a second end of said first optical fiber and measuring the light output at the corres-ponding end of said second optical fiber.

16. A method of measuring the displacement of an object, as defined in Claim 15, further characterized by:
determining the ratio of said measured output at said corresponding end of said second optical fiber to the sum of said measured outputs at both said corresponding end of said second optical fiber and said second end of said first opti-cal fiber.

17. A method of measuring the displacement of an object, as defined in Claim 15, characterized in that said illuminating step comprises providing modulated illumination and wherein said light output measuring step comprises measuring the peak output of illumination at said second end of said first optical fiber and said corresponding end of said second optical fiber and inte-grating said peak output over time to reduce the bandwidth of said measuring step.