Method and Apparatus for Remote Position Sensing Using Edge Diffraction

This invention relates to the application of laser light properties to a diffraction edge remote position sensor for detecting the position of a surface or interface, and in particular, but not exclusively, to such a sensor for determining the position of the vertex of a partially fabricated contact lens during the manufacture thereof.

In one known method of manufacture of contact lenses, for example, that disclosed in U.S. Pat. No. 4,455,901, a cylindrical blank of plastics material is machined on a lathe to form the inner concave surface of the lens to be produced. Manufacture of the lens is completed by attaching the machined concave surface to a domed metal arbour by means of wax or other suitable meltable material, mounting the metal arbour in the chuck of the lathe and machining the exposed material to form a convex surface defining the outer surface of the lens and then removing the lens from the arbour.

It will be appreciated that, in order to ensure that the lens is neither too thick nor too thin, it is necessary to be able to determine the vertex position--that is, the point of intersection of the axis of symmetry of the lens with the inner surface--with a high degree of accuracy. If the lens is too thick or too thin it will need to be scrapped because it will not perform to its optical design specifications. At present, the position of the vertex of the lens is determined by using a micrometer to measure the distance between the vertex of the concave surface and the plain end of the blank, and then marking this dimension on the arbour. Knowing this dimension, the blank is machined to form the concave surface defining the outer surface of the lens at a distance to provide a lens of the required thickness. This process is, however, time consuming and does not provide sufficient accuracy for production of thin lenses. In the particular manufacturing technique studied, it was necessary to be able to determine the position of the vertex with an accuracy in the region of .+-.10.0 microns.

A need therefore exists for a sensor capable of determining the vertex position of the formed surface of a partially fabricated contact lens with sufficient accuracy to enable the other surface to be formed to produce a contact lens having an acceptable thickness. Diffraction phenomena generated by laser sources have previously been used for demonstrations of the wave nature of light. Such applications of diffraction have been mostly limited to image processing via spatial filtering and have not been previously used for determining the vertex position of a partially fabricated contact lens to within the required accuracy. However, in the parent application, the present inventor recognized that the properties of laser sources, namely, the high beam intensity and sub-millimeter beam diameter, are ideally suited for application to edge and position sensing. The present invention relates to further enhancements of the method therein disclosed.

According to one aspect of this invention, there is provided an apparatus for detecting the presence of a surface or interface of an object at a predetermined position within a sensing region with respect to the apparatus, the apparatus comprising:

illuminating means for directing radiation along a sensing axis through the sensing region; and

radiation detector means for detecting radiation from the illuminating means which has been deviated by the surface or interface of the object, the radiation detector means being disposed such that radiation which is not deviated by the surface or interface does not reach the radiation detector means, whereby the presence of the surface or interface of the object may be determined from the radiation detected by the radiation detector means.

By this arrangement, when the surface or interface is in a predetermined position within the sensing region, the beam of radiation is caused to be deflected by a surface diffraction effect, thus altering the intensity of the radiation detected by the optical radiation detector means. The inventor has observed that as a surface is moved through the sensing region in a direction transverse to the sensing axis a curve representing intensity versus axial position reaches a sharply defined, symmetric peak at a unique position which is thought to be caused by surface diffraction effects akin to those which give rise to the sunburst phenomenon observed at dusk or dawn. It will be appreciated that the sharpness of the peak allows data position to be defined with a high degree of accuracy, and since the peak is symmetrical, its exact location may be determined by taking the mean position of two equal intensity points, one on either side of the peak.

Preferably, the illuminating means includes a laser source and the optical radiation detector means comprises an objective lens adapted to form an image of the sensing region at an image plane and an eyepiece adapted to focus the image onto a detector element. In this arrangement, a slit may advantageously be provided at the image plane to select a predetermined portion of the sensing region for examination.

The present invention relates to a further development of the above technique whereby under optimum conditions positions can be located to <.+-.1 .mu.m at 30 mm; <.+-.5 .mu.m at 3 meters and <.+-.10 .mu.m at 8 meters. A closed loop system also promises a further significant increase in resolution and a means of obtaining high resolution profile measurements. However, the technique herein disclosed is not limited to test pieces of any particular range of refractive indices such as contact lenses and may be used for other objects having flat surfaces to curvatures <<1 mm radius.

As in the parent application, the basis of the technique is the projection of an image of an edge diffraction intensity distribution through a slit to a detector while the non-diffracted component of the beam is prevented from reaching the detector. The high intensity laser diffraction edge pattern generates a narrow symmetric output signal from the detector as the edge transverses a slit which enables test piece positions to be located along the translation axis. The present invention relates to particular configurations whereby the above technique is performed with or without a beam stop or some other device for preventing the non-deflected beam which has travelled through the sensing region from reaching the radiation detector.

By way of example only, several specific embodiments of this invention will now be described in detail, reference being made to the accompanying drawings, in which:

FIG. 1 is a schematic view from above of part of a contact lens lathe, illustrating a first embodiment of a position sensor;

FIG. 2 is a graph plotting intensity against axial position for the arrangement of FIG. 1;

FIG. 3 is a view showing the view through the microscope of FIG. 1;

FIG. 4 is a schematic view of a second embodiment of the position sensor;

FIG. 5 is a schematic view of a third embodiment of the position sensor showing an optimum configuration (b) where a beam stop or knife edge are not necessary;

FIG. 6 shows a schematic view of a closed loop system for assessing the limiting performance of the technique of the invention of the third embodiment;

FIGS. 7(a)-7(c) and 8(a)-8(c) show photographs of the diffraction edge patterns for various test pieces taken by projecting the image from the microscope eyepiece onto a frosted glass screen placed some 700 mm distant;

FIG. 9 shows the results of a test on a diamond tool having a radius of 1.524 mm whereby sub-micron resolution is possible; and

FIGS. 10(a)-(c) show photographs of a diamond machine tool edge for different light sources.

By way of example only, the position sensors illustrated in the Figures are designed to be attached to the tool slide of a computer controlled contact lens lathe to allow the vertex position of the concave inner surface of a contact lens to be determined with sufficient accuracy to enable a lens of the correct thickness to be produced. A lathe of this type is of conventional construction, including a chuck for holding a workpiece and a toolslide carrying a tool for machining a workpiece held in the chuck. In the manufacture of contact lenses, a cylinder blank of plastics material is initially machined to provide the inner concave surface of the lens. The concave surface of the lens is then attached to the convex surface of a brass holder known as an arbour using a meltable wax adhesive, and the remainder of the blank is machined to provide the convex outer surface. The machined lens is then removed from the arbour.

The sensors illustrated in the Figures are designed to provide to the computer controlling the lathe a data signal representative of the vertex position of the concave inner surface so that the computer may determine the appropriate path for the tool 10a to follow when machining the lens to ensure that the finished lens is of the correct thickness.

In FIG. 1, a partially formed contact lens blank 10 is shown attached to the convex end surface of a brass arbour 11. The contact lens blank 10 is of transparent plastics material. The other end of the arbour 11 is gripped by the chuck (not shown) of a lathe (not shown) for rotation about axis `A`.

The position sensor comprises a Helium-Neon (He--Ne) Laser 12, a knife edge 13, a microscope 14 and a photodiode 15 all secured to the tool slide 23 of the lathe. The tool slide is movable backwards and forwards parallel to axis `A`.

The laser 12 emits a collimated beam 16, typically 2 mm wide, along a sensing axis B which lies perpendicular to axis A and intersects it within a sensing region 17. The microscope 14 is aligned with sensing axis B and includes an .times.5 objective lens system 18 for producing an image of the intersection of axes `A` and `B` at an image plane 19. A slit 20 is also located at the image plane to select only a slot region extending in a direction orthogonal to axes A and B. The image at slit 20 is then passed on to the photodiode 15 by means of an .times.10 eyepiece system 21 of the microscope.

Adjacent the outer lens of the objective lens system 18 is located the knife edge 13 which extends orthogonal to axes `A` and `B` and which is positioned so that it protrudes into the beam 16 just sufficiently to prevent it from reaching the microscope when there is nothing in sensing region 17 to cause the beam to deviate.

In use, the tool slide is moved towards the arbour in the direction of axis `A` while the output of the photodiode 15 is monitored to record the intensity of this incident radiation. When there is no material in the sensing region, the incident radiation is at a low background level because it is almost wholly obscured by the knife edge 13. When, however, the outer end region of the blank 10 is adjacent or coincident with the beam 16, a surface diffraction effect causes the beam to deviate sufficiently to cause an increase in intensity of the radiation incident upon photodiode 15, and consequently its output rises to indicate the presence of the surface. Continued movement of the tool slide towards the chuck causes the laser beam 16 to impinge on the blank 10 intermediate its ends so that no surface effects are apparent. Thus, there is no deviation of the beam 16, and the intensity of this radiation incident upon the photodetector is therefore relatively low. When, however, the beam 16 approaches the convex interface with the concave surface of the lens blank 10 and the wax adhering the blank to the arbour 11 approaches the beam 16, the variation of intensity incident on photodiode 15 with its axial position is as shown in FIG. 2. Referring to this figure, it will be noted that the intensity reaches a sharply-defined symmetric peak. By monitoring the output of the photodiode 15, the presence of the interface, and thus the position of the vertex of the concave surface of the blank can be determined precisely. In tests carried out, it was found that the position of the vertex could be determined to an accuracy of within .+-.3 microns using a laser beam of width 2 mm.

The output from the photodiode 15 may be monitored by supplying it to a meter (not shown) and then adjusting the position of the arbour to obtain maximum intensity and then reading off or marking the longitudinal position of the tool slide with respect to the lathe bed. Alternatively, the position may be determined by recording the longitudinal position of the tool slide in relation to the intensity detected by the photodiode as the tool slide is moved towards the arbour. In this case, the data position may be determined as before by sensing the position of maximum sensed intensity or by noting the position of two equal intensity points on either side of the maximum and then calculating the mid-point which will correspond to the position of the maximum.

The technique described above allows a sharply defined data point to be fixed with respect to the lathe bed which corresponds to the axial position of the vertex of the concave surface of the partially formed lens blank 10. The lathe can then be calibrated to enable lenses of a particular thickness to be manufactured without the need manually to measure the axial length of the partially formed blank. Calibration may be effected by cutting the outer concave surface of a lens with the tool slide in approximately the correct longitudinal position, measuring the actual thickness of the lens produced and calculating the offset (if any) required to produce a lens of the required thickness.

FIG. 3 shows the view seen by the photodiode with the slit removed; the position of the edges of the slit are shown in dotted lines. The Figure illustrates the position of the blank 10 and arbour 11 viewed by the microscope when the tool slide is at the data position. A surface diffraction effect causes a large part of the concave interface to glisten. The slit 20 selects only that part of the image which is influenced by the presence of the interface. This enables the signal to noise ratio of the device to be relatively high.

In FIG. 4, the second embodiment of the position sensor is shown. It has a similar form to the first embodiment, and similar items will be given similar references.

The position sensor of FIG. 4 comprises a mounting bracket 30 attached to the tool slide of the lathe. The mounting bracket supports the Helium-Neon laser 12, first and second turning mirrors 31, 32, each for deflecting the beam 16 emitted from the laser 12 through 90.degree., and the microscope 14. The microscope 14 mounts at its front end a knife edge 13 arranged, as in the previous embodiment, to block the beam 16 when it is undeviated. The microscope includes an .times.5 objective lens system 18, an .times.10 eyepiece lens system comprising a field lens 33 and an eye lens 34, a slit 20 located between the eye lens 34 and the field lens 33, and a photodiode 15 to measure the output.

As before, the mounting bracket is moved towards the arbour 11 while the output from photodiode 15 is monitored, and the data position is set at the longitudinal position corresponding to the sharply defined symmetrical peak shown in FIG. 2.

FIG. 5(a) shows the basic configuration of the third embodiment of the position sensor of the invention. A randomly polarized low power He--Ne laser 50 (approximately 1 mW) incident on the test piece (edge) 51 produces the edge diffraction imaged by an objective lens 52 through a slit 53 to an eyepiece 54 and a photodiode detector 55. The non-diffracted through beam is stopped by an opaque disc (beam stop) 56 or edge placed on the objective. The microscope view to the naked eye is shown in FIG. 5(b).

Laboratory tests have indicated that the optimum configuration of the elements is as shown in (b) of FIG. 5(a), where the laser and/or detector end are rotated about the test piece vertical axis to remove the need for a beam stop 56, thereby improving image resolution. In accordance with such an arrangement, input .theta. and output .phi. angles are limited by the surface optical quality, but the technique may be used for an angle up to 45.degree. with minimal degradation of the output peak for aluminum and brass turned surfaces. Moreover, in the case of a knife-edge 51, these angles may be extended up to 90.degree. (either .theta. or .phi.=90.degree.).

A further gain in resolution will also result if the input beam is focused onto the edge 51 by focusing optics 57. This gain results from the definition of a smaller area on the edge 51 and a higher beam intensity. In addition, increasing the numerical aperture (NA) of the output objective improves the fringe (i.e., reduces electromagnetic interference) and hence position resolution. This results in a reduction in the width of the first bright maxima from approximately 12-15 .mu.m to full resolution on the first fringe at approximately 1.5 .mu.m. The invention of this embodiment has been tested for `remote` range test distances of up to 8.23 meters by using a telescope in series with the microscope to provide the required resolution.

FIG. 6 shows a closed loop system 100 for assessing the limiting performance of the technique set forth above, and such a closed loop system may be used, for example, for diamond tool profile measurement tests. The He--Ne laser 110 output beam is dithered by the piezo controlled mirror (FLAT) so as to modulate the beam about the test piece edge. The beam then passes through an input objective 120 focused on the tool edge, and the diffraction intensity distribution is then relayed by output optics 140 to a photodiode detector 130 via a slit. The output of the detector is then fed into a phase sensitivity detector 150, the phase sensitivity detector 150 being referenced to the dither mirror frequency. The phase sensitive detector 150 locks onto the AC signal minimum output from AC signal generator 190 that occurs with a peak signal of the photodiode. Any deviations from this position result in error signals which are fed out to the integrator 160 which provides a DC offset signal with a set time constant to the high voltage DC operational amplifier 170. Amplifier 170 returns the mirror FLAT to the nominal position by providing a drive signal from piezoelectric (PZT) driver 180, thus completing the loop. The "error" signal from the integrator 160 is in this instance a measure of the tool position or surface profile if the tool were rotated about its mean radius of curvature. Of course, similar circuitry may be used for closed loop control, and other types of measurements may also be conducted.

FIGS. 7 and 8 show photographs of the diffraction edge patterns for various test pieces taken by projecting the image from the microscope eyepiece 54 onto a frosted glass screen placed some 700 mm distance from the eyepiece. The test pieces are to the left of the pictures. The knife-edge pattern shows the symmetric intensity of the fringes about the maximum. It can be seen that increasing the test piece radius has the effect of introducing an asymmetry which apparently varies in proportion to the test piece radius.

In addition, the visual impression for the viewer is confirmed in the photographs. Namely, for laser illumination the bulk of the diffraction energy appears to be channeled into the first few fringes such that for a low resolution .times.5 objective, the bright rim image is seen as a single bright fringe in the case of knife-edge test pieces.

The use of a very high power .times.63 input objective does not improve the resolution, and instead, the optimum resolution may be obtained by working at an input angle of approximately 25.degree. with an .times.5 objective while using an .times.10 (0.25 NA) objective to relay the image to the slit plane. This results in reduction of the full width half maximum (FWHM) to approximately 3-4 .mu.m with a 25 .mu.m slit while the use of a 5 .mu.m slit on a diamond tool may produce the same result with an FWHM of approximately 11/2 .mu.m, i.e, on the order of the first fringe FWHM. In addition, for a given edge and optical system the slit size does not appear to have any marked effect on the output peak FWHM, although the radius of the test piece does.

Use of an input objective to define a smaller target area and to provide a beam on the order of NA to the output optics has been demonstrated to provide an improved resolution performance when matched to the output optics NA with an .times.5 (0.12 NA) objective providing the most practical solution. In addition, an .times.10 (0.25 NA) output objective has been found to be the best compromise between resolution and working distance (which is on the order of 7 mm) for the output optics, and in theory such an objective should provide a full resolution of the diffraction pattern around the first bright fringe.

This technique can also been extended with suitable optics to ranges on the order of meters. For a remote test a basic system consisting of a light collecting lens and a slit only at a distance of 11/2 meters has been found to have a resolution .ltoreq..+-.5 .mu.m. In addition, an arrangement with a large collecting doublet and a microscope image magnifying system may also have a resolution <.+-.5 .mu.m, but in addition, such a system shows a high degree of sensitivity to atmospheric turbulence in the beam path in front of the doublet. The output of this system when displayed on an oscilloscope shows the frequency components of the beam turbulence (mostly in the 10 Hz to 2 kHz range) when the beam is disturbed at various distances from the doublet. Accordingly, atmospheric noise input should be considered when using a sensor arrangement at these ranges in an open air environment.

Further tests carried out on a diamond tool of radius 1.524 mm have indicated that sub-micron resolution is possible and that a sensor using a knife-edge may be capable of locating and operating about the peak maximum, i.e., the modulation minimum, as shown in FIG. 9.

An additional application of the technique for diamond tool and edge profiling is generally suggested by the photographs of FIG. 10. FIG. 10 shows a diamond machine tool edge measured in accordance with the invention. FIG. 10(a) shows the diamond machine tool edge with incoherent (0.88 .mu.m) illumination, where FIG. 10(b) shows illumination with the raw input of an He--Ne laser source. FIG. 10(c), on the other hand, shows a high power .times.400 image processed profile made using an intellect software package. The combination of the edge sensing technique with such intellect image processing software can be used to provide rapid analysis of profile roughness and curvatures in accordance with the invention. In addition, the magnification of the .times.400 view of FIG. 10(c) may be increased to provide details on the scale of .lambda..

The third embodiment of the invention described above may be mounted in an assembly similar to that described with respect to FIGS. 1 and 4. In addition, the third embodiment of the invention may be modified so as to employ a dual wavelength system such as a Krypton laser line pair at 676 nm and 647 nm such that the resulting diffraction pattern might be able to provide an edge line width .DELTA.S.perspectiveto.V.sqroot..lambda..sub.1 -.lambda..sub.2, where in this case .DELTA.S is approximately equal to 20 nm. By using the intellect image processing software to obtain frame difference intensities for .lambda..sub.1 and .lambda..sub.2, the resolution might go some way towards the theoretical limit for the edge profile thickness.

Moire techniques may also be used in accordance with the present invention in conjunction with a two edge system, and the relative orientation of the edges may be obtainable from the resulting Moire fringes as well as the edge separation. The limits of these techniques are set by laser noise in the form of spurious fringe patterns from dust particles and edge artifacts. However, such a system results in rather low resolution.

Finally, a two-color edge variant may be applied to the present invention by using .lambda.=633 nm and .lambda.=546 nm beams incident on opposite sides of the test piece and focused to a common spot on the detector plane so that any movements in the test piece in the X-Y plane may show up by a color change at the detector (assuming a fixed vertical axis orientation). Such a technique can be applied to remotely measure test piece sizes and positions along two axes with a resolution approximately equal to the basic single .lambda. technique described above.

The edge diffraction techniques discussed above are best suited to special applications that call for remote sensing and in particular those required to be carried out within optical media. However, it will be appreciated that the present invention may have numerous applications outside of the optical media. For example, a different radiation source such as X-rays may be used. The main advantage of this technique compared to the large number of commercial sensors and transducer systems capable of sub-micron operation is the cost. The basic system of the present invention is relatively inexpensive, and in addition, the availability of the laser diode source should reduce both cost and dimensions of the sensor units.

Although the invention has been described in connection with several sample embodiments, it should be understood that the invention is not limited to such embodiments. The invention is intended to cover all modifications and equivalents as may be included within the scope and spirit of the invention as defined by the appended claims.