GB2175687A - Interferometric-incremental device for testing flatness - Google Patents

Abstract

The test object having the surface (11) whose flatness is to be checked is disposed on a rotatably mounted object stage (16) and an interferometer I is mounted on a precision translatory or rotatable guide (18) for incremental scanning, in such a way that the measurement beam (IM) of the interferometer is directed perpendicularly to the test surface (11). The interferometer comprises a monochromatic laser beam 4, interferometric divider 7 and beam divider 2 and has either a diaphragm 3 located between diodes 2 and 7 or has photoelectric detectors whose active surfaces are disposed relative to one another and to the beam dividers in such a way that they are only impinged upon by homologous beam pairs. The interferometer also incorporates integrated CCD arrays 38, 40. <IMAGE>

Description

SPECIFICATION Interferometric-incremental device for testing flatness The present invention relates to a device for testing flatness, in particular to a device for the interferometric testing flatness, more particularly of polished surfaces having diameters of up to 400 mm.

It is often necessary in such cases to scan the surfaces in a non-contact manner in order to avoid damaging these surfaces, which may include, for example, planar reflectors, silicon wafers, disc stores or metallic or ceramic planar lapped parts.

Interferometric devices for testing flatness are already known. In a first group of such devices (Feinwerktechnik+Messtechnik 82 (1974), No. 7, pages 353-354; German Offenlegungsschrift No. 2 406 184), a triangular glass prism is used whose hypotenuse surface is a highly planar surface and thus serves as a flatness reference. The surface whose flatness is to be tested is placed at a slight distance of approximately 1 mm from this hypotenuse surface. A monochromatic ray bundle enters the glass prism through one of the short sides and falls upon the hypotenuse, where one part is reflected and the other part passes through the hypotenuse, is refracted and strikes the surface to be tested at a relatively large angle of incidence.The beam is then repeatedly reflected backwards and forwards between the surface to be tested and the reference surface, and multiple-beam interferences are formed.

A major disadvantage of this type of method is that the interference structure produced is the result of multiple-beam interferences and it is thus not possible to interpret the information contained in the interference figure point-by-point, but rather it is always the result of an areal integration of the mutually interfering beams. The resulting possible misinterpretations of the surface structure using the interference structure can only be kept to a minimum if the gap between the reference surface and the surface to be tested is small. The fact that this condition must be met considerably impairs the useability of this type of device with respect to technology.A further disadvantage of these methods resides in the fact that, as a result of the oblique incidence of light onto the surface to be tested, the height difference to be assigned to the order clearance is A 2 In a second group of methods (German Auslegeschrift No. 25 37 162 and German Auslegeschrift No. 26 36 211) diffraction gratings are used which are disposed in a plane parallel to the surface whose flatness is to be tested. This produces interference between the diffraction order directly reflected at the diffraction grating and the radiation passing through the diffraction grating, being reflected at the surface to be tested, passing through the diffraction grating and then again being diffracted.The disadvantage of German Auslegeschrift No. 25 37 162 is that the necessary gap of approximately 250 Am between the grating plane and the surface whose flatness is to be tested is utenably small for technical applications. This disadvantage is partly overcome in German Auslegeschrift No. 26 36 211, in which larger gaps between the test surface and the grating plane are permissible. However, the intensity of the interference pattern alters with the variation in the gap between the grating plane and test surface in accordance with a periodic beat, and in the beat maxima the height difference to be assigned to the order clearance is approximately 2 and in the beat minima, however, it is approximately 4 In order to be able to interpret the interference pattern correctly, the gap between the grating plane and the test surface must therefore be known.Furthermore, the interference structure generated by this method is also the result of the superposition of four beams, three of which contact the test surface at different points. The larger the gap between the test surface and the grating plane, the greater the areal integrating character of the interference structure.

In a third method according to German Auslegeschrift No. 26 36 498, a diffraction grating is again used which is disposed in this case perpendicularly to the surface whose flatness is to be tested. One part of the monochromatic, coherent and parallel light bundle used to test the flatness falls directly onto the diffraction grating in such a way that the first order spreads out normally in relation to the grating plane. The other part of the light bundle impinges the surface to be tested, receives the phase information of the surface structure and is reflected at the diffraction grating, and the direction of the first diffraction order generated on striking the diffraction grating also falls normal to the diffraction grating. Thus the interference of the first two orders takes place in the normal direction of the diffraction grating.

The present arrangement eliminates major disadvantages of the above-mentioned inventions, although the height difference to be assigned to the order clearance is relatively low as a result of the oblique incidence.

All methods of testing flatness, including those not mentioned above, have a further disadvantage. This is due to the fact that the randomness of the surface structure of the test surface is reflected in the randomness of the obtained interference structure. This means that the interference structure can be linear, circular, elliptical or any other shape and, in particular, the order clearance can be altered within the interference structure from point to point It is also possible to view an interference structure as an optical grating and the order clearance as a grating constant. When testing flatness, this means that the grating spacing changes from one point of the surface to be tested to the next.This is the reason why it has not been possible to apply the incremental method, which has successfully been used in linear and angular metrology, to any of the known methods of testing surfaces interferometrically, as it is a prerequisite that the selected grating constant be invariable.

For this reason, in the known methods of flatness testing, the interference pattern containing the surface information is formed on multidiode arrays or multidiode vidicons and is discretely scanned point-by-point. This reveals the limitations of all the interferometers functioning with this kind of scanning of the interferogram plane. As soon as the fringe spacing in the interferometer becomes smaller than the scanning space, the measured phase ambiguously becomes modulo 2K. This limitation becomes noticeable when measuring surfaces with large phase gradients in narrow zones.

An interferometer, in particular for the incremental scanning of variable interference structures, has already been proposed in which a diaphragm is located between the interferometer divider and the beam divider, and the beam is divided by the diaphragm in the beam divider into partial beams, and a photoelectric detector is disposed in the beam paths of each of these partial beams. Furthermore, a first beam splitter and a second beam splitter are disposed between the interferometer divider and the beam divider, the beam entering the second beam splitter is split into partial beams, a linear arrangement of integrated photoelectric scanning elements is disposed in the beam paths of each of these partial beams, and these two lines are disposed perpendicular to one another.

It is an aim of the present invention to provide a device for testing flatness with which it is possible to measure relatively large surfaces with a simple interferometric structure, high resolution of linear measurement, large phase gradients in narrow zones and as dense a number of measuring points as required in an optically contactless manner. It must be possible to interpret the interferometric surface information obtained in a strictly point-by-point manner, and the measurement error should be at a minimum. Accuracy of measurement and resolution capability should not be impaired by large gaps between the interferometer and the test surface.

It is a further aim of the present invention to provide an interferometric-incremental device for testing flatness which can be used to test surfaces having diameters of up to 400 mm with a height difference to be assigned to the order clearance of exactly A 2 The gap between the test surface and the interferometer can be of any dimension within broad limits, but at least up to 50 mm. The maximum gradient of the surface profile can be 20 minutes and more.

In accordance with the present invention, there is provided a device for checking the flatness of a surface of an object, the device comprising a rotatably mounted object stage for mounting the test object which has the surface whose flatness is to be checked, an interferometer for incremental scanning of variable interference structures, a monochromatic laser beam source for the interferometer, the latter having a diaphragm between an interferometer divider and a beam divider or having photQelectric detectors whose photoelectrically active surfaces are disposed relative to one another and to the beam-dividing and beam-reflecting elements in such a way that they are only impinged by homologous beam pairs, the interferometer also having CCD arrays integrated therein and a precision guide opposite the object stage, on which guide the interferometer is mounted such that the measurement beam of the interferometer is directed substantially perpendicularly to the surface whose flatness is to be tested.

In one embodiment, the guide opposite the object stage is a rotatable precision bearing and in another embodiment, the guide opposite the object stage is a translational precision bearing.

The insertion of a diaphragm between the interferometer divider and the beam divider in the interferometer for incremental scanning of variable interference structures enables the interference structure to be scanned point-by-point, whereby changes in the grating constant of the interference structure during measurement do not interfere with the acquisition of incremental measurement signals. The same effect is achieved if the photoelectrically active surfaces of the photoelectric detector scanning the incremental measuring signals are disposed relative to one another and to the beam-dividing and beam-reflecting elements in such a way that they are only struck by homologous beam pairs.

If the test object having the surface whose flatness is to be tested is disposed on a rotatably mounted object stage and if a translatory guide is disposed above this object stage, with an interferometer for incremental scanning of variable interference structures secured to the object stage, in such a way that the measurement beam is directed perpendicuarly to the test surface, then, as the object stage turns and the interferometer simultaneously makes a translatory movement beginning at the centre of the rotating stage and directed towards its edge, the measurement beam describes an Archimedean spiral on the test surface, beginning in the middle and continuing towards the outside.If the rotating bearing for the revolving stage is disigned such that it rotates in a play-free and tilt-free manner, and if the translatory guide for the interferometer fulfils the same conditions, the joint action of the translatory movement of the interferometer and the rotary movement of the object stage causes a plane to be fixed in a notional sense above the surface to be tested which assumes the function of the planar reference surface. By mutually coordinating the angular velocity of the object stage and the linear speed of the interferometer movement, the test surface can, if necessary, be scanned completely and, from a metrological point of view, in a precise point-by-point manner, with the measurement beam being directed exactly perpendicular to the test surface. Thus, basic metrological requirements for testing flatness are strictly fulfilled.

By way of example only, specific embodiments of the present invention will now be described, with reference to the accompanying drawings, in which: Figure 1 is a first embodiment of interferometric-incremental device in accordance with the present invention, for testing flatness, in which the interferometer has a translatory guide; and Figure 2 is a second embodiment of interferometric-incremental device in accordance with the present invention, for testing flatness, in which the interferometer has a rotating guide.

In Fig. 1, the test object 5 with the surface 11 to be tested is disposed on an object stage 16. The object stage 16 is carried by a precision bearing 17 in a form of a rotating bearing.

Precision bearings include, for example, pneumatic or hydraulic bearings. A precision guide 18 is disposed opposite the object stage 16 and is in the form of a translatory guide. The translatory precision guide 18 is preferably pneumatic. A push rod 19 guided in the precision guide 18 is connected to a base plate 6 on which the interferometer I for incremental scanning of variable interference structures is disposed. The expanded laser beam 4 is divided in the interferometer divider 1 at the divider layer 7 into measurement beam 1M and reference beam 1R The reference beam 1R falls on to the reference reflector 12 which is in the form of a planar reflector, and the measurement beam 1M falls onto the surface 11 whose flatness is to be tested.Following reflection of both beams, they are rejoined at the divider layer 7, whereby the interference structure 20 is formed. Between the first beam splitter 30 and the beam divider 2 there is a diaphragm 3 through which the interference structure 20 is scanned point-by-point. The bundle 13 passing through the diaphragm 3 is divided in the beam divider 2 into parial bundles 14 and 15 which are passed on to photoelectric detectors (not shown). As a result of the point-by-point scanning of the interference structure 20 through the diaphragm 3, an invariance in the acquisition of measured values is obtained relative to changes in the interference structure 20 of any kind as caused by a surface profile. Consequently, the optical signals of the partial bundles 14 and 15 are exactly in-phase.The 90" phase shift required for the incremental method is obtained from the optical signals 14 and 15 using polarizing optical means (not shown). These optical signals, too, are, independent of the shape of the surface profile, phase-constant at any moment with regard to their 90" phase position.

The beam entering the first beam splitter 30 and modulated with the interference structure 20 is split in the first beam splitter 30 into beams 32 and 33, and beam 32 enters a second beam splitter 34, on whose divided layer it is split into beams 35 and 36. Beam 36 falls on the CCD line 38 and beam 35 on the CCD line 40. The linear arrangements of the photoelectric scanning elements on the CCD lines 38 and 40 are directed perpendicular to one another. Thus the surface structure in each scanning point can be of any shape whatever, as the phase gradient in each case can be accurately determined.

Fig. 2 shows an arrangement in whcih the interferometer I is rotationally guided above the test surface 11. The bearing pin 21 is pivotally mounted in the rotatably precision bearing 22 and an arm 23, which carries the laser 24 and the interferometer I for scanning variable interference structures, is secured to the bearing pin 21. Fig. 2a) shows a plan view of the test surface 11 of the arrangement and Fig. 2b) shows the arrangement in the direction of arrow A. The interferometer I in Fig. 2 is identical to the interferometer I in Fig. 1.

If the object stage 16 and the test object 5 in the arrangements in Fig. 1 and Fig. 2 are rotated by a suitable drive, and if the interferometer I is moved over the surface 11 to be tested either by means of the translatory precision guide 18 or the rotating precision bearing 22, the point of incidence of the measurement beam 1M on the surface 11 describes a spiral, and, by coordinating the two relative movements, the pitch of the spiral can be varied and the sequence of measurement points adapted to the task in question.

In accordance with the prior art, precision guides 18 and precision bearings 17 and 22 can be made which have a guide accuracy which meets the interferometric requirements. Also, it is not necessary in the present case for the rectilinear movement caused by the movement of the precision guide 18 or the circular movement caused by the rotation of the precision bearing 22 to take place in strict parallel with the surface of the object stage 16, because a certain angular deviation can be ascertained either by measurement against the object stage 16 itself or against a plane-parallel plate placed thereon and can, if necessary, be subtracted in further measurements.

List of reference numerals used - -- interferometer for incremental scanning of variable interference structures 1M - measuring beam 1R - reference beam 1 - interferometer divider 2 - beam divider 3 - diaphragm 4 - monochromatic laser beam 5 - test object 6 - base plate to which interferometer i is secured 7 - divider layer in interferometer divider 1 8 - divider cube 9 - partial bundle at exit of divider cube 8 10 -- partial bundle at exit of divider cube 8 11 - surface whose flatness is to be tested, test surface 12 -- reference reflector 13 - beam through diaphragm 3 14 - partial bundle in beam divider 2 15 - partial bundle in beam divider 2 16 - object stage 17 -- rotatable precision bearing for rotating object stage 16 18 -- translatory precision guide 19 - push rod 20 -- interference structure at exit of interferometer divider 1 21 - bearing pin 22 -- rotatable precision bearing for rotating the interferometer I 23 - arm 24 - laser 30 -- first beam splitter 32 - partial beam in first beam splitter 30 34 - second beam splitter 35 - partial beam in second beam splitter 36 - partial beam in second beam splitter 38 - CCD line 40 -- CCD line

Claims (4)

1. A device for checking the flatness of a surface of an object, the device comprising a rotatably mounted object stage for mounting the test object which has the surface whose flatness is to be checked, an interferometer for incremental scanning of variable interference structures, a monochromatic laser beam source for the interferometer, the later having a diaphragm between an interferometer divider and a beam divider or having photoelectric detectors whose photoelectrically active surfaces are disposed relative to one another and to the beamdividing and beam-reflecting elements in such a way that they are only impinged by homologous beam pairs, the interferometer also having CCD arrays integrated therein and a precision guide opposite the object stage, on which guide the interferometer is mounted such that the measurement beam of the interferometer is directed substantially perpendicularly to the surface whose flatness is to be tested.

2. A device for checking flatness as claimed in Claim 1, wherein the guide opposite the object stage is a rotatable precision bearing.

3. A device for testing flatness as claimed in Claim 1, wherein the guide opposite the object stage is a translational precision bearing.

4. A device for testing flatness, substantially as herein described, with reference to, and as illustrade in Fig. 1 or Fig. 2 of the accompanying drawings.