EP2920544A1

EP2920544A1 - Optical measuring method and measuring device having a measuring head for capturing a surface topography by calibrating the orientation of the measuring head

Info

Publication number: EP2920544A1
Application number: EP13814625.3A
Authority: EP
Inventors: Martin SCHÖNLEBER; Berthold Michelt; Matthias Kunkel
Original assignee: Precitec Optronik GmbH
Current assignee: Precitec Optronik GmbH
Priority date: 2012-11-15
Filing date: 2013-11-14
Publication date: 2015-09-23
Also published as: US20170234678A1; CN104797903B; CN104797903A; US9677871B2; US9982994B2; WO2014076649A1; US20150260504A1; DE102012111008A1; DE102012111008B4

Abstract

The invention relates to an optical measuring method for capturing a surface topography (1) of a measured object (2). To this end, a measuring device (3) is provided, comprising a measuring head (4) in a measuring head guiding device (5), for chromatic confocal capture of the surface topography (1) or for a spectral interferometric OCT distance measurement to the surface topography (1). Spectral light of a light source (6) is applied to the measured object (2) from a fibre array (7) comprising i fibres (8) from i measurement spots (12 to 15) via a common measuring head lens (10), forming a spot array (11) from i measurement spots (12 to 15). The i reflection spectra of the i measurement channels are then captured and digitalised. The digitalised reflection spectra are evaluated by calculating temporal variations of systematic measurement errors and temporally induced deviation movements of the measuring head guiding device (5), comprising the following steps: - capture of geometric distance values (a, b, c) of the i measurement channels and of the three-dimensional position values for the i measurement spots on a measured object surface at the time t(j); - capture of a local inclination of the measured object surface (16) relative to the measuring head (4) comprising at least three measurement spots (12, 13, 14) of a triangle (17), which are projected onto the measured object surface (16) for correction of the measurement values; - correlation of the local topographies by separating temporally induced deviation movements of the measuring head guiding device (5) by means of a three-dimensional acceleration sensor on the measuring head (4); - creation of the correct local topographies.

Description

description

OPTICAL MEASURING METHOD AND MEASURING DEVICE WITH A MEASURING HEAD FOR DETECTING SURFACE TOPOGRAPHY BY CALIBRATING THE MEASUREMENT HEAD ORIENTATION

The invention relates to an optical measuring method for detecting a surface topography of a Messob ektes. For this purpose, a measuring device with a measuring head in a measuring head guiding device for detecting the surface topography is provided.

Such a measuring device for measuring a surface is known from the document DE 10 2008 041 062 AI. The known measuring device generates a measuring light beam which, after passing through at least three separately focusing optical components, impinges on the surface of the object, is reflected therefrom and is detected by a spatially resolving light detector together with reference light after interfering superimposition.

For this purpose, the known measuring device to an optical assembly, which comprises the at least three separately focusing opti ^¬ cal components. The main axes of these separately focusing optical components are offset from each other and arranged side by side. Moreover, the known measuring device has a beam splitter arranged in a beam ^{path of} the measuring light beam. In addition, a reference surface is provided for the known device and a spatially resolving light detector.

The light source, the beam splitter and the optical assembly are arranged relative to each other such that emitted from the light source and the focusing optical components passing measuring light hits the surface and is reflected by this and hits the focusing via the focusing optical components on the detector. In addition, the known measuring device has an evaluation system for receiving image data from the spatially resolving light detector and for outputting measurement data representing a surface shape of the surface. For this purpose, distance values representing a distance of a location of the surface from the focusing optical components are detected. From these distance values, the evaluation system forms parameters which represent the surface shape of the surface.

In addition, the above document discloses a method for measuring a surface of an object, which essentially comprises subsequent method steps. First, a measuring light is generated. From the measuring light, three converging partial beams of a first part of the measuring light are formed in order to illuminate three spaced apart areas of the surface of the object. The reflected light and the three partial beams of light reflected from the surface of light are directed together with a second part of the measurement ^¬ light on a position-sensitive detector to form therein interference. These interferences are finally analyzed by a detector detecting light intensities to represent the surface shape of the surface of the object by corresponding measurement data.

A disadvantage of the known device is that it requires a large amount of space due to their separate and adjacent optical components. Another disadvantage is that no precautions are taken, systematic measurement errors, long-term changes in the absolute distance measurement or deviations from a desired path on which the measurement object is guided in the known embodiment, to take into account in any form during the evaluation or to correct the measurement results accordingly. Consequently, the known measuring system is not able to provide reliable data, in particular in the nanometer range, in order to detect surface topographies in such dimensions.

Chromatic confocal distance measurement is understood herein to mean a method that utilizes the effect that lenses have different focal points for different wavelengths of light. The chromatic-confocal distance measurement uses the dispersion of spectrally broadband light in an optical imaging system to determine the distance of a reflective surface to the measuring head. A spectrally broadband point light source, which is normally realized by a first pinhole or an optical fiber end, is focused on the object with the optical imaging system. The distance of the focus from the imaging system is a unique, well-defined function of the wavelength. The reflected light is imaged again on the same imaging system and coupled out from the illumination beam path and imaged on a Lochblen de, which is arranged at the mirror point of a beam splitter. Alternatively, the reflected light can also be returned directly into the first pinhole and then decoupled. A detector behind the pinhole then determines the dominant wavelength of the reflected light. From the knowledge of the focal lengths of the individual wavelengths, the object distance can be determined directly from the dominant wavelength. An advantage of this method is the lack of moving components. In a preferred embodiment, the light of the light source is coupled into an optical waveguide, passes through a fiber coupler and emerges from the measuring head at a fiber end. The light returning from the measurement object re-enters the fiber end and is branched off at the fiber coupler in the direction of the detector. The fiber end simultaneously forms the light point for the object illumination as well as the pinhole for the filtering of the measuring light.

Optical Coherence Tomography (OCT) is the name given to an investigation method in which spectrally broadband light is used to measure the distance of objects with the aid of an interferometer. The examination object is scanned point by point. In this case, an arm with a known optical path length is used as a reference to a measuring arm. The interference of the partial waves from both arms then gives a pattern from which one can read the difference of the optical path length of the two arms. A distinction is made here between two spectral interferometric measuring and evaluation methods, the Time Domain OCT and the Frequency Domain OCT, which is why one speaks of the signal in the time domain (TD) and the signal in the frequency domain (FD). Simply put, this means that either the reference arm varies in length and continuously the intensity of the interference measures, without referring to the spectrum consideration (time do ^¬ main), or the interference of the individual spectral Comp ^¬ components detected (frequency domain) ,

An object of the invention is to provide an optical measuring method for detecting surface topographies of a measuring object, with which the measuring accuracy up to the Nanometer range is improved, and to create a suitable measuring device.

This object is achieved with the subject matter of the independent claims. Advantageous developments emerge from the dependent claims.

In a first aspect of the invention, an optical measuring method for detecting a surface topography of a measurement object, in particular in the nanometer range, is provided. For this purpose, a measuring device with a measuring head is provided in a measuring head guiding device for a chromatic-confocal detection of the surface topography or for a spectral interferometric OCT distance detection to the surface topography.

First, spectrally broadband light of a light source from a fiber array with i fibers of i measurement channels is applied to the measurement object via a common measurement head optics, forming a spot array of i measurement spots. Then i reflection spectra of the i measurement channels are acquired and digitized. Then the reflection spectra of each measurement channel are evaluated separately and a distance value is determined. Thereafter, the totality of the distance values is evaluated in combination with different measurement channels and times in order to eliminate temporal variations of systematic measurement errors and time-dependent deviation movements.

An advantage of this measurement method is that the displayed measured local topography measurements are checked for temporal variations of systematic errors and timing deviations of the probe head apparatus, so that a separation between a sample surface real surface topography and real information about measurement errors and deviations movements of the measuring head guide device can be done.

For this purpose, a plurality of evaluation steps are required, which require in detail a detection of geometric distance values of the i measurement channels at time t (j). In addition, detection of three-dimensional position values for the i measurement spots on the measurement object surface is carried out at time t (j). In addition, a local inclination of the measuring object surface is detected relative to the measuring head. This is followed by correlating temporal variations of systematic measurement errors based on the detected slope. This step is followed by creating local topographies for the redundant i measurement channels.

At the balance correlating the local topographies under separation of temporally-related differential movements of the measuring head guide device by separating a bumpiness of positions and orientations of the measuring head or its dependent scan of the measuring light in the different measurement channels on the basis of the measuring head guide ^¬ apparatus true of the real or Oberflächentopogra ^¬ phy with a measurement resolution in the nanometer range. Finally, there is an output an adjusted Oberflä ^¬ chentopographie and a real path and orientation of the measuring head guide device of the measuring head. These evaluation steps are determined by comparing the different samples.

In general, it is possible to scan at different clock rates or to scan in a line with the measuring spots arranged in the scanning direction at different intervals. Do this to avoid artifacts by sub-sampling.

Another technique is used to determine a local inclination of the measuring object surface relative to the measuring head by means of three measuring spots, wherein preferably the three measuring spots are arranged on the measuring object surface in an isosceles triangle. In this case, a normal vector of the triangle, which represents the local inclination, can be determined from distance values in the triangle and, subsequently, the inclination errors of the measuring head guiding device can then be determined and eliminated, for example, via an evaluation table.

A further variant for correction of the measured values is to fix a three-dimensional acceleration sensor on the measuring head guide device or on the measuring head and three-dimensionally to detect the seasonal deviate ^¬ monitoring movements in situ, with which the measured values of the surface topography measurement can be corrected accordingly.

Moreover, it is possible by means of a vector model to detect sensor movements by vectorial determination of the yawing, pitching or rolling of the measuring head on the measuring head guide device. In this case, a yaw means a pivoting of the measuring head about its vertical axis, a pitch means a pivoting of the measuring head about its transverse axis and a rolling results when pivoting the measuring head about its longitudinal axis.

Furthermore, it is possible to carry out a determination of local slopes of the measurement object surface by height difference formation between i measurement spots and integration of the entirety of the local slopes into a surface topography. This differential scanning method measures path differences between two measuring spots of a measuring head with a fiber end and the measuring spots of two focusing lenses. In this case, the optical path difference to the two measurement spots is measured by means of spectral interferometry (OCT) and from this a surface topography is derived.

Another aspect of the invention relates to a precision optical measuring device for detecting a surface topography of a measurement object, in particular in the nanometer range. For this purpose, the precision measuring device has a device with a measuring head in a measuring head guiding device for a chromatic-confocal detection of the surface topography or for a spectral interferometric OCT distance detection to the surface topography. In the measuring device, it is possible to provide spectrally broadband light sources which supply fibers with broadband light via i Y couplers. It is also possible to supply fibers with broadband light with a single broadband light source via a lXi coupler and then further iY couplers.

A fiber array with the i fibers for measuring channels is arranged in the measuring head. Furthermore, a common measuring head optics, which forms the spot array with i measuring spots on the measuring object, is present in the measuring head. Furthermore, means for detecting and digitizing i reflection spectra of the i measurement channels in i spectrometers are provided. Furthermore, the optical precision measuring device has an evaluation unit for the digitized reflection spectra for calculating out temporal variations of systematic measurement errors and time-dependent deviation movements of the measuring head guidance device. In this optical precision measuring device tion, the measuring head moves on a Meßkopfführungsvor- direction on a linear target path.

In addition to an option for multi-channel detection with i spectrometers, a spectrometer with a fiber array input and reading several spectra with a matrix CCD can also be used. In addition, it is possible to use a single temporal channel multiplexed spectrometer as described below in relation to one embodiment. As an alternative to the time multiplexer, several measurement channels can also be combined in one spectrum. For this purpose, the measuring head is designed in such a way that the distance values of the channels occupy a fixed order of priority, in that the smallest value is always detected by channel 1, the next larger value by channel 2 and so on. This method can be carried out both for a chromatic-confocal measurement with spectral peak position and for an OCT measurement with peak position in a Fourier transformation of the equalized spectrum.

However, a guide device can also be provided for the measurement object, with which the measurement object is guided along under a stationary measuring head. In both cases, un ^¬ terliegen the probe guiding devices with their Move ^¬ handy components seasonal variation movements relative to the setpoint position. Time-related system errors such as temporally variable inclinations of the measuring head relative to the orthogonal to the measuring object surface can also occur, which can be determined with the aid of the measures described above, for example by arranging three measuring spots in an isosceles triangle and the measured values can be corrected accordingly. Instead of a plurality of i reflection spectra, the i measurement channels can also be supplied to a multiplexer and recorded in a single spectrometer and then digitized.

For calculating out temporal variations of systematic measurement errors and time-related deviation movements of the measuring head guiding device from the reflection spectra, the optical precision measuring device furthermore has the following means. Means are provided which are designed to detect geometric distance values of the i measurement channels at time t (j) and means which are designed to acquire three-dimensional position values for the i measurement spots on the measurement object surface at time t (j). Further detection means serve to determine a local inclination of the measuring object surface relative to the measuring head. For detecting the local tilt, the measuring device comprises means adapted to correct for temporal variations of systematic measurement errors based on the detected tilt. Furthermore, means are provided which are formed from ^¬ for creating local topographies for the redundant i measurement channels.

Finally, means are configured for correlating the local topographies by separating the true Oberflächentopo ^¬ chromatography of seasonal variation movements of the measuring head guide device by separating a seasonal bumpiness a position and a seasonal bumpiness an orientation of the measuring head in the measuring head guide device. Finally, means are provided which are designed to output a cleaned surface topography and further means which are designed to output a real web and a real orientation of the Meßkopfführungsvorrichtung. These aforementioned means are summarized in the evaluation unit in order to determine real measured values at i positions on the measurement object from the i reference spectra and to separate temporal variations of measuring system errors and time-dependent motion deviations from the real precision measured value in order to extract a measured value in the nanoscale. Thus, an extremely precise measured value is advantageously peeled out of the measured raw data of the i spectrometers with the aid of this precision measuring device, whereby at the same time the shell already provides a real value for the size of the time-related measuring system errors and for the size of the time-dependent deviation movements of the measuring guide device.

The invention will now be explained in more detail with reference to the accompanying figures.

FIG. 1 shows schematically a flowchart of an optical system

Measuring method for detecting a surface topography according to a first embodiment of the invention;

Figure 2 schematically shows a block diagram of a Auswer ^¬ teeinheit a measuring device for Präzisionsmes ^¬ sen according to the first operation example;

FIG. 3 shows by way of example a measurement result of a calibration run for a chromatic-confocal measuring head;

Figure 4 shows a schematic diagram of a measuring device for

Measuring a surface topography according to an embodiment of the invention; FIG. 5 shows a schematic diagram of a measuring device for measuring a surface topography according to a further embodiment of the invention.

FIG. 6 shows a schematic diagram of a measuring path for explaining a further exemplary embodiment of the measuring method.

FIG. 1 schematically shows a flow diagram 50 of a nanometer-scale optical measuring method for detecting a surface topography according to a first exemplary embodiment of the invention. For this purpose, the optical measuring method is started with the start block 100.

In method step 101, a measuring device with a measuring head is provided in a measuring head guiding device for one for chromatic-confocal detection of the surface topography or for spectral interferometric OCT distance detection to the surface topography and a measuring head in the measuring head guiding device is moved over the measuring object. This is followed by the method step 102, in which an application of spectrally broadband light of a light source from a fiber array with i fibers of i measurement channels via a common measuring head optics to form a spot array of i measurement spots on the measurement object when guiding the measuring head is performed on the measurement object , Position values of the actuators and the measured values of the i measuring channels are provided with a time stamp and recorded.

The position values of the actuators can be determined as follows: a) A target value for the controlled measuring head position, possibly taking into account reproducible time-related path deviations, which were determined in a calibration run, is used.

b) Real position values of actuator elements measured with encoders are used. Encoder values are based on glass scales or strain gauges or optical interferometers.

c) Measured values of acceleration sensors

integrated and correlated with the position values determined under a) and b).

d) Distance differences between two measuring points are measured directly and integrated into a topography that is not distorted by jumps in the measuring head.

In method step 103, acquisition and digitization of i reflection spectra of the i measurement channels takes place. Finally, in the box framed by a dot-dash line, a multi-unit process step 104 is carried out for evaluating the digitized reflection spectra, excluding temporal variations of systematic measurement errors and time-dependent deviation movements of the Meßkopfführungsvorriehtung.

This evaluation in method step 104 comprises a method step 105 for acquiring geometric distance values of the i measurement channels at time t (j). Subsequent method step 106 provides detection of three-dimensional position values for the i measurement spots on the measurement object surface at time t (j). The process step 107 may then follow, in which a detection of a local inclination of the Measurement object surface relative to the measuring head takes place, and then the evaluation process can proceed to the method step 108, in which a correlation of measured variations based on the detected slope temporal variations of systematic measurement errors is performed.

Method step 109 is used to create local topographies for the redundant measurement channels. This step is followed by method step 110 with correlating the local topographies, in which a separation of time-dependent deviation movements of the measuring head guiding device by separating a bumpiness of a sensor selection and a bumpiness of a sensor orientation of the measuring head in the Meßkopfführungsvorrichtung from the true surface topography. Finally, in method step 111, an adjusted surface topography and a real path and orientation of the measuring head guiding device of the measuring head are output, so that method step 112 can terminate the method.

FIG. 2 schematically shows a block diagram 60 of an evaluation unit 20 which is required in a measuring device for precision measurement in a nanometer range according to the first implementation example. The evaluation unit 20 detects in a first block a means 21 which is designed to detect geometric distance values of the i measurement channels to the t (j). The means 21 of this block cooperates with means 22 and 23, wherein the means 22 is designed for detecting three-dimensional position values for the i measurement spots on the measurement object surface at time t (j), and the means 23 in the adjacent block for detecting a local Inclination of the measuring object surface is formed relative to the measuring head. The means 23 moves on to a block with the means 24, which is designed to associate the detected tendency to temporal variations of systematic measurement errors by correlating temporal patterns. The means 25 is designed to take into account the measurement errors when creating local topographies for the redundant i measurement channels.

From means 25 it transitions to the block 26 having means 26 adapted to correlate the local topographies, separating timing deviation movements of the probe head by separating a bump of a sensor trace and a bump of a sensor orientation of the probe in the probe head from the true surface topography , Finally, starting two output blocks are ver ^¬ ensures by the means 26, again with the agent 27, which is adapted to output an adjusted surface topography, as well as a further block to the means 38 which is adapted to output a real path and an actual orientation of the Measuring head guiding device.

FIG. 3 shows the measurement result of a calibration run for a chromatic-confocal measuring head. The measuring head of Messvor ^¬ direction as 4 and 5 show the subsequent figures, is directed in this calibration run on a flat glass, which is mounted on a linear table. The relative distance of the Planglases to the measuring head can be measured with a control interferometer with nm (nanometer) accuracy. The calibration travel determines the relationship between the spectrometer signal in the form of the peak position of the spectral peak and the distance, and this is carried out in the entire spectral range of the spectrometer. This is the characteristic curve "Distance over pixel position", which is valid for the combination of spectrometer and measuring head.

The diagram shown in FIG. 3 provides the time-dependent deviation of the actual value from the expected value for two repetition measurements. For this purpose, a range of ± 0.1 μπι deviation over a range of 100 μπι (micrometers) is shown as a function of path and time. From this diagram according to FIG. 3, the following effects are visible. The dashed curve a is generally a few nanometers lower than the solid curve b, which corresponds to a time-dependent long-term drift of the working distance, for example due to temperature influence. The curves show a coarse ripple of ± 10 nm in comparison to the calibration run of the time-dependent long-term drift of the measuring range, as shown by the dotted curve c. The two curves a and b have a dominant periodicity of 2 μπι and are fairly closely correlated. To determine a characteristic, such as the dotted characteristic c, a smoothing filter can be used which removes such fine-wave periodicities. Such periodicities may be caused by tilting movements of the driving axis, by varying the distance at the control interferometer or at the measuring head or may represent a temporally variable systematic behavior of the measuring head or of the evaluation unit, which would be smoothed out during the production of the characteristic curve.

Remaining fine-wave time-related deviations between the two curves a and b are due in part to measured value noise, partly to other time-related vibrations and to fluctuations in the measurement setup and the control interferometer distance measurement. With the method previously discussed with reference to FIGS. 1 and 2 and the following Following measuring devices 3 and 40, it is now possible to differentiate in a profile measurement in nanometer precision these various disturbances, namely the long-term drift, the temporal variations systematic error of the distance sensor, the positioning errors of the control axes and the vibrations of Messob ectes and the measured noise and from the To calculate the profile of the test object.

For this purpose, according to the invention, an optical measuring head with a plurality of measuring channels is provided, as shown in the following FIGS. 4 and 5, wherein the foci of the spots of the measuring channels are arranged at least along a scanning direction, which is referred to as the main line. In the case of the scan, the profile of the measurement object in each measurement channel is measured, and the totality of the time and positionally offset profiles is assembled by means of correlation methods to form an average overall profile.

Since the profile deviations in the individual channels by temporal and spatial patterns of movement in all profiles to be identical may be prepared from the profiles deviation ^¬ movements of the measuring head guide device as well as temporal variations of systematic measurement errors are eliminated, and a random noise can be reduced by a plurality of measurements of the underlying surface topography. The provision of at least one additional measuring channel transverse to the main line can already as as example ^¬ is shown in the following Figures 4A to 4C and 5A to 5C, serve to measure seasonal tilting movements of the measuring head at the measurement head transfer device and herauszurechnen. It is also possible to calculate local slopes of the surface in order to calculate systematic measurement errors of the measuring head that depend on the inclination of the surface of the sensor. For this purpose, for example, the time-dependent phase position of the measurement signal of two channels can already be calculated and thus a time-dependent phase difference can be determined. The difference phase can in turn be converted into a local height difference as a gradient between two measuring points. In addition, other measurement channels can be overlaid interferometrically to measure the optical path difference corresponding to the difference phase. From the entirety of the local slopes, a topography can then be integrated. This differential scan competes with an absolute topography acquisition derived from the ideal actuator trajectory and the measured distance values.

This error correction method also works for steep surfaces when the quality of spektralinterfero ^¬ metric distance measurement by seasonal spacing ^¬ change is affected. Thus, a robust measurement is provided for chromatic-confocal detection of the surface topography.

In order to eliminate a measuring system error with respect to an inclination of the measuring head relative to the orthogonal of the measuring surface, as already mentioned above, a measuring spot arranged transversely to the main line is sufficient. However, the inclination or the inclination angle can be determined more accurately if, instead of the one transverse measuring spot, three measuring spots are arranged to the main line such that they form an equilateral triangle as in the following FIGS. 4A and 5A forming, by setting up the normal vector of the triangle out of any inclinations is possible.

In addition to the deviation by an angle of inclination, which can be equated to a time-dependent pivot or rotational direction about a longitudinal axis in the direction of travel and is also referred to as a roll angle, there are other time-related orientation deviations, which are referred to as pitch and a rotation angle Θ about the transverse axis of the Mark the measuring head guide device or the measuring head. Yawing is possible as the third direction of rotation and known as the angle of rotation Φ about the vertical axis. The three solid angles of the time-related orientation deviations during scanning, namely, and, are also known as Euler angles or as La ^¬ angles.

For the time-related orientation deviation, the transformation matrix then consists of the three individual rotation matrices for the respective angles. The rotation ^¬ sequence in the order Φ, Θ and Φ is specified in the nachfol ^¬ constricting transformation matrix.

In linearized form with angles given in radians then the transformation matrix has the form: [1 φ - Θ

D = _{ - Φ + * θ 1 + * θ * Φ

[g - Φ + θ * Φ 1

If the product terms are neglected, which is possible with good Aktorik the Meßkopffführungsvorrichtung, the simplified torque matrix results:

- Θ

D = - Φ Φ

Θ - Φ

Thus, a unit matrix plus an antisymmetric matrix in the three Euler angles can be used as vector models of time-related orientation deviations.

It should be noted that such actuators or Meßkopfführungsvorrichtungen often represent a coupling between path deviations and orientation deviations. In a rigid guide results for a wavy path z (x) or a time-dependent wavy course of the orientation with the pitch angle Θ. With a rigid guide, the measuring head always follows parallel to the guide, so that Θ (theta) corresponds to the pitch of the measuring head guide, with:

Theta (x) (d / dx) z (x)

A measuring head, which is rigidly attached to such Meßkopfführung, thus makes the orientation changes in Ratio 1: 1 so that its trajectory rl (x) follows the trajectory rO (x) of a bearing point according to rl (x) r0 (x) + D (x) * R1, so that the relative path deviation motion rl (x) - rO (x) = D (x) * Rl.

The movement of the support point is therefore completely described by:

• z (x) or theta (x)

• y (x) resp. Psi (x)

• Phi (x)

The missing angles of such a rotation matrix are:

Theta (x) (d / dx) z (x)

Psi (x) (d / dx) y (x)

The path parameter in the precision measuring device is not x, but the variable time t = x / v, which corresponds to the distance traveled s.

Thus, if the lever vector Rl between bearing and measuring point is known, it can be concluded from the orientation at the measuring point on the position error at the measuring point.

If there is no rigid guide for the Meßkopfführungsvorrichtung, but a sprung guide with a certain delay and a certain damping, which applies in time conditional equation of motion for a forced damped oscillation with a displacement of x (t). x ^u (t) + i omegaO x '(t) + M omegaO ^A 2 * x (t) = F (t) with the constraining force

F (t) = M x O (t)

Where F is the derivative of F by time t.

This means that in the case of a sprung guide of the measuring head on the measuring head guiding device, the response to a guide waviness is phase-shifted and changed

Amplitude takes place. The parameters of the suspension are the resonance frequency omega and the damping.

For long-wave deviation movements of the measuring head guiding device or the measuring head, this time-dependent deviation movement is transmitted in a ratio of 1: 1. If the time-dependent deviation movement is close to a resonance frequency, this leads to more or less excessive time-dependent deviation movements with an approximately 90 ° phase shift. However, in the case of short-wave, time-dependent deviation movements, these are integrated away.

Thus, a measuring run with frequency analysis of the measured distance values provides information about the natural vibrations of the actuator and the measuring head holder of the measuring head guiding device. A non-optimal control of the scanning drive can even lead to resonant timing deviations. In this case, the amplitude of a resonant deviation movement can depend on the position of the measuring head with respect to a partial actuator system. However, lateral deviation movements only become visible when measured on structured objects. A straight line grid then looks wavy. On a vertical plane mirror is a pivoting about the vertical axis or a yaw as discussed above not detectable. However, deviations in the axial direction are clearly visible, so that pitching about the transverse axis and rolling about the longitudinal axis will cause ripples in the apparent topography. The multi-channel measurement according to the invention thus distinguishes pitching and rolling caused by the measuring head guidance device.

Further correction possibilities result from the fact that redundant measuring points, as shown in the following FIGS. 4A, 4B and 4C as well as 5A, 5B and 5C, are arranged differently with their distances from each other. This avoids that periodic fluctuations of the topography with a period shorter than

Tl = distance / driving speed, not to be perceived as sub-sampled space frequencies. The comparison with a measuring point with distance 2> distance 1 shows the difference immediately.

In addition, it is possible to use even unequal frequency measurement points to detect a subsampling in the temporal range. All that is required is a temporal sampling with two different clock rates.

In addition, it is possible to compensate for measurement errors due to object tilt. For this purpose, a calibration tion of the measuring head by distance measurement on a precision ball made. The deviation of the measured topography from the desired shape is then determined and a table is generated about deviations with respect to object slopes. For this purpose, the measuring head is preferably at least three measuring points, as shown in the following figures 4A, 4B, 4C and 5A, 5B, 5C, which ideally form an equilateral triangle 17, as shown in the following figures 4A and 5A, on the measuring object project, so that a local inclination of the object surface can be determined from a normal vector of the triangle. In a measurement evaluation, only a calculation of a local object tilt is performed, wherein the correction of the measured distances can be carried out with the aid of the table "deviation via object tilt".

Furthermore, the time-related deviations from temporal variations of systematic measurement errors shown in FIG. 3 can be corrected by recording temporally induced lateral movements of the measuring head by acceleration sensors. The acceleration sensors measure the movement of the measuring head relative to the space, transverse to the optical axis (in the x and y direction), and integrate this acceleration into a time-dependent path deviation dx (t) and dy (t). This can also be done in the z direction. In this case, the multi-channel measuring head, as shown in FIGS. 4 and 5, measures the local object tilt. The correction of the measuring points can on the one hand be done by interpolating the object surface to points lying on the desired path and thus enabling a correction of the distance values or it can be an indication of the surface coordinates (x, y and z) with the measured path X = X_soll + dX instead of the set path X_set (t) = [x_set (t), y_set (t)]. A vibration with the amplitude xO and the frequency f gives an amplitude of the acceleration of: aO = xO * omega ²

4pi ² * x * f ² with omega = 2pi * f. The conversion factor for the conversion of the acceleration a as a multiple of the gravitational acceleration g = 9.81 m / s ² is then given

F = 4pi ² / (9,81ms-2) / (1μπ \) / (1Hz) ² = 4,02 * 10 ^"6 and thus:

AO (in g) = xO (in μηι) * f (in Hz) ^Λ 2 * 4 millionth

Table of acceleration for typical vibrations

times the gravitational acceleration g

From vibrations of the orientation can be calculated back to vibrations of the situation. For this purpose, the vibrations of the orientation measured using the multipoint measuring head according to FIGS. 4 and 5 are sorted into proportions of typical natural oscillations of the actuators. A geometrical model gives the relation between measuring head orientation and measuring head position. tively to the actuators for the natural oscillations. This is determined in a calibration procedure. In the subsequent measurement evaluation, the measured deflection of the orientation, specifically separated according to natural oscillations and based on the geometric model, is converted into a deflection of the measuring head position. The thus determined deflection of the measuring head position is then used to correct the coordinates of the measured points on the object surface.

For example, in order to enable detection of sub-sampled ripples, the distance of the measurement points in a spatial direction can not be kept constant in order to detect periodic ripples in a sub-scan. In addition, it is provided, as already indicated above, to represent the measuring clock differently for different measuring points in order to make periodic ripples in the sub-sampling likewise recognizable.

FIG. 4 shows a schematic diagram of a measuring device 3 for the precision measurement of a surface topography 1 according to an embodiment of a precision measuring device in the nanometer range. For this purpose, the measuring device 3 has a measuring head 4 in a measuring head guiding device 5 for a chromatic-confocal detection of the surface topography 1 or for a spectral interferometric OCT distance detection of the distance e between a measuring head optical system 10 of the measuring head 4 and a measuring object surface 16.

The measuring head 4 can be supplied by a spectrally broadband light source, a so-called SOA (solid state optical amplifier). Preferably, however, a superluminescent diode (SLD) derived from the SOA light source is used. The SLD supercontinuum light source has more power per bandwidth. It is therefore suitable as a single light source, the light can be distributed via LXi coupler on many channels.

As an alternative to a broadband light source in cooperation with a line spectrometer, a wavelength-tunable light source with time-sequential recording of the spectrum is also applicable. However, additional time-delay effects may occur.

In the embodiment according to FIG. 4, the broadband light of an SLD light source 6 is coupled into fibers 8 of a light fiber bundle, which in this embodiment has four optical fibers by way of example, via four Y couplers 28, 29, 30 and 31. The broadband light of the SLD light source 6 is thus distributed to the four optical fibers and fed to the measuring head 4 at the Meßkopffführungsvorrichtung 5.

Fiber ends 19 form a fiber array 7 over a head optics 10 and the head optics 10 project a spot array 11 on a target surface 16 which is reflected back and fed via the Y couplers 28, 29, 30 and 31 to four spectrometers 32, 33, 34 and 35 , The reflectance spectra which are formed in the spectrometers 32, 33, 34 and 35, can be to the effect out by an evaluation unit 20 ^¬ evaluates then that temporal variations of system errors and seasonal variation movements of the measuring head guide device and of the measuring head from the surface to be measured topography be separated.

In order to allow such a correction of the measured values when separating the measuring errors, transverse to a central measuring spot 15 on a main line of the scanning directions x or F and G further measuring spots 12, 13 and 14 are arranged, which form the spot array 11, wherein the three measuring spots 12, 13 and 14 form an equilateral triangle transverse to the main line, for example, if the angle of inclination or the angle of rotation about the longitudinal axis x of the Meßkopfführungsvorrichtung 5 is zero, as shown in FIG. 4A.

If this inclination or roll angle is not zero, but deviates from the zero position, then, as FIG. 4B shows, the triangle from the measurement spots 12, 13 and 14 can be distorted. From this distortion, for example, a normal vector of the triangle and thus an angle of inclination can be calculated via the distances of the measuring spots 12, 13 and 14 from each other and thus the measured value for the surface topography can be corrected. For this purpose, the inclination of the object can be used for the correction of distance values of all measuring points or it can be determined from local topographies an inclination for each measuring point. While in FIG. 4B the measurement spot 12 deviates extremely from the main line, FIG. 4C shows another equilateral triangle distortion, as originally shown in FIG. 4A, in which both the measurement spot 12 and the measurement spot 13 are opposite to the positions of a zero roll angle are shifted.

This inclination or roll angle Φ is just one example of a system error. The same applies to the other two Euler angles, namely the yaw angle Φ for a rotation around the

The vertical axis z and the pitch angle Θ for a rotation about the transverse axis y, as discussed above with respect to the vector model of the temporal deviations movements.

FIG. 5 shows a schematic diagram of a measuring device 40 for the precision measurement of a surface topography 1 in accordance with FIG another embodiment of the invention. Components having the same functions as in FIG. 4 are identified by the same reference numerals and will not be discussed separately.

The embodiment according to FIG. 5 differs from the embodiment according to FIG. 4 in that a multiplexer 18 is used, with which measuring channels 9 are supplied with a time offset to a single spectrometer 32, wherein the multiplexer 18 ensures that the multiplicity of signals shown in FIG shown spectrometer can be reduced. By means of the multiplexer 18, the Y couplers shown in FIG. 4 can also be dispensed with at the same time, since the broadband light source 6 can also be coupled to the individual optical fibers in a time-offset manner via the multiplexer 18. One advantage is that the light intensity is not divided by the number of optical fibers, but each individual optical fiber receives the full light intensity via the multiplexer 18.

Figure 6 shows a schematic diagram of a measuring section for Erläu ^¬ esterification of a further embodiment of the Messverfah ^¬ proceedings, in which a typical application of an optical sensor lines ^¬ on Cartesian axes driving as a special example ^¬ be written is. In principle, it is a matter of repeatedly measuring the same surface points with a linear sensor in several measuring runs with crossed linear axes and thereby highlighting the two disturbing effects of profile measurement in the form of vibrations and deterministic axle collars of the x-axis, which can be determined with one or more calibration runs to add up.

For this purpose, an optical measuring head 4, as shown in FIG. 6, passes through several measuring points which are equidistant at a distance L are arranged on a line and each measure a distance in the z-direction. An actuator system consisting of a linear positioner 41 in the x-direction on an x-axis and a positioner 42 in the y-direction on a y-axis move either the measuring object 2 or the measuring head 4. Preferably, x, y , as a driving axis perpendicular to each other. A measuring head holder is provided, with which the measuring head 4 is rotatable about the z-axis, so that the line of the measuring points is at an angle α at an angle to the x-axis.

FIG. 6 also shows that the measuring object 2 is arranged in the coordinate system substantially in a plane spanned by the x and y axes.

To carry out the exemplary measuring method, the measuring head 4 moves at uniform speed vx = dx / dt in the x direction. For each sensor measuring cycle dt, the distance dx is covered, while the y-axis remains fixed. The sensor performs a measurement at times t = M * dt + tO * N. The position of the x-axis or the y-axis is:

X (M) = dx * M

Y (N) = dy * M

Each measuring point i follows the axis positions and supplies a measured distance profile z_mess (i, M, N) in an equidistant grid

X (i, M, N) = DX * i + M * dx

Y (I, M, N) = DY * i + N * dy

and with

DX = L * cos (alpha)

DY = L * sin (alpha) The position of the x-axis is then:

X (M) = dx * M

For the path deviation of the x-axis a characteristic z (X) is assumed, which hardly changes from drive to drive.

At the end of the line, the x-axis is returned to start position and the y-axis by an amount

YSTEP = di * DY

offset with integer di.

The same point (x, y) is now measured not in the data set (i, M, N), but in a data set (i-di, M-dM, N + 1)

di = YSTEP / DY

d = DX / dx.

The measured value consists of topography (zO), axis bumpers (zx, zy) and vibrations zt such that: z (i, M, N) = zO (x, y) + zx (M) + zy ( N) + zt (tl), z (i-di, M-dM, N + 1) = zO (x, y) + zx (M-dM) + zy (N + 1) + zt (t2)

The following assumptions were made:

1) of the measuring point i and the measurement point i-di measure the moving distance ^¬ Chen,

2) the failure of the x-axis towards reaching ^¬ exactly reproducible in neighboring rides and therefore does not depend on N, and

3) The error of the y-axis is constant during a scan and therefore does not depend on M.

Following these assumptions, the following evaluation is possible, in which first the difference of the measured distance z (x, y) is formed between two adjacent line scans N and N + 1:

N) -z (i-di, M-dM, N + 1) zx (M) -zx (M-dM)

zy (N) - zy (N + 1)

zt (tl) - zt (t2)

Subsequently, the sum over dz is obtained from M = 1 to MO, assuming the above assumptions 1 to 3:

M0 * dm * [zy (N) -cy (N + 1)]

Ύ ^■ Ιΰ'Ν ....! Θ · Ν + Μ * ώ) zt ( ^v t) '- Yί-ι (ί = ΐΰ · (Ν + \) ... ίϋ * (Ν + \) + Μ * ώ)

In this case, the expression in the first line is composed of a sum over the first pixels, which is constant, minus a moving average over dM points to the axis tracer zx (m).

The second line gives a difference between the axis errors of the y-axis, weighted by the number dM and increasing linearly with M. Finally, in the third line there is a time sum via vibration deflections. This can hardly be greater than the sum over half an oscillation period because the oscillation moves around 0.

While the shape of this time fraction is different for each N, zx_mittel (m) changes only slightly over many N's. Therefore, it can be expected that, considering several N, the course of zx can be extracted out. For dissecting out or filtering out the temporal component, an averaging over several line scans can take place in noise-like vibration patterns, and a band-stop filter can be used for periodic vibration patterns of a known period, whereby care must be taken that periodic topographies of the same period are not leveled.

Furthermore, in a next step, a profile which consists of several overlapping parts can be assembled into an overall profile in the y-direction, which is also called "stitching". For this purpose, the overlapping profiles of two adjacent journeys, z (i, N) and z (i-di, N + l), are brought into coincidence and the y-axis errors (zy (N) -zy (N + l) and its increase in the i-direction.

A plurality of variations of this embodiment may be made by those skilled in the art without ^{departing from} the scope of the ^appended claims. The axes of the Cartesian squareness may vary with ^¬ play as. Furthermore, instead of a rotation of the measuring head 4 about the z-axis, a rotational movement about the y-axis, as in the case of a pressure roller, can be generated as a measuring object. Even with imperfect overlap interpolation can be pre ^¬ taken. Furthermore, special cases of vibration elimination are possible where = 0. In the case (a = 0), a maximum number with imax = number of measuring points in the measuring head of redundant profiles is obtained. In this case, a summation over all measuring points i provides the imax of three parts, namely a current oscillation deflection, a mean value of the topography and an average value of the Axis error, where both the mean of the topography and the mean of the axis error, as mentioned above, barely changes.

The method according to the invention and the device according to the invention are thus outstandingly suitable for the measurement of surface topographies in the micrometer and nanometer range. Specifically, the method and apparatus of the present invention can be used to qualitatively and / or quantitatively measure the roughness, waviness, flatness and porosity of metallic and non-metallic surfaces.

In particular, the surfaces of metallic workpieces precision from machine such as turbine blades, ^¬ clutch and transmission components come tracht in loading.

Furthermore, the surface topographies of optical components such as aspherical precision glass bodies can be detected qualitatively and quantitatively. Such precision glass body can be provided with vapor-deposited multi-layer system of metals and / or dielectrics.

In addition, the inventive method and the inventive apparatus in the field of medical technology may be especially in areas used in the field of ophthalmology Example, in the qualitative and quantitative detection of the Oberflä ^¬ chentopographie a cornea.

Although at least one exemplary implementation of the method has been described in the foregoing description, various changes and modifications can be made become. The foregoing implementations of the precision method are merely exemplary and are not intended to limit the scope, applicability, or configuration in any way. Rather, the foregoing description exemplifies those skilled in the art to practice at least one implementation of the method, and numerous changes in the function and arrangement of further exemplary embodiments of a measuring device may be made without departing from the scope of the appended claims and their claims to leave legal equivalents.

LIST OF REFERENCE NUMBERS

1 surface topography

2 measuring object

3 measuring device

4 measuring head

5 measuring head guide device

6 light source

7 fiber array

8 fiber

9 measuring channel

10 measuring head optics

11 spot array

12 measuring spot

13 measuring spot

14 measuring spot

15 measuring spot

16 measuring object surface

17 equilateral triangle

18 multiplexers

19 fiber end

20 evaluation unit

21 funds

22 funds

23 funds

24 funds

25 funds

26 funds

27 funds

28 Y coupler

29 Y coupler

30 Y coupler

31 Y coupler spectrometer

spectrometer

medium

Measuring device (2nd embodiment of the invention) Positioner in the x direction

Positioner in y-direction

flow chart

Block diagram start block

step

Claims

claims

1. An optical measuring method for detecting a surface topography (1) of a measuring object (2), comprising the following method steps;

Providing a measuring device (3) with a measuring head (4) in a measuring head guiding device (5) for a chromatic-confocal detection of the surface topography (1);

Application of spectrally broadband light of a light source (6) from a fiber array (7) with i fibers (8) of i measuring channels (9) via a common measuring head optics (10) to form a spot array (11) of i measuring spots (12 to 15) on the measuring object (2); Detecting and digitizing i reflection spectra of the i measuring channels (9) and

Evaluating the digitized reflection spectra by taking out temporal variations of systematic measurement errors and time-related deviation movements of the measuring head guiding device (5).

2. Optical measuring method for detecting a surface topography (1) of a measuring object (2), comprising the following method steps;

Providing a measuring device (3) with a measuring head (4) in a measuring head guiding device (5) for a spectral interferometric OCT distance detection to the surface topography (1);

Application of spectrally broadband light of a light source (6) from a fiber array (7) with i fibers (8) of i measuring channels (9) via a common measuring head optics (10) to form a spot array (11) from i measuring spots (12 to 15) on the measuring object (2); Detecting and digitizing i reflection spectra of the i measuring channels (9) and

3. Optical measuring method according to claim 1 or 2, wherein furthermore the following evaluation steps are carried out for taking out temporal variations of systematic measuring errors and time-related deviation movements of the measuring head guiding device (5):

Acquiring geometric distance values (a, b, c) of the i measurement channels at time t (j);

Acquiring three-dimensional position values for the i measurement spots on a measurement object surface (16) at time t (j);

Detecting a local inclination of the target surface (16) relative to the probe (4);

Correcting temporal variations based on the detected slope systematic measurement errors; Creating local topographies for the redundant i measurement channels;

Correlating the local topographies with separation of temporal deviations movements of the Meßkopfführungsvorrichtung (5) by separating a bumpiness of a sensor track and a bumpiness of a sensor orientation of the measuring head (4) in the Meßkopfführungungsvorrichtung (5) of the true surface topography (1) and

Output of a cleaned surface topography (1) and a real trajectory and orientation of the measuring head guiding device (5) of the measuring head (4). Optical measuring method according to claim 3, wherein for detecting the local inclination of the measuring object surface (16) relative to the measuring head (4) at least three measuring spots (12, 13, 14) of a triangle (17) are projected onto the measuring object surface (16).

Optical measuring method according to claim 4, wherein for detecting the local inclination of the measuring object surface (16) relative to the measuring head (4) at least three measuring spots (12, 13, 14) of an equilateral triangle (17) are projected onto the measuring object surface (16) and from distance values of the measurement spots (12, 13, 14) to each other, the inclination is determined.

Optical measuring method according to one of claims 1 to

5, wherein the seasonal variation movements of the measuring head guide device (5) by means of a three-dimensional acceleration sensor the measuring head (4) summarizes ER and the measured values of Oberflächentopogra ^¬ chromatography (1) to be corrected accordingly.

Optical measuring method according to one of claims 1 to

6, wherein by means of a vector model measuring head movements by determining the yawing, pitching and Rol ^¬ lens of the measuring head (4) on the Meßkopfführungsvorrichtung (5) are detected.

Optical measuring method according to one of the preceding claims 1 to 4, wherein a differential scanning method is used, comprising the method steps: Determining local slopes of the measuring object surface (16) by means of height difference formation between i measuring spots (12 to 15) and

- Integrating the total of local slopes to a surface topography (1).

9. Optical measuring method according to claim 8, wherein the differential scanning method detects path differences between two measuring spots of a measuring head (4) with a fiber end and the measuring spots (12, 13) of two focusing lenses, and wherein the spectral interferometry (OCT) the optical path difference to the two measurement spots (12, 13) is measured.

10. Optical measuring method according to one of claims 1 to

9, wherein measuring spots (13, 14) arranged in a line in the scanning direction are arranged at different distances.

11. Optical measuring method according to one of claims 1 to

10, wherein temporally successive sampling pulses occur at different time intervals.

12. Optical measuring method according to one of claims 1 to 10, wherein the temporal sampling of the measuring spots (13, 14) takes place with different clock rates.

13. Optical measuring device for detecting a surface topography (1) of a test object (2) comprising:

a measuring device (3) having a measuring head (4) in a measuring head guiding device (5) for a chromatic-confocal detection of the surface topography (1), a spectrally broadband light source (6) which supplies broadband light via i Y couplers (28, 29, 30, 31) i fibers (8),

a fiber array (7) with the i fibers (8) for i measuring channels (9) arranged in the measuring head (4), a measuring head optics (10) having a spot array (11) with i measuring spots (12, 13, 14 , 15) on the measurement object (2);

Means for detecting and digitizing i reflection spectra of the i measuring channels (9) in i spectrometers (32, 33, 34, 35) and

an evaluation unit (20) for the digitized reflection spectra for the purpose of calculating temporal variations of systematic measurement errors and time-dependent deviation movements of the measurement head guidance device (5).

Optical measuring device for detecting a surface topography (1) of a measuring object (2) comprising:

a measuring device (3) with a measuring head (4) in a measuring head guiding device (5) for a spectral interferometric OCT distance detection to the surface topography (1),

a spectrally broadband light source (6), via i Y coupler (28, 29, 30, 31). i fibers (8) supplied with broadband light,

a fiber array (7) with the i fibers (8) for i Messka ^¬ ducts (9) in the measuring head (4) is arranged, a common measuring head optical system (10) comprising a spot array (11) having i measurement spots (12, 13, 14, 15) on the Messob ^¬ jekt (2) is formed; Means for detecting and digitizing i reflection spectra of the i measuring channels (9) in i spectrometers (32, 33, 34, 35) and

15. An optical measuring device according to claim 13 or 14, wherein the i measuring channels (9) via a multiplexer (18) with a spectrometer (32) are in communication.

16. A precision optical measuring device according to claim 15, wherein the measuring head is designed such that the distance values of the channels occupy a fixed ranking.

17. Optical precision measuring device according to one of claims 13 to 16, which further comprises the following means for calculating out time variations of systematic measurement errors and time-related deviation movements of the Meßkopffführungsvorrichtung (5) from the i reflection spectra:

Means (21) adapted to detect geometric distance values of the i measurement channels (9) at time t (j);

Means (22) adapted to acquire three-dimensional position values for the i measurement spots (12 to 15) on the measurement object surface (16) at time t (j); Means (23) adapted to detect a local inclination of the target surface (16) relative to the probe (4); Means (24) adapted to correct for temporal variations of systematic measurement errors based on the detected slope;

Means (25) adapted to create local To ^¬ pographies for the redundant i measuring channels (9);

Means (26) adapted to correlate the local topographies while separating time-related deviation movements of the Meßkopfführungsvorrichtung

(5) by separating a bumpiness of a sensor track and a bumpiness of a sensor orientation of the measuring head (4) in the measuring head guiding device

(5) from the true surface topography and

Means (27) adapted to output a cleaned surface topography (1) and

Means (38) adapted to output a real web and a real orientation of the Meßkopfführungsvorrichtung (5).