DE102013018364B4 - Method for detecting and / or measuring surface defects of a component - Google Patents

Abstract

Method for detecting and / or measuring surface defects (5) of a component (2) comprising the steps of: a) measuring a component (2) to be tested for obtaining actual data (3) from a plurality of measuring points of the component to be tested (3) 2), b) formation of total actual area data (3) from the plurality of measurement points; c) providing target data (1) of a surface of the component to be tested (2), d) aligning the total actual area data (3) to the target data (1); e) dismembering the actual data (3) into a multiplicity of actual data tiles (10), f) individually aligning all the actual data tiles (10) to the target area corresponding to the actual data of the tiles (10) Data (1) and g) determining a plurality of surface deviations between measuring points of the aligned actual data tiles (10) and the target data corresponding to these actual data tiles (10) (h) output of the step g) determined surface deviations.

Description

The invention relates to a method for detecting and / or measuring surface defects of a component, in particular a body shell component.

In the case of a plurality of components, for example components which form an outer skin of a vehicle, essentially two different deviations in shape are distinguished in the quality determination of the components. The first shape deviation represent so-called shape errors of the component. Examples of such shape errors are shape deviations of the component, for example with regard to its length, width or its curvature. Such form errors are characterized in that they relate to the entire component or at least a large portion of the component.

A second shape deviation, which is essentially responsible for the visual appearance of a component, are so-called surface defects. The characteristic of these surface defects is that their geometric expression and / or extent is relatively small compared to the entire component and its shape errors. If a component is now measured, for example by means of optical measuring devices, then the surface defects interfere with the shape errors of a component and can be lost in a comparison of nominal data with actual data of the component to be tested, since usually the magnitude of the shape deviations from surface defects is significantly lower are as the order of magnitude of the shape error of a component. However, surface defects, even though they are substantially lower in terms of their absolute error dimensions than shape errors of the component for the visual impression of the component, are highly relevant in terms of their surface quality and by no means negligible. Such surface defects are, for example, dents, dents, sink marks, ripples, constrictions or cracks in the corresponding components. This results in the problem that a direct, simple use of the actual data of the component for checking the dimensional accuracy, that is to say the shape errors, for example with an optical coordinate measuring machine, is not always easy to test, even with regard to surface defects.

In the prior art, therefore, for determining shape errors, actual data of the component to be tested, which were detected by means of a suitable measuring device, for example an optical coordinate measuring machine or computer tomographic measuring method, for determining the shape errors as a whole by means of an orientation in a reference point system or on the basis of a global That is, the entire best-fit Bestfit alignment to the target data of the component or the target data aligned to the component surface.

Forming errors can be easily read from such a global orientation of the total actual data compared to the total target data. Surface defects, on the other hand, are often invisible to the evaluating measurement technician almost invisibly, since the surface defects are usually orders of magnitude smaller than the shape errors.

Different methods are used in the prior art to determine surface defects.

A first method is z. B. to be carried out by an employee manual stripping a body panel with a whetstone. This procedure is time consuming. The result, in particular the assessment of the result, depends essentially on the experiential knowledge of the employee. After manual removal of the sheet metal part according to a usually component-dependent, predetermined manner, surface defects are visible on the surface of the component and can be associated with some experience of the employee also possible causes that have led to the surface defects. A numerical evaluation of the surface defects in the form of measurements or deviation values from the target data or to a reference workpiece is not possible.

Another possibility of surface analysis and / or surface evaluation are software-based methods that offer a computer-aided simulation of the real peel-off with a virtual whetstone. Such a method using so-called "virtual deduction stones" or "digital deduction stones" is time-consuming and requires a high computing power.

Furthermore, it is known in the prior art that specially trained sensors or measuring devices for surface inspection are used in addition to the virtual stripping. Such sensors perform, for example, a comparison of the actual surface of a component to be tested with the surface of an object defined by the user as a reference part or boundary pattern. Not all of these sensors have the ability to take measurements on the recorded surface areas of the component to be tested and output deviation values exactly. Often only qualitative detection of surface defects is possible. Some sensors are optimized for the detection and detection of very small deviations. They do, however relatively small measuring field, so that for a complete inspection of a component, in particular a large-area component, such. B. a roof skin or a door outer skin of a motor vehicle incurred a lot of time.

In order to reduce the high time required for a complete inspection of a test object, such sensor-based surface detections are often only used in a predefined surface area of the component to be tested. However, this presupposes that an experienced employee already selects the surface areas of the component to be tested prior to the measurement, which according to experience are particularly susceptible to surface defects. It may happen that surface areas of the component to be tested are not tested, although these surface areas have surface defects. In such a case, then surface imperfections remain undetected and undetected.

From the EP 2 282 166 A1 For example, a method for displaying the surface of a test object has become known. In such a method, to display the surface of an object, 3-D IST data of the surface of the object are determined. 3-D-SOLL data of the surface, for example CAD data of the surface of the object to be tested are modified on the basis of the 3-D-ACTUAL data of the surface of the object. The 3-D-SOLL data of the surface of the object and the modified 3-D-SOLL data of the surface of the object are used as 3-D display data for visualizing the surface. To detect surface defects in the actual data, this document proposes to use the method of "virtual stripping" already explained above. In order to obtain the 3-D display data, the 3-D DESIRED data are modified at least in certain areas.

From the EP 1 476 797 B1 For example, a method is known for determining deviations D from a present surface and a nominal surface and for generating a topographic pattern from the deviations D, wherein the topographic pattern is projected onto the component to be tested by means of a laser projection. Such a method requires a high computational effort.

From the DE 44 02 414 C2 For example, a method for three-dimensional measurement of a surface of objects is known.

Another measuring method for optical 3-D measurement is from the EP 1 065 628 B1 known.

From the US Pat. No. 8,233,694 B2 It is known to qualitatively detect surface defects by comparing reflections or other light effects on a reference part and the test object to be tested. The generation of the light effects takes place in that a specific pattern is applied to the reference part and then to the measurement object and, based on the differences in the two figures, it is concluded that there are existing surface defects. In this method, it is also necessary that a certain amount of experience in the evaluating persons is available, which can be used for the evaluation of the visually displayed surface defects.

From the US 2005/0021163 A1 For example, a method for aligning freeform surfaces with arbitrary start points is known. In this method, regions are first selected and generated on the basis of surface characteristics, for example curvatures, wherein these regions are formed both in a free-form surface of a reference model and in the free-form surface of the measured surface. Based on the surface properties, corresponding regions are determined and overlapped. Subsequently, a coarse adjustment of the measuring surface to the reference surface. This coarse alignment is based on less, over the entire areas to be considered distributed points pairs. Subsequently, a fine alignment of the two surfaces (measuring and reference surface) is carried out as a whole, wherein the basis for the fine alignment of the two surfaces is an increased number of pairs of points, which lie in particular within the defined regions. The coarse alignment and the fine alignment is based on a common transformation matrix. In such a method, there is the disadvantage that although form errors can be detected, surface errors remain undetected.

The object of the invention is to provide a simple and reliable method for the determination of surface defects, which allows a qualitative detection and a quantitative measurement of surface defects.

Furthermore, the required measuring effort of a part to be tested should be minimized.

Furthermore, the method should provide reliable results regardless of the experience of an employee or independent of predetermined component areas whose selection is based on empirical values.

This object is achieved by a method having the features of claim 1. Advantageous embodiments are specified in the subclaims.

• a) Vermessen eines zu prüfenden Bauteils (2) zum Erhalt von IST-Daten (3) aus einer Vielzahl von Messpunkten des zu prüfenden Bauteils (2),
• b) Bildung von Gesamt-IST-Flächendaten (3) aus der Vielzahl von Messpunkten;
• c) Bereitstellen von SOLL-Daten (1) einer Oberfläche des zu prüfenden Bauteils (2),
• d) Ausrichten der Gesamt-IST-Flächendaten (3) zu den SOLL-Daten (1);
• e) Zerstückeln der IST-Daten (2) in eine Vielzahl von IST-Daten-Kacheln (10),
• f) Individuelles Ausrichten aller IST-Daten-Kacheln (10) auf zu den IST-Daten der Kacheln korrespondierende SOLL-Daten (1) und
• g) Ermitteln einer Vielzahl von Flächenabweichungen zwischen Messpunkten der ausgerichteten IST-Daten-Kacheln (10) und den zu diesen IST-Daten-Kacheln (10) korrespondierenden SOLL-Daten (1),
• h) Ausgabe der im Schritt g) ermittelten Flächenabweichungen.

An inventive method for detecting and / or measuring surface defects of a component comprises the following steps:

• a) measuring a component to be tested ( 2 ) for receiving actual data ( 3 ) from a plurality of measuring points of the component to be tested ( 2 )
• b) Formation of total IST area data ( 3 ) from the plurality of measuring points;
• c) providing target data ( 1 ) a surface of the component to be tested ( 2 )
• d) aligning the total actual area data ( 3 ) to the target data ( 1 );
• e) dismembering the actual data ( 2 ) into a plurality of actual data tiles ( 10 )
• f) Individual alignment of all actual data tiles ( 10 ) to target data corresponding to the actual data of the tiles ( 1 ) and
• g) determining a plurality of surface deviations between measuring points of the aligned IST data tiles ( 10 ) and the actual data tiles ( 10 ) corresponding target data ( 1 )
• h) output of the surface deviations determined in step g).

The method according to the invention is therefore fundamentally based on a comparison of actual data tiles formed from measured total actual area data of the component to be tested with target data of the component to be tested, in particular with respect to a surface of the component to be tested.

It is essential to the invention that the total actual surface area data of the component formed from measured measuring points are fragmented into a multiplicity of actual data tiles, which is possible in a simple manner with little computation effort. After dismembering the total actual data into the actual data tiles, all resulting actual data tiles are individually aligned, in particular optimally aligned, to the desired data corresponding to the data of the actual data tile (s).

As a result of this measure, it is possible to individually adapt the actual data tiles, which are much smaller than the entire component surface, to the corresponding nominal data. As a result, form errors, which are usually orders of magnitude larger than surface errors, are masked out, since the alignment of the actual data tiles with the nominal data takes place in much smaller area sections. Thus, when comparing the aligned IST data tiles relative to the corresponding DESIRED data, smaller deviations, particularly deviations that represent surface imperfections, become more apparent. According to the invention, therefore, a modification of the total actual surface area data of the component to be tested, to the effect that it sections, z. B. be aligned in area sections in the size of the IST data tiles on the corresponding target data. Each of the aligned IST data tiles is thus alignable with a smaller error, ie much more accurately to the corresponding DESIRED data than the total ACTUAL area data on the totality of the DESIRED data. If, in a comparison of the aligned actual data tiles with the corresponding target data, then still determined surface deviations are present, then the qualitative presence of surface defects can be deduced. When determining the surface deviations between the aligned actual data tiles and the corresponding desired data, it can then be determined in a simple manner from which amount of the determined surface deviations these are to be output.

The determination of the surface deviations is carried out as described below: In the case of a multiplicity of points of the at least one aligned IST data tile, in each case one solder is passed through these points onto the DESIRED data 1 and in each case the distance between the points of the actual data tiles and the intersection of the solder with the DESIRED data is determined. The distance between the points of the IST data tiles and the corresponding intersections of the solders with the DESIRED data is a measure of the area deviation of the IST data tiles to the DESIRED data at the respective point of the ACTUAL data tile. It is possible to specify the solders in two ways. In a first alternative, the solders are aligned through the respective points of the IST data tiles orthogonal to the actual data tiles in the respective points. In a second alternative, the solders are aligned orthogonal to the DESIRED data and each contain a point of the IST data tile. The two alternatives are basically equally applicable, but should be formed within a component to be tested, in particular within an IST data tile, in the same manner. The points of the actual data tiles can be measuring points which, together with the orientation of the actual data tiles, were moved to the desired data. Alternatively, the points can also be corner points and / or nodes of a triangular grid, which were used in the determination of the actual data from the measuring points to form the total actual area data. In the orientation of the actual data tiles, these corner and / or node points have in any case been moved during the alignment of the actual data tiles.

In the method according to the invention, it is advantageous that the component to be tested only has to be measured once. The total IST data obtained from measuring points can be be used in two ways. On the one hand, the entirety of the actual data obtained can be optimally aligned with the entirety of the desired data, for example with a bestfit alignment, wherein in particular a comparison of the aligned overall actual data to the target data includes dimensional errors of the component to be tested can be determined.

In a preferred manner, a copy of the total actual data obtained from the measurement points can be made, which are then used as a basis for the further method according to the invention and the total actual data, as described above, firstly into a multiplicity of actual data. Data tiles are dismembered. It has been found that the method according to the invention thereby ensures an expense advantage, since the component to be tested only has to be measured once.

Second, it has been shown that the data processing of the total IST data, in particular the dismantling of the total actual data into a plurality of actual data tiles and the subsequent alignment of the individual actual data tiles on the corresponding target Data generated a relatively low computational effort and thus thereby also a time advantage or an advantage in terms of the required hardware performance is achievable.

The measurement of the component to be tested preferably takes place by means of optical or computed tomographic measurement methods, which have sufficient accuracy in the measurement of the surface of a component. As a result of these measuring methods, which have a significantly higher resolution compared to tactile measuring methods with a comparable measuring duration, there is a larger plurality of measuring points from which the total actual area data are generated.

In a preferred manner, all IST data tiles are individually and individually aligned to the desired data corresponding to the actual data of the tiles. Alternatively, it is of course also possible to align only a subset of the IST data tiles in the areas of the surface of the component to be tested on the target data, in which it z. B. is known that surface defects occur frequently.

Bestfit alignment has proven to be a possible way of aligning the individual IST data tiles with the target data, such that the sum of all surface deviations of an actual data tile compared to the corresponding target data is minimal.

According to the invention, a bestfit orientation is understood to mean that one or more of the actual data tiles is optimally adapted to the desired data via one or more translations and / or one or more rotations, such that the sum of all surface deviations with respect to the corresponding target Data is minimal. The translations can be made along spatial axes, z. As an X, Y or Z axis of a coordinate system or along any vector in this coordinate system. Likewise, the rotations may be around the X, Y, and / or Z axes, and any vectors in a coordinate system (X, Y, Z system) may act as a rotation axis.

In addition, it may be expedient to limit the degrees of freedom of the translational alignments and / or the degrees of freedom of the rotational orientations. The restrictions of one or more degrees of freedom, in particular with regard to the translational orientation and / or the rotational orientation, are particularly useful if, by translation or rotation in the direction of the relevant degree of freedom or about an axis in the direction of the relevant degree of freedom, an overlap of the aligned actual data Tiles would be degraded by such a shift from the corresponding target data.

This measure makes it easier to detect surface errors representing small deviations between an actual data tile and the corresponding target data and to hide the larger and larger shape errors with regard to their amount. The IST data tile is thus optimally applied or adapted to the corresponding target data, as a result of which smaller-area surface defects, which usually also have a smaller amount of error than unresolved area deviations, are left over.

To simplify the computational effort, it may make sense to make the alignment of the individual actual data tiles on the corresponding target data only by means of a translational shift, the actual data tiles specifically in a coordinate system along an X, Y and / or Z-axis takes place. Such a shift makes sense especially when the actual data tiles have only small curvatures.

In the case of larger curvatures of the actual data tiles, the orientation of the individual actual data tiles to the corresponding desired data can be rotary, expedient manner in each case take place via IST data tile-specific rotations about coordinate axes, for example an X, Y and / or Z axis.

In addition, it is preferable to treat the IST data tiles during alignment with the corresponding desired data such as rigid bodies or rigid surfaces. They are moved as a whole undeformed and / or twisted and / or adjusted in the context of a Bestfit alignment to the target data. This also reduces the required computational effort.

In another embodiment of the invention, the alignment of the individual actual data tiles on the corresponding target data in a direction perpendicular to the actual data tile or in a direction perpendicular to a bestfit layer, which formed in advance and to the actual Data tile is aligned. This means that at least one or at least one selection of actual data tiles is first created a bestfit level. Upon receipt of this bestfit layer, a solder is created on this plane. The shift, d. H. Alignment of the relevant actual data tile is then carried out only along the solder or the perpendicular direction until the sum of the surface deviations between the actual data tile and the corresponding target data is minimal.

Furthermore, it may be expedient to align the individual actual data tiles with the corresponding target data by a combination of the measures "displacement along coordinate axes", "rotations about coordinate axes" and / or "displacements along surface normals at bestfit levels". perform.

Conveniently, breaking up the total IST data into a plurality of IST data tiles, iteratively, increases the number of IST data tiles from one iteration step to the next iteration step until a predefined termination condition is reached. As a predefined termination condition, it is possible, for example, to fall below a predetermined, maximum permissible value for the edge length of a tile. As a possible further termination condition, for example, falls below a maximum allowable value for the sum of all surface deviations of all tiles in question. The former results in a relatively low computation effort, since it is not necessary to carry out an area deviation determination after each iteration step. The latter of the two named alternatives results in a more exact adaptation of the IST data tiles to the corresponding DESIRED data. This measure is particularly recommended in areas where it is known that surface defects occur preferentially there.

In addition, a predefined termination condition may be a number of iteration steps predetermined by the measurement personnel or test personnel, the number essentially depending on the component geometry, in particular the geometry of the surface of the component. So it may be z. B. in a less curved component already sufficient to perform a smaller number of iterations, where it is useful in a very highly curved component or in the region of highly curved zones of a component, optionally to perform a higher number of iteration steps.

In this case, it may be expedient to divide a selection of existing IST data tiles from one iteration step to the next in each case into further, smaller tiles. This measure can be particularly advantageous if not over the entire component or over the entire surface of the component, but only in certain areas of the component, for example having strong curvatures, a higher number of iterations is performed, as in areas of the surface, the have a lower curvature.

Alternatively, it may also be expedient to first combine the already existing IST data tiles into a complete modified total IST data set for each iteration step, and then to reconstruct the modified total ACTUAL data composed of the ACTUAL data tiles by means of finer fragmentation of the modified total IST data into a higher plurality of IST data tiles.

In the method according to the invention is of particular advantage that the measurement of the component to be tested can be done with a single meter in a single measurement run and the total IST data obtained from this can be used to detect and / or measurement of surface defects of the component. A copy of the total ACTUAL data may e.g. B. then be used in a parallel process to determine form deviations of the component to be tested.

Thus, both surface defects and shape errors can be detected and measured in parallel with a single measurement run in suitable computer structures.

Is, in particular after a sufficient iterative fragmentation, in a first step, a position of a surface defect with respect determined qualitatively of the component to be tested, it may be that this surface defect is present in a transition region of two IST data tiles. If this is the case, it is expedient in a second step to carry out a further dismembering of the actual data in the area of the detected surface defect, in which case the dismemberment of the tiles takes place in such a way that a qualitatively detected surface defect does not occur in the edge region of one of the reconstitutions emerges actual data tiles lies or is traversed by such an edge region of an actual data tile. This measure ensures that a surface error lies as far as possible within an actual data tile, whereby after the alignment of the actual data tile to the corresponding target data then within a single tile the surface defect as a whole lies and with less computational effort can be detected and / or measured.

In addition, the detection and / or determination of the size of the surface defect in the context of such a measure is not affected by the fact that in the border area between two IST-data tiles this due to the tile-individual alignment with the target data, if necessary, a minimum piece overlap or a gap or a discontinuity (jump) arises. Such effects can be masked out with this measure.

In general, it has proven expedient in areas of the surface to be tested which have larger curvatures, the area, ie. H. make the size of the surface of the IST data tiles smaller than in areas of the surface to be inspected with areas that are less curved, d. H. have a lower curvature.

The output of the surface deviations determined in the method according to the invention for visualizing the magnitude of the surface defects can be carried out in a particularly advantageous manner graphically in the form of a false-color representation of the component to be tested.

In the following the invention will be explained in more detail by way of example with reference to the drawing.

Show it:

1 schematically shows a section through SOLL data of a component to be measured;

2 : the cut according to 1 by total actual data of the component obtained during a measurement 1 ;

3 : an overlay of the target data 1 and the total IS data off 2 , where the total actual data is off 2 as a whole as part of a bestfit adaptation to the target data 1 are aligned;

4a FIG. 2: schematically an alignment of a multiplicity of actual data tiles with corresponding target data of the component to be measured; FIG.

4b schematically modified IST data of the component to be tested, which are composed of individual IST data tiles, which were aligned on corresponding target data;

5 : a representation comparable to 3 in which, within an oval mark, in a sectional representation of an alignment of a subarea of the actual data, ie one or more actual data tiles, has been performed on corresponding target data;

6 By way of example, a graphical representation of surface defects of a component to be tested whose position determination and their quantification with the inventive method has taken place.

1 shows a section through SOLL data 1 a component to be tested 2 in the form of a solid line. This TARGET data 1 of the component 2 are, for example, CAD data of the component to be tested 2 ,

In 2 is by a dashed line the same section as in 1 through a real component to be tested 2 shown, where the intersection of total actual data 3 in the context of a measurement of the component to be tested 2 were formed. Within a mark 4 in 2 is a local defect as a surface defect 5 shown schematically. In addition, the real component has 2 according to 2 compared to the target data 1 according to 1 Shape deviations, ie shape errors. Such a shape error is schematically indicated by the fact that a distance x between end points of the cutting line according to 1 is smaller than a corresponding distance x 'between the corresponding endpoints of the slice through the total ACTUAL data 3 according to 2 , Also, the section line according to 2 through the total actual data 3 a flatter curvature than the section line according to 1 through the TARGET data 1 ,

Now the cutting line according to 2 in the form of a bestfit alignment as a whole to the TARGET data 1 according to 1 aligned to best match or minimize the sum of all surface variations obtained, we obtain a superimposed course according to 3 ,

Out 3 it can be seen that the position of the distance x to the distance x 'is different. Furthermore, the distance x 'is significantly greater than the distance x. Such form deviations, which by means of hatching 6 are shown in some areas are significantly larger than differences between the total actual data 3 and the target data 1 in the area of the surface defect 5 (crosshatch 7 ). Thus, there is a risk that with such a superposition of DESIRED data 1 and overall IS data 3 surface defects 5 go into the background and may not be detected.

In 4a is shown that the measured total actual data 3 into a variety of actual data tiles 10 are dismembered. Each of the IST data tiles 10 is optimally aligned to a corresponding target data area. Such alignment may be by way of a so-called bestfit alignment, a merely translational alignment or a merely rotational alignment of the individual IST data tiles 10 relative to the target data 1 respectively. In 4a Such an alignment is simplistic by means of arrows 11 shown. The mutually spaced representation of the IST data tiles 10 in 4a has only representational reasons. As part of the 4a Shall show that each of the IST data tiles 10 to the target data 1 , what a 4a by a representation of a door outer skin 12 is represented, is aligned. In selected areas 13 For example, in the area of a door handle recess, the IST data tiles 10 be chosen smaller to achieve a higher resolution. In the present embodiment, the door handle recess is to be understood merely as an example. It is possible to select regions selected according to empirical values in which surface defects are to be expected, or, for example, regions in which a specific curvature or curvature region exists. The areas where the actual data tiles 10 are finely fragmented, are different from component to component and in particular it may be appropriate, in visible areas of an outer skin component, a higher degree of fragmentation of the IST data tiles 10 than in areas of outer skin parts that are not or less visible.

After all actual data tiles 10 are aligned to their corresponding target data, is expediently from all actual data tiles 10 a modified total IST data set 15 composed, which with the target data 1 is compared. An aligned IST data tiles 10 composite modified total IS data set 15 is exemplary in 4b shown.

In a sectional view according to 5 through the TARGET data 1 and by the modified overall IST record 15 where, for example, a bestfit alignment of the actual data tiles 10 relative to the target data 1 has occurred, surface defects occur 5 (crosshatch 7 ) more clearly and can be detected by means of a data comparison. In such a data comparison in which the maximum surface deviations are determined by the orientation of the individual IST data tiles 10 from which the modified actual data set was obtained, the shape errors of the component to be tested 2 hidden. The shape error of the component to be tested 2 So no longer superimpose the surface defects in terms of their magnitude 5 of the component to be tested. To display the location and size of surface defects 5 in the component to be tested then offers a false color representation, as in 6 is shown on. Local surface defects 5 occur more clearly through the inventive method of processing the total ACTUAL data that has been measured and can be easily graphically visualized. Depending on the amount of deviation of the aligned IST data tiles 10 to the target data 1 For example, the deviation amount can be assigned a specific color value, from which an absolute scale for the surface defect, for example the depth of the surface defect, can be determined on the basis of a corresponding scale.

In the method according to the invention is of particular advantage that there is only one, z. B. optical measurement of the component to be tested needs. From the total actual data obtained, in a simple manner, as is known, by aligning the total measured total actual data to the target data 1 in a simple way, a form error detection take place.

By fragmenting the measured total IST data according to the invention into IST data tiles 10 and their individual orientation to corresponding target data 1 , makes it possible in a simple way, the measured total actual data 3 also for visualization, ie for the detection and measurement of surface defects 5 to use. For this purpose, it is expedient to make a copy of the measured total actual data before the dismemberment of the measured total actual data, wherein the copy of the measured total actual data, for example, to determine the shape error of the component to be tested 2 can be used.

LIST OF REFERENCE NUMBERS

1

NOMINAL data

2

component

3

(Total) ACTUAL data

4

mark

5

surface defects

6

hatching

7

crosshatch

10

Actual data tile

11

arrows

15

modified total IST data set

x, x '

distance

v → w →

vectors

Claims (16)

Method for detecting and / or measuring surface defects ( 5 ) of a component ( 2 ) comprising the steps of: a) measuring a component to be tested ( 2 ) for receiving actual data ( 3 ) from a plurality of measuring points of the component to be tested ( 2 ), b) formation of total actual area data ( 3 ) from the plurality of measuring points; c) providing target data ( 1 ) a surface of the component to be tested ( 2 ), d) aligning the total actual area data ( 3 ) to the target data ( 1 ); e) dismembering the actual data ( 3 ) into a plurality of actual data tiles ( 10 ), f) Individual alignment of all actual data tiles ( 10 ) to the actual data of the tiles ( 10 ) corresponding target data ( 1 ) and g) determining a plurality of surface deviations between measuring points of the aligned IST data tiles ( 10 ) and the actual data tiles ( 10 ) corresponding target data ( 1 h) output of the surface deviations determined in step g). Method according to claim 1, characterized in that the measuring of the component to be tested ( 2 ) by means of optical measuring method, or computer tomographic measurement method takes place. Method according to one of the preceding claims, characterized in that the orientation of the individual IST data tiles ( 10 ) to the target data ( 1 ) via a bestfit alignment, such that the sum of all area deviations of an actual data tile ( 10 ) compared to the corresponding target data ( 1 ) is minimal. Method according to one of the preceding claims, characterized in that the orientation of the individual IST data tiles ( 10 ) to the corresponding target data ( 1 ) individually in each case translationally by means of an IST data-cell-specific displacement along an X, Y and / or Z axis of a coordinate system or along a displacement vector v → having an x and / or a y and / or a z component takes place. Method according to one of the preceding claims, characterized in that the orientation of the individual IST data tiles ( 10 ) to the corresponding target data ( 1 ) Individually in each case via at least one actual data tile-specific rotation about the X, Y and / or Z axis or only an arbitrary vector w → takes place in the coordinate system. Method according to one of the preceding claims, characterized in that the orientation of the individual IST data tiles ( 10 ) to the corresponding target data ( 1 ) in a displacement direction along a surface normal to the IST data tile ( 10 ) or a displacement direction along a surface normal to a best-fit level of the IST data tile (FIG. 10 ) to the corresponding target data ( 1 ) he follows. Method according to one of the preceding claims, characterized in that the orientation of the individual IST data tiles ( 10 ) to the corresponding target data ( 1 ) is carried out by a combination of the measures according to claims 5 to 7, wherein the orientation is due to translations, rotations and / or displacements along surface normals in any order. Method according to one of the preceding claims, characterized in that the IST data tiles ( 10 ) are treated during alignment like rigid bodies or rigid surfaces and are displaced and / or twisted as a whole. Method according to one of the preceding claims, characterized in that the dismemberment of the total actual area data ( 3 ) into a plurality of actual data tiles ( 10 iteratively increasing the number of IST data tiles ( 10 ) from one iteration step to the next iteration step until a predefined termination condition is reached. Method according to one of the preceding claims, characterized in that the predefined termination condition falls below a predetermined page length of an IST data tile ( 10 ) or a predetermined number of iterations or falls below a maximum allowable value of the sum of all surface deviations. Method according to one of the preceding claims, characterized in that already existing IST data tiles ( 10 ) in each iteration step are each divided into further, smaller tiles. Method according to one of the preceding claims, characterized in that before or at each iteration step, the already existing IST data tiles ( 10 ) first again to a modified total IST area data set ( 15 ) and then from the IST data tiles ( 10 ) composite modified total IST area data set ( 15 ) again by finer fragmentation of the composite modified total IST area data set ( 15 ) into a higher plurality of IST data tiles ( 10 ) are dismembered. Method according to one of the preceding claims, characterized in that the measuring of the component to be tested ( 2 ) takes place with a single measuring device in a single measuring pass and the total actual area data ( 3 ) for the detection and / or measurement of surface defects ( 5 ) of the component ( 2 ) and / or wherein, after step d) of claim 1, the pre-directed total IST area data ( 3 ) for determining deviations in shape of the component to be tested ( 2 ) to be provided. Method according to one of the preceding claims, characterized in that a position of a detected surface defect ( 5 ) with regard to the component to be tested ( 2 ) and if the surface defect ( 5 ) in the edge area of an IST data tile ( 10 ) or is traversed by an edge of an IST data tile, at least those IST data tiles ( 10 ), which detects the detected surface defect ( 5 ) and in a predefined environment around the surface error ( 5 ), combined into a partial IST record, and re-fragmenting the sub-actual record into a plurality of actual data tiles ( 10 ) is performed such that a detected surface defect ( 5 ) not in the edge area of an IST data tile ( 10 ) or from an edge of an IST data tile ( 10 ) is traversed. Method according to one of the preceding claims, characterized in that in areas of the surface of the component to be tested ( 2 ) with larger curvatures, an area of the IST data tiles ( 10 ) smaller than in areas of the surface of the component to be tested ( 2 ) with less curved areas. Method according to one of the preceding claims, characterized in that the output of the determined surface deviations is displayed graphically in the form of a false-color representation of the component to be tested ( 2 ) he follows.

Images