DE19815098A1 - Measuring rotary table deviations for co-ordinate measurement machines - Google Patents

Abstract

Test body is mounted at three measurement points on a rotary table, with at least two of the rotary table positions measured. The method involves mounting a test body at three measurement points on the rotary table and measuring it for at least two positions of the rotary table. The test body is mounted at a first measurement point away from the center of the table and near a table surface. A second measurement point is at least approximately in the center of the table and near the table surface. The third measurement point is at least approximately in the center of the table and at a distance from the table surface. The deviation lines of the rotary table are derived from the three measurement sequences for the three measurement points.

Description

The invention relates to a method for measuring Rotary table deviations.

According to the prior art, rotary tables on Ko ordinate measuring devices used. For a use of rotation tables on coordinate measuring machines, it is necessary that the angle values given by the turntable and the deviation axis of the turntable from an ideal rotation axis are as small as possible and the measurement uncertainty of the Ko Do not enlarge the ordinate measuring device.

The deviations of a turntable can be divided into the following components:

• 1. angular position deviations,
• 2. run deviations of the axis of rotation,
  • a) axial deviations,
  • b) radial deviations,
  • c) wobble

According to the prior art (DE 36 37 410 C2) is a Known methods for measuring rotary table deviations where a test specimen with at least three balls is used is measured in several positions on the table becomes.

This has the prior art method the disadvantage that a special test specimen with a Hal is needed. The test specimen must on the one hand Have minimum accuracy. Second, it is expensive to transport this test specimen to the respective customer. The state of the art method has the white  ter disadvantage that a variety of measurements in ver different measuring positions is necessary, which is a certain order set-up time required. Due to the multitude of measurements moreover, the measuring time is very long. Beyond that a complicated evaluation through statistical calculation all measurements required.

The technical problem underlying the invention is a method of measuring turntable bar to specify softs, on the one hand a simple test body is sufficient, in addition to which only a few Measurements are required with a simple retrofit and a short set-up time, which is also a short Has measurement duration, and in which he has only a simple evaluation is required.

This technical problem is caused by the characteristics of the Claim 1 solved.

The fact that a test specimen, for example a Ku gel, measured in three measuring positions, are only three Series of measurements required. For the individual measurements is only a very simple changeover with a short set-up time required. In addition, the evaluation is simple. Another advantage of the method according to the invention is that only low costs for the test specimen, the holder and so on incurred further.

According to the inventive method, the ball for example, off-center at the edge of the turntable posed, as close as possible to the table top surface. The ball can, for example, simply by means of usually attached to her stem in the with profit table tops provided with holes are screwed in.  

In this position, the error caused by the Win setting of the turntable comes about. That is, the correction for the scaling of the angle knife is made.

In the next step, the ball is in the middle of the Tables arranged directly above the table top. Hereby translation errors can be determined. About the Hö hike will be the error of the Z degree of freedom Right.

In the third measuring position, namely in the middle of the Turntables, but at a distance from the turntable, are remaining errors determined, namely the wobble. The fact that the ball in the measuring positions II and III itself when measuring the associated measuring lines hardly in the measuring volume moving the coordinate measuring machine, you have the advantage that geometry errors of the coordinate measuring machine in the measuring Lines II and III do not enter, but only in the measurement line I are important.

The different measuring positions as well as the associated ones Measurement series can also be entered in a different order be taken.

The test specimen is advantageously designed as a ball. But it is also possible to test specimens of other geometries use.

The method according to the invention can advantageously be used tels of the software so that the Kugelqua not received, namely by turning the ball screw sters.  

Further details of the invention are the Unteran to take sayings.

In the drawing is an embodiment of the He shown, namely:

Fig. 1 shows a first measuring position;

Fig. 2 shows a second measurement position;

Fig. 3 shows a third measuring position;

FIG. 4 shows the determination of the reference circle;

Fig. 5 shows the determination of the wobble.

Fig. 1 shows a turntable ( 1 ) which is rotatably mounted in the direction of arrow (A). A test specimen ( 2 ) in the form of a ball is arranged on the turntable ( 1 ) in the vicinity of a surface ( 3 ) of the turntable ( 1 ) in a measuring position (I). The ball ( 2 ) is arranged in the edge region of the rotary table ( 1 ). The ball ( 2 ) is now measured, for example, in 36 positions of the turntable ( 1 ) by the coordinate measuring machine (not shown). This corresponds to 10 ° steps.

The ball ( 2 ) is measured in a first position ( 4 ) (zero position). Each spherical measurement must consist of at least four probes of the spherical surface in order to obtain a center evaluation of the current spherical position. Each ball measurement advantageously consists of nine to fifteen points.

After this measurement, the ball ( 2 ) is moved into a second position ( 5 ) (shown in dashed lines) by rotating the turntable ( 1 ) in the direction of the arrow (A), and the ball ( 2 ) is measured again. This is carried out in further positions ( 5 ', 5 '') until the ball ( 2 ) is again in the zero position ( 4 ).

Referring to FIG. 2, the ball (2) is arranged in the center (6) of the turntable (1) in the measuring position (II), directly on the table surface (3). Here, too, the ball ( 2 ) is again measured, for example, in the 36 positions of the turntable ( 1 ), each ball measurement in turn consisting of at least four points, whereby the position of the ball center point of the ball ( 2 ) is detected.

Referring to FIG. 3, the ball (2) is again arranged in the center (6) of the turntable. The ball ( 2 ) is now arranged at a greater distance via an extension ( 7 ) to the surface ( 3 ) of the turntable ( 1 ) in the measuring position (III). The extension ( 7 ) can be a commercially available extension that is screwed into the center ( 6 ) of the rotary table ( 1 ). A set-up time of less than two minutes is required for this.

In this position, the ball ( 2 ) is again measured in, for example, 36 positions of the turntable ( 1 ) and the position of the center of the ball of the ball ( 2 ) is detected.

The six angle-dependent deviation lines of the rotary table ( 1 ) (see VDI 2617, part 4) can be calculated from the three measurement series of the measurement positions (I, II, III).

According to Fig. 4, the results from which it most test series is using (series of measurements according to Fig. 1) formed a best total Watched circuit (8), the reference circle is referred to (8) below as, whose surface normals with the center an axis (9 ) Are defined. All the following measurements relate to the axis ( 9 ). It is referred to below as the reference axis ( 9 ).

The center of the sphere ( 2 ) was determined for the zero position ( 4 ) for a point ( 10 ) in the coordinate system. For the position ( 5 ), the center of the sphere ( 2 ) was determined on a position ( 11 ) of the coordinate system. The point ( 11 ) does not lie on the reference circle ( 8 ), so that there is a deviation of the actual position ( 11 ) from a target position ( 12 ). There is also a deviation of the actual position from the target position in the other positions ( 5 ', 5 ''). For position ( 4 ), target and actual position ( 10 ) match.

A measuring point ( 14 ), which represents a measured center of the sphere ( 2 ), shows that there is not only a radial deviation from the reference circle ( 8 ), but also an angular deviation dϕ. The measuring point ( 14 ) has an actual position ( 15 ) with the coordinates (x _I , y _I , z _I ) with a deviation (dx, dy, dz) from a target position ( 16 ) with the coordinates (x _S , y _S , z _S ).

The angular position-dependent radial position deviations of the table center point ( 6 ) are determined from the radial deviations of the center position of the second measuring row according to FIG. 2 to the reference axis.

The axial deviations (pitch) are calculated from the axial (in the direction of the reference axis) deviations of the Mit point positions to the reference circle determined.  

Referring to FIG. 5, the angular position-dependent dew be melabweichungen of the table (1) from differences between the ra Dialen center deviations of each of the second and third series of measurements (Fig. 2 and 3) to the reference axis (9) on the resultant by the difference in height lever (h) certainly.

The wobble angle δ is determined from the deviation (b) of the ball center of the ball ( 2 ) in position (II) according to FIG. 2 and the deviation (a) of the ball center of the ball ( 2 ) in position (III) according to FIG . 3 as well as of the lever (h) according to the following formula:

The angular position deviations of the turntable ( 1 ) are determined from the center positions of the first measurement series ( Fig. 1) to the respective position on the reference circle, taking into account the angular position-dependent radial position deviations of the table center in the X and Y directions ( Fig. 1).

With

= Vector to the target point,
= Vector to the actual point.

The deviation of the center of the sphere is determined as follows:

x = x _I -x _S
y = y _I -y _S
z = z _I -z _S (3).

Characterized in that the ball ( 2 ) of the second and third series of measurements ( Fig. 2 and 3) in the middle ( 6 ) of the rotary table ( 1 ) is close to the axis of rotation and during the 36 measuring positions, for example, hardly in the measuring volume of the coordinate measuring machine ( not shown) moves, the geometry errors of the coordinate measuring machine are not included here.

Five of the six deviation lines of the table are already calculated from the second and third measurements ( FIGS. 2 and 3).

Only when the first measuring line ( FIG. 1) is recorded is the ball ( 2 ) moved through the measuring volume, so that local deviations in the accuracy of the coordinate measuring machine influence the result. By several measurements of the first series of measurements ( Fig. 1) in different angular positions ( 4 , 5 , 5 ', 5 '') of the ball ( 2 ) off-center on the table top and averaging over the measurements, the results can be improved. Here there is a dependency on the geometry errors of the coordinate measuring machine.

The results of the second and third measurements ( FIGS. 2 and 3) can also be stabilized if the results are recorded and averaged several times. The resulting somewhat improved stability of the results is not necessarily of great advantage if you consider the increased measuring time. One round per measuring line is normally sufficient.

The six error components of the rotary table obtained at the measured angular positions ( 4 , 5 , 5 ', 5 '') are long-period errors. Interpolation values can be obtained by interpolation.

The total duration of a single measurement (one revolution per measuring line) with conversion to the three spherical positions takes about 45 minutes. Outliers when measuring the ball ( 2 ), for example due to dirt on the ball surface, can be identified by checking the ball shape. According to the invention, the influences of the spherical shape itself can be reduced by rotating the probing pattern on the sphere ( 2 ) using appropriate software. This means that probing is always carried out at the same points on the surface of the sphere. So balls of poorer quality could also be used.

The correction is independent of the position of the rotary table ( 1 ) in the measuring volume of the coordinate measuring machine. The process can also be carried out very easily, for example, on rotary tables with a horizontal axis, since the balls only have to be screwed into the table top. No special bracket is necessary.

Reference list

1

Turntable

2nd

Bullet

3rd

surface

4th

Zero position

5

,

5

',

5

'' Positions

6

Focus

7

renewal

8th

Reference circle

9

Reference axis

10th

Measuring point

11

Actual position (measuring point)

12th

Target position

14

Measuring point

15

Actual position

16

Target position
A arrow
dϕ angular deviation
d, dx, dy deviation
h lever
δ wobble angle
a, b deviations
I, II, III measuring positions
Vector to the target point
Vector to the actual point

Claims (14)

1. A method for measuring rotary table deviations characterized in

• - That a test specimen ( 2 ) is arranged and measured in three measuring positions (I, II, III) on a rotary table ( 1 ),
• - The test specimen ( 2 ) is arranged in a first measuring position (I) outside a center ( 6 ) of the turntable ( 1 ) and in the vicinity of a turntable surface ( 3 ),
• - That the test specimen ( 2 ) in a second measuring position (II) at least approximately in the middle ( 6 ) of the turntable ( 1 ) and in the vicinity of the turntable surface ( 3 ) is net angeord,
• - And that in a third measuring position (III) the test specimen ( 2 ) is arranged at least approximately in the middle ( 6 ) of the rotary table ( 1 ) and at a distance from the surface of the rotary table ( 3 ),
• - And that the deviation lines of the turntable are determined from the three series of measurements of the three measuring positions (I, II, III).

2. The method according to claim I, characterized in that the arrangement and subsequent measurement of the test specimen ( 2 ) in the three measuring positions (I, II, III) takes place in a different order. 3. The method according to claim 1 or 2, characterized in that the test body ( 2 ) in each measuring position (I, II, III) in at least two positions ( 4 , 5 , 5 ', 5 '') of the rotary table (I) is measured. 4. The method according to any one of the preceding claims, characterized in that the test body ( 2 ) is designed as a ball. 5. The method according to claim 4, characterized in that a ball measurement from measurements of at least four Points on a spherical surface. 6. The method according to claim 1 or 2, characterized in that the six angle-dependent deviation lines of the rotary table ( 1 ) are calculated. 7. The method according to claim 6, characterized in that with the help of the results from the series of measurements (I) with the test specimen arrangement eccentrically of the rotary table ( 1 ) a well-fitted circle (reference circle) ( 8 ) is formed, the surface normal together with the center point an axis (Reference axis) ( 9 ) defined. 8. The method according to claim 6, characterized in that the angular position-dependent radial position deviations of the rotary table center ( 6 ) from the radial deviations from the center positions of the series of measurements (II) with the test specimen arrangement in the middle and in the vicinity of the surface of the rotary table are determined. 9. The method according to claim 6, characterized in that the axial deviations (pitch) from the axial deviations of the center positions of the test specimen ( 2 ) to the reference circle ( 8 ) are determined. 10. The method according to claim 6, characterized in that the angular position-dependent wobble deviations of the rotary table ( 1 ) from the differences in the radial center point deviations from the series of measurements (II) with the test body arrangement in the middle ( 6 ) of the rotary table ( 1 ) and in Proximity of the surface ( 3 ) and from the measurement series (III) with the test specimen arrangement in the center ( 6 ) of the rotary table ( 1 ) and at a distance from the surface ( 3 ) to the reference axis ( 9 ) can be determined. 11. The method according to claim 6, characterized in that the angular position deviations of the turntable ( 1 ) from the center positions of the measurement series (I) with the test specimen arrangement eccentrically of the turntable ( 1 ) to the respective position on the reference circle ( 8 ) are determined. 12. The method according to claim 1 or 2, characterized in that a rotation of the probing pattern of the test body pers ( 2 ) is carried out by means of software. 13. The method according to claim 1 or 2, characterized in that an interpolation of values between the recorded positions of the turntable ( 1 ) is carried out. 14. The method according to claim 1 or 2, characterized in that a shape monitoring of the test body ( 2 ) is carried out for the detection of measurement outliers.