JP2010223887A - Device for measuring error in five degrees of freedom - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a device for detecting an error in five degrees of freedom which enables high-accuracy measurement of a movement error in the directions of five axes in all which is produced with rotation of a rotating body. <P>SOLUTION: A diffraction grating surface 3 formed of a plurality of circumferential grooves provided in the direction of the axis of rotation of the rotating body 2, with a prescribed space left between, is provided on the outer peripheral surface of the rotating body 2. Two-axes interference sensor units S1, S2 and S3 which detect the relative movement displacement of the diffraction grating surface along a first axis of a sensor coordinate system corresponding to the radial direction of the rotating body 2 and a second axis of the sensor coordinate system corresponding to the axial direction of the rotating body 2, are disposed fixedly at three different positions, at least, in the circumferential direction of the rotating body 2, spaced outside in the radial direction from the outer peripheral surface of the rotating body 2 and prevented from overlapping each other, and the movement errors Δx, Δy, Δz, Δu and Δv in the directions of five axes in all are determined simultaneously, based on interference signals on the first axis of the sensor coordinate system and those on the second axis of the sensor coordinate system of the three two-axes interference sensor units S1, S2 and S3. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  According to the present invention, the rotating body mounted on the bearing or the rotating body constituting the bearing itself or a part of the rotating body rotates, and the rotating body has two axes in the radial direction, one axis in the axial direction, and two axes in the inclined direction. The present invention relates to a 5-degree-of-freedom error measurement device that simultaneously measures motion errors occurring in the system.

  The allowable range of motion errors that accompany rotation of rotating bodies such as end mills and rotating wheels mounted on various types of bearings and various NC machine tools, etc. is required to raise the level of machining accuracy. In recent years, it has become increasingly severe.

The motion errors that occur in the rotating body include radial motion that occurs in the plane of two orthogonal axes in the radial direction of the rotating body, axial motion that occurs along one axis in the axial direction of the rotating body, and the axis of the rotating body is tilted There is an error in the direction of 5 axes such as an angular motion that occurs around the 2 axes in the direction of movement.
It is desirable to keep the motion error of radial motion and axial motion within the nanometer level and the motion error of angular motion within the second level.

  Conventionally, the measurement of motion error of a rotating body has been performed by a multi-step method or a multi-probe method including the inversion method using a reference sample with high roundness and sphericity. Requires a procedure for separating the roundness and sphericity errors of the reference sample, and the post-processing is complicated.

On the other hand, as a method for measuring an error related to the rotation of the rotating body in a non-contact manner, an axial encoder composed of an encoder scale or an optical head having alternating linear band-shaped light reflecting portions and light transmitting portions is already known. In addition, by using an interference signal detected by a three-axis displacement sensor composed of a scale element having a fine shape on two axes and a sensor head, the displacement of XY in two orthogonal axes or XY Non-Patent Document 2 proposes a diffracted light interference type three-axis nano-displacement sensor that simultaneously measures the displacement of -Z in three orthogonal axes.
The former axial encoder is specialized for measuring a motion error in one axial direction related to the axial motion of a rotating body.
The latter diffracted light interference type three-axis nano-displacement sensor can measure the motion error in the three-axis direction related to radial motion and axial motion. Since the configuration is such that the positional error is specified and the motion error in the three-axis direction is measured, as it is, the two axes related to the angular motion generated in the orientation change of the scale element, that is, the direction in which the axis of the rotating body is inclined. There is an inconvenience that it is impossible to measure the movement error in the direction.

Co-authored by Seiji Kimoto, Mitsuyoshi Nomura, Takayuki Shibata, Yoshihiko Murakami, Jun Horiuchi and Masami Hamada . 583 Jointly written by Akihide Kimura, Yoshikazu Arai and Kowei, "Proceedings of Academic Lecture Meeting of the 2008 Annual Meeting of Precision Engineering," Study on Diffractive Interferometric 3-Axis Nano Displacement Sensor (3rd Report) ", Japan Society for Precision Engineering, p. 233

  Therefore, an object of the present invention is to measure a movement error in a convenient five-axis direction with high accuracy including a radial motion, an axial motion, and an angular motion generated with the rotation of a rotating body without requiring complicated post-processing. An object of the present invention is to provide a 5-degree-of-freedom error measuring device having a compact structure capable of providing

The 5-degree-of-freedom error measuring device according to the present invention includes a rotating body attached to a bearing or a bearing or a rotating body constituting a part thereof, and two axial axes of the rotating body and one axial axis. And a five-degree-of-freedom error measuring device for measuring a motion error occurring in two axes in the tilt direction.
On the outer circumferential surface of the rotating body, provided with a diffraction grating surface formed from a plurality of circumferential grooves provided at a certain interval in the direction of the rotation axis of the rotating body,
A biaxial interference sensor unit for detecting a relative displacement of a diffraction grating surface along a sensor coordinate system first axis corresponding to a radial direction of the rotating body and a sensor coordinate system second axis corresponding to an arrangement direction of the circumferential grooves; Fixedly arranged at at least three different positions in the circumferential direction of the rotating body with a gap from the outer peripheral surface to the outer side in the radial direction of the rotating body and not overlapping each other in the radial direction of the rotating body And
The diameter of the rotating body based on the amount of movement obtained from each of the interference signals of the first axis of the sensor coordinate system output from at least two of the two-axis interference sensor units that are not simultaneously positioned along the diameter direction of the rotating body. Determining a motion error occurring in the two axes in the direction, and based on at least one of the movement amounts obtained from each of the interference signals of the second axis of the sensor coordinate system output from at least three of the two-axis interference sensor units. A motion error occurring in one axis in the axial direction of the rotating body is obtained, and at least 2 of each movement amount obtained from each of the interference signals of the sensor coordinate system second axis output from at least three of the two-axis interference sensor units. There is provided an error calculating means for obtaining a motion error generated in two axes in the tilt direction of the rotating body based on each combination.

In this way, each of the two-axis interference sensor units is along the sensor coordinate system second axis corresponding to the alignment direction of the sensor coordinate system first axis corresponding to the radial direction of the rotating body and the circumferential groove, that is, the direction of the rotation axis. It is configured to detect the relative displacement of the diffraction grating surface, each of which is spaced from the outer peripheral surface of the rotating body in the radial direction of the rotating body and does not overlap each other in the radial direction of the rotating body. Since the rotating body is fixedly disposed at at least three different positions in the circumferential direction, the sensor coordinate system output from at least two two-axis interference sensor units that are not positioned along the same diameter direction of the rotating body. With each of the uniaxial interference signals, at least two radial motion errors having different directions can be measured among the motion errors generated by the rotation of the rotating body. Since the angle formed by the first axis of the sensor coordinate system of the two two-axis interference sensor units is known at the design stage, by combining these at least two motion errors by error calculation means, the orthogonal 2 in the radial direction of the rotating body is obtained. It is possible to obtain motion errors Δx and Δy related to radial motion occurring in the plane of the axes (X and Y axes).
Further, since the interference signal of the second axis of the sensor coordinate system output from at least three two-axis interference sensor units is a value corresponding to the movement displacement in the circumferential groove arrangement direction, that is, the rotation axis direction, at least Based on at least one of the movement amounts obtained from the interference signals of the second axis of the sensor coordinate system output from the three two-axis interference sensor units, this occurs in one axis (Z axis) in the axial direction of the rotating body. The motion error Δz in one axis direction related to the axial motion can be obtained.
Further, since at least three two-axis interference sensor units detect movement displacement in the circumferential groove arrangement direction, that is, the direction of the rotating shaft, at different positions outside the rotating body in the radial direction, A combination of at least two movement amounts obtained from each of the interference signals of the second axis of the sensor coordinate system output from at least three two-axis interference sensor units fixedly arranged at different positions, more specifically, Sensor coordinate system second axis obtained from a biaxial interference sensor unit spaced apart in the X-axis direction and a deviation in the movement amount of the sensor coordinate system second axis obtained from the biaxial interference sensor unit separated from each other in the Y-axis direction Motion errors Δu and Δv related to the angular motion generated around the two orthogonal axes (X and Y axes) in the direction in which the axis of the rotating body is inclined, ie, Y An attitude change Δu around the X axis related to the deviation of the movement amount obtained from each of the interference signals of the second axis of the sensor coordinate system output from at least two two-axis interference sensor units having different positions in the axial direction, and the X-axis It is possible to determine the posture change Δv about the Y axis related to the deviation of the movement amount obtained from each of the interference signals of the second axis of the sensor coordinate system output from at least two two-axis interference sensor units having different positions in the direction. It becomes.
Further, since the two-axis interference sensor unit is arranged on the one end face side of the rotating body, there is no inconvenience that the overall structure is enlarged.

More specifically, two of the three two-axis interference sensor units are arranged along the diameter direction of the rotating body, and the other one is arranged on a radius orthogonal to the diameter direction. ,
The error calculation means includes
The average of the movement amounts obtained from the interference signals of the first axis of the sensor coordinate system output from the two two-axis interference sensor units arranged along the diameter direction is set as one axis in the radial direction of the rotating body. The amount of movement obtained from the interference signal of the first axis of the sensor coordinate system that is output from the one two-axis interference sensor unit arranged on the radius as the motion error to be generated is the other one in the radial direction of the rotating body. A motion error that occurs in the shaft,
The average of each movement amount obtained from each of the interference signals of the sensor coordinate system second axis output from the three two-axis interference sensor units is a motion error generated in one axis in the axial direction of the rotating body,
The average of each moving amount obtained from each of the interference signals of the second axis of the sensor coordinate system output from the two two-axis interference sensor units arranged along the diameter direction and the one two arranged on the radius Based on the deviation from the movement amount obtained from the interference signal of the sensor coordinate system second axis output from the axis interference sensor unit, the motion error occurring in one axis in the tilt direction of the rotating body is obtained, and along the diameter direction Based on the deviation of each movement amount obtained from each of the interference signals of the second axis of the sensor coordinate system output from the two two-axis interference sensor units arranged in the same manner, It is desirable to configure to determine the resulting motion error.

Thus, in a configuration in which two of the three two-axis interference sensor units are arranged along the diametrical direction of the rotating body and the other one is arranged on a radius orthogonal to the diametrical direction. The average of the movement amounts obtained from each of the interference signals of the first axis of the sensor coordinate system output from the two two-axis interference sensor units arranged along the diameter direction is set as one axis in the radial direction of the rotating body. For example, the motion error Δx in the X-axis direction and the interference signal of the first axis of the sensor coordinate system output from another one of the two-axis interference sensor units arranged on the radial direction orthogonal to the Y axis, that is, the Y axis Can be used as a movement error Δy in the other axis in the radial direction of the rotating body, that is, in the Y-axis direction. The sensor coordinate system first axis of the two two-axis interference sensor units is along one axis (X axis) in the radial direction of the rotating body. At the same time, the sensor coordinate system first axis of the other one of the two-axis interference sensor units is Since it is arranged along one other axis (Y axis) that is orthogonal to one axis, the radial in the two orthogonal axes (X, Y axes) in the radial direction of the rotating body without performing processing such as coordinate conversion The motion error of the motion can be obtained immediately, and the time required for the arithmetic processing is shortened.
In addition, an axial motion in which an average of each moving amount obtained from each of the interference signals of the sensor coordinate system second axis output from the three two-axis interference sensor units is generated on one axis (Z axis) in the axial direction of the rotating body. It is assumed that the motion error Δz related to. Since the average of the movement amounts measured at three different positions in the circumferential direction of the rotating body is the motion error Δz of the axial motion generated in one axis (Z axis) of the rotating body, the rotating body is in the angular direction. Even when the posture changes Δu and Δv occur, the motion error Δz of the axial motion can be accurately obtained.
Further, the average of the respective movement amounts obtained from each of the interference signals of the second axis of the sensor coordinate system output from the two two-axis interference sensor units arranged along one axis (X axis) in the diametrical direction, and this Based on a deviation from the movement amount obtained from the interference signal of the second axis of the sensor coordinate system output from the one two-axis interference sensor unit arranged along one radial axis (Y-axis) orthogonal to this. The posture change Δu, which is a motion error related to the angular motion generated around one of the two orthogonal axes (X axis) in the direction in which the axis of the rotating body is inclined, is obtained, and one axis in the diametrical direction (X axis) Is orthogonal in the direction in which the axis of the rotating body is inclined based on the deviation of each moving amount obtained from each of the interference signals of the second axis of the sensor coordinate system output from the two two-axis interference sensor units arranged along Around the other one of the two axes (Y axis) Attitude change Δv is a movement error associated with angular motion can be obtained. As in the case of the motion error of radial motion, the first axis of the sensor coordinate system of the two two-axis interference sensor units is along one axis (X axis) in the radial direction of the rotating body, and at the same time, another one of the two-axis interference sensors. Since the first axis of the sensor coordinate system of the unit is arranged along the other one axis (Y axis) orthogonal to the one axis, there is no need to perform complicated coordinate conversion processing, and around the X axis The posture change Δu and the posture change Δv around the Y axis can be easily obtained, and the time required for the arithmetic processing is shortened.

  In addition to the above-described configuration, a rotation driving means for rotating the rotating body is further provided, and the error calculation means is configured so that each time the rotating body rotates [1 / n], two radial axes and an axial direction of the rotating body A data storage means for obtaining a motion error occurring on one axis and two axes in the tilt direction and storing the error in correspondence with the absolute rotation angle of the rotating body, and the data read from the data storage means are visually displayed. Data display means may be provided.

  The rotating body is rotated in increments of [1 / n], and the rotational error of the rotating body is determined by calculating the motion error that occurs in the two axial axes of the rotating body, one axial axis, and two inclined axes at each rotational angle. By storing the data in the data storage means in correspondence with the specific rotation angle, and further reading the data in the data storage means and displaying it visually on the data display means, it is possible to easily detect the motion error in the 5-axis direction that occurs at each rotation angle of the rotating body. Can grasp.

  Further, every time the rotating body rotates [m + (1 / n)], a motion error generated in the two radial axes of the rotating body, the one axial axis, and the two tilted axes is obtained. You may comprise so that it may memorize | store corresponding to an absolute rotation angle.

When such a configuration is applied, the motion error in the 5-axis direction is measured by shifting the rotation angle corresponding to [1 / n] rotations every m rotations of the rotating body, and this value is stored in the data storage means. Therefore, even if the time required for the processing required for measurement and data writing is long, or even when the rotational speed of the rotating body is considerably high, the computation processing is performed without difficulty and each rotation It is possible to easily grasp the motion error in the five-axis direction caused by the angle.
For example, when m = 2 and n = 36, the measurement and calculation of motion error and the data storage process need only be performed every time the rotating body rotates [2 × 360 + 10] °, that is, 730 °. In spite of the measurement of the motion error at an absolute rotation angle of 10 °, the calculation and data storage processing may be performed while the rotating body rotates 730 °.

  The five-degree-of-freedom error measuring device of the present invention is provided with a diffraction grating surface formed of a plurality of circumferential grooves provided at a certain interval in the direction of the rotation axis of the rotating body on the outer peripheral surface of the rotating body. A two-axis interference sensor unit for detecting the relative displacement of the diffraction grating surface along the sensor coordinate system first axis corresponding to the radial direction of the body and the sensor coordinate system second axis corresponding to the alignment direction of the circumferential grooves is provided on the rotating body. The at least three two-axis interference sensor units are fixedly arranged at at least three different positions in the circumferential direction of the rotating body so as not to overlap each other with a gap from the outer peripheral surface to the radial outer side of the rotating body. Radial motion generated in the plane of two axes perpendicular to the radial direction of the rotating body based on the moving amount obtained from the interference signal of the first axis of the sensor coordinate system, that is, the moving amount having different directionality in the radial direction of the rotating body Related luck Based on at least one of the amount of movement obtained from each of the interference signals of the second axis of the sensor coordinate system output from the at least three two-axis interference sensor units, that is, the amount of movement generated in the axial direction of the rotating body. The motion error in one axis direction related to the axial motion is obtained, and the second axis interference of the sensor coordinate system output from at least three two-axis interference sensor units fixedly arranged at different positions in the circumferential direction of the rotating body. This is a motion error related to angular motion generated around two orthogonal axes in the radial direction of the rotating body in the direction in which the axis of the rotating body is inclined based on a combination of at least two movement amounts obtained from each of the signals. Since the posture change around the axis is calculated, there is no need for complicated post-processing such as separating the roundness and sphericity errors of the reference sample. Radial motion caused by the rotation of the rotating body, an axial motion, convenient five-axis direction of the combined angular motion a motion error can be measured directly in high precision.

  In particular, two of the three two-axis interference sensor units are arranged along the diametrical direction of the rotating body and the other one is arranged on a radius orthogonal to the diametrical direction, along the diametrical direction. While the average of the movement amounts obtained from each of the interference signals of the first axis of the sensor coordinate system output from the two two-axis interference sensor units deployed is a motion error that occurs in one axis in the radial direction of the rotating body, The amount of movement obtained from the interference signal of the first axis of the sensor coordinate system that is output from the other one of the two-axis interference sensor units arranged on the radius is defined as a movement error that occurs on the other axis in the radial direction of the rotating body, The average of the amount of movement obtained from each of the interference signals of the second axis of the sensor coordinate system output from the three two-axis interference sensor units is set as the motion error generated in one axis in the axial direction of the rotating body, and in the diameter direction. The two two deployed along Output from the other one of the two-axis interference sensor units arranged on a radius orthogonal to the average of the movement amounts obtained from the interference signals of the second axis of the sensor coordinate system output from the interference sensor unit. The movement error generated in one axis in the tilt direction of the rotating body is obtained based on the deviation from the movement amount obtained from the interference signal of the second axis of the sensor coordinate system, and further, the two of the two arranged along the diameter direction are obtained. Based on the deviation of each movement amount obtained from each of the interference signals of the second axis of the sensor coordinate system output from the two-axis interference sensor unit, a motion error generated in the other one axis in the tilt direction of the rotating body is obtained. Therefore, without performing complicated processing such as coordinate transformation, the motion error of radial motion in two orthogonal axes in the radial direction of the rotating body, the posture change around one axis, and the posture change around the other one axis A Gyula motion it is possible to easily determine the motion error in a short time, moreover, it is possible to determine the motion error associated with axial motion occurring in one axis in the axial direction of the rotating body more accurately.

  Further, a rotation driving means for rotating the rotating body is provided, and each time the rotating body rotates [1 / n], a motion error that occurs in the two axial axes of the rotating body, one axial axis, and two tilting axes is detected. The five-axis directions generated at each rotation angle are obtained by storing the data in the data storage means in correspondence with the absolute rotation angle of the rotating body and reading the data in the data storage means and displaying the data on the data display means. The movement error can be easily grasped.

  Also, every time the rotating body rotates [m + (1 / n)], the motion error generated in the two radial axes of the rotating body, the one axial axis, and the two tilted axes is obtained to obtain the absolute value of the rotating body. By adopting a configuration that stores data corresponding to the rotation angle, even if the processing time required for measurement and data writing is long, or even when the rotational speed of the rotating body is considerably high, it is reasonable. By performing arithmetic processing, it becomes possible to easily grasp the motion error in the five-axis directions that occurs at each rotation angle.

FIG. 6 is a schematic view showing the positional relationship between three biaxial interference sensor units constituting part of the five-degree-of-freedom error measuring apparatus 1 according to an embodiment of the present invention and the diffraction grating surface provided on the outer peripheral surface of the rotating body. FIG. 1A is a plan view of the rotating body viewed from above, and FIG. 1B is an elevational view of the rotating body viewed from the front. It is the functional block diagram which showed simply about the principle of operation of the biaxial interference sensor unit in the embodiment. An example of another two-axis interference sensor unit in which interference light having a phase difference of 90 ° is generated at the position of the photodiode of the two-axis interference sensor unit and an interference signal is measured by another photodiode is shown in a simplified manner. FIG. 3A is a configuration diagram, in which FIG. 3A shows a two-axis interference sensor unit from a viewpoint along the Y axis, and FIG. 3B shows a two-axis interference sensor unit from a viewpoint along the Z axis. Is shown. NC machine tool that rotationally drives a chuck that is a kind of rotating body, a numerical control device that drives and controls the NC machine tool, and three two-axis interference sensor units that are fixedly arranged on the column head of the NC machine tool and the chuck side The block diagram which simplified and showed the connection relation of the main body part of the 5-degree-of-freedom error measuring device comprised as a main part, and the personal computer etc. which function as an error calculating means of a 5-degree-of-freedom error measuring device It is. It is the flowchart shown about the outline | summary of the program for an error measurement installed in the error calculating means. It is a continuation of the flowchart shown about the outline | summary of the program for an error measurement. It is a continuation of the flowchart shown about the outline | summary of the program for an error measurement. It is a continuation of the flowchart shown about the outline | summary of the program for an error measurement. It is the conceptual diagram which showed the logic structure of the data storage file produced | generated by the data storage means of an error calculating means. 10A is a conceptual diagram showing the relationship between the absolute rotation angle of the chuck and the total rotation angle when the absolute rotation angle of the chuck when one rotation signal is output is defined as 0 °, and FIG. 10A shows the absolute rotation angle. Is the state when the total rotation angle is 0 ° and FIG. 10B is the state when the absolute rotation angle is 10 ° and the total rotation angle is [360 × m + 10] °, and FIG. 10C is the absolute state. The state is shown when the rotation angle is 350 ° and the total rotation angle is [360 × (n−1) · m + (n−1) · 10] °.

  Embodiments of the present invention will be specifically described below with reference to the drawings.

  FIG. 1 shows three biaxial interference sensor units S1, S2 and S3 constituting a part of a five-degree-of-freedom error measuring apparatus 1 according to an embodiment to which the present invention is applied, and an outer peripheral surface of a chuck 2 which is a kind of rotating body. FIG. 1A is a conceptual diagram showing a simplified correspondence between positions with the diffraction grating surface 3 provided in FIG. 1. FIG. 1A is a plan view of the chuck 2 as viewed from above, and FIG. ) Is an elevation view of the chuck 2 as seen from the front.

  As shown in FIG. 1B, the diffraction grating surface 3 is formed on the outer peripheral surface of the chuck 2 by a plurality of peripheral grooves provided at a constant interval in the direction of the rotation axis CL of the chuck 2. ing. The outer peripheral surface of the chuck 2 that forms the diffraction grating surface 3 must be processed into an accurate cylindrical shape as much as possible, and the pitch of the peripheral grooves needs to be processed as constant as possible.

Each of the biaxial interference sensor units S1, S2, and S3 detects a relative movement displacement of the measurement object in two orthogonal directions of the sensor coordinate system first axis and the sensor coordinate system second axis. Each of the interference sensor units S1 and S2 has its sensor coordinate system first axis along one of the diameter directions of the chuck 2, that is, along the X-axis direction shown in FIG. As shown in b), the chuck 2 is fixedly arranged on a column head of an NC machine tool (not shown) at a certain distance radially outward from the outer peripheral surface of the chuck 2.
In addition, the biaxial interference sensor unit S3 has the first axis of the sensor coordinate system along the Y axis, which is the one radial direction described above, that is, the other radial direction orthogonal to the X axis. As shown in FIG. 2, the chuck 2 is fixedly arranged on a column head of an NC machine tool (not shown) at a certain distance from the outer peripheral surface of the chuck 2 to the outside in the radial direction.

  That is, the sensor coordinate system first axis of the three two-axis interference sensor units S1, S2, and S3 detects the relative displacement of the diffraction grating surface 3 that is the measurement object in the direction corresponding to the radial direction of the chuck 2. The second axes of the sensor coordinate systems of all the biaxial interference sensor units S1, S2, and S3 correspond to the Z-axis direction that coincides with the alignment direction of the circumferential grooves on the diffraction grating surface 3, that is, the axial direction of the chuck 2. The relative displacement of the diffraction grating surface 3 that is the object to be measured is detected in the direction, each of which has a gap from the outer peripheral surface of the chuck 2 to the radially outer side of the chuck 2, and the chuck 2. The chucks 2 are fixedly arranged at at least three different positions in the circumferential direction of the chuck 2 so as not to overlap each other in the radial direction.

  FIG. 2 is a functional block diagram simply showing the operation principle of the biaxial interference sensor units S1, S2, and S3.

  Since the configurations of the biaxial interference sensor units S1, S2, and S3 are the same, the relative displacement of the diffraction grating surface 3 in the X-axis direction is detected by the sensor coordinate system first axis and the sensor coordinate system second axis is used here. The configuration will be described by taking as an example a biaxial interference sensor unit S1 (XZ sensor) that detects the relative displacement of the diffraction grating surface 3 in the Z-axis direction.

  The diffraction grating surface 3 shown in FIG. 2 shows a part of the diffraction grating surface 3 in FIGS. 1 (a) and 1 (b). The arrangement direction of the grooves on the diffraction grating surface 3 is shown in FIG. 2, the direction is the Z axis.

  As shown in FIG. 2, the biaxial interference sensor unit S <b> 1 generally includes a laser diode 4 serving as an output source of laser light, reflecting mirrors 5 to 8 and beam splitters 9 to 10, and a quarter wavelength plate 11. ˜16 and ½ wavelength plate 17, reference fixed mirror 18, and photodiodes PD <b> 11 to PD <b> 12.

  In the above configuration, the laser light A emitted from the laser diode 4 is incident on the beam splitter 9 via the reflecting mirrors 5 to 6 and along the X axis along with the laser light A1 directed to the reference fixed mirror 18. The laser beam A <b> 2 is directed toward the diffraction grating surface 3.

  Of these, the laser beam A1 directed to the fixed mirror 18 is irradiated to the fixed mirror 18 via the quarter-wave plate 16 and reflected by the fixed mirror 18, and follows the reverse path as it is to the quarter wavelength. The light enters the beam splitter 9 through the plate 16.

  Further, the laser light A2 directed toward the diffraction grating surface 3 along the X axis is irradiated to the diffraction grating surface 3 via the quarter-wave plate 11, and the zero-order light B and ± The light is divided into primary lights B1 and B2.

  Of these, the zero-order light B follows the reverse path as it is, enters the beam splitter 9 through the quarter-wave plate 11, and is combined with the laser light A1 reflected by the fixed mirror 18 to become interference light. Then, the light enters the photodiode PD12 via the quarter-wave plate 15.

  Further, the + 1st order light B1 reaches the reflecting mirror 8 via the quarter wavelength plate 13 and the half wavelength plate 17, is reflected by the reflecting mirror 8 and enters the beam splitter 10, and the −1st order light B2 Reaches the reflecting mirror 7 via the quarter-wave plate 12, is reflected by the reflecting mirror 7, and enters the beam splitter 10. These ± primary lights B1 and B2 are combined by the beam splitter 10 to become interference light, and this interference light enters the photodiode PD11 via the quarter-wave plate 14.

The interference signals I _PD11 , I _PD12 detected by the photodiodes PD11, PD12 are the displacement Δz in the Z-axis direction and the displacement Δx in the X-axis direction of the diffraction grating relative to the biaxial interference sensor unit S1, and the pitch d of the diffraction grating and Depending on the wavelength λ of the light source, it can be expressed by the following equations (1) and (2) (where A and B are arbitrary amplitudes).

  Accordingly, it is possible to measure the displacement Δx in the X-axis direction and the displacement Δz in the Z direction by solving the equations (1) and (2) for Δx and Δz.

  Since the arithmetic processing itself is already known in Non-Patent Document 2, etc., detailed description is omitted.

  Furthermore, if interference light having a phase difference of 90 ° is generated at the positions of the photodiodes PD11 and PD12 and detected by a new photodiode, it is possible to determine the moving direction of the diffraction grating surface 3 and to perform phase interpolation. For example, when the wavelength of the light source is 650 nm and the pitch of the diffraction grating is 1 μm, if phase interpolation of 1/2000 is possible at θm = 40.5 °, 0.25 nm in the Z-axis direction and 0 in the X-axis direction. A resolution of .16 nm is obtained.

  In addition to the configuration of FIG. 2, FIG. 3 further shows a two-axis interference sensor unit S1 in which interference light having a phase difference of 90 ° is generated at the positions of the photodiodes PD11 and PD12 and an interference signal is measured by another photodiode. FIG. 3A is a simplified configuration diagram showing another example, in which FIG. 3A shows the biaxial interference sensor unit S1 from a viewpoint along the Y axis, and FIG. 3B shows biaxial interference. The sensor unit S1 is shown from a viewpoint in the direction along the Z axis.

  The biaxial interference sensor unit S1 shown in FIG. 3 changes the optical system of the biaxial interference sensor unit S1 shown in FIG. 2 to omit the reflecting mirrors 5 to 6, and further, the biaxial interference sensor unit shown in FIG. Another photodiode is provided by newly arranging a beam splitter 21 for branching a part of the interference light between the quarter wavelength plate 14 and the photodiode PD11 in S1 and providing the half wavelength plate 19 together. An interference signal having a phase difference of 90 ° with respect to the interference signal detected by the photodiode PD11 is measured by the PD 11 ′ (see FIG. 3A), and in the biaxial interference sensor unit S1 shown in FIG. A beam splitter 22 for branching a part of the interference light is newly disposed between the quarter wavelength plate 15 and the photodiode PD12, and the half wavelength plate 20 is additionally provided. The Rukoto by another photodiode PD 12 ', is obtained so as to measure different interference signals 90 ° phase relative to the interference signal detected by the photodiode PD 12.

As already described, the configuration of the other two-axis interference sensor units S2 (XZ sensor) and S3 (YZ sensor) is the same as that of the above-described two-axis interference sensor unit S1, and therefore a detailed description will be given. Omitted.
Also for the biaxial interference sensor units S2 and S3, the moving direction of the diffraction grating plane 3 and phase interpolation can be performed by applying the configuration shown in FIG. 3 as described above.
The biaxial interference sensor unit S2 includes photodiodes PD21 to PD22 corresponding to the photodiodes PD11 to PD12 of the biaxial interference sensor unit S1, and the biaxial interference sensor unit S3 is a photodiode PD11 of the biaxial interference sensor unit S1. Photodiodes PD31 to PD32 corresponding to PD12 are provided.

Hereinafter, the interference signal detected by the photodiode PD21~PD22 biaxial interferometric sensor unit S2 is defined as _I PD 21 _{~I PD 22,} the interference signal detected by the photodiode PD31~PD32 biaxial interferometric sensor unit S3 I _{PD31 to IPD32} .

  FIG. 4 shows an NC machine tool 23 having a chuck 2 which is a kind of rotating body, a numerical control device 24 for driving and controlling the NC machine tool 23, and the NC machine tool 23 under the installation conditions already described. The main body 25 of the 5-degree-of-freedom error measuring apparatus 1 including the two-axis interference sensor units S1, S2, S3 fixed to the column head and the diffraction grating surface 3 on the chuck 2 side as main parts, and 5 freedom 2 is a block diagram showing a simplified connection relationship of a personal computer or the like that functions as the error calculation means 26 of the degree error measuring apparatus 1, and the combination of the main body portion 25 and the error calculation means 26 has five degrees of freedom. This is an error measurement device 1.

  The NC machine tool 23 is, for example, a vertical milling machine. In the NC machine tool 23 such as a surface grinder using a rotating grindstone, the spindle (main shaft) corresponds to the chuck 2, that is, the rotating body. Of course, an end mill, a drill bit, or a rotating grindstone may be considered as a rotating body.

  In general, the NC machine tool 23 performs a strict positioning process for the table feed in each axis X, Y, and Z direction using the process of the position, speed, and current loop by the servo motor of each axis and the numerical controller 24. However, with respect to the rotation of the chuck 2 and the spindle, the strict rotation angle is not controlled, only the speed control in the open loop is performed, and the rotation angle of the chuck 2 and the spindle is When the position coincides with the specified position, one rotation signal is sent from the NC machine tool 23 to the numerical controller 24. The rotation speed of the chuck 2 and the spindle can be set manually using a manual data input (MDI) provided in the numerical controller 24 or set using an NC machining program such as APT.

  In this embodiment, a spindle motor (not shown) of the NC machine tool 23 that rotates the chuck 2 functions as a rotation driving means in the 5-degree-of-freedom error measuring device 1.

  The error calculation means 26 can be configured by a commercially available personal computer, a workstation, a dedicated control device, or the like.

  The main part of the error calculation means 26 includes a microprocessor 27 for calculation processing, a ROM 28 storing a basic control program for the microprocessor 27, a RAM 29 used for temporary storage of calculation data, and various application programs. And the like, and an interface 31 used for connection with other external devices, etc., and an input / output circuit 32 of the microprocessor 27 includes a keyboard 33 and a display 34 that function as a man-machine interface. Is connected.

  In this embodiment, the display 34 functions as data display means of the error calculation means 26, and the RAM 29 and the hard disk 30 function as data storage means of the error calculation means 26.

Interference signals I _{PD11 to} I _PD12 and photodiodes PD21 to PD21 of the biaxial interference sensor unit S2 detected by the photodiodes PD11 to PD12 of the biaxial interference sensor unit S1 disposed in the main body 25 of the 5-degree-of-freedom error measuring device 1. each of the interference signal _I PD 31 _{~I PD 32} that is detected by the photodiode PD31~PD32 of the interference signal _I PD 21 _{~I PD22} biaxial interference sensor unit S3, detected by PD22, the error calculating unit via the input and output circuit 32 The one rotation signal that is read by the 26 microprocessor 27 and sent from the NC machine tool 23 to the numerical control device 24 is also read by the microprocessor 27 of the error calculating means 26 via the input / output circuit 32. ing.

  5 to 8 are flowcharts showing an outline of an error measurement program installed as one of application programs on the hard disk 30 of the error calculation means 26. FIG.

  Next, the overall processing operation of the 5-degree-of-freedom error measurement device 1 will be described in detail with reference to FIGS.

  However, at this time, the NC machine tool 23 and the numerical control device 24 are already in operation, and the chuck 2 is rotationally driven by the spindle motor of the NC machine tool 23, and error measurement as shown in FIGS. It is assumed that the program for the program is read out as an execution target program and expanded from the hard disk 30 of the error calculation means 1 including a personal computer or the like to the RAM 29.

  The microprocessor 27 that has started to execute the error measurement program first reads the current value of the spindle speed set for the chuck 2 from the numerical controller 24 via the input / output circuit 32 (step a1). ).

  Next, the microprocessor 27 selects, from the data setting file (not shown) registered in the hard disk 30 according to the magnitude of the spindle speed, the value of the standby count m required for error measurement according to the spindle speed. Further, the time required for one rotation of the chuck 2 is obtained based on the spindle speed read in the process of step a1, and the value of the measurement division number n input by the user via the keyboard 33 and the time required for one rotation described above are obtained. Based on the above, a basic sampling period Δt is calculated (step a2).

  Further, the spindle rotational speed is determined by the processing speed of the microprocessor 27 so that the value of the standby count m increases as the spindle rotational speed increases, and decreases as the spindle rotational speed decreases. If it is smaller than a certain value, it is registered in the data setting file so that it becomes zero.

Therefore, if the spindle rotational speed is smaller than a certain value and the processing speed of the microprocessor 27 is sufficient in comparison with the rotational speed of the chuck 2, the microprocessor 27 that constitutes the main part of the error calculating means 26 is set in the standby count. The value 0 is selected from the data setting file as the value of m, and the error is measured every time the chuck 2 rotates [1 / n], that is, every basic sampling period Δt.
On the other hand, when the spindle rotational speed is large and the processing speed of the microprocessor 27 is insufficient in comparison with the rotational speed of the chuck 2, the micro that constitutes the main part of the error calculating means 26 is used. The processor 27 selects one of the integer values m = 1, 2, 3,... From the data setting file according to the magnitude of the spindle rotational speed, and the chuck 2 rotates [m + (1 / n)]. Error measurement is performed every time. In this case, the substantial sampling period is [required time for one rotation × m + Δt].

  Therefore, it is possible to cope with an NC machine tool 23 such as a milling machine having a relatively small spindle rotational speed and an NC machine tool 23 such as a surface grinder having a relatively large spindle rotational speed. It is also possible to cope with a high-speed rotating NC machine tool 23 such as a jig grinder.

  When the initialization process related to the standby count m and the basic sampling period Δt is completed in this way, the microprocessor 27 counts the value of the counter i for counting the number of rotations of the chuck 2 and the number of times for measuring the error. Both the values of j are initialized to 0 (step a3 to step a4), and it is determined whether the value 0 is selected as the standby count m or an integer value other than that is selected (step a5).

  When the determination result in step a5 is true and it is clear that the value of the standby count m is 0, every [1 / n] rotation of the chuck 2, in other words, at the basic sampling period Δt. This means that the error measurement is set to be executed, and if the determination result in step a5 is false and it is clear that the value of the standby count m is a value other than 0, the chuck 2 Means that the error is measured every time [m + (1 / n)] rotates, that is, [one rotation required time × m + Δt].

  Accordingly, when the determination result in step a5 becomes true, the microprocessor 27 firstly receives the first one-turn signal from the NC machine tool 23 via the numerical controller 24 and the input / output circuit 32. The system waits (step a22), and when the input of the first rotation signal is confirmed, the processing of SUB (A) in FIG. 8 relating to error measurement is immediately executed (step a23).

The microprocessor 27 that has started the processing of SUB (A) starts with the photodiodes PD11 to PD12 and PD21 to the biaxial interference sensor units S1, S2, and S3 provided in the main body 25 of the 5-degree-of-freedom error measuring device 1. PD 22, the interference signal _I PD 11 has been detected in _{PD31~PD32 ~I PD12, I PD21 ~I PD22} , reads each I _{PD31 ~I PD32} (step b1 in FIG. 8).

Next, the microprocessor 27 that constitutes the main part of the error calculation means 26 _determines the values of the interference signals I _{PD11 to} I _PD12 detected by the photodiodes PD11 to PD12 of the two-axis interference sensor unit S1 and the above-described equation (1). Based on the equation (2), the displacement Δx1 in the X-axis direction detected by the biaxial interference sensor unit S1 and the displacement Δz1 in the Z direction detected by the biaxial interference sensor unit S1 are obtained.
The microprocessor 27, two-axis interferometric sensors interference is detected by the photodiode PD21~PD22 units the S2 signal _I PD 21 _~I value of the aforementioned formula _{PD 22} (1), _{I PD 11} ~I in equation (2) _By replacing the _{PD 12} and executing the same processing as described above, the displacement Δx2 in the X-axis direction detected by the biaxial interference sensor unit S2 and the displacement Δz2 in the Z direction detected by the biaxial interference sensor unit S2 are obtained. Ask.
Furthermore, the microprocessor 27, two-axis interferometric sensors interference is detected by the photodiode PD31~PD32 unit signal S3 _I PD 31 _~I value of the aforementioned formula _{PD 32} (1), _{I PD 11} ~I in equation (2) _By replacing the _{PD 12} and executing the same processing as above, the displacement Δy3 in the Y-axis direction detected by the biaxial interference sensor unit S3 and the displacement Δz3 in the Z direction detected by the biaxial interference sensor unit S3 are obtained. Obtained (step b2).

Next, the microprocessor 27 that constitutes the main part of the error calculating means 26 is from each of the two-axis interference sensor units S1 and S2 arranged along the X-axis that is one of the diameter directions of the chuck 2 that is a rotating body. An average [(Δx1 + Δx2) / 2] of the movement amounts Δx1, Δx2 in the first axis direction of the sensor coordinate system obtained is obtained, and this value is used as a motion error Δx ( Displacement).
Further, the microprocessor 27 calculates the movement amount Δy3 in the first axis direction of the sensor coordinate system obtained from the two-axis interference sensor unit S3 provided along the Y direction in the radial direction orthogonal to the diameter direction, that is, the X axis. It is assumed that the movement error Δy (displacement) occurs in the other axis of the chuck 2 in the radial direction, that is, in the Y-axis direction.
Then, the microprocessor 27 obtains an average [(Δz1 + Δz2 + Δz3) / 3] of the movement amounts Δz1, Δz2, Δz3 in the second axis direction of the sensor coordinate system obtained from the three two-axis interference sensor units S1, S2, S3. This value is defined as a movement error Δz (displacement) that occurs in one axis in the axial direction of the chuck 2, that is, in the Z-axis direction.
Further, the microprocessor 27 calculates the average [(Δz1 + Δz2) of the movement amounts Δz1, Δz2 in the sensor coordinate system second axis direction obtained from the two-axis interference sensor units S1, S2 arranged along the diameter direction, that is, the X axis. ) / 2] and the deviation Δz ′ between the movement amount Δz3 in the second direction of the sensor coordinate system obtained from the biaxial interference sensor unit S3 arranged along the Y axis in the radial direction perpendicular thereto, ie, [Δz3 − (Δz1 + Δz2) / 2] and the radius r (see FIG. 1A) of the outer peripheral surface of the chuck 2 where the diffraction grating surface 3 is formed, one axis in the tilt direction of the chuck 2, that is, X A motion error Δu (angle) of the posture change generated around the axis is obtained. Specifically, the function f for obtaining the motion error Δu is tan ⁻¹ (Δz ′ / r).
Then, the microprocessor 27 further calculates the movement amounts Δz1 and Δz2 in the sensor coordinate system second axis direction obtained from the two-axis interference sensor units S1 and S2 arranged along the X axis which is one of the diameter directions. Based on the deviation Δz ”, that is, [Δz1−Δz2] and the diameter 2r (see FIG. 1A) of the outer peripheral surface of the chuck 2 where the diffraction grating surface 3 is formed, the other axis of the chuck 2 in the tilt direction. That is, the motion error Δv (angle) of the posture change that occurs around the Y axis is obtained, and the function g for obtaining the motion error Δv is specifically tan ⁻¹ (Δz ″ / 2r) (step b3 above). ).

  In this way, the processing of SUB (A) is completed, and the movement errors Δx and Δy related to the radial motion generated in the plane of the two orthogonal axes X and Y in the radial direction of the chuck 2 and the axial direction of the chuck 2, that is, the Z axis Motion error Δz related to the axial motion generated along the axis 2 and tilt direction motion errors Δu, Δv related to the angular motion generated around the two axes X and Y in the direction in which the central axis of the chuck 2 tilts are obtained. Based on the current value of the counter j, the microprocessor 27 stores Δx, Δy, Δz,... In the field of address j of the data storage file (see FIG. 9) in the RAM 29 that functions as data storage means of the error calculation means 26. The values of Δu and Δv are written (step a24 in FIG. 6), and the basic sampling period Δt is set in the sampling period measuring timer. Timekeeping is started (step a25), and the system waits for the sampling period measurement timer to finish measuring the set time (step a26).

  Then, the time measured by the sampling period measurement timer is finished, the determination result in step a26 becomes true, and the elapsed time since the previous error measurement processing as shown in SUB (A) is performed is the basic sampling period Δt. , That is, when it is confirmed that the chuck 2 has been incrementally rotated by [1 / n] rotation, the microprocessor 27 increments the value of the counter j by 1 (step a27), and the current value of the counter j It is determined whether or not has reached the measurement division number n (step a28).

  If the current value of the counter j has not reached the measurement division number n and the determination result in step a28 is true, error measurement for the number of times corresponding to one rotation of the chuck 2 has not been completed yet. Therefore, the microprocessor 27 repeatedly executes the processing from step a23 to step a28 in the same manner as described above, and obtains the values of Δx, Δy, Δz, Δu, Δv by the error measurement processing of SUB (A). Based on the updated current value of the counter j, the values of Δx, Δy, Δz, Δu, Δv are written in the field of the address j of the data storage file in FIG. 9, and the basic sampling period Δt elapses again. (Steps a23 to a28).

  Therefore, every time the chuck 2 rotates [1 / n] after the basic sampling period Δt has elapsed, the error measurement process shown in SUB (A) is repeatedly executed, and finally the current value of the counter j When the measurement division number n is reached, the determination result of step a28 becomes false, and error measurement for the number of times corresponding to one rotation of the chuck 2 is completed.

  Therefore, if the value of the measurement division number n is set to 36, the values of Δx, Δy, Δz, Δu, Δv are measured every time the chuck 2 rotates 10 °, and these data are Is written in the field of the address j of the data storage file in FIG. 9 based on the current value of the counter j. For example, if the absolute rotation angle of the chuck 2 when one rotation signal is output is defined as 0 °, the time when the rotation signal is output, that is, the rotation angle of the chuck 2 is 0 ° (see FIG. 10A). Δx, Δy, Δz, Δu, Δv are written in the address 0 field of the data storage file of FIG. 9, and the absolute rotation angle of the chuck 2 is 10 ° (see FIG. 10B). The values of Δx, Δy, Δz, Δu, Δv are written in the address 1 field of the data storage file in FIG. 9, and Δx, Δy, Δz, Δu, Δv when the absolute rotation angle of the chuck 2 is 20 °. 9 is written in the address 2 field of the data storage file of FIG. 9, and finally, Δx, when the absolute rotation angle of the chuck 2 is 350 ° (see FIG. 10C). The values of Δy, Δz, Δu, and Δv are the same as the data storage format shown in FIG. Is written in the field of the address 35 (= n-1) of the file, and the value of the counter j is incremented by 1 within the same processing cycle to become j = 36 (= n), and the determination result of step a28 becomes false. .

  In this way, when the determination result in step a28 is false and it is confirmed that the error measurement corresponding to one rotation of the chuck 2 has been completed, the microprocessor 27 checks the address 0 of the data storage file in FIG. Read all the data of Δx, Δy, Δz, Δu, Δv stored in the address n−1, and read the address of the data storage file, that is, the absolute rotation angle of the chuck 2 and Δx, Δy, Δz, Δu, Δv. The data is associated with each other and displayed visually on the display 34 functioning as data display means of the error calculation means 26 (step a17 in FIG. 7).

  When displaying data, it is possible to numerically display the data contents as shown in FIG. 9, but this display is updated every time the chuck 2 is rotated. It is difficult to follow, and it is desirable to display a graph using a spline curve or the like. X, Y, Z, U (rotation around the X axis) and V (rotation around the Y axis) are paralleled to display the correspondence between the absolute rotation angle and Δx, Δy, Δz, Δu, Δv. Alternatively, the X, Y, Z, U, and V axes may be shared and the graph display color for each axis may be changed to correspond to the absolute rotation angle and Δx, Δy, Δz, Δu, Δv. Relationships may be displayed. This type of monitor display is already well known.

  Next, the microprocessor 27 constituting the main part of the error calculation means 26 stores n sets of error measurement data Δx, Δy, Δz, Δu, equivalent to one rotation of the chuck 2 on the hard disk 30 functioning as data storage means. The data storage file of FIG. 9 in which Δv is registered is stored (step a18).

  The hard disk 30 is reserved in advance for a storage area for storing the information of the data storage file of FIG. 9 for k sets, and the contents of the data storage file for the latest k sets are always stored. More specifically, every time the display update process in step a17 is completed, the microprocessor stores the data storage files for storage having the same file format as in FIG. 9 in the reserved storage area of the hard disk 30. 27 cyclically increments the value of the file selection index w in the range of 0 to [k−1], and stores the data shown in FIG. 9 in the storage data storage file [w] corresponding to the current value of the file selection index w. By overwriting and saving the file information, the data storage file of FIG. 9 is cyclically overwritten and saved to k storage data storage files [w] of w = 0 to [k−1]. Data storage file [0], storage data storage file [1], storage data storage file [2], ..., storage data storage file [k-1] The contents of the data storage file of the k-set amount is adapted to save.

  Next, the microprocessor 27 constituting the main part of the error calculating means 26 determines whether or not a measurement end command is input from the keyboard 33 by the user's operation (step a19).

  If no measurement end command is input and the determination result in step a19 is false, the microprocessor 27 returns to the process in step a4 and initializes the value of the counter j to 0. Similarly, the processing of step a5, step a22 to step a28, and step a17 to step a19 is repeatedly executed.

Finally, when the determination result in step a19 becomes true and it is confirmed that the measurement end command is input from the keyboard 33 by the user's operation, the microprocessor 27 serves as the data storage means of the error calculation means 26. Read all k set storage data storage files from the functioning hard disk 30 and average the data for each absolute rotation angle of the chuck 2, for example, every 10 ° for each Δx, Δy, Δz, Δu, Δv, Related to the average value of Δx and Δy, which are motion errors related to radial motion that occurs in the plane of two orthogonal axes X and Y in the radial direction of the chuck 2, and axial motion that occurs along the axial direction of the chuck 2, that is, along the Z axis. Angular motion generated around the two axes X and Y in the direction in which the central axis of the chuck 2 is inclined. The average values of Δu and Δv, which are motion errors in the tilt direction related to the angle, are obtained (step a20), and these values are used as absolute rotation angles of the chuck 2 with reference to the rotation angle at the time of outputting one rotation signal, for example, , 0 °, 10 °, 20 °,..., 350 °, and is visually displayed on the display 34 functioning as data display means of the error calculation means 26 (step a21).
As for the display form, it is possible to display a graph using a spline curve or the like as described above, but in this case, since it is a final result display in particular, as shown in FIG. It is desirable to display the average of the data contents in a format.

  On the other hand, the result of the determination process in step a5, which is performed when the initial setting process related to the standby count m and the basic sampling period Δt is completed, is false, and the standby count m is set to a value other than 0. If it becomes clear that a numerical value has been selected, this means that it is necessary to perform error measurement every time the chuck 2 rotates [m + (1 / n)]. Then, it waits for one rotation signal to be input from the NC machine tool 23 via the numerical controller 24 and the input / output circuit 32 (step a6), and increments the value of the counter i by 1 each time one rotation signal is input. (Step a7) Wait until the current value of the counter i reaches the standby count m (Step a8). During this time, error measurement processing is not performed.

  When it is confirmed in step a8 that the current value of the counter i has reached the standby count m, the microprocessor 27 initializes the value of the counter i for counting the number of rotations of the chuck 2 to 0 again. (Step a9) Based on the current value of the counter j that counts the number of error measurements, a value obtained by multiplying the basic sampling period Δt by j is obtained (Step a10), and this value is set in the sampling period measurement timer. After starting the time measurement (step a11), the process waits until the sampling period measurement timer finishes measuring the set time (step a12).

When the time measured by the sampling period measuring timer ends and the determination result in step a12 becomes true, the microprocessor 27 executes the processing of SUB (A) in the same manner as described above, and Δx, Δy, Δz, Δu, Δv. (Step a13), and based on the current value of the counter j for counting the number of error measurements, the values of Δx, Δy, Δz, Δu, Δv are stored in the address j field of the data storage file in FIG. Write (step a14).
Since the initial value of the counter j is 0 immediately after the error measurement program is started (see step a4), in the process of step a10 that is first executed after the error measurement program is started, [Δt × 0 That is, a value of 0 is set in the sampling period measurement timer, and the determination process in step a12 is immediately true. Therefore, the SUB (A) process is actually executed simultaneously with the input of one rotation signal. Based on the current value 0 of j, in the address 0 field of the data storage file in FIG. 9, Δx, Δy, Δz, Δu, corresponding to the rotation angle of the chuck 2 at the time of inputting one rotation signal, that is, the absolute rotation angle 0 °. The value of Δv will be written.

  Next, the microprocessor 27 increments the value of the counter j by 1 (step a15), and determines whether or not the current value of the counter j has reached the measurement division number n (step a16).

  If the current value of the counter j has not reached the measurement division number n and the determination result at step a16 is true, error measurement for the number of times corresponding to one rotation of the chuck 2 has not been completed yet. Means that.

  Therefore, in this case, the microprocessor 27 returns to the process of step a6, and then repeatedly executes the processes of step a6 to step a8 in the same manner as described above, and inputs one rotation signal a new number of times corresponding to the standby count m. (Steps a6 to a8).

  Then, the determination result of step a8 is false and the current value of the counter i has reached the standby count m, that is, the chuck 2 after the previous error measurement processing as shown in SUB (A) is performed. Is confirmed to have reached m times, the microprocessor 27 constituting the main part of the error calculating means 26 again initializes the value of the counter i to 0 (step a9), and the current value of the counter j Based on the value, a value obtained by multiplying the basic sampling period Δt by j is obtained (step a10), and this value is set in the sampling period measurement timer to start timing (step a11). (Step a12).

When the time measured by the sampling period measurement timer is completed and the determination result in step a12 becomes true, the microprocessor 27 constituting the main part of the error calculation means 26 performs the error measurement process of SUB (A) as described above. The values of Δx, Δy, Δz, Δu, and Δv are obtained by executing (step a13), and based on the current value of the counter j, the fields of the address j of the data storage file in FIG. 9 are Δx, Δy, Δz, Δu, The value of Δv is written (step a14).
At this stage, the value of the counter j is updated from the initial value 0 to 1 (see step a15), so in the process of step a10, [Δt × 1], that is, the value of Δt is set in the sampling period measurement timer. The Therefore, the determination result in step a12 is that the chuck is measured when Δt has passed since the m-th rotation signal is confirmed, that is, after the previous error measurement process shown in SUB (A). It becomes true when the rotational speed of 2 reaches [m + (1 / n)]. Then, the processing of SUB (A) is executed in the same manner as described above, and based on the current value 1 of the counter j, the rotation speed of the chuck 2 is set to [m + (1 / N)] The values of Δx, Δy, Δz, Δu, Δv corresponding to the rotation angle of the chuck 2 at the time when the rotation is reached, that is, the absolute rotation angle is 10 ° (the rotation amount is m × 360 ° + 10 °). Will be written.

  Next, the microprocessor 27 further increments the value of the counter j by 1 (step a15), and determines whether or not the current value of the counter j has reached the measurement division number n (step a16).

  If the current value of the counter j has not reached the measurement division number n and the determination result at step a16 is true, the error measurement is still performed for the number of times corresponding to one rotation of the chuck 2 at this time. Since it means that the processing has not been completed, the microprocessor 27 constituting the main part of the error calculating means 26 returns to the processing of step a6 and then repeatedly executes the processing of step a6 to step a16 in the same manner as described above. .

While the processing of step a6 to step a16 is repeatedly executed, the value of the counter j is sequentially incremented by 1 unit. Therefore, the standby time of step a12 is also sequentially increased by Δt, and the m-th rotation signal for each time is The standby time from the detection until the error measurement processing shown in SUB (A) is performed increases in order of 0, Δt, 2Δt,..., (N−1) Δt. The rotation amount of the chuck 2 based on the absolute rotation angle 0 ° at the time of detection is also [0 / n] = 0 rotation, [1 / n] rotation (eg 10 °), [2 / n] rotation (eg 20 °) , [3 / n] rotation (for example, 30 °)..., [(N-1) / n] rotation (for example, 350 °).
As a result, every time the chuck 2 rotates [m + (1 / n)], the values of Δx, Δy, Δz, Δu, Δv are measured, and these data are shown on the basis of the current value of the counter j each time. 9 is written in the field of address j of the data storage file. For example, if the absolute rotation angle of the chuck 2 when one rotation signal is output is defined as n = 36 with 0 °, the absolute rotation angle of the chuck 2 is 0 °, for example, the total rotation angle of the chuck 2 is 0 °. The values of Δx, Δy, Δz, Δu, Δv at the time (see FIG. 10A) are written in the address 0 field of the data storage file of FIG. 9, and the absolute rotation angle of the chuck 2 is 10 °, for example, the chuck 2 The values of Δx, Δy, Δz, Δu, Δv when the total rotation angle of [360 × m + 10] ° (see FIG. 10B) is written in the address 1 field of the data storage file of FIG. The values of Δx, Δy, Δz, Δu, and Δv when the absolute rotation angle of 2 is 20 ° and the total rotation angle of the chuck 2 is [360 × 2 · m + 2 · 10] ° are the addresses of the data storage file in FIG. Write in field 2 In rare cases, Δx, Δy, when the absolute rotation angle of the chuck 2 is 340 ° and the total rotation angle of the chuck 2 is [360 × (n−2) · m + (n−2) · 10] °, for example. The values of Δz, Δu, and Δv are written in the field of the address 34 (= n−2) of the data storage file in FIG. 9. Finally, the absolute rotation angle of the chuck 2 is 350 °, for example, the total rotation angle of the chuck 2 9 is [360 × (n−1) · m + (n−1) · 10] ° (see FIG. 10C), the values of Δx, Δy, Δz, Δu, Δv of the data storage file of FIG. It is written in the field of address 35 (= n−1), the value of the counter j is incremented by 1 within the same processing cycle as this, and j = 36 (= n), and the determination result in step a16 becomes false.

The absolute rotation angle of the chuck 2 is 340 ° from the total rotation angle [360 × (n−1) · m + (n−1) · 10] ° when the absolute rotation angle of the chuck 2 is 350 °. As is apparent from the value [360 · m + 10] ° obtained by subtracting [360 × (n−2) · m + (n−2) · 10] °, the total rotation angle of the chuck 2 is actually SUB The cycle in which the error measurement process shown in (A) is performed is the same as the cycle in which the chuck 2 rotates [m + (1 / n)].
Therefore, the error measurement process shown in SUB (A), in particular, the relatively complicated process of step b2 required to obtain Δx1, Δz1, Δx2, Δz2, Δy3, and Δz3, [1 / n ] In the case where a time longer than the required rotation time is required, that is, when the required time for processing required for measurement and data writing is long, or when the rotation speed of the chuck 2 is considerably high, The microprocessor 27 that constitutes the main part of the error calculation means 26 performs calculation processing without difficulty, and ensures motion errors Δx, Δy, Δz, Δu, Δv in five axes at every rotation angle, for example, every 10 °. Can be measured.
Further, the processing of step b1 relating to reading of the interference signal detected by the photodiodes PD11 to PD12, PD21 to PD22, PD31 to PD32 of the biaxial interference sensor units S1, S2, S3 is performed by using the value of the interference signal in the RAM 29. Since this is a simple process that only latches temporarily in the storage area and does not require processing time unlike the measurement process in step b2, there is no fear that this process will cause a delay.

  In this way, when the determination result in step a16 is false and it is confirmed that the error measurement corresponding to one rotation of the chuck 2 is completed, the microprocessor 27 checks the address 0 of the data storage file in FIG. Read all the data of Δx, Δy, Δz, Δu, Δv stored in the address n−1, and read the address of the data storage file, that is, the absolute rotation angle of the chuck 2 and Δx, Δy, Δz, Δu, Δv. Corresponding to the data, it is visually displayed on the display 34 functioning as the data display means of the error calculation means 26 (step a17).

  When displaying the data, it is possible to display the data content as shown in FIG. 9 as it is, but since this display is updated every m rotations of the chuck 2, the change in the value can be visually observed. It is difficult to follow, and it is desirable to display a graph using a spline curve or the like. The X, Y, Z, U, and V axes may be arranged in parallel to display the correspondence relationship between the absolute rotation angle and Δx, Δy, Δz, Δu, Δv, or X, Y, Z, U , V may be used in common, and the display color of the graph for each axis may be changed to display the correspondence between the absolute rotation angle and Δx, Δy, Δz, Δu, Δv.

  Next, the microprocessor 27 constituting the main part of the error calculation means 26 stores n sets of error measurement data Δx, Δy, Δz, Δu, equivalent to one rotation of the chuck 2 on the hard disk 30 functioning as data storage means. The data storage file of FIG. 9 in which Δv is registered is stored (step a18), and it is determined whether or not a measurement end command is input from the keyboard 33 by a user operation (step a19).

  If no measurement end command is input and the determination result in step a19 is false, the microprocessor 27 returns to the process in step a4 and initializes the value of the counter j to 0. Similarly, the processing of step a5 to step a19 is repeatedly executed.

Finally, when the determination result in step a19 becomes true and it is confirmed that the measurement end command is input from the keyboard 33 by the user's operation, the microprocessor 27 serves as the data storage means of the error calculation means 26. Read all k set storage data storage files from the functioning hard disk 30 and average the data for each absolute rotation angle of the chuck 2, for example, every 10 ° for each Δx, Δy, Δz, Δu, Δv, Related to the average value of Δx and Δy, which are motion errors related to radial motion that occurs in the plane of two orthogonal axes X and Y in the radial direction of the chuck 2, and axial motion that occurs along the axial direction of the chuck 2, that is, along the Z axis. Angular motion generated around the two axes X and Y in the direction in which the central axis of the chuck 2 is inclined. The average values of Δu and Δv, which are motion errors in the tilt direction related to the rotation angle, are obtained (step a20), and these values are used as absolute rotation angles of the chuck 2, for example, based on the rotation angle when one rotation signal is output, for example, , 0 °, 10 °, 20 °,..., 350 °, and is visually displayed on the display 34 functioning as data display means of the error calculation means 26 (step a21).
As for the display form, a graph display using a spline curve or the like is possible as described above. However, in this case, since it is a final result display in particular, as shown in FIG. It is desirable to display the average of the data contents in a format.

In the above embodiment, the sensor coordinate system first axis of the two two-axis interference sensor units S1, S2 is set along one axis (X axis) in the radial direction of the chuck 2, and at the same time, another one of the two-axis interference sensors. By arranging the sensor coordinate system first axis of the unit S3 along the other axis (Y axis), two orthogonal axes (X, X, X) in the radial direction of the chuck 2 can be obtained without performing processing such as coordinate conversion. Y) The motion errors Δx, Δy of the radial motion in Y) and the motion errors Δu, Δv of the angular motion in the two axes (U, V) in the tilt direction of the axis of the chuck 2 are obtained to shorten the time required for the calculation process. However, in principle, if the two-axis interference sensor units S1, S2, S3 are fixedly arranged at three different positions in the circumferential direction of the chuck 2 so as not to overlap each other, the mounting position is not limited. Na .
For example, the sensor coordinate system first axis of two two-axis interference sensor units S1 and S2 intersects the X axis at an angle, and the sensor coordinate system first axis of the two-axis interference sensor unit S3 is relative to the Y axis. Even if they intersect at a certain angle, coordinate transformation processing such as multiplying the amount of movement of the biaxial interference sensor units S1, S2, S3 in the sensor coordinate system first axis direction by a determinant is performed. For example, the output of the first axis of the sensor coordinate system of these two-axis interference sensor units S1 to S3 can be converted into displacement in the X-axis direction or displacement in the Y-axis direction.

  Further, although the accuracy is reduced, the two-axis interference sensor unit S1, which is two XZ sensors arranged along the X-axis which is one of the diameter directions of the chuck 2 in the process of step b3 shown in FIG. Instead of obtaining the average [(Δx1 + Δx2) / 2] of movement amounts Δx1, Δx2 in the first axis direction of the sensor coordinate system obtained from each of S2 as the movement error Δx, two two axes arranged along the diameter direction The movement amount Δx1 or Δx2 in the first axis direction of the sensor coordinate system of any one of the interference sensor units S1 and S2 can be immediately set as the motion error Δx in the X axis direction.

  Furthermore, the average of the movement amounts Δz1, Δz2, Δz3 in the sensor coordinate system second axis direction output from the three two-axis interference sensor units S1, S2, S3 in the process of step b3 shown in FIG. Instead of obtaining the related movement error Δz, the average of the movement amounts in the second axis direction of the sensor coordinate system output from any two biaxial interference sensor units, or the output from any one biaxial interference sensor unit The movement amount itself in the second axis direction of the sensor coordinate system can be set as the motion error Δz of the axial motion. However, in such a case, when the posture change (Δu, Δv) in the angular direction occurs in the chuck 2, the value of the motion error Δz of the axial motion becomes inaccurate.

  Further, without obtaining the average [(Δz1 + Δz2) / 2] of the movement amounts Δz1, Δz2 in the second axis direction of the sensor coordinate system obtained from the biaxial interference sensor units S1, S2 in the process of step b3 shown in FIG. , The movement amount Δz1 or Δz2 in the second axis direction of either one of the two two-axis interference sensor units S1 and S2 arranged along the X axis and the two axes arranged along the Y axis Based on the deviation from the movement amount in the second axis direction of the sensor coordinate system output from the interference sensor unit S3, that is, [Δz3-Δz1] or [Δz3-Δz2], the chuck 2 is tilted around the X axis. It is also possible to obtain the posture change Δu, which is a motion error related to the angular motion generated in. However, in such a case, the value of the movement error Δu becomes inaccurate when the posture change Δv in the angular direction occurs in the chuck 2.

In FIG. 4, the error calculation means 26 of the 5-degree-of-freedom error measuring device 1 is constituted by a personal computer or the like independent of the numerical control device 24. However, an error measuring program as shown in FIGS. Other multitask processing in the numerical controller 24, for example, reading one block of the NC program, preprocessing related to creation of execution data, pulse distribution processing for controlling the servo motor of each axis, etc. If executed substantially in parallel, the error calculation means 26 comprising an independent personal computer or the like is unnecessary.
In such a case, the microprocessor itself of the numerical control device 24 functions as the error calculating means 26 of the 5-degree-of-freedom error measuring device 1.

  In any case, the 5-axis direction is used under the same conditions as when the rotating body such as the chuck 2 is actually used, for example, under the same conditions as when an end mill, a rotating grindstone, or a drill bit is mounted and the machining program is executed. Motion errors Δx, Δy, Δz, Δu, Δv can be measured, so that radial motion, reflecting various processing conditions such as differences in cutting styles such as up-cutting and down-cutting, and the amount of cutting at one time, There is a merit that each error of axial motion and angular motion can be measured.

  Further, by using the measured values of the movement errors Δx, Δy, Δz, Δu, Δv obtained in real time to correct and control the cutting amount and the tool path by the numerical control device 24, more precise machining can be realized. It is also possible to do.

Of course, instead of using the spindle motor of the NC machine tool 23 as a rotation driving means for rotating the rotating body such as the chuck 2, the rotating body such as the chuck 2 is rotated by a test apparatus using a dedicated stepping motor or the like. The main body 25 of the 5-degree-of-freedom error measuring device 1, that is, the two-axis interference sensor units S1, S2, and S3, may be attached to the test device to measure the motion errors Δx, Δy, Δz, Δu, and Δv. .
In this case, since the actual machining is not performed, the rotation speed of the rotating body such as the chuck 2 can be set to an arbitrary setting speed. For example, the rotation of the rotating body is completely stopped every [1 / n]. Since the error measurement process can be performed, the measurement accuracy is further improved.

  In any case, it is not necessary to perform error measurement using another reference sample (for example, a cylinder or a sphere), and the interference signals from the non-contact type two-axis interference sensor units S1, S2, and S3 are used as they are. Since the motion errors Δx, Δy, Δz, Δu, Δv can be obtained by performing the arithmetic processing in this way, there is no troublesome post-processing for separating the roundness and sphericity errors of the reference sample. In addition, there is no concern that the measurement error increases due to the intervention of the reference sample.

1 5-degree-of-freedom error measuring device 2 chuck (rotating body)
3 Diffraction grating surface 4 Laser diodes 5 to 8 Reflecting mirrors 9 to 10 Beam splitters 11 to 16 1/4 wavelength plate 17 1/2 wavelength plate 18 Fixed mirror 19 to 20 1/2 wavelength plate 21 to 22 Beam splitter 23 NC work Machine (Rotary drive means)
24 Numerical Control Device 25 Body Unit of 5-degree-of-freedom Error Measuring Device 26 Error Calculation Means (Personal Computer, etc.)
27 Microprocessor 28 ROM
29 RAM (data storage means)
30 Hard disk (data storage means)
31 Interface 32 Input / output circuit 33 Keyboard 34 Display (data display means)
A, A1, A2 Laser light B 0th-order light B1 + 1st-order light B2 -1st-order light r Radius 2r of the outer peripheral surface of the chuck where the diffraction grating surface is formed Diameter of the outer peripheral surface of the chuck where the diffraction grating surface is formed CL Center of rotation of rotating body PD11-PD12 Photodiode PD11'-PD12 'Photodiode S1 2-axis interference sensor unit S2 2-axis interference sensor unit S3 2-axis interference sensor unit

Claims (4)

As a rotating body attached to the bearing, or a bearing or a rotating body constituting a part of the rotating body, a rotational error that occurs in the two axial axes of the rotating body, one axial axis, and two tilting axes is generated. A 5-degree-of-freedom error measuring device for measuring,
On the outer circumferential surface of the rotating body, provided with a diffraction grating surface formed from a plurality of circumferential grooves provided at a certain interval in the direction of the rotation axis of the rotating body,
A biaxial interference sensor unit for detecting a relative displacement of a diffraction grating surface along a sensor coordinate system first axis corresponding to a radial direction of the rotating body and a sensor coordinate system second axis corresponding to an arrangement direction of the circumferential grooves; Fixedly arranged at at least three different positions in the circumferential direction of the rotating body with a gap from the outer peripheral surface to the outer side in the radial direction of the rotating body and not overlapping each other in the radial direction of the rotating body And
The diameter of the rotating body based on the amount of movement obtained from each of the interference signals of the first axis of the sensor coordinate system output from at least two of the two-axis interference sensor units that are not simultaneously positioned along the diameter direction of the rotating body. Determining a motion error occurring in the two axes in the direction, and based on at least one of the movement amounts obtained from each of the interference signals of the second axis of the sensor coordinate system output from at least three of the two-axis interference sensor units. A motion error occurring in one axis in the axial direction of the rotating body is obtained, and at least 2 of each movement amount obtained from each of the interference signals of the sensor coordinate system second axis output from at least three of the two-axis interference sensor units. A 5-degree-of-freedom error measuring device comprising error calculating means for obtaining a motion error occurring in two axes in the tilt direction of the rotating body based on each combination. Two of the three two-axis interference sensor units are arranged along the diameter direction of the rotating body and the other one is arranged on a radius orthogonal to the diameter direction;
The error calculation means includes
The average of the movement amounts obtained from the interference signals of the first axis of the sensor coordinate system output from the two two-axis interference sensor units arranged along the diameter direction is set as one axis in the radial direction of the rotating body. The amount of movement obtained from the interference signal of the first axis of the sensor coordinate system that is output from the one two-axis interference sensor unit arranged on the radius as the motion error to be generated is the other one in the radial direction of the rotating body. A motion error that occurs in the shaft,
The average of each movement amount obtained from each of the interference signals of the sensor coordinate system second axis output from the three two-axis interference sensor units is a motion error generated in one axis in the axial direction of the rotating body,
The average of each moving amount obtained from each of the interference signals of the second axis of the sensor coordinate system output from the two two-axis interference sensor units arranged along the diameter direction and the one two arranged on the radius Based on the deviation from the movement amount obtained from the interference signal of the sensor coordinate system second axis output from the axis interference sensor unit, the motion error occurring in one axis in the tilt direction of the rotating body is obtained, and along the diameter direction Based on the deviation of each movement amount obtained from each of the interference signals of the second axis of the sensor coordinate system output from the two two-axis interference sensor units arranged in the same manner, The 5-degree-of-freedom error measuring device according to claim 1, wherein the apparatus is configured to obtain a generated motion error.   Rotating drive means for rotating the rotating body, and the error calculating means includes two radial axes of the rotating body, one axial axis and an inclination direction each time the rotating body rotates [1 / n]. Data storage means for obtaining a motion error occurring in the two axes and storing it in correspondence with the absolute rotation angle of the rotating body, and data display means for visually displaying the data read from the data storage means. The 5-degree-of-freedom error measuring device according to claim 1, wherein the 5-degree-of-freedom error measuring device is provided.   Rotation drive means for rotating the rotator is further provided, and the error calculation means is configured such that each time the rotator rotates [m + (1 / n)], two radial axes and one axial axis of the rotator. And a data storage means for obtaining a motion error occurring in the two axes in the tilt direction and storing it in correspondence with an absolute rotation angle of the rotating body, and a data display means for visually displaying the data read from the data storage means The 5-degree-of-freedom error measuring device according to claim 1, wherein the 5-degree-of-freedom error measuring device is provided.

Images