JP2005069775A - Three dimensional-shape measuring method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a three-dimensional shape measuring method capable of measuring three-dimensional shapes of the surface and the back of a measuring object highly accurately in a short time without redoing setting of the measuring object. <P>SOLUTION: Ahead of three-dimensional shape measurement of the surface and the back of the measuring object, a parallel flat plate whose surface and back are reference planes is set to a holding part for rotating the measuring object so that the normal of the reference plane becomes approximately parallel to the rotation axis of the holding part. The parallel flat plate is rotated, and shape measurement data of the surface and the back of the parallel flat plate are acquired respectively by two probes, and a correction value of an inclination error in each measuring coordinate system is calculated from the acquired shape measurement data and design shape data of the parallel flat plate. The parallel flat plate is replaced with a reference ball whose surface and back are reference spherical surfaces, and the reference ball is rotated, and shape measurement data of the surface and the back of the reference ball are acquired respectively by two probes, and a correction value of a perpendicularity error in each measuring coordinate system and a correction value of the relative distance between the two measuring coordinate systems are calculated from the acquired shape measurement data and design shape data of the reference ball. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a three-dimensional shape measuring method for measuring the shape of the front and back surfaces of an object to be measured by scanning a probe on the surface to be measured.

  A three-dimensional shape measuring machine that measures the shape of the front and back surfaces of an object to be measured that is a rotationally symmetric shape such as a lens is known.

  For example, Japanese Patent Laid-Open No. 2002-71344 discloses one such three-dimensional shape measuring machine. As shown in FIG. 23, the three-dimensional shape measuring machine is configured such that the measurement probe 514 scans the measurement surface 511a of the lens 511, which is the measurement object, in the X direction and the Y direction, so that the XY of the measurement probe 514 is obtained. The Z coordinate at the coordinate position is obtained, and the shape of the measured surface 511a is measured. Here, the lens 511, which is the object to be measured, is supported by a measuring jig 512 having a configuration in which three positioning balls 513 protrude, so that the front surface 511a and the back surface 511b of the lens can be seen, and is fixed on a surface plate.

  In this three-dimensional shape measuring machine, first, the lens surface 511a is set up, and the shape of the three positioning spheres 513 and the surface 511a of the lens 511 is measured by the measurement probe 514, and the measurement data of the three positioning spheres 513 is measured. Then, the reference coordinates of the measurement jig 512 are determined, and the deviation of the surface 511a of the lens 511 from the reference coordinates of the measurement jig 512 is obtained. Thereafter, the measurement jig 512 is rotated 180 degrees (inverted) and fixed on the surface plate with the back surface 511b of the lens 511 facing up, and the shapes of the three positioning balls 513 and the back surface 511b of the lens 511 are measured, and the three positionings are performed. The reference coordinates of the measurement jig 512 are determined from the measurement data of the sphere 513, and the deviation of the back surface 511b of the lens 511 with respect to the reference coordinates of the measurement jig 512 is obtained. The relative positional deviation between the front surface 511a and the rear surface 511b of the lens 511 is obtained from the deviation of the front surface 511a and the rear surface 511b of the lens 511 with respect to the reference coordinates of the measuring jig 512.

  Japanese Patent Application Laid-Open No. 2001-324311 discloses another three-dimensional shape measuring machine. As shown in FIG. 24, this three-dimensional shape measuring machine has two optical probes 523a and 523b facing the front surface 522a and the back surface 522b of the aspherical lens 522 that is a measurement object supported by the θ stage 521. Has been placed.

In this three-dimensional shape measuring machine, the aspherical lens 522 is set so that the axis of the object to be measured substantially coincides with the rotation axis of the θ stage 521. With respect to the aspheric lens 522 rotated by the θ stage 521, the two optical probes 523a and 523b follow the change in the Z direction of the front surface 522a and the back surface 522b of the aspheric lens 522, respectively, while rotating the rotation axis of the θ stage 521. Scan in the X direction orthogonal to. The position coordinates of the optical probes 523a and 523b are measured by the laser length measuring devices 524a and 524b. At this time, the shape of the front surface 522a and the back surface 522b of the aspherical lens 522 is measured from the θ coordinate of the θ stage 521 and the X coordinate and Z coordinate of the optical probes 523a and 523b.
JP 2002-71344 A JP 2001-324111 A

  In the three-dimensional shape measuring machine shown in FIG. 23, not only the shape data of the surface 511a of the lens 511 but also the surface shape data of the three positioning spheres 513 are required when measuring the surface shape of the lens 511 that is the object to be measured. It is. Therefore, each time the object to be measured is exchanged or the object to be measured is attached to or detached from the measuring jig 512, it is necessary to measure the surface of the object to be measured and the three positioning balls 513. In addition, in order to measure the shape of the back surface 511b of the lens 511, it is necessary to rotate the measurement jig 512 by 180 degrees (reversing the front and back) and set the measurement jig 512 again. This complicates the process and increases the measurement time. Further, since the probe 514 is configured to scan the surface to be measured in the X direction and the Y direction, for example, when measuring the shape of the entire surface to be measured having an axisymmetric shape such as an aspherical surface, Since the change in the Z direction becomes larger than the change in the X direction or the Y direction of the probe 514 (the number of times the probe goes up and down increases), the moving speed of the probe cannot be increased and the measurement time becomes longer. .

  On the other hand, the three-dimensional shape measuring machine shown in FIG. 24 can measure the shape of the front surface 522a and the back surface 522b of the aspheric lens 522 without resetting the setting of the aspheric lens 522 as the object to be measured. In addition, since the object to be measured is rotated with the axis of the aspheric lens 522 as the object to be measured substantially aligned with the axis of rotation of the θ stage 521, the shape of the object to be measured on the same radius becomes substantially constant. Since the position control of 523a and 523b is facilitated and the scanning speed can be increased, the measurement time can be shortened as compared with the scanning in the X direction and the Y direction.

However, in the three-dimensional shape measuring machine shown in FIG. 24, for measuring a first measurement coordinate system composed of X ₁ axis and Z ₁ axis to measure the XZ coordinates of the optical probe 523a, the XZ coordinates of the optical probe 523b The second measurement coordinate system composed of the X ₂ axis and the Z ₂ axis needs to coincide with the reference coordinate system with the rotation axis of the θ stage 521 as a reference with high accuracy. If there is an error in each measurement coordinate system, the three-dimensional shape including the relative positional relationship between the front and back surfaces of the object to be measured cannot be measured accurately.

25 and 26 show errors in each measurement coordinate system. As shown in FIG. 25, the reference coordinate system, the rotation axis direction of the θ stage 521 and Z ₀ axis, it is defined from the X ₀ axis as θ-axis is the rotation direction orthogonal to the Z ₀ axis. The error of the first measurement coordinate system, the X ₁ axis in the first measurement coordinate system with respect to X ₀ axis of the reference coordinate system and the tilt error of the first measurement coordinate system for Z ₀ axis of the reference coordinate system Z ₁ axis There is an inclination error (this is a squareness error between the X ₁ axis and the Z ₁ axis in the first measurement coordinate system). There is a similar error for the second measurement coordinate system. Further, as shown in FIG. 26, there is a relative error Δθ in the θ direction between the X ₁ axis of the first measurement coordinate system and the X ₂ axis of the second measurement coordinate system.

  Summarizing the above, errors in each measurement coordinate system include (a) X-axis tilt error of each measurement coordinate system with respect to a plane orthogonal to the rotation axis, and (b) direct difference between the X-axis and Z-axis of each measurement coordinate system. There is an angle error, (c) an angle error in the θ direction of the X axis of each measurement coordinate system, and the errors in the measurement coordinate system of (a) to (c) are measurement errors. Particularly, the error (a) and the error (b) are not only accurate measurement of the relative positional relationship between the front surface 522a and the back surface 522b of the aspherical lens 522, but the shape of each surface is also accurate. It is very important because it will be impossible to grasp.

  Further, if there is an error (c), an angle measurement error in the θ direction is caused as a relative positional relationship between the front surface 522a and the back surface 522b. Regarding the error (c), high measurement accuracy is not required depending on the object to be measured, and there are cases where the required measurement accuracy can be satisfied by mechanical or optical adjustment of each measurement coordinate system. However, depending on the accuracy required for the object to be measured, it may not be possible to achieve the required measurement accuracy only by adjusting each measurement coordinate system.

  The present invention has been made paying attention to the problems of these prior arts, and its purpose is to quickly change the three-dimensional shape of the front and back surfaces of the object to be measured without having to reset the object to be measured. It is to provide a three-dimensional shape measuring method capable of measuring with high accuracy.

  The present invention is directed to a method of measuring the three-dimensional shape of the front and back surfaces of the object to be measured with the first probe and the second probe, respectively, and the three-dimensional shape measuring method listed in the following items. Contains.

  1. Prior to the measurement, the three-dimensional shape measuring method of the present invention includes a coordinate system defined by a holding unit that rotates the object to be measured, and a first probe defined by a first probe that measures the surface of the object to be measured. A correction step for correcting an error between the measurement coordinate system and the second measurement coordinate system defined by the second probe for measuring the back surface of the object to be measured; the correction step includes a first correction step and a second correction step; The first correction step uses a first reference object having two reference planes parallel to each other, and is a method of one plane of the first reference object. One of the first reference objects is held by holding the first reference object so that the line is substantially parallel to the rotation axis of the holding part, and rotating the first reference object around the rotation axis. The shape measurement data A1 is measured by measuring the reference plane of the first probe. The other reference plane of the first reference object is measured with the second probe to obtain shape measurement data A2, and the correction coefficient L1 is calculated from the shape measurement data A1 and the design shape data of the first reference object. Then, the correction coefficient L2 is calculated from the shape measurement data A2 and the design shape data of the first reference object, and the second correction step includes two reference spheres that are part of the same sphere. A second reference object having the second reference object, holding the second reference object, rotating the second reference object around the rotation axis, and one reference spherical surface of the second reference object. Is measured with the first probe to obtain the shape measurement data B1, and the other reference spherical surface of the second reference object is measured with the second probe to obtain the shape measurement data B2. The shape measurement data B1 The correction coefficient L1 is added to the new shape Measurement data B1 ′ is obtained, and the correction coefficient L2 is added to the shape measurement data B2 to obtain new shape measurement data B2 ′. The correction coefficient is obtained from the shape measurement data B1 ′ and the design shape data of the second reference object. M1 is calculated, a correction coefficient M2 is calculated from the shape measurement data B2 ′ and the design shape data of the second reference object, and the shape measurement data B1 ′, the shape measurement data B2 ′, and the second reference object are calculated. The correction value ΔN is calculated from the design shape data (more precisely, the design shape data of the two reference spheres of the second reference object).

  In this three-dimensional shape measurement method, the first measurement coordinate system has an x-axis, a y-axis, and a z-axis that are substantially orthogonal to each other, and the z-axis is set substantially parallel to the rotation axis of the holding unit. The probe is moved along the x-axis of the first measurement coordinate system. The second measurement coordinate system has an x-axis, a y-axis, and a z-axis that are substantially orthogonal to each other, the z-axis is set substantially parallel to the rotation axis of the holding portion, and the second probe has the second measurement coordinate It is moved along the x-axis of the system.

  In this three-dimensional shape measuring method, the tilt error of the x-axis of the first measurement coordinate system with respect to the plane orthogonal to the rotation axis of the holding unit can be corrected based on the correction value L1, and based on the correction value L2. The tilt error of the x-axis of the second measurement coordinate system with respect to the plane orthogonal to the rotation axis of the holding unit is calculated based on the correction value M1, and the z-axis of the first measurement coordinate system is directly related to the x-axis of the first measurement coordinate system. Based on the correction value M2, the square error of the z-axis of the second measurement coordinate system with respect to the x-axis of the second measurement coordinate system is determined based on the correction value ΔN. The relative distance error along the z-axis of the coordinate system can be corrected.

  2. Another three-dimensional shape measuring method of the present invention is the three-dimensional shape measuring method according to the first item, wherein the correction step further includes a third correction step, and the third correction step is the first reference. The first reference object is held such that a normal line of one reference plane of the first reference object is inclined with respect to the rotation axis of the holding unit, and the first reference object is The reference object is rotated around the rotation axis, and one reference plane of the first reference object is measured by the first probe to obtain shape measurement data C1, and the other reference object of the first reference object is obtained. The shape measurement data C2 is obtained by measuring the reference plane with the second probe, and the shape measurement data C1 ′ is obtained by adding the correction coefficient L1 and the correction coefficient M1 to the shape measurement data C1. The correction coefficient L2 and the correction coefficient M2 are added to the data C2. New shape measurement data C2 ′ is obtained, and an angle error correction value Δθ is calculated from the shape measurement data C1 ′, the shape measurement data C2 ′, and the design shape data of the first reference object. And

  In this three-dimensional shape measurement method, the angle error in the θ direction of the x axis of the second measurement coordinate system relative to the x axis of the first measurement coordinate system can be corrected based on the angle error correction value Δθ.

  3. Another three-dimensional shape measurement method of the present invention includes a coordinate system defined by a holding unit that rotates a measurement object, a first probe that measures the surface of the measurement object, and the measurement object before measurement. A correction step for correcting a relative positional error with the second probe for measuring the back surface of the object, and this correction step uses a reference object having a reference surface on the front and back surfaces, A reference object is held, the reference object is rotated around the rotation axis of the holding part, and one reference surface of the reference object is measured with a first probe to obtain shape measurement data D1, The other reference surface is measured with a second probe to obtain shape measurement data D2, and a correction value ΔX1 is calculated from the shape measurement data D1 and the design shape data of the reference object, and the shape measurement data D2 and the reference object are calculated. Correction from design shape data And performing processing for calculating the .DELTA.X2.

  In this three-dimensional shape measuring method, it is possible to correct the origin position error of the first probe with respect to the rotation axis of the rotation means and the origin position error of the second probe with respect to the rotation axis of the rotation means.

  4). Another three-dimensional shape measuring method of the present invention measures a coordinate system defined by a holding unit for rotating the object to be measured, a first probe for measuring the surface of the object to be measured, and the back surface of the object to be measured. A correction step for correcting a relative position error with respect to the second probe. The correction step holds the object to be measured, and rotates the object to be measured around the rotation axis of the holding unit. Then, of one surface of the object to be measured, a range on one side with respect to the rotation axis is measured with a first probe to obtain shape measurement data E1, and a range on the other side with respect to the rotation axis Is measured with the first probe to obtain the shape measurement data E1 ′, and the shape of the other surface of the object to be measured is measured with the second probe on one side of the rotation axis as a reference. Data E2 is obtained and the other side of the rotation axis is used as a reference. The shape is measured with a second probe to obtain shape measurement data E2 ′, a correction value ΔX1 ′ is calculated from the shape measurement data E1, the shape measurement data E1 ′, and the design shape data of the object to be measured, and the shape A process of calculating a correction value ΔX2 ′ from the measurement data E2, the shape measurement data E2 ′, and the design shape data of the object to be measured is performed.

  In this three-dimensional shape measuring method, it is possible to correct the origin position error of the first probe with respect to the rotation axis of the rotation means and the origin position error of the second probe with respect to the rotation axis of the rotation means without using a reference object. .

  ADVANTAGE OF THE INVENTION According to this invention, the three-dimensional shape measuring method which can measure the three-dimensional shape of the surface of a to-be-measured object and a back surface with high precision in a short time, without having to reset the to-be-measured object is provided.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[First embodiment]
As described above, depending on the object to be measured, the measurement accuracy required for the angle between the front and back surfaces of the object to be measured is not so high, and the measurement accuracy required by mechanical adjustment of the measurement coordinate system is achieved. There are cases where it is possible.

  The present embodiment is directed to three-dimensional shape measurement effective in such a case. In other words, this embodiment takes into account (a) the tilt error of the X axis of each measurement coordinate system with respect to a plane orthogonal to the rotation axis, and (b) the squareness error of the X axis and Z axis of each measurement coordinate system. It is directed to 3D shape measurement.

  FIG. 1 is a partial cross-sectional view schematically showing the configuration of the main part of the three-dimensional shape measuring machine according to the first embodiment of the present invention.

  As shown in FIG. 1, the three-dimensional shape measuring machine according to this embodiment includes a holding unit 102 that holds an object 104 to be measured, an air spindle 101 that is a rotating unit that rotates the holding unit 102, and an object 104 to be measured. There are provided a probe 103a and a probe 103b that are arranged to face each other. The measured object 104 has a known design shape of the front surface 104 a and the back surface 104 b, and is fixed to the air spindle 101 via the holding unit 102. The air spindle 101 rotates with high accuracy around the rotation shaft 101A and outputs the rotation angle (θ). The probes 103a and 103b are arranged so as to face the measurement surface 104a and the measurement surface 104b of the measurement object 104, respectively.

  The position of the probe 103a can be controlled in a three-dimensional direction by an unillustrated x stage, y stage, and z stage. The z stage can also perform contact pressure control that keeps the contact pressure between the probe 103a and the measured surface 104a constant. Similarly, the position of the probe 103b can be controlled in a three-dimensional direction by an unillustrated x stage, y stage, and z stage. The z stage can also perform contact pressure control that keeps the contact pressure between the probe 103b and the measured surface 104b constant.

  Here, a first measurement coordinate system having an x-axis, a y-axis, and a z-axis that are substantially orthogonal to each other is set with respect to the probe 103a, and the z-axis is substantially parallel to the rotation axis 101A of the air spindle 101. Similarly, for the probe 103b, a second measurement coordinate system having an x axis, a y axis, and a z axis substantially orthogonal to each other and the z axis being substantially parallel to the rotation axis 101A of the air spindle 101 is set. The three-dimensional shape measuring machine measures three-dimensional coordinates (x, y, z) in the first measurement coordinate system of the probe 103a and three-dimensional coordinates (x, y, z) in the second measurement coordinate system of the probe 103b. Can do.

  Hereinafter, the three-dimensional shape measurement method of this embodiment will be described.

  In this embodiment, the shape of the measurement surface is measured by moving the probe along the x axis. By using a high-precision stage for each stage that moves the probe, the probe does not move along the y-axis direction, or can be kept to a minute amount even if it moves. The positions of the probe 103a and the probe 103b in the y-axis direction are assumed to be positioned by the y stage so as to coincide with the rotation axis 101A of the air spindle 101 (y = 0). Therefore, it is assumed that the y-coordinate error is basically negligible.

(1) Correction of X-axis tilt error of each measurement coordinate system with respect to a plane orthogonal to the rotation axis FIG. 2 shows the correction of the x-axis tilt error of the measurement coordinate system to the three-dimensional shape measuring machine shown in FIG. It shows a state in which a parallel plane plate, which is a reference object, is attached. FIG. 4 shows a flowchart of a procedure for obtaining inclination error correction values of the two measurement coordinate systems in the three-dimensional shape measurement in the first embodiment of the present invention.

  First, as shown in FIG. 2, the plane parallel plate 105 is fixed to the air spindle 101 via the holding portion 102 as a first reference object (Step 1). The plane parallel plate 105 is a plane with high surface accuracy (hereinafter referred to as “reference planes 105a and 105b”), and the reference plane 105a and the reference plane 105b have good parallelism. That is, the plane parallel plate 105 has two reference planes 105a and 105b parallel to each other. The plane parallel plate 105 is fixed to the air spindle 101 so that the normal line of the reference plane 105a is substantially parallel to the rotation axis 101A. Since the parallel plane plate 105 has good parallelism between the front surface and the back surface, the normal line of the reference plane 105b is also substantially parallel to the rotation axis 101A.

Next, the parallel flat plate 105 is rotated by the air spindle 101, and contact pressure control is performed so that the probe 103a follows the change along the z axis of the rotating reference plane 105a. Further, the probe 103a is moved along the x axis substantially orthogonal to the rotation shaft 101A of the air spindle 101. Meanwhile, the xz coordinate (x, z) of the probe 103a and the rotation angle (θ) of the air spindle 101 are measured, and the shape measurement data (x _c , z _c , θ _c ) (c = 1, 2) of the reference plane 105a. , 3... Are acquired as point sequence data. Similarly, the shape measurement data (x _d , z _d , θ _d ) (d = 1, 2, 3,...) Of the reference plane 105b is acquired by the probe 103b (Step 2).

  In Step 2, the movement of the probe 103a and the probe 103b along the x-axis may be performed intermittently or continuously. That is, when the positions of the probes 103a and 103b in the x-axis direction are fixed and measured at a plurality of arbitrary positions, concentric shape measurement data centering on the rotation shaft 101A of the air spindle 101 is obtained. Further, when the probes 103a and 103b are measured while moving in the x-axis direction, spiral shape measurement data is obtained. The acquired shape measurement data may be such concentric shape measurement data or spiral shape measurement data, or may be shape measurement data obtained by combining these. Moreover, the movement of the probe 103a and the probe 103b may be performed simultaneously or separately.

Next, the shape measurement data (x _c , z _c , θ _c ) and (x _d , z _d , θ _d ) of the reference planes 105a and 105b acquired in Step 2 are respectively measured with the measurement data (X _c , Y _c , Z _c ), (X _d , Y _d , Z _d ) (Step 3).

The measurement data of the reference planes 105a and 105b obtained in Step 3 are coordinate transformed to calculate measurement error data (X _c , Y _c , ΔZ _c ) and (X _d , Y _d , ΔZ _d ) of the reference planes 105a and 105b. (Step 4).

  The coordinate transformation of Step 4 is, for example, as compared with design shape data and measurement data in consideration of the radius of curvature of the tip sphere of the probe in the design equation of the surface to be measured as disclosed in JP 2002-1116019 A, Using known methods such as the least square method and the Newton method, the measurement data is such that the error between the design shape data of the measured surface defined with reference to the rotation axis 101A of the air spindle 101 and the measurement data is minimized. Is a coordinate transformation.

  This coordinate conversion is a conversion in which when the surface to be measured is a plane, it is rotated around the X axis and the Y axis and translated along the Z axis. When the surface to be measured is a spherical surface, the conversion is performed by translation along the X, Y, and Z axes. When the surface to be measured is an aspherical surface, it is a conversion in which the surface to be measured is rotationally moved around the X axis and the Y axis and is translated along the X axis, the Y axis, and the Z axis. In Step 4, since the surface to be measured is a flat surface, the rotation is moved around the X axis and the Y axis and translated along the Z axis.

  The surface to be measured is a reference plane with good surface accuracy, and the normal lines of the reference plane 105a and the reference plane 105b are arranged so as to be substantially parallel to the rotation axis 101A. The z coordinates of 103a and probe 103b hardly change. For this reason, there is almost no measurement error due to the squareness error between the x-axis and the z-axis of each measurement coordinate system and the angle error in the θ direction of the x-axis of each measurement coordinate system. Accordingly, the measurement error included in the measurement error data obtained in Step 4 is a measurement error due to an x-axis tilt error of each measurement coordinate system with respect to a plane orthogonal to the rotation axis.

The measurement error data obtained by using the reference plane when there is an x-axis tilt error of the measurement coordinate system with respect to the plane orthogonal to the rotation axis 101A of the air spindle 101 is, for example, a graph as shown in FIG. . FIG. 7 is a graph showing measurement error data (X _c , Y _c , ΔZ _c ) obtained from concentric shape measurement data with X _c on the horizontal axis and ΔZ _c which is a measurement error on the vertical axis. Yes, a measurement error ΔZ _c proportional to X _c occurs.

Next, the mean square value of the measurement error ΔZ _c of the measurement error data (X _c , Y _c , ΔZ _c ) of the reference plane 105a calculated in Step 4 is used as an evaluation value, and the mean square value of the measurement error ΔZ _c is designed. If it is larger than the allowable value, it is determined that there is an inclination error of the x-axis of the first measurement coordinate system, and the process proceeds to Step 6. Similarly, the measurement error ΔZ _d of the measurement error data (X _d , Y _d , ΔZ _d ) on the reference plane 105b is also determined (Step 5).

When the mean square value of the measurement error ΔZ _c is larger than the allowable value, new shape measurement data (for the reference plane 105a) is added to the shape measurement data for the reference plane 105a using the correction value z = Cx for correcting the tilt error as a parameter. x _c , z _c + Cx _c , θ _c ) and return to Step 3 again. Further, when the square mean value of the measurement error [Delta] Z _d is greater than the allowable value, in addition to the shape measurement data of the reference plane 105b of the correction value z = Dx for correcting the tilt error as a parameter, a new form measuring data of the reference plane 105b (X _d , z _d + Dx _d , θ _d ) and return to Step 3 again (Step 6). Here, C is a correction coefficient for correcting the tilt error of the x-axis of the first measurement coordinate system with respect to the plane orthogonal to the rotation axis 101A of the air spindle 101, and D is the first correction coefficient relative to the plane orthogonal to the rotation axis 101A of the air spindle 101. This is a correction coefficient for correcting the tilt error of the x-axis of the two measurement coordinate system.

The loop of Step 3 to Step 6 is repeated until the respective mean square values of the measurement errors ΔZ _c and ΔZ _d become smaller than the allowable value. When the mean square value of the measurement errors ΔZ _c and ΔZ _d becomes smaller than the allowable value, the process proceeds from Step 5 to Step 7.

In addition, as an allowable value which is a judgment criterion in Step 5, an allowable tolerance (P-V value, RMS value, etc.) with respect to the shape of the surface to be measured, or an error of each measurement coordinate system in the three-dimensional shape measuring machine (described above ( a) to (c) error), etc., and a mean square value of measurement error ΔZ _c or ΔZ _d generated by a noise component included in measurement error data acquired in a state where there is no measurement error (ie, a three-dimensional shape measuring machine) ) And taking into account design tolerances such as Alternatively, whether or not the mean square value of the measurement error ΔZ _c or ΔZ _d is the minimum value may be used as a tolerance. In this case, the correction coefficient C or D for correcting the inclination is sequentially changed by a predetermined amount, and whether or not the mean square value of the measurement errors ΔZ _c and ΔZ _d before and after the change of the inclination correction value is the minimum value is used as a criterion. When the tilt correction value is calculated, an optimal tilt error correction value can be obtained even when the noise component of the three-dimensional shape measuring machine is large.

The correction coefficients C and D of the inclination errors of the probes 103a and 103b in which the mean square values of the measurement errors ΔZ _c and ΔZ _d obtained in the loop of Step 3 to Step 6 are smaller than the allowable values are stored (Step 7).

As described above, the shape measurement data of the measurement surface measured by the probe _{103a (x c, z c,} θ c) correction value z coordinates (z = Cx) adding to the shape measurement data (x _c, z _c + Cx _c , θ _c ), it is possible to correct an inclination error of the x axis of the first measurement coordinate system with respect to a plane orthogonal to the rotation axis 101A of the air spindle 101. For the shape measurement data measured by the probe 103b in the same manner, the correction value of the z-coordinate (z = Dx) adding to the shape measurement data _{(x d, z d + Dx} d, θ d) by a, an air spindle The tilt error of the x-axis of the second measurement coordinate system with respect to the plane orthogonal to the rotation axis 101A of 101 can be corrected.

  In the present embodiment, the shape is measured by moving the probe 103a and the probe 103b along the x-axis, so the probes 103a and 103b do not move or move along the y-axis orthogonal to the x-axis and the z-axis. Even a minute amount. Therefore, the y coordinate of the probe 103a and the probe 103b is described as being negligible (y = 0). However, when the probe moves along the y-axis and affects the shape measurement data, it is preferable to measure because the y-coordinate of the probe cannot be ignored.

For example, when the y coordinate of the probe 103a cannot be ignored, in Step 2, the xyz coordinates (x, y, z) of the probe 103a and the rotation angle (θ) of the air spindle 101 are measured, and the shape measurement data of the reference plane 105a is measured. (X _c , y _c , z _c , θ _c ) (c = 1, 2, 3...) Is acquired as point sequence data. In Step 3, the shape measurement data (x _c , y _c , z _c , θ _c ) is converted into measurement data (X _c , Y _c , Z _c ) in an orthogonal coordinate system. Step 4 and subsequent steps are performed in the same manner with the y coordinate taken into account.

  When the probe moves greatly along the y-axis, the tilt error of the y-axis of the measurement coordinate system with respect to the plane orthogonal to the rotation axis may not be ignored.

For example, when the y-axis tilt error of the first measurement coordinate system cannot be ignored, the y-axis tilt error of the first measurement coordinate system is calculated in the same procedure as the calculation of the x-axis tilt error of the first measurement coordinate system. A correction value z = Ey is obtained. Here, E is a correction coefficient for correcting the tilt error of the y axis of the first measurement coordinate system with respect to the plane orthogonal to the rotation axis 101A of the air spindle 101. Therefore, the correction value of the z coordinate with respect to the x coordinate and the y coordinate is z = Cx + Ey. The correction value (z = Cx + Ey) of the z coordinate is added to the shape measurement data (x _c , y _c , z _c , θ _c ) of the surface to be measured, and the shape measurement data (x _c , y _c , z _c + Cx) _c + Ey _c , θ _c ), the x-axis tilt error and the y-axis tilt error of the first measurement coordinate system with respect to the plane orthogonal to the rotation axis 101A of the air spindle 101 can be corrected.

(2) Correction of perpendicularity error of x-axis and z-axis of each measurement coordinate system FIG. 3 shows correction of perpendicularity error of x-axis and z-axis of measurement coordinate system to the three-dimensional shape measuring machine shown in FIG. It shows a state in which a reference sphere that is a reference object for is attached. FIG. 5 shows a procedure for obtaining the squareness error correction value of each of the two measurement coordinate systems and the correction value of the relative distance in the z direction of the two measurement coordinate systems in the three-dimensional shape measurement in the first embodiment of the present invention. The flowchart is shown.

  Subsequent to Step 7, instead of the plane parallel plate 105, as shown in FIG. 3, a reference sphere 106 is fixed to the air spindle 101 via the holding portion 102 as a second reference object (Step 8). The diameter of the reference sphere 106 is known, and the front surface and the back surface are spherical surfaces with good surface accuracy (hereinafter referred to as “reference spherical surfaces 106a and 106b”). That is, the reference sphere 106 has two reference spheres 106a and 106b, both of which are part of the same sphere.

Similarly to Step 2, the shape measurement data of the reference spherical surface 106a is acquired as point sequence data (x _e , z _e , θ _e ) (e = 1, 2, 3,...) By the probe 103a. Similarly, the shape measurement data (x _f , z _f , θ _f ) (f = 1, 2, 3...) Of the reference spherical surface 106b is acquired by the probe 103b (Step 9).

The correction coefficient C for the x-axis tilt error of the first measurement coordinate system stored in Step 7 is read out and added to the shape measurement data (x _e , z _e , θ _e ) of the reference spherical surface 106a acquired in Step 9 to obtain the shape measurement data. (X _e , z _e + Cx _e , θ _e ). Similarly, the shape measurement data of the reference spherical surface 106b is set as shape measurement data (x _f , z _f + Dx _f , θ _f ) (Step 10).

The shape measurement data (x _e , z _e + Cx _e , θ _e ) of the reference spherical surface 106a acquired in Step 10 is converted into measurement data (X _e , Y _e , Z _e ) in the XYZ orthogonal coordinate system. Similarly, the shape measurement data of the reference spherical surface 106b is converted into measurement data (X _f , Y _f , Z _f ) in the XYZ orthogonal coordinate system (Step 11).

Similar to Step 4, using known methods such as the least square method and the Newton method, the measurement data of the reference sphere 106a is minimized so that the error between the design shape data of the reference sphere that is the surface to be measured and the measurement data is minimized. Is transformed to calculate measurement error data (X _e , Y _e , ΔZ _e ) of the reference spherical surface 106a. Similarly, measurement error data (X _f , Y _f , ΔZ _f ) of the reference spherical surface 106b is calculated (Step 12). The coordinate conversion in Step 12 is a conversion that translates along the X, Y, and Z axes because the surface to be measured is a spherical surface.

  Here, the inclination error of the x-axis of each measurement coordinate system with respect to the plane orthogonal to the rotation axis 101A has already been corrected. Therefore, the measurement error included in the measurement error data obtained in Step 12 is a measurement error due to the squareness error between the x-axis and the z-axis of each measurement coordinate system and the angle error in the θ direction of the x-axis of each measurement coordinate system. However, since the object to be measured is a spherical surface with good surface accuracy, the angle error in the θ direction of the x axis of the measurement coordinate system does not affect. Therefore, the measurement error included in the measurement error data obtained in Step 12 is a measurement error due to the squareness error between the x-axis and the z-axis of each measurement coordinate system.

FIG. 8 shows an example of measurement error data obtained using the reference spherical surface when there is a squareness error of the z axis with respect to the x axis of the measurement coordinate system. FIG. 8 is a graph showing measurement error data (X _e , Y _e , ΔZ _e ) obtained from concentric circular shape measurement data with X _e as the horizontal axis and ΔZ _e as the measurement error as the vertical axis. is there.

Next, the mean square value of the measurement error ΔZ _e of the measurement error data (X _e , Y _e , ΔZ _e ) of the reference spherical surface 106a calculated in Step 12 is used as an evaluation value, and the mean square value of the measurement error ΔZ _e is designed by design. If it is larger than the allowable value, it is determined that there is a squareness error in the first measurement coordinate system, and the process proceeds to Step 14. Similarly, the measurement error ΔZ _f of the measurement error data (X _f , Y _f , ΔZ _f ) of the reference spherical surface 106b is also determined (Step 13).

When the root mean square value of the measurement error ΔZ _e is larger than the allowable value, the correction value x = Gz for correcting the squareness error of the z axis of the first measurement coordinate system with respect to the x axis of the first measurement coordinate system is used as a parameter. in addition to the shape measurement data of the spherical 106a, a new form measuring data of the reference spherical surface _{106a (x e + Gz e,} z e, θ e) and, returns to Step10. When the mean square value of the measurement error ΔZ _f is larger than the allowable value, the correction value x = Hz for correcting the squareness error of the z axis of the second measurement coordinate system with respect to the x axis of the second measurement coordinate system is used as a parameter. In addition to the shape measurement data of the reference spherical surface 106b, new shape measurement data (x _f + Hz _f , z _f , θ _f ) of the reference spherical surface 106b is set, and the process returns to Step 10 again (Step 14). Here, G is a correction coefficient for approximately correcting the squareness error of the z axis of the first measurement coordinate system with respect to the x axis of the first measurement coordinate system, and H is the second measurement with respect to the x axis of the second measurement coordinate system. This is a correction coefficient that approximately corrects the squareness error of the z-axis of the coordinate system.

The loop of Step 10 to Step 14 is repeated until the respective mean square values of the measurement errors ΔZ _e and ΔZ _f become smaller than the allowable value. When the square mean value of the measurement error [Delta] Z _e and [Delta] Z _f is smaller than the allowable value proceeds from Step13 to Step 15.

Note that the allowable value, which is the determination criterion in Step 13, is also set in consideration of the design allowable range, as in Step 5. Alternatively mean square value of the measurement error [Delta] Z _e or [Delta] Z _f as an allowable value may be determined based on whether the minimum value.

Correction coefficient G that approximately corrects the squareness error of the z-axis with respect to the x-axis of each measurement coordinate system in which the mean square value of the measurement errors ΔZ _e and ΔZ _f is smaller than the allowable value, obtained in the loop of Step 10 to Step 14 , H are stored (Step 15).

As described above, the shape measurement data of the measurement surface measured by the probe _{103a (x e, z e,} θ e) the correction value of x coordinates (x = Gz) and adds shape measurement data (x _e + Gz _e , Z _e , θ _e ), the squareness error of the z axis of the first measurement coordinate system with respect to the x axis of the first measurement coordinate system can be corrected. Similarly, the shape measurement data measured by the probe 103b is also added to the x coordinate correction value (x = Hz) to obtain shape measurement data (x _f + Hz _f , z _f , θ _f ). The squareness error of the z axis of the second measurement coordinate system with respect to the x axis of the measurement coordinate system can be corrected. Therefore, the z-axis of the first measurement coordinate system and the z-axis of the second measurement coordinate system can be matched with high accuracy on the basis of the rotation axis 101A of the air spindle 101.

  Since the reference spherical surface 106a and the reference spherical surface 106b are a part of the reference spherical surface 106, the centers of curvature of the reference spherical surface 106a and the reference spherical surface 106b coincide with each other. Therefore, the relative distance in the z direction between the probe 103a and the probe 103b can be calculated by the following procedure.

From the measurement error data (X _e , Y _e , ΔZ _e ) of the reference spherical surface 106a calculated in Step 12, the z coordinate (Z _e0 ) of the center of curvature of the reference spherical surface 106a is obtained. Similarly, the z coordinate (Z _f0 ) of the center of curvature of the reference spherical surface 106b is obtained from the measurement error data (X _f , Y _f , ΔZ _f ) of the reference spherical surface 106b (Step 16).

If there is no relative distance error in the z direction between the first measurement coordinate system and the second measurement coordinate system, the z coordinate of the center of curvature of the reference spherical surface 106a and the z coordinate of the center of curvature of the reference spherical surface 106b match, so Z _e0 = It should be Z _f0 . If there is a relative distance error in the z direction, the z coordinate of the center of curvature of the reference sphere 106a and the z coordinate of the center of curvature of the reference sphere 106b do not match and a difference occurs. This difference is a relative distance error in the z-axis direction between the first measurement coordinate system and the second measurement coordinate system. Therefore, the correction value Δz = Z _e0 −Z _f0 of the relative distance error in the z direction between the first measurement coordinate system and the second measurement coordinate system is calculated and stored (Step 17).

The correction value Δz = Z _e0 −Z _f0 of the relative distance error in the z direction calculated in Step 17 is added to, for example, the shape measurement data (x _f , z _f , θ _f ) of the measured surface measured by the probe 103b. By using the measurement data (x _f , z _f + Δz, θ _f ), the relative distance error in the z direction between the first measurement coordinate system and the second measurement coordinate system can be corrected.

  In the present embodiment, the shape is measured by moving the probe 103a and the probe 103b along the x-axis, so the probes 103a and 103b do not move or move along the y-axis orthogonal to the x-axis and the z-axis. Even a minute amount. Therefore, the y coordinate of the probe 103a and the probe 103b is described as being negligible (y = 0). However, when the probe moves along the y-axis and affects the shape measurement data, it is preferable to measure because the y-coordinate of the probe cannot be ignored.

For example, if the y-coordinate of the probe 103a cannot be ignored, in step 9, the xyz coordinates (x, y, z) of the probe 103a and the rotation angle (θ) of the air spindle 101 are measured, and the shape measurement data of the reference spherical surface 106a. (X _e , y _e , z _e , θ _e ) (e = 1, 2, 3...) Is acquired as point sequence data. Then, the shape measurement data with the tilt error corrected is converted into measurement data (X _e , Y _e , Z _e ) in the orthogonal coordinate system in Step 11. Step 12 and the subsequent steps are performed in the same manner, taking the y coordinate into consideration.

  In addition, when the probe moves greatly along the y axis, the squareness error of the z axis with respect to the y axis may not be ignored.

For example, when the squareness error of the z axis with respect to the y axis of the first measurement coordinate system cannot be ignored, the z axis relative to the y axis is calculated in the same procedure as the calculation of the squareness error of the z axis with respect to the x axis of the measurement coordinate system. A correction value y = Iz of the squareness error is obtained. Here, I is a correction coefficient for approximately correcting the squareness error of the z axis of the first measurement coordinate system with respect to the y axis of the first measurement coordinate system. The correction value (x = Gz) of the z-axis squareness error with respect to the x-axis and the correction value (y = Iz) of the z-axis squareness error with respect to the y-axis are measured with the shape measurement data (x _e , y _e , z _e , θ by adding to _e), the shape measurement data _{(x e + Gz e, y} e + Iz e, z e, by the theta _e), the z axis of the first measurement coordinate system with respect to the x axis of the first measurement coordinate system The squareness error and the squareness error of the z axis of the first measurement coordinate system with respect to the y axis of the first measurement coordinate system can be corrected.

(3) Aspherical Lens Shape Measurement FIG. 6 shows a flowchart of a procedure for evaluating the three-dimensional shape of the front and back surfaces of the object to be measured in the three-dimensional shape measurement in the first embodiment of the present invention.

  Subsequent to Step 17, instead of the reference sphere 106, as shown in FIG. 1, as an object to be measured, for example, an aspherical lens 104 whose front surface 104 a and rear surface 104 b are both aspherical surfaces is connected to the air spindle via the holding unit 102. It is fixed to the air spindle 101 so as to be substantially coincident with the rotation shaft 101A of 101 (Step 18).

Next, the aspherical lens 104 is rotated by the air spindle 101, and contact pressure control is performed so that the probe 103a follows the change along the z axis of the rotating aspherical surface 104a. Further, the probe 103a is moved along the x axis. Meanwhile, the xz coordinate (x, z) of the probe 103a and the rotation angle (θ) of the air spindle 101 are measured, and the shape measurement data of the aspherical surface 104a is converted into point sequence data (x _i , z _i , θ _i ) (i = 1, 2, 3 ...). Similarly, the shape measurement data of the aspherical surface 104b is acquired by the probe 103b as point sequence data (x _j , z _j , θ _j ) (j = 1, 2, 3...) (Step 19).

  Here, since the axis of the aspherical lens 104 and the rotation axis 101A of the air spindle 101 are substantially coincident with each other, the shape of the aspherical lens 104 on the same radius is substantially constant, and therefore the contact between the probe 103a and the probe 103b. Since the pressure control is facilitated and the scanning speed can be increased, the measurement time can be shortened compared with the scanning in the x-axis direction and the y-axis direction as in the three-dimensional shape measuring machine shown in FIG.

  In Step 19, as in Step 2, the movement of the probe along the x axis may be performed intermittently or continuously. Moreover, the movement of the probe 103a and the probe 103b may be performed simultaneously or separately.

Correction coefficients C and D for the tilt error of the x-axis of the first measurement coordinate system and the second measurement coordinate system stored in Step 7 and the z-axis with respect to the x-axis of the first measurement coordinate system and the second measurement coordinate system stored in Step 15 The correction coefficients G and H for correcting the squareness error of, and the correction value Δz of the relative distance error in the z direction of the z axis of the first measurement coordinate system and the z axis of the second measurement coordinate system stored in Step 17, The shape measurement data (x _i , z _i , θ _i ) of the aspheric surface 104a acquired in Step 19 and the shape measurement data (x _j , z _j , θ _j ) of the aspheric surface 104b are added to the shape measurement data of the aspheric surface 104a. (X _i + Gz _i , z _i + Cx _i , θ _i ) and shape measurement data (x _j + Hz _j , z _j + Dx _j + Δz, θ _j ) of the aspheric surface 104b are set (Step 20).

Thereby, the shape measurement data (x _i + Gz _i , z _i + Cx _i , θ _i ) of the aspheric surface 104a and the shape measurement data (x _j + Hz _j , z _j + Dx _j + Δz, θ _j ) of the aspheric surface 104b are the same. It can be regarded as shape measurement data by the measurement coordinate system.

The shape measurement data (x _i + Gz _i , z _i + Cx _i , θ _i ) and the shape measurement data (x _j + Hz _j , z _j + Dx _j + Δz, θ _j ) of the aspheric surface 104b acquired in Step 20 are Each is converted into measurement data (X _i , Y _i , Z _i ) and (X _j , Y _j , Z _j ) in the XYZ orthogonal coordinate system (Step 21).

Based on the measurement data (X _i , Y _i , Z _i ) of the aspheric surface 104a calculated in Step 21 and the measurement data (X _j , Y _j , Z _j ) of the aspheric surface 104b, the relative positional relationship between the aspheric surface 104a and the aspheric surface 104b. The three-dimensional shape of the aspherical lens 104 including is understood (Step 22).

  Note that the relative positional relationship between the aspherical axis of the aspherical surface 104a and the aspherical surface of the aspherical surface 104b of the aspherical lens 104 can be obtained, for example, as follows.

Using known methods such as the least square method and the Newton method, the measurement data is coordinate-transformed so that the error between the design shape data of the aspherical surface 104a that is the surface to be measured and the measurement data of the aspherical surface is minimized. Thus, the shape error data (X _i , Y _i , ΔZ _i ) of the aspheric surface 104a is calculated. Similarly, shape error data (X _j , Y _j , ΔZ _j ) of the aspherical surface 104b is calculated.

  This coordinate conversion is a conversion in which the surface to be measured is an aspherical surface and is rotated around the X and Y axes and translated along the X, Y, and Z axes.

From the result of this coordinate conversion, the coordinate conversion amount (A _i , B _i , C _i , α _i of the measurement data of the aspherical surface 104a with respect to the design shape data of the aspherical surface 104a defined with reference to the rotation axis 101A of the air spindle 101 is used. , Β _i ). Here, A _i is the translational movement amount in the X-axis direction of the measurement data of the aspherical surface 104a, B _i is the translational movement amount in the Y-axis direction, C _i is the translational movement amount in the Z-axis direction, and α _i is around the X-axis. The rotational movement amount β _i is the rotational movement amount around the Y axis. Similarly, coordinate conversion amounts (A _j , B _j , C _j , α _j , β _j ) of the measurement data of the aspheric surface 104b are obtained. Here, A _j is the translational movement amounts in the X-axis direction of the measurement data of the aspheric 104b, translation of B _j is the Y-axis direction, C _j is the translational movement amount in the Z axis direction, alpha _j is about the X axis The rotational movement amount, β _j is the rotational movement amount around the Y axis.

The difference between the coordinate transformation amount (A _i , B _i , C _i , α _i , β _i ) of the aspheric surface 104a and the coordinate transformation amount (A _j , B _j , C _j , α _j , β _j ) of the aspheric surface 104b ( _{a i -A j, B i -B} j, C i -C j, α i -α j, from β _i -β _j), the relative aspheric axis of the aspheric surface 104b with respect to the aspherical axis of the aspherical surface 104a You can see the positional relationship.

  According to the present embodiment, it is possible to measure the shape of the aspherical surface 104a and the aspherical surface 104b in a short time without re-setting the aspherical lens 104 that is the object to be measured. In addition, the measurement error due to the measurement coordinate system error (the errors of (a) and (b) described above) is calculated by using the parallel flat plate 105 having a good surface accuracy on the front and back surfaces and a parallelism on the front and back surfaces, and a diameter. Since the shape measurement data of the aspherical surface 104a and the aspherical surface 104b is corrected by calculating using a known and accurate reference sphere 106 and taking this into consideration, the non-uniformity including the relative positional relationship between the front surface and the back surface is included. The three-dimensional shape of the spherical lens 104 can be evaluated with high accuracy.

  In connection with the description of the three-dimensional shape measuring machine in this embodiment, the position control system (x stage, y stage, and z stage) of the probes 103a and 103b and the control system of the three-dimensional shape measuring machine (probes 103a and 103b). The processing apparatus for processing the obtained measurement data is not particularly illustrated, but the same configuration as that of the three-dimensional shape measuring machine of the fourth embodiment described later can be applied to them.

[Second Embodiment]
In the first embodiment, (a) the X-axis tilt error of each measurement coordinate system with respect to a plane orthogonal to the rotation axis and (b) the squareness error between the X-axis and Z-axis of each measurement coordinate system are considered. However, (c) the angle error in the θ direction of the x axis of each measurement coordinate system is not considered. However, depending on the object to be measured, high measurement accuracy may be required for the angle between the front surface and the back surface of the object to be measured in the θ direction.

  The present embodiment is directed to three-dimensional shape measurement that can meet such demands. That is, in this embodiment, in addition to (a) the tilt error of the X axis of each measurement coordinate system with respect to the plane orthogonal to the rotation axis, and (b) the squareness error of the X axis and Z axis of each measurement coordinate system, (C) It is directed to three-dimensional shape measurement in consideration of an angle error in the θ direction of the x axis of each measurement coordinate system.

  The three-dimensional shape measurement method of the present embodiment is the same as that of the first embodiment until halfway, and differs from the measurement method of the first embodiment after Step 17. Below, the measurement procedure after Step17 which is different from 1st embodiment is demonstrated about the three-dimensional shape measuring method of this embodiment.

(4) Correction of the angle error in the θ direction of the x axis of each measurement coordinate system FIG. 9 is a diagram for correcting the angle error in the θ direction of the x axis of the measurement coordinate system shown in FIG. Fig. 5 shows a state in which the plane parallel plate is attached at an angle. FIG. 10 shows a flowchart of a procedure for obtaining the θ-direction angle error correction value in the three-dimensional shape measurement in the second embodiment of the present invention.

  Subsequent to Step 17, instead of the reference sphere 106, as shown in FIG. 9, the parallel flat plate 105 is moved to the air spindle via the holding portion 102 so that the normal line of the reference plane 105a is inclined with respect to the rotation shaft 101A. It fixes to 101 (Step31).

Similarly to Step 2, the shape measurement data of the reference plane 105a is acquired as point sequence data (x _g , z _g , θ _g ) (g = 1, 2, 3...) By the probe 103a. Similarly, the shape measurement data (x _h , z _h , θ _h ) (h = 1, 2, 3...) Of the reference plane 105b are acquired by the probe 103b (Step 32).

Correction for correcting the inclination error of the x-axis of the first measurement coordinate system stored at Step 7 and the squareness error of the z-axis of the first measurement coordinate system with respect to the x-axis of the first measurement coordinate system stored at Step 15 The coefficient G is read and added to the shape measurement data (x _g , z _g , θ _g ) of the reference plane 105a acquired at Step 32, and the shape measurement data (x _g + Gz _g , z _g + Cx _g , θ _g ) of the reference plane 105a is obtained. ). Similarly, the shape measurement data of the reference plane 105b is set as shape measurement data (x _h + Hz _h , z _h + Dx _h , θ _h ) (Step 33).

The shape measurement data (x _g + Gz _g , z _g + Cx _g , θ _g ) acquired in Step 33 is converted into measurement data (X _g , Y _g , Z _g ) in the XYZ orthogonal coordinate system. Similarly, the shape measurement data of the reference plane 105b is converted into measurement data (X _h , Y _h , Z _h ) in the XYZ orthogonal coordinate system (Step 34).

As in Step 4, using a known method such as the least square method or the Newton method, the measurement of the reference plane 105a is performed so that the error between the design shape data and the measurement data of the reference plane 105a that is the surface to be measured is minimized. The data is subjected to coordinate transformation to calculate measurement error data (X _g , Y _g , ΔZ _g ) of the reference plane 105a. Similarly, measurement error data (X _h , Y _h , ΔZ _h ) of the reference plane 105b is calculated (Step 35).

From the result of the coordinate conversion performed in Step 35, the rotational movement amount α _g around the X axis and the rotational movement amount β _g around the Y axis of the measurement data of the reference plane 105a are known. Similarly, the rotational movement amount α _h around the X axis and the rotational movement amount β _h around the Y axis of the measurement data of the reference plane 105b obtained by the probe 103b are known (Step 36).

Here, the inclination error of the x axis of each measurement coordinate system with respect to the plane orthogonal to the rotation axis 101A and the squareness error of the z axis with respect to the x axis of each measurement coordinate system are corrected. Further, since the object to be measured is the parallel plane plate 105, the directions of the normal lines (surface inclinations) of the reference plane 105a and the reference plane 105b coincide with each other with high accuracy. Therefore, if there is no relative angle error in the θ direction between the x axis of the first measurement coordinate system and the x axis of the second measurement coordinate system, the rotational movement amount of the measurement data of the reference plane 105a obtained in Step 36. (Α _g , β _g ) and the rotational movement amount (α _h , β _h ) of the measurement data on the reference plane 105b should match with high accuracy. When there is a relative angle error in the θ direction, the rotational movement amount (α _g , β _g ) of the measurement data on the reference plane 105a and the rotational movement amount (α _h , β _h ) of the measurement data on the reference plane 105b are different values. It becomes.

Next, the difference (α _g −α _h ) between the rotational movement amount (α _g , β _g ) of the measurement data of the reference plane 105a calculated in Step 36 and the rotational movement amount (α _h , β _h ) of the measurement data of the reference plane 105b. ) ² + (β _g -β _h ) ² is used as an evaluation value, and it is determined whether the difference (α _g -α _h ) ² + (β _g -β _h ) ² is smaller than the design tolerance. If it is larger than the allowable value, it is assumed that there is a relative angular error in the θ direction, and the process proceeds to Step 38 (Step 37).

When the difference (α _g −α _h ) ² + (β _g −β _h ) ^{2 of the} rotational movement amount is larger than the allowable value, for example, θ that corrects the angle error in the θ direction of the x axis of the second measurement coordinate system In addition to the shape measurement data of the reference plane 105b using the direction angle error correction value Δθ as a parameter, new shape measurement data (x _h , z _h , θ _h + Δθ) of the reference plane 105b is set, and the process returns to Step 33 again (Step 38).

The loop of Step 33 to Step 38 is repeated until the difference in rotational movement amount (α _g −α _h ) ² + (β _g −β _h ) ² becomes smaller than the allowable value. When the difference (α _g −α _h ) ² + (β _g −β _h ) ^{2 in the} rotational movement amount becomes smaller than the allowable value, the process proceeds from Step 37 to Step 39.

Note that the allowable value, which is the determination criterion in Step 37, is set in consideration of the design allowable range, as in Step 5. Alternatively, as a permissible value, whether or not the difference (α _g −α _h ) ² + (β _g −β _h ) ^{2 in the} rotational movement amount is a minimum value may be used as a criterion.

The angle in the θ direction of the x-axis of the second measurement coordinate system in which the difference (α _g −α _h ) ² + (β _g −β _h ) ² obtained in the loop of Step 33 to Step 38 becomes smaller than the allowable value. The error correction value Δθ is stored (Step 39).

As described above, the shape measurement data of the measurement surface measured by the probe _{103b (x h, z h,} θ h) in the by adding the correction value Δθ of theta direction angular error shape measurement data (x _h, z _h , Θ _h + Δθ), the angle error in the θ direction of the x axis of the second measurement coordinate system with respect to the x axis of the first measurement coordinate system can be corrected. Therefore, the x axis of the first measurement coordinate system and the x axis of the second measurement coordinate system can be matched with high accuracy. The angle error in the θ direction may be corrected for the first measurement coordinate system or for both the first measurement coordinate system and the second measurement coordinate system.

  As described above, by correcting the measurement error caused by the error in the measurement coordinate system (the errors from (a) to (c) described above), the first measurement coordinate is based on the rotation axis 101A of the air spindle 101. The system and the second measurement coordinate system can be matched with high accuracy. Therefore, the measurement data obtained by the probe 103a and the measurement data obtained by the probe 103b can be regarded as measurement data by the same measurement coordinate system.

(5) Shape Measurement of Aspherical Lens FIG. 11 shows a flowchart of a procedure for evaluating the three-dimensional shape of the front and back surfaces of the object to be measured in the three-dimensional shape measurement in the second embodiment of the present invention.

  Subsequent to Step 39, instead of the plane parallel plate 105, as shown in FIG. 1, for example, an aspherical lens 104 having both aspherical surfaces 104 a and 104 b as a measured object is inserted into the air through the holding unit 102. The rotation axis 101A of the spindle 101 is substantially aligned with the air spindle 101 (Step 40).

Next, the aspherical lens 104 is rotated by the air spindle 101, and contact pressure control is performed so that the probe 103a follows the change along the z axis of the rotating aspherical surface 104a. Further, the probe 103a is moved along the x axis. Meanwhile, the xz coordinate (x, z) of the probe 103a and the rotation angle (θ) of the air spindle 101 are measured, and the shape measurement data of the aspherical surface 104a is converted into point sequence data (x _i , z _i , θ _i ) (i = 1, 2, 3 ...). Similarly, the shape measurement data of the aspherical surface 104b is acquired by the probe 103b as point sequence data (x _j , z _j , θ _j ) (j = 1, 2, 3...) (Step 41).

  Here, since the axis of the aspherical lens 104 and the rotation axis 101A of the air spindle 101 are substantially coincident with each other, the shape of the aspherical lens 104 on the same radius is substantially constant, and therefore the contact between the probe 103a and the probe 103b. Since the pressure control is facilitated and the scanning speed can be increased, the measurement time can be shortened compared with the scanning in the x-axis direction and the y-axis direction as in the three-dimensional shape measuring machine shown in FIG.

  Also in Step 41, as in Step 2, the movement of the probe along the x axis may be performed intermittently or continuously. Moreover, the movement of the probe 103a and the probe 103b may be performed simultaneously or separately.

Correction coefficients C and D for the tilt error of the x-axis of the first measurement coordinate system and the second measurement coordinate system stored in Step 7 and the z-axis with respect to the x-axis of the first measurement coordinate system and the second measurement coordinate system stored in Step 15 Correction coefficients G and H for correcting the squareness error in step S17, a correction value Δz of the relative distance error in the z direction between the z axis of the first measurement coordinate system and the z axis of the second measurement coordinate system stored in Step 17, and Step 39 Read the angle error correction value Δθ in the θ direction of the x-axis of the second measurement coordinate system with respect to the stored x-axis of the first measurement coordinate system, and shape measurement data (x _i , z _i , θ of the aspherical surface 104a acquired in Step 41 _i ) and the shape measurement data (x _j , z _j , θ _j ) of the aspheric surface 104b, and the shape measurement data (x _i + Gz _i , z _i + Cx _i , θ _i ) of the aspheric surface 104a and the aspheric surface 104b. shape measurement data (x _j + Hz _{j z j + Dx j + Δz,} θ j + Δθ) to (Step42).

As a result, the shape measurement data (x _i + Gz _i , z _i + Cx _i , θ _i ) of the aspheric surface 104a and the shape measurement data (x _j + Hz _j , z _j + Dx _j + Δz, θ _j + Δθ) of the aspheric surface 104b are It can be regarded as shape measurement data by the same measurement coordinate system.

The shape measurement data (x _i + Gz _i , z _i + Cx _i , θ _i ) and the shape measurement data (x _j + Hz _j , z _j + Dx _j + Δz, θ _j + Δθ) of the aspheric surface 104b acquired in Step 42 are used. These are converted into measurement data (X _i , Y _i , Z _i ) and (X _j , Y _j , Z _j ) in the XYZ orthogonal coordinate system (Step 43).

Based on the measurement data (X _i , Y _i , Z _i ) of the aspheric surface 104a calculated at Step 43 and the measurement data (X _j , Y _j , Z _j ) of the aspheric surface 104b, the relative positional relationship between the aspheric surface 104a and the aspheric surface 104b. The three-dimensional shape of the aspherical lens 104 including is understood (Step 44).

  According to the present embodiment, it is possible to measure the shape of the aspherical surface 104a and the aspherical surface 104b in a short time without re-setting the aspherical lens 104 that is the object to be measured. In addition, the measurement error due to the measurement coordinate system error (the errors of (a) to (c) described above) is calculated by using the parallel flat plate 105 having a good surface accuracy on the front and back surfaces and a parallelism on the front and back surfaces, and a diameter. Since the shape measurement data of the aspherical surface 104a and the aspherical surface 104b is corrected by calculating using a known and accurate reference sphere 106 and taking this into consideration, the non-uniformity including the relative positional relationship between the front surface and the back surface is included. The three-dimensional shape of the spherical lens 104 can be evaluated with high accuracy.

[Third embodiment]
In the three-dimensional shape measuring machine shown in FIG. 24 described above, when the shape of the surface to be measured is measured by scanning the probe with respect to the surface to be rotated, the optical probe 523a and the light with respect to the rotation axis of the θ stage 521 are measured. It is necessary to accurately grasp the origin position of the probe 523b. If there is an origin position error of each probe with respect to the rotation axis, a measurement error occurs, and as a result, the three-dimensional shape including the relative positional relationship between the front surface and the back surface of the object to be measured cannot be measured accurately.

  The present embodiment is directed to three-dimensional shape measurement that is effective when there is an origin position error of the probe with respect to the rotation axis of the rotating means. FIG. 12 is a partial cross-sectional view showing a schematic configuration of the three-dimensional shape measuring machine in the third embodiment of the present invention.

  As shown in FIG. 12, the three-dimensional shape measuring machine of this embodiment is different from the three-dimensional shape measuring machine of the first embodiment in that a contact type probe 103a and a probe 103b are respectively known optical probes 107a. And an optical probe 107b. The optical probe 107a is so-called autofocus controlled so as to irradiate the surface to be measured with the laser beam 108a, detect the reflected light, and keep the distance between the optical probe 107a and the surface to be measured 104a constant. For this optical probe 107 a, a first measurement coordinate system having an x axis, a y axis, and a z axis substantially orthogonal to each other and the z axis being substantially parallel to the rotation axis 101 A of the air spindle 101 is set. The three-dimensional shape measuring machine can measure three-dimensional coordinates (x, y, z) in the first measurement coordinate system of the optical probe 107a. The optical probe 107b has the same configuration. Other configurations are the same as those in the first embodiment, and a description thereof will be omitted.

  Hereinafter, the three-dimensional shape measurement method of this embodiment will be described.

  In the present embodiment, the shape of the measurement surface is measured by moving the optical probe along the x axis. In addition, by using a highly accurate stage for each stage that moves the optical probe, the optical probe does not move along the y-axis direction, or can be suppressed to a minute amount even if it moves. Further, it is assumed that the positions of the optical probe 107a and the optical probe 107b in the y-axis direction are positioned by a y stage (not shown) so as to coincide with the rotation axis 101A of the air spindle 101 (y = 0). Therefore, it is assumed that the y-coordinate error is basically negligible.

  In the following, for simplification, the measurement coordinate system error is corrected using the measurement method of the first embodiment or the second embodiment, and errors other than the origin position error can be ignored.

  FIG. 13 shows a flowchart of the three-dimensional shape measurement in the third embodiment of the present invention.

  First, a reference object having a reference surface whose design shape data on the front and back surfaces is known is fixed to the air spindle 101 via the holding unit 102 (Step 51).

  Here, the reference sphere 106 or the parallel flat plate 105 described in the first embodiment is used as the reference object. In the case of the reference sphere 106, it is supported as shown in FIG. 3, and in the case of the plane parallel plate 105, it is supported as shown in FIG.

Next, the optical probe 107a is caused to follow the change along the z axis of the surface of the reference object while rotating the reference object by the air spindle 101. Further, the optical probe 107a is moved along the x-axis substantially orthogonal to the rotation axis 101A of the air spindle 101. At this time, the xz coordinates (x, z) of the optical probe 107a and the rotation angle (θ) of the air spindle 101 are taken in as point sequence data, and the shape measurement data (x _m , z _m , θ _m ) of the surface of the reference object. (M = 1, 2, 3...) Similarly, the shape measurement data (x _n , z _n , θ _n ) (n = 1, 2, 3...) Of the back surface of the reference object is acquired by the optical probe 107b (Step 52).

  Also in Step 52, as in Step 2 of the first embodiment, the movement of the optical probes 107a and 107b along the x axis may be performed intermittently or continuously. Moreover, the movement of the optical probe 107a and the optical probe 107b may be performed simultaneously or separately.

Next, the shape measurement data (x _m , z _m , θ _m ) and (x _n , z _n , θ _n ) of the front and back surfaces of the reference object acquired in Step 52 are respectively measured in the XYZ orthogonal coordinate system (X _m , Y _m , Z _m ), (X _n , Y _n , Z _n ) are converted (Step 53).

Further, similarly to Step 4 of the first embodiment, using known methods such as the least square method and the Newton method, the measurement data on the front and back surfaces of the reference object obtained in Step 53 are coordinate-transformed to obtain measurement error data ( _{X m, Y m, ΔZ m} ), calculates the _{(X n, Y n, ΔZ} n) (Step54).

In Step 54, coordinate conversion is performed so that the difference in the Z direction with respect to the X coordinate and the Y coordinate is minimized. Therefore, if there is no origin position error, ΔZ _m of the measurement error data on the surface of the reference object and the measurement error data of the back surface ΔZ _n is only a noise component included in the measurement point sequence data, and is small enough to be ignored. However, if there is an origin position error, ΔZ _m and ΔZ _n of the measurement error data become large, resulting in a measurement error.

For example, when the optical probe 107a has an origin position error in the x-axis direction, an example of measurement error data obtained using the reference spherical surface is shown in FIG. 14, and an example of measurement error data obtained using the reference plane is shown in FIG. Show. 14 and 15 are both measured error data obtained from the concentric shape measurement data _{(X m, Y m, ΔZ} m) on the horizontal axis X _m of and a [Delta] Z _m is the measurement error on the vertical axis graph It is displayed. Without the optical probe 107a origin position error of x-axis direction measurement error [Delta] Z _m are eliminated.

Next, the mean square value of the measurement error ΔZ _m of the measurement error data (X _m , Y _m , ΔZ _m ) on the surface of the reference object calculated in Step 54 is used as an evaluation value, and the mean square value of the measurement error ΔZ _m is designed. If it is smaller than the above allowable value, it is determined that there is an origin position error in the x-axis direction of the optical probe 107a, and the process proceeds to Step 56. Similarly, the measurement error ΔZ _n of the measurement error data (X _n , Y _n , ΔZ _n ) on the back surface of the reference object is also determined (Step 55).

If the mean square value of the measurement error ΔZ _m is larger than the allowable value, the offset value Δx _a for correcting the origin position error in the x-axis direction of the optical probe 107a is added as _a parameter to the shape measurement data of the surface of the reference object, and newly Shape measurement data (x _m + Δx _a , z _m , θ _m ) and return to Step 53 again. _If the mean square value of the measurement error ΔZ _n is larger than the allowable value, the offset value Δx _b for correcting the origin position error in the x-axis direction of the optical probe 107b is added as a parameter to the shape measurement data on the back surface of the reference object, New shape measurement data (x _n + Δx _b , z _n , θ _n ) is set, and the process returns to Step 53 again (Step 56).

The loop of Step 53 to Step 56 is repeated until the respective mean square values of the measurement errors ΔZ _m and ΔZ _n become smaller than the allowable value. When the mean square value of the measurement errors ΔZ _m and ΔZ _n becomes smaller than the allowable value, the process proceeds from Step 55 to Step 57.

In addition, as an allowable value which is a judgment criterion of Step 55, there is no measurement error such as an allowable tolerance (PV value, RMS value, etc.) with respect to the shape of the surface to be measured or an origin position error in this three-dimensional shape measuring machine. This is set in consideration of a design tolerance such as a mean square value of measurement error ΔZ _m or ΔZ _n (that is, measurement capability of the three-dimensional shape measuring machine) generated by a noise component included in the measurement error data acquired in step (1). Alternatively, whether or not the mean square value of the measurement error ΔZ _m or ΔZ _n is a minimum value may be used as a criterion. In this case, the offset value Δx _a or Δx _b is sequentially changed by a predetermined amount, and the offset value is determined based on whether or not the mean square value of the measurement errors ΔZ _m and ΔZ _n before and after the offset value change is the minimum value. When calculated, an optimum correction value of the origin position error is obtained even when the noise component of the three-dimensional shape measuring machine is large.

The offset value Δx _a of the optical probe 107a in which the mean square value of the measurement error ΔZ _m obtained in the loop of Step 53 to Step 56 is smaller than the allowable value is stored as the origin position correction value (Δx _a ) of the optical probe 107a. Similarly, the origin position correction value (Δx _b ) of the optical probe 107b is stored (Step 57).

That is, the x-coordinate origin position correction value (Δx _a ) and the optical probe of the optical probe 107a that makes the evaluation value of the measurement error data on the front and back surfaces of the reference object smaller than the design tolerance by the loop of Step 53 to Step 56. The origin position correction value (Δx _b ) of the x coordinate of 107b is obtained.

  Next, in place of the reference object, for example, an aspherical lens 104 having an aspheric surface 104a and a back surface 104b as an object to be measured is substantially aligned with the rotating shaft 101A of the air spindle 101 via the holding unit 102, and air It is fixed to the spindle 101 (Step 58).

Next, while rotating the aspheric lens 104 by the air spindle 101, the optical probe 107a is caused to follow the change along the z-axis of the aspheric surface 104a that is the measurement target surface. Further, the optical probe 107a is moved along the x-axis substantially orthogonal to the rotation axis 101A of the air spindle 101. At this time, the xz coordinates (x, z) of the optical probe 107a and the rotation angle (θ) of the air spindle 101 are taken in as point sequence data, and the shape measurement data (x _p , z _p , θ _p ) ( p = 1, 2, 3... Similarly, the shape measurement data (x _q , z _q , θ _q ) (q = 1, 2, 3...) Of the aspherical surface 104b is acquired by the optical probe 107b (Step 59).

  Here, since the axis of the aspherical lens 104 and the rotation axis 101A of the air spindle 101 are substantially coincident with each other, the shape of the aspherical lens 104 on the same radius is substantially constant, so the optical probe 107a and the optical probe 107b. Since the position control becomes easy and the scanning speed can be increased, the measurement time can be shortened as compared with the scanning in the x-axis direction and the y-axis direction as in the three-dimensional shape measuring machine shown in FIG.

  Also in Step 59, as in Step 2 of the first embodiment, the movement of the probe along the x axis may be performed intermittently or continuously. Moreover, the movement of the optical probe 107a and the optical probe 107b may be performed simultaneously or separately.

The origin position correction values (Δx _a ) and (Δx _b ) of the optical probe 107 a and optical probe 107 b stored in Step 57 are read, and the shape measurement data (x _p , z _p , θ _p ) of the aspheric surface 104 a acquired in Step 59 and The shape measurement data (x _p + Δx _a , z _p , θ _p ) of the aspheric surface 104a and the shape measurement data ( _{a of} the aspheric surface 104b) are added to the shape measurement data (x _q , z _q , θ _q ) of the aspheric surface 104b. x _q + Δx _b , z _q , θ _q ) (Step 60).

The shape measurement data (x _p + Δx _a , z _p , θ _p ) and the shape measurement data (x _q + Δx _b , z _q , θ _q ) of the aspheric surface 104b acquired in Step 60 are respectively in an XYZ orthogonal coordinate system. measurement data _{(X p, Y p, Z} p) and _{(X q, Y q, Z} q) is converted into (Step 61).

Based on the measurement data (X _p , Y _p , Z _p ) of the aspheric surface 104a and the measurement data (X _q , Y _q , Z _q ) of the aspheric surface 104b calculated in Step 61, the relative positional relationship between the aspheric surface 104a and the aspheric surface 104b. The three-dimensional shape of the aspherical lens 104 including is understood (Step 62).

  In this embodiment, since the shape measurement is performed by moving the optical probe 107a and the optical probe 107b along the x axis, the optical probes 107a and 107b do not move along the y axis orthogonal to the x axis and the z axis. Or even if it moves, it is a minute amount. Therefore, the y coordinate of the optical probe 107a and the optical probe 107b has been described as being negligible (y = 0). However, when the probe moves along the y-axis and affects the shape measurement data, it is preferable to measure because the y-coordinate of the probe cannot be ignored.

For example, when the y coordinate of the optical probe 107a cannot be ignored, in step 52, the xyz coordinates (x, y, z) of the optical probe 107a and the rotation angle (θ) of the air spindle 101 are captured as point sequence data, and the reference object Surface shape measurement data (x _m , y _m , z _m , θ _m ) (m = 1, 2, 3,...) Are acquired. Then in STEP 53, and converts the shape measurement data _{(x m, y m, z} m, θ m) measured data of the orthogonal coordinate system _{(X m, Y m, Z} m) to. Step 54 and the subsequent steps are performed in the same manner with the y coordinate taken into account.

  Further, in the configuration in which the probe is moved along the x-axis and the z-axis to measure the shape of the surface to be measured as in this embodiment, the origin position error in the y-axis direction is greater than the origin position error in the x-axis direction. Has little effect on shape measurement data. Therefore, by adjusting the position of the optical probe 107a, the optical probe 107b, and the air spindle 101 in the y-axis direction, it is possible to reduce the influence of the origin position error in the y-axis direction. The error was described as negligible. However, depending on the measurement accuracy required for the three-dimensional shape measuring machine, it is necessary to correct the origin position error in the y-axis direction.

For example, FIG. 16 shows an example of measurement error data obtained using the reference spherical surface when there is an origin position error in the y-axis direction of the optical probe 107a. FIG. 16 is a graph showing measurement error data (X _m , Y _m , ΔZ _m ) obtained from concentric shape measurement data with X _m as the horizontal axis and ΔZ _m as the measurement error as the vertical axis. is there. Without y-axis direction of the home position error in the optical probe 107a measurement error [Delta] Z _m are eliminated. For even y-axis direction of the home position error, the same procedure as the calculation of the origin position error of x-axis direction, obtains the offset value [Delta] y _a for correcting the y-axis direction of the origin position error of the optical probe 107a. Then, adding an offset value [Delta] y _a for correcting the home position error offset value [Delta] x _a and y-axis direction to correct the x-axis direction of the origin position error of the optical probe 107a in the shape measurement data as parameters, the shape measurement of the reference spherical surface 106a Data (x _m + Δx _a , y _m + Δy _a , z _m , θ _m ). In this way, not only the origin position error in the x-axis direction of the probe but also the origin position error in the y-axis direction can be corrected.

  According to the present embodiment, it is possible to measure the shape of the aspherical surface 104a and the aspherical surface 104b in a short time without re-setting the aspherical lens 104 that is the object to be measured. Further, the origin position error in the x-axis direction and the y-axis direction of the optical probe 107a and the optical probe 107b with respect to the rotation axis 101A of the air spindle 101 is determined based on the reference sphere 106 having a known diameter and good surface accuracy on the front and back surfaces or Since the parallel plane plate 105 having good surface accuracy on the back surface is used as a reference object, and the shape measurement data of the aspheric surface 104a and the aspheric surface 104b is corrected by taking this into consideration, the relative positional relationship between the front surface and the back surface 3D shape of the aspherical lens 104 including that can be evaluated with high accuracy.

  In this embodiment, the shape measurement data of the object to be measured is corrected based on the origin position correction values of the optical probe 107a and the optical probe 107b obtained using the reference sphere 106 or the parallel flat plate 105. The measurement may be performed after correcting the initial positions of the optical probe 107a and the optical probe 107b, that is, the origin position of the measurement coordinate system based on the position correction value.

  In this embodiment, the reference object is a reference sphere with a known diameter and good surface accuracy on the front and back surfaces, or a parallel flat plate with good surface accuracy on the front and back surfaces. Is known, an aspheric lens may be used as a reference object.

  In the present embodiment, the non-contact type three-dimensional shape measuring machine using the optical probe has been described as an example. However, the contact type three-dimensional shape measuring machine as in the first embodiment may be used. .

[Fourth embodiment]
The present embodiment is directed to effective three-dimensional shape measurement when there is an origin position error of each probe with respect to the rotation axis of the rotating means. In particular, it is directed to three-dimensional shape measurement that is effective when the position reproducibility of the measurement origin of the measurement coordinate system for measuring the position coordinates of the probe is poor or when the measurement origin fluctuates.

  For example, as in the prior art of FIG. 24, a laser length measuring device may be used as means for measuring the position coordinates of the probe of the three-dimensional shape measuring machine. A laser length measuring instrument can measure the displacement of an object to be measured with high resolution and high accuracy, but it measures the relative displacement from a certain state. The laser length measuring instrument has a measurement origin. do not do. Even if the measurement origin is determined in a certain state, if the measurement environment or the like varies, a so-called drift may occur with time and the measurement origin may change. In such a case, it is necessary to accurately grasp the origin position of the probe at the time of measuring the shape of the surface to be measured.

  FIG. 17 is a partial cross-sectional view showing the configuration of the three-dimensional shape measuring machine in the fourth embodiment of the present invention. FIG. 18 is an enlarged view of the object to be measured shown in FIG. 17 and one of the probes.

  As shown in FIG. 17, the three-dimensional shape measuring machine of the present embodiment is installed on a base 214 arranged horizontally. On the base 214, an air spindle 201, which is a rotating means, is fixed via a frame 203. The air spindle 201 includes a rotating unit 201B and a rotation support unit 201C, and a holding unit 202 that holds an object 204 to be measured is provided inside the rotating unit 201B. The rotating part 201B of the air spindle 201 can rotate with high accuracy about the rotating shaft 201A. Accordingly, the DUT 204 supported by the holding unit 202 can be rotated about the rotation shaft 201A. Although not shown, the air spindle 201 includes a rotary motor and a rotary encoder that outputs a rotation angle θ.

  On the base 214, two x stages 213a and 213b are installed with the mount 203 interposed therebetween. The x stages 213a and 213b are adjusted so as to be movable in a direction substantially coincident with the x-axis direction orthogonal to the paper surface of FIG. 17, and have a position detector composed of a motor (not shown) and a laser length measuring device. doing. On the x stages 213a and 213b, y stages 212a and 212b that are movable in a direction substantially coinciding with the y-axis direction shown in FIG. 17 are respectively mounted. It has the position detector which consists of. Further, two probes 205a and 205b are mounted on the y stages 212a and 212b so as to face the measurement surface 204a that is the surface of the measurement object 204 and the measurement surface 204b that is the back surface, respectively. . Accordingly, the probes 205a and 205b can be individually moved in the x-axis direction and the y-axis direction by the x-stages 213a and 213b and the y-stages 212a and 212b.

  As the probe 205a, for example, a contact probe as described in Japanese Patent Application No. 2001-271500 is used. An outline of the contact probe will be described.

  The probe 205a is provided with an air slide 208a having a shaft 206a with a ruby ball 207a at the tip. The air slide 208a has a shaft 206a that can move with a very small frictional force in the directions of arrows A and B shown in FIGS. 17 and 18, but is highly rigid in the direction orthogonal to the directions of arrows A and B and does not displace. It has become. There is also a position detector 209a comprising a laser length measuring device or the like for detecting the amount of displacement of the shaft 206a in the directions of arrows A and B relative to the air slide 208a. The air slide 208a and the position detector 209a are placed on the support plate 210a, and the support plate 210a is placed on the y stage 212a via the angle adjustment unit 211a. Then, the ruby ball 207a is placed on the surface of the surface to be measured 204a by adjusting the angle adjusting unit 211a to give the shaft 206a a slight inclination angle φ with respect to the z-axis direction as shown in FIG. It is comprised so that it may contact by the dead weight of the axis | shaft 206a. By configuring the probe 205a in such a configuration, the contact pressure of the ruby ball 207a with respect to the surface of the surface to be measured 204a can be made extremely small by giving a minute angle φ, and the contact pressure is reduced by the shaft 206a. There is a feature that even if it moves in the directions of arrows A and B, it does not change.

  The probe 205b is the same contact type probe as that of the probe 205a with the orientation changed and will not be described. However, in FIG. 17, the suffix a is changed to b.

  FIG. 19 is a block diagram showing a control system of the three-dimensional shape measuring machine shown in FIG. As shown in FIG. 19, the control system of the three-dimensional shape measuring machine according to the present embodiment includes a control unit 215 that performs overall control, and a program memory that stores a control program for controlling the three-dimensional shape measuring machine. 216, input means 217 for inputting design formulas, measurement conditions, various data and the like of the surface to be measured, storage means 218 for storing various data such as shape measurement data, and coordinates for performing coordinate conversion of the shape measurement data and the like Conversion means 219, correction value calculation means 220 for calculating a correction value for correcting a measurement error included in the shape measurement data, data correction means 221 for correcting the shape measurement data, and display for displaying various information such as measurement results Means 222.

  In FIG. 19, a control means 215, a program memory 216, a storage means 218, a coordinate conversion means 219, a correction value calculation means 220, and a data correction means 221 constitute a processing device for processing measurement data obtained by the probes 205a and 205b. doing.

  In such a three-dimensional shape measuring machine, the probe 205a includes a first measurement coordinate by a position detector (not shown) of the x stage 213a, a position detector (not shown) of the y stage 212a, and a position detector 209a of the shaft 206a. A system is set up and three-dimensional coordinates (x, y, z) can be measured. Similarly, the probe 205b is also set with the second measurement coordinate system and can measure the three-dimensional coordinates (x, y, z). However, there are errors in the measurement coordinate system (errors (a) to (c) described above) in the three-dimensional coordinates (x, y, z) of the first measurement coordinate system and the second measurement coordinate system. Further, since the position detectors 209a and 209b have an inclination angle φ with respect to the z-axis, a squareness error of the z-axis with respect to the y-axis of each measurement coordinate system is generated and becomes an error of the measurement coordinate system. However, errors in these measurement coordinate systems can be corrected by using the measurement method of the first embodiment or the second embodiment.

  In the following, for simplification, the errors of the first measurement coordinate system and the second measurement coordinate system are corrected using the measurement method of the first embodiment or the second embodiment, and errors other than the origin position error are ignored. It shall be possible. However, each position detector that measures the three-dimensional coordinates (x, y, z) of the probe uses a laser length measuring device, and the origin position of the measurement coordinate system cannot be accurately grasped. In this case, the relative distance in the z-axis direction between the z-axis of the first measurement coordinate system and the z-axis of the second measurement coordinate system cannot be accurately determined. This point will be described later.

  FIG. 20 shows a flowchart of the three-dimensional shape measurement in this embodiment. FIG. 21 shows a measurement range in the three-dimensional shape measurement of the present embodiment. Hereinafter, the three-dimensional shape measurement method of the present embodiment will be described with reference to the flowchart of FIG.

As shown in FIGS. 17 and 21, the measured object 204 having the measured surface 204 a and the measured surface 204 b is substantially aligned with the rotating shaft 201 </ b> A of the air spindle 201 via the holding unit 202, so that the rotating unit 201 </ b> B is aligned. Fix (Step 71). Here, the effective radii (measurement radii) of the measured surface 204a and the measured surface 204b are R _a and R _b , respectively.

The control system input means 217 inputs design formulas, effective radii (measurement radii) R _a and R _b , measurement conditions, and the like of the front surface 204a and the back surface 204b of the object 204 to be measured. The inputted design formula and the like are stored in the storage unit 218 based on the control of the control unit 215.

  Next, based on the control of the control means 215, the x stage 213a and the y stage 212a are driven to move the probe 205a, and the ruby ball 207a is measured near the rotation axis 201A of the air spindle 201 (position 205A in FIG. 21). The surface 204a is brought into contact. Similarly, for the probe 205b, the ruby ball 207b is brought into contact with the surface to be measured 204b in the vicinity of the rotation axis 201A of the air spindle 201 (position 205A ′ in FIG. 21). The xyz coordinate values of the probes 205a and 205b and the rotation angle of the rotary encoder of the air spindle 201 are reset (Step 72).

  In Step 72, the positions of the probe 205a and the probe 205b do not need to be exactly aligned with the rotation axis 201A of the air spindle 201. Good. Thereafter, the positions of the y stage 212a and the y stage 212b are not moved while being held.

  Also in this embodiment, as in the first embodiment, the shape of the measurement surface is measured by moving the probe along the x axis. By making each stage that moves the probe with high accuracy, the probe does not move along the y-axis direction, or even if it moves, the amount is very small. Therefore, in the following description, it is assumed that the y-coordinates of the probes 205a and 205b can be ignored (y = 0).

Next, based on the control of the control means 215, the probe 205a is caused to follow the displacement of the surface to be measured 204a in the z-axis direction while rotating the object to be measured 204 by the air spindle 201. Further, the x stage 213a is driven in _a range of ± Ra (the range from the position 205B to the position 205C in FIG. 21) that is the measurement range of the measured surface 204a, and the probe 205a is moved along the x axis. At this time, the xz coordinates (x, z) of the probe 205a and the rotation angle (θ) of the air spindle 201 are taken in as point sequence data, and the shape measurement data (x _r , z _r , θ _r ) ( r = 1, 2, 3,...) and stored in the storage means 218. Similarly, the shape measurement data (x _s , z _s , θ _s ) in the range of ± R _b (the range from position 205B ′ to position 205C ′ in FIG. 21) that is the measurement range of the measurement target surface 204b by the probe 205b. (S = 1, 2, 3,...) Is stored (Step 73).

  Also in Step 73, similarly to Step 2 of the first embodiment, the movement of the probe 205a and the probe 205b along the x axis may be performed intermittently or continuously. Further, the movement of the probe 205a and the probe 205b may be performed simultaneously or separately.

  In the shape measurement data of the measured surface 204a acquired in Step 73, the shape measurement data in the range where the x coordinate of the probe 205a is negative (position 205A to position 205B in FIG. 21) and the x coordinate of the probe 205a are positive (see FIG. The shape measurement data in the range of 21 positions 205A to 205C) is included. That is, in Step 73, the shape measurement data (x, z, θ) of the measured surface 204a in which the x coordinate of the probe 205a is positive and the shape measurement data of the measured surface 204a in which the x coordinate of the probe 205a is negative ( x, z, θ) is acquired. The same applies to the shape measurement data of the surface 204b to be measured by the probe 205b.

  In other words, in Step 73, the probe 205a measures the range on one side of the measured surface 204a with respect to the rotation axis 201A and the range on the other side with respect to the rotation axis 201A. Shape measurement data (x, z, θ) is acquired. Further, in the surface to be measured 204b, the range on one side with respect to the rotating shaft 201A and the range on the other side with respect to the rotating shaft 201A are measured by the probe 205b and the shape measurement data (x, z, θ) is obtained.

Next, the shape measurement data (x _r , z _r , θ _r ) and the shape measurement data (x _s , z _s , θ _s ) of the measured surface 204a acquired in Step 73 are converted into the coordinate conversion means 219. Are converted into measurement data (X _r , Y _r , Z _r ) and (X _s , Y _s , Z _s ) in the XYZ orthogonal coordinate system (Step 74).

Further, the coordinate conversion means 219 uses a known method such as the least square method or Newton method as in Step 4 of the first embodiment, and the error between the design shape data of the measured surface 204a and the measured data of the measured surface 204a. There was a coordinate transformation of measurement data of the measurement surface 204a so as to minimize the shape error data of the surface to be measured _{204a (X r, Y r,} ΔZ r) stored in the storage unit 218 to calculate a. Similarly, the shape error data (X _s , Y _s , ΔZ _s ) of the measured surface 204b is calculated and stored in the storage means 218 (Step 75).

The shape error data (X _r , Y _r , ΔZ _r ) of the surface to be measured 204 a obtained in Step 75 and the shape error data (X _s , Y _s , ΔZ _s ) of the surface to be measured 204 b are the rotation axis of the air spindle 201. A measurement error due to an origin position error in the x-axis direction of the probe 205a and the probe 205b with respect to 201A is included. FIG. 22 shows an example of shape error data when the probe 205a has an origin position error in the x-axis direction. Figure 22 is one in which the shape error data of the surface to be measured _{204a (X r, Y r,} ΔZ r) on the horizontal axis X _r of the displayed graph by the [Delta] Z _r is a shape error on the vertical axis. If the probe 205a has an origin position error in the x-axis direction, the shape error data 223 in the range where the x coordinate of the probe 205a is negative (position 205A to position 205B in FIG. 21) as in the shape error data in FIG. And the shape error data 224 in the range of plus (position 205A to position 205C in FIG. 21). If there is no origin position error in the probe 205a, the shape error data 223 and the shape error data 224 overlap.

Next, in the correction value calculation means 220, the shape error data 223 in the range where the x coordinate of the probe 205a of the shape error data (X _r , Y _r , ΔZ _r ) of the measured surface 204a calculated in Step 75 is in the minus range and the plus range. The difference of the shape error data 224 at is calculated. Further, when the mean square value of the difference between the shape error data 223 and the shape error data 224 is used as an evaluation value, it is determined whether the mean square value of the difference between the shape error data is smaller than a design allowable value. Since there is an origin position error, the process proceeds to Step 77. Similarly, the shape error data (X _s , Y _s , ΔZ _s ) of the surface 204b to be measured by the probe 205b is also determined (Step 76).

If the mean square value of the difference in the shape error data of the measured surface 204a is larger than the allowable value, the data correction means 221 uses the offset value Δx _a ′ for correcting the origin position error in the x-axis direction of the probe 205a as _a parameter. In addition to the shape measurement data, the shape measurement data (x _r + Δx _a ', z _r , θ _r ) of the measured surface 204a is stored in the storage means 218, and the process returns to Step 74 again. When the mean square value of the difference in the shape error data of the measured surface 204b is larger than the allowable value, the offset value Δx _b ′ for correcting the origin position error in the x-axis direction of the probe 205b is added to the shape measurement data as a parameter. The shape measurement data (x _s + Δx _b ′, z _s , θ _s ) of the measured surface 204b is stored in the storage means 218, and the process returns to Step 74 again (Step 77).

In the loop of Step 74 to Step 77, the mean square value of the difference between the shape error data in the range where the x coordinate of the measured surface 204a is minus and the shape error data in the plus range, and the x coordinate of the measured surface 204b are in the minus range. It repeats until the mean square value of the difference between the shape error data and the shape error data in the plus range is smaller than the design tolerance. If the mean square value of the difference in shape error of the measured surface 204a and the mean square value of the difference in shape error of the measured surface 204b are both smaller than the allowable values, the offset values Δx _a ′ and Δx _b ′ at this time Is stored in the storage means 218, and the process proceeds from Step 76 to Step 78.

In the loop of Step 74 to Step 77, the origin position correction value Δx _a ′ of the x coordinate of the probe 205a that makes the evaluation value of the difference between the two shape error data 223 and 224 of the measured surface 204a smaller than the design allowable value is obtained. The origin position correction value Δx _a ′ is added to the shape measurement data (x, z, θ) of the measured surface 204a to correct the shape measurement data, and the corrected shape measurement data (x + Δx _a ′) of the measured surface 204a is corrected. , Z, θ) are converted into shape measurement data (X, Y, Z) in an XYZ orthogonal coordinate system. The same applies to the shape measurement data of the measurement target surface 204b.

In addition, as an allowable value which is a judgment criterion of Step 76, an allowable tolerance (PV value, RMS value, etc.) with respect to the shape of the measured surface 204a or 204b, or a measurement error such as an origin position error in this three-dimensional shape measuring machine. Taking into account design tolerances such as the mean square value of measurement error ΔZ _r or ΔZ _s (that is, the measurement capability of a three-dimensional shape measuring machine) caused by noise components included in the shape error data acquired in the absence of Set. Alternatively, it may be determined whether the mean square value of the difference between the shape errors ΔZ _r or ΔZ _s is the minimum value as an allowable value. In this case, the offset value Δx _a ′ or Δx _b ′ is sequentially changed by a predetermined amount, and a criterion for determining whether the root mean square value of the difference between the measurement errors ΔZ _r and ΔZ _s before and after the offset value change is the minimum value. If the offset value is calculated as follows, the optimum correction value for the origin position error can be obtained even when the noise component of the three-dimensional shape measuring machine is large.

From the measurement data (X _r , Y _r , Z _r ) of the measured surface 204a calculated in Step 76 and the measured data (X _s , Y _s , Z _s ) of the measured surface 204b, the measured surface 204a and the measured surface 204b Relative distance T is calculated and stored in the storage means 218 (Step 78).

  Looking at the relative distance T between the measured surface 204a and the measured surface 204b calculated in Step 78, the measurement error due to the origin position errors in the x-axis direction and the y-axis direction of the probes 205a and 205b is corrected. However, the measurement error due to the origin position error of the probes 205a and 205b in the z-axis direction (the relative distance error in the z-axis direction between the z-axis of the first measurement coordinate system and the z-axis of the second measurement coordinate system) is not corrected. Therefore, the relative distance T includes a measurement error due to an origin position error in the z-axis direction.

Therefore, in the present embodiment, the value T ₀ measured by another measuring means is used as the distance between the measured surface 204a of the measured object 204 and the measured surface 204b (so-called thickness of the measured object 204). As a means for measuring the thickness of the object to be measured 204, there are a number of known means of non-contact type or contact type, and the thickness can be measured with high accuracy. Alternatively, when the measurement accuracy of the thickness of the measurement object 204 is not important, the design value of the measurement object can be used.

In the data correction means 221, the measurement data (X _r , Y _r , Z _r ) of the measurement target surface 204a or the measurement target surface is set so that the relative distance T of the measurement target 204 calculated in Step 78 matches the thickness T _0. The measurement data (X _s , Y _s , Z _s ) of 204b or both are corrected to measure the measurement data (X _r ′, Y _r ′, Z _r ′) of the measured surface 204a and the measured data (X _s ′, Y _s ′, Z _s ′) and stored in the storage means 218 (Step 79).

Measurement data of the measurement surface 204a calculated in _{Step79 (X r ', Y r} ', Z r ') measurement data of the measured surface _{204b (X s', Y s} ', Z s') , the surface to be measured The three-dimensional shape of the object 204 to be measured including the relative positional relationship between 204a and the surface to be measured 204b can be found. The measurement result of the object to be measured 204 is displayed on the display means 222 as a desired format such as a numerical value or a two-dimensional or three-dimensional figure (Step 80).

  In the present embodiment, the shape is measured by moving the probes 205a and 205b along the x-axis, so that the probes 205a and 205b do not move along the y-axis, or the amount of movement is small. Therefore, the y coordinate of the probe 205a and the probe 205b has been described as being negligible (y = 0). However, when the probe moves along the y-axis and affects the shape measurement data, it is preferable to measure because the y-coordinate of the probe cannot be ignored.

For example, when the y coordinate of the probe 205a cannot be ignored, in step 73, the xyz coordinate (x, y, z) of the probe 205a and the rotation angle (θ) of the air spindle 201 are fetched as point sequence data, and the measured surface 204a is measured. Shape measurement data (x _r , y _r , z _r , θ _r ) (r = 1, 2, 3,...) Are acquired. In Step 74, the shape measurement data (x _r , y _r , z _r , θ _r ) is converted into measurement data (X _r , Y _r , Z _r ) in an orthogonal coordinate system. For Step 75 and subsequent steps, the same operation is performed with the y coordinate taken into account.

  Further, in the configuration in which the probe is moved along the x-axis and the z-axis as in this embodiment, the influence of the origin position error in the y-axis direction on the shape measurement data is smaller than the origin position error in the x-axis direction. small. Therefore, by adjusting the position of the probe 205a, probe 205b, and air spindle 201 in the y-axis direction by using the y stage 212a and the y stage 212b, it is possible to reduce the influence of the origin position error in the y-axis direction. It has been described that the origin position error in the y-axis direction can be ignored. However, depending on the measurement accuracy required for the three-dimensional shape measuring machine, it is preferable to correct the origin position error in the y-axis direction.

For example, when the origin position error of the probe 205a in the y-axis direction cannot be ignored, in Step 76, an offset for correcting the origin position error of the probe 205a in the y-axis direction is performed in the same procedure as the calculation of the origin position error in the x-axis direction. The value Δy _a 'is obtained. In Step 77, the offset value Δx _a ′ for correcting the origin position error in the x-axis direction of the probe 205a and the offset value Δy _a ′ for correcting the origin position error in the y-axis direction are added as parameters to the shape measurement data, and the surface to be measured 204a of the shape measurement data _{(x r + Δx a ',} y r + Δy a', z r, θ r) and. In this way, not only the origin position error in the x-axis direction of the probe but also the origin position error in the y-axis direction can be corrected.

  According to the present embodiment, the origin position errors in the x-axis direction and the y-axis direction of the probe 205a and the probe 205b with respect to the rotation axis 201A of the air spindle 201 are calculated using the shape measurement data of the object to be measured 204. Therefore, the three-dimensional shape of the object to be measured including the relative positional relationship between the front surface and the back surface can be evaluated with high accuracy without using a reference sphere or plane parallel plate with good surface accuracy.

  In the present embodiment, the shape measurement data of the object to be measured is corrected based on the origin position correction values of the probe 205a and the probe 205b obtained using the object to be measured 204, but based on the origin position correction value. Measurement may be performed after correcting the initial positions of the probes 205a and 205b, that is, the origin position of the measurement coordinate system.

  In the present embodiment, the three-dimensional shape measuring machine using the contact type probe has been described as an example. However, the non-contact type three-dimensional shape measuring machine using the known optical probe as in the third embodiment. It may be.

Further, in this embodiment, in Step 72, when the position of the probe 205a is at the position 205A in FIG. 21, the xyz coordinate value of the probe 205a is reset, but the xyz coordinate of the probe 205a is near the position 205B in FIG. Can be reset to measure the shape of the measured surface 204a. In this case, the initial value of the xyz coordinate of the probe 205a is set to (−R _a , 0, 0). Then, the shape measurement data is acquired by moving the probe 205a along the x-axis direction by a series of operations from the position 205B to the position 205C in FIG. In this way, the movement along the x-axis of the probe 205a at the time of shape measurement only needs to be performed in one direction (from the position 205B to the position 205C in FIG. 21), so that the measurement time can be further shortened.

  Further, after acquiring the shape measurement data in the range where the x coordinate of the probe 205a is negative, the probe 205a is once separated from the surface to be measured 204a, and moved by the x stage 213a until the x coordinate of the probe 205a becomes a positive range. The probe 205a may be brought into contact with the surface to be measured 204a again, and shape measurement data in the range where the x coordinate of the probe 205a is plus may be acquired. In this way, it is possible to measure the shape of a donut-like object to be measured in which a circular hole is opened in the center part (near the rotating shaft 201A) of the surface to be measured 204a.

  In Step 73 of the present embodiment, the shape measurement data is acquired as point sequence data. However, in the following Step 74 to Step 80, the shape measurement data in which the x coordinates of the probes 205a and 205b are in the minus range and the plus range. The approximate surface equation corresponding to each of the shape measurement data may be obtained as appropriate, and the correction value may be calculated based on the obtained approximate surface equation.

  Although the embodiments of the present invention have been described above with reference to the drawings, the present invention is not limited to these embodiments, and various modifications and changes can be made without departing from the scope of the present invention. May be.

It is a fragmentary sectional view which shows roughly the structure of the principal part of the three-dimensional shape measuring machine in 1st embodiment of this invention. 1 shows a state in which a parallel plane plate as a reference object for correcting an inclination error of the x-axis of the measurement coordinate system is attached to the three-dimensional shape measuring machine shown in FIG. 1 shows a state in which a reference sphere, which is a reference object for correcting the squareness error between the x-axis and the z-axis of the measurement coordinate system, is attached to the three-dimensional shape measuring machine shown in FIG. The flowchart of the procedure which calculates | requires each inclination error correction value of two measurement coordinate systems in the three-dimensional shape measurement in 1st embodiment of this invention is shown. The flowchart of the procedure which calculates | requires the squareness error correction value of each of two measurement coordinate systems and the correction value of the relative distance of z direction of two measurement coordinate systems in the three-dimensional shape measurement in 1st embodiment of this invention is shown. Yes. The flowchart of the procedure which evaluates the three-dimensional shape of the surface of a to-be-measured object and a back surface in the three-dimensional shape measurement in 1st embodiment of this invention is shown. In the first embodiment of the present invention, measurement error data obtained using a reference plane when there is an inclination error of the x-axis of the measurement coordinate system with respect to a plane orthogonal to the rotation axis of the air spindle is shown. In the first embodiment of the present invention, measurement error data obtained using a reference spherical surface when there is a squareness error of the z axis with respect to the x axis of the measurement coordinate system is shown. 1 shows a state in which a parallel plane plate is attached to the three-dimensional shape measuring machine shown in FIG. 1 in order to correct an angle error in the θ direction of the x axis of the measurement coordinate system. The flowchart of the procedure which calculates | requires (theta) direction angle error correction value in the three-dimensional shape measurement in 2nd embodiment of this invention is shown. The flowchart of the procedure which evaluates the three-dimensional shape of the surface of a to-be-measured object and a back surface in the three-dimensional shape measurement in 2nd embodiment of this invention is shown. It is a fragmentary sectional view which shows roughly the structure of the principal part of the three-dimensional shape measuring machine in 3rd embodiment of this invention. The flowchart of the three-dimensional shape measurement in 3rd embodiment of this invention is shown. In the third embodiment of the present invention, measurement error data obtained using a reference spherical surface when the optical probe has an origin position error in the x-axis direction is shown. In the third embodiment of the present invention, measurement error data obtained using a reference plane when the optical probe has an origin position error in the x-axis direction is shown. In the third embodiment of the present invention, measurement error data obtained using a reference spherical surface when the optical probe has an origin position error in the y-axis direction is shown. It is a fragmentary sectional view which shows the structure of the three-dimensional shape measuring machine in 4th embodiment of this invention. FIG. 18 shows an object to be measured and a probe shown in FIG. 17 in an enlarged manner. It is a block diagram which shows the control system of the three-dimensional shape measuring machine shown by FIG. The flowchart of the three-dimensional shape measurement in 4th embodiment of this invention is shown. It is a figure explaining the measurement range in the three-dimensional shape measurement of 4th embodiment of this invention. The shape error data in the case where the probe has an origin position error in the x-axis direction in the three-dimensional shape measurement of the fourth embodiment of the present invention are shown. 1 shows one three-dimensional shape measuring machine according to a conventional example. 3 shows another three-dimensional shape measuring machine according to a conventional example. The tilt error between the first measurement coordinate system and the second measurement coordinate system with respect to the reference coordinate system is shown. Shows the relative error Δθ of the X ₁ axis and θ directions of the X ₂ axis of the second measurement coordinate system of the first measurement coordinate system.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 ... Air spindle, 101A ... Rotating shaft, 102 ... Holding part, 103a, 103b ... Probe, 104 ... Object to be measured, 105 ... Parallel plane plate, 105a, 105b ... Reference plane, 106 ... Reference sphere, 106a, 106b ... Reference Spherical surface, 107a, 107b ... optical probe, 201 ... air spindle, 201A ... rotating shaft, 202 ... holding part, 205a, 205b ... probe, 207a, 207b ... ruby ball, 208a, 208b ... air slide, 209a, 209b ... position detection 212a, 212b ... y stage, 213a, 213b ... x stage, 215 ... control means, 216 ... program memory, 218 ... storage means, 219 ... coordinate conversion means, 220 ... correction value calculation means, 221 ... data correction means.

Claims (4)

A method of measuring the three-dimensional shape of the front and back surfaces of the object to be measured with the first probe and the second probe, respectively, and coordinates defined by a holding unit that rotates the object to be measured prior to the measurement. A first measurement coordinate system defined by a first probe that measures the surface of the object to be measured and a second measurement coordinate system defined by a second probe that measures the back surface of the object to be measured A correction step for correcting the error, and the correction step includes a first correction step and a second correction step;
The first correction step uses a first reference object having two reference planes parallel to each other,
Holding the first reference object so that the normal of one plane of the first reference object is substantially parallel to the rotation axis of the holding unit;
Rotating the first reference object around the rotation axis;
One reference plane of the first reference object is measured with a first probe to obtain shape measurement data A1,
The other reference plane of the first reference object is measured with a second probe to obtain shape measurement data A2,
A correction coefficient L1 is calculated from the shape measurement data A1 and the design shape data of the first reference object,
A process of calculating a correction coefficient L2 from the shape measurement data A2 and the design shape data of the first reference object is performed,
The second correction step uses a second reference object having two reference spheres that are part of the same sphere,
Holding the second reference,
Rotating the second reference object around the axis of rotation;
Measuring one reference spherical surface of the second reference object with the first probe to obtain shape measurement data B1,
Measuring the other reference spherical surface of the second reference object with the second probe to obtain the shape measurement data B2,
The correction coefficient L1 is added to the shape measurement data B1 to obtain new shape measurement data B1 ′,
The correction coefficient L2 is added to the shape measurement data B2 to obtain new shape measurement data B2 ′,
A correction coefficient M1 is calculated from the shape measurement data B1 ′ and the design shape data of the second reference object,
A correction coefficient M2 is calculated from the shape measurement data B2 ′ and the design shape data of the second reference object,
A three-dimensional shape measurement method, comprising: performing a process of calculating a correction value ΔN from the shape measurement data B1 ′, the shape measurement data B2 ′, and the design shape data of the second reference object. In claim 1, the correction step further includes a third correction step, the third correction step uses the first reference object,
Holding the first reference object so that the normal of one reference plane of the first reference object is inclined with respect to the rotation axis of the holding unit;
Rotating the first reference object around the rotation axis;
One reference plane of the first reference object is measured with the first probe to obtain shape measurement data C1,
Measuring the other reference plane of the first reference object with the second probe to obtain the shape measurement data C2,
Adding the correction coefficient L1 and the correction coefficient M1 to the shape measurement data C1 to obtain new shape measurement data C1 ′;
Adding the correction coefficient L2 and the correction coefficient M2 to the shape measurement data C2 to obtain new shape measurement data C2 ′;
A three-dimensional shape measurement method, comprising: performing an angle error correction value Δθ processing from the shape measurement data C1 ′, the shape measurement data C2 ′, and the design shape data of the first reference object. A method of measuring the three-dimensional shape of the front and back surfaces of the object to be measured with the first probe and the second probe, respectively, and coordinates defined by a holding unit that rotates the object to be measured prior to the measurement. A correction step for correcting a relative position error between the system and a first probe for measuring the surface of the object to be measured and a second probe for measuring the back surface of the object to be measured. , Using a reference object having a reference surface on the opposite front and back,
Holding the reference,
Rotating the reference object around the rotation axis of the holding unit;
One reference surface of the reference object is measured with a first probe to obtain shape measurement data D1,
The other reference surface of the reference object is measured with a second probe to obtain shape measurement data D2,
A correction value ΔX1 is calculated from the shape measurement data D1 and the design shape data of the reference object,
A three-dimensional shape measurement method, comprising: performing a process of calculating a correction value ΔX2 from the shape measurement data D2 and the design shape data of the reference object. A method for measuring a three-dimensional shape of a front surface and a back surface of a measurement object, a coordinate system defined by a holding unit that rotates the measurement object, a first probe that measures the surface of the measurement object, and A correction step of correcting a relative position error with the second probe for measuring the back surface of the object to be measured;
Holding the object to be measured;
Rotate the object to be measured around the rotation axis of the holding unit,
Of one surface of the object to be measured, a range on one side with respect to the rotation axis is measured with a first probe to obtain shape measurement data E1, and a range on the other side with respect to the rotation axis is determined. The shape measurement data E1 ′ is obtained by measuring with the first probe,
Of the other surface of the object to be measured, a range on one side with respect to the rotation axis is measured with a second probe to obtain shape measurement data E2, and a range on the other side with respect to the rotation axis is determined. The shape measurement data E2 ′ is obtained by measuring with the second probe,
A correction value ΔX1 ′ is calculated from the shape measurement data E1, the shape measurement data E1 ′, and the design shape data of the object to be measured,
A three-dimensional shape measurement method, comprising: performing a process of calculating a correction value ΔX2 ′ from the shape measurement data E2, the shape measurement data E2 ′, and the design shape data of the object to be measured.

Images