JP2006242606A - Method for measuring runout of rotating tool and its measuring apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To measure runout of cutting edges of rotary tools with accuracy of submicron order in a high-speed rotation range between extremely low-speed rotation and high-speed rotation at 100,000/min or faster. <P>SOLUTION: By arranging a lens B in such a way that a cutting edge of a rotary tool may be matched with a front focus point of the lens B, arranging a dimmer filter at a rear focal point of the lens B, matching a front focal point of a lens C with the rear focal point of the lens B, and arranging an imaging element at a rear focal point of the lens C, Fraunhofer images or shadow diffraction images of the cutting edge are acquired from the imaging element to improve accuracy in calculating the position of the center of a ball of the rotary tool and compute runout. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

The present invention relates to precise measurement of runout of a rotary tool in precision machining using a rotary tool, and aims to suppress runout based on the measurement result of the runout.

Precision machining using a rotary tool is generally performed. However, if there is runout due to eccentric error or chucking error of the rotary tool, the machining shape accuracy deteriorates and the tool life decreases. This is an obstacle to machining.
In response to this problem, for example, in Japanese Patent Laid-Open No. 10-141934, the shape accuracy and run-out of the cutting edge of the ball end mill are obtained by non-contact measurement using laser light with reference to the motion accuracy of the processing machine. However, this method has problems such as sequential measurement, which takes time and makes the measurement accuracy dependent on the machine accuracy.

In Japanese Patent Laid-Open No. 2002-205246, a point on a cutting edge ridge line of a ball end mill is measured to detect a radius accuracy, and this measurement method can also be applied to a runout measurement aimed at by the present invention. However, this measurement method also has a problem that it takes time for measurement because of sequential measurement, as in JP-A-10-141934.
In addition to the above, many methods have been proposed in which the laser beam is irradiated to the rotary tool, and the fluctuations in the position of the rotary tool that accompany the rotation are measured one by one to obtain the deflection.

However, in all of these conventional vibration measurement methods using laser light, the coherence of the laser light is very good. Therefore, due to the mutual interference between the diffracted light generated at the cutting edge of the rotary tool and the transmitted light, the cutting edge There is a problem that the boundary of is unclear and the identification accuracy decreases.
In addition, when the identification accuracy of the edge of the cutting edge is low in this way, if severe machining shape accuracy such as micron or submicron is required,
1. 1. measure and correct for runout of the rotating tool; Trial cutting with the corrected rotating tool, and measuring and evaluating machining shape accuracy,
If it is not repeated many times, the required machining shape accuracy cannot be secured.

  In particular, as the rotary tool becomes smaller in diameter and the machining shape accuracy becomes more severe, it is necessary to increase the measurement accuracy and correction accuracy of the cutting edge ridge runout of the rotary tool. Will have to spend a lot of time. In addition, the smaller the diameter of the rotating tool, the higher the rotational speed of the tool, but the higher the speed, the more difficult it is to measure the shape of the cutting edge.

  By limiting the amount of light in the transmission optical system, the shape of the cutting edge ridge can be measured with high accuracy.

13 and 14 are explanatory diagrams for the effects of the present invention.
FIG. 13 shows the relationship between the run-out of the ball end mill measured using the run-out measuring device of the present invention and the rotational run-out obtained from the measurement data of the three-dimensional shape of the concave spherical surface cut by the ball end mill. Both show a very good linear relationship. By using the run-out measuring device, the run-out of the tool can be accurately measured to 1 μm or less, and a good concave spherical surface can be obtained by correcting the measured run-out. It is clear that can be processed.

The single-crystal diamond single-blade ball end mill 17 is chucked on the rotation axis 18 of the rotating tool of the inclined machining unit 39 of the portal jig boring machine 35 shown in FIG. 7, and the ball end mill rotates. Fig. 14 shows a micrograph of the machined surface of the workpiece 1 when it is first cut with a feed, then revolved to 180 ° around the spindle, and then retracted. FIG. 14A shows a result of cutting without correcting the shake, and the shake is large.
1) It has a large diameter in the lateral direction. 2) At the first 0 ° position and 180 ° end position, there is a marked radial boundary in the lower half.
This is confirmed.

By the way, the tool runout at this time was measured with the runout measuring device of the present invention and found to be 70.5 μm. On the other hand, when the concave spherical shape of FIG. 14A is measured with a three-dimensional shape measuring instrument and the geometric processing shape error of the concave spherical surface is calculated based on the measured data, the tool runout is about 70 μm. It was found that the measurement results with the shake measurement device of the present invention corresponded well.

  Based on the measurement result of the runout, the runout of the tool was corrected to 1 μm or less by the self-revolving unit described in JP-A-2002-361510. Fig. 14 (b) shows the processing result of cutting in this state. In the concave spherical surface, the boundary due to run-out is not confirmed, and it is clear that the run-out of the tool is small. As described above, when the tool runout was calculated from the geometric machining shape error based on the measurement data of the three-dimensional shape of the concave spherical surface, it was confirmed that the runout was suppressed to 1 μm or less. .

In addition, it has been confirmed that even with a single blade ball end mill with a ball radius of 0.5 mm rotating at 100,000 / min, the deflection measurement is possible by using the deflection measuring device.
As described above, by using the run-out measuring device of the present invention, it is possible to measure run-out of a rotating tool even with a rotating tool that rotates at high speed. In addition, while seeing the Fraunhofer image as shown in FIG. 8 or the shadow diffraction image as shown in FIG.
Thus, in the shake measurement method using the neutral density filter of the present invention,
1) A high-contrast concentric francopha image or shadow diffraction image can be obtained.
2) The position of the center of the ball of the rotating tool is obtained from the individual fringes of the concentric francophor image or shadow diffraction image, and the image is distorted as shown in FIG. The center position calculation accuracy can be improved even in
It has the following features.
The machining shapes targeted by the present invention are shown in FIGS. FIG. 1 shows a concave spherical surface 2 having a radius of curvature R and having a rotationally symmetric shape around a central axis 3. Further, the processing object is a concave cylindrical surface whose cross section shown in FIG. 2 is a part of a circle 6 and a concave elliptic cylindrical surface 7 which is a part of an ellipse 8 in FIG. When machining these concave spherical surface 2, concave cylindrical surface 5 or concave elliptic cylindrical surface 7, a ball end mill is used as a rotating tool.

  Fig. 4 shows the AA cross section in Fig. 1. Here, the ball center Q ′ of the ball end mill having the ball radius R does not exist strictly on the rotational axis 4 of the tool due to the eccentricity of the cutting edge of the ball end mill or a chucking error. Have Here, if there is a distance from the rotation axis 4 of the tool to the ball center Q ′, that is, a tool runout ε, the ball end mill swings around the rotation axis 4. As a result, a concave spherical processed shape having a radius of curvature R ′ different from the design value R and a shape error Δe is obtained. In FIG. 4, these relationships are exaggerated for the sake of explanation. Similarly, the concave cylindrical surface 5 shown in FIG. 2 to the concave elliptic cylindrical surface 7 shown in FIG. 3 have a problem that if the tool is shaken, a similar curvature radius error or shape error is caused.

Thus, in order to precisely machine a concave spherical surface, a concave cylindrical surface, or a concave elliptical cylindrical surface, it is indispensable to measure the tool runout ε with high accuracy and correct the runout ε to zero as much as possible. .
In order to deal with the above problems, the object of the present invention is to measure the runout of the cutting edge of a rotating tool with a high accuracy on the micron or submicron level in a wide rotational speed range from very low speed to several hundred thousand revolutions per minute. The purpose of this is to provide a measurement method that can be used to compensate for the fluctuation of the cutting edge of a rotating tool.

  In response to this problem, a neutral density filter that limits the amount of zero-order light and high-order light is inserted in the incident optical system so that a part of the transmitted light emitted from the light source does not reach the image sensor. A Francophor image or a shadow diffraction image of the cutting edge of a rotating tool is obtained. As a result, the calculation accuracy of the ball center of the rotating tool can be improved, and the runout of the cutting edge of the rotating tool can be obtained with high accuracy.

  An embodiment of the present invention will be described with reference to FIGS. FIG. 5 shows an embodiment of the shake measuring apparatus of the present invention, FIG. 6 is a diagram for explaining the principle thereof, and FIG. 7 is a portal jig boring machine equipped with the shake measuring apparatus shown in FIG. The board 35 is shown.

  The shake measuring device is composed of a shake measuring unit 9, a light source module 11, and a personal computer 20. In the shake measurement unit 9, a light source module 11, a lens barrel module 13, and a detection module 14 are arranged on a base 10 in this order with a common optical axis 16, and these are covered by a cover 15 as a whole. ing. A window is opened on a part of the upper surface of the cover, and a ball end mill 17 to be measured passes through the window, and the ball center of the ball end mill is assembled so as to substantially coincide with the optical axis 16. It has been. Further, the detection end of the origin angle detector 19 is arranged close to the outer periphery of the ball end mill 17 and the other end is fixed to a stand standing on the base 10 to detect the origin of the ball end mill 17 in the rotation angle direction. It is possible.

The light source power source 12 is electrically connected to the light source module 11 and supplies power to the light source module. The image signal from the detection module 14 and the origin signal from the origin angle detector 19 are electrically connected to the personal computer 20.
FIG. 7 shows a case where the optical axis 16 of the shake measuring device is taken in parallel to the X-axis direction of the portal jig boring machine 35, and the center of the ball of the ball end mill 17 described above is shown in FIG. The predetermined position of the measuring device is matched with NC control commands in the X-axis, Y-axis and Z-axis directions of the portal jig boring machine 35.

Next, the measurement principle of the present invention is described.
FIG. 6 shows an optical system in which key parts of the shake measurement unit 9 shown in FIG. 5 are arranged on a plane in order to explain the measurement principle of the present invention. The laser oscillator 21 and the lens group A built in the light source module 11, the lens B built in the lens barrel module 13, the neutral density filter 25 and the lens C, and the image pickup device 26 built in one end of the detection module 14 are arranged in this order. Therefore, it is arranged on the optical axis 16 common to the laser beam emitted from the laser oscillator 21. Here, the rear focal point 28 of the lens B coincides with the front focal point 28 ′ of the lens C, which is called confocal here. At this confocal position, the amount of transmitted light 30 that converges on the rear focal point 28 through the lens B, the so-called zero-order light amount, and the light amount of higher-order diffracted light 29 can be reduced to 1: 1. A neutral density filter 25 is arranged. The rear focal point 26 ′ of the lens C coincides with the detection surface of the image sensor 26. Further, the front focal point 27 ′ of the lens B and the cutting edge ridge 27 of the ball end mill 17 to be measured are made to coincide.

With this configuration of the optical system, when the laser beam 31 emitted from the laser oscillator 21 is collimated through the lens group A and irradiated onto the cutting edge ridge 27 of the ball end mill 17, the cutting edge Multi-order diffracted light 29 is generated at the edge 27. In addition, the transmitted light 30 of the 0th-order light component that passes through the portion of the laser light 31 away from the cutting edge ridge 27 enters the lens B as parallel light. Since the cutting edge ridge 27 coincides with the front focal point 27 ′ of the lens B, the transmitted light 30 and the diffracted light 29 that pass through the periphery of the cutting edge ridge 27 are reflected by the lens B by a Fourier transform action. At the rear focal point 28, it is decomposed into spatial frequencies. That is, the transmitted light 30 is collected, but the diffracted light 29 is not collected. At the position of the rear focal point 28 of the lens B, a neutral density filter 25 having a filter function for reducing the amount of zero-order light and high-order light to 1: 1 is installed. Is dimmed and reaches the image sensor 26 disposed at the rear focal point 26 'of the lens C.

On the other hand, the high-frequency component diffracted light 29 transmitted through the neutral density filter 25 is coupled onto the image pickup device 26 disposed at the rear focal point 26 'of the lens C by the inverse Fourier transform action of the lens C after being attenuated. Image. At this time, since the diffracted light 29 and the transmitted light 30 have the same amount of light in the image sensor 26, as shown in FIG. 8 or FIG. 9, a concentric interference image having a good contrast generated by the cutting edge ridge 27 is obtained. A Francophor image 32 as shown in FIG. 8 is extracted outside the boundary of the cutting edge ridge 27 of the ball end mill 17 or a shadow diffraction image 33 as shown in FIG. 9 is extracted inside.

Next, how to obtain the maximum value of the deflection of the rotating cutting edge ridge 27 and the rotation angle at which the deflection takes the maximum value will be described with reference to FIGS. . First, out of the shadow diffraction images detected at regular time intervals, a shadow diffraction image 32 ′ having a maximum shake value and a shadow diffraction image 32 ″ at a rotation angle of 180 ° relative to this are obtained. Next, it is idealized as a concentric circle group 40, and for example, moire fringes 41 such as those shown in Fig. 10 generated by superimposing the left and right concentric circle groups are corrected by correcting the ball end mill 17 in a direction to reduce the number of the fringes. 11 can calculate and correct the position of the cutting edge ridge 27. Fig. 11 shows a state where there is no moire fringes, that is, a state where the position of the cutting edge ridge 27 is calculated and corrected with high precision. The shadow diffraction image can also be calculated and corrected in the same way.

  Since the origin in the rotation direction of the ball end mill is detected by the origin angle detector 19, the rotation angle from the shadow diffraction image 32 ′ to the shadow diffraction image 32 ″ where the shake is the maximum is determined from the origin signal. Can be calculated from the time to reach the maximum value.

In addition, the shadow diffraction image from the image sensor can be captured at a high sampling rate of 1 μs or less with an electronic shutter, and shake measurement errors caused by intermittent digital data at each sampling interval can be ignored.
Even if the sampling rate is slow, the sine wave approximation based on intermittent digital data at each sampling interval and interpolation of the data can reduce the calculation error.

FIG. 7 shows an example in which the run-out measuring device shown in FIG. 5 is mounted on a portal jig boring board 35. The shake measurement unit 9 is fixed to the X-axis table 36 so that the optical axis 16 is parallel to the X-axis direction.
On the other hand, an inclination machining unit 39 capable of alignment correction with an accuracy within 1 μm is coupled to the main shaft 37 of the portal jig boring machine 35 as described in JP-A-2002-361510. Revolution around 38 is possible. In addition, the inclined machining unit has a built-in tool rotation shaft 18 ′ having a rotation axis 18 of the tool inclined with respect to the main shaft axis 38 and capable of rotating, and a single blade at one end of the tool rotation shaft. A ball end mill 17 having a cutting edge ridge 27 made of single crystal diamond is chucked, and minute alignment correction can be made in a direction perpendicular to the tool rotation axis 18 '.

Figure 12 shows the procedure for measuring and correcting the shake.
First, the runout measurement unit 9 is installed on the X-axis table 36 of the portal jig boring board 35. Next, the ball center of the ball end mill 17 is aligned with the front focal point of the lens B on the optical axis of the shake measuring unit. Subsequently, the projected image of the cutting edge and the origin angle of the rotary tool are taken into the personal computer. Based on these projection image data and origin angle data, the rotation angle at which the shake is at its maximum value and the maximum value of the shake at that time are obtained by analysis, and the shake is zero at the obtained rotation angle position. , To correct the shake.

  The concave cylindrical surface 5 shown in FIG. 2 can be machined by setting the same rotary tool and portal jig boring machine as the concave spherical surface 2 shown in FIG. Further, the concave elliptic cylindrical surface 7 in FIG. 3 is arranged so that the rotation axis 4 of the tool in the feed direction in the set-up state of the rotary tool and the portal jig boring machine similar to the processing of the concave spherical surface shown in FIG. It can be machined by tilting and feeding it. This can be achieved, for example, by feeding a rotating tool that rotates in a state where the inclined machining unit attached to the main shaft of the portal type boring machine is rotated at a certain angle in the revolution direction. .

The present invention can be applied to a square end mill, a grindstone, or a drill instead of the ball end mill 17 as a rotating tool, and the same runout measurement and runout correction can be performed.

Even if an encoder is used instead of the origin angle detector 19, it is possible to detect the rotation angle at which the maximum value of the shake is obtained, and the same effect can be obtained.

The run-out measuring device of the present invention can be mounted not only on the portal jig boring machine 35 described above but also on a general-purpose milling machine, a multi-axis machine, a jig grinder, etc., and the same effects as described above can be obtained.

Example of concave spherical surface to be processed Example of concave cylindrical surface to be machined Example of concave elliptic cylinder surface to be machined Illustration of shape error due to runout of rotating tool The figure which shows one Example of the shake measuring apparatus of this invention Explanatory drawing showing the principle of the present invention FIG. 5 is an explanatory diagram of a gate-type jig boring machine equipped with the embodiment of FIG. Fraunforfa example Example of shadow diffraction image Example of moire fringes that occur before adjusting the edge position Example of no moire fringe after cutting edge ridge position adjustment Flow chart showing run-out measurement / correction procedure Diagram showing the effect of this measurement Comparison of concave spherical surfaces due to runout differences Procedure for obtaining the ball center position with high accuracy from a distorted image

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Workpiece 2 Concave spherical surface 3 Center axis of concave spherical surface 4 Tool rotation axis 5 Concave cylindrical surface 6 Circle 7 Concave elliptical cylindrical surface 8 Ellipse 9 Shaking measurement unit 10 Base 11 Light source module 12 Power source for light source 13 Lens barrel 14 Detection Module 15 Cover 16 Optical axis 17 Ball end mill 18 Tool rotation axis 18 ′ Tool rotation axis 19 Origin angle detector 20, 20 ′ PC 21 Laser oscillator 22 Lens group A
23 Lens B
24 Lens C
25 Neutral filter 26 Imaging element 26 ′ Rear focus 27 of lens C Cutting edge ridge 27 ′ Front focus 28 of lens B Rear focus 28 ′ of lens B Front focus 29 of lens C Diffracted light 30 Transmitted light 31 Laser light 32 , 32 ′, 32 ″ Francophor image 33, 33 ′, 33 ″ shadow diffraction image 35 Gate-type jig boring machine 36 X-axis table 37 Spindle 38 Spindle axis 39 Inclined machining unit 40 ′, 40 ″ Concentric circle group 41 Moire fringe 42 Cross section 43 passing through the axis of the concave elliptic cylinder surface and perpendicular to the workpiece surface 43

O The center of the projection circle of the concave sphere P The apparent center of curvature of the concave sphere with an eccentric error
Q Center of curvature of concave sphere Q ', Q "Ball center of tool R Radius of curvature of concave sphere R' Apparent radius of curvature of concave sphere with eccentric error ε Eccentricity of tool ball center relative to tool rotation axis , That is, runout Δe shape error θ inclination angle of rotation axis in cross section 42

Claims (2)

A light source module that emits parallel light, a lens barrel and an image sensor that share the optical axis and axis of the parallel light, and a microcomputer having an arithmetic processing function, and a rotation placed at a position that blocks the parallel light In a measurement system that measures tool runout,
The lens barrel includes at least two lenses having a confocal point and a neutral density filter for limiting the amount of zero-order light of incident light disposed in the vicinity of the confocal point. Of the remaining two focal points, the cutting edge of the rotary tool is disposed at the focal point on the light source module side, and the imaging element is disposed at the other focal point so as to be orthogonal to the optical axis, and 2) Two Francophor images showing the maximum shake at rotational angles different from each other by taking an electrical signal of the Francophor image or shadow diffraction image of the cutting edge ridge projected on the image sensor by a microcomputer The runout of the cutting edge of the rotary tool can be calculated from the difference between the two or the shadow diffraction image,
Measuring method and measuring device of rotating tool characterized by The measuring system according to claim 1, further comprising an angle detector or an origin angle detector, wherein the rotational angle of the rotary tool can be detected.

Images