JP2000263391A - Surface shape evaluation method and high-precision smooth polishing method using the same - Google Patents

Abstract

(57) [Summary] The present invention relates to a method for evaluating the surface shape of a molded article surface,
And a polishing tool for polishing a polished surface of a workpiece by using the polishing tool, and a high-precision smooth polishing method for polishing a polished surface with high accuracy. The undulation is accurately evaluated, and the undulation having a main undulation wavelength. The purpose is to remove easily and reliably. SOLUTION: In a high-precision smooth polishing method for polishing a polished surface of a workpiece with a polishing tool and smooth-polishing the polished surface with high precision, after measuring a precise shape of the polished surface and obtaining shape measurement data. Frequency analysis of the shape measurement data, extraction of a main undulation wavelength on the polished surface, and polishing of the polished surface with a polishing tool having a diameter of twice or more the main undulation wavelength.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for evaluating the surface shape of a surface of a molded product, and a method for polishing a polished surface of a workpiece using a polishing tool using the method, and for smoothing the polished surface with high precision. It relates to a smooth polishing method.

[0002]

2. Description of the Related Art Generally, very high precision and quality are required for optical elements such as lenses and reflection mirrors, which are ultraprecision optical parts used for precision equipment, and ultraprecision dies for molding lenses. Have been. In the finish processing of such an optical element or a mold, undulations formed on a surface to be processed have been conventionally removed with high accuracy by a polishing tool such as a polisher.

Conventionally, as a high-precision smooth polishing method for removing such waviness existing on a surface to be processed, for example, a method disclosed in Japanese Patent Application Laid-Open No. 4-256562 is known. In this high-precision smooth polishing method, as shown in FIG. 8, a polishing tool 1 having a diameter D that is at least twice the undulation period P to be removed is used, and undulation is applied to the polishing tool 1 to remove undulation. .

Therefore, in such a high-precision smooth polishing method, when selecting the tool diameter D of the polishing tool 1, it is necessary to accurately determine the undulation wavelength to be removed. Conventionally,
Such a measurement of the undulation wavelength has been performed by directly measuring the interval between individual undulations, for example, by observing an interference fringe shape obtained from a laser interferometer on a monitor.

[0005] For example, the shape measurement data obtained from the measuring device is displayed in a three-dimensional figure by a computer, and the wavelength of the undulation to be removed is estimated from the obtained three-dimensional figure. Further, for example, from the obtained shape measurement data, extracting the center cross section of the workpiece, estimating the undulation from the cross-sectional shape, is performed by directly measuring the interval of the estimated undulation convex or concave portions. I have.

[0006]

However, in the above-described method of observing the interference fringe shape on a monitor, it is very difficult to determine a swell having a height (amplitude) of several tens of nanometers, which is a problem in, for example, ultra-precision machining. It was difficult. In addition, there is also a problem that human variation among observers is large, and an error is likely to occur in specifying the swell wavelength.

That is, it is easy to specify the swell which is a single sine wave schematically shown in FIG. 8, but the actual work surface is a sine wave having a single period (wavelength). It is seldom composed of only undulations, and as shown in FIG. 9, an aggregate (polymer) of various undulations having a plurality of periods (wavelengths)
It has become. Therefore, it is not easy to specify a dominant swell wavelength (hereinafter referred to as a main swell wavelength) to be removed from the swell component by a direct observation method, and individual differences and variations are likely to occur. .

FIG. 9 shows the shape measurement data of the center cross section of the data obtained by measuring the surface to be processed with a Fizeau laser interferometer and calculating the difference (shape error) from the design shape. . Generally, in a normal processing step, the undulation polishing step is often performed at a relatively early stage of the polishing step as a step before the shape correction step (final step).

Incidentally, undulation polishing removes undulations from a high frequency to an intermediate frequency range, and shape correction processing removes undulations in a low frequency range. FIG. 10 shows shape measurement data of the center cross section of the workpiece.
The shape of the surface to be processed is represented by a position where the vertical axis is 0 (the value of the vertical axis is 0).
Is wavy up and down from the design shape)
It can be seen that the deviation from the design shape is large and the shape error is large. Usually, when the height (amplitude) of the waveform is several tens of nm, it is estimated that undulation has occurred, and this value is used as the undulation component. Since the height of the waveform in FIG. 10 is about 1000 nm (shape error component), the workpiece shown in FIG. 10 has a shape error component several to several tens times larger than the undulation component. You can see that.

In the stage where the surface accuracy is poor (the shape error component is large) as described above, when observing the interference fringe shape on a monitor, a plurality of interference fringes corresponding to the deviation of the surface accuracy become obstacles, and minute irregularities are formed. It is very difficult to determine the undulation. The same applies to the case where the shape error data obtained from the measuring device is stereoscopically displayed by a computer or the like, for example,
When a shape error component of several hundred nanometers remains on the surface to be processed in a PV (Peak to Valley (Max-Mini)) value,
Since the amplitude of the normal undulation component is several tens of nm, the amplitude of the undulation is one tenth of the shape error component.

Accordingly, from the three-dimensional figure and the shape of the central section,
It is very difficult to determine the main undulation wavelength, and the fact that the main undulation wavelength is not known accurately has been a major problem when selecting the diameter of the polishing tool. For example, if the main undulation wavelength, which is the dominant undulation wavelength to be removed, is 16
mm, it cannot be specified. For example, when polishing is performed using a polishing tool having a tool diameter of 30 mm, it is difficult to effectively remove waviness.

That is, in this case, the polishing tool cannot simultaneously contact at least two or more ridges (convex) of the undulation to be removed, and the polishing work surface is stable with respect to the work surface. It becomes difficult to move. For this reason, it is not possible to selectively process the peak (convex) portion of the undulation to be removed, and it becomes difficult to effectively remove the undulation having the main undulation wavelength.

The present invention has been made to solve such a conventional problem, and a method for evaluating the surface shape of a molded article surface, and a method for easily and reliably removing undulations having a main undulation wavelength. An object is to provide a high-precision smooth polishing method.

[0014]

According to a first aspect of the present invention, there is provided a surface shape evaluation method comprising: measuring a shape of a surface of a molded body; obtaining shape measurement data; It is characterized in that a main swell wavelength is obtained.

According to a second aspect of the present invention, there is provided a high-precision smooth polishing method for polishing a polished surface of a workpiece with a polishing tool and smooth-polishing the polished surface with high precision. After obtaining the shape measurement data, the shape measurement data is subjected to frequency analysis and evaluation, the main undulation wavelength on the polished surface is extracted, and the polished surface is polished with a polishing tool having a diameter of twice or more the main undulation wavelength. Is polished.

A high-precision smooth polishing method according to a third aspect is characterized in that, in the high-precision smooth polishing method according to the second aspect, the polished surface of the workpiece has an aspherical shape. A high-precision smooth polishing method according to a fourth aspect of the present invention is the high-precision smooth polishing method according to the second or third aspect, wherein the shape measurement data is subjected to a high-pass filter processing to extract a swell component which is a frequency range to be removed. After that, a frequency analysis evaluation is performed to extract a main undulation wavelength on the polished surface.

A high-precision smooth polishing method according to a fifth aspect of the present invention is the high-precision smooth polishing method according to any one of the second to fourth aspects, wherein the shape measurement data is subjected to a band-pass filter processing to remove the shape measurement data. After extracting a swell component having a frequency range of, frequency analysis and evaluation are performed, and a main swell wavelength on the polished surface is extracted.

(Function) In the surface shape evaluation method of the first aspect, after measuring the shape of the surface of the molded body and obtaining the shape measurement data, the shape measurement data is subjected to frequency analysis and evaluation, thereby obtaining the main undulation of the surface of the molded body. The wavelength is determined, and the swell is accurately evaluated.

According to the high-precision smooth polishing method of the present invention, the main undulation wavelength on the polished surface is extracted by measuring the precise shape of the polished surface, obtaining the shape measurement data, and performing frequency analysis evaluation of the shape measurement data. Is done.

The “main waviness wavelength” in the present invention refers to the wavelength of a wavelength which is considered to have the largest power spectral density in the shape data of the center section (for example, FIG. 2) and to be greatly involved in the improvement of the shape error. That is. Then, by polishing the polished surface with a polishing tool having a diameter of at least twice the main undulation wavelength, the undulation having the main undulation wavelength is reliably removed.

In the high-precision smooth polishing method according to the third aspect, the polished surface of the workpiece has an aspherical shape, and undulations having a main undulation wavelength existing on the polished surface having the aspherical shape are removed by the polishing tool. You. In the high-precision smooth polishing method according to the fourth aspect, the shape measurement data is subjected to high-pass filtering to remove the shape error component in the low frequency range from the shape measurement data, and only the undulation component in the frequency range to be removed is extracted. You.

In the high-precision smooth polishing method according to the fifth aspect, the shape measurement data is subjected to band-pass filter processing,
Components such as noise, roughness, and flaws are removed from the shape measurement data, and only the undulation component that is the frequency range to be removed is extracted.

[0023]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the high-precision smooth polishing method of the present invention will be described in detail with reference to the drawings.

FIG. 1 shows an embodiment of the high-precision smooth polishing method of the present invention. In this embodiment, as shown in FIG.
The present invention is applied to polishing (1). In this embodiment, first, as shown in FIG.
Is irradiated with laser light 15 from a laser interferometer 13, and interference fringes 17 are generated on the workpiece 11.

The interference fringes 17 are, as shown in FIG.
The image is captured by the CCD image sensor 19 and displayed on a CRT (not shown). It is stored as a digital signal in a microcomputer. In this embodiment, fringe scan interferometry for performing interference measurement by changing the relative distance between the workpiece 11 and the laser interferometer 13 is performed.
The image is captured as an image by the CD image sensor 19, displayed on a CRT (not shown), and stored as a digital signal in the microcomputer.

Next, as shown in (d), image processing is performed in the microcomputer based on the plurality of interference fringes 17 stored as digital signals in the microcomputer, and a three-dimensional figure 21 is generated. Then, in this embodiment, by taking one central cross section of the three-dimensional figure 21, two-dimensional data as shown in FIG.

Next, in this embodiment, as shown in (f), the shape measurement data 23 is
By doing so, a low-frequency shape error component is removed from the shape measurement data 23. Then, as shown in (g), only the swell component which is the frequency range to be removed is extracted, and the high-pass filter processing data 27 is generated.

The high-pass filter processing 25 includes:
This is performed by processing the shape measurement data 23 using a high-pass filter in the same manner as in the analysis of electrical signals and audio signals. In general, there are various definitions of the undulation, but here, empirically, an error component having a wavelength of 20 mm or less is considered to be an undulation, and an averaging filter of 20 mm is employed.

Next, in this embodiment, if necessary,
As shown in (h), bandpass filter processing 29 is performed on the high-pass filter processing data 27 shown in (g). As a result, as shown in (i), components such as noise, roughness, and scratches are removed from the high-pass filter processing data 27,
Only the swell component that is the frequency range to be removed is extracted,
Band-pass filter processing data 31 is generated.

That is, with respect to the shape measurement data 23 obtained from the measuring device having a high resolution, in order to remove noise, roughness, and flaw components, for example,
Bandpass filter processing using a bandpass filter of 20 mm or less and 0.1 mm or more is performed. Next, the band-pass filter processing data 31 thus filtered is subjected to frequency analysis evaluation (FFT analysis) 33 as shown in (j), thereby obtaining frequency analysis evaluation data as shown in (k). 35 is generated.

Then, from the frequency analysis evaluation data 35, a frequency 37 corresponding to the main undulation wavelength on the polished surface 11a is extracted, and the main undulation wavelength is calculated as described later. Next, a polishing tool 39 having a diameter that is at least twice the main undulation wavelength is selected, and as shown in (m), the polished surface 11a is polished by the polishing tool 39, whereby the undulation having the main undulation wavelength is obtained. Is removed.

Example 1 FIG. 2 shows the result of measuring the polished surface 11a of the workpiece 11 with a Fizeau laser interferometer and determining the difference from the design shape (shape error). An example is shown. Further, in the present embodiment, the shape analysis data obtained based on the laser interferometer measurement is subjected to the frequency analysis evaluation without filtering.

In FIG. 2, the vertical axis indicates the shape error component (deviation from the design value), and the horizontal axis indicates the diameter of the workpiece. From this, it can be seen that the undulation on the surface of the workpiece used in this embodiment exists within a range of about ± 10 nm from the design value where the value on the vertical axis is 0. The main undulation wavelength is obtained by performing frequency analysis evaluation of the shape measurement data of the center cross section of the workpiece.

FIG. 3 shows the result of frequency analysis evaluation (FFT analysis) of the shape measurement data of the central section of FIG. Looking at this, the power spectrum density (vertical axis) becomes the largest when the frequency (horizontal axis) is 0.125 cycle / m.
m. From the results of this frequency analysis evaluation,
The frequency of the main undulation wavelength which is the dominant undulation wavelength to be removed is f = 0.125 cycle / mm, and the main undulation wavelength is easily determined to be T = 1 / f = 8 mm. it can.

Therefore, by selecting a polisher diameter that can be brought into contact with a polisher working surface having a wavelength twice or more the main undulation wavelength, undulation polishing can be performed. Here, the main swell wavelength is 8mm
A polisher diameter of 20 mm, which is twice as large as that of the above, was selected, and undulation polishing was performed to successfully remove the undulation. [Second Embodiment] In the second embodiment, the obtained shape measurement data is filtered.

FIG. 4 shows the shape measurement data 23 when the shape error component is larger than the waviness component in the initial polishing step. FIG. 5 shows that a shape error component (low frequency region) is removed from the shape measurement data 23 by applying a high-pass filter processing 29 to the shape measurement data 23 shown in FIG. This is an example of finding.

Referring to FIG. 5, the waviness of the surface of the workpiece used in this embodiment is about ± 10% from the design value where the value on the vertical axis is 0.
It can be seen that it exists within the range of nm. The main undulation wavelength is obtained by performing frequency analysis evaluation of the shape measurement data of the center cross section of the workpiece. There are several theories about the division of the frequency band between the shape error component and the undulation component. Here, the error shape component having a wavelength of 20 mm or less is adopted by performing a high-pass filtering process 25 using a 20 mm filter. Thought swell.

For the filters, they are simply averaged using the effective data points within the filter size,
An averaging filter was used to exchange data at the center point. FIG. 6 shows the high-pass filter processing 25 shown in FIG.
8 shows the result of frequency analysis evaluation (FFT analysis) performed on the data subjected to. Looking at this,
The largest power spectrum density (vertical axis)
It can be seen that the frequency (horizontal axis) is 0.125 cycle / mm.

As in the first embodiment, the frequency 37 of the main undulation wavelength starts from the stage where the shape error component in the initial polishing step is large.
Is f = 0.125 cycle / mm, and the main swell wavelength is T = 1 / f = 8 mm. In Example 2, as in Example 1, a polisher diameter of 20 mm, which is twice or more the main undulation wavelength of 8 mm, was selected, undulation polishing was performed, and undulation was successfully removed.

In the present invention, when the shape error component is large, it is preferable that the shape measurement data is subjected to a high-pass filter or band-pass filter processing. May be subjected to frequency analysis. When performing the filtering process, both of the filtering processes may be performed, or one of the filtering processes may be performed.

[Embodiment 3] In this embodiment, a case in which a workpiece whose polishing surface is an aspheric surface is polished will be described. In addition, polishers that can be used in the present invention will be described.
FIG. 7 shows a undulation polisher which is an example of the above-described polishing tool 39. With this undulation polisher,
A felt is adhered to the bottom surface of the tool plate body 41 as an elastic member 43.

The elastic member 43 has a multilayer structure in which a straight asphalt pitch 45 serving as a polishing work surface is arranged on the bottom surface. In addition, sponge rubber, nylon, etc. may be used for the elastic member 43 other than felt. Further, the tool diameter can be variously selected according to the aspherical amount of the polished surface 11a of the workpiece 11.

In this embodiment, the elastic area of the elastic member 43 arranged below the undulation polisher is the amount of aspherical surface in the polishing surface area within the range of the polisher diameter (the diameter of the polishing tool 39), that is, the reference. The polisher diameter was considered so as not to exceed the deviation of the aspherical shape to be obtained from the spherical shape. Then, the main undulation wavelength was derived in the same manner as in Examples 1 and 2, and a tool diameter of 50 mm which was approximately 6 times the main undulation wavelength of 8 mm was selected.

As described above, by selecting a polisher diameter of 50 mm and a tool diameter approximately six times the main undulation wavelength,
The number of undulations in contact with each polisher increased, and the processing efficiency was improved. In this embodiment, cerium oxide abrasive grains are used for polishing, and the polished surface 11a is used.
The grinding process was performed by making the tool work surface that rotates and the tool work surface that rotates rotate relative to each other.

As described above, in the high-precision smooth polishing method of the present invention, after measuring the precise shape of the polished surface 11a and obtaining the shape measurement data 23, the shape measurement data 23 is subjected to frequency analysis evaluation 33, and the polished surface is evaluated. Since the main undulation wavelength in 11a was extracted and the polishing surface 11a was polished by the polishing tool 39 having a diameter twice or more the main undulation wavelength,
The undulation having the main undulation wavelength can be easily and reliably removed.

In the high-precision smooth polishing method of the present invention,
The undulation having the main undulation wavelength existing on the aspherical polished surface 11a of the workpiece 11 can be reliably removed. Furthermore, in the high-precision smooth polishing method of the second embodiment,
Since the shape measurement data 23 is subjected to the high-pass filter processing 25, it is possible to remove the low-frequency shape error component from the shape measurement data 23, and to reliably extract only the undulation component in the frequency range to be removed.

Accordingly, the main undulation wavelength on the polished surface 11a can be easily extracted by the frequency analysis evaluation 33. Further, in the high-precision smooth polishing method of the present invention, the shape measurement data 23 can be subjected to band-pass filter processing 29, and components such as noise, roughness, and flaws are removed from the shape measurement data 23 to remove the object to be removed. Only the swell component in the frequency range can be reliably extracted.

Therefore, the main swell wavelength on the polished surface 11a can be easily extracted by the frequency analysis evaluation 33. In the above-described embodiment and Example 3,
The example in which the present invention is applied to polishing of the workpiece 11 made of an aspherical lens has been described. However, the present invention is not limited to these examples.
The present invention can also be applied to the workpiece 11 having 1a.

In the above-described embodiments and examples, examples in which the present invention is applied to polishing of the workpiece 11 made of a lens have been described. However, the present invention is not limited to these examples. It can be widely applied to polishing of an optical element such as a mirror or an ultra-precision mold for molding a lens. Further, in the above-described embodiments and examples, the example in which the shape measurement data 23 of the surface to be processed is obtained by the laser interferometer 13 has been described. However, the present invention is not limited thereto. Shape measurement data can be obtained by a container or the like.

[0050]

As described above, according to the surface shape evaluation method of the first aspect, the shape of the surface of the molded body is measured, and the shape measurement data is obtained. Since the main swell wavelength of the body surface is determined, the swell can be accurately evaluated.

In other words, when the surface undulation is observed by a direct observation method as in the prior art, variations are caused by the observer, and the determination of the surface undulation is ambiguous. According to the method, it is possible to accurately evaluate the undulation of the surface without variation among observers. In the high-precision smooth polishing method of claim 2, after measuring the precise shape of the polished surface and obtaining the shape measurement data, the shape measurement data is subjected to frequency analysis and evaluation, and the main undulation wavelength on the polished surface is extracted,
Since the polished surface is polished by a polishing tool having a diameter twice or more the main undulation wavelength, undulations having the main undulation wavelength can be easily and reliably removed.

According to the high-precision smooth polishing method of the third aspect, it is possible to reliably remove the waviness having the main waviness wavelength existing on the aspherical polished surface of the workpiece. In the high-precision smooth polishing method according to the fourth aspect, since the shape measurement data is subjected to the high-pass filter processing, the low-frequency shape error component is removed from the shape measurement data, and only the waviness component in the frequency range to be removed is reliably obtained. Can be extracted.

Therefore, the main undulation wavelength on the polished surface can be easily extracted by the frequency analysis evaluation. In the high-precision smooth polishing method according to claim 5, the shape measurement data is
Since the band-pass filter processing is performed, components such as noise, roughness, and flaws are removed from the shape measurement data, and only the undulation component in the frequency range to be removed can be reliably extracted. Therefore, the main undulation wavelength on the polished surface can be easily extracted by the frequency analysis evaluation.
Further, according to the present invention, work time and work cost can be reduced, and work efficiency can be improved.

[Brief description of the drawings]

FIG. 1 is an explanatory view showing one embodiment of a high-precision smooth polishing method of the present invention.

FIG. 2 is an explanatory diagram showing shape measurement data of a center cross section.

FIG. 3 is an explanatory diagram showing a graph obtained by performing frequency analysis evaluation on the shape measurement data of FIG. 2;

FIG. 4 is an explanatory diagram showing shape measurement data of a center cross section of a workpiece having a large shape error component as compared with a waviness component.

FIG. 5 is an explanatory diagram showing a state where the shape measurement data of FIG. 4 has been subjected to a high-pass filter process.

FIG. 6 is an explanatory diagram showing a graph obtained by performing frequency analysis evaluation on the data of FIG. 5;

FIG. 7 is an explanatory view showing an example of a polishing tool used in the present invention.

FIG. 8 is an explanatory view showing a conventional high-precision smooth polishing method.

FIG. 9 is an explanatory diagram showing shape measurement data of a center cross section.

FIG. 10 is an explanatory diagram showing shape measurement data of a center cross section of a workpiece having a large shape error component compared to a waviness component.

[Explanation of symbols]

 11 Workpiece 11a Polished surface 23 Shape measurement data 25 High pass filter processing 29 Band pass filter processing 33 Frequency analysis evaluation 39 Polishing tool

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G01B 21/30 102 G01B 21/30 102 F term (Reference) 2F065 AA45 AA49 BB05 CC00 CC22 FF01 FF04 FF51 GG04 JJ03 JJ26 QQ23 QQ31 QQ33 QQ34 SS02 SS13 2F069 AA51 BB40 CC10 GG04 GG07 HH30 NN02 QQ10 3C034 BB93 CA22 DD20 3C049 AA06 AA16 BA07 CB01

Claims (5)

[Claims] 1. After measuring the shape of the surface of a molded body and obtaining shape measurement data, the shape measurement data is subjected to frequency analysis and evaluation.
A surface shape evaluation method, wherein a main undulation wavelength of the surface of the molded body is obtained. 2. In a high-precision smooth polishing method for polishing a polished surface of a workpiece with a polishing tool and polishing the polished surface with high precision, a precise shape of the polished surface is measured to obtain shape measurement data. After that, the shape measurement data is subjected to frequency analysis and evaluation, a main undulation wavelength on the polished surface is extracted, and the polished surface is polished with a polishing tool having a diameter twice or more the main undulation wavelength. Precision smooth polishing method. 3. The high-precision smooth polishing method according to claim 2, wherein the polished surface of the workpiece has an aspherical shape. 4. The high-precision smooth polishing method according to claim 2, wherein the shape measurement data is subjected to a high-pass filter processing to extract a swell component which is a frequency range to be removed, and then to perform a frequency analysis evaluation. And extracting a main undulation wavelength on the polished surface. 5. The high-precision smooth polishing method according to claim 2, wherein the shape measurement data is subjected to a band-pass filter processing.
A high-precision smooth polishing method characterized by performing a frequency analysis evaluation after extracting a swell component which is a frequency range to be removed, and extracting a main swell wavelength on the polished surface.