JP5009653B2 - Method and apparatus for measuring surface shape of long body - Google Patents

Description

The present invention relates to a method for measuring a surface shape of a long body and a surface shape measuring apparatus. The present invention is suitable for measuring the surface shape (straightness) in the linear direction of a machined long body on a table of a machine tool.

In numerical control processing machines that require nano-level precision processing, such as photomask quartz glass substrate, stainless steel resin extrusion T-die, zirconia cast coating T-die, etc., long objects such as work tables and tool tables It is required that the surface straightness be 1.0 μm / m or less and the machined workpiece surface straightness be 0.5 μm / m or less. Therefore, the straightness of the table surface and the straightness of the machined workpiece surface in the vertical or horizontal direction are measured using the autocollimator or a plurality of displacement meter sensors to measure the surface shape of the long body (straightness shape 9). Things have been done.

For example, a displacement that has a corner cube for displacement detection and a reflecting mirror for angle detection on the object to be measured, and a position sensor consisting of a laser generator, a solid-state imaging device camera, and light-receiving glass. In addition to an angle measurement device, a position sensor for detecting the displacement and angle of the angle measurement device, a reference ruler that moves together with the object to be measured, and a solid-state image sensor camera for reading the scale Then, the displacement and angle associated with the linear motion of the measured object measured by the displacement detection corner cube and the angle detection reflector are corrected based on the attitude of the displacement / angle measurement device detected by the position sensor. There has been proposed a straightness measuring apparatus characterized by the above (see, for example, Patent Document 1).

Further, the input curve to be measured is x, the output is v, the average sensitivity is Sm, and the linear error is g (x), and the calibration curve is expressed as f (x) = v = Sm · x + g (x). A sensor self-calibration method, wherein each output at a plurality of sampling points obtained by measuring a predetermined calibration range by the displacement sensor is set to vi, and a zeroth-order approximate value x0i≈ of an input value at each sampling point. Using the step of obtaining vi / Sm and the output at the sampling point where the minute change Δx is given to each sampling point as vi +, the difference Δvi = vi + −vi between the outputs of the two points is used, and the linear error g ( a step of obtaining a zeroth-order approximation g′0 (x0i) ≈Δvi / Δx−Sm of a derivative of x) and a zeroth-order approximation g′0 (x0i) of the derivative obtained in this step. Numerical integration to the zeroth order of the linear error g (x) Using the step of obtaining the value g0 (x) = Σg′0 (x0i) Δx and the 0th order approximate value g0 (x0i) of the linear error g (x) obtained in this step, at each sampling point The approximate value of the input value is corrected, the sampling point of the derivative is corrected using the corrected value, and the correction result is numerically integrated to correct the approximate value of the linear error as many times as necessary. There has been proposed an autonomous calibration method of a displacement sensor characterized by comprising steps (see, for example, Patent Document 2).

Also, in the two-point method using two displacement sensors, in order to remove the influence of the relative tilt (pitching) between the displacement sensor probe and the object to be measured during scanning, the pitching in the relative motion is measured and corrected with an autocollimator. 6. Specifically, as shown in FIG. 6, the sensor unit holding the displacement sensor A3 and the displacement sensor B4 is scanned along the X-axis slide 7, and the tilt angle of the sensor unit being scanned is determined by the X-axis slide. 7 is measured by a laser length measuring device comprising a laser length measuring light source 10, a laser length measuring interferometer 11 and a target mirror 12 on the sensor unit side disposed at one end, and each of them is measured by a displacement sensor A3 and a displacement sensor B4. based on the Z-axis direction of the distance data obtained for each sensor interval, the inclination angle of the surface relief of the measuring object W ₁ Mel. At each measurement point for each sensor spacing, angle data by a laser length measuring machine adds to the angle data based on the difference between the outputs of the displacement sensors 3, the shape data representing the surface shape of the measurement object W ₁ from the accumulated A calculation method has been proposed (see, for example, Patent Document 3 and Patent Document 4).

Specifically, in FIG. 6, the height position data of the surface of the measuring object W ₁ by the displacement sensor A at the point Xn moved in the X-axis direction by the sensor interval p between the displacement sensor A and the displacement sensor B is represented by m _A ( Xn), the height position data of the surface of the measuring object W ₁ by the displacement sensor B at the point Xn is m _B (Xn), and the angle data from the laser length measuring interferometer 11 at the point Xn is Lp (Xn).

In this case, the shape data of the measuring object W ₁ at Xm point f (Xm), the Z-axis direction movement error of the X-axis stage at the Xn point E (Xn), Z by pitching the X-axis stage in Xn point When the axial component error is S (Xn) and the measured values of the displacement sensor A and displacement sensor B at the Xm point are used as a reference, m _A (Xm) = f (Xm) (1) m _B (Xm) = f (Xm + p) ··· (2) m A (Xn) = f (Xm + p) + E (Xn) ··· (3) m B (Xn) = f (Xm + 2p) + E (Xn) + S (Xn) (4) E (Xn) = m _B (Xm) −m _A (Xn) (5) {m _A (Xn) and m _B (Xm) are data at the same point} ( 4) substituting (5) into equation when _{m B (Xn) = f (} Xm + 2p) + m B (Xm) -m a (Xn) + S (Xn) ··· (6) f (Xm + 2p) _{= m B (Xn) + m} a (Xn) becomes _{-m B (Xm) -S (Xn} ) ··· (7), pitching error S (Xn) remains.

And angle data Lp at the Xm point (Xm)
From the angle data Lp (Xn) at the point Xn, S (Xn) = p × sin (Lp (Xn) −Lp (Xm)) (8) is obtained. _{(Xm + 2p) = m B} (Xn) + m a (Xn) -m B (Xm) -p × sin (Lp (Xn) -Lp (Xm))) ··· (9) , and the X-axis stage Straightness shape data obtained by removing both the translation error E (Xn) and the pitching error S (Xn) due to the movement can be obtained.

Further, as shown in FIG. 7, the measured object 100 and the displacement sensors 31 to 33 are relatively moved by the sequential three-point method using the three displacement sensors 31 to 33, and the detection outputs of the displacement sensors 31 to 33 are obtained. Based on the sequential three-point method, the surface shape of the DUT 100 is calculated. Based on the detection output of the displacement sensor, the object to be measured 100, 200 and the displacement sensors 41, 42, 32 are relatively moved before and after the object to be measured (reference ruler) 200 is inverted. Then, the surface shape of the DUT 100 by the inversion method is calculated. A method has also been proposed in which a zero error compensation amount is calculated based on the calculation result of the surface shape of the object 100 to be measured by the sequential three-point method and the inversion method, and the zero-point error correction is sequentially performed by the zero-point error compensation amount. (For example, refer to Patent Document 5).

As disclosed in Patent Document 5, the sequential three-point method uses three displacement sensors 31 to 33 to simultaneously measure the object surface shape and the movement error during scanning at a constant interval (displacement sensor interval). This is a method of separating the straightness shape information of the object to be measured and the movement error by processing the displacement output of each displacement sensor.

In FIG. 8, three displacement sensors 31 to 33 that measure the displacement in the z direction with respect to the measurement surface of the DUT 100 are fixed to the sensor head (arm) 30 </ b> A at a distance d. The reference point of the sensor head 30A is the center displacement sensor 32. The measured object has a measured surface shape in the z direction of g (x), the table 25 on which the measured object is placed is moved in the x direction perpendicular to the z direction, and the sensor head 30A is sent by d. The displacement output from each displacement sensor is taken in every time.

In this case, either the sensor head 30A or the table 25 on which the object to be measured is moved may be moved. However, when the table 25 on which the object to be measured is moved, the displacement sensor is opposite to the moving direction. The displacement output is taken in the direction.

At this time, the output from the displacement sensor 31 of the forward reading the information of the surface to be measured above and S _3F, the output from the middle of the displacement sensor 32 and S _30, the output from the rear of the displacement sensor 33 and S _3R To do. The movement of the table 25 usually has a movement error. Among the motion errors, a translational motion error component in the z direction is represented by e _z , and a rotational motion error component (pitching error) about the y axis is represented by eθ.

The table 25 on which the object to be measured is placed is moved, and the output of each displacement sensor when the center displacement sensor 32 serving as a reference comes to the position of a point x _i is as follows. S ₃₀ (x _i ) = g (x _i ) + e _z (x _i ) (10) S _3F (x _i ) = g (x _{i + 1} ) + e _z (x _i ) + deθ (x _i ). _{(11) S 3R (x i} ) = g (x i-1) + and _{e z (x i) -deθ (} x i) ··· (12) the center of the displacement sensor 32, the output of the rear of the displacement sensor 33 When the difference is ΔS _3R (x _i ), the following equation is obtained from the equations (10)-(12).

ΔS _3R (x _i ) = S ₃₀ (x _i ) −S _3R (x _i ) = g (x _i ) −g (x _i−1 ) + deθ (x _i ) (13) g (x _i ) −g (x _i−1 ) is a true shape difference between x _i and x _i−1 , and if this is Δg (x _i−1 ), equation (13) is expressed by the following equation (14) and Become.

ΔS _3R (x _i ) = Δg (x _i−1 ) + deθ (x _i ) (14) On the other hand, the output difference between the center displacement sensor 32 and the front displacement sensor 31 is expressed as ΔS _3F (x _i ). Then, from the formula (11) -the formula (10), the following formula (15) is obtained.

ΔS _3F (x _i ) = S _3F (x _i ) −S ₃₀ (x _i ) = g (x _{i + 1} ) −g (x _i ) + deθ (x _i ) (15) g (x _{i + 1} ) −g Since (x _i ) is a true shape difference between x _i and x _{i + 1} , if this is Δg (x _i ), equation (15) becomes equation (16) below.

ΔS _3F (x _i ) = Δg (x _i ) + deθ (x _i ) (16) Here, when equation (14) is transferred, the rotational motion error differential value around the y-axis is expressed by the following equation (17) It is represented by deθ (x _i ) = ΔS _3R (x _i ) −Δg (x _i−1 ) (17)

By substituting this equation (17) into equation (16), the true shape difference Δg (x _i ) between x _i and x _{i + 1} is expressed by the following equation (18). Δg (x _i ) = ΔS _3F (x _i ) −ΔΔS _3R (x _i ) + Δg (x _i−1 ) (18)

As shown in FIG. 4, the surface shape g (x _i ) at the position of a certain point x _i is the surface shape g (x _i−1 ) at the position (step) of the previous point x _i−1. ) And Δg (x _i-1 ) obtained from the output of each displacement sensor when the center displacement sensor 32 comes to the position of the point x _i-1 . g (x _i ) = g (x _i−1 ) + Δg (x _i−1 ) (19)

The surface shape at the position (step) of the next point x _{i + 1} can be expressed by the following equation (20) from the equation (18) + the equation (19). g (x _{i + 1} ) = g (x _i ) + Δg (x _i ) (20) Therefore, the currently obtained information (18) and the already known information (19) are added. By proceeding, it is possible to obtain information at the next position.

Next, when the surface shape at each position (step) is measured such that i = 1, 2, 3,..., N at the position of a certain point x _i , from the measurement start point to the point x _n How the surface shape g ₃ (x _n ) is sequentially calculated by the three-point method will be described.

(When i = 1) When i = 1, the following is obtained from the equations (19) and (18). g (x ₁ ) = g (x ₀ ) + Δg (x ₀ ) (21) Δg (x ₁ ) = ΔS _3F (x ₁ ) −ΔS _3R (x ₁ ) + Δg (x ₀ ) (...) 22) Here, g (x ₀ ) is a direct current component that is the reference height of the surface to be measured, so it is set as G ₀ , and Δg (x ₀ ) is the _value of x ₁ and x ₀ at the start of measurement. If this is set as ΔG ₀ , Equation (21) and Equation (22) are as follows. g (x ₁ ) = G ₀ + ΔG ₀ (23) Δg (x ₁ ) = ΔS _3F (x ₁ ) −ΔS _3R (x ₁ ) + ΔG ₀ (24)

(When i = 2) When i = 2, the following is obtained from the equations (19) and (18). g (x ₂ ) = g (x ₁ ) + Δg (x ₁ ) (25) Δg (x ₂ ) = ΔS _3F (x ₂ ) −ΔS _3R (x ₂ ) + Δg (x ₁ ) ( 26) Substituting equation (24) into equation (25) yields equation (27) below. g (x ₂ ) = G ₀ + 2ΔG ₀ + {ΔS _3F (x ₁ ) −ΔS _3R (x ₁ )} (27)

Further, when Expression (24) is substituted into Expression (26), the following Expression (28) is obtained. Δg (x ₂ ) = {ΔS _3F (x ₂ ) −ΔS _3R (x ₂ )} + {ΔS _3F (x ₁ ) −ΔS _3R (x ₁ )} + ΔG ₀ (28)

(When i = 3) Also in the case of i = 3, as in the case of i = 2, the equations (19) and (18) are used to relate to g (x ₃ ) and Δg (x ₃ ). The following equations (29) and (30) are obtained by obtaining equations and substituting equations (17) and (18) into the equations. g (x ₃ ) = G ₀ + 3ΔG ₀ +2 {ΔS _3F (x ₁ ) −ΔS _3R (x ₁ )} + {ΔS _3F (x ₂ ) −ΔS _3R (x ₂ ) (29) Δg (x ₃ ) = {ΔS _3F (x ₃ ) −ΔS _3R (x ₃ )} + {ΔS _3F (x ₂ ) −ΔS _3R (x ₂ )} + {ΔS _3F (x ₁ ) −ΔS _3R (x ₁ )} + ΔG ₀ (30)

(When i = n) When i = n, the following equations (1-22) and (1-23) are _{expressed in} terms of g (x _n ) and Δg (x _n ) in the same manner as described above. Obtainable.

従って、式(１−２２)、式(１−２３)からも分かるように、g_３(x_ｎ)，Δg(x_ｎ)は、並運動誤差成分や、ｙ軸まわりの回転運動誤差成分を含む運動誤差を含まず、被測定物の表面形状情報を、変位センサ３１，３２，３３の出力に基づいて逐次求めることができる。

Therefore, as can be seen from the equations (1-22) and (1-23), g ₃ (x _n ) and Δg (x _n ) are the translational motion error component and the rotational motion error component around the y axis. The surface shape information of the object to be measured can be sequentially obtained based on the outputs of the displacement sensors 31, 32, 33 without including the motion error.

Since the invention of Patent Document 6 uses a plurality of displacement sensors in the sequential three-point method described above, a zero error when the position height of the electrode below each displacement sensor of the measurement error factor generated between the displacement sensors is different. The method of performing the correction of the above using the standard ruler 200 is disclosed.

Using the reference ruler and the two-point method, two displacement sensors A and B of the same measurement range, sensitivity, and resolution are used so that the reference displacement meter detects n times the displacement of the displacement sensor to be calibrated. A self-supporting two-point calibration method is also proposed in which the support base of the reference side displacement sensor A is movable and the calibration measurement is performed while increasing the tilt of the lever (see Non-Patent Document 1).

Specifically, when the output of the reference side displacement sensor A reaches the end of the measurement range, the calibrated displacement sensor B has only been inspected for 1 / n of the measurement range. After moving the support so that the output returns to the minimum value, the next 1 / n range is inspected, and this is repeated to inspect the entire measurement range of the calibrated displacement sensor. Next, the arrangement of the displacement sensor A and the displacement sensor B is exchanged, the displacement sensor A is calibrated with reference to the displacement sensor B, and the error from the average sensitivity line included in the first calibration curve is multiplied by 1 / n ² times. Attenuate. By exchanging the arrangement of the displacement sensor A and the displacement sensor B and repeating the calibration procedure k (k is an integer of 1 to 30) times, the error is 1 / n ^k times and is as close to 0 as possible. Is a measurement method using

Japanese Patent No. 2650830 Japanese Patent Laid-Open No. 10-332420 JP 2005-114549 A JP 2005-121370 A JP 2006-337112 A Akira Kiyono, Satoshi Sugisaki, Ken Morishima, “Autonomous Calibration of Linear Error of Displacement Meter”, Journal of Precision Engineering, Vol.59, No.12, p.131-136, published on December 5, 1993

Conventionally, the surface shape was measured on the table of a surface shape measuring instrument installed in a temperature-controlled room, but now the surface is measured on the work table of a machine tool installed in the factory. It is desired to directly measure the surface shape such as straightness of a long machined body (object to be measured) provided with a shape measuring instrument and fixed on a work table. That is, when the straightness of the object to be measured does not satisfy the user's reference value, it is possible to continue precision machining that satisfies the user's reference value at the site. The conventional object is removed from the work table and transferred to a temperature-controlled room where the straightness of the object to be measured is measured, and when the straightness does not meet the user's reference value, the object to be measured is again attached to the machine tool. Compared with the method of continuing precision machining that satisfies the user's reference value by fixing on the work table, there is an advantage that the labor of measurement and the setup work for continuing precision machining can be saved again.

In order to eliminate the influence of the relative inclination (pitching) of the sensor probe and the object under measurement in the two-point method described in Patent Document 3 and Patent Document 4, the pitching in the relative motion is measured and corrected by an autocollimator. The method can guarantee the accuracy of a straight shape of about 10 nm if the length of the object to be measured is within 1 m, but if the length of the object to be measured exceeds 1 m, high-precision measurement by laser length measurement is difficult. Therefore, the length of the object to be measured cannot be used for measuring a long body such as a length of 2 to 10 m.

The five displacement sensors 31, 32, 33, 41 are used to measure the straightness of the object to be measured on the work table of the machine tool by the sequential three-point method using the five displacement sensors described in Patent Document 5. , 42 are simultaneously linearly moved to one end side of the DUT 100 and the standard ruler 200, and the left side surface 200a and the right side surface 200b of the reference ruler 200 need to be reversed. Coordinate alignment is difficult and causes measurement errors.

The first object of the present invention is that the straightness shape of an object having a length exceeding 1 m in the method using the sequential two-point method and the autocollimator described in Patent Document 3 and Patent Document 4 cannot be measured with high accuracy. In the operating range where the autocollimator can measure with high accuracy (for example, within 1 m), the pitching is corrected using the value measured by the autocollimator. For measuring the straightness of the longer part, another third displacement sensor is added to the top of the linear extension where two first and second displacement sensors are aligned, and there is concern about the accuracy of pitching measurement by the autocollimator. Make a measurement standard with this third displacement sensor in the range exceeding 1m, and expand the measurement range one after another, correct the straightness from the measured value of these three displacement sensors, It is to provide a method for obtaining a straightness curve of high accuracy's straightness curve coalesce.

According to a second object of the present invention, when the pitch distance p of each of the three displacement sensors is as large as 5000 mm or 1,000 mm, the buyer of the processed object to be measured “the measurement point density is low. May be pointed out. Therefore, the fourth displacement sensor is arranged between the pitches p of the first and second displacement sensors, and the pitch distance between the first and fourth displacement sensors is p / n (where p / n is an integer of 2 to 30). And a method of deriving a straightness curve in which the density of measurement points is increased by interpolating the straightness values measured by the fourth displacement sensor.

The third object of the present invention is to use a standard straight ruler with a known straightness shape in advance and the three displacement sensors instead of using the pitching correction by the autocollimator, and to know the straightness shape of the object to be measured. Using the straightness value in the range, the pitching and translation in the scanning motion are measured separately by the first and second displacement sensors, and the effects of the pitching and translation error components included in the output of the preceding third displacement sensor are corrected. It is to provide a method for deriving a straightness curve.

A fourth object of the present invention is to provide a straightness curve of an object to be measured that is fixed on a work table movable in a linear direction of a machine tool. An object of the present invention is to provide a surface shape measuring apparatus including a ruler, a linear scale, input / output means, measurement control means, straightness calculation means, storage means, and recording means.

The fifth object of the present invention is to measure the straightness shape value of the reference ruler that is worn due to the long-term use of a standard straight ruler whose straightness shape is known, and the straightness shape of the surface has changed to a minute amount, It is to provide a method for correcting the straightness shape value of a new standard straight ruler.

In the first aspect of the invention, the three displacement sensors A, B, and C are fixed to the arm in series at every pitch interval p, and the straightness of the object to be measured fixed on the table is measured in the direction of the arm or the table linearly. In the method of measuring the straightness of an object to be measured by movement, the operating range in which the autocollimator can measure with high accuracy is detected by the two displacement sensors A and B using the values measured by the autocollimator. In addition, the straightness value data string is obtained by correcting the straightness value pitching, and in the measurement of the straightness of the measured object in the operating range where the autocollimator cannot measure with high accuracy, 1) the two first displacements An output detected by a third displacement sensor C provided at a pitch interval p between the second displacement sensor B and the second displacement sensor B at the upper end of the linear extension in which the sensor A and the second displacement sensor B are arranged, and the first displacement sensor A Force is correlated with the output of the second displacement sensor B, the initial straightness value data string already obtained is regarded as a measurement standard, and the last two points obtained in the straightness value data string 2) to obtain a straightness value output at points on the extension of the initial straightness value data string and separated by the pitch p. 2) The newly obtained straightness value output is also obtained as the initial straightness value. Using the extended straightness value data string added to the value data string as a new measurement standard, the pitching correction for adding the straightness value output at a point further away by the next pitch p to the extended straightness value data string is repeatedly measured one after another. The present invention provides a method for measuring the surface shape of a long body, characterized in that an extended straightness value data string is obtained by expanding the range and outputting a straightness curve.

According to the first aspect of the present invention, in the measurement of the straightness in the operating range where the autocollimator cannot measure with high accuracy, the measurement reference is made by the straightness value detected by the third displacement sensor C, and the measurement of this measurement reference is made. Pitching is detected from the output of the two-point method by the displacement sensors A and B with reference to the value of the point, and while using it to correct the pitching, the measurement range is repeatedly expanded to obtain an extended straightness value data string. Since the straightness curve is output, the straightness curve can be derived with high accuracy even if the object to be measured is a long body exceeding 1 m.

In the invention of claim 2, three displacement sensors A, B, and C are fixed to the arm in series at every pitch interval p, and the fourth displacement sensor is disposed between the first and second displacement sensors. sensors and the pitch distance of the fourth displacement sensor of p / n (where, the value of p / n is an integer of 2 to 30.) was inserted in the position where the, the four displacement sensors a, B , C, D in series, or a method of measuring the straightness of the object to be measured fixed on the table by moving the table for fixing the object to be measured.
In the operating range where the autocollimator can measure with high accuracy, the initial straightness is corrected by correcting the pitching of the straightness values detected by the two-point method by the two displacement sensors A and B using the values measured by the autocollimator. In the measurement of the straightness of the object to be measured in the operating range in which the autocollimator cannot measure with high accuracy by obtaining a value data string, 1) The linearly extended upper end where the two first displacement sensors A and the second displacement sensor B are arranged The second displacement sensor B is associated with the output detected by the third displacement sensor C provided at the pitch interval p, the output of the first displacement sensor A, and the output of the second displacement sensor B. The obtained initial straightness value data string is regarded as a measurement standard, and is synthesized with the two output values obtained at the end of the straightness value data string, thereby obtaining the initial straightness value data string. On extension The straightness value output of points separated by the pitch p is obtained. 2) The newly obtained straightness value output is also added to the initial straightness value data row, and the extended straightness value data row is further used as a new measurement standard. Pitching correction for adding the straightness value output at the point separated by the next pitch p to the extended straightness value data sequence is repeated one after another to widen the measurement range to obtain an extended straightness value data sequence. 3) The straightness curve The straightness value data for each p / n between the pitch intervals p is the straightness output detected by the two-point method of the pitch interval p / n by the first displacement sensor A and the fourth displacement sensor D as the autocollimator. 4) Output the straightness value data string as a straightness curve by interpolating into the initial straightness value data string or the extended straightness value data string obtained in step 4 and increasing the measurement density. Of the long body It is to provide a surface shape measuring method.

According to the second aspect of the present invention, the coordinate positions 2p, 3p, 4p,..., (L-3p), (L-3p) of the object to be measured having the length L detected by the three displacement sensors A, B, C L-2p) is corrected based on the straightness value detected by the fourth displacement sensor D, and an error of the straightness value at the coordinate position is subtracted as a correction value from (L-2p) inter-coordinate position ( Since interpolation is performed as a straightness correction value at (L−2p−p / n) from (p + p / n), a straightness curve with an increased measurement point density can be derived.

The invention of claim 3 uses a standard straight ruler whose straightness shape is known in advance, and three displacement sensors A, B, C are fixed to the arm in series at every pitch interval p, and are fixed on the table. In the method of measuring the straightness of the object to be measured by moving the arm or the table in the linear direction, the straightness shape of the object to be measured is used in the scanning motion by using the straightness value in a known range. A method in which pitching and translation are separately measured by the first and second displacement sensors A and B, and the effects of pitching and translation error components included in the output of the preceding third displacement sensor C are corrected to output a straightness curve. Is to provide.

According to the invention of claim 3, a high-precision reference straight ruler is fixed to the table, the straightness of the reference ruler is detected by the three displacement sensor probes A, B and C, and the straightness of the known reference ruler. The pitching and translation in the scanning motion are separated by the first and second displacement sensors A and B from the error, and the effects of the pitching and translation error components included in the output of the preceding third displacement sensor C are compared. to correct. Even without using an autocollimator, it is possible to measure the straightness shape of the object to be measured on the work table of the machine tool.

The invention according to claim 4 is a work table movable in a linear direction on which an object to be measured can be placed, and three displacement sensors A, B, and C are connected in series at a pitch interval p via a tool head or arm. Three displacement sensors A, B, and C fixed to the head, a standard straight ruler with a known straightness shape, linear scale, input / output means, measurement control means, straightness calculation means, storage means, and recording means It is an object of the present invention to provide an apparatus for measuring the surface shape of an object fixed on a work table.

According to the invention of claim 4, the straightness shape of the object to be measured fixed on a work table such as a grinding device, a lapping device, or a polishing device can be output by carrying out the measuring method of claim 1. it can.

According to the invention of claim 5, the straightness value of a reference straight ruler whose straightness shape is known is detected by a plurality of displacement sensors, the straightness value data string detected and the straightness of the known reference straight ruler. The value data strings are compared, and when the difference between both exceeds 10 nm, the detected straightness value data string is recorded as a new straightness shape value of the reference straight ruler. The surface shape measuring method of the standard straight ruler according to Item 2 is provided.

According to the invention of claim 5, the degree of wear of the surface of the standard straight ruler whose straightness shape is known can be measured, and the correction time of the straightness shape value of the stored standard straight ruler can be determined. Further, by correcting the straightness shape value of the reference straight ruler, the measurement error of the straightness shape of the object to be measured newly becomes small.

Hereinafter, the present invention will be described in more detail with reference to the drawings. FIG. 1 is a perspective view used for explaining the autocollimator-based displacement two-point method and stitching, FIG. 2 is a diagram used for sequentially explaining three-point method steps, and FIG.
FIG. 4 is a diagram showing a two-point displacement method based on an autocollimator and interpolation stitching, FIG. 4 is a diagram showing straightness shape at each position coordinate in stitching, and FIG. 5 is provided with a workpiece surface shape measuring device. It is a top view which shows the principal part of a surface grinding machine.

A numerical control (NC) surface grinding machine 300 shown in FIG. 5 includes a variable hydraulic hydrostatic bearing linear motion carriage (work table and tool table) disclosed as FIG. 1 of JP-A-2006-102867. It is a grinding machine. In the figure, w is a work, 303 is a grinding wheel, and 304 is a work table that can be reciprocated in the left-right direction. 305 is a bed, 306 is a saddle that can reciprocate in the front-rear direction, 307 is a column, 308 is an operation panel, 309 is a grinding wheel head, and 310 is an electromagnetic chuck.

An arm 311 for fixing three non-contact capacitive displacement meters (displacement sensor probes) A, B, and C at equal intervals in the left-right linear direction to the grinding wheel head (tool head) 309 of the numerically controlled surface grinding machine 300. And the arm 312 which fixes the three non-contact capacitance type displacement meters A, B, and C at equal intervals in the front-rear straight line direction is attached. These non-contact capacitance type displacement meters A, B, and C are attached to the side wall of the grinding wheel protective cover 320 by an arm 311 and an arm 312 for fixing the other non-contact capacitance type displacement meters A, B, and C. Is attached to the other side wall of the grinding wheel protection cover 320. Further, the standard straight rulers 200X and 200Y are fixed on the work table 304 or on the side surfaces, and the linear scales 330X and 330Y are fixed on the work table 304 or on the side surfaces.

The surface shape measuring apparatus 2 includes the work table 304, the arms 311 and 312 in which the three displacement sensor probes A, B, and C are fixed in series at every pitch interval p, and the standard straight rulers 200X and 200Y with known straightness shapes. , Linear scales 330X and 330Y, input / output means 313, storage means 314, measurement control means 315, straightness calculation means 316, and recording means 317. The non-contact capacitive displacement meters A, B, and C have a resolution higher than 10 nm, and preferably have the same measurement range, sensitivity, and resolution.

As non-contact type capacitance displacement meters, there are Microsense 5000 series (trade name) manufactured by Nippon D.I. Co., Ltd., and capacitance displacement meter VE-521 (trade name) manufactured by Ono Kousen Co., Ltd. ), AccuMeasure (trade name) of Koyo Seisakusho Co., Ltd., ST-3571A (trade name) of Iwatori Measurement Co., Ltd., ATS series, ATM series (trade name) of Techno System Co., Ltd., etc. can be used. Instead of the non-contact type capacitive displacement meter, a non-contact type laser beam displacement meter may be used.

As a material of the standard straight ruler, single crystal quartz is generally used. Recently, things like zirconia standard straight rulers are on the market. Therefore, you may use the same kind of material as the workpiece processed. Of course, the material of the straight ruler 200 is selected to be close to the coefficient of thermal expansion of the workpiece material. However, since the resolution of the displacement sensor is 10 nm or less and the straightness required by the user is 1.0 μm / 1 m or 0.5 μm / 1 m, the straightness of the standard straight ruler is selected to be 10 nm or less. To do.

The movement of the lower head portion of the displacement sensor probe above the standard straight ruler 200X fixed on the work table 304 is driven by a driving means for moving the saddle 306 mounted with the grindstone shaft to which the arms 311 and 312 are mounted back and forth. To align. The movement of the lower head portion of the displacement sensor probe above the reference straight ruler 200Y fixed on the work table 304 is driven by the driving means for moving the work table 304 to the left and right for alignment.

For machining the workpiece w fixed on the work table 304 via the electromagnetic chuck 310, the column 307 that supports the grinding wheel head 309 including the grinding wheel 303 that rotates and the horizontal movement of the work table 304 is fixed. The saddle (tool table) 306 is moved in the front-rear direction and the wheel head 309 is moved in the up-down direction.

The non-contact type capacitance displacement meter forms a capacitance (C) between a sensor electrode having an area S of the displacement sensor probe and a measured object (distance d), and the distance from the sensor to the measured object is measured. It is an instrument that measures the displacement of the surface of the object to be measured by changing the capacitance according to the change (C = e · S / d). Note that e in the equation represents a dielectric constant.

Next, a method for measuring the straightness shape of the object to be measured according to claim 1 of the present invention will be described with reference to FIGS. 1 to 3 are diagrams for explaining stitching in the two-point displacement method based on the autocollimator. A reflecting mirror (target mirror) 400 is attached to the left end of the arm 311 in which the three displacement sensor probes A, B, and C are fixed in series at every pitch interval p, and an autocollimator 500 is installed on the left rear side. ing. The pitch distance between the probes A and B is p, and the pitch distance d between the probes A and C is 2p.

In the operating range in which the autocollimator can measure with high accuracy, the value measured by the autocollimator 500 as shown in FIG. 1 is used, and the straightness detected by the two-point method by the two displacement sensors A and B shown in FIG. An initial straightness value data string is obtained by correcting pitch value pitching.

In FIG. 1, the measurement start point of the displacement sensor probe A is X = 0, the distance between the probe lower part of the displacement sensor A and the surface of the object w is f (X ₀ ) = z ₀ , and the measurement point of the displacement sensor A probe is X _i, and the distance between the probe lower and the measured object w surface at the measurement points X _i and _{f (X i) = z i} , the measurement starting point of the displacement sensor B is X = p, the measurement starting point of the displacement sensor C Becomes X = d = 2p.

E _{z (X)} a translation error of the displacement sensor A, e p _(X) pitching error, S _B is zero point offset of the displacement sensor B relative to the displacement sensor A (the inclination of the zero point of the autocollimator with respect to the optical axis), Let _SC be the zero point offset of displacement sensor C with reference to displacement sensor A. The outputs m _A (x), m _B (x), and m _C (x) of the displacement sensors A, B, and C are expressed by the following equations.

_{m A (X) = f (} X) + e z (X) ··· (31) m B (X) = f (X + p) + e z (X) + e p (X) · p + S B ··· ( _{32) m C (X) =} f (X + 2p) + e z (X) + e p (X) · 2p + S C ··· (33)

The differential output of the displacement sensors A and B is the following equation (34), the differential output of the displacement sensors A and C is the following equation (35), and the differential output of the displacement sensors B and C is the following equation ( 36). _{m B (X) -m A (} X) = f (X + p) -f (X) + e p (X) · p + S B ··· (34) m C (X) -m A (X) = f (X + 2p) -f (X ) + e p (X) · 2p + S C ··· (35) m C (X) -m B (X) = f (X + 2p) -f (X + p) + e _p (X) · 2p + S _C -S _B (36)

Straightness shape of the workpiece w at a distance within 1m from the left edge of the object w (d + w) is the auto-collimator 500 and the displacement sensor A, measured by B, as pitching error e _{p (X)} is corrected, sequential straightness shape f the equation (33) (X + p) + S B and f (X + 2p) + Request S _B. f (X + p) = m B (X p) + m A (X p) -m B (X 0) -psin (S B (X p -S B (X 0)) ··· (37) f ( _{X + 2p) = m B (} X 2p) + m A (X 2p) -m B (X p) -psin (S B (X 2p -S B (X p)) ··· (38)

The above values including zero point offset S _B. That, S optical axis displacement sensor A of autocollimator the effect of _B, the slope of zero line B are included in the straightness shape.

Thus, the auto that three displacement sensors A shown in FIG. 2, B, Step 1 of the process steps of the three points sequential procedure using C _{(k i)} and (k ₁₎ pitching in the relative movement in Step 2 _{(k 2)} Since it can be measured and corrected by the collimator, the influence of the accidental error is ignored, and attention is paid to the fact that each of the displacement sensors A, B, and C is first used redundantly at the X coordinate 2p. When the influence of the pitching of the displacement sensor B is corrected, the following equation (39) is obtained. m _B (2p) −m _A (0) = f (X _2p ) −f (X ₀ ) + 2p · S _B (39)

On the other hand, the following equation (40) is obtained from the differential output using the displacement sensor A and the displacement sensor C. m _C (0) −m _A (0) = f (X _2p ) −f (X ₀ ) + S _C (40)

The following equation (41) is derived from the equations (39) and (40), and it is understood that this is equivalent to the zeros of the displacement sensors A, B, and C being aligned. m _C (0) −m _B (2p) = S _C −2p · S _B (41)

Therefore, the inclination of the shape obtained by the sequential two-point method using displacement sensor A and displacement sensor B {m _B (2p) −m _A (0)} / 2p = {f (X _2p ) −f (X ₀ ) + 2p · S _B } / 2p (42) is the inclination of the zero line of displacement sensor A and displacement sensor C {m _C (0) −m _A (0)} / 2p = {f (X _2p ) −f (X ₀ ) + S _C } / 2p... (43)

Expression (44) is derived from Expression (42) and Expression (43). {F (X _2p ) −f (X ₀ ) + 2p · S _B } / 2p = {f (X _2p ) −f (X ₀ ) + S _C } / 2p (44) Therefore, autocollimator, displacement sensor a, if the output (measurement), correction of the displacement sensor B is S _B known, S _C is capable of pitching correction in step 3 following shape measurement results, the output of the displacement sensor C, S _C a correction Calculated.

In the measurement of the straightness of the object to be measured in the operating range where the autocollimator 500 cannot measure with high accuracy, 1) the second displacement sensor is located at the top of the linear extension lined up with the two first displacement sensors A and B. B is related to the output detected by the third displacement sensor C provided at the pitch interval p, the output of the first displacement sensor A, and the output of the second displacement sensor B. The initial straightness value data string is regarded as a measurement standard, and is combined with the last two output values obtained in the straightness value data string, thereby extending the initial straightness value data string and the pitch p. 2) The newly obtained straightness value output is also added to the initial straightness value data row as the new straightness value data row and the new straightness value data row is used as a new measurement standard. Straightness of points separated by pitch p The resulting extended straightness value data string to expand the measuring range is repeated one after another pitching correction applied to the extension straightness value data string, and outputs the straightness curve.

After step 3 of pitching correction based on the result of the straightness shape measurement shown in FIG. 2, the straightness shape is also measured by the displacement sensors B and C at the measurement points (3p, 4p, 5p...) Of the displacement sensor A. . After step k is step 3, the known f (X ₀ ), f (X _p ), f (X _2p ),..., F (X _{(k−1) p} ) are converted into two displacement sensors. Measure with A and B, calculate and correct the pitching, and make the corrected pitching known as f (X _2p ), f (X _3p ), f (X _4p ), the straightness shape can be measured without using an autocollimator in the measurement step of step 3 (distance 2p or more) or more. Therefore, even if the length of the object to be measured exceeds 1 m, for example, a 10 m ground workpiece, the straightness shape can be measured with an error accuracy of 10 nm or less.

Therefore, from the results derived from the above-described equation groups output and corrected by the displacement sensors A, B, C and the autocollimator, Δf (X) = {f (X _kp ) −f using the displacement sensor B as a reference sensor. (X _{(k−1) p} )} (45) By integrating and outputting the value of (45), the straightness shape between the distance of the left end p of the object to be measured and the distance of (k−1) · p is expressed as p. Can be shot on recording paper at intervals, or displayed on a personal computer display screen.

Next, the accumulation of the chance error σ in stitching in the two-point displacement method based on the autocollimator in FIG. 1 will be discussed below.

Step 1 (two-point method with pitching corrected by autocollimator) Accidental error of f (X _P ) at position coordinate p σ ₁₁ ² = σ ₁₀ ² + (2σ _s ² + p ² σ _A ² ) = (2σ _s ² + P ² σ _A ² ) Accidental error of Δf (0) = f (Xp) −f (X ₀ ) σ _Δ0 ² = (2σ _s ² + p ² σ _A ² ) Random of f (X _2p ) at position coordinate 2p Error σ ₁₂ ² = σ ₁₁ ² + (2σ _s ² + p ² σ _A ² )

Step 2 (The shape of the first step is transferred by a two-point method in which pitching is corrected.) The coincidence error of Δf (p) = f (X _2p ) −f (X _p ) is expressed by the following equation. σ _Δ1 ² = 2σ _s ² + p ² σ _A ² The accidental error of f (X _2p ) at the position coordinate 2p is expressed by the following equation. The coincidence error of σ ₂₁ ² = σ ₁₁ ² + (2σ _s ² + σ _Δ1 ² ) Δf (2p) = f (X _2p ) −f (X _p ) is expressed by the following equation. σ _Δ2 ² = σ _Δ1 ² + 2σ _s ² The coincidence error of f (X _3p ) at the position coordinate 3p is expressed by the following equation. ^{_{σ 22 2 = σ 21 2 +}} (2σ s 2 + σ Δ2 2)

After step 3 (pitching is measured by the two-point method using a known shape one step before step k.) Accidental error of Δf (kp) = f (X _kp ) −f (X _kp−1 ) Is expressed by the following equation. σ _Δk ² = 2σ _s ² + σ _{Δ (k−1} ) The accidental error of f (X _kp ) at the ^2- position coordinate kp is expressed by the following equation. The coincidence error of σ _k1 ² = σ _{(k−1) 2} ² + σ _Δk ² Δf (kp + p) = f (X _{(k + 1) p} ) −f (X _kp ) is expressed by the following equation. σ _{Δ (k + 1)} ² = σ _Δk ² + 2σ _s ² The coincidence error of f (X _{(k + 1) p} ) at the position coordinate (k + 1) p is expressed by the following equation. σ _k2 ² = σ _k1 ² + σ _{Δ (k + 1)} ² )

As a combination of random error in the autocollimator and the displacement sensor (displacement meter), if the _{^{σ Ad 2 = 2σ s 2 =}} p 2 σ A 2, σ 11 2 = 2σ Ad 2, σ 12 2 = 4σ Ad 2, σ Δk _{^{2 = σ Δ (k-1}} ) 2 + kσ Ad 2, a _σ Δ0 ² = σ _Ad ^2.

Assuming k> 1 (where the starting value of k is 2) σ _k2 ² = σ _{(k−1) 2} ² + 2σ _Δk ² + σ _Ad ² = σ _{(k−1) 2} ² +2 (k + 1) σ _Ad ² + σ _Ad ² = σ ₁₂ ² + Σ (2k−1) σ _Ad ²

_{^{Therefore, σ 22 2 = 7σ Ad 2}} , σ 32 2 = 12σ Ad 2, the σ ₄₂ ² = 19σ _Ad ^2.

When an object to be measured having a length of 3 m, 4 m, and 5 m is sequentially measured by a two-point method using an autocollimator having a working distance of 10 nm accuracy of 1 m, the uncertainties are about 26 nm, 35 nm, and 45 nm, respectively. At 10 m, the uncertainty is less than 100 nm. If the uncertainty is less than 100 nm = 0.1 μm, it sufficiently satisfies the user's desire for a straightness shape measurement error of less than 0.5 μm, and these accidental errors are ignored (treated as σ = 0). It is understood that it does not matter.

The evaluation error of the zero difference (S _C -S _B ) is a parabolic error. From Equation (36) and Equation (37), if the accidental error included in the differential outputs of S _C and S _B is omitted in Step 1 because the error in reading the autocollimator is common, S _C = because ^{_{(p 2 σ a 2 + 4σ}} s 2) and may be evaluated as _S B = _{2σ s} ^2, zero point error due to parabola error is as follows. (P ² σ _A ² + 6σ _s ² ) = 4σ _Ad ²

When the length of the object to be measured is 10 m and the pitch interval p between the displacement sensors A, B, and C is 1 m, the measurement point density can be obtained even if the overall straightness is known with only nine straightness shapes. There may be users who are too low and anxious. In that case, as shown in FIG. 3, the displacement sensor D is added, the interpolation point reference | standard synthesis method is implemented, and the density of a measurement point is raised.

In the interpolation point reference synthesis method, the probes of the three displacement sensors A, B, and C are fixed to the arm in series at every pitch interval p, and between the probes of the first displacement sensor A and the second displacement sensor B, The probe of the fourth displacement sensor D is inserted at a position where the pitch distance between the first displacement sensor A and the fourth displacement sensor D is p / n (however, the value of p / n is an integer of 2 to 30). The four displacement sensor probes A, D, B, and C are fixed in series, or the table for fixing the object to be measured is moved to measure the straightness of the object to be measured fixed on the table. Is the method.

In the operating range in which the autocollimator can measure with high accuracy of 10 nm, the pitching of the straightness value detected by the two-point method is corrected by the two displacement sensors A and B using the value measured by the autocollimator 500. An initial straightness value data string is obtained.

In measuring the straightness of the object to be measured in the operating range where the autocollimator cannot measure with high accuracy, as described above, 1) on the linear extension in which the two first displacement sensors A and the second displacement sensor B are arranged. First, the second displacement sensor B associates the output detected by the third displacement sensor C provided at the pitch interval p, the output of the first displacement sensor A, and the output of the second displacement sensor B, The initial straightness value data string is regarded as a measurement standard, and the initial straightness value data string is synthesized with the last two output values in the straightness value data string. 2) The straightness value output of the point separated by the pitch p is obtained, and 2) The newly obtained straightness value output is also added to the initial straightness value data row to obtain a new straightness value data row. As a measurement standard, it is further separated by the next pitch p. Obtaining an extension straightness value data string straight degree value output to expand the measuring range is repeated one after another pitching correction applied to the extension straightness binary data string.

3) Straightness value data for each p / n between the pitch intervals p of the straightness curve is obtained by the two-point method of the pitch interval p / n by the first displacement sensor A and the fourth displacement sensor D. The detected straightness output is interpolated into the initial straightness value data string or the extended straightness value data string obtained by the autocollimator to obtain a straightness value data string with an increased measurement density. 4) These straightness values The value data string is output as a straightness curve.

For example, the pitch interval p between the displacement sensor probe A and the displacement sensor B probe, the pitch interval 0.1p between the displacement sensor A probe and the displacement sensor D probe, and the pitch interval 0.9p between the displacement sensor B probe and the displacement sensor D probe. Then, the pitch interval of the sensor probe is changed and the above-described two-point method is used to move the table or the arm every pitch interval of 0.1 p to obtain a corrected value 2 for the signals output from the displacement sensors A, D, B. A straightness curve with a higher measurement density is output by the point method.

In order to measure the workpiece w on the work table 304 of the NC surface grinding machine 300 equipped with the surface shape measuring apparatus shown in FIG. 5, three non-contact capacitive displacement meters (displacement sensors) are mounted on the grinding wheel head 309. Arms 311 for fixing the probes A, B, and C at equal pitches p in the left and right linear directions and three non-contact capacitance displacement meters A, B, and C are fixed at equal intervals in the front and rear linear directions. The arm 312 is attached, and these non-contact capacitive displacement meters A, B, C are attached to the side wall of the grinding wheel protective cover 320 by the arm 311 and the other non-contact capacitive displacement meters A, B, C. Is fixed to the other side wall of the grinding wheel protection cover 320. Further, the reference straight rulers 200X and 200Y are fixed on or on the side of the work table 304, and the work table 304 is fixed. Or fixing the linear scale 330X and 330Y to the side.

The detection of the straightness shape f (X) in the width direction (front-rear direction: Y-axis direction) of the workpiece is performed by using the non-contact type capacitance displacement meters A, B, and C subjected to autonomous calibration as the grinding wheel protective cover 320. The displacement gauges A, B, and C are supported on the side wall of the arm 312 so that they are aligned in the front-rear direction, and the displacement heights of the workpiece are detected using these displacement sensors A, B, and C. Similar to the process, detection is performed by moving the work table 304 or the tool table 306 on which the grinding wheel head 309 is mounted. The measurement of the height displacement in the front-rear direction (Y-axis direction) of the work piece (which may be a reference straight ruler) uses a reference straight ruler 200Y placed on the work table 304 in the Y-axis direction. As described, the autocollimator 500, the displacement sensor A, and the displacement sensor B are used for the measurement within 1 m in the length direction of the workpiece or the first (X = 1) and second (X = 2p) measurement points. The measurement is performed by the point method, and the part beyond 1 m in the length direction of the workpiece that cannot be measured with high accuracy by the autocollimator 500, or the measurement after the third measurement point (X = 3p) is performed sequentially using displacement sensors A, B, and C. A method of correcting the measured value is performed by a point method or an interpolation method using the displacement sensors A, D, and B.

Grinding of a new workpiece is performed by removing the displacement sensors A, B, C and the standard straight ruler from the work table, placing the new workpiece on the work table, and rotating the grinding wheel 303 of the numerically controlled surface grinding machine. Grind the workpiece. After grinding, the displacement sensors A, B, and C are attached to the grinding wheel protective cover 320 and the reference straight ruler is attached to the work table 304, and the above-described displacement height measuring method of the workpiece to be ground (measured object) is measured. The straightness shape displacement function f (X) is projected on the personal computer screen, or the straightness shape is stamped on the recording paper. On the grinding machine 300, it is preferable to move the work table 304 and the tool table 306 without moving the displacement sensors A, B, C and the autocollimator 500.

Examples of the material for the workpiece include stainless steel, metals such as brass and hastelloy, ceramics such as zirconia, alumina, and silicon nitride, glass, quartz, meteorite, and sapphire substrate.

If the straightness shape f ′ (X) of the reference straight ruler 200X, 200Y is measured in advance using the displacement sensors A, B, C or the displacement sensors A, D, B, C, the straightness shape f '(X) value is adopted as the standard straightness shape value of the standard straight ruler, and the straightness shape f () of the object measured using the displacement sensors A, B, C or the displacement sensors A, D, B, C X) By subtracting from the value, the displacement sensors A and B or the pitching and translation in the scanning motion of the object to be measured are separately measured, and the pitching and translation included in the output of the preceding third displacement sensor C That is, the straightness shape graph is output after correcting the influence of the error component. Therefore, the straightness shape {f (X) −f ′ (X)} of the object to be measured can be displayed on the personal computer screen or output on the recording paper without using the autocollimator 500.

The reference straight rulers 200X and 200Y sometimes wear the ruler surface exceeding 10 nm while being used for straightness shape measurement for a long time. Therefore, from time to time, the straightness shape f of the reference straight ruler using the displacement sensors A, B, and C or the displacement sensors A, D, B, and C measured the reference straightness shape f ′ (X) of the reference straight ruler. “When (X) is measured and there is a part where the difference between the straightness shape f ′ (X) measured last time and the straightness shape f” (X) of the standard straightness ruler measured this time exceeds 10 nm, It is judged that it cannot be adopted as a standard straight ruler due to progress of wear, and is replaced with a new standard straight ruler, or the surface of this worn standard straight ruler is ground or polished to become a standard straight ruler showing a straightness shape of less than 10 nm. Let it play.

The straightness shape of the regenerated standard straight ruler or the replaced standard straight ruler is measured using the displacement sensors A, B, C or the displacement sensors A, D, B, C, and a new standard straightness of the standard straight ruler. If the degree shape f ′ (X) value is stored in the recording unit of the personal computer, the straightness shape measurement of the object to be measured on the work table can be performed without an autocollimator. Further, it can be used as comparison data for measuring the degree of wear of the standard straight ruler at a later date.

It became possible to measure the straightness shape of a long body exceeding 1 m in the operating range exceeding the accuracy of 10 nm of the autocollimator. In addition, it is possible to determine the replacement time by measuring the degree of wear of the standard straight ruler.

It is a perspective view used in order to explain a two-point displacement method and stitching based on an autocollimator. It is a figure used in order to demonstrate a sequential 3 point method step. It is a figure which shows the two-point displacement method and interpolation stitching based on an autocollimator. It is a figure which shows the straightness shape of the to-be-measured object in each position of the X coordinate in stitching. (Known) It is a top view which shows the principal part of a surface grinding machine provided with the surface shape measuring apparatus of a workpiece. It is a perspective view which shows the surface shape measuring apparatus of the to-be-measured object by the sequential two-point method using an autocollimator. (Known) It is a perspective view which shows the surface shape measuring apparatus of the to-be-measured object by a reference | standard ruler and a sequential 3 point method. (Known) It is a figure explaining the surface shape measuring method of the to-be-measured object by a sequential 3 point method. (Known)

Explanation of symbols

w Object to be Measured A Capacitance displacement sensor probe B Capacitance displacement sensor probe C Capacitance displacement sensor probe D Capacitance displacement sensor probe 200X Standard straight ruler 200Y Standard Straight ruler 300 Numerically controlled surface grinding machine 304 Work table 306 Tool table 311 Arm 312 Arm 330X Linear scale 330Y Linear scale 400 Target mirror 500 Autocollimator

Claims (5)

Three displacement sensors A, B, and C are fixed to the arm in series at every pitch interval p, and the straightness of the object to be measured fixed on the table is adjusted by moving the arm or the table in the linear direction. In the method of measuring the degree,
In the operating range where the autocollimator can measure with high accuracy, the initial straightness is corrected by correcting the pitching of the straightness values detected by the two-point method by the two displacement sensors A and B using the values measured by the autocollimator. In the measurement of the straightness of the object to be measured in the operating range in which the autocollimator cannot measure with high accuracy by obtaining a value data string, 1) The linearly extended upper end where the two first displacement sensors A and the second displacement sensor B are arranged The second displacement sensor B is associated with the output detected by the third displacement sensor C provided at the pitch interval p, the output of the first displacement sensor A, and the output of the second displacement sensor B. The obtained initial straightness value data string is regarded as a measurement standard, and is synthesized with the two output values obtained at the end of the straightness value data string, thereby obtaining the initial straightness value data string. On extension The straightness value output of points separated by the pitch p is obtained. 2) The newly obtained straightness value output is also added to the initial straightness value data row, and the extended straightness value data row is further used as a new measurement standard. The pitch value correction for adding the straightness value output of the point separated by the next pitch p to the extended straightness value data sequence is repeated one after another to expand the measurement range to obtain the extended straightness value data sequence, and output the straightness curve. A method for measuring the surface shape of a long body. Three displacement sensors A, B, and C are fixed to the arm in series at every pitch interval p, and the fourth displacement sensor is connected to the first displacement sensor and the fourth displacement sensor between the first and second displacement sensors. pitch distance p / n (where, the value of p / n is an integer of 2 to 30.) of was inserted in the position where the, the four displacement sensors a, B, C, and D in series In the method of measuring the straightness of the measurement object fixed on the table by moving the fixing arm or the table for fixing the measurement object,
In the operating range where the autocollimator can measure with high accuracy, the initial straightness is corrected by correcting the pitching of the straightness values detected by the two-point method by the two displacement sensors A and B using the values measured by the autocollimator. In the measurement of the straightness of the object to be measured in the operating range in which the autocollimator cannot measure with high accuracy by obtaining a value data string, 1) The linearly extended upper end where the two first displacement sensors A and the second displacement sensor B are arranged The second displacement sensor B is associated with the output detected by the third displacement sensor C provided at the pitch interval p, the output of the first displacement sensor A, and the output of the second displacement sensor B. The obtained initial straightness value data string is regarded as a measurement standard, and is synthesized with the two output values obtained at the end of the straightness value data string, thereby obtaining the initial straightness value data string. On extension The straightness value output of points separated by the pitch p is obtained. 2) The newly obtained straightness value output is also added to the initial straightness value data row, and the extended straightness value data row is further used as a new measurement standard. Pitching correction for adding the straightness value output at the point separated by the next pitch p to the extended straightness value data sequence is repeated one after another to widen the measurement range to obtain an extended straightness value data sequence. 3) The straightness curve The straightness value data for each p / n between the pitch intervals p is the straightness output detected by the two-point method of the pitch interval p / n by the first displacement sensor A and the fourth displacement sensor D as the autocollimator. 4) Output the straightness value data string as a straightness curve by interpolating into the initial straightness value data string or the extended straightness value data string obtained in step 4 and increasing the measurement density. Of the long body Surface shape measuring method. Using a standard straight ruler whose shape of straightness is known in advance, three displacement sensors A, B, and C are fixed to the arm in series at every pitch interval p, and the straightness of the object to be measured fixed on the table is determined as described above. In the method of measuring the straightness of the object to be measured by moving the arm or table in the linear direction,
The pitching and translation in the scanning motion are separately measured by the first and second displacement sensors A and B using the straightness value in a range in which the straightness shape of the object to be measured is known, and the preceding third displacement sensor C A method of outputting a straightness curve by correcting the effects of pitching and translation error components included in the output.   Three worktables movable in a linear direction on which the object to be measured can be placed, and three displacement sensors A, B, and C are fixed to the tool head via a tool head or an arm in series at every pitch interval p. Displacement sensors A, B, C, a standard straight ruler with a known straightness shape, linear scale, input / output means, measurement control means, straightness calculation means, storage means, and recording means A device for measuring the surface shape of an object to be measured fixed on a table.   The straightness value of a reference straight ruler with a known straightness shape is detected by a plurality of displacement sensors, the detected straightness straightness value data string is compared with the straightness value data string of the known reference straight ruler, 3. The reference straight line according to claim 1, wherein when the difference between the two exceeds 10 nm, the detected straightness value data string is recorded as a new straightness shape value of the reference straight ruler. Ruler surface shape measurement method.

Images