JP6430874B2 - Measuring method - Google Patents

Description

  The present invention relates to a measuring method, and more particularly to a measuring method capable of measuring a straight shape and a surface shape of an object to be measured with high accuracy.

  In order to accurately measure the surface shape and the cross-sectional linear shape of an object to be measured such as a long object, a comparative measurement with a standard straight ruler may be performed. Alternatively, a method of calculating a linear shape by measuring an inclination of a mirror on a table that is in contact with a surface to be measured at two points in the scanning direction with an autocollimator on the basis of linearity of the optical axis is also used. Further, when the reference cannot be used, a method of separating the motion error and the shape error by a multipoint method using a multipoint probe is employed. Furthermore, there is also a method for obtaining a linear shape with a level or a Talibel that abuts at two points.

  As the measuring object of straight shape or planar shape becomes larger, the reference ruler becomes longer and its creation becomes difficult. In addition, the optical axis reference may not be sufficiently accurate due to the influence of light fluctuations in the air. Against this background, the need for measurement using the multipoint method is increasing, but the multipoint method has a problem of parabolic error due to zero point adjustment error, and the longer the number becomes, the more the number of successive parabolic errors increases. There is a problem that becomes larger.

  In Patent Document 1, for example, the stage inclination is measured when the movement start point and the end point are stationary in shape measurement, and is included in the difference in inclination between both ends in a straight shape measured and evaluated by a multipoint probe. Disclosed is a technology that can perform so-called in-situ calibration, which can calibrate the zero point of the multipoint probe from the target shape measurement data itself, by using the fact that the influence of the parabolic error due to the zero adjustment error of the probe can be extracted. Yes.

JP 2009-281768 A

  By the way, in actual measurement, it is necessary to relatively move the long object to be measured and the probe by the drive system, but rather to relatively move the stage on which the relatively heavy object to be measured is placed. However, it may be advantageous to install the sensor holder for holding the probe on a portal frame or the like and move the sensor holder relative to the stage relative to the frame because the load on the drive system is reduced. However, when moving the frame, there may be slight deformation of the frame due to the influence of the driving force or the weight of the sensor holder, which causes the stage to tilt during the measurement, causing the probe and the object to be measured to move. The difference in inclination at both ends in the scanning direction changes, which may adversely affect shape measurement. However, increasing the rigidity of the frame has the problem of increasing the load on the drive system, and even if the rigidity of the frame is increased, it is difficult to completely eliminate the tilt of the stage during measurement. On the other hand, even if the stage side is relatively moved instead of fixing the sensor holder side, the stage may be inclined by relatively moving the stage on which the relatively heavy object to be measured is placed.

  The present invention has been made in view of such problems, and an object thereof is to provide a measurement method capable of measuring the shape of a long object with high accuracy while having a simple configuration.

In the measurement method according to claim 1, a stage on which an object to be measured is placed and a sensor holder that holds three probes are relatively movable, and the object to be measured is used by using the probe. In the measurement method by the sequential three-point method for scanning and measuring the linear shape of
Using the gravitational acceleration direction as a reference, the inclination angle in the scanning direction of the sensor holder is measured at the start and end points of scanning measurement, and the difference is obtained. The inclination angle in the scanning direction of the stage is measured and the difference is obtained. Therefore, the linear shape of the object to be measured is obtained based on the obtained two differences while eliminating the influence of the zero point error of the probe.

  According to the present invention, the difference between the start point and the end point of the scan measurement of the tilt angle in the scanning direction of the sensor holder, and the difference between the start point and the end point of the scan measurement of the tilt angle in the scanning direction of the stage. Based on this, by eliminating the influence of the zero point error of the probe, even if the stage or the sensor holder is tilted during measurement, it is possible to perform highly accurate shape measurement. Small and simple.

  According to a second aspect of the present invention, there is provided the measurement method according to the first aspect, wherein the tilt angle of the sensor holder or the tilt angle of the stage is a light source that emits a light beam and the light beam in a suspended state according to gravitational acceleration. Is measured by an angle measuring unit having a reference mirror that reflects the light and a photodetector having a light receiving surface on which the light beam reflected by the reference mirror is incident.

  According to the present invention, since the tilt angle in the scanning direction of the sensor holder and the stage is measured at the start point and end point of scanning measurement, the sensor holder and the stage are continuously moved relative to each other in the meantime. In addition, since the sensor holder and the stage can be kept stationary at the start and end points of scanning measurement, the reference mirror of the angle measurement unit can be prevented from shaking. Accurate tilt angle can be measured.

  The measuring method according to claim 3 is the invention according to claim 1, wherein the tilt angle of the sensor holder or the tilt angle of the stage is suspended in a light source that emits a light beam and in a liquid in a container, Measured by a level having a float having a reference mirror that reflects a light beam provided on at least a part of the surface, and a photodetector having a light receiving surface on which the light beam reflected by the reference mirror is incident. And

  According to the present invention, since the tilt angle in the scanning direction of the sensor holder and the stage is measured at the start point and end point of scanning measurement, the sensor holder and the stage are continuously moved relative to each other in the meantime. In addition, it is possible to perform an efficient measurement at the start point and the end point of scanning measurement. Tilt angle can be measured.

  ADVANTAGE OF THE INVENTION According to this invention, although it is a simple structure, the measuring method which can measure the shape of a long thing accurately can be provided.

It is a schematic diagram of measuring device MD concerning this embodiment. It is a figure which shows schematic structure of the angle measurement unit AMU. It is a figure which shows schematic structure of the level LV. It is a schematic diagram which shows the relationship between a sensor holder, a to-be-measured object, and a stage.

  Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a perspective view of a measuring apparatus MD capable of realizing the measuring method according to the present embodiment. Here, the scanning direction with respect to the measurement object OBJ is defined as the X direction, the vertical direction orthogonal to the X direction is defined as the Z direction, and the direction orthogonal to the X direction and the Z direction is defined as the Y direction. The measuring device MD drives a stage ST that supports an object OBJ having a length L, a portal frame FR that is movable in the X direction with respect to the stage ST, and the portal frame FR. A drive mechanism DR including a drive source and a guide, a sensor holder SH supported at the center of the portal frame FR, and three probes PB1, PB2, and PB3 that are displacement sensors held by the sensor holder SH are provided. doing. The three probes PB1, PB2, and PB3 are arranged in this order along the X direction at intervals d, and the detection sensitivity axis direction is the Z direction.

  An angle measuring unit AMU is disposed on the sensor holder SH so as to be integrally tilted. FIG. 2 is a diagram showing a schematic configuration of the angle measurement unit AMU. A support portion SP that holds the case CA is fixed on the sensor holder SH.

  A light source LD that emits a collimated light beam DL, a beam splitter BS, and a photodetector PD that detects the light beam DL are fixed in the case CA. In addition, a reference mirror SR is connected to the lower end of a thread SG (which is more stable if there are two) suspended from the ceiling surface of the case CA. Thereby, the reflecting surface of the reference mirror SR is parallel to the gravitational acceleration direction. The light source LD, the reference mirror SR, and the photodetector PD constitute a light projecting / receiving system.

  Calibration of the angle measuring unit AMU is performed with the sensor holder SH horizontal. At this time, the light beam DL emitted from the light source LD passes through the beam splitter BS and enters the reference mirror SR. The light beam DL reflected by the reference mirror SR is reflected by the beam splitter BS and enters the light receiving surface PDa of the photodetector PD. The incident position at this time is set as the origin and stored in a memory (not shown) or the like. When actually measuring the tilt angle of the sensor holder SH, the case CA is tilted together with the sensor holder SH, whereas the reference mirror SR always extends in the direction of gravitational acceleration, and therefore the light beam DL incident on the reference mirror SR. Is changed, and the position where the reflected light beam DL is incident on the light receiving surface PDa of the photodetector PD deviates from the origin. Since this deviation amount corresponds to the inclination angle of the sensor holder SH, the inclination angle of the sensor holder SH can be determined by detecting the deviation amount.

  FIG. 3 is a diagram showing a schematic configuration of a level LV for measuring the tilt angle of the stage. The level LV shown in FIG. 3 includes a support portion SP including a pair of leg portions SPa placed on the upper surface of the stage ST. A case CA is disposed on the support portion SP.

  A container VL is formed at the bottom of the case CA. A high-viscosity liquid LQ such as silicon oil is stored in the container VL, and an inverted conical float FT is floated in the container. The bottom of the float FT and the bottom surface of the container VL are connected by the thread SG, and the tilt of the float FT is allowed but the movement is limited. The upper surface of the float FT is exposed above the high-viscosity liquid LQ, and the reference mirror SR is attached here.

  On the other hand, in the case CA, a light source LD, a beam splitter BS, and a photodetector PD are attached above the container VL. The light source LD, the reference mirror SR, and the photodetector PD constitute a light projecting / receiving system.

  Calibration of the level LV is performed by placing it on a horizontal surface plate or the like. At this time, the collimated light beam DL emitted from the light source LD is reflected by the beam splitter BS, is directed to the reference mirror SR on the float FT, is reflected here, passes through the beam splitter BS, and is incident on the photodetector PD. Incident. The incident position at this time is set as the origin and stored in a memory (not shown) or the like. When actually measuring the tilt angle of the stage ST, the case CA also tilts according to the tilt of the stage ST, but the reference mirror SR is always horizontal, so the incident angle of the light beam DL incident on the reference mirror SR changes, The position where the reflected light beam DL is incident on the light receiving surface PDa of the photodetector PD deviates from the origin. Since this deviation amount corresponds to the inclination angle of the stage ST, the inclination angle of the stage ST can be known by detecting the deviation amount. The level LV is fixedly disposed on the stage ST at the center (x = L / 2) of the object OBJ, and is mounted at two points so as not to affect the bending deformation of the stage ST itself. It is preferably mounted so that the inclination of the OBJ to be measured and the inclination of the level LV are the same. However, when the stage ST can be regarded as a rigid body, the level LV is not limited to the center of the object to be measured OBJ on the stage ST, and may be arranged at any position. Further, the tilt angle of the sensor holder SH may be measured using the level LV shown in FIG. 3, and the tilt angle of the stage ST may be measured using the angle measuring unit AMU shown in FIG.

  Next, a method for measuring an object to be measured in the present embodiment will be described. Here, while the frame FR is moved relative to the stage ST, the shape of the upper surface of the object OBJ is measured by the sequential three-point method.

If a minute deformation or tilt occurs when moving the frame FR, a motion error component is generated due to the entire sensor holder SH moving or tilting in the z direction. Here, the surface shape of the object OBJ to be measured is f (x), the eccentric error in the Z direction of the sensor holder SH is _ez (x), the inclination error in the scanning direction is E _p (x), and each probe The outputs m ₁ (x), m ₂ (x), and m ₃ (x) of PB1, PB2, and PB3 can be expressed by the following equations.
m ₁ (x) = f (x−d) + e _z (x) −d · E _p (x) (1)
m ₂ (x) = f (x) + e _z (x) (2)
m ₃ (x) = f (x + d) + e _z (x) + d · E _p (x) (3)

Further, the eccentric error component is eliminated from the outputs of adjacent probes to obtain a differential output of the following formula.
μ ₁ (x) = m ₃ (x) −m ₂ (x) = f (x + d) −f (x) + d · E _p (x) (4)
μ ₂ (x) = m ₂ (x) −m ₁ (x) = f (x) −f (x−d) + d · E _p (x) (5)

Further, when the difference between the expressions (4) and (5) is Δμ (x), the following expression from which the inclination error component is removed is obtained.
Δμ (x) = μ ₁ (x) −μ ₂ (x) = f (x + d) −2f (x) + f (x−d) (6)

On the other hand, when the second-order difference of f (x) is obtained from the equations (1) to (3), the following equation (7) is obtained.
Δ ² f (x)
= {F (x + d) −2 · f (x) + f (x−d)} / d ²
= [{F (x + d) -f (x)}-{f (x) -f (x-d)}] / d ²
= {M ₃ (x) −2 · m ₂ (x) −m ₁ (x)} / d ² (7)

Therefore, Δ ² f (x) is not affected by the translation error e _z (x) and the tilt error E _p (x) of the stage ST, and the probe outputs m ₁ (x), m ₂ (x), m ₃ (x) and the distance d.

That is, it is possible to know the surface shape f (x) of the upper surface of the object OBJ by second-order integration of Δ ² f (x) obtained from the measured values m ₁ (x) to m ₃ (x). it can. Note that the terms below the first order of f (x) represent the average distance and inclination of the measurement part of the object OBJ, and can be ignored in the shape measurement.

However, in practice, each probe PB1, PB2, PB3 supported by the sensor holder SH has a reference point shift at the time of measurement, that is, a so-called zero point shift. For example, when the deviations of the probes PB1, PB2, and PB3 from the reference point in the z direction are set to k ₁ , k ₂ , and k ₃ , respectively, and the equations (1) to (3) are recalculated, the following equation ( 1) ′ to (3) ′.
m ₁ (x) = f (x−d) + e _z (x) −d · E _p (x) + k ₁ (1) ′
m ₂ (x) = f (x) + e _z (x) + k ₂ (2) ′
m ₃ (x) = f (x + d) + e _z (x) + d · E _p (x) + k ₃ (3) ′

Further, taking the second-order difference of f (x), the following equation (7) ′ is obtained.
Δ ² f (x)
= {F (x + d) −2 · f (x) + f (x−d)} / d ²
= {M ₃ (x) −2 · m ₂ (x) −m ₁ (x)} / d ² − {k ₃ −2 · k ₂ + k ₁ } / d ²
= {M ₃ (x) −2 · m ₂ (x) −m ₁ (x)} − k ₁₂₃ / d ² (7) ′
However, in Equation (7) ′, k ₃ −2 · k ₂ + k ₁ = k ₁₂₃ is set.

Further, when Δ ² f (x) is second-order integrated based on the equation (7) ′, k ₁₂₃ / 2d ² is taken as a coefficient in addition to the terms such as the measured values m ₁ (x) to m ₃ (x). A term proportional to x ² is generated. Therefore, the value obtained from the measured values m ₁ (x) to m ₃ (x) is shifted from the surface shape f (x) by k ₁₂₃ · x ² / 2d ² , which is a so-called parabolic error. It is an error caused by a zero point shift known as. Let this parabolic error be g (x). That is, from the output values of the probes PB1, PB2, and PB3, the error inherent shape h (x) = f (), in which the parabolic error g (x) is superimposed on the true top surface shape f (x) of the object OBJ. x) + g (x) is obtained. Therefore, unless the parabolic error g (x) is obtained, it can be said that the upper surface shape f (x) of the true object OBJ cannot be obtained.

Therefore, consider eliminating the parabolic error using a level. For the differential output of equations (4) and (5), the zero point error term α is added to equation (5), and the differential output of equation (4) is shifted by d to obtain the following equation: .
μ ₁ (x + d) = f (x + 2d) −f (x + d) + d · E _p (x + d) (4) ′
μ ₂ (x) = f (x) −f (x−d) + d · E _p (x) + α (5) ′

Here, α is a zero point error in which the shift error in the Z direction due to the fact that the line connecting the measurement ends of two adjacent probes is not parallel corresponds to the angle. Taking the difference between the equations (4) ′ and (5) ′, the following equation is obtained.
ΔEp (x) ≡d (E _p (x + d) −E _p (x)) = μ ₁ (x + d) −μ ₂ (x) + α (8)

  Since equation (8) represents the difference in tilt error between adjacent probes, the following equation (9) is obtained by sequentially adding N points.

  Ep (0) on the left side of equation (9) is a tilt error at the measurement start point (x = 0), and Ep (Nd) is a tilt error at the measurement end point (x = Nd = L). That is, if the inclination of the sensor holder SH at the measurement start point and the measurement end point is measured by the angle measurement unit AMU, the value on the right side, that is, the zero point error α can be theoretically obtained.

  On the other hand, the present inventors have realized that the influence of the zero point error cannot be completely eliminated only by measuring the inclination of the sensor holder SH. Previously, a stage formed from a fairly thick metal was considered a rigid body, and it was estimated that there would be no deformation of the stage when moving the sensor holder side, even if it was negligible. In this case, it is found that the movement of the frame causes a non-negligible deformation on the stage side, the posture relationship between the probes PB1 to PB3 and the object to be measured OBJ changes, and there is a possibility of affecting the zero point error. It was. The same applies to the case where the frame is fixed and the stage is moved.

Based on this knowledge, referring to FIG. 4, the difference Δ1 in the inclination of the sensor holder SH measured by the angle measurement unit AMU at the measurement start point and the measurement end point, and the level at the measurement start point and the measurement end point The difference Δ2 in the inclination of the stage ST measured by LV is obtained, and the following equation (10) can be obtained. In other words, if the differences Δ1 and Δ2 are known, the zero point error α can be obtained from the equation (9). Therefore, the right side of the equation (5) ′ is determined, and this allows the shape of the object OBJ to be accurately measured. It can be sought.
E _p (Nd) −E _p (0) = Δ1 + Δ2 (10)

  In an experiment conducted by the present inventors, the inclination of the whole measuring apparatus at the time of measurement may be inclined by about 0.005 mm / m by moving the stage per meter, and using a probe with a measurement axis interval d = 75 mm pitch. When evaluating the shape of an object to be measured having a length L = 1800 mm by the sequential three-point method, if there is an inclination error of 0.005 mm / m due to the inclination difference between the front and rear ends of the probe, the straightness changes by 1.23 μm. I know. According to the present embodiment, the stage and the sensor holder can be checked with the level at the same time, so that the measurement with the error removed can be performed.

  In the case of the sequential three-point method in which shape measurement is performed using three probes, in-situ calibration is possible only by measuring the inclination angles of the stage ST and the sensor holder SH in the vicinity of the measurement start end and measurement end. In the case of the sequential two-point method in which shape measurement is performed using two probes, errors are eliminated by measuring the tilt angles of the stage ST and the sensor holder SH at or near the measurement point for every shape measurement over the entire scanning measurement. It is desirable.

  The present invention is not limited to the embodiments described in the specification, and other embodiments and modifications are apparent to those skilled in the art from the embodiments and ideas described in the present specification. It is. The description and examples are for illustrative purposes only, and the scope of the invention is indicated by the following claims. For example, in the above-described embodiment, the sensor holder that holds the probe is moved, but the stage side may be moved. In addition, the angle detection unit and / or the level is not limited to the above type, and a general type may be used.

BS Beam splitter CA Case DR Drive mechanism FR Portal frame FT Float DL Light beam LD Light source LQ High viscosity liquid LV Level MD Measuring device OBJ Object to be measured PB1, PB2, PB3 Probe PD Photodetector PDa Light receiving surface SG Thread SH sensor Holder SP Support part SPa Leg SR Reference mirror ST Stage VL Container

Claims (3)

The stage on which the object to be measured is placed and the sensor holder that holds the three probes are relatively movable, and three successive points for scanning and measuring the linear shape of the object to be measured using the probe In the measurement method by the method,
Using the gravitational acceleration direction as a reference, the inclination angle in the scanning direction of the sensor holder is measured at the start and end points of scanning measurement, and the difference is obtained. The inclination angle in the scanning direction of the stage is measured and the difference is obtained. And measuring the linear shape of the object to be measured while eliminating the influence of the zero point error of the probe based on the obtained two differences.   The tilt angle of the sensor holder or the tilt angle of the stage includes a light source that emits a light beam, a reference mirror that reflects the light beam in a suspended state according to gravitational acceleration, and a light receiving light that is incident on the light beam reflected by the reference mirror. The measurement method according to claim 1, wherein the measurement is performed by an angle measurement unit having a photodetector with a surface.   The tilt angle of the sensor holder or the tilt angle of the stage includes a light source that emits a light beam, a float that floats in a liquid in a container, and a reference mirror that reflects the light beam is provided on at least a part of the surface; The measurement method according to claim 1, wherein the measurement is performed by a level having a light receiving surface on which the light beam reflected by the reference mirror is incident.

Images