KR102627963B1 - Polishing device - Google Patents

Abstract

The purpose of the present invention is to provide a polishing device that can stop polishing of a workpiece at a timing when a desired workpiece shape is achieved or at a timing when a desired workpiece shape is achieved, based on the change in shape of the workpiece being polished.
Through a polishing machine 10 that polishes the workpiece W by a rotating lower surface plate 11 and an upper surface plate 12, and a measuring hole 19 formed in the upper surface plate 12. A shape measuring device 20 that measures the shape of the workpiece W, a memory 30 that stores shape information of the workpiece W measured by the shape measuring device 20, and a shape measuring device 20 measuring the shape of the workpiece W. It is provided with an indicator 40 that displays shape information of the workpiece W and a control unit 50 that controls the displayed content of the indicator 40, and the control unit 50 is configured to control the current measured by the shape measuring device 20. A first drawing P1 is created in which the shape drawing of the workpiece Wα, which is a workpiece being polished, is arranged in time series, and this first drawing P1 is displayed on the display 40.

Description

Polishing device {POLISHING DEVICE}

The present invention relates to a polishing device for polishing the surface of a workpiece, such as a silicon wafer.

Conventionally, it is provided with an upper surface plate, a lower surface plate, a sun gear, an internal gear, and a carrier plate, and is used to polish the surface of a workpiece such as a silicon wafer held on the carrier plate. Polishing devices are known (see, for example, Patent Document 1). This polishing device has a measuring device that measures the thickness of the workpiece being polished in real time through a through hole formed in the upper surface plate, and determines the stop timing of the polishing process based on the measurement result of the workpiece thickness by this measuring device.

[Patent Document 1] Japanese Patent Publication No. 2015-47656

However, in a conventional polishing apparatus, the stop timing of polishing is determined based on the measurement result of the thickness of the workpiece. However, it is difficult to predict the future shape change of the workpiece when polishing is continued from the temporary measurement results of the workpiece thickness. Therefore, a problem arises in that it is impossible to determine whether the shape of the workpiece approaches the desired shape when polishing is continued, making it difficult to stop polishing the workpiece at the timing when the desired shape has been achieved. In addition, it is believed that differences in various conditions related to polishing not only affect the shape of the workpiece after completion of polishing, but also affect the shape transition of the workpiece during polishing. However, until now, the time-series change in workpiece shape following polishing has largely depended on the user's skills, which has been an obstacle to improving the efficiency of process improvement.

The present invention was made with attention to the above problem, and provides a polishing device that can stop polishing of a workpiece at the timing when the desired workpiece shape is achieved or at the timing when the desired workpiece shape is achieved, based on the change in shape of the workpiece being polished. The task is to do this.

In order to achieve the above object, the polishing device of the present invention includes a polishing machine that polishes the workpiece by a rotating surface plate, a shape measuring device that measures the shape of the workpiece through a measuring hole formed in the surface plate, and a shape measuring device that measures the shape of the workpiece through a measuring hole formed in the surface plate. It is provided with a memory that stores the shape information of the workpiece, an indicator that displays the shape information of the workpiece measured by the shape measuring device, and a control unit that controls the display contents of the indicator.

Then, the control unit generates a first drawing that arranges in time series the shape drawing of the workpiece being polished, which is the workpiece currently being polished, as measured by the shape measuring device, and displays this first drawing on the display.

As a result, based on the change in shape of the workpiece being polished, polishing of the workpiece can be stopped at the timing when the desired workpiece shape is achieved or at the timing when the desired workpiece shape is achieved.

1 is an explanatory diagram schematically showing the overall configuration of a polishing device of Example 1.
Figure 2 is an explanatory diagram showing the positional relationship between the sun gear, internal gear, and carrier plate in Example 1.
FIG. 3A is an explanatory diagram showing a passage trajectory when a measurement hole passes over a workpiece in the polishing device of Example 1.
FIG. 3B is an explanatory diagram showing a cross-sectional shape line indicating the cross-sectional shape of a workpiece in the polishing device of Example 1.
Fig. 4 is an explanatory diagram showing a first drawing produced by the polishing apparatus of Example 1.
Fig. 5 is an explanatory diagram showing a second drawing produced by the polishing apparatus of Example 1.
Fig. 6 is a flowchart showing the flow of the polishing stop determination process performed in Example 1.
Fig. 7 is a flowchart showing the flow of the second drawing generation process performed in Example 1.
Fig. 8 is an explanatory diagram showing the screen of the display in the polishing apparatus of Example 1.
Fig. 9A is an explanatory diagram arranged in time series of shape drawings when the first workpiece is polished.
FIG. 9B is an explanatory diagram arranged in time series of shape drawings when the second workpiece is polished.
Fig. 10A is an explanatory diagram arranged in time series of shape drawings when the third workpiece is polished.
FIG. 10B is an explanatory diagram arranged in time series of shape drawings when the fourth workpiece is polished.
Figure 11 is an explanatory diagram schematically showing the overall configuration of the polishing apparatus of Example 2.
Fig. 12 is a flowchart showing the flow of the polishing stop determination process performed in Example 2.
Fig. 13A is an explanatory diagram arranged in time series of shape drawings when the fifth workpiece is polished.
FIG. 13B is an explanatory diagram arranged in time series of shape drawings when the sixth workpiece is polished.
Figure 14a is an explanatory diagram showing the relationship between the workpiece polishing time and the flatness of the central portion of the workpiece.
Figure 14b is an explanatory diagram showing the relationship between the workpiece polishing time and the flatness of the outer peripheral area of the workpiece.

Hereinafter, a mode for implementing the polishing apparatus of the present invention will be described based on Examples 1 and 2 shown in the drawings.

(Example 1)

Hereinafter, the configuration of the polishing device 1 of Example 1 is described as “overall configuration,” “detailed configuration of the polisher,” “detailed configuration of the shape measuring device,” “detailed configuration of the memory,” “detailed configuration of the indicator,” and “control unit.” Detailed configuration,” “Polishing stop determination processing configuration,” and “Second drawing generation processing configuration.”

[Full configuration]

The polishing device 1 of Example 1 is a double-sided polishing device that polishes both the front and back sides of a thin plate-shaped workpiece W, such as a semiconductor wafer, quartz wafer, sapphire wafer, glass wafer, or ceramic wafer. As shown in FIG. 1, the polishing device 1 includes a polisher 10, a shape measuring device 20, a memory 30, an indicator 40, and a control unit 50.

[Detailed configuration of the polisher]

The polishing machine 10 polishes the workpiece W using a rotating lower surface plate 11 and an upper surface plate 12. The polishing machine 10 includes a disc-shaped lower grinding plate 11 and an upper grinding plate 12 arranged concentrically with the axis L1 as the center, and a sun gear rotatably disposed at the center of the lower grinding plate 11 ( 13), an internal gear 14 disposed on the outer peripheral side of the lower platen 11, and a workpiece holding hole 15a disposed between the lower platen 11 and the upper platen 12 (see Fig. 2). It has a carrier plate 15 formed thereon. Additionally, a polishing pad 11a is attached to the upper surface of the lower surface 11, and a polishing pad 12a is attached to the lower surface of the upper surface 12. Additionally, a supply hole (not shown) through which the polishing slurry is supplied is formed in the upper surface plate 12.

Here, the carrier plate 15 is engaged with the sun gear 13 and the internal gear 14, as shown in FIG. 2. Then, the carrier plate 15 rotates (revolutions) around the axis L1 while the sun gear 13 and the internal gear 14 rotate.

The workpiece W is disposed within the workpiece holding hole 15a of the carrier plate 15. Then, the carrier plate 15 rotates and rotates while being sandwiched between the polishing pad 11a attached to the rotating lower surface 11 and the polishing pad 12a attached to the rotating upper surface 12, thereby producing the workpiece. (W) is polished by the polishing pad 11a and the polishing pad 12a. That is, the surface of the polishing pad 11a and the polishing pad 12a becomes a polishing surface for polishing the workpiece W.

The upper surface plate 12 is fixed to the rod 16 through support studs 16a and attachment members 16b attached to the upper surface. The rod 16 is expanded and contracted in the vertical direction by the fifth drive device M5. That is, the upper plate 12 moves up and down as the rod 16 expands and contracts.

At the center of the polisher 10, a first drive shaft 17a is disposed upright along the axis L1. The first drive shaft 17a is a shaft rotated by the first drive device M1. A driver 18 is fixed to the upper end of the first drive shaft 17a. As a result, the driver 18 is rotated integrally with the first drive shaft 17a by the first drive device M1.

The driver 18 has a groove (not shown) formed on its outer peripheral surface into which the hook 12b installed on the top plate 12 engages. Then, the rod 16 extends, the upper surface plate 12 moves downward, and the hook 12b is engaged with the groove of the driver 18, so that the driver 18 and the upper surface plate 12 become integrated and rotate. do. That is, the upper surface plate 12 is rotated integrally with the first drive shaft 17a by the first drive device M1.

The second drive shaft 17b is fixed to the hole 13a in the central portion of the sun gear 13 in a penetrating state. The second drive shaft 17b is a hollow tube open at both ends, and rotatably penetrates the first drive shaft 17a. Additionally, the second drive shaft 17b is rotated by the second drive device M2. As a result, the sun gear 13 is rotated integrally with the second drive shaft 17b by the second drive device M2.

A third drive shaft 17c is formed in the lower part of the central portion of the lower surface 11. The third drive shaft 17c is a hollow tube open at both ends, and the second drive shaft 17b is rotatably penetrated therethrough. Additionally, the third drive shaft 17c is rotated by the third drive device M3. Thereby, the lower platen 11 rotates integrally with the third drive shaft 17c by the third drive device M3.

A fourth drive shaft 17d is formed on the internal gear 14. The fourth drive shaft 17d is a hollow tube open at both ends, and the third drive shaft 17c is rotatably penetrated therethrough. Additionally, the fourth drive shaft 17d is rotated by the fourth drive device M4. Thereby, the internal gear 14 rotates integrally with the fourth drive shaft 17d by the fourth drive device M4.

Additionally, a measurement hole 19 is formed in the upper surface plate 12 at a position that is a predetermined distance away from the center along the radial direction. This measurement hole 19 is equipped with a window member 19a that penetrates the upper plate 12 and the polishing pad 12a and transmits the laser light, which is the measurement light.

[Detailed configuration of the shape measuring device]

The shape measuring device 20 irradiates measurement light toward the workpiece W, receives measurement light reflected from the workpiece W, and measures the thickness of the workpiece W being polished. Additionally, this shape measuring device 20 determines the cross-sectional shape of the workpiece W from the measured thickness of the workpiece W. The shape measuring device 20 has a measuring unit 21, a thickness measuring unit 22, and a shape calculating unit 23.

The measurement unit 21 is attached to the upper surface plate 12 and rotates integrally with the upper surface plate 12. In addition, the measurement unit 21 includes a laser light source (not shown) that irradiates laser light, which is the measurement light, toward the workpiece W through the window member 19a of the measurement hole 19 of the upper surface plate 12, It has a light receiving part (not shown) that receives reflected light reflected from the workpiece W. The light receiving signal received by the light receiving unit is transmitted to the thickness measuring unit 22 by the transmitting unit 21a.

The thickness measuring unit 22 measures the thickness of the workpiece W using, for example, an optical reflection interference method. This thickness measuring unit 22 has a receiving unit 22a that receives a light-receiving signal transmitted from the measuring unit 21, and determines the thickness of the workpiece W based on the light-receiving signal received by this receiving unit 22a. .

Here, due to the rotation of the upper plate 12, as shown in FIG. 3A, the laser light from the measurement unit 21 is emitted during the period when the measurement hole 19 is passing on the surface of the workpiece W. The surface of the workpiece W is continuously irradiated. Therefore, the thickness measurement unit 22 continuously measures the thickness of each in-plane position of the workpiece W on the passage traces Na to Nc of the measurement hole 19. And, this thickness measurement unit 22 is used while the measurement hole 19 is passing through each passage trajectory Na to Nc (from one end W1a to W3a of the workpiece W to the other end W1b to W3b). [during the passage period of the measurement hole 19], a data string consisting of a large number of consecutive thickness data is output for each passage. As a result, the thickness measuring unit 22 consists of a plurality of continuous data measuring the thickness of each in-plane position of the workpiece W each time the measurement hole 19 passes on the surface of the workpiece W. Outputs a data string. Meanwhile, the data string output from the thickness measurement unit 22 is stored in the memory 30.

The shape calculation unit 23 determines the cross-sectional shape of the workpiece W. The interval for obtaining the cross-sectional shape of the workpiece W can be set arbitrarily. In Example 1, for example, the cross-sectional shape of the workpiece W is obtained based on a data stream acquired over 15 seconds, and the cross-sectional shape of the workpiece W is newly obtained at intervals of 15 seconds. In addition, shape information such as the cross-sectional shape of the workpiece created by the shape calculation unit 23, workpiece shape patterns obtained by processing the shape information, conditional properties when the workpiece W is polished, and the workpiece W The workpiece shape pattern, etc. generated based on the learning result of the correlation between shape information is stored in the memory 30.

Additionally, the shape calculation unit 23 obtains a cross-sectional shape line T1 as shown in FIG. 3B. This cross-sectional shape line T1 is a shape drawing showing the cross-sectional shape of the workpiece W. The cross-sectional shape line T1 is obtained each time the thickness of the workpiece W is measured by the thickness measurement unit 22. As a result, by arranging the cross-sectional shape lines T1 obtained for the same workpiece W in time series, the change in shape of the workpiece W appears. In addition, processing result information, which is the final workpiece shape of the workpiece W, is displayed by the cross-sectional shape line T1 at the end of polishing of the workpiece W. On the other hand, information on the cross-sectional shape line T1 (information on the cross-sectional shape of the workpiece W) obtained by the shape calculation unit 23 is stored in the memory 30.

[Detailed configuration of memory]

The memory 30 is a storage device capable of recording and reading data from the shape measuring device 20 and the control unit 50. This memory 30 contains information on the thickness of the workpiece W determined by the thickness measurement unit 22 and information on the cross-sectional shape of the workpiece W determined by the shape calculation unit 23 (hereinafter referred to as “workpiece W”). ), etc. are stored.

Additionally, this memory 30 stores the shape information of the workpiece W in relation to the conditional attributes when the workpiece W is polished. Here, “conditional properties” are various parameters that affect the polishing process of the workpiece W, such as polishing conditions, polishing environment, and device characteristics, and have a correlation with the polishing state of the workpiece W. “Conditional attributes” include, for example, operating conditions of the polishing machine 10, polishing slurry conditions, polishing pad conditions, carrier plate conditions, workpiece conditions, and polishing process conditions.

Meanwhile, the operating conditions of the polishing machine 10 include, for example, the rotation speed of the lower grinding plate 11 and the upper grinding plate 12, the rotation speed of the sun gear 13 and the internal gear 14, and the processing of the upper grinding plate 12. Load set value and unit pressure, load slope, coolant temperature of the lower platen 11 or upper platen 12, rotation and revolution speeds of the carrier plate 15, vibration state and inclination characteristics of the polisher 10, etc. . Polishing slurry conditions include, for example, the type, temperature, and flow rate of the polishing slurry, slurry life, and slurry pH value. The polishing pad conditions include, for example, the type, thickness, groove shape, surface roughness, polishing pad life, degree of denatured material deposition, seasoning conditions, etc. of the polishing pad 11a or polishing pad 12a. Carrier plate conditions include the material and thickness of the carrier plate 15, the shape and bending characteristics of the workpiece holding hole 15a and the discard hole, the life of the carrier plate, and the area where wear occurs. Workpiece conditions include the type of workpiece W, thickness at the start of polishing, shape at the start of polishing, and variations in workpiece thickness within a batch. The polishing process conditions include information on the change in shape within the batch, the number of consecutive polishing times, the amount of polishing of the workpiece, the polishing time, the thickness difference between the carrier plate 15 and the workpiece W, etc.

[Detailed configuration of indicator]

Based on a display command from the control unit 50, the indicator 40 provides shape information of the workpiece W currently being polished, shape information of the workpiece W polished in the past, and shape information of the workpiece W. A workpiece shape pattern obtained by calculating and processing a workpiece (W), a workpiece shape pattern created based on the learning results of the correlation between the conditional properties when the workpiece (W) is polished and the shape information of the workpiece (W). Displays arbitrary information such as a decision to stop polishing. The indicator 40 is attached to the polishing machine 10, for example. This indicator 40 has a screen 40a (see Fig. 8) that can be viewed by the user of the polishing machine 10.

[Detailed configuration of the control unit]

The control unit 50 includes a control operation unit 51 consisting of a CPU (Central Processing Unit), a sub-memory 52, an input device 53, and the like. This control unit 50 is based on the program stored in the sub-memory 52 or the processing target or conditional properties of the workpiece W input by the user of the polishing machine 10 through the input device 53, Control commands are output from the control calculation unit 51 to the first driving devices M1 to the fifth driving devices M5 to control the operation of the polishing machine 10.

In addition, this control calculation unit 51 displays the first drawing P1, which arranges the shape drawing of the workpiece W being polished in time series, on the display 40 and predicts the transition of the workpiece shape based on the shape information. And, based on this prediction result, a polishing stop determination process is performed to determine whether or not to stop polishing the workpiece W. That is, this control operation unit 51 includes a first drawing generation unit 54, a second drawing generation unit 55, a display control unit 56, a shape transition prediction unit 57, and a state determination unit ( 58).

The first drawing generation unit 54 extracts shape information of the workpiece currently being polished (hereinafter referred to as “workpiece during polishing (Wα)”) measured by the shape measuring device 20 from the memory 30. . Here, the shape information extracted from the memory 30 is shape information obtained between the start of polishing of the workpiece Wα during polishing and the measurement performed immediately before extraction of the shape information. Then, in this first drawing generation unit 54, based on the extracted shape information of the workpiece Wα during polishing, the shape drawing (cross-sectional shape line T1) of the workpiece Wα during polishing is arranged in order in time series. A first drawing P1 (see FIG. 4) is created.

Meanwhile, the shape information of the workpiece Wα during polishing increases as the number of measurements increases as the polishing process of the workpiece Wα during polishing progresses. Therefore, the first drawing P1 gradually changes from the drawing shown on the left side of FIG. 4 to the drawing shown on the right side of FIG. 4 according to the number of measurements.

In the second drawing generation unit 55, after acquiring the conditional properties of the workpiece Wα during polishing, among the workpieces W polished in the past, that is, before polishing the workpiece Wα during polishing, the workpiece during polishing ( Shape information of the workpiece (hereinafter referred to as “shape reference workpiece (Wβ)”) related to the conditional property matching the conditional property of Wα), or conditional property matching the conditional property of the workpiece (Wα) during polishing. Typical shape information related to is extracted from the memory 30.

Here, the shape information extracted from the memory 30 is extracted from a desired range among the shape information stored in the memory 30.

On the other hand, “typical shape information” is the typical shape transition when the workpiece W is polished, which is abstract and representative shape information obtained arithmetically, and the conditional properties and workpiece when the workpiece W is polished. (W) This is a workpiece shape pattern created based on the learning results of the correlation between the shape information. In addition, hereinafter, it is referred to as “shape information of the selection master” including the shape information or typical shape information of the shape reference workpiece Wβ.

In addition, “matches the conditional properties of the workpiece (Wα) during polishing” means that the conditional properties during polishing are at least partially the same as the conditional properties of the workpiece (Wα) during polishing, or the conditional properties of the workpiece (Wα) during polishing are the same. ) refers to a case where at least some of the conditional properties are similar. For example, as conditional properties of the workpiece Wα during polishing, “rotation speed of lower grinding plate 11 = A”, “rotation speed of upper grinding plate 12 = B”, “slurry type = C”, “carrier material” =D”, “Rotation speed of lower surface plate 11 = A ± x”, “Rotation speed of upper surface plate 12 = B ± y”, “Slurry type = C or C’”, “ Conditional attributes such as "Carrier material = D or D'" are determined to "match the conditional attributes of the workpiece (Wα) during polishing", and the shape information of the selection master related to these conditional attributes is stored from the memory 30. is extracted. On the other hand, the judgment standard for whether the conditional attribute matches can be set arbitrarily.

Then, the second drawing generation unit 55 draws the shape of the selected master (cross-sectional shape line T1) based on the shape information of the selected master extracted based on the conditional properties of the workpiece Wα during polishing. A second drawing P2 (see FIG. 5) is generated in which the are arranged sequentially in time series from the start of polishing to the stop of polishing. Meanwhile, the second drawing generation unit 55 always monitors the conditional properties of the workpiece Wα during polishing during polishing of the workpiece W. And, for example, if the conditional properties of the workpiece Wα during polishing deviate from the original setting or assumed state during the progress of a certain batch due to an abnormality in the slurry flow rate, etc., the conditional properties of the workpiece Wα during polishing may be changed. Edit a new combination of conditional attributes based on the attribute's deviation pattern. Then, the shape information of the selection master related to the newly edited combination of conditional attributes is extracted from the memory 30. Then, the second drawing P2 is generated again based on the newly extracted shape information of the selected master. In addition, when regenerating the second drawing P2, the second drawing generation unit 55 may generate the second drawing P2 based on the virtual pattern of the shape drawing derived by the learning function.

In the display control unit 56, the first drawing P1 generated by the first drawing generating unit 54 and the second drawing P2 generated by the second drawing generating unit 55 are displayed on the display 40. ) A control command to be displayed on the screen 40a is output to the display 40. Additionally, when the status determination unit 58 determines that the polishing process by the polishing machine 10 is to be stopped, the display control unit 56 displays a screen 40a of the display 40 indicating that the polishing stop determination has been made. A control command to be displayed is output to the display 40.

In the shape transition prediction unit 57, the time series change in the shape information of the workpiece Wα during polishing extracted by the first drawing generating unit 54 and the selection master extracted by the second drawing generating unit 55 Compare and calculate time series changes in shape information. Then, this shape change prediction unit 57 predicts the future shape change of the workpiece Wα during polishing based on the result of the comparison operation. On the other hand, the shape transition of the workpiece Wα during polishing predicted by the shape transition prediction unit 57 is the transition of the workpiece shape obtained for each measurement including the final polishing shape.

Then, the comparison calculation of time series changes in shape information by this shape transition prediction unit 57 is performed, for example, in the following procedure. That is, the cross-sectional shape line (T1) of the selection master is arranged in time series for each conditional attribute. Then, a shape transition pattern of the selection master for each conditional attribute is generated, and a database regarding the shape transition pattern is constructed. Here, in the selection master, the conditional properties during polishing match the conditional properties of the workpiece Wα during polishing. Therefore, it is believed that the shape transition of the workpiece Wα during polishing becomes the same as that of the selected master.

Therefore, the shape transition prediction unit 57 compares the cross-sectional shape line T1 of the workpiece Wα during polishing with the databased shape transition pattern through pattern recognition. Then, with reference to the shape transition of the selected master, it is estimated which polishing stage of the workpiece Wα during current polishing is between the start of polishing and the stop of polishing. Additionally, the shape transition prediction unit 57 predicts the future shape transition of the workpiece Wα during polishing based on the current polishing stage of the workpiece Wα during polishing and the shape transition of the selected master.

On the other hand, this shape transition prediction unit 57 is equipped with a machine learning function, and frequently updates the shape transition pattern and temporal variation pattern using machine learning. Additionally, as the polishing process progresses, when the conditional properties monitored during the polishing process of the workpiece Wα change beyond a negligible range, the shape change prediction unit 57 immediately sets a new condition. Based on the properties, the shape of the workpiece Wα during subsequent polishing is predicted, calculated, and output.

The state determination unit 58 determines the current polishing state of the workpiece Wα during polishing based on the future shape transition of the workpiece Wα during polishing predicted by the shape change prediction unit 57 . Here, the “polishing state” refers to a polishing stop state in which the workpiece shape of the workpiece Wα during polishing has reached a workpiece shape in which polishing can be stopped, or a polishing continuation state in which polishing processing by the polishing machine 10 needs to be continued. etc. are included.

[Polishing stop judgment processing configuration]

FIG. 6 is a flowchart showing the flow of the polishing stop determination process executed in the control calculation unit 51 of the control unit 50 in the first embodiment. Hereinafter, based on FIG. 6, each step of the polishing stop determination process in Example 1 will be described.

In step S1, it is determined whether or not polishing processing of the workpiece W by the polishing machine 10 is being performed. If YES (polishing workpiece), proceed to step S2. If NO (no workpiece polishing), repeat step S1.

Here, the determination of whether to carry out polishing of the workpiece by the polisher 10 is based on the control command from the control operation unit 51 to the first to fifth drive devices M1 to M5, and determines whether the polishing command flag is set. It is carried out based on whether or not

In step S2, following the determination in step S1 that the workpiece is being polished, the shape information of the workpiece Wα during polishing measured by the shape measuring device 20 is extracted from the memory 30, and the process proceeds to step S3.

In step S3, following the extraction of the shape information of the workpiece Wα during polishing in step S2, the shape of the workpiece Wα during polishing is drawn based on the shape information of the workpiece Wα during polishing extracted in this step S2. The first drawing P1 (see FIG. 4) arranged sequentially in time series from the start of polishing by the polisher 10 to the measurement performed immediately before information extraction is generated, and the process proceeds to step S4.

In step S4, following the generation of the first drawing P1 in step S3, the first drawing P1 generated in this step S3 and the second drawing P2 generated in the second drawing generation process described later ( 5) is read, and the process proceeds to step S5.

Here, the process for generating the second drawing (P2) (second drawing generation process) is executed in parallel with each step of the polishing stop determination process shown in FIG. 6, and the second drawing (P2) is performed on the workpiece during polishing. If the conditional attribute changes during polishing of (Wα), it is replaced appropriately.

In step S5, following the reading of the first drawing P1 and the second drawing P2 in step S4, the first drawing P1 generated in step S3 and read in step S4 is read in step S4. A control command for simultaneously displaying the drawn second drawing P2 on the screen 40a of the display 40 is output to the display 40, and the process proceeds to step S6.

Meanwhile, in Example 1, as shown in FIG. 8, the indicator 40 equalizes the ratios of the X and Z axes of the first drawing (P1) and the second drawing (P2), and The drawing P1 and the second drawing P2 are arranged horizontally and displayed on the screen 40a. Additionally, the display 40 simultaneously displays the conditional properties (some or all) of the workpiece Wα during polishing on the screen 40a. On the other hand, whether to display all conditional attributes can be set by taking into consideration the display space of the screen 40a and the convenience of monitoring the conditional attributes. Additionally, the conditional properties of the workpiece Wα during polishing do not need to be displayed on the screen 40a.

In step S6, following the output of the control command to the display 40 in step S5, the time series change in the shape information of the selected master extracted when generating the second drawing P2 and the workpiece during polishing (Wα) extracted in step S2 ) Compare and calculate the time series change in shape information, predict the future shape transition of the workpiece Wα during polishing from the results of this comparison calculation, and proceed to step S7.

On the other hand, when the second drawing P2 is replaced, the shape information of the selected master extracted when generating the second drawing P2 changes according to the replaced second drawing P2.

In step S7, following the prediction of the shape transition of the workpiece Wα during polishing in step S6, the polishing state of the workpiece Wα during polishing is determined based on the predicted shape transition, and the process proceeds to step S8.

Here, the polishing state of the workpiece Wα during polishing is determined to be either a “polishing stop state” in which a workpiece shape that can stop polishing processing has been reached, or a “polishing continuation state” in which polishing processing needs to be continued.

In step S8, following the determination of the polishing state of the workpiece Wα during polishing in step S7, the polishing process of the workpiece Wα during polishing is performed by the polishing machine 10 based on the determination of the polishing state performed in this step S7. Determine whether to stop. If YES (polishing stop), proceed to step S9. If NO (continue polishing), proceed to step S2.

Here, the determination that polishing of the workpiece Wα during polishing has been stopped is made when it is determined in step S7 that “polishing has stopped.”

In step S9, following the determination of polishing stop in step S8, a control command is sent to the display 40 to display on the screen 40a of the display 40 that the polishing stop of the workpiece Wα during polishing has been determined. The output is output, a polishing stop decision is notified, and the process proceeds to step S10.

In step S10, following the notification of the polishing stop determination in step S9, polishing of the workpiece Wα during polishing by the polisher 10 is stopped, and the process proceeds to the end.

Here, the polishing process by the polishing machine 10 is stopped by outputting a stop control command from the control operation unit 51 to the first to fifth driving devices M1 to M5.

[Second drawing generation processing configuration]

FIG. 7 is a flowchart showing the flow of the second drawing generation process executed in the second drawing generating unit 55 of the control unit 50 in the first embodiment. Below, each step of the second drawing generation process of Example 1 will be explained based on FIG. 7.

In step S11, it is determined whether or not polishing processing of the workpiece W by the polishing machine 10 is being performed. If YES (polishing workpiece), proceed to step S12. If NO (no workpiece polishing), step S11 is repeated.

Here, the determination of whether to polish the workpiece by the polisher 10 is performed in the same manner as step S1 in the polishing stop determination process.

In step S12, following the determination in step S11 that the workpiece is being polished, the conditional properties of the workpiece Wα currently being polished are acquired, and the process proceeds to step S13.

Here, the conditional properties of the workpiece Wα during polishing are input through the input device 53 by the user of the polishing machine 10 or stored in the sub-memory 52 at the start of polishing of the workpiece Wα during polishing. It is stored in advance, or the change status is monitored by sensors etc. through the CPU.

In step S13, following the acquisition of the conditional attribute of the workpiece Wα during polishing in step S12, selection related to the conditional attribute matching the conditional attribute acquired in step S12 among the workpiece W that has been polished in the past is selected. The master shape information (shape information of the shape reference workpiece Wβ, or typical shape information) is extracted from the memory 30, and the process proceeds to step S14.

In step S14, following the extraction of the shape information of the selected master in step S13, the shape drawing of the selected master is performed from the start of polishing by the polishing machine 10 to the stop of polishing, based on the shape information of the selected master extracted in step S13. A second drawing P2 (see FIG. 5) arranged sequentially in time series is generated, and the process proceeds to step S15.

In step S15, following the generation of the second drawing P2 in step S14, it is determined whether or not to continue polishing the workpiece Wα during polishing. If YES (polishing continues), proceed to step S16. In the case of NO (polishing stop), the process proceeds to the end without requesting generation or replacement of the second drawing P2, and the second drawing generation process ends.

Here, the determination that polishing of the workpiece Wα during polishing is continued is made when it is determined that “polishing continues” requiring continuation of polishing based on the polishing state of the workpiece Wα during polishing.

In step S16, following the determination that polishing continues in step S15, the conditional properties of the workpiece Wα during polishing are acquired again, and the process proceeds to step S17.

In step S17, following the reacquisition of the conditional attributes of the workpiece Wα during polishing in step S16, it is determined whether or not a change has occurred in the state of the conditional attributes of the workpiece Wα during polishing. If YES (there is a change), the process proceeds to step S18. If NO (no change), the process returns to step S15.

Here, whether a change in state has occurred in the conditional properties of the workpiece Wα during polishing is determined by the conditional properties of the workpiece Wα during polishing obtained in step S16 and the conditions of the workpiece Wα during polishing obtained before that. Compare enemy attributes and make judgments based on the differences. On the other hand, a case in which a state change occurs in the conditional attribute of the workpiece (Wα) during polishing refers to a case in which the conditional attribute deviates significantly from the initially set state or the conditional attribute significantly changes from the assumed state as the polishing process progresses. In one case, etc.

In step S18, following the determination that there is a change in the state of the conditional attribute in step S17, the conditional attribute of the selected master is edited based on the changed conditional attribute of the workpiece Wα after the change obtained in step S16. , proceed to step S19.

Here, the editing of the conditional attribute is a conditional attribute of the selected master based on the conditional attribute of the workpiece Wα during polishing, and is a conditional attribute that has a high impact on the polishing process of the workpiece W, or a specific condition. It is to select or replace conditional properties according to.

In step S19, following the editing of the conditional attribute in step S18, the shape information of the selection master related to the conditional attribute that best matches the conditional attribute edited in step S19 [shape information of the shape reference workpiece Wβ, or typical shape information] is extracted from the memory 30, and the process proceeds to step S20.

In step S20, following the extraction of the shape information of the selected master in step S19, the second drawing P2 is generated again based on the shape information of the selected master extracted in step S19. Then, the second drawing P2 generated up to the time of re-extracting the shape information of the selected master is replaced with this newly generated second drawing P2, and the process returns to step S15.

Below, the operation will be explained.

First, the “problem when stopping polishing of a workpiece” will be explained, and then the action of the polishing device 1 of Example 1 will be explained by dividing it into “polishing stop action” and “action of improving prediction accuracy of shape transition.”

[Issues when grinding workpieces are stopped]

When both surfaces of the workpiece W are polished by the polisher 10 of the polishing device 1, the thickness and cross-sectional shape of the workpiece W gradually change as the polishing process progresses. In particular, the cross-sectional shape of the workpiece W is generally determined by the difference in thickness between the carrier plate 15 and the workpiece W when other conditional properties are kept in a desirable state, but as the polishing process progresses, for example, It transitions from a “central convex/outer periphery roll-off-(minus) shape” to a “flat shape” to a “central concave/outer periphery upright shape”. On the other hand, the “central convex shape” is a shape in which the thickness of the central portion of the workpiece W is greater than that of the outer peripheral region. Additionally, the “outer peripheral sagging shape” is a shape in which the thickness gradually decreases toward the outer peripheral edge of the workpiece W. Additionally, the “flat shape” is a shape in which the entire surface of the workpiece W is substantially flat. Additionally, the “central concave shape” is a shape in which the thickness of the central portion of the workpiece W is smaller than that of the outer peripheral region. Additionally, the “outer peripheral standing shape” is a shape in which the thickness gradually increases toward the outer peripheral edge of the workpiece W.

Then, by stopping the polishing process when the thickness of the workpiece W falls within the target thickness range (T1 ≤ thickness ≤ T2), the workpiece W becomes the desired thickness. On the other hand, the cross-sectional shape of the workpiece W varies depending on the settings in the subsequent machining process, but in general, it is often desirable for the workpiece W to have a “flat shape” in which the entire surface is substantially flat. Therefore, it is desired that the polishing process of the workpiece W be stopped when the thickness falls within the target thickness range and the cross-sectional shape becomes a “flat shape.”

In this regard, the thickness of the workpiece W is measured in real time, and a shape drawing of the workpiece W being polished is generated based on the measurement results each time it is measured. Then, the user of the polishing machine 10 monitors the drawing of the shape of the workpiece W, and at a timing when it is considered that the thickness of the workpiece W falls within the target thickness range and the cross-sectional shape has reached a “flat shape,” the polisher ( 10) is done to stop.

However, due to the influence of differences in conditional properties during polishing of the workpiece W, the process of change (shape transition) in the cross-sectional shape of the workpiece W may be different. In addition, the cross-sectional shape of the workpiece W may not have a desired shape such as a “flat shape” due to correlation with the conditional properties during polishing of the workpiece W, and in that case, the second allowable It becomes necessary to stop polishing processing at the cross-sectional shape.

On the other hand, it is difficult to predict how the shape of the workpiece W will change in the future simply by temporarily monitoring the shape drawing of the workpiece W. That is, for example, in the case of the workpiece W having a “slightly convex shape” at the present time, there are cases where it becomes a “flat shape” and a case where it becomes an “upright shape on the outer periphery” by continuing the polishing process. By only temporarily monitoring the current workpiece shape, “approximately central convex shape,” the subsequent workpiece shape is unclear, and as a result, polishing processing cannot be stopped at an appropriate timing, and the workpiece W has an “outer periphery standing shape.” As a result, there is a risk that the site front least squares range (SFQR) in the outer peripheral area of the workpiece will deteriorate. In addition, even if the workpiece W has a “flat shape,” the timing of stopping polishing may be delayed from the right time.

In other words, it is not possible to determine the change in shape of the workpiece W being polished simply by temporarily monitoring the shape drawing of the workpiece W. Therefore, based on the change in shape of the workpiece W, a problem arises in that polishing of the workpiece W cannot be stopped at the timing at which the desired workpiece shape is achieved or at the timing at which the desired workpiece shape is achieved.

[Abrasive stop action]

In the polishing device 1 of Example 1, the thickness and cross-sectional shape of the workpiece W are measured by the shape measuring device 20 while the workpiece W is being polished by the polisher 10. Then, this polishing device 1 stores information on the thickness and cross-sectional shape of the workpiece W measured by the shape measuring device 20 in the memory 30 .

On the other hand, when polishing of the workpiece W is performed by the polisher 10, the control operation unit 51 of the control unit 50 determines that the workpiece W is being polished, and from step S1 shown in the flowchart of FIG. 6 Each process of step S2, step S3, and step S4 is performed in order. That is, the first drawing generation unit 54 extracts the shape information of the workpiece Wα during polishing from the memory 30 and creates the first drawing P1 based on the extracted shape information of the workpiece Wα during polishing. Create. Additionally, the display control unit 56 reads the first drawing P1 generated by the first drawing generating unit 54 and the second drawing P2 generated by the second drawing generating unit 55, and displays the display. A control command is output to display the first drawing (P1) and the second drawing (P2) on the screen 40a of (40).

As a result, the first drawing P1 and the second drawing P2 are displayed simultaneously on the screen 40a of the display 40. In this way, in the polishing device 1 of Example 1, when the shape of the workpiece Wα during polishing is measured by the shape measuring device 20, the first drawing P1 is displayed on the display 40.

Here, the first drawing P1 is a drawing of the shape of the workpiece Wα during polishing (cross-sectional shape line T1) arranged in order in time series. Therefore, the user of the polishing machine 10 can recognize at a glance the continuously drawn shape drawing of the workpiece Wα during polishing. As a result, the user can grasp the change in shape of the workpiece Wα during polishing from the start of polishing to the present (time of first drawing generation). As a result, the user can predict the future shape change of the workpiece Wα during polishing based on the change in shape of the workpiece Wα during polishing. Therefore, even when the polishing machine 10 is controlled by manual operation by the user to stop the polishing process, it is easy to stop the polishing process at the timing when the desired workpiece shape is achieved or at the timing when the desired workpiece shape is achieved.

In addition, since the user can at a glance recognize the shape drawing of the workpiece Wα during polishing in a state related to conditional properties, the user can design the device of the polisher 10 and design auxiliary materials such as slurry. , it is possible to easily improve the conditional properties when polishing a workpiece, such as selection of subsidiary materials and selection of processing conditions. In addition, users can more efficiently plan process improvements and further improve the productivity of wafers that have the desired final workpiece shape.

Moreover, in this Embodiment 1, the first drawing P1 and the second drawing P2 are displayed simultaneously on the screen 40a of the display 40. Therefore, the user of the polishing machine 10 can simultaneously monitor the first drawing P1 and the second drawing P2 by visually viewing the screen 40a.

Here, the second drawing (P2) is based on the shape information of the selected master extracted based on the conditional properties of the workpiece (Wα) during polishing, and draws the shape drawing (cross-sectional shape line (T1)) of this selected master in time series. It is arranged in order. Therefore, the user of the polishing machine 10 monitors the first drawing P1 and simultaneously confirms the second drawing P2, with reference to the shape transition of the selected master shown in the second drawing P2. , it is possible to more accurately predict the future shape change of the workpiece (Wα) during polishing. As a result, when the user controls the polishing machine 10 by manual operation to stop the polishing process, the polishing process can be stopped at a more appropriate timing. Additionally, in some cases, the final workpiece shape or timing of completion of polishing may be adjusted by changing the conditional properties.

On the other hand, in the polishing apparatus 1 of this Example 1, after displaying the first drawing P1 and the second drawing P2 on the screen 40a of the display 40, step S6 shown in the flowchart of FIG. 6 Each process from step S7 to step S8 is performed in order. That is, the shape transition prediction unit 57 compares and calculates the time series change in the shape information of the workpiece Wα during polishing and the time series change of the selected master, and determines the future shape of the workpiece Wα during polishing based on the result. Predict trends. In addition, the state determination unit 58 determines the current polishing state of the workpiece Wα during polishing based on the shape transition of the workpiece Wα during polishing predicted by the shape change prediction unit 57, and polishes the workpiece Wα during polishing. It is determined whether or not to stop polishing processing based on the status determination result.

When the state determination unit 58 determines that polishing of the workpiece Wα during polishing is to be stopped, the process of step S9 shown in the flowchart of FIG. 6 is performed. That is, the display control unit 56 outputs a control command to display on the screen 40a of the display 40 that it has determined that polishing of the workpiece Wα during polishing has been stopped. Then, on the screen 40a of the display 40, a decision to stop polishing of the workpiece Wα during polishing is displayed, and the decision to stop polishing is notified.

As a result, the user of the polishing machine 10 can determine that polishing has been stopped by visually viewing the screen 40a. As a result, even when the user controls the polishing machine 10 through manual operation to stop polishing, polishing can be stopped at an appropriate timing. Meanwhile, as will be described later, even when the polishing machine 10 is stopped by a stop control command from the control unit 50, it becomes possible for the user to recognize the stopping operation of the polishing machine 10.

After that, the process of step S10 shown in the flowchart of FIG. 6 is performed. That is, the control operation unit 51 of the control unit 50 outputs various outputs for finishing polishing processing, such as a stop control command, to the first driving devices M1 to the fifth driving devices M5. As a result, the polishing machine 10 automatically stops after going through a predetermined sequence, and the polishing process of the workpiece Wα during polishing is completed. As a result, in the polishing apparatus 1 of Example 1, the timing of stopping the polishing process can be prevented from being delayed beyond the appropriate time, and the polishing process can be automatically stopped at an appropriate timing.

In the polishing device 1 of this first embodiment, the shape transition prediction unit 57 and the state determination unit 58 of the control calculation unit 51 of the control unit 50 determine the state of the workpiece Wα during polishing. Compare and calculate the time series change of the shape information and the time series change of the selected master. Additionally, the change in shape of the workpiece Wα during polishing is predicted based on the results of this comparison operation. And, based on the change in shape of the workpiece Wα during polishing, the polishing state of the workpiece Wα during polishing is automatically recognized.

Here, the shape information of the selected master is related to conditional properties that match the conditional properties of the workpiece Wα during polishing. Therefore, the time series change of this selected master reflects the correlation between the conditional properties and shape transition of the workpiece Wα during polishing. As a result, in this polishing apparatus 1, the prediction accuracy of the shape transition of the workpiece Wα during polishing can be improved, and the state of the workpiece Wα during polishing can be appropriately determined. And, by appropriately determining the state of the workpiece Wα during polishing, the polishing process can be stopped at the optimal timing when the workpiece Wα during polishing has reached the desired shape or when the workpiece Wα during polishing has reached the desired shape. there is.

Additionally, in the polishing device 1 of this first embodiment, it is determined whether to stop or continue the polishing process depending on the polishing state of the workpiece Wα during polishing, and the polisher 10 is automatically operated at a timing determined to be appropriate. Stop. By automatically stopping polishing in this way, it is possible to prevent the timing of stopping polishing from being delayed from the right time. Additionally, depending on how the device control is structured, the final workpiece shape or timing of completion of polishing may be adjusted by intentionally changing the conditional properties of the workpiece Wα during polishing.

In addition, in this Example 1, the second drawing generation process is performed in the second drawing generation unit 55 in parallel with the polishing stop determination process, and the period between the start and end of polishing of the workpiece Wα during polishing is performed. It is monitored whether the conditional properties of the workpiece Wα change during polishing. And, when the conditional properties of the workpiece Wα change during polishing, the second drawing P2 is replaced regardless of the stage at that point in the polishing stop determination process.

That is, when the workpiece W is polished by the polisher 10, the second drawing generation unit 55 determines that the workpiece W is being polished, and proceeds from step S11 to step S12 shown in the flowchart of FIG. 7. , Step S13, and Step S14 are performed in order. That is, the second drawing generation unit 55 acquires the conditional attributes of the workpiece Wα during polishing, and generates shape information of the selection master related to the conditional attributes matching the acquired conditional attributes [shape reference workpiece ( [Wβ) shape information or typical shape information] is extracted from the memory 30. Then, this second drawing generation unit 55 generates the second drawing P2 based on the shape information of the extracted selected master.

If the second drawing P2 has been generated, the second drawing generating unit 55 performs the process of step S15 and determines whether polishing of the workpiece W? is continued during polishing. Then, when polishing continues, each process of step S16 and step S17 is performed in order, the conditional attribute of the workpiece Wα during polishing is acquired again, and it is determined whether or not a change has occurred in this conditional attribute. do.

Then, when it is determined that a change has occurred in the conditional properties of the workpiece Wα during polishing, it is determined that the conditional properties have changed during polishing, and steps S18 through S19 and S20 shown in the flowchart of FIG. 7 are performed. Perform processing. That is, the conditional attributes of the selection master are re-edited based on the re-acquired conditional attributes of the workpiece Wα during polishing, and the selection master related to the conditional attribute that best matches the re-edited conditional attributes is edited. , generate a new second drawing (P2). Then, the second drawing (P2) up to that point is replaced with a new one.

Accordingly, even if the conditional properties of the workpiece Wα during polishing change as the polishing process progresses, the second drawing P2 can be made the latest drawing according to the change in the conditional properties of the workpiece Wα during polishing. The shape of the workpiece Wα can be appropriately predicted during subsequent polishing.

Specific examples are given below.

As shown in FIG. 9A, in the initial polishing step A, the first workpiece W1, which has a cross-sectional shape of a “strong concave shape” with a relatively large central portion, is subjected to polishing, and the thickness is reduced. In the polishing step B immediately after reaching the target thickness range (T1 ≤ thickness ≤ T2), the cross-sectional shape becomes “slightly concave (with a small depression in the central portion).” Then, the polishing process continues, and in the polishing step C when the thickness approaches the lower limit T1 of the target thickness range, the cross-sectional shape of the first workpiece W1 becomes a “flat shape.”

Meanwhile, as shown in FIG. 9B, in the polishing step A at the initial stage of polishing, the second workpiece W2, which has a “slightly convex” cross-sectional shape with a relatively small protruding central portion, undergoes polishing, and the thickness decreases. In the polishing step B immediately after reaching the target thickness range (T1 ≤ thickness ≤ T2), the cross-sectional shape becomes “slightly concave (with a small depression in the central portion).” However, the polishing process is continued thereafter, and in the polishing step C when the thickness approaches the lower limit (T1) of the target thickness range, the cross-sectional shape of the second workpiece W2 is “strongly concave (largely depressed in the center)”. It becomes this.

In contrast, in the polishing device 1 of Example 1, a first drawing P1 is generated in which the shape drawings of the first workpiece W1 and the second workpiece W2 are arranged in order in time series, and the first drawing P1 is The drawing (P1) is displayed on the indicator (40). Therefore, the shape transition of each of the first workpiece W1 and the second workpiece W2 can be grasped from this first drawing P1, and subsequent shape changes can be predicted.

That is, in the polishing device 1 of Example 1, at polishing step B, it is determined that, for the first workpiece W1, “since the depression in the central part is gradually becoming shallower, it is better to polish until polishing step C.” You can. On the other hand, in the second workpiece W2, it can be determined that “since the depression in the central part is gradually deepening, it is better to stop polishing at the polishing step B rather than polishing to the polishing step C.” In this way, the polishing device 1 of Example 1 can appropriately determine the optimal timing for stopping polishing according to the shape of the workpiece at the start of polishing and polish it to a desired workpiece shape.

Additionally, in the polishing apparatus 1 of Example 1, the time series change in the shape information of the workpiece Wα during polishing and the time series change in the selected master are compared and calculated. For this reason, even in cases where the future change in shape of the workpiece cannot be predicted based solely on the shape change of the current arrangement, the shape change can be appropriately predicted.

As shown in FIG. 10A, the third workpiece W3, which has a “strongly convex shape” with a relatively large central portion protruding at the start of polishing, is polished by a polisher 10 with a “conditional property that the peripheral portion of the workpiece is difficult to deform (bend).” A case in which polishing processing is performed will be described. In the polishing step A at the beginning of polishing, the cross-sectional shape of the third workpiece W3 of “strongly convex shape” is the polishing step immediately after polishing progresses and the thickness reaches the target thickness range (T1 ≤ thickness ≤ T2). In B, it has a “slightly convex shape (the central portion protrudes slightly).” Furthermore, the polishing process continues, and in the polishing step C when the thickness approaches the lower limit T1 of the target thickness range, the cross-sectional shape of the third workpiece W3 becomes a “flat shape.”

Meanwhile, as shown in FIG. 10B, the fourth workpiece W4, which has a “strongly convex shape” with a relatively large central portion protruding at the start of polishing, is polished by the polisher 10 with a “conditional property that the peripheral portion of the workpiece is prone to bending.” A case in which processing is performed will be explained. In the polishing step A at the beginning of polishing, the cross-sectional shape of the fourth workpiece W4 of “strongly convex shape” is the polishing step immediately after the polishing process progresses and the thickness reaches the target thickness range (T1 ≤ thickness ≤ T2). In B, it has a “slightly convex shape (the central part protrudes slightly).” In addition, in the polishing step C where the polishing process continues and the thickness approaches the lower limit T1 of the target thickness range, the cross-sectional shape of this fourth workpiece W4 is "slightly convex and slightly erect" (the center and peripheral portions are each small and It becomes a “protruding state”).

Here, when the accumulation of data is still insufficient and it is not possible to predict the shape transition with a strong correlation with the conditional properties, for example, polishing is performed immediately before polishing the workpiece Wα during polishing (e.g., before one batch). From the shape transition of the machined workpiece, it is possible to predict the shape transition of the workpiece Wα during polishing after a certain point in the current batch, which shows the same shape trend as the current batch. That is, in the polishing apparatus 1 of Example 1, for example, a workpiece polished before one batch of the workpiece Wα during polishing is used as the selection master. Then, by comparing and calculating the time series change of the selection master and the time series change of the shape information of the third workpiece W3 or the fourth workpiece W4, the polishing machine 10 used in the current batch or in batches performed before and after it. Conditional properties can be estimated.

That is, in the polishing device 1 of Example 1, when polishing is performed with the polishing machine 10 with the “conditional property that the peripheral portion of the workpiece is difficult to bend,” in polishing step B, “even if the protrusion in the central portion becomes shallow, the outer peripheral region Since it is difficult for standing to occur, it can be determined that it is better to carry out polishing up to polishing step C. On the other hand, when performing polishing with the polishing machine 10, which has “the conditional property that the peripheral part of the workpiece is prone to bending,” in polishing step B, “standing progresses in the outer peripheral area as polishing progresses, so up to polishing step C.” It can be determined that “it is better to stop the polishing process at polishing step B rather than polishing.” In this way, the polishing device 1 of Example 1 appropriately determines the optimal timing of polishing stop according to the conditional attribute of the polisher 10 selected by the user or the conditional attribute of the given polisher 10, and achieves the desired polishing stop. It can be polished to the shape of the workpiece.

[Improvement of prediction accuracy of shape transition]

In the polishing apparatus 1 of Example 1, the shape information of the workpiece W is stored in the memory 30 in association with the conditional attributes during polishing of the workpiece W. Then, the shape information of the selection master related to the conditional attribute matching the conditional attribute of the workpiece Wα during polishing is extracted from the memory 30, and a second drawing (P2) is performed based on the shape information of the extracted selection master. ) is created.

Therefore, the shape transition of the selected master shown by the second drawing P2 reflects the correlation between the conditional properties and the shape transition in the workpiece Wα during polishing. And, by displaying this second drawing P2 simultaneously with the first drawing P1, the prediction accuracy of the shape transition of the workpiece Wα during polishing by the user who monitors these drawings can be improved.

In addition, as the shape information of the selection master, a workpiece shape pattern (typical shape information) generated based on the learning result of the correlation between the shape information of the workpiece W and the conditional properties when the workpiece W is polished. In the case of using the shape information of the selection master, the transition accuracy of the shape of the workpiece shown by the second drawing P2 can be improved compared to the case of using the shape information of the shape reference workpiece Wβ. Therefore, it becomes possible to more accurately predict the shape transition of the workpiece Wα during polishing and to stop polishing at a more appropriate timing.

Additionally, in Example 1, the shape transition prediction unit 57 is equipped with a machine learning function and frequently updates the databased shape transition prediction pattern using machine learning. As a result, the prediction of the shape transition of the workpiece Wα during polishing from the current point onwards automatically becomes more accurate.

In addition, as the prediction accuracy of the shape transition is improved, when determining the state of the workpiece Wα during polishing based on the prediction of the transition of the workpiece shape, the state of the workpiece Wα can be determined more appropriately. As a result, it becomes possible to stop polishing at the optimal timing with higher precision.

Next, the effect is explained.

In the polishing device 1 of Example 1, the effects listed below can be obtained.

(1) a polishing machine 10 that polishes the workpiece W by a rotating surface plate (lower surface plate 11 and upper surface plate 12);

A shape measuring device 20 that measures the shape of the workpiece W through a measuring hole 19 formed in the surface plate (top plate 12),

A memory 30 that stores shape information of the workpiece W measured by the shape measuring device 20,

An indicator (40) that displays shape information of the workpiece (W) measured by the shape measuring device (20),

Provided with a control unit 50 that controls the display contents of the indicator 40,

The control unit 50 generates a first drawing P1 in which the shape drawing of the workpiece Wα during polishing, which is the workpiece currently being polished, measured by the shape measuring device 20, is arranged in time series, and this first drawing P1 ) was configured to display on the indicator 40.

Accordingly, based on the change in shape of the workpiece Wα during polishing, the polishing process of the workpiece Wα during polishing can be stopped at the timing at which the desired workpiece shape is achieved or at the timing at which the desired workpiece shape is achieved.

(2) The memory 30 stores the shape information of the workpiece W in relation to the conditional properties when the workpiece W is polished, and the control unit 50 stores the workpiece Wα during polishing. Based on the results of comparison calculation of the time series change of the shape information of the workpiece (Wα) during polishing and the time series change of the shape information (shape information of the selected master) related to the conditional properties matching the conditional properties of the workpiece (Wα) during polishing ( The shape transition of the workpiece Wα during polishing is predicted, and the state of the workpiece Wα during polishing is determined based on the prediction of the shape transition of the workpiece Wα during polishing.

As a result, the prediction accuracy of the shape transition of the workpiece Wα during polishing can be improved to determine the state appropriately, and the optimal timing when the workpiece Wα during polishing has the desired shape, or the workpiece Wα during polishing has the desired shape. Polishing can be stopped at the optimal timing for shaping.

(3) a polishing machine 10 that polishes the workpiece W by a rotating surface plate (lower surface plate 11 and upper surface plate 12);

A shape measuring device 20 that measures the shape of the workpiece W through a measuring hole 19 formed in the surface plate (top plate 12),

A memory 30 that stores shape information of the workpiece W measured by the shape measuring device 20,

An indicator (40) that displays shape information of the workpiece (W) measured by the shape measuring device (20),

Provided with a control unit 50 that controls the display contents of the indicator 40,

The control unit 50 provides a first drawing P1 in which the shape drawing of the workpiece Wα during polishing, which is the workpiece currently being polished, measured by the shape measuring device 20, is arranged in time series, and a polishing drawing P1 of the workpiece Wα during polishing. Generating a second drawing (P2) in which the shape drawing of a previously polished workpiece (selection master) is arranged in time series, and displaying the first drawing (P1) and the second drawing (P2) on the display 40 at the same time. It was composed.

Accordingly, based on the change in shape of the workpiece Wα during polishing, the polishing process of the workpiece Wα during polishing can be stopped at the timing when the desired workpiece shape is achieved or at the timing when the workpiece Wα during polishing becomes the desired shape. You can.

(4) The memory 30 stores the shape information of the workpiece W by associating it with conditional properties when the workpiece W is polished,

The control unit 50 is configured to generate the second drawing P2 based on the shape information of the workpiece (selection master) related to the conditional attribute matching the conditional attribute of the workpiece Wα during polishing.

As a result, it is possible to improve the prediction accuracy of the shape transition of the workpiece Wα during polishing.

(5) The control unit 50 generates a workpiece shape pattern (typical shape information) based on the correlation between the conditional properties when the workpiece W is polished and the shape information of the workpiece W. 2 Create a drawing (P2).

As a result, the accuracy of the transition of the shape of the workpiece shown by the second drawing P2 can be improved, and the shape transition of the workpiece Wα during polishing can be predicted more accurately.

(6) The memory 30 stores the shape information of the workpiece W by associating the conditional properties when the workpiece W is polished,

The control unit 50 controls the time series change of the shape information of the workpiece Wα during polishing and the time series change of shape information (shape information of the selection master) related to the conditional attribute matching the conditional attribute of the workpiece Wα during polishing. The shape transition of the workpiece Wα during polishing is predicted based on the results of the comparison operation, and the state of the workpiece Wα during polishing is determined based on the prediction of the shape transition of the workpiece Wα during polishing. .

As a result, the prediction accuracy of the shape transition of the workpiece Wα during polishing can be improved to determine the state appropriately, and the workpiece Wα during polishing has become the desired shape, or the workpiece Wα during polishing has the desired shape. Polishing processing can be stopped at the optimal timing.

(7) When the control unit 50 determines that polishing of the workpiece Wα during polishing is to be stopped as a result of determining the state of the workpiece Wα during polishing, the control unit 50 stops the polishing process of the workpiece Wα during polishing, It is configured to notify that the polishing process of the workpiece Wα during polishing has been stopped.

Thereby, the polishing stop of the workpiece Wα during polishing can be automatically performed at an appropriate timing, and the user of the polishing machine 10 can be notified of the stoppage of polishing processing.

(Example 2)

The polishing apparatus of Example 2 is an example in which the responsible parameters are specified when the final workpiece shape of the workpiece Wα during polishing becomes a secondary acceptable workpiece shape, and the responsible parameters are notified. Hereinafter, the polishing apparatus of Example 2 will be described. Meanwhile, for structures equivalent to those of the polishing device 1 in Example 1, the same reference numerals as in Example 1 are given, and detailed descriptions are omitted.

In the polishing apparatus 1A of Example 2, as shown in FIG. 11, the control operation unit 51A of the control unit 50A includes a first drawing generating unit 54, a second drawing generating unit 55, and , a display control unit 56A, a shape transition prediction unit 57A, a state determination unit 58A, a parameter specification unit 59, and a correlation data processing unit 60.

The shape transition prediction unit 57A of Example 2 compares and calculates the time series change in the shape information of the workpiece Wα during polishing and the time series change in the shape information of the selected master, and based on the results of this comparison operation, polishing Predict the future shape trend of the heavy workpiece (Wα). In addition, in this shape transition prediction unit 57A, based on the prediction of the future shape transition of the workpiece Wα during polishing, the workpiece Wα during polishing is supported while receiving support from the correlation data processing unit 60 as necessary. It is predicted whether the final workpiece shape (hereinafter referred to as “final workpiece shape”) can be the desired workpiece shape.

Here, the “desired workpiece shape (hereinafter referred to as “desired state”)” is a shape that satisfies the first shape condition set in advance. On the other hand, when it is predicted that the final workpiece shape cannot be in the desired state, the shape transition prediction unit 57A receives support from the correlation data processing unit 60 as necessary and determines that the final workpiece shape is secondarily acceptable. Predict whether the shape will be possible or not. On the other hand, “secondarily acceptable workpiece shape (hereinafter referred to as “secondary acceptable state”)” refers to a determination that the final workpiece shape cannot reach the desired state, that is, the first shape condition is not satisfied even if polishing processing continues. It is a shape that satisfies the second shape condition set when

The state determination unit 58A of Example 2 determines the current polishing state of the workpiece Wα during polishing based on the future shape transition of the workpiece Wα during polishing predicted by the shape change prediction unit 57A. do. Here, the “polishing state” determined by the state determination unit 58A includes the first polishing stop state in which the workpiece shape of the workpiece Wα during polishing has reached the desired state, or the workpiece shape of the workpiece Wα during polishing. These include a second polishing stop state that has reached this secondary allowable state, a third polishing stop state that requires immediate polishing stop, and a polishing continuation state that requires continuation of polishing by the polishing machine 10.

In the parameter specifying unit 59, when the shape transition prediction unit 57A determines that the final workpiece shape may enter the secondary acceptable state, the final workpiece shape of the workpiece Wα during polishing enters the secondary acceptable state ( A highly correlated conditional attribute (hereinafter referred to as a “responsibility parameter”) is specified for a condition that cannot be in the desired state. Additionally, in this parameter specifying unit 59, responsible parameters may be listed in order of highest correlation strength.

Specification of the responsible parameters by this parameter specification unit 59 is performed, for example, in the following procedure. That is, the data of the conditional attribute searched by the correlation data processing unit 60 and having a high correlation strength with the abnormal state of the workpiece shape stored in the memory 30, and the polishing process determined to be in a secondary acceptable state. Contrast the conditional properties of the artifact (Wα). Then, a conditional attribute with a relatively high correlation strength is specified as a “responsibility parameter.” Meanwhile, there may be one or more of these responsibility parameters.

In addition, the parameter specification unit 59 enumerates the responsible parameters in order of high correlation and strength, for example, in the following order. In other words, data of conditional properties that have a high degree of correlation with the ideal state of the workpiece shape are contrasted with the conditional properties of the workpiece (Wα) during polishing that has been determined to be in a secondary acceptable state. Then, multiple conditional attributes are selected in order of relatively high correlation strength, and responsibility parameters are listed in the selected order. Meanwhile, the number of conditional attributes to select may be two or more.

On the other hand, it is also possible to analyze the correlation between abnormalities in state information monitored during polishing of the workpiece W, such as data on temperature distribution on the polishing pad and data on vibration and temperature of the bearing, and the abnormal state of the shape of the workpiece. Therefore, by monitoring not only the single workpiece shape trend, but also the correlation between the workpiece shape trend trend over multiple batches and the trend trend of the conditional attribute, the specific precision of the responsible parameter and the correlation strength of the responsible parameter can be achieved. It can be improved. On the other hand, multivariate analysis, artificial intelligence (machine learning, deep learning), etc. can be used to update the correlation analysis and the prediction model based on it, or to update the prediction accuracy from time to time.

In addition, in the display control unit 56A of Example 2, when outputting a control command to the display 40 to display on the screen 40a of the display 40 that a polishing stop determination has been performed, the status determination unit 58A ) is determined to be the “first polishing stop state”, a control command indicating that the workpiece shape is in the desired state is output. In addition, when the state determination unit 58A determines that it is a “second polishing stop state,” the workpiece shape is in a secondary allowable state, information on the responsible parameter specified by the parameter specification unit 59, or correlation Outputs a control command that displays information on the responsible parameters listed in descending order of intensity. In addition, when the state determination unit 58A determines the “third polishing stop state”, it outputs a control command indicating that the shape of the workpiece is outside the allowable shape.

The correlation data processing unit 60 searches for the conditional properties of the workpiece W, the shape transition of the workpiece W during polishing, and the correlation strength of the final workpiece shape. The correlation strength search by the correlation data processing unit 60 is performed, for example, in the following procedure.

That is, based on the results of polishing the workpiece W performed in the past, recognition of the abnormal state when an abnormality (for example, discontinuity in slurry flow rate) occurs with respect to a predetermined conditional attribute, and the shape of the workpiece Wα during polishing. The relationship between shape transitions (hereinafter referred to as “abnormal state of workpiece shape”) when entering the secondary acceptable state is calculated computationally. On the other hand, for example, recognition of the state of discontinuity in the slurry flow rate is performed by comparing the length of time when the slurry flow rate was below a predetermined value with a threshold value.

Then, for each conditional attribute when an abnormal state of the workpiece shape occurs, the strength of correlation with the abnormal state of the workpiece shape is searched, for example, by calculating a correlation coefficient through regression analysis. Additionally, from the results of polishing the workpiece W performed in the past, the relationship between predetermined conditional properties and the shape transition of the workpiece W at that time is obtained. And, based on the relationship between this predetermined conditional property and the workpiece shape transition, the parameter that is suspicious as the cause of the discrepancy between the desired shape transition or desired final workpiece shape following polishing processing and the actual shape trend or final workpiece shape occurs. By specifying , the correlation strength between the conditional attribute and the workpiece shape transition and the final workpiece shape may be searched. This correlation strength data is stored in the memory 30 in association with the recognition and conditional attributes of a specific abnormal state.

On the other hand, this correlation data processing unit 60 is a dedicated calculation unit that exclusively searches the conditional properties of the workpiece W, the shape transition of the workpiece W during polishing, and the correlation strength of the final workpiece shape. Therefore, this correlation data processing unit 60 can perform correlation search calculation regardless of whether the workpiece W is being polished or not.

Next, the polishing stop determination process performed in the polishing apparatus 1A of Example 2 will be described using the flowchart shown in FIG. 12. Processes that are the same as the polishing stop determination process in Example 1 are given the same reference numerals as in Example 1, and detailed descriptions are omitted. Meanwhile, also in the polishing apparatus 1A of Example 2, a second drawing generation process for generating a second drawing P2 is performed in parallel with the polishing stop determination process shown in FIG. 12. In the polishing stop determination process performed in Example 2, the second drawing P2 generated by the second drawing generation process is read at a required timing (step S4).

In the polishing stop determination process performed in Example 2, in step S6, the time series change in the shape information of the selected master and the time series change in the shape information of the workpiece Wα during polishing are compared and calculated. Then, if the future shape transition of the workpiece Wα during polishing is predicted from the results of this comparison operation, the process proceeds to step S61.

In step S61, it is determined whether the final workpiece shape of the workpiece Wα during polishing can become the desired workpiece shape, based on the prediction of the future shape transition of the workpiece Wα during polishing. If YES (the desired state can be achieved), the process proceeds to step S71. If NO (cannot be in the desired state), the process proceeds to step S62.

In step S62, following the determination in step S61 that the final workpiece shape cannot be in the desired state, the final workpiece shape of the workpiece Wα during polishing is determined based on the prediction of the future shape transition of the workpiece Wα during polishing. Secondly, it is determined whether an acceptable workpiece shape can be achieved. If YES (secondary allowable state may be reached), polishing processing continues and the process proceeds to step S63. If NO (secondary permission state cannot be reached), the process proceeds to step S71.

In step S63, following the determination in step S62 that the final workpiece shape can be in a secondary acceptable state, a conditional attribute with a high degree of correlation is applied to the final workpiece shape of the workpiece Wα during polishing being in a secondary acceptable state. The responsible parameters are specified, or the responsible parameters are listed in order of highest correlation strength, and the process proceeds to step S71.

In step S71, it is determined that the final workpiece shape in step S61 can be in the desired state, in step S62, it is determined that the final workpiece shape cannot be in either the desired state or the secondary acceptable state, and in step S63, the responsibility parameter is determined. Following any one of the enumeration of responsible parameters in order of specific or high correlation strength, the polishing state of the workpiece Wα during polishing is determined based on the prediction of the future shape transition of the workpiece Wα during polishing, and the steps are: Proceed to S81.

Here, the polishing state of the workpiece Wα during polishing is a “first polishing stop state” in which the workpiece shape of the workpiece Wα during polishing has reached a desired state, and the workpiece shape of the workpiece Wα during polishing is in a secondary allowable state. It is determined as one of the “second polishing stop state” that has been reached, the “third polishing stop state” that immediately stops polishing, and the “polishing continuation state” that requires continuation of polishing.

On the other hand, the determination of whether it is the “first polishing stop state” is made when it is determined in step S61 that the final workpiece shape can be in the desired state. In addition, determination of whether the "second polishing stop state" is made is made in step S63 when the responsible parameters are specified or the responsible parameters are enumerated in order of high correlation strength. Additionally, a determination of “third polishing stop state” is made when it is determined in step S62 that the final workpiece shape cannot be in either the desired state or the secondary allowable state. In addition, in the judgment of “polishing continued state”, it is determined that the final workpiece shape is either the desired state or the secondary acceptable state, but the current workpiece shape of the workpiece Wα during polishing reaches the desired state or the secondary acceptable state. It is done when it is not there.

In step S81, following the determination of the polishing state of the workpiece Wα during polishing in step S71, the workpiece Wα during polishing is polished by the polishing machine 10 based on the determination of the polishing state performed in step S71. Determine whether to stop. If YES (polishing stop), proceed to step S91. If NO (continue polishing), return to step S2.

Here, the determination that the polishing of the workpiece Wα during polishing is stopped is when one of the “first polishing stop state,” “second polishing stop state,” and “third polishing stop state” is made in step S71. It is done.

In step S91, following the determination of polishing stop in step S81, the state of the workpiece Wα during polishing is displayed on the screen 40a of the display 40 along with the effect that the polishing stop of the workpiece Wα during polishing has been determined. A control command to be displayed is output to the display 40, a polishing stop decision is notified, and the process proceeds to step S10.

Here, when a determination of “first polishing stop state” is made in step S71, a control command indicating that polishing stop has been determined and indicating that the workpiece Wα during polishing is in a “desired state” is output. In addition, when a determination of “second polishing stop state” is made in step S71, a control command indicating that polishing stop has been determined and that the workpiece Wα during polishing is in a “secondary allowable state” is output. In addition, when a determination of “third polishing stop state” is made in step S71, along with the effect that polishing stop has been determined, the workpiece Wα during polishing is “allowed,” meaning that it is neither the desired state nor the secondary allowable state. Outputs a control command that displays the “external shape.”

Next, the operation of the polishing device 1A of Example 2 will be explained.

In the polishing device 1A of Example 2, when the workpiece W is polished by the polisher 10, the first drawing P1 and the second drawing P2 are displayed as indicators ( 40) is displayed on the screen 40a. After that, each process of step S6 and step S61 shown in the flowchart of FIG. 12 is performed in order. That is, the shape transition prediction unit 57A compares and calculates the time-series change in the shape information of the workpiece Wα during polishing and the time-series change in the shape information of the selected master, and based on the result, the time-series change in the shape information of the workpiece Wα during polishing is calculated. Predict future shape trends. Then, based on the prediction of this shape transition, it is determined whether the final workpiece shape can be in the desired state.

When it is determined that the final workpiece shape can be in the desired state, the processes from step S71 to step S81 shown in the flowchart of FIG. 12 are sequentially performed. That is, the state determination unit 58A determines the current polishing state of the workpiece Wα during polishing based on the shape transition of the workpiece Wα during polishing predicted by the shape change prediction unit 57A, and polishes the workpiece Wα during polishing. It is determined whether or not to stop polishing processing based on the status determination result.

When the state determination unit 58A determines that the polishing state of the workpiece Wα during polishing is “the first polishing stop state,” the process of step S91 shown in the flowchart of FIG. 12 is performed. That is, the display control unit 56A controls to display on the screen 40a of the display 40 that the workpiece Wα during polishing is in a “desired state” along with the effect that it has determined that polishing of the workpiece Wα during polishing has been stopped. Outputs a command. And, on the screen 40a of the display 40, in addition to the decision to stop polishing of the workpiece Wα during polishing, it is displayed that the workpiece Wα during polishing is in the “desired state”, and a decision to stop polishing is notified. do.

As a result, the user of the polishing machine 10 can determine that polishing has been stopped and the shape of the workpiece Wα during polishing by visually viewing the screen 40a. As a result, even when the user controls the polishing machine 10 through manual operation to stop polishing, polishing can be stopped at an appropriate timing. Additionally, even when the polishing machine 10 is stopped by a stop control command from the control unit 50A, the user can recognize the stopping operation of the polishing machine 10.

After that, by performing the process of step S10, the control operation unit 51A of the control unit 50A sends various outputs for finishing polishing processing, such as a stop control command, to the first drive devices M1 to the fifth drive devices M5. do it As a result, the polishing machine 10 automatically stops after going through a predetermined sequence, and the polishing process of the workpiece Wα during polishing is completed.

On the other hand, when it is determined that the final workpiece shape cannot be in the desired state based on the prediction of the future shape transition of the workpiece Wα during polishing, the process of step S62 shown in the flowchart of FIG. 12 is performed. That is, the shape transition prediction unit 57A determines whether the final workpiece shape can enter the secondary acceptable state based on the prediction of the future shape transition of the workpiece Wα during polishing.

When it is determined that the final workpiece shape can enter the secondary acceptable state, the process of step S63 shown in the flowchart of FIG. 12 is performed. In other words, the parameter specification unit 59 specifies the parameters responsible for the fact that the final workpiece shape of the workpiece Wα is in a secondary acceptable state during polishing (the final workpiece shape cannot be in the desired state). , or list the responsible parameters in order of highest correlation strength.

Then, when the responsible parameters are specified or listed, the processes from step S71 to step S81 to step S91 shown in the flowchart of FIG. 12 are performed in order. That is, the state determination unit 58A determines the current polishing state of the workpiece Wα during polishing based on the shape transition of the workpiece Wα during polishing, and performs polishing processing based on the determination result of the polishing state. Determine whether to stop. Then, when the state determination unit 58A determines that the polishing state of the workpiece Wα during polishing is a “second polishing stop state,” the display control section 56A determines whether the polishing state of the workpiece Wα during polishing is stopped. In addition to this, the workpiece Wα during polishing is in a “secondary acceptable state” and the responsible parameters specified by the parameter specification unit 59 or the responsible parameters listed in order of high correlation are displayed in the indicator 40. A control command displayed on the screen 40a is output. And, on the screen 40a of the display 40, it is determined that the polishing stop of the workpiece Wα during polishing has been determined, the workpiece Wα during polishing is in a “secondary acceptable state,” and the responsibility parameter or correlation strength is high. Responsible parameters listed in order are respectively displayed, and a polishing stop decision is notified.

After that, by performing the process of step S10, the control operation unit 51A of the control unit 50A sends various outputs for finishing polishing processing, such as a stop control command, to the first drive devices M1 to the fifth drive devices M5. do it As a result, the polishing machine 10 automatically stops after going through a predetermined sequence, and the polishing process of the workpiece Wα during polishing is completed.

As a result, by visually viewing the screen 40a, the user of the polishing machine 10 determines that polishing has stopped and, in addition to the workpiece shape of the workpiece Wα during polishing, the workpiece Wα during polishing is in a secondary allowable state. Responsibility parameters with a high degree of correlation can be identified. As a result, even when the user controls the polishing machine 10 through manual operation to stop polishing, polishing can be stopped at an appropriate timing. Additionally, even when the polishing machine 10 is stopped by a stop control command from the control unit 50A, the user can recognize the stopping operation of the polishing machine 10.

Furthermore, by being able to identify the responsible parameters or their candidates, it is possible to rationally devise necessary measures (improvement of conditional properties, improvement of the polishing machine 10, etc.) to bring the final workpiece shape to the desired state. And, it becomes possible to efficiently obtain the workpiece W in the desired state. In addition, it is possible to urge expansion of experience data according to the degree of discrepancy between the desired final workpiece shape and the actual workpiece shape, and suggestions for improvement of the polishing device 1 to resolve the gap.

On the other hand, in the process of step S62 shown in the flowchart of FIG. 12, when it is determined that the final workpiece shape cannot be brought into the secondary acceptable state, the processes of step S62, step S71, step S81, and step S91 are sequentially performed. That is, the state determination unit 58A determines the current polishing state of the workpiece Wα during polishing, and determines that the polishing state of the workpiece Wα during polishing is the “third polishing stop state” and thus the polishing process is stopped. . Then, the display control unit 56A displays on the screen 40a of the display 40 that the workpiece Wα during polishing has an “unacceptable shape” in addition to the effect that it has determined that polishing of the workpiece Wα during polishing has been stopped. Outputs control commands. Then, on the screen 40a of the display 40, it is displayed that the polishing stop of the workpiece Wα during polishing has been determined and that the workpiece Wα during polishing has an “unacceptable shape”, and the decision to stop polishing is notified. .

After that, by performing the process of step S10, the control operation unit 51A of the control unit 50A sends various outputs for finishing polishing processing, such as a stop control command, to the first drive devices M1 to the fifth drive devices M5. do it As a result, the polishing machine 10 automatically stops after going through a predetermined sequence, and the polishing process of the workpiece Wα during polishing is completed.

As a result, the user of the polishing machine 10 can determine that polishing has been stopped and the shape of the workpiece Wα during polishing by visually viewing the screen 40a. As a result, even when the user controls the polishing machine 10 through manual operation to stop the polishing process, the polishing process can be stopped immediately. Additionally, even when the polishing machine 10 is stopped by a stop control command from the control unit 50A, the user can recognize the stopping operation of the polishing machine 10.

In addition, when a plurality of responsible parameters are specified by the parameter specification unit 59 and these responsible parameters are displayed on the screen 40a of the display 40 in the order of high correlation and strength, the workpiece Wα during polishing It becomes easier to identify the conditional attributes that have the greatest impact on being in this secondary acceptable state. As a result, it becomes possible to more rationally devise necessary measures (improvement of conditional properties, improvement of the polishing machine 10, etc.) to bring the final workpiece shape to the desired state.

Specific examples are given below.

As shown in FIG. 13A, the fifth workpiece W5, which has a “convex sagging shape” whose outer peripheral region is relatively polished at the start of the polishing process, is polished with the polisher 10 of the “first batch after dressing.” Explain. The cross-sectional shape of the fifth workpiece W5 with a "convex sagging shape" in the polishing step A at the beginning of polishing is the polishing step B immediately after the polishing process progresses and the thickness reaches the target thickness range (T1 ≤ thickness ≤ T2). In this case, it has a “flat sagging shape (the central part is flat and the outer peripheral area is excessively polished).” Thereafter, the polishing process continues, and in the polishing step C when the thickness approaches the lower limit T1 of the target thickness range, the cross-sectional shape of the fifth workpiece W5 becomes a desired “flat shape.”

Next, as shown in FIG. 13B, the sixth workpiece W6, which has a “convex sagging shape” whose outer peripheral region has been relatively polished at the start of the polishing process, is polished with the polisher 10 in the “10th batch after dressing.” Explain the case. The cross-sectional shape of the sixth workpiece W6 with a “convex sagging shape” in the polishing step A at the beginning of polishing is the polishing step B immediately after the polishing process progresses and the thickness reaches the target thickness range (T1 ≤ thickness ≤ T2). In this case, it has a “flat sagging shape (the central part is flat and the outer peripheral area is excessively polished).” However, in the polishing step C where the polishing process is further continued and the thickness approaches the lower limit (T1) of the target thickness range, the cross-sectional shape of this sixth workpiece W6 is a “concave sagging shape (largely dented in the center and excessive in the periphery”). It becomes a “polished state”.

Here, when polishing is performed with the polishing machine 10 in the “first batch after dressing,” in the polishing step C, the cross-sectional shape of the fifth workpiece W5 becomes a “flat shape,” so the final workpiece shape is desired. It is predicted that this may be the case. On the other hand, when polishing is performed with the polishing machine 10 “in the 10th batch after dressing,” the cross-sectional shape of the workpiece does not become “flat” no matter which polishing step. Therefore, it is predicted that when polishing the sixth workpiece W6, the final workpiece shape cannot be in the desired state. However, since the “flat/sagging shape” in polishing step B corresponds to the secondary acceptable state, it is predicted that the final workpiece shape of this sixth workpiece W6 may be in the secondary acceptable state.

In addition, a highly correlated conditional attribute (responsible parameter ) is specified as, for example, deterioration of the surface of the polishing pad that occurs as the number of batches increases.

Therefore, in the polishing apparatus 1A of Example 2, when the shape transition of the sixth workpiece W6 is predicted during polishing of the sixth workpiece W6, the final workpiece shape is in the desired state based on this prediction ( It is determined that it cannot be in a flat shape) and can be in a secondary acceptable state (flat/sagging shape). Then, as the polishing process progresses, a polishing stop determination is notified at the timing when the workpiece shape of the sixth workpiece W6 reaches the secondary acceptable state (flat/sagging shape), and the sixth workpiece W6 is in the secondary acceptable state ( Flat, sagging shape) and “surface deterioration of the polishing pad” are notified as a parameter that is highly responsible for entering a secondary acceptable state.

Thereby, the optimal timing for stopping polishing can be appropriately determined, allowing the final workpiece shape to enter a secondary allowable state that satisfies the second shape condition even though it is not in the desired state. In addition, since the responsible parameters or their candidates can be identified, this contributes to the improvement of conditional properties during workpiece polishing and allows the user to plan a more efficient process. Additionally, the polishing device 1 itself can make suggestions for improvement.

Next, the effect is explained.

In the polishing device 1A of Example 2, the effects listed below can be obtained.

(8) When the control unit 50A determines, based on the prediction of the shape transition of the workpiece Wα during polishing, that the workpiece Wα during polishing cannot be brought to the desired workpiece state, the workpiece Wα during polishing When in the secondary allowable state, the polishing process of the workpiece Wα during polishing is stopped, and a decision to stop the polishing process is notified.

As a result, even if the final workpiece shape cannot be in the desired state, the polishing process can be stopped at an appropriate timing and a delay in stopping the polishing process can be prevented. In addition, it is possible to prevent the timing of stopping polishing from being delayed beyond the appropriate time, and to automatically stop polishing at an appropriate timing.

(9) The control unit 50A specifies a conditional attribute (responsible parameter) with a high degree of correlation with respect to the appearance of the secondary acceptance state of the workpiece Wα during polishing, or the appearance of the secondary acceptance state of the workpiece Wα during polishing. The conditional attributes are listed in order of high correlation, and the specified or listed conditional attributes (responsibility parameters or candidates thereof) with high correlation are notified.

As a result, the user can identify the responsible parameters or their candidates and reasonably devise necessary measures to bring the final workpiece shape to the desired state.

Above, the polishing device of the present invention has been described based on Example 1 and Example 2, but the specific configuration is not limited to these Examples and does not deviate from the gist of the invention related to each claim in the scope of the patent claims. Unless otherwise specified, changes or additions to the design are permitted.

In the polishing device 1 of Example 1, an example is shown in which the memory 30 stores the shape information of the workpiece W by associating it with conditional attributes when the workpiece W is polished. However, it is not limited to this. For example, when storing the shape information of the workpiece W in the memory 30, conditional attributes created through learning may be associated with the shape information of the workpiece W. Here, the “learning-generated conditional attribute” is a conditional attribute obtained as a result of learning and calculating the relationship (tendency) between the shape information and the conditional attribute acquired during polishing performed in the past. This “learning-generated conditional attribute” is, for example, based on conditional attributes related to the shape information of the workpiece W accumulated during polishing performed in the past, and is based on the shape information of the workpiece W. Calculations such as learning the tendency of the conditional attribute, automatically calculating the magnitude of influence in the polishing results of each parameter of the conditional attribute from a given complex condition system, and using the result to weight the influence prediction. The conditional properties output, etc. are taken into consideration.

In the case where the shape information of the selected master is extracted by matching the conditional properties of the workpiece Wα during polishing and the conditional properties generated through learning, and the shape transition of the workpiece Wα during polishing is predicted, Users can make more optimal predictions by going beyond the prediction patterns or trends that were easy to make in the past.

In other words, by extracting the shape information of the selected master based on the shape information of the workpiece (W) related to the conditional properties created through learning, the degree of influence of a specific parameter under certain conditions is voluntarily discovered and presented. , a combination of moderate effects can be performed. In addition, depending on how the learning algorithm is structured to combine the degrees of influence, the output purely as a calculation result can expand the prediction range of the shape transition of the workpiece (Wα) during polishing, beyond the range of prediction by the user. .

As a result, the accuracy of predicting the shape of the workpiece Wα during polishing, which was conventionally easy to overlook, is significantly improved compared to the accuracy of prediction by the user. In addition, it becomes possible to objectively predict the correlation between the conditional properties and workpiece shape accuracy under specific conditions with high precision, and in the early stages of polishing or before polishing of the workpiece Wα during polishing, the conditional properties Advance changes can be requested as needed, contributing to improving and stabilizing the yield of good products.

Additionally, in the polishing device 1 of Example 1, the memory 30 stores the shape information of the workpiece W in association with the conditional attributes when the workpiece W is polished. Then, in the shape transition prediction unit 57, the time series change in the shape information of the selected master extracted based on the conditional properties of the workpiece Wα during polishing is compared with the time series change in the shape information of the workpiece Wα during polishing. An example of calculating and predicting the shape transition of the workpiece (Wα) during polishing is shown. That is, in Example 1, the shape information of the shape reference workpiece Wβ or the shape information of the workpiece W is generated based on the learning results of the correlation between the conditional properties when the workpiece W is polished and the shape information of the workpiece W. The shape transition of the workpiece (Wα) during polishing is estimated based on typical shape information, which is a workpiece shape pattern. However, it is not limited to this.

A workpiece shape pattern obtained by processing the shape information of the workpiece W, with support from the correlation data processing unit 60 as needed, and a desired workpiece obtained by processing information having the shape characteristics of the workpiece W. You can also use a shape pattern. In this case, in the memory 30, all the conditional attributes when the workpiece W is polished are associated with the workpiece shape pattern having the shape characteristics of the workpiece W, and the conditional attributes are adjusted as necessary. Group or remember by subdividing according to the progress of learning. In addition, the shape transition prediction unit 57 may predict the shape transition of the workpiece Wα during polishing based on a workpiece shape pattern related to a conditional attribute matching the conditional attribute of the workpiece Wα during polishing. .

That is, information containing at least one of the shape information of the workpiece W and the workpiece shape pattern obtained by calculating the shape information of the workpiece W is called “prediction information,” and the memory 30 stores this prediction information. , the conditional properties when the workpiece (W) is polished and processed, or the conditional properties created through learning are related and memorized. In addition, the shape transition prediction unit 57 of the control unit 50 is related to the time series change in the prediction information of the workpiece Wα during polishing and the conditional attribute matching the conditional attribute of the workpiece Wα during polishing. Based on the results of a comparison operation of time series changes in the prediction information stored in the memory 30, the shape transition of the workpiece Wα during polishing is predicted.

Here, “computation processing” means, for example, averaging the cross-sectional shape lines of a plurality of workpieces W selected for each conditional attribute and obtaining the average cross-sectional shape line for each conditional attribute, or calculating the average cross-sectional shape line for each conditional attribute. Calculating flatness from the difference between the maximum and minimum values of thickness, obtaining the desired cross-sectional shape line using the mode or median value of a plurality of cross-sectional shape lines, and the correlation between the conditional attribute and the shape of the workpiece, Calculating a workpiece shape related to a group of shape features of the workpiece W or a newly statistically and computationally generated group as a typical shape for the group.

That is, for example, the flatness of the central portion of the workpiece W, which is a workpiece shape pattern obtained from shape information of the workpiece W polished in the past, is calculated for each workpiece polishing time. Subsequently, regression analysis and the like are performed, and the relationship between the workpiece polishing time and the flatness of the central portion of the workpiece is generated as a first flatness prediction line La, as shown in FIG. 14A. Then, the shape of the workpiece Wα during polishing may be predicted by comparing this first flatness prediction line La with the change in flatness of the central portion of the workpiece Wα during polishing.

In addition, the flatness of the outer peripheral area of the workpiece W, which is the workpiece shape pattern obtained from the shape information of the workpiece W polished in the past, is calculated for each workpiece polishing time. Subsequently, regression analysis and the like are performed, and the relationship between the workpiece polishing time and the flatness of the outer peripheral region of the workpiece is generated as a second flatness prediction line Lb, as shown in FIG. 14B. Then, the shape of the workpiece Wα during polishing may be predicted by comparing the second flatness prediction line Lb with the flatness trend of the outer peripheral region of the workpiece Wα during polishing.

Meanwhile, from the first flatness prediction line La, it can be seen that the flatness of the central portion of the workpiece gradually decreases with the workpiece polishing time, and deteriorates when the predetermined time Ta is exceeded. Therefore, it is believed that a workpiece with good flatness in the central portion can be obtained by stopping the polishing process of the workpiece Wα during polishing when the workpiece polishing time reaches the predetermined time Ta. In addition, from the second flatness prediction line Lb, the flatness of the outer peripheral area of the workpiece maintains an almost constant value on the negative side until the workpiece polishing time reaches the predetermined time Tb, and It can be seen that when the time (Tb) is exceeded, it becomes positive and gradually increases. Therefore, it is believed that a workpiece with good flatness in the outer peripheral region can be obtained by stopping the polishing process of the workpiece Wα during polishing at the timing when the workpiece polishing time reaches the predetermined time Tb.

Additionally, in Example 1, the second drawing generation unit 55 creates the second drawing P2 based on the shape information of the selection master related to the conditional attribute matching the conditional attribute of the workpiece Wα during polishing. An example of creation is shown. However, the shape information of the workpiece W used when generating the second drawing P2 is not limited to this. For example, regardless of the conditional properties of the workpiece Wα during polishing, it is based on the shape information of the workpiece W that was polished immediately before the workpiece Wα during polishing by a polisher that polishes the workpiece Wα during polishing. Thus, the second drawing (P2) may be generated. In addition, regardless of the conditional properties of the workpiece Wα during polishing, the shape information of the workpiece W that was polished before the workpiece Wα during polishing by the polisher that polished the workpiece Wα during polishing The second drawing P2 may be generated based on .

Additionally, in Examples 1 and 2, the second drawing generation unit 55 generates the second drawing P2 based on the shape information of the selected master extracted based on the conditional properties of the workpiece Wα during polishing. Although an example of generation is shown, it is not limited to this, and a plurality of drawings are created based on the conditional properties of the workpiece Wα during polishing, such as a second drawing based on the shape information of the selected master, a third drawing, etc. Extract the selection masters, generate a plurality of second drawings based on the shape information of these selection masters, and display the plurality of second drawings on the screen 40a of the display 40, or refer to the plurality of second drawings. Thus, it is possible to predict the future shape change of the workpiece Wα during polishing.

In addition, the shape information of the workpiece W stored in the memory 30 is extracted from a desired range, the shape information of the extracted workpiece W is averaged, or specific data is acquired from the shape information of the extracted workpiece W. The second drawing P2 may be generated based on a workpiece shape pattern having shape characteristics obtained by performing computational processing such as generating desired shape information.

Even in this case, the accuracy of the transition of the shape of the workpiece shown by the second drawing P2 can be improved compared to the case where the second drawing P2 is generated using the shape information of the selected master. Therefore, it becomes possible to more accurately predict the shape transition of the workpiece Wα during polishing and to stop polishing at a more appropriate timing.

And, the shape information of the workpiece W is a measurement value obtained by receiving the results of comparison with the measurement value data of the workpiece shape by a shape measurement dedicated machine installed separately from the polisher 10 (e.g., a separate flatness measurement machine, etc.) The data may be corrected and adjusted as necessary.

Additionally, in the polishing device 1 of Example 1, the first drawing P1 and the second drawing P2 are generated respectively, and the first drawing P1 and the second drawing P2 are displayed on the display 40. An example of simultaneous display is shown. However, it is not limited to this. Only the first drawing P1, which arranges the shape drawing of the workpiece Wα during polishing in time series, may be displayed on the display 40. Even in this case, the user can determine the change in shape of the workpiece Wα during polishing. Then, based on the change in shape of the workpiece Wα during polishing, the polishing process of the workpiece Wα during polishing can be stopped at the timing when the desired workpiece shape is achieved.

Additionally, in the polishing device 1 of Example 1, when it is determined that polishing of the workpiece Wα during polishing has been stopped, this decision to stop polishing is displayed on the indicator 40 to notify the user, and the polisher 10 is stopped. An example was shown. However, for example, it is sufficient to simply notify that a decision to stop the polishing process has been made, or to stop the polishing process of the workpiece Wα during polishing by controlling the polisher 10 to stop without notifying the stop decision.

Additionally, in Example 2, when it is determined that the final workpiece shape cannot be in the desired state based on the prediction of the shape transition of the workpiece Wα during polishing, when the workpiece Wα during polishing is in a secondary acceptable state, this An example in which the decision to stop polishing processing is displayed on the indicator 40 to notify the user and the polishing machine 10 is stopped is shown. However, when the workpiece Wα during polishing is in a secondary allowable state, for example, it is sufficient to notify that a stop determination of polishing processing has been made, and the polisher 10 is controlled to stop without notifying the stop decision, and the workpiece Wα during polishing is ) is sufficient just to stop the polishing process.

In Example 1, an example in which the measurement unit 21 is attached to the upper table 12 is shown, but the present invention is not limited to this. For example, laser light, which is measurement light, may be irradiated from an optical head installed above the top plate 12. In this case, a plurality of measurement holes are formed along the circumferential direction of the upper plate 12, and the laser light is irradiated each time each measurement hole comes directly under the optical head by rotation of the upper plate 12, Measure the thickness of the workpiece. On the other hand, a measurement hole may be formed in the lower surface plate 11, and the thickness may be measured by irradiating a laser beam to the lower surface of the workpiece W from below the lower surface plate 11.

In addition, in Example 1, when obtaining the cross-sectional shape of the workpiece W, the thickness data of the obtained data string is averaged by moving average processing or polynomial approximation curve drawing processing, etc., but this is not limited to this and the Any method may be used as long as the cross-sectional shape can be visualized.

Additionally, in Examples 1 and 2, a second drawing generation process is executed in parallel with the polishing stop determination process, changes in the conditional properties of the workpiece Wα during polishing are monitored, and the second drawing P2 is appropriately created. An example of replacing is shown. However, it is not limited to this, and for example, the second drawing P2 once generated based on the conditional attributes of the workpiece Wα during polishing acquired after determining that polishing is in progress may be maintained until the end of polishing. Additionally, in this case, it is not necessary to perform the second drawing generation process in parallel with the polishing stop determination process, and it may be performed during the polishing stop determination process (e.g., between steps S1 and S2, between steps S3 and S4, etc.). , each process of step S12, step S13, and step S14 in the second drawing generation process may be executed.

In addition, in Examples 1 and 2, a double-sided polishing device having a lower surface plate 11 and an upper surface plate 12 was shown and capable of simultaneously polishing both sides of the workpiece W, but only one side of the workpiece W The present invention can be applied even if it is a single-side polishing device that polishes only one side.

1, 1A: polishing device 10: polishing machine
11: Lower class 12: Upper class
20: shape measuring device 19: measuring hole
30: memory 40: indicator
40a: screen 50, 50A: control unit
51, 51A: control operation unit 54: first drawing generation unit
55: second drawing generation unit 56, 56A: display control unit
57, 57A: Shape trend prediction unit 58, 58A: Status determination unit
59: Parameter specification unit 60: Correlation data processing unit

Claims (12)

A polishing machine that polishes a workpiece with a rotating surface,
A shape measuring device that measures the shape of the workpiece through a measuring hole formed in the surface,
a memory that stores shape information of the workpiece measured by the shape measuring device;
an indicator that displays shape information of the workpiece measured by the shape measuring device;
Provided with a control unit that controls the display contents of the indicator,
The control unit generates a first drawing that displays a change trend of a plurality of shape drawings for a workpiece currently being polished, which is a workpiece currently being polished, as measured by the shape measuring device, over time, and generates the first drawing. A polishing device characterized in that the display is displayed on the indicator. According to clause 1,
The memory includes conditional attributes or learning-generated information when the workpiece is polished, in prediction information including at least one of the shape information of the workpiece and the workpiece shape pattern obtained by computationally processing the shape information of the workpiece. Remember to associate conditional properties,
The control unit is configured to compare the time series change in the prediction information of the workpiece during polishing and the time series change in prediction information related to the conditional attribute matching the conditional attribute of the workpiece during polishing. A polishing device, characterized in that it predicts the shape transition and determines the state of the workpiece during polishing based on the prediction of the shape transition of the workpiece during polishing. According to clause 2,
When the control unit determines that polishing of the workpiece during polishing is to be stopped as a result of the determination of the state of the workpiece during polishing, the control unit notifies the stop of polishing processing of the workpiece during polishing and the decision to stop polishing processing of the workpiece during polishing. A polishing device characterized in that it performs at least one of the following. According to clause 2,
When the control unit determines that the workpiece during polishing cannot be brought into a desired workpiece state based on the prediction of the shape transition of the workpiece during polishing, if the workpiece during polishing is in a secondary allowable state, the control unit determines that the workpiece during polishing cannot be in a desired workpiece state. A polishing device, characterized in that it performs at least one of the following: stop and notification of a decision to stop polishing of a workpiece being polished. According to clause 4,
The control unit specifies a conditional attribute having a correlation greater than or equal to a predetermined correlation with respect to the appearance of a secondary acceptable state of the workpiece during polishing, or a correlation value with respect to the appearance of a secondary acceptable state of the workpiece during polishing. A polishing device characterized in that it lists conditional attributes in high order and notifies specific or listed conditional attributes. A polishing machine that polishes a workpiece using a rotating surface plate,
A shape measuring device that measures the shape of the workpiece through a measuring hole formed in the surface,
a memory that stores shape information of the workpiece measured by the shape measuring device;
an indicator that displays shape information of the workpiece measured by the shape measuring device;
Provided with a control unit that controls the display contents of the indicator,
The control unit is configured to display a first drawing that displays a change trend of a plurality of shape drawings of a workpiece currently being polished, which is a workpiece currently being polished, as measured by the shape measuring device over time, and Generating a second drawing that displays the change trend of a plurality of shape drawings over time for a workpiece different from the workpiece during polishing that was previously polished, and simultaneously drawing the first drawing and the second drawing. A polishing device characterized in that the display is displayed on the indicator. According to clause 6,
The memory stores the shape information of the workpiece by associating it with a conditional attribute when the workpiece is polished or a learning-generated conditional attribute,
The control unit generates the second drawing based on shape information of the workpiece related to conditional properties that match the conditional properties of the workpiece during polishing. According to claim 6 or 7,
The control unit is based on a workpiece shape pattern obtained by processing the shape information of the workpiece stored in the memory, or a learning result of the correlation between the conditional attribute when the workpiece is polished and the shape information of the workpiece. A polishing device, characterized in that the second drawing is generated based on the workpiece shape pattern created. According to claim 6 or 7,
The memory includes conditional attributes or learning-generated information when the workpiece is polished, in prediction information including at least one of the shape information of the workpiece and the workpiece shape pattern obtained by computationally processing the shape information of the workpiece. Remember to associate conditional properties,
The control unit is configured to compare the time series change in the prediction information of the workpiece during polishing and the time series change in prediction information related to the conditional attribute matching the conditional attribute of the workpiece during polishing. A polishing device, characterized in that it predicts a shape transition and determines the state of the workpiece during polishing based on the prediction of the shape transition. According to clause 9,
When the control unit determines that polishing of the workpiece during polishing is to be stopped as a result of the determination of the state of the workpiece during polishing, the control unit notifies the stop of polishing processing of the workpiece during polishing and the decision to stop polishing processing of the workpiece during polishing. A polishing device characterized in that it performs at least one of the following. According to clause 9,
When the control unit determines that the workpiece during polishing cannot be brought into a desired workpiece state based on the prediction of the shape transition of the workpiece during polishing, if the workpiece during polishing is in a secondary allowable state, the control unit determines that the workpiece during polishing cannot be in a desired workpiece state. A polishing device, characterized in that it performs at least one of the following: stop and notification of a decision to stop polishing of a workpiece being polished. According to clause 11,
The control unit specifies a conditional attribute having a correlation greater than or equal to a predetermined correlation with respect to the appearance of a secondary acceptable state of the workpiece during polishing, or a correlation value with respect to the appearance of a secondary acceptable state of the workpiece during polishing. A polishing device characterized in that it lists conditional attributes in high order and notifies specific or listed conditional attributes.

Images