JP7613133B2 - Surface Texture Estimation System - Google Patents

Description

The present invention relates to a surface texture estimation system.

Grinding is performed, for example, by bringing a rapidly rotating tool into contact with a rotating workpiece. When a workpiece is machined with a rotating tool, if chatter vibration occurs, the machining surface accuracy may decrease and an excessive load may act on the tool. Conventionally, a method has been used to detect the occurrence of chatter vibration during machining by checking the surface condition of the workpiece after grinding. The surface condition of the workpiece is measured by a roundness measuring device after grinding is completed. Because the grinding device and the surface condition measuring device are separated, even if chatter vibration is found to be occurring on the workpiece surface, there is a problem of a time lag in feeding this back to the machining conditions, etc.

In response to this, a method has been proposed in which chatter vibration occurrence is detected in-process, as in Patent Document 1. A chatter vibration detector measures, for example, the vibration acceleration and vibration displacement of the grinding machine or the workpiece, and determines that chatter vibration has occurred when vibration exceeding a predetermined threshold is detected. By detecting the occurrence of chatter vibration on the grinding machine, if chatter vibration is detected, the processing conditions can be changed to suppress the chatter vibration.

JP 2000-233368 A

The chatter detection method of Patent Document 1 makes it possible to know whether chatter is occurring or not, but in order to more accurately grasp the quality of the workpiece and the wear state of the grinding wheel, it is desirable to know the surface properties of the workpiece over a certain range or more. However, measuring the surface properties of a workpiece takes time, and it is particularly difficult to obtain the surface properties in-process.

The present invention aims to provide a surface quality estimation system that quickly and accurately estimates the surface quality of a workpiece during the in-process grinding process.

The surface texture estimation system includes a detection device that outputs detection data corresponding to the surface condition of a workpiece ground by a grinding wheel in a grinding device, and a surface texture generation device that generates the surface texture of the workpiece based on the detection data detected by the detection device by moving the measurement position of the detection device on the surface of the workpiece at least in the circumferential direction relative to the workpiece. The surface texture generation device calculates a surface texture caused by the grinding wheel, which is a transfer of the surface condition of the grinding surface of the grinding wheel, based on first detection data, which is detection data detected by moving the measurement position on the surface of the workpiece in a spiral shape in the circumferential and axial directions of the workpiece, and second detection data, which is detection data detected by moving the measurement position in the circumferential direction at the same position in the axial direction of the workpiece, and calculates a surface texture caused by center-to-center relative vibration, which is a transfer of vibration generated by relative center-to-center fluctuation between the grinding wheel and the workpiece, based on the second detection data, and generates the surface texture of the workpiece by adding the surface texture caused by the grinding wheel and the surface texture caused by the center-to-center relative vibration.

According to the surface texture estimation system, the first detection data detected in a spiral shape and the second detection data detected at the same axial position can be used to calculate the surface texture caused by the grinding wheel and the surface texture caused by the intercenter relative vibration. This makes it possible to reduce the number of first detection data pieces required to calculate the surface condition of the workpiece compared to when only the first detection data is used, for example, and as a result, the time required to generate and estimate the surface texture of the workpiece can be shortened. In addition, because the surface texture of the workpiece can be generated using the first detection data and the second detection data, the surface texture of the workpiece can be generated with high accuracy even when the number of data pieces is reduced.

FIG. 2 is a plan view showing the configuration of the grinding device. FIG. 2 is a diagram for explaining a detection device. 4 is a flowchart showing a grinding process of the grinding device. FIG. 2 is a diagram for explaining the surface properties of a workpiece. FIG. 1 is a diagram for explaining surface properties caused by a grindstone. FIG. 1 is a diagram for explaining surface properties caused by a grindstone. FIG. 13 is a diagram for explaining surface texture caused by intercenter relative vibration. FIG. 1 is a block diagram showing a configuration of a surface texture estimation system. 11 is a diagram for explaining detection of first basic data (first detection data). FIG. 13 is a diagram for explaining detection of second basic data (second detection data). FIG. FIG. 2 is a block diagram showing a configuration of a first data processing unit of the surface texture generating device. FIG. 2 is a block diagram showing a configuration of a second data processing unit of the surface texture generating device. FIG. 2 is a block diagram showing a configuration of a mapping processing unit of the surface texture generating device. FIG. 13 is a diagram showing surface characteristics due to a grindstone mapped by a mapping processing unit. 11A and 11B are diagrams for explaining a data position correction process performed by a mapping processor; 11 is a diagram for explaining a data synthesis process performed by a mapping processor; FIG. FIG. 13 is a diagram showing surface characteristics due to a grindstone mapped by a mapping processing unit. FIG. 13 is a diagram showing surface texture (high frequency components) caused by intercenter relative vibration mapped by a mapping processor. FIG. 13 is a diagram showing surface texture (low frequency components) caused by intercenter relative vibration mapped by a mapping processor. FIG. 4 is a diagram showing the surface texture of a workpiece mapped by a mapping processing unit.

The surface texture estimation system H will be described below with reference to the drawings. As shown in FIG. 1, the surface texture estimation system H includes a grinding device 10, a surface texture generation device 20, and a detection device 30. The surface texture estimation system H of this example also includes an image output device 40. In the surface texture estimation system H of this example, the detection device 30 detects a detectable displacement value after or during grinding by the grinding device 10, and the surface texture generation device 20 uses detection data corresponding to the surface state of the workpiece W detected by the detection device 30 to map and output the surface texture of the workpiece W. The image output device 40 of this example then outputs the map representing the surface texture of the workpiece W generated by the surface texture generation device 20 as an image.

Here, examples of the detection data detected by the detection device 30 include displacement (amplitude) obtained by conversion based on the acceleration of vibrations occurring in response to the surface state (surface shape) of the workpiece W, and the displacement (amplitude) of vibration. The detection device 30 is capable of detecting detection data by directly or indirectly contacting the workpiece W, and is also capable of detecting detection data without contacting the workpiece W (non-contact). In this example, an example is shown in which the detection device 30 detects a displacement obtained by conversion based on the acceleration on the surface of the workpiece W, and a case is described in which the surface texture generating device 20 generates a surface texture, i.e., a map, using the detected displacement.

(1. Configuration of Grinding Apparatus 10)
1 and 2, the grinding device 10 mainly comprises a bed 11, a grinding wheel 12, a grinding wheel head 13, a headstock 14, a tailstock 15, a spindle table 16, and a sizing device 17. A workpiece W is supported at both ends in the rotational axis direction by the headstock 14 and the tailstock 15 and rotates. Note that in this example, the workpiece W is cylindrical or columnar. The grinding device 10 brings the grinding wheel 12 into contact with the surface (outer circumferential surface) of the rotating workpiece W and grinds it to form the shape of the workpiece W.

The grinding wheel 12 is supported by the wheel head 13 so as to be rotatable around an axis parallel to the Z axis. A wheel head guide 11a is fixed on the bed 11, and the wheel head 13 is supported by the wheel head guide 11a so as to be movable in the X axis direction. A rotational drive force is applied to the grinding wheel 12 from a wheel rotation motor 12a controlled by a controller 18, causing the grinding wheel 12 to rotate around the rotation axis. As the wheel head 13 moves in the X axis direction, the grinding wheel 12 approaches the workpiece W placed at a distance in the X axis direction and grinds the workpiece W.

On the bed 11, the spindle table guide 11b is fixed at a position separated from the wheel head guide 11a in the X-axis direction. The spindle table guide 11b supports the spindle table 16 so that it can move in the Z-axis direction. The spindle head 14 and tailstock 15 are arranged opposite each other on the spindle table 16. The workpiece W is rotatably supported at both ends by the spindle head 14 and tailstock 15, and is rotated by the rotational driving force applied from the spindle rotation motor 14a controlled by the controller 18.

As shown in FIG. 2, the sizing device 17, which is an outer diameter measuring device, has a pair of probes 17a that are contact parts that come into contact with the surface of the workpiece W, and a pair of fingers 17b that support the probes 17a. The probes 17a are arranged to abut against the surface of the workpiece W at two points on either side of the center of rotation O of the workpiece W. The sizing device 17 detects the outer diameter of the workpiece W by converting the mechanical displacement of the probes 17a into an electrical signal. The sizing device 17 is supported by an axial movement device 17c and can move in the axial direction of the workpiece W, i.e., in the Z-axis direction. The movement of the sizing device 17 in the Z-axis direction is controlled by an axial movement control unit 17d.

The detection device 30 is attached to at least one of the pair of fingers 17b of the sizing device 17 of the grinding device 10. The detection device 30 in this example is mainly equipped with an acceleration sensor, and detects the displacement obtained by conversion based on the acceleration. In other words, the detection device 30 detects the displacement that occurs in the finger 17b when the measuring probe 17a of the sizing device 17 moves relative to the workpiece W while in contact with the surface of the workpiece W.

The grinding device 10 grinds the workpiece W through the process shown in FIG. 3. The grinding process is divided according to the difference in the grindstone feed speed, and is performed in the order of rough grinding process St1, fine grinding process St2, fine grinding process St3, and spark-out process St4. The grindstone feed speed in each process is rough grinding process St1>fine grinding process St2>fine grinding process St3>spark-out process St4. In the rough grinding process St1, the rough shape of the workpiece W is formed. In the subsequent fine grinding process St2 and fine grinding process St3, the grindstone feed speed is reduced while the surface shape of the workpiece W is adjusted. In the final spark-out process St4, the surface of the workpiece W is finished to complete the workpiece W.

Here, it is preferable for the surface texture estimation system H to estimate the surface texture of the workpiece W after the spark-out process St4 when grinding is completed. The surface texture estimation system H estimates the surface texture of the workpiece W in-process, which refers to the period until the workpiece W is removed from the grinding device 10 and includes the period after the spark-out process St4. It is preferable for the surface texture estimation system H to estimate the surface texture S of the workpiece W, which will be described later, in a state in which the rotation of the workpiece W during grinding is maintained after grinding of the workpiece W is completed.

(2. Overview of Surface Texture Generation Device 20)
Next, the configuration of the surface texture generating device 20 will be described. As shown in Fig. 4, it can be said that the surface texture S of the workpiece W ground by the grinding device 10 is determined due to various factors. That is, it can be said that the surface texture S of the workpiece W is a combination of a grindstone-induced surface texture S1 to which the surface state of the grinding surface of the grinding wheel 12 is transferred, as shown by the solid line and the two-dot chain line in Fig. 4, and a center-to-center relative vibration-induced surface texture S2 to which vibrations generated by relative fluctuations between the grinding wheel 12 and the workpiece W, i.e., the center-to-center, are transferred, as shown by the dashed line in Fig. 4.

In other words, the surface texture S of the workpiece W can be separated into surface texture S1 caused by the grinding wheel and surface texture S2 caused by the intercenter relative vibration. Therefore, surface texture S1 and surface texture S2 are generated separately, and finally, surface texture S1 and surface texture S2 are synthesized (added) to generate the final surface texture S.

The surface texture S1 caused by the grinding wheel can be further separated into a surface texture S11 in which the grinding wheel surface irregularities are transferred, and a surface texture S12 at the static workpiece reference radius, as shown in Figures 5 and 6. The surface texture S2 caused by the intercenter relative vibration contains low-frequency and high-frequency components, as shown in Figure 7, and can be separated into a surface texture S21 caused by the high-frequency intercenter relative vibration waveform, which is a high-frequency component, and a surface texture S22 caused by the low-frequency intercenter relative vibration waveform, which is a low-frequency component, by extracting them from the frequency characteristics of the detection data by the detection device 30.

Therefore, the surface texture S1 can be generated by combining (adding) the surface textures S11 and S12, the surface texture S2 can be generated by combining (adding) the surface textures S21 and S22, and finally the surface texture S can be generated by further combining (adding) the surface textures S1 and S2. Also, by appropriately selecting the surface textures S1, S11, S12, S2, S21, and S22 to be combined as required, it is possible to arbitrarily generate the required surface texture S.

The surface texture generating device 20 in this example acquires the displacement detected by the detection device 30 as detection data, and extracts low-frequency and high-frequency components from the frequency characteristics of the acquired detection data. The surface texture generating device 20 then performs various data processing on the extracted low-frequency and high-frequency components, and generates maps representing the surface textures S11, S12 (i.e., surface texture S1) and surface textures S21, S22 (i.e., surface texture S2) based on the results of the various data processing. The surface texture generating device 20 then synthesizes (adds) the maps representing the surface textures S1 and S2 to finally generate the surface texture S of the workpiece W, specifically, a map representing the surface texture S.

(2-1. Configuration of Surface Texture Generation Device 20)
As shown in FIG. 8, the surface texture generating device 20 includes a basic data acquiring unit 21, a first data processing unit 22, a second data processing unit 23, and a mapping processing unit 24 serving as a surface texture generating unit.

(2-2. Basic Data Acquisition Unit 21)
The basic data acquiring unit 21 acquires detection data (displacement) detected by the detecting device 30 after or during grinding. Specifically, as shown in Fig. 8, the basic data acquiring unit 21 acquires first detection data K1 output from the detecting device 30 as first basic data D1, and acquires second detection data K2 as second basic data D2.

Here, the detection device 30 first detects the first detection data K1 when the measurement position for measuring the vibration displacement on the surface of the workpiece W is moved spirally in the circumferential and axial directions relative to the workpiece W as shown in FIG. 9, and outputs the first detection data K1 to the basic data acquisition unit 21. That is, when acquiring the first basic data D1, the measuring probe 17a of the sizing device 17 is brought into contact with the surface of the workpiece W while the workpiece W is rotated, and the sizing device 17 is continuously moved in the axial direction of the workpiece W by the axial movement device 17c. Here, the measurement position in this example is the position where the measuring probe 17a of the sizing device 17 is in contact with the surface of the workpiece W. It is preferable that the moving speed of the sizing device 17 in the axial direction of the workpiece W is about 1 mm per rotation of the workpiece W, and in this case, the moving speed of the sizing device 17 can be controlled by the axial movement control unit 17d.

As shown in FIG. 10, the detection device 30 detects second detection data K2 for one revolution of the outer circumferential surface of the workpiece W at the same measurement position in the axial direction (same axial position) without moving the measurement position in a spiral manner, and outputs the data to the basic data acquisition unit 21. That is, while the workpiece W is rotating, the probe 17a of the sizing device 17 is brought into contact with the surface of the workpiece W, and the axial movement device 17c stops the sizing device 17 at the same axial position of the workpiece W.

The basic data acquisition unit 21 acquires the first detection data K1 detected in a spiral shape as the first basic data D1. The basic data acquisition unit 21 also acquires one revolution of the second detection data K2 acquired at the same axial position as the second basic data D2. The basic data acquisition unit 21 then outputs the first basic data D1 and the second basic data D2 to the first data processing unit 22 and the second data processing unit 23, respectively.

Here, the first basic data D1 and the second basic data D2 are time series data related to displacement. Note that the first basic data D1 and the second basic data D2 are generally acquired as data based on a time axis, but may be converted from time and the rotational speed of the workpiece W into data based on the rotational angle of the workpiece W.

(2-3. First Data Processing Unit 22)
The first data processing unit 22 extracts low-frequency components from the frequency characteristics of the first basic data D1 and the second basic data D2, and performs various data processing described later on the extracted low-frequency components. For this reason, the first data processing unit 22 mainly includes a low-frequency component extraction unit 221, a spiral low-frequency waveform generation unit 222, a low-frequency center-to-center relative vibration waveform generation unit 223, and a workpiece reference radius calculation unit 224, as shown in FIG.

The low-frequency component extraction unit 221 performs a fast Fourier transform (hereinafter referred to as "FFT") on the first basic data D1 acquired from the basic data acquisition unit 21, and extracts a low-frequency component D11, which is a first low-frequency component and excludes the 1 peak/period component, from the frequency characteristics of the first basic data D1. The 1 peak/period component is a component corresponding to the rotation frequency of the workpiece W. The low-frequency component extraction unit 221 also performs an FFT on the second basic data D2 acquired from the basic data acquisition unit 21, and extracts a low-frequency component D21, which is a second low-frequency component and excludes the 1 peak/period component, from the frequency characteristics of the second basic data D2. The 1 peak/period component is excluded here because, for example, when the rotation axis of the workpiece W is misaligned, the 1 peak/period component may be strongly detected. The low-frequency component extraction unit 221 extracts low-frequency components from the first basic data D1 and the second basic data D2 that are in a frequency range of less than 50 Hz (approximately 15 to 50 peaks in the waveform), for example.

The spiral low-frequency waveform generating unit 222 performs an inverse fast Fourier transform (hereinafter referred to as "inverse FFT") on the low-frequency components D11 of the first basic data D1 extracted by the low-frequency component extracting unit 221. Here, the first basic data D1 is the first detection data K1 (displacement) detected in a spiral shape along the outer circumferential surface (surface) of the workpiece W by the detecting device 30. As a result, the spiral low-frequency waveform generating unit 222 generates a spiral low-frequency waveform SLW that represents the waveform of the low-frequency components D11 of the displacement fluctuation, i.e., vibration, in the spiral direction of the workpiece W.

The low-frequency intercenter relative vibration waveform generating unit 223 performs an inverse FFT on the low-frequency component D21 of the second basic data D2 extracted by the low-frequency component extracting unit 221. Here, the second basic data D2 is the second detection data K2 (displacement) detected by the detection device 30 at the same position in the axial direction of the workpiece W. As a result, when an inverse FFT is performed on the low-frequency component D21 of the second basic data D2, a one-section low-frequency waveform is obtained that represents the displacement fluctuation in the circumferential direction (one revolution) of the workpiece W, i.e., the low-frequency component D21 of the vibration.

The single cross-section low-frequency waveform represents the relative vibration (low-frequency intercenter relative vibration) that occurs due to changes in the relative position, i.e., the center distance, between the grinding wheel 12 and the workpiece W, which does not change significantly during grinding of one workpiece W, such as pump pulsation in the grinding device 10 or the setting accuracy of the workpiece W, and can be considered to be the same for one workpiece W along the axial direction of the workpiece W. Therefore, the low-frequency intercenter relative vibration waveform generating unit 223 generates the single cross-section low-frequency waveform obtained by performing the inverse FFT as the low-frequency intercenter relative vibration waveform LDV.

The workpiece reference radius calculation unit 224 uses the spiral low-frequency waveform SLW generated by the spiral low-frequency waveform generation unit 222 and the low-frequency intercenter relative vibration waveform LDV generated by the low-frequency intercenter relative vibration waveform generation unit 223 to calculate the workpiece reference radius R resulting from the transfer of the surface condition of the grinding surface of the grinding wheel 12 to the outer peripheral surface of the ground workpiece W. Specifically, the workpiece reference radius calculation unit 224 calculates the workpiece reference radius R by subtracting the low-frequency intercenter relative vibration waveform LDV from the spiral low-frequency waveform SLW.

Here, as described above, the low-frequency intercore relative vibration waveform LDV is a single cross-sectional low-frequency waveform considered to be the same in the axial direction of the workpiece W. For this reason, the workpiece reference radius calculation unit 224 calculates the workpiece reference radius R by adding (copying) the low-frequency intercore relative vibration waveform LDV a number of times equal to the number of spirals C of the spiral low-frequency waveform SLW and subtracting it from the spiral low-frequency waveform SLW in accordance with the following formula 1.
R=SLW-C×LDV...Formula 1

(2-4. Second Data Processing Unit 23)
The second data processing unit 23 extracts high-frequency components from the frequency characteristics of the first basic data D1 and the second basic data D2, and performs various data processing described later on the extracted high-frequency components. For this reason, the second data processing unit 23 mainly includes a spiral high-frequency component extractor 231, a single cross-section high-frequency component extractor 232, a spiral high-frequency waveform generator 233, a high-frequency intercenter relative vibration waveform generator 234, and a grindstone surface unevenness calculator 235, as shown in FIG.

The spiral high-frequency component extraction unit 231 performs an FFT on the first basic data D1 acquired from the basic data acquisition unit 21, and extracts the first high-frequency component from the frequency characteristics of the first basic data D1 as the spiral high-frequency component D12. Here, the first basic data D1 is the first detection data K1 (displacement) detected in a spiral shape along the outer peripheral surface of the workpiece W by the detection device 30. In addition, the spiral high-frequency component extraction unit 231 extracts, as the spiral high-frequency component D12, the frequency characteristics of the high-frequency component, for example, in a frequency range of 50 Hz or more and less than the upper limit of detection by the detection device 30 (approximately 50 to 500 peaks in the waveform).

The single cross-section high frequency component extraction unit 232 performs FFT on the second basic data D2 acquired from the basic data acquisition unit 21, and extracts high frequency components from the frequency characteristics of the second basic data D2. Furthermore, the single cross-section high frequency component extraction unit 232 extracts high frequency components, i.e., second high frequency components, obtained by excluding the grinding wheel rotation frequency component corresponding to the rotation speed of the grinding wheel 12 from the extracted high frequency components, as the single cross-section high frequency components D22. The above-mentioned grinding wheel rotation frequency components are frequency components consisting of the grinding wheel rotation frequency and its harmonics.

Here, the second basic data D2 is the second detection data K2 (displacement) detected by the detection device 30 at the same position in the axial direction of the workpiece W. As a result, the high-frequency components extracted from the second basic data D2 correspond to one revolution in the circumferential direction of the workpiece W, i.e., one cross section of the workpiece W. The one-cross section high-frequency component extraction unit 232 also extracts, as the high-frequency components, a frequency range of, for example, 50 Hz or more and less than the upper limit of detection by the detection device 30 (approximately 50 to 500 peaks in the waveform) as the one-cross section high-frequency components D22.

The spiral high-frequency waveform generating unit 233 performs an inverse FFT on the spiral high-frequency component D12 of the first basic data D1 extracted by the spiral high-frequency component extracting unit 231. As a result, the spiral high-frequency waveform generating unit 233 generates a spiral high-frequency waveform SHW that represents the waveform of the spiral high-frequency component D12 of the displacement fluctuation, i.e., the vibration, in the spiral direction of the workpiece W.

The high frequency center-to-center relative vibration waveform generating unit 234 performs an inverse FFT on the single cross-section high frequency component D22 of the second basic data D2 extracted by the single cross-section high frequency component extracting unit 232. As a result, when the inverse FFT is performed on the single cross-section high frequency component D22 obtained by excluding the grinding wheel rotation frequency component from the high frequency components of the second basic data D2, a single cross-section high frequency waveform is obtained that represents the displacement fluctuation in the circumferential direction (one revolution) of the workpiece W, i.e., the single cross-section high frequency component D22 of the vibration.

The single cross-section high-frequency component D22 does not include the grinding wheel rotation frequency component corresponding to the rotation speed of the grinding wheel 12. Therefore, the single cross-section high-frequency waveform represents vibrations that affect the surface quality S of the workpiece W (more specifically, the surface quality S2 caused by the center-to-center relative vibration) in addition to the grinding wheel rotation frequency component corresponding to the rotation speed of the grinding wheel 12. Here, examples of vibrations that affect the surface quality S of the workpiece W include the rotation of the servo motor that controls the movement of the grinding wheel head 13 and the spindle table 16, externally applied vibrations, self-excited chatter, etc.

Therefore, the single cross-section high-frequency waveform represents the relative vibration (high-frequency inter-center relative vibration) that occurs due to changes in the relative position between the grinding wheel 12 and the workpiece W, i.e., the center-to-center distance, in the high-frequency region, and, like the single cross-section low-frequency waveform, can be considered to be the same for one workpiece W along the axial direction of the workpiece W. Therefore, the high-frequency inter-center relative vibration waveform generating unit 234 generates the single cross-section high-frequency waveform obtained by performing an inverse FFT as the high-frequency inter-center relative vibration waveform HDV.

The grinding wheel surface unevenness calculation unit 235 uses the spiral high-frequency waveform SHW generated by the spiral high-frequency waveform generation unit 233 and the high-frequency inter-center relative vibration waveform HDV generated by the high-frequency inter-center relative vibration waveform generation unit 234 to calculate the grinding wheel surface unevenness P caused by the surface condition of the grinding surface of the grinding wheel 12 being transferred to the outer peripheral surface of the ground workpiece W. Specifically, the grinding wheel surface unevenness calculation unit 235 calculates the grinding wheel surface unevenness P by subtracting the high-frequency inter-center relative vibration waveform HDV from the spiral high-frequency waveform SHW.

Here, as described above, the high frequency intercore relative vibration waveform HDV is a single cross-sectional high frequency waveform that is considered to be the same in the axial direction of the workpiece W. For this reason, the grindstone surface unevenness calculation unit 235 calculates the grindstone surface unevenness P by adding (copying) the high frequency intercore relative vibration waveform HDV a number of times that matches the number of spirals C of the spiral high frequency waveform SHW and subtracting it from the spiral high frequency waveform SHW in accordance with the following formula 2.
P=SHW-C×HDV...Formula 2

(2-5. Mapping Processing Unit 24)
The mapping processor 24 generates maps M1, M2, M3, and M4 corresponding to the above-mentioned surface textures S11, S12, S21, and S22 (see FIGS. 4-7), respectively, based on the above-mentioned various data processing results by the first data processor 22 and the second data processor 23. Then, the mapping processor 24 generates a surface texture map M representing the surface texture S of the workpiece W by combining (adding) the generated maps M1, M2, M3, and M4.

As shown in FIG. 13, the mapping processing unit 24 mainly includes a grinding wheel surface unevenness map generating unit 241, a workpiece reference radius map generating unit 242, a high-frequency intercenter relative vibration map generating unit 243, a low-frequency intercenter relative vibration map generating unit 244, and a surface texture map generating unit 245.

The grindstone surface unevenness map generator 241 acquires the grindstone surface unevenness P from the second data processor 23 (grindstone surface unevenness calculator 235). Then, as shown in FIG. 14, the grindstone surface unevenness map generator 241 generates a map M1 that represents the surface quality S11 due to the grindstone surface unevenness P caused by the grindstone.

The grinding wheel surface unevenness P output from the grinding wheel surface unevenness calculation unit 235 is calculated using the spiral high-frequency waveform SHW generated by the spiral high-frequency waveform generation unit 233, as described above. The spiral high-frequency waveform SHW is based on the first basic data D1 detected in a spiral shape. Therefore, since the waveform representing the grinding wheel surface unevenness P continues in a spiral shape along the outer peripheral surface of the workpiece W, when generating the map M1, it is necessary to divide the waveform representing the grinding wheel surface unevenness P into division sections including, for example, an arbitrary number of peaks (specifically, a multiple of the grinding wheel rotation period, etc.), and to arrange each division section in the axial direction of the workpiece W.

In this way, when the grinding wheel surface irregularities P are divided into any number of divisions, the grinding wheel surface irregularities P1 of each division (hereinafter referred to as "divided grinding wheel surface irregularities P1") will have different angles relative to the rotation axis of the workpiece W. That is, as shown in FIG. 15, the divided grinding wheel surface irregularities P1 at adjacent axial positions correspond to positions shifted from each other in the circumferential direction of the workpiece W. Here, the grinding wheel surface irregularities P of the workpiece W are due to the surface condition of the grinding surface of the grinding wheel 12, and therefore are repeatedly seen on the surface of the workpiece W with each grinding wheel rotation cycle of the grinding wheel 12.

The grindstone surface unevenness map generating unit 241 of this example therefore considers the grindstone surface unevenness on the same circumference of the workpiece W to be repeated periodically. That is, the grindstone surface unevenness map generating unit 241 of this example considers that when divided grindstone surface unevenness P1 for different division sections, i.e., different angles, are arranged in the axial direction, they are divided grindstone surface unevenness P1 at the same angle of the workpiece W. For this reason, the grindstone surface unevenness map generating unit 241 moves each divided grindstone surface unevenness P1 in the circumferential direction (the direction of the arrow in FIG. 15) and aligns them side by side as shown in FIG. 16. As a result, the grindstone surface unevenness map generating unit 241 generates the map M1 shown in FIG. 14.

When generating the map M1 by dividing the continuous grinding wheel surface irregularities P into divided sections to generate the divided grinding wheel surface irregularities P1, the division may occur at a position that is offset from the period of the grinding wheel surface irregularities P. In other words, in this case, if the divided grinding wheel surface irregularities P1 have a fraction, and the divided grinding wheel surface irregularities P1 containing the fraction are arranged in the axial direction of the workpiece W, it may not be possible to properly connect the divided grinding wheel surface irregularities P1 to each other.

For example, in the case of an imbalance of the grinding wheel 12 associated with the grinding wheel surface irregularities P, the waveform phases of the divided grinding wheel surface irregularities P1 coincide in the axial direction of the workpiece W in the region corresponding to the width of the grinding wheel 12. That is, the waveforms of the divided grinding wheel surface irregularities P1 at adjacent axial positions have continuity, with peaks and valleys adjacent to each other in the axial direction. In order to reproduce such continuity when generating the map M1, the grinding wheel surface irregularity map generation unit 241 corrects the relative circumferential positions of each divided grinding wheel surface irregularity P1, i.e., performs fractional processing, so that the waveform phases at the end points of the multiple divided grinding wheel surface irregularities P1 coincide. By performing fractional processing by the grinding wheel surface irregularity map generation unit 241, it is possible to smooth the connection of the divided grinding wheel surface irregularities P1 at each axial position to the axial direction of the workpiece W, and improve the accuracy of the generated map M1.

The workpiece reference radius map generating unit 242 acquires the workpiece reference radius R from the first data processing unit 22 (workpiece reference radius calculating unit 224). Then, as shown in FIG. 17, the workpiece reference radius map generating unit 242 generates a map M2 that represents the surface texture S12 due to the workpiece reference radius R, which is caused by the grinding wheel.

Here, the workpiece reference radius R output from the workpiece reference radius calculation unit 224 is calculated using the spiral low-frequency waveform SLW as described above. The spiral low-frequency waveform SLW is based on the first basic data D1 detected in a spiral shape. Note that since the waveform representing the workpiece reference radius R is a change in the workpiece radius that continues in the axial direction of the workpiece W, when generating map M2, it is necessary to arrange the same number of points as the number of points arranged in the circumferential direction of map M1 for the grinding wheel surface irregularities P.

For this reason, the workpiece reference radius map generating unit 242 copies the axially continuous workpiece reference radii R by the number of circumferential arrangements at the time of fractional processing into which the grinding wheel surface irregularities P are divided. Then, the workpiece reference radius map generating unit 242 generates map M2 by arranging the copied workpiece reference radii R.

The high frequency intercenter relative vibration map generating unit 243 acquires the high frequency intercenter relative vibration waveform HDV from the second data processing unit 23 (high frequency intercenter relative vibration waveform generating unit 234). Then, as shown in FIG. 18, the high frequency intercenter relative vibration map generating unit 243 generates a map M3 that represents the surface texture S21 due to the high frequency intercenter relative vibration waveform HDV caused by the intercenter relative vibration. When generating the map M3, the high frequency intercenter relative vibration map generating unit 243 selects (extracts) any of the multiple division sections into which the grinding wheel surface unevenness P is divided. Then, the high frequency intercenter relative vibration waveform generating unit 234 generates the map M3 by arranging the high frequency intercenter relative vibration waveforms HDV corresponding to the selected (extracted) division sections in the axial direction of the workpiece W in the same number as the number of the map M2.

The low-frequency intercenter relative vibration map generating unit 244 acquires the low-frequency intercenter relative vibration waveform LDV from the first data processing unit 22 (low-frequency intercenter relative vibration waveform generating unit 223). Then, as shown in FIG. 19, the low-frequency intercenter relative vibration map generating unit 244 generates a map M4 that represents the surface texture S22 due to the low-frequency intercenter relative vibration waveform LDV caused by the intercenter relative vibration. When generating the map M4, the low-frequency intercenter relative vibration map generating unit 244 selects (extracts) a division section that is the same as the division section of the map M3 generated by the high-frequency intercenter relative vibration map generating unit 243 from among the multiple division sections into which the grinding wheel surface unevenness P is divided. Then, the low-frequency intercenter relative vibration map generating unit 244 generates the map M4 by arranging the low-frequency intercenter relative vibration waveforms LDV corresponding to the selected (extracted) division sections in the axial direction of the workpiece W in the same number as the number of arrangements of the map M2.

20, the surface texture map generator 245 generates a surface texture map M representing the surface texture S of the workpiece W. That is, the surface texture map generator 245 generates the surface texture map M representing the surface texture S of the workpiece W by synthesizing (adding) the map M1 generated by the grinding wheel surface unevenness map generator 241, the map M2 generated by the workpiece reference radius map generator 242, the map M3 generated by the high frequency intercenter relative vibration map generator 243, and the map M4 generated by the low frequency intercenter relative vibration map generator 244.

Here, map M1 generated by grindstone surface unevenness map generator 241 and map M2 generated by workpiece reference radius map generator 242 are both caused by the grindstone. Therefore, surface texture map generator 245 can generate a map (not shown) representing surface texture S1 caused by the grindstone as surface texture map M representing surface texture S of workpiece W by synthesizing (adding) only map M1 and map M2 as necessary.

In addition, map M3 generated by the high-frequency intercenter relative vibration map generating unit 243 and map M4 generated by the low-frequency intercenter relative vibration map generating unit 244 are both caused by intercenter relative vibration. Therefore, the surface texture map generating unit 245 can generate a map (not shown) representing the surface texture S2 caused by intercenter relative vibration as the surface texture map M representing the surface texture S of the workpiece W by synthesizing (adding) only maps M3 and M4 as necessary.

Then, the surface texture map generator 245 outputs the generated surface texture map M to the image output device 40. As a result, the image output device 40 displays the acquired surface texture map M, for example, as an image on a display.

As can be understood from the above explanation, according to the surface texture estimation system H, the surface texture generating device 20 can calculate the grindstone surface irregularities P and the workpiece reference radius R that form the surface texture S1 caused by the grindstone based on the first detection data K1, i.e., the first basic data D1, detected in a spiral shape, and the second detection data K2, i.e., the second basic data D2, detected at the same axial position. In addition, the surface texture generating device 20 can calculate the low-frequency intercenter relative vibration waveform LDV and the high-frequency intercenter relative vibration waveform HDV that form the surface texture S2 caused by the intercenter relative vibration based on the second detection data K2, i.e., the second basic data D2, detected at the same axial position. Then, the surface texture S of the workpiece W can be generated by adding the surface texture S1 caused by the grindstone and the surface texture S2 caused by the intercenter relative vibration.

More specifically, the surface texture generating device 20 can calculate the surface texture S1 caused by the grindstone based on the low-frequency component D11 as the first low-frequency component and the spiral high-frequency component D12 as the first high-frequency component among the frequency components of the first basic data D1 (first detection data K1) detected in a spiral shape, and the low-frequency component D21 as the second low-frequency component and the one-section high-frequency component D22 as the second high-frequency component among the frequency components of the second basic data D2 (second detection data K2) detected at the same axial position. The surface texture generating device 20 can also calculate the surface texture S2 caused by the center-to-center relative vibration based on the low-frequency component D21 and the one-section high-frequency component D22 of the second basic data D2 (second detection data K2).

This makes it possible to reduce the number of data items (the number of peaks in the waveform) required to calculate the grindstone-induced surface quality S1, which is calculated using the first basic data D1 and the second basic data D2. As a result, it is possible to reduce the time required to estimate the surface quality S of the workpiece W, specifically, the time required to detect (collect) the first basic data D1 (first detection data K1).

When calculating the surface quality S1 due to the grindstone, the second basic data D2 (second detection data K2) can be used, which corresponds to the same axial position as the first basic data D1 (first detection data K1) detected in a spiral shape, i.e., one revolution of the workpiece W. This improves the calculation accuracy of the surface quality S1 due to the grindstone, and as a result, improves the estimation accuracy of the surface quality S of the final workpiece W.

In addition, when the surface quality S1 caused by the grinding wheel is calculated, for example, after the workpiece W is machined, the surface quality S1 can be calculated with high accuracy in a short time. Therefore, the surface quality S1 can be used to determine the maintenance of the grinding device 10, for example, the truing interval, etc.

Furthermore, the surface texture S2 caused by the center-to-center relative vibration can be calculated using the second basic data D2 (second detection data K2) detected at the same axial position. This reduces the amount of data required to calculate the surface texture S2 caused by the center-to-center relative vibration. As a result, it becomes possible to calculate the surface texture S2 caused by the center-to-center relative vibration even while the workpiece W is being ground. This allows the calculated surface texture S2 caused by the center-to-center relative vibration to be used to monitor the grinding accuracy and the operating state of the grinding device 10, and can be used for quality checks and preventive maintenance of the workpiece W, or for abnormality detection, etc.

(3. Other Examples)
In the above-described embodiment, the detection device 30 is mainly equipped with an acceleration sensor, and detects displacement obtained by conversion based on acceleration as the first detection data and the second detection data. The detection device 30 is not limited to being mainly equipped with an acceleration sensor, and may be mainly equipped with a displacement sensor that detects displacement caused by unevenness on the surface of the workpiece W.

Examples of the displacement sensor equipped in the detection device 30 include a contact-type sizing device 17 and a linear gauge, or a non-contact type laser sensor, optical sensor, eddy current sensor, etc. The contact-type sizing device 17 and linear gauge have a contact member such as a probe 17a that contacts the surface of the workpiece W, and detect the displacement of the vibration of the contact member that occurs with the rotation of the workpiece W. In addition, in order to enable the sizing device 17 to detect high-frequency components, for example, an analog output amplifier that omits the low-pass filter provided in the sizing device 17 or a high-frequency digital output amplifier can be used. The non-contact type laser sensor, optical sensor, and eddy current sensor are arranged so as to be in non-contact with the surface of the workpiece W, and detect the displacement from the reference position to the surface of the workpiece W that occurs with the rotation of the workpiece W.

The displacement of the vibration of the contact member detected by the contact sensor and the displacement detected by the non-contact sensor are both detection data (time series data) that indicate the displacement of the unevenness on the surface of the workpiece. Therefore, even in this case, the detection data (displacement) output from the detection device 30 is time series data, and the basic data acquisition unit 21 acquires the first detection data K1 and the second detection data K2 output from the detection device 30 as the first basic data D1 and the second basic data D2, respectively.

The linear gauge has a probe that contacts the workpiece W and an arm that supports the probe, and detects the displacement of the surface of the workpiece W with the probe in contact with the rotating workpiece W. The linear gauge is supported by an axial movement device, similar to the sizing device 17, and can be moved in the axial direction of the workpiece W, i.e., in the Z direction.

Furthermore, in the above-mentioned example, the low-frequency component extraction unit 221 of the first data processing unit 22 performs FFT, and the spiral low-frequency waveform generation unit 222 and the low-frequency intercenter relative vibration waveform generation unit 223 perform inverse FFT. Also, the spiral high-frequency component extraction unit 231 and the single cross-section high-frequency component extraction unit 232 of the second data processing unit 23 perform FFT, and the spiral high-frequency waveform generation unit 233 and the high-frequency intercenter relative vibration waveform generation unit 234 perform inverse FFT.

In this way, in order to omit performing FFT or inverse FFT, it is also possible to provide each of the above sections with a filter capable of extracting the desired frequency components. Examples of filters include a low-pass filter, a high-pass filter, a band-pass filter, a Gaussian filter, etc.

10...grinding device, 11...bed, 11a...grinding wheel head guide unit, 11b...spindle table guide unit, 12...grinding wheel, 12a...grinding wheel rotation motor, 13...grinding wheel head, 14...spindle head, 14a...spindle rotation motor, 15...tailstock, 16...spindle table, 17...sizing device, 17a...measuring probe, 17b...finger, 17c...axial movement device, 17d...axial movement control unit, 18...controller, 20...surface texture generating device, 21...basic data acquisition unit, 22...first data processing unit , 221... low frequency component extraction unit, 222... spiral low frequency waveform generation unit, 223... low frequency center-to-center relative vibration waveform generation unit, 224... workpiece reference radius calculation unit, 23... second data processing unit, 231... spiral high frequency component extraction unit, 232... one cross section high frequency component extraction unit, 233... spiral high frequency waveform generation unit, 234... high frequency center-to-center relative vibration waveform generation unit, 235... grindstone surface unevenness calculation unit, 24... mapping processing unit, 241... grindstone surface unevenness map generation unit, 242... work Crop reference radius map generator, 243...high frequency intercenter relative vibration map generator, 244...low frequency intercenter relative vibration map generator, 245...surface texture map generator, 30...detection device, 40...image output device, C...number of spirals, K1...first detection data, K2...second detection data, D1...first basic data, D11...low frequency component, D12...spiral high frequency component, D2...second basic data, D21...low frequency component, D22...first cross section high frequency component, HDV...high frequency intercenter relative Vibration waveform, LDV...low-frequency intercenter relative vibration waveform, SHW...spiral high-frequency waveform, SLW...spiral low-frequency waveform, M...surface quality map, M1, M2, M3, M4...map, O...rotation center, P...grinding wheel surface unevenness, P1...divided grinding wheel surface unevenness, R...workpiece reference radius, S...surface quality, S1...(grinding wheel-induced) surface quality, S2...(intercenter relative vibration-induced) surface quality, S11, S12, S21, S22...surface quality, H...surface quality estimation system, W...workpiece

Claims (9)

a detection device that outputs detection data corresponding to a surface condition of a workpiece ground by a grinding wheel in a grinding device;
a surface texture generating device that generates a surface texture of the workpiece based on the detection data detected by the detection device by moving a measurement position of the detection device on the surface of the workpiece relative to the workpiece at least in a circumferential direction,
The surface texture generating device comprises:
a surface texture caused by the grinding wheel, into which the surface state of the grinding surface of the grinding wheel is transferred, is calculated based on first detection data, which is the detection data detected by helically moving the measurement position on the surface of the workpiece in the circumferential and axial directions of the workpiece, and second detection data, which is the detection data detected by moving the measurement position in the circumferential direction at the same position in the axial direction of the workpiece, and a surface texture caused by center-to-center relative vibration, into which vibration generated by relative center-to-center fluctuation between the grinding wheel and the workpiece is transferred, based on the second detection data;
a surface texture estimating system that generates a surface texture of the workpiece by adding the surface texture caused by the grinding wheel and the surface texture caused by the intercenter relative vibration. The surface quality caused by the grindstone is
a surface texture due to the grindstone of a low frequency component based on a first low frequency component which is a low frequency component among frequency components of the first detection data and a second low frequency component which is a low frequency component among frequency components of the second detection data;
2. The surface texture estimating system according to claim 1, further comprising: a first high frequency component that is a high frequency component in a higher frequency range than the first low frequency component among the frequency components of the first detection data; and a surface texture caused by the grindstone based on a second high frequency component obtained by excluding a grindstone rotation frequency component of the grinding wheel from high frequency components in a higher frequency range than the second low frequency component among the frequency components of the second detection data. The surface quality caused by the relative vibration between the centers is as follows:
a surface texture caused by the intercenter relative vibration based on a second low-frequency component that is a low-frequency component among frequency components of the second detection data; and
and a surface texture caused by the intercenter relative vibration based on a second high-frequency component in a frequency range higher than the second low-frequency component among frequency components of the second detection data. The surface texture generating device comprises:
a basic data acquisition unit that acquires the first detection data as first basic data and acquires the second detection data as second basic data;
a first data processing unit that extracts a first low-frequency component, which is a low-frequency component in the frequency characteristics of the first basic data, and a second low-frequency component, which is a low-frequency component in the frequency characteristics of the second basic data, calculates a reference radius of the workpiece by subtracting the second low-frequency component from the extracted first low-frequency component, and calculates a low-frequency intercenter relative vibration, which is an intercenter relative vibration between the grinding wheel and the workpiece, using the second low-frequency component;
a second data processing unit that extracts a first high-frequency component in a higher frequency range than the first low-frequency component from the frequency characteristics of the first basic data and a second high-frequency component in a higher frequency range than the second low-frequency component from the frequency characteristics of the second basic data, calculates a high-frequency intercenter relative vibration that is an intercenter relative vibration between the grinding wheel and the workpiece using the extracted second high-frequency component, and calculates a grinding wheel surface irregularity in which the surface state of the grinding surface of the grinding wheel is transferred to the surface of the workpiece by subtracting the high-frequency intercenter relative vibration from the extracted first high-frequency component;
generating a surface texture due to the grindstone using the reference radius of the workpiece calculated by the first data processing unit and the grindstone surface irregularity calculated by the second data processing unit, and generating a surface texture due to the intercore relative vibration using the low-frequency intercore relative vibration calculated by the first data processing unit and the high-frequency intercore relative vibration calculated by the second data processing unit;
a surface texture generating unit that generates a surface texture of the workpiece by adding the generated surface texture caused by the grindstone and the generated surface texture caused by the intercenter relative vibration;
The surface texture estimation system according to any one of claims 1 to 3, comprising: The surface texture generating unit includes:
5. The surface texture estimation system according to claim 4, wherein the reference radius of the workpiece, the grinding wheel surface irregularities, the low-frequency inter-center relative vibration, and the high-frequency inter-center relative vibration are each mapped, and a map representing the surface texture of the workpiece is generated by adding up each map. The detection device includes:
The surface texture estimating system according to any one of claims 1 to 5, wherein time series data relating to acceleration or displacement is output as the detection data. The detection device includes:
The surface texture estimation system according to any one of claims 1 to 6, which is provided in an outer diameter measuring device that measures the outer diameter of the workpiece in the grinding device. The first data processing unit is
6. The surface texture estimating system according to claim 4 or 5, wherein the first low-frequency component is extracted by excluding one peak/period component from the low-frequency components in the frequency characteristics of the first basic data, and the second low-frequency component is extracted by excluding one peak/period component from the low-frequency components in the frequency characteristics of the second basic data. The surface texture estimation system according to any one of claims 1 to 8, comprising an image output device that outputs the surface texture of the workpiece generated by the surface texture generation device as an image.

Images