Surface texture estimation device and grinding machine

Abstract

To provide a surface property estimation device that can improve accuracy in estimating a surface property, by mitigating effect of noise due to a gain unstable frequency band with a natural vibration property.SOLUTION: A calculation processing device 20 of a surface property estimation device 2 is provided with: a signal data obtaining part 21 that obtains first spiral data D1 on an acceleration sensor 32 of a sizing instrument 3; an evaluation information storage part 23 that stores evaluation information I in which a gain stable frequency band K0 and a gain unstable frequency band K1 are set in a natural vibration property K of the acceleration sensor 32; a rotation speed setting part 25 that sets a measurement-time rotation speed Vw2 of the work-piece W, on the basis of a relation between a processing-time rotation speed Vt1 of a grinding wheel 12 and a processing-time rotation speed Vw1 of the work-piece W; and a surface property estimation part 22 that estimates, as a surface property S, a concave-convex surface state of a ground peripheral surface W1 of the work-piece W, on the basis of the first spiral data D1 obtained by the signal data obtaining part 21 when the work-piece W is rotated at the measurement-time rotation speed Vw2, at the time of measurement.SELECTED DRAWING: Figure 6

Description

Technical Field

[0001] The present invention relates to a surface property estimation device and a grinding machine.

Background Art

[0002] A surface property estimation device is used to estimate (measure) surface properties such as unevenness on the surface of a workpiece rotatably supported on a grinding machine. On the grinding wheel surface of the grinding wheel of the grinding machine, unevenness due to a large number of abrasive grains is formed, and during grinding, the unevenness due to these abrasive grains is reflected on the surface of the workpiece. Then, depending on the state of the abrasive grains of the grinding wheel, processing conditions such as the rotational speed of the grinding wheel and the rotational speed of the workpiece, and the shape of the workpiece in the grinding machine, the surface properties formed on the surface of the workpiece change. In the surface property estimation device, an unevenness measuring device having a sensor or the like is used to measure the unevenness generated in the circumferential direction and the axial direction on the surface of the workpiece, and the surface properties of the surface of the workpiece are estimated using the measurement results of this unevenness.

[0003] For example, in the surface property estimation system of Patent Document 1, when the measurement position on the surface of the workpiece is moved spirally in the circumferential direction and the axial direction, unevenness on the surface of the workpiece is detected by a sensor, and based on the time-series data related to this detection, the surface properties of the surface of the workpiece are generated. The time-series data is acquired corresponding to the spiral positions at each predetermined angle with respect to the rotation axis of the workpiece, and based on the time-series data, circumferential chatter is generated at different angles of the workpiece and at different axial positions of the workpiece. Then, the circumferential chatter is regarded as a plurality of circumferential chatters at the same angle of the workpiece, and a planar chatter in which the plurality of circumferential chatters are arranged in parallel in the axial direction is generated.

Prior Art Documents

[0005] Sensors that make up a sizing device, etc., are affected by vibrations generated in various parts of the grinding machine and have natural vibration characteristics as frequency response characteristics. These natural vibration characteristics affect the measurement of surface irregularities of a workpiece by the sizing device, etc., as noise and become a source of error in estimating the surface properties of the workpiece. The surface property estimation system in Patent Document 1 also does not take these natural vibration characteristics into consideration, and there is a risk of errors in the generation (estimation) of surface properties. Therefore, further improvements are needed to improve the accuracy of estimating the surface properties of a workpiece by the surface property estimation device.

[0006] This invention has been made in view of the above problems, and aims to provide a surface texture estimation device and a grinding machine that can improve the accuracy of surface texture estimation by mitigating the influence of noise due to the gain instability frequency band of the natural vibration characteristics. [Means for solving the problem]

[0007] One aspect of the present invention is, A surface texture estimation device for estimating the surface texture of the ground surface of a workpiece that has been ground by a grinding wheel on a grinding machine, A measuring device provided in the grinding machine for measuring the diameter of the grinding surface while grinding the workpiece while it is rotating, A calculation processing device that estimates the surface properties of the grinding surface of the workpiece, based on signal data output from a sensor provided in the measuring device during measurement after grinding of the workpiece, and which is caused by the grinding wheel onto which the uneven surface of the grinding wheel has been transferred. Equipped with, The aforementioned arithmetic processing unit is A signal data acquisition unit acquires the signal data when the workpiece is rotated and the sizing device and the workpiece are moved relative to each other in the axial direction of the workpiece, corresponding to the uneven surface state of the grinding circumferential surface. An evaluation information storage unit stores evaluation information in which, when a vibration with a varied frequency is input to the measuring device, the natural vibration characteristics as the frequency response characteristics of the output data output from the sensor include a gain-stable frequency band in which the variation in the gain of the natural vibration characteristics is within a predetermined value, and a gain-unstable frequency band in which the frequency is lower than the gain-stable frequency band and the variation in the gain of the natural vibration characteristics exceeds the predetermined value. A machining condition acquisition unit that acquires the machining rotation speed Vt1 of the grinding wheel and the machining rotation speed Vw1 of the workpiece during grinding of the workpiece by the grinding wheel, Based on the relationship between the machining rotation speed Vt1 of the grinding wheel and the machining rotation speed Vw1 of the workpiece, when assuming the measurement rotation speed Vw2 of the workpiece at the time of measurement, the fundamental frequency components that will be included in the signal data are determined. Based on A rotational speed setting unit sets the rotational speed Vw2 of the workpiece during measurement so that at least two selected frequencies are included in the gain stabilization frequency band, A surface texture estimation unit estimates the uneven surface state of the grinding circumferential surface as the surface texture based on the signal data acquired by the signal data acquisition unit when the workpiece is rotated at the measurement rotation speed Vw2 during the measurement, Equipped with 、 The aforementioned at least two selected frequencies are, The fundamental frequency component and at least one harmonic component among the plurality of harmonic components of the fundamental frequency component, Alternatively, These are at least two harmonic components among the multiple harmonic components of the aforementioned fundamental frequency component. It is located in the surface texture estimation device.

[0008] Another aspect of the present invention is a grinding machine equipped with a surface texture estimation device according to one aspect of the present invention. [Effects of the Invention]

[0009] In the surface texture estimation device according to one embodiment described above, evaluation information of the natural vibration characteristics obtained by a sensor is stored, and the influence of the natural vibration characteristics on the estimation of the surface texture is mitigated by intentionally changing the rotational speed (rotational speed) Vw2 of the workpiece during measurement relative to the rotational speed (rotational speed) Vw1 of the workpiece during machining, taking into account the natural vibration characteristics. Specifically, in the evaluation information stored in the evaluation information unit, a gain stable frequency band and a gain unstable frequency band for the natural vibration characteristics are set.

[0010] Then, in the rotation speed setting unit, based on the relationship between the machining rotation speed Vt1 of the grinding wheel and the machining rotation speed Vw1 of the workpiece, the machined rotation speed Vw2 of the workpiece is set such that, assuming the machined rotation speed Vw2 of the workpiece during measurement, at least two selected frequencies among the fundamental frequency components and harmonic components that will be included in the signal data of the signal data acquisition unit are included in the gain stable frequency band.

[0011] By setting the workpiece's rotational speed Vw2 during measurement in this way, the signal data acquired by the signal data acquisition unit during measurement can be prevented from being affected by noise in the gain instability frequency band of the natural vibration characteristics. Furthermore, by reducing the error factors, the accuracy of the surface texture estimation by the surface texture estimation unit can be improved.

[0012] Therefore, according to the surface texture estimation device of the above embodiment, the influence of noise due to the gain instability frequency band of the natural vibration characteristics can be excluded, thereby improving the accuracy of surface texture estimation.

[0013] According to the other embodiment of the grinding machine described above, the same effects as those of the surface texture estimation device of the one embodiment described above are achieved. [Brief explanation of the drawing]

[0014] [Figure 1] An explanatory diagram showing a grinding machine including a surface texture estimation device according to an embodiment. [Figure 2]Explanatory drawing showing a surface property estimation device at the time of measurement after grinding of a workpiece according to an embodiment. [Figure 3] Explanatory drawing showing a surface property estimation device during grinding of a workpiece according to an embodiment. [Figure 4] Flowchart showing a grinding process of a grinding machine according to an embodiment. [Figure 5] Explanatory drawing showing the surface property of the grinding peripheral surface of a workpiece ground by a grinding machine according to an embodiment. [Figure 6] Explanatory drawing showing the configuration of a surface property estimation device according to an embodiment. [Figure 7] Explanatory drawing showing divided data of first spiral data by an acceleration sensor according to an embodiment. [Figure 8] Explanatory drawing showing phase-matched data of first spiral data by an acceleration sensor according to an embodiment. [Figure 9] Explanatory drawing showing a chatter map based on first spiral data by an acceleration sensor according to an embodiment. [Figure 10] Explanatory drawing showing a radius map based on second spiral data and rotation data by a displacement sensor according to an embodiment. [Figure 11] Explanatory drawing showing a surface property map in which a chatter map and a radius map are combined according to an embodiment. [Figure 12] Graph showing the natural vibration characteristics of an acceleration sensor according to an embodiment. [Figure 13] Graph showing Fourier transform data after fast Fourier transform of first spiral data by an acceleration sensor according to an embodiment. [Figure 14] Graph showing Fourier transform data after fast Fourier transform of first spiral data by an acceleration sensor according to an embodiment.

Mode for Carrying Out the Invention

[0015] A preferred embodiment of the surface texture estimation device and the grinding machine equipped with the surface texture estimation device described above will be explained with reference to the drawings. (Embodiment)

[0016] 1. Configuration of grinding machine 1 The configuration of the grinding machine 1 will now be described. As shown in Figure 1, the grinding machine 1 is equipped with a surface texture estimation device 2. The surface texture estimation device 2 is used as an analysis device for the grinding state in the grinding machine 1. In particular, the surface texture estimation device 2 estimates the surface texture S of the ground surface W1 of the workpiece W that has been ground by the grinding wheel 12 in the grinding machine 1. The grinding machine 1 is equipped with an image output device 4 that outputs the surface texture S of the ground surface W1 of the workpiece W, estimated by the surface texture estimation device 2, as a surface texture map M as a mapped image.

[0017] The grinding machine 1 grinds the grinding surface W1 of the workpiece W by rotating the workpiece W and the grinding wheel 12 while moving the workpiece W and the grinding wheel 12 relative to each other. The grinding machine 1 can be applied to a table traverse type grinding machine in which the workpiece W is traversed, or to a grinding wheel base traverse type grinding machine in which the grinding wheel base 13 supporting the grinding wheel 12 is traversed. Below, a table traverse type grinding machine will be described as an example of the grinding machine 1.

[0018] The grinding machine 1 comprises a bed 11, a grinding wheel 12, a grinding wheel base 13, a headstock 14, a tailstock 15, a spindle table 16, a control device 10, and a surface texture estimation device 2. The workpiece W is supported at both ends in the axial direction by the headstock 14 and the tailstock 15, and rotates driven by the headstock 14. The shape of the workpiece W is not particularly limited and may be cylindrical, cylindrical, or the like. As shown in Figure 2, the grinding surface W1 of the workpiece W may be not only the outer surface of a cylindrical or cylindrical workpiece W, but also the inner surface of a cylindrical workpiece W.

[0019] In this embodiment, the case in which the workpiece W is cylindrical is illustrated. The grinding machine 1 brings the grinding wheel 12 into contact with the outer circumferential surface W1 of the rotating workpiece W, and grinds the outer circumferential surface of the workpiece W with the grinding wheel 12.

[0020] Here, the direction parallel to the axial direction of the workpiece W supported by the headstock 14 and tailstock 15 is called the Z-axis direction, and the direction perpendicular to the Z-axis direction in which the grinding wheel 12 approaches the workpiece W is called the X-axis direction. In each figure, the Z-axis direction is indicated by the symbol Z, and the X-axis direction is indicated by the symbol X. The grinding wheel 12 is supported on the grinding wheel base 13 so as to be rotatable around an axis parallel to the Z-axis direction. A grinding wheel base guide section 11a is provided on the bed 11, and the grinding wheel base 13 is supported on the grinding wheel base guide section 11a so as to be movable in the X-axis direction, and moves in the X-axis direction by a drive source (not shown) controlled by the control device 10.

[0021] The grinding wheel 12 is driven and rotated by the grinding wheel rotation motor 12a of the grinding wheel base 13, which is controlled by the control device 10. As the grinding wheel base 13 moves in the X-axis direction, the grinding wheel 12 approaches the grinding surface W1 of the workpiece W and grinds the grinding surface W1 of the workpiece W.

[0022] The spindle table 16 is supported on the bed 11 by a spindle table guide 11b so as to be movable in the Z-axis direction, and moves in the Z-axis direction by a drive source (not shown) controlled by the control device 10. The headstock 14 and tailstock 15 are arranged opposite each other on the spindle table 16. The workpiece W is rotated by the spindle rotation motor 14a of the headstock 14, which is controlled by the control device 10.

[0023] 2. Overview of Surface Texture Estimation Device 2 The surface texture estimation device 2 will now be described. Figure 2 shows the surface texture estimation device 2 during measurement after grinding of the workpiece W, and Figure 3 shows the surface texture estimation device 2 during grinding of the workpiece W (sometimes simply referred to as during processing). As shown in Figures 2 and 3, the surface texture estimation device 2 estimates the surface texture S of the ground surface W1 of the workpiece W, which has been ground by the grinding wheel 12 on the grinding machine 1.

[0024] The surface texture estimation device 2 comprises a sizing device 3 equipped with an acceleration sensor 32 as a sensor, and a calculation processing device 20. The sizing device 3 is installed on the grinding machine 1 and measures the diameter of the grinding surface W1 of the workpiece W while grinding the workpiece W while it is rotating. When measuring the workpiece W after grinding, the calculation processing device 20 estimates the uneven surface state of the grinding surface W1 of the workpiece W, which is caused by the grinding wheel 12 onto which the uneven surface of the grinding wheel 12 has been transferred, as the surface texture S, based on the signal data output from the acceleration sensor 32 of the sizing device 3.

[0025] 3. Configuration of the sizing device 3 of the surface texture estimation device 2 Next, the sizing device 3 will be described. In this embodiment, as shown in Figure 2, the outer surface of the cylindrical workpiece W is ground as the grinding surface W1, and the sizing device 3 measures the outer diameter of the outer surface of the workpiece W as an outer diameter measuring device. The sizing device 3 may also measure the inner diameter of the inner surface of the workpiece W as the grinding surface W1. The diameter of the grinding surface W1 of the workpiece W measured by the sizing device 3 is used for switching the grinding process. However, in this embodiment, the sizing device 3 is also used for estimating the surface properties S of the grinding surface W1 of the workpiece W by the surface properties estimation device 2.

[0026] The sizing device 3 includes a pair of contact members 31 that contact the grinding surface W1 of the workpiece W and displace according to the uneven surface condition of the grinding surface W1, and an acceleration sensor 32 positioned on one of the contact members 31. By using the acceleration sensor 32, it becomes easy to measure the high-frequency components of the uneven surface on the grinding surface W1 of the workpiece W.

[0027] The tips of a pair of contact members 31 are provided with contact elements 33 that contact the grinding surface W1 of the workpiece W. The contact elements 33 abut the grinding surface W1 of the workpiece W at two points on either side of the rotation center O of the workpiece W, and displace in accordance with the uneven surface formed on the grinding surface W1. The displacement of the contact elements 33 is measured by a displacement sensor 34. The displacement sensor 34 may be, for example, a differential transformer that converts mechanical linear motion into an electrical signal as a displacement.

[0028] The sizing device 3 measures the outer diameter of the grinding surface W1 of the workpiece W by converting the mechanical displacement of the displacement sensor 34 into electrical signal data, and also measures the uneven surface condition of the grinding surface W1 of the workpiece W by converting the acceleration detected by the acceleration sensor 32 into displacement and electrical signal data. As shown in Figure 1, the sizing device 3 is movable in the Z-axis direction parallel to the central axis of the workpiece W by an axial movement device 35 controlled by the control device 10.

[0029] The uneven surface condition of the ground surface W1 of the workpiece W is measured by rotating the workpiece W after grinding and measuring the uneven surface condition of the ground surface W1 of the workpiece W as follows: low-frequency component uneven surface detected based on the signal data of the displacement sensor 34 and high-frequency component uneven surface detected based on the signal data of the acceleration sensor 32. The low-frequency component detected based on the signal data of the displacement sensor 34 is detected as a frequency of, for example, 10 Hz or less, and the high-frequency component detected based on the signal data of the acceleration sensor 32 is detected as a frequency of, for example, more than 10 Hz and less than or equal to 2500 Hz.

[0030] 4. Grinding process of grinding machine 1 Next, the grinding process of the grinding machine will be explained. As shown in Figure 4, the grinding process is divided according to the difference in the feed rate of the grinding wheel 12 in the X-axis direction, and is performed in the order of rough grinding process St1, fine grinding process St2, micro grinding process St3, and spark-out process St4. The feed rate of the grinding wheel 12 in each process is rough grinding process St1 > fine grinding process St2 > micro 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 micro grinding process St3, the surface shape of the workpiece W is refined while reducing the feed rate of the grinding wheel 12. In the final spark-out process St4, the surface of the workpiece W is finished, completing the workpiece W.

[0031] The diameter of the grinding surface W1 of the workpiece W, measured by the sizing device 3, is used to switch between grinding processes. The control device 10 of the grinding machine 1 is set to determine the diameter of the workpiece W when switching between the rough grinding process St1 and the fine grinding process St2, when switching between the fine grinding process St2 and the micro-grinding process St3, and when switching between the micro-grinding process St3 and the spark-out process St4. During grinding, when the diameter of the workpiece W, measured by the sizing device 3, reaches the set diameter, the control device 10 of the grinding machine 1 switches the grinding process.

[0032] In this configuration, the surface texture estimation device 2 preferably estimates the surface texture S of the workpiece W after the spark-out process St4 in which grinding is completed. The surface texture estimation device 2 estimates the surface texture S of the workpiece W in-process. In-process refers to the period from when the workpiece W is ground until it is removed from the grinding machine 1, and includes the period after the spark-out process St4. In this embodiment, the surface texture estimation device 2 maintains the rotational state of the workpiece W during grinding after the grinding of the workpiece W is completed, and appropriately increases the rotational speed Vw2 of the workpiece W to estimate the surface texture S of the workpiece W.

[0033] 5. Surface properties S of the grinding circumferential surface W1 of the workpiece W. Next, the surface properties S of the grinding surface W1 of the workpiece W will be described. As shown in Figure 5, the surface properties S of the grinding surface W1 of the workpiece W, which is ground by the grinding machine 1, are formed due to various factors. In this embodiment, the surface properties S will be those caused by the grinding wheel 12, which is the transfer of the uneven surface as the surface state of the grinding outer surface of the grinding wheel 12.

[0034] The surface properties S of the grinding surface W1 caused by the grinding wheel 12 are separated into surface properties S1, which are transferred from the uneven surface of the grinding wheel 12, and surface properties S2, which are due to variations in the reference radius in the circumferential direction C of the workpiece W. The surface properties S are generated by combining (adding) surface properties S1 and S2. The variation in the reference radius in the circumferential direction C of the workpiece W is expressed as roundness, etc., and is thought to be mainly caused by a shift in the distance between the center of the grinding wheel 12 and the center of the workpiece W during grinding.

[0035] 6. Configuration of the arithmetic processing unit 20 of the surface texture estimation device 2 Next, the calculation processing unit 20 of the surface texture estimation device 2 will be described. As shown in Figure 6, the calculation processing unit 20 includes a signal data acquisition unit 21, an evaluation information storage unit 23, a machining condition acquisition unit 24, a rotation speed setting unit 25, a workpiece spindle drive unit 26, and a surface texture estimation unit 22.

[0036] The signal data acquisition unit 21 acquires signal data from the acceleration sensor 32 and the displacement sensor 34 of the measuring device 3, corresponding to the uneven surface state of the grinding surface W1, when the workpiece W is rotated and the measuring device 3 and the workpiece W are moved relative to each other in the Z-axis direction of the workpiece W, in other words, when the position where the contact element 33 of the measuring device 3 contacts the workpiece W is moved relative to each other in a spiral manner in the Z-axis direction. The signal data from the acceleration sensor 32 is acquired as first spiral data D1 of the high frequency component, and the signal data from the displacement sensor 34 is acquired as second spiral data D2 of the low frequency component. In addition, when the workpiece W is rotated while the relative position of the measuring device 3 and the workpiece W in the Z-axis direction is fixed, the signal data acquisition unit 21 acquires signal data from the displacement sensor 34 as rotation data D3.

[0037] In Figure 5, the states in which the first spiral data D1 and second spiral data D2 are measured by the sizing device 3, and the states in which the rotation data D3 is measured by the sizing device 3, are shown by dashed lines. The first spiral data D1 and second spiral data D2 are simultaneously acquired as data for a portion of the Z-axis direction and a portion of the circumferential direction C of the grinding surface W1 of the workpiece W. The portion of the circumferential direction C of the first spiral data D1 is extracted as the area in which the grinding wheel surface 121 of the grinding wheel 12 is transferred multiple times to the grinding surface W1 of the workpiece W. Within this portion of the circumferential direction C of the grinding surface W1, grinding marks from each abrasive grain of the grinding wheel 12 repeatedly appear. The first spiral data D1 from the acceleration sensor 32 is used to measure the uneven surface condition of the grinding surface W1 of the workpiece W.

[0038] The second spiral data D2 obtained from the displacement sensor 34 is used to measure the difference in radius of the workpiece W in the Z-axis direction. The difference in radius of the workpiece W in the Z-axis direction can, in principle, be measured by moving the sizing device 3 relative to the workpiece W in the Z-axis direction. However, the second spiral data D2 is acquired when the contact element 33 moves spirally relative to the circumferential C and Z-axis directions simultaneously with the acceleration sensor 32. Therefore, the change in radius of the grinding surface W1 of the workpiece W in the Z-axis direction is obtained by excluding the change in radius of the grinding surface W1 of the workpiece W in the circumferential C direction obtained from the rotation data D3 from the change in radius of the grinding surface W1 of the workpiece W in the Z-axis direction and the change in radius of the grinding surface W1 of the workpiece W obtained from the second spiral data D2.

[0039] Rotational data D3 is acquired as data for the entire circumference of the circumferential direction C at the same position in the Z-axis direction that overlaps with a portion of the Z-axis range of the first spiral data D1 and the second spiral data D2. The change in radius of the grinding surface W1 of the workpiece W in the circumferential direction C, as indicated by roundness, is considered to be maintained in a similar state in the Z-axis direction. Therefore, rotational data D3 only needs to be acquired at one position in the Z-axis direction. However, to improve accuracy, rotational data D3 may be acquired at multiple positions in the Z-axis direction within the Z-axis range of the first spiral data D1 and the second spiral data D2.

[0040] As shown in Figure 6, the surface texture estimation unit 22 estimates the uneven surface state of the grinding surface W1 as the surface texture S based on the signal data acquired by the signal data acquisition unit 21 when the workpiece W is rotated at the measurement rotation speed Vw2 during measurement. The surface texture estimation unit 22 performs signal processing on the first spiral data D1, the second spiral data D2, and the rotation data D3 to estimate the surface texture S of the grinding surface W1 of the workpiece W, and creates a surface texture map M that visually represents the surface texture S by differences in color. The surface texture map M is created by combining a chatter map M1 that shows the surface texture S1 of the grinding surface W1 caused by the uneven surface of the grinding wheel 12, and a radius map M2 that shows the surface texture S2 of the grinding surface W1 caused by errors such as roundness and radius of the grinding surface W1.

[0041] The tremor map M1 is created by performing signal processing on the first spiral data D1, which is signal data from the acceleration sensor 32. This signal processing is performed by the following steps: gain correction processing 221a, division processing 222, high-frequency component analysis 223a, and phase alignment processing 224. In the gain correction processing 221a, it is taken into consideration that the acceleration sensor 32 has input / output characteristics that attenuate the output signal for input signals exceeding a certain frequency, and the output signal for each frequency is corrected so that the variation in the output signal of the acceleration sensor 32 is reduced.

[0042] As shown in Figure 7, in the division process 222, the first spiral data D1 of the grinding surface W1 of the workpiece W is divided into multiple divided data D11, which are data for each position in the Z-axis direction set at predetermined intervals in the Z-axis direction and for each position in the circumferential direction C set at predetermined intervals in the circumferential direction C. In the high-frequency component analysis 223a, a fast Fourier transform (FFT) is performed on each divided data D11. Then, regarding the uneven surface state of the grinding surface W1 on which abrasive marks are repeatedly formed by each abrasive grain of the grinding wheel 12, the formation state of each abrasive mark is extracted as Fourier transform data of high-frequency components represented by frequency (or period) and amplitude.

[0043] Subsequently, specific high-frequency components are extracted to remove noise components from the Fourier transform data of the high-frequency components, and an inverse fast Fourier transform (inverse FFT) is performed on the extracted high-frequency components. This removes the noise components from each divided data D11. The noise components include noise caused by mechanical vibrations, etc.

[0044] In this embodiment, as will be described later, by adjusting the rotational speed Vw2 of the workpiece W during measurement in the increasing direction, the number of repeated occurrences of each abrasive grain mark on the grinding surface W1 is increased, and the frequency when the first spiral data D1 is Fourier transformed in the high-frequency component analysis 223a is adjusted in the increasing direction.

[0045] As shown in Figure 8, in the phase alignment process 224, the divided data D11, which has had noise components removed and is set to be shifted at each position in the Z-axis direction and each position in the circumferential direction C, is phase-aligned so that it is located at the same part in the circumferential direction C, and phase-aligned data D12 is created. When this phase alignment is performed, the phase in the circumferential direction C is finely adjusted so that the uneven surface state of the grinding circumferential surface W1 in the phase-aligned data D12 is not discontinuous in the Z-axis direction. In this way, as shown in Figure 9, a chatter map M1 for the first spiral data D1 from the acceleration sensor 32 is created.

[0046] The radius map M2 is created by signal processing the second spiral data D2 and rotation data D3, which are signal data from the displacement sensor 34 of the sizing device 3. This signal processing is divided into gain correction processing 221b and low-frequency component analysis 223b for the second spiral data D2, and gain correction processing 221c and roundness analysis 225 for the rotation data D3. Then, as a roundness component removal process 226, the second spiral data D2 after low-frequency component analysis 223b and the rotation data D3 after roundness analysis 225 are used to obtain the change in radius in the Z-axis direction on the grinding surface W1 of the workpiece W.

[0047] In the gain correction processing 221b and 221c of the second spiral data D2 and rotation data D3, it is taken into consideration that the displacement sensor 34 has input / output characteristics that attenuate the output signal for input signals exceeding a specific frequency, and the output signal for each frequency is corrected so that the variation in the output signal of the displacement sensor 34 is reduced.

[0048] In the low-frequency component analysis 223b of the second spiral data D2, a Fast Fourier Transform (FFT) is performed on the second spiral data D2. The change in radius of the grinding surface W1 of the workpiece W in the Z-axis direction and the circumferential direction C is then extracted as Fourier transform data of the low-frequency component, expressed in terms of frequency (or period) and amplitude.

[0049] Subsequently, specific low-frequency components are extracted to remove noise components from the Fourier transform data of the low-frequency components, and an inverse fast Fourier transform (inverse FFT) is performed on the extracted low-frequency components. This removes the noise components from the second spiral data D2. The noise components include noise caused by mechanical vibrations, etc. The second spiral data D2 also includes the change in radius in the Z-axis direction of the grinding surface W1 of the workpiece W.

[0050] In the roundness analysis 225 of the rotation data D3, the extent to which irregularities due to roundness errors occur on the uneven surface state of the grinding circumferential surface W1 of the workpiece W is analyzed. Then, in the roundness removal process 226, the rotation data D3 after the roundness analysis 225 is compared with the second spiral data D2 after the low-frequency component analysis 223b is performed, and the influence of irregularities in the circumferential C and Z axis directions, as well as the influence of irregularities due to roundness errors, is removed from the second spiral data D2. Then, radius data showing the difference in radius at each position in the Z axis direction is obtained, and a radius map M2 is created based on the radius data, as shown in Figure 10.

[0051] Subsequently, as shown in Figure 11, the chatter map M1 and the radius map M2 are combined to create a surface texture map M. In the surface texture map M, the uneven surface condition of the grinding surface W1 of the workpiece W, which reflects the difference in radius at each position in the Z-axis direction, is visualized by color coding, etc. In Figure 11, the width transferred from one rotation of the grinding wheel 12 in the circumferential direction C of the grinding surface W1 of the workpiece W is shown.

[0052] 7. Adjustment of the rotational speed Vw2 during measurement of the workpiece W. Next, the adjustment of the rotational speed Vw2 during measurement of the workpiece W will be described. In this embodiment, the surface texture estimation unit 22 adjusts the rotational speed Vw2 during measurement of the workpiece W, taking into account the natural vibration characteristics K of the acceleration sensor 32, and performs high-frequency component analysis 223a on the first spiral data D1 obtained by the acceleration sensor 32. Specifically, as shown in Figure 6, the evaluation information storage unit 23, machining condition acquisition unit 24, rotational speed setting unit 25, and workpiece spindle drive unit 26 of the surface texture estimation device 2 adjust the rotational speed Vw2 during measurement of the uneven surface state of the grinding circumferential surface W1 of the workpiece W, thereby reducing the influence of noise due to the natural vibration characteristics K of the acceleration sensor 32 on the estimation of the surface texture S.

[0053] As shown in Figure 12, the evaluation information storage unit 23 stores evaluation information I, which is set for the natural vibration characteristic K, which is the frequency response characteristic of the output data output from the acceleration sensor 32 of the dimensionality device 3, where the variation ΔG of the gain G of the natural vibration characteristic K is within a predetermined value, and the variation ΔG of the gain G of the natural vibration characteristic K is lower than the gain stable frequency band K0 and exceeds a predetermined value.

[0054] In the gain-instability frequency band K1, not only is the gain G, which indicates the input-output ratio of the acceleration sensor 32, significantly greater than 1, but the variation ΔG of the gain G is also large. In the high-frequency component analysis 223a of the surface texture estimation unit 22, it is preferable not to use the first spiral data D1 of frequencies included in the gain-instability frequency band K1.

[0055] Furthermore, in this embodiment, the evaluation information I in the evaluation information storage unit 23 includes, in addition to the gain stable frequency band K0 and the gain unstable frequency band K1, an anti-resonant frequency band K2 that has a higher frequency than the gain stable frequency band K0 and where the gain G of the natural vibration characteristic K fluctuates up and down due to the anti-resonant characteristics of the acceleration sensor 32, making gain compensation impossible. Fluctuating gain G means that when the frequency changes, there are consecutive cases where the gain G fluctuates more than 1 and cases where the gain G fluctuates less than 1. Gain compensation impossible means that it is difficult to correct the gain G as the input / output ratio of the acceleration sensor 32 to be close to 1. In the high-frequency component analysis 223a of the surface texture estimation unit 22, it is preferable not to use the first spiral data D1 of frequencies included in the anti-resonant frequency band K2.

[0056] To obtain the natural vibration characteristics K of the acceleration sensor 32, before the operation of the surface texture estimation device 2 and the grinding machine 1, vibrations of varying frequencies are applied to the contact member 31 of the sizing device 3 by the excitation device 5 shown in Figure 2, and the change in output by the acceleration sensor 32 is measured. The degree of increase or decrease in the amplitude of the output of the acceleration sensor 32 with respect to the amplitude of the vibration input to the acceleration sensor 32 by the excitation device 5 is obtained as the frequency response characteristic of the acceleration sensor 32 with respect to the gain G. The natural vibration characteristics K are determined by focusing on the magnitude of the variation (fluctuation) ΔG of the gain G for each frequency in this frequency response characteristic with respect to the gain G, and the gain unstable frequency band K1, the gain stable frequency band K0, and the anti-resonance frequency band K2 are set in order from the lower frequency side.

[0057] Gain G is defined as 1 when the output amplitude is equal to the input amplitude, greater than 1 when the output amplitude increases, and less than 1 when the output amplitude decreases. The variation in gain G, ΔG, indicates the range of variation in gain G within a predetermined frequency range.

[0058] The gain instability frequency band K1 is set as the frequency range where the gain G approaches 1 from a large value around 100, and where the variation ΔG exceeds a predetermined value. The gain stability frequency band K0 is set as the frequency range where the gain G is close to 1, and where the variation ΔG falls within a predetermined value. The anti-resonance frequency band K2 is set as the frequency range where the gain G swings significantly between the side greater than 1 and the side less than 1. The predetermined value of the variation ΔG is determined by focusing on the allowable range of gain G variation within a predetermined frequency range. The allowable range of gain G variation is set, for example, as a numerical range such as ±1.2 times the gain G variation.

[0059] In this embodiment, the gain-stable frequency band K0 of the evaluation information I in the evaluation information storage unit 23 is set to a gain-compensable resonance frequency band K3 in which the gain G of the natural vibration characteristic K fluctuates up and down due to the resonance characteristics of the acceleration sensor 32. In the resonance frequency band K3, the gain G fluctuates up and down, while the variation ΔG of the gain G is within a predetermined value. The resonance frequency band K3 is a frequency band in which it is possible to correct the gain G as the input / output ratio of the acceleration sensor 32 to be close to 1, and the first spiral data D1 of the frequencies included in the resonance frequency band K3 may be used in the high-frequency component analysis 223a of the surface texture estimation unit 22. However, the first spiral data D1 of the frequencies included in the resonance frequency band K3 may not be used in order to simplify the correction of the gain G.

[0060] As shown in Figure 6, the machining condition acquisition unit 24 acquires the machining rotation speed Vt1 of the grinding wheel 12 and the machining rotation speed Vw1 of the workpiece W during grinding by the grinding wheel 12. The rotation speed of the grinding wheel 12 is faster than the rotation speed of the workpiece W, and the machining rotation speed Vt1 of the grinding wheel 12 is greater than the machining rotation speed Vw1 of the workpiece W.

[0061] As shown in Figures 6 and 12, the rotational speed setting unit 25 sets the rotational speed Vw2 of the workpiece W to the fundamental frequency component F1 and the harmonic component F2 of the fundamental frequency component F1, which will be included in the first spiral data D1, which is the signal data of the acceleration sensor 32, when the rotational speed Vw2 of the workpiece W during measurement is assumed to be included in the gain stable frequency band K0, based on the relationship between the rotational speed Vt1 of the grinding wheel 12 during machining and the rotational speed Vw1 of the workpiece W during machining, which is acquired by the machining condition acquisition unit 24.

[0062] The fundamental frequency component F1 represents the frequency at which the accelerometer 32 of the measuring device 3 repeatedly detects the irregularities resulting from grinding marks by numerous abrasive grains of the grinding wheel 12 on the grinding surface W1 of the workpiece W when the workpiece W is rotated at the measurement rotation speed Vw2. The harmonic component F2 is represented as an integer multiple of the fundamental frequency component F1, and is expressed as a frequency of the second order or higher when the fundamental frequency component F1 is considered the first order component. The harmonic component F2 is generated by the characteristics of the accelerometer 32.

[0063] The high-frequency component analysis 223a of the surface texture estimation unit 22 sets the rotational speed Vw2 of the workpiece W during measurement to be faster than the rotational speed Vw1 of the workpiece W during machining, when the uneven surface condition of the grinding surface W1 of the workpiece W is measured by the sizing device 3. By increasing the rotational speed Vw2 of the workpiece W during measurement, the frequency of the frequency component in the divided data D11 of the first spiral data D1 obtained by the acceleration sensor 32, which is used in the high-frequency component analysis 223a, shifts to a higher frequency. In other words, by increasing the rotational speed Vw2 of the workpiece W during measurement, the period in which the unevenness of the abrasive marks caused by the numerous abrasive grains of the grinding wheel 12 is repeatedly measured when the uneven surface condition of the grinding surface W1 of the workpiece W is measured by the acceleration sensor 32 is shortened.

[0064] Furthermore, in this embodiment, the rotation speed setting unit 25 sets the rotation speed Vw2 of the workpiece W during measurement such that at least two selected frequencies Fs are excluded from the gain unstable frequency band K1 and the anti-resonant frequency band K2, and are included in the gain stable frequency band K0. With this configuration, at least two selected frequencies Fs are also excluded from the anti-resonant frequency band K2, improving the accuracy of the high-frequency component analysis 223a.

[0065] Figure 13 shows the Fourier transform data of high-frequency components, expressed by the relationship between frequency and amplitude, obtained by performing a fast Laplace transform on the first spiral data D1 in the high-frequency component analysis 223a of the surface texture estimation unit 22. Figure 13 shows the state in which the four selectable frequencies Fs—harmonic component F21 as the second-order component, harmonic component F22 as the third-order component, harmonic component F23 as the fourth-order component, and harmonic component F24 as the fifth-order component—are included in the gain-stable frequency band K0, which includes the resonant frequency band K3. If it is acceptable for the selectable frequencies Fs to be included in the resonant frequency band K3, the selection of the selectable frequencies Fs is relatively easy.

[0066] Furthermore, in this embodiment, the rotation speed setting unit 25 sets the rotation speed Vw2 during measurement of the workpiece W such that at least two selected frequencies Fs are included in the gain stable frequency band K0, excluding the gain unstable frequency band K1 and the anti-resonant frequency band K2, and excluding the resonant frequency band K3. With this configuration, at least two selected frequencies Fs are also excluded from the resonant frequency band K3, further improving the accuracy of the high-frequency component analysis 223a.

[0067] Figure 14, similar to Figure 13, shows the Fourier transform data of the high-frequency components, expressed by the relationship between frequency and amplitude, obtained by performing a fast Laplace transform on the first spiral data D1 in the high-frequency component analysis 223a of the surface texture estimation unit 22. Figure 14 shows the state in which the three selected frequencies Fs—the second-order harmonic component F21, the third-order harmonic component F22, and the fifth-order harmonic component F24—are included in the gain-stable frequency band K0, excluding the resonant frequency band K3.

[0068] The high-frequency component analysis 223a of the surface texture estimation unit 22 extracts a plurality of calculation frequencies Fc within a predetermined range, each containing a selected frequency Fs, based on the first spiral data D1, which is signal data from the acceleration sensor 32 acquired by the signal data acquisition unit 21 when the workpiece W is rotated at the measurement rotation speed Vw2 during measurement. Based on the amplitudes of the plurality of calculation frequencies Fc, the uneven surface state of the grinding circumferential surface W1 is estimated as the surface texture S.

[0069] As shown in Figures 13 and 14, in the high-frequency component analysis 223a of this embodiment, a plurality of predetermined range calculation frequencies Fc are extracted from the divided data D11 of the first spiral data D1 that has undergone a Fast Fourier Transform. The calculation frequencies Fc are extracted as a range of frequencies with a predetermined width for at least two selected frequencies Fs from the fundamental frequency component F1 and the harmonic components F2 which are integer multiples of the fundamental frequency component F1. For example, the calculation frequencies Fc may be extracted by setting predetermined widths on the higher and lower sides of the frequency centered on the frequency of the fundamental frequency component F1 or the harmonic component F2.

[0070] Furthermore, a portion of the calculation frequency Fc, which is a frequency range with a predetermined width, may be included in the resonant frequency band K3. Alternatively, the entire calculation frequency Fc, which is a frequency range with a predetermined width, may not be included in the resonant frequency band K3. It is also preferable that the gain instability frequency band K1 and the anti-resonant frequency band K2 not include any portion of the calculation frequency Fc.

[0071] In the high-frequency component analysis 223a, an inverse fast Fourier transform is performed on the frequency range extracted as multiple calculation frequencies Fc from the divided data D11 of the first spiral data D1 which has undergone a fast Fourier transform. This allows the chatter map M1 created by the surface texture estimation unit 22 to show the irregularities as abrasive marks caused by numerous abrasive grains that appear on the grinding surface W1 of the workpiece W with higher accuracy, after removing the noise component.

[0072] In this embodiment, the rotation speed setting unit 25 sets the rotation speed Vw2 during measurement of the workpiece W such that at least two selected frequencies Fs from the fundamental frequency component F1 as the first-order component and the harmonic components F2 (F21, F22, F23, F24) which are 2 to 5 times the fundamental frequency component F1 as second to fifth-order components are included in the gain-stable frequency band K0, excluding the resonant frequency band K3. By limiting the harmonic components F2 to the fifth-order component, the accuracy of creating the surface texture map M can be maintained.

[0073] Furthermore, it is preferable that the rotation speed setting unit 25 sets the rotation speed Vw2 during measurement of the workpiece W such that the four selectable frequencies Fs, which are harmonic components F2 (F21, F22, F23, F24) that are 2 to 5 times the fundamental frequency component F1 as a second to fifth order component, are included in the gain-stable frequency band K0, excluding the resonant frequency band K3.

[0074] In experiments, when the rotational speed Vw2 of the workpiece W during measurement was significantly faster than the rotational speed Vw1 of the workpiece W during processing, such that the fundamental frequency component F1 as the first-order component was included in the gain-stable frequency band K0, vibrations occurred in the components of the grinding machine 1, making it impossible to estimate the surface texture S with high accuracy. It has been found that in order to estimate the surface texture S with high accuracy, it is sometimes better to set the rotational speed Vw2 of the workpiece W during measurement so that the fundamental frequency component F1 as the first-order component is included in the gain-unstable frequency band K1. Therefore, by creating a surface texture map M based on the harmonic components F2 from the second to the fifth order, the accuracy of the surface texture map M can be easily maintained.

[0075] For example, it may not be easy to ensure that three or four selectable frequencies Fs, which are harmonic components F2 (F21, F22, F23, F24) from the second to fifth order, are included in the gain-stable frequency band K0, excluding the gain-unstable frequency band K1, the anti-resonant frequency band K2, and the resonant frequency band K3. Taking such cases into consideration, the rotation speed setting unit 25 may repeatedly perform calculations by changing the rotation speed Vw2 during measurement of the assumed workpiece W so that at least two selectable frequencies Fs are included in the gain-stable frequency band K0, and determine the rotation speed Vw2 during measurement of the workpiece W. For example, it is preferable to repeatedly perform calculations by changing the rotation speed Vw2 during measurement of the assumed workpiece W so that four selectable frequencies Fs, which are harmonic components F2 (F21, F22, F23, F24) from the second to fifth order, are included in the gain-stable frequency band K0, excluding the resonant frequency band K3.

[0076] As shown in Figure 6, the workpiece spindle drive unit 26 maintains the rotational speed of the workpiece W at the measurement rotational speed Vw2 when the uneven surface condition of the grinding circumferential surface W1 of the workpiece W is measured by the sizing device 3. The workpiece spindle drive unit 26 is configured within the control device 10 and acquires information on the measurement rotational speed Vw2 of the workpiece W from the rotational speed setting unit 25, and rotates the spindle rotation motor 14a of the headstock 14 at the measurement rotational speed Vw2.

[0077] The measurement rotational speed Vw2 of the workpiece W, controlled by the evaluation information storage unit 23, the machining condition acquisition unit 24, the rotational speed setting unit 25, and the workpiece spindle drive unit 26, only needs to be adjusted when at least one of the machining conditions, Vt1 of the grinding wheel 12 and Vw1 of the workpiece W, is changed. Once the machining rotational speed Vw1 of the workpiece W has been set, adjustment of the measurement rotational speed Vw2 of the workpiece W does not need to be performed until at least one of the machining conditions, Vt1 of the grinding wheel 12 and Vw1 of the workpiece W, is changed.

[0078] 8. Relationship between the grinding wheel 12 and the uneven surface of the grinding circumferential surface W1 of the workpiece W. Next, the relationship between the grinding wheel 12 and the uneven surface of the grinding circumferential surface W1 of the workpiece W will be explained. The grinding wheel 12 is composed of numerous abrasive grains, a binder that binds the abrasive grains together, and pores. The grinding wheel surface (outer surface) 121 of the grinding wheel 12 has an uneven surface formed by numerous abrasive grains. The grinding circumferential surface W1 of the workpiece W, which is ground by the outer surface of the grinding wheel 12, has an uneven surface state formed by the transfer of the uneven surface formed by numerous abrasive grains.

[0079] As shown in Figure 3, the outer diameter of the grinding wheel 12 is larger than the outer diameter of the grinding surface W1 of the workpiece W. Also, the rotational speed Vt1 of the grinding wheel 12 during machining is larger than the rotational speed Vw1 of the workpiece W during machining. During machining, the grinding wheel 12 rotates multiple times while the workpiece W rotates once. During one rotation of the workpiece W, irregularities are repeatedly formed on the grinding surface W1 of the workpiece W as a result of the collision of numerous abrasive grains on the outer surface of the grinding wheel 12.

[0080] Figure 3 shows the case where the grinding wheel 12 and the workpiece W rotate in the same direction at their contact point. The grinding wheel 12 and the workpiece W may rotate in opposite directions at their contact point.

[0081] During machining, when using the machining rotation speed (rotational speed) Vt1 [rps] of the grinding wheel 12 and the machining rotation speed (rotational speed) Vw1 [rps] of the workpiece W, each abrasive grain of the grinding wheel 12 collides with the grinding surface W1 of the workpiece W a number of times determined based on Vt1 / Vw1 during one rotation of the workpiece W. The machining rotation speed Vw1 of the workpiece W should preferably be the rotation speed of the workpiece W at the end of machining. Grinding wheel surface information for the number of times the abrasive grains collide is transferred to the grinding surface W1 of the workpiece W, and this is measured as vibration of a specific frequency by the acceleration sensor 32 of the sizing device 3.

[0082] On the other hand, if the rotational speed Vw2 of the workpiece W during measurement is changed from the rotational speed Vw1 of the workpiece W during machining, vibrations of a frequency determined based on Vt1 / Vw1 × Vw2 will be measured by the acceleration sensor 32 of the sizing device 3. In the surface texture estimation device 2, vibrations of this frequency are measured as the fundamental frequency component F1. The rotational speed Vw2 of the workpiece W during measurement may be changed not only by a large amount compared to the rotational speed Vw1 of the workpiece W during machining, but may also be changed to a smaller amount compared to the rotational speed Vw1 of the workpiece W during machining.

[0083] The rotational speed setting unit 25 of this embodiment determines the fundamental frequency component F1 by multiplying the rotational speed ratio Vr(Vt1 / Vw1), which is the ratio of the rotational speed Vt1 of the grinding wheel 12 to the rotational speed Vw1 of the workpiece W during machining, by the measured rotational speed (rotational speed) Vw2 [rps] of the workpiece W. The fundamental frequency component F1 is determined based on Vr × Vw2.

[0084] 9. Other components The signal data acquisition unit 21 has a configuration that prevents frequency bands of 1 / 2 or more of the sampling frequency from affecting the results of the high-frequency component analysis 223a of the surface texture estimation unit 22 as aliasing noise when measuring the uneven surface condition of the grinding surface W1 of the workpiece W by the sizing device 3. Therefore, it is preferable to further include a low-pass filter 211 in the surface texture estimation device 2.

[0085] Specifically, as shown in Figure 6, the signal data acquisition unit 21 acquires signal data via a low-pass filter 211 whose cutoff frequency is set to a frequency higher than the gain stability frequency band K0. The cutoff frequency should be set to the upper limit frequency of the natural vibration characteristic K stored in the evaluation information storage unit 23. In this embodiment, the evaluation information storage unit 23 stores natural vibration characteristics K up to 2500 Hz. The cutoff frequency of the low-pass filter 211 should be set to sufficiently attenuate frequency components above 2500 Hz.

[0086] Furthermore, the signal data acquisition unit 21 acquires signal data at a sampling frequency of at least twice the cutoff frequency. In this embodiment, the sampling frequency of the signal data acquisition unit 21 is set to 5000 Hz or higher, which is at least twice the upper limit frequency of the natural vibration characteristic K stored in the evaluation information storage unit 23.

[0087] As shown in Figure 1, the grinding machine 1 may have a determination unit 101 that performs various judgments using a surface texture map M estimated and created by a surface texture estimation device 2. The determination unit 101 may determine whether the grinding of the workpiece W is good or bad based on the surface texture map M. The surface texture estimation device 2 stores a reference surface texture map that serves as a standard for when good grinding is performed on the workpiece W, and the determination unit 101 compares the surface texture map M created each time the workpiece W is ground in the grinding machine 1 or a predetermined number of times with the reference surface texture map. The determination unit 101 may then determine that the grinding is poor if the surface texture S of the ground surface W1 of the ground workpiece W has deteriorated compared with the surface texture in the reference surface texture map.

[0088] The determination unit 101 may determine, based on the surface condition map M, whether it is necessary to adjust the machining conditions of the workpiece W to be machined in subsequent times. The machining conditions of the workpiece W include the rotational speed Vt1 of the grinding wheel 12 during machining and the rotational speed Vw1 of the workpiece W during machining, as well as the feed rate (depth of cut) of the grinding wheel 12 in the X-axis direction. In this case as well, the surface condition estimation device 2 stores a reference surface condition map, and the determination unit 101 compares the surface condition map M, which is created each time the workpiece W is ground in the grinding machine 1 or a predetermined number of times, with the reference surface condition map. The determination unit 101 may then determine that it is necessary to adjust the machining conditions if the surface condition S of the ground surface W1 of the ground workpiece W has deteriorated compared with the surface condition in the reference surface condition map.

[0089] The determination unit 101 may determine the need to adjust the timing of the grinding wheel 12's correction based on the surface condition map M. The timing of the grinding wheel 12's correction indicates the frequency of correction of the grinding wheel 12. Corrections of the grinding wheel 12 include truing (shaping) as correction of the runout and shape of the grinding wheel 12, and dressing (straightening) as correction of the amount of abrasive grain protrusion and the creation of cutting edges of the abrasive grains. In this case as well, the surface condition estimation device 2 stores a reference surface condition map, and the determination unit 101 compares the surface condition map M, which is created each time the workpiece W is ground in the grinding machine 1, with the reference surface condition map. The determination unit 101 may then determine that it is necessary to shorten the frequency of correction of the grinding wheel 12 when the surface condition S of the ground surface W1 of the ground workpiece W has deteriorated compared with the surface condition in the reference surface condition map.

[0090] 10. Effects In the surface texture estimation device 2 of this embodiment, evaluation information I of the natural vibration characteristic K obtained by the acceleration sensor 32 is stored, and the influence of the natural vibration characteristic K on the estimation of the surface texture S is mitigated by intentionally changing the rotational speed Vw2 of the workpiece W during measurement relative to the rotational speed Vw1 of the workpiece W, taking the natural vibration characteristic K into consideration. Specifically, in the evaluation information I of the evaluation information storage unit 23, the gain stable frequency band K0, the gain unstable frequency band K1, the anti-resonance frequency band K2, and the resonance frequency band K3 for the natural vibration characteristic K are set.

[0091] Then, the rotation speed setting unit 25 sets the rotation speed Vw2 of the workpiece W during measurement based on the relationship between the machining rotation speed Vt1 of the grinding wheel 12 and the machining rotation speed Vw1 of the workpiece W, such that at least two selected frequencies Fs of the harmonic components F2 from the second to fifth order components, which will be included in the first spiral data D1 of the signal data acquisition unit 21, are included in the gain stable frequency band K0, excluding the resonant frequency band K3.

[0092] By setting the rotational speed Vw2 of the workpiece W during measurement in this way, the first spiral data D1 acquired by the signal data acquisition unit 21 during measurement can be prevented from being affected by noise from the gain instability frequency band K1, anti-resonance frequency band K2, and resonance frequency band K3 of the natural vibration characteristic K. As a result, the error factors are reduced, which improves the accuracy of the surface texture estimation unit 22's estimation of the surface texture S and improves the accuracy of the surface texture map M.

[0093] Therefore, according to the surface texture estimation device 2 of this embodiment, the influence of noise due to the gain instability frequency band K1 of the natural vibration characteristic K can be excluded, thereby improving the accuracy of the surface texture map M.

[0094] The present invention is not limited to these embodiments, and further different embodiments can be constructed without departing from the spirit of the invention. Furthermore, the present invention includes various modifications, modifications within the equivalent range, etc. Moreover, various combinations and forms of components conceivable from the present invention are also included in the technical concept of the present invention. [Explanation of Symbols]

[0095] 1. Grinding machine 12 Grinding Wheel 2 Surface texture estimation device 20 Arithmetic Processing Unit 21 Signal data acquisition unit 22 Surface texture estimation section 23 Evaluation Information Storage Unit 24 Machining condition acquisition section 25 Rotation speed setting section 3 Sizing device 32 Accelerometer 34 Displacement Sensor W Workpiece W1 Grinding surface S surface texture F1 Fundamental frequency component F2 harmonic component Fs Selectable Frequency Fc calculation frequency M Surface properties map G Gain ΔG variation K Natural vibration characteristics I. Evaluation Information K0 Gain Stability Frequency Band K1 Gain Instability Frequency Band K2 anti-resonance frequency band K3 Resonant Frequency Band Vt1 Grinding wheel rotation speed during machining Vw1 Workpiece rotational speed Vw2 Workpiece measurement rotation speed

Claims

1. A surface texture estimation device for estimating the surface texture of the ground surface of a workpiece that has been ground by a grinding wheel on a grinding machine, A measuring device provided in the grinding machine for measuring the diameter of the grinding surface while grinding the workpiece while it is rotating, A calculation processing device that estimates the surface properties of the grinding surface of the workpiece, based on signal data output from a sensor provided in the measuring device during measurement after grinding of the workpiece, and which is caused by the grinding wheel onto which the uneven surface of the grinding wheel has been transferred. Equipped with, The aforementioned arithmetic processing unit is A signal data acquisition unit acquires the signal data when the workpiece is rotated and the sizing device and the workpiece are moved relative to each other in the axial direction of the workpiece, corresponding to the uneven surface state of the grinding circumferential surface. An evaluation information storage unit stores evaluation information in which, when a vibration with a varied frequency is input to the constant-distance device, the natural vibration characteristics as the frequency response characteristics of the output data output from the sensor include a gain-stable frequency band in which the variation in the gain of the natural vibration characteristics is within a predetermined value, and a gain-unstable frequency band in which the frequency is lower than the gain-stable frequency band and the variation in the gain of the natural vibration characteristics exceeds the predetermined value. A processing condition acquisition unit that acquires the processing rotation speed Vt1 of the grinding wheel and the processing rotation speed Vw1 of the workpiece during the grinding process of the workpiece by the grinding wheel, A rotation speed setting unit sets the rotation speed Vw2 of the workpiece during measurement, based on the relationship between the rotation speed Vt1 of the grinding wheel during machining and the rotation speed Vw1 of the workpiece during machining, such that at least two selected frequencies obtained based on the fundamental frequency components included in the signal data are included in the gain stabilization frequency band, when the rotation speed Vw2 of the workpiece during measurement is assumed to be at the time of measurement. A surface texture estimation unit estimates the uneven surface state of the grinding circumferential surface as the surface texture based on the signal data acquired by the signal data acquisition unit when the workpiece is rotated at the measurement rotation speed Vw2 during the measurement, Equipped with, The at least two selected frequencies are, The fundamental frequency component and at least one harmonic component among the plurality of harmonic components of the fundamental frequency component, Alternatively, A surface property estimation device comprising at least two harmonic components from among multiple harmonic components of the aforementioned fundamental frequency component.

2. The aforementioned rotation speed setting unit is The surface texture estimation apparatus according to claim 1, wherein the rotational speed Vw2 of the workpiece during measurement is determined by repeatedly performing calculations with the assumed rotational speed Vw2 of the workpiece changed such that at least two of the selected frequencies are included in the gain-stabilizing frequency band.

3. The surface texture estimation unit is, The surface texture estimation device according to claim 1 or 2, wherein, during the measurement, when the workpiece is rotated at the measurement rotation speed Vw2, a plurality of calculation frequencies within a predetermined range, each including the selected frequency, are extracted based on the signal data acquired by the signal data acquisition unit, and the uneven surface state of the grinding circumferential surface is estimated as the surface texture based on the amplitude of the calculation frequencies.

4. The aforementioned rotation speed setting unit is The surface texture estimation apparatus according to claim 1 or 2, wherein the fundamental frequency component is determined by multiplying the rotational speed ratio Vr, which is the ratio of the rotational speed Vt1 of the grinding wheel during machining to the rotational speed Vw1 of the workpiece during machining, by the rotational speed Vw2 of the workpiece during measurement.

5. The evaluation information in the evaluation information storage unit includes: In addition to the gain-stable frequency band and the gain-unstable frequency band, there is an anti-resonant frequency band that has a higher frequency than the gain-stable frequency band and in which the gain of the natural vibration characteristic fluctuates up and down due to the anti-resonant characteristic, making gain compensation impossible. The aforementioned rotation speed setting unit is The surface texture estimation apparatus according to claim 1 or 2, wherein the rotational speed Vw2 of the workpiece during measurement is set such that at least two of the selected frequencies are included in the gain stable frequency band, excluding the gain unstable frequency band and the anti-resonant frequency band.

6. The gain stabilization frequency band of the evaluation information in the evaluation information storage unit includes: The resonant frequency band includes a gain-compensable resonant frequency range in which the gain of the aforementioned natural vibration characteristics fluctuates up and down due to the resonant characteristics. The aforementioned rotation speed setting unit is The surface texture estimation apparatus according to claim 1 or 2, wherein the rotational speed Vw2 of the workpiece during measurement is set such that at least two of the selected frequencies are included in the gain stable frequency band, excluding the gain unstable frequency band and excluding the resonant frequency band.

7. The aforementioned rotation speed setting unit is The surface texture estimation apparatus according to claim 1 or 2, wherein the rotational speed Vw2 of the workpiece during measurement is set such that at least two of the selected frequencies, which are the fundamental frequency component as a first-order component and the harmonic components that are 2 to 5 times the fundamental frequency component as second to fifth-order components, are included in the gain-stable frequency band.

8. The aforementioned rotation speed setting unit is The surface texture estimation apparatus according to claim 1 or 2, wherein the rotational speed Vw2 of the workpiece during measurement is set such that four selected frequencies, which are harmonic components that are 2 to 5 times the fundamental frequency components as 2nd to 5th order components, are included in the gain-stable frequency band.

9. The signal data acquisition unit, The surface texture estimation apparatus according to claim 1 or 2, wherein the signal data is acquired through a low-pass filter whose cutoff frequency is set to a frequency higher than the gain-stabilizing frequency band, and the signal data is acquired with a sampling frequency of twice or more the cutoff frequency.

10. The surface texture estimation device according to claim 1 or 2, wherein the sizing device comprises a contact member that contacts the grinding circumferential surface of the workpiece and displaces in accordance with the uneven surface condition of the grinding circumferential surface, and an acceleration sensor as the sensor disposed on the contact member.

11. A grinding machine equipped with a surface texture estimation device according to claim 1 or 2.

12. The grinding machine according to claim 11, further comprising a determination unit that performs at least one of the following based on the surface properties estimated by the surface properties estimation device: determining whether the grinding of the workpiece is good or bad, determining whether it is necessary to adjust the processing conditions of the workpiece to be processed in subsequent times, and determining whether it is necessary to adjust the timing of the grinding wheel correction.

Images