Method and apparatus for detecting the mounting status of a tool holder, and machine tools

Abstract

To accurately align a phase of normal state shape data (master data) with a phase of actual actuation shape data (measurement data) regardless of notch and accuracy of a flange part of a tool holder.SOLUTION: A tool holder fitting state detecting device comprises: a sensor 1 for measuring a shape of a flange part 11B; phase restricted correlation function calculating means 30 for separately performing discrete Fourier transformation on master data 31 and measurement data 32 based on normal state shape data which are measured in a normal fitting state in advance and extracting only a phase component, and then performing reverse discrete Fourier transformation on the acquired phase difference signal, to calculate a phase restricted correlation function; waveform position detecting means 36 for determining a phase deviation amount between the master data 31 and the measurement data 32 from the phase restricted correlation function; waveform position aligning means 37 for, on the basis of the phase deviation amount, performing phase alignment of the master data 31 and the measurement data 32; and fitting abnormality determining means 38 for determining fitting abnormality, after performing the phase alignment.SELECTED DRAWING: Figure 6

Description

[Technical Field]

[0001] The present invention relates to machine tools, including NC (numerical control) machining centers and machining machines, that process workpieces (objects to be processed, objects to be measured), and more particularly to a method and apparatus for detecting the mounting status of a tool holder, which is equipped with an automatic tool changer (ATC) that appropriately selects and attaches / detaches tools, and which determines whether the tool holder to which the tool is attached is properly mounted on the spindle, as well as to a machine tool. [Background technology]

[0002] A machining center (MC) is a device that automatically selects various tools according to the machining process and automatically mounts them on the spindle to perform various types of machining. In this MC, tool changes are automated. This process is performed by an automatic tool changer (ATC), which automatically retrieves the tool holder with the tool attached from the tool magazine and mounts it to the spindle. The tool holder has a conical fitting portion, which is fitted into a conical fitting portion formed on the spindle. However, if chips or other debris adhere to this fitting portion, the spindle will be mounted in a bent position. If machining is performed in this state, tool runout will occur, and the machining accuracy of the workpiece will be significantly reduced.

[0003] Machining centers are evolving to include models that rotate a worktable at high speed and can perform turning by attaching a cutting tool to the spindle, as well as models that use probes for dimensional measurement instead of tools. It is desirable to equip the machining center with probes for dimensional measurement to automatically measure the dimensions of the workpiece during machining or once the machining process is complete. In particular, in order to reflect machining errors in the workpiece in the machining process, it is known that optical measuring instruments, which have a light source that emits light for measurement and a light detection means for detection, can be attached to and detached from the tool spindle position of the machining center.

[0004] Patent Document 1 describes a system that provides a sensor to measure the shape of the outer circumferential surface of the flange portion of a tool holder to which a probe is attached, and determines an abnormal attachment based on the difference between normal state shape data, which is shape data measured in advance in a normal attachment state, and actual operation shape data, which is measured during actual operation.

[0005] Furthermore, Patent Document 2 describes performing FFT analysis on normal state shape data and actual operating shape data to calculate the amplitude and phase of each peak component, determine the true eccentricity, and then determine whether there is an abnormality in the installation based on that.

[0006] Furthermore, Patent Document 3 describes that when there are two notches in the flange portion of the tool holder, the detection data of the surface position of the outer circumferential surface of the flange detected by the sensor at a predetermined period is interpolated to calculate interpolated detection data with an interpolation period shorter than the predetermined period, and then the data invalidation period is determined.

[0007] Furthermore, Patent Document 4 describes a waveform alignment method using the Phase-Only Correlation (POC) method, which involves performing a Discrete Fourier Transform on the input waveform signal and the reference waveform signal to separate them into amplitude and phase components, extracting only the phase component, and then performing an inverse Discrete Fourier Transform on the phase difference signal to calculate the phase-only correlation function. [Prior art documents] [Patent Documents]

[0008] [Patent Document 1] Japanese Patent Publication No. 2018-89738 [Patent Document 2] Japanese Patent Publication No. 2002-200542 [Patent Document 3] Japanese Patent Publication No. 2008-93750 [Patent Document 4] Japanese Patent Publication No. 2007-240170 [Overview of the project] [Problems that the invention aims to solve]

[0009] In the above-mentioned prior art, the devices described in Patent Documents 1 and 2 make it difficult to accurately measure runout caused by mounting errors that occur when attaching the probe to the spindle position if there is a notch in the flange portion of the tool holder.

[0010] Furthermore, even if the description in Patent Document 3 is applied to the methods described in Patent Documents 1 and 2, the error in runout measurement due to the effect of notches in the flange can be reduced by the use of complementary data. However, if there is a shift in the sampling start position or unevenness in the rotational speed when detecting the actual operating shape data, it becomes difficult to align the phase between the normal state shape data and the actual operating shape data, making it impossible to measure mounting errors with high accuracy.

[0011] Furthermore, simply applying the method described in Patent Document 4 does not allow for easy and accurate phase alignment and precise detection of mounting errors. For example, if runout occurs due to chip jamming, if there are two notches and the sampling start position is significantly shifted, or if there is a lot of noise, complex processing is required, such as fitting a correlation peak model to a phase-limited correlation function, using a phase-limited correlation function with a spectral weighting function, or using a band-limited phase-limited correlation function with a limited frequency band.

[0012] The objective of the present invention is to solve the problems of the prior art described above, to enable easy and highly accurate alignment of the phase between normal state shape data (master data) and actual operating shape data (measurement data), regardless of the notch and precision of the flange portion of the tool holder, and to improve the detection accuracy of abnormal mounting of the tool holder to the spindle (chuck error). [Means for solving the problem]

[0013] To achieve the above objective, the present invention provides a tool holder mounting state detection method for determining whether a tool holder is properly mounted on a spindle, comprising: a sensor for measuring the shape of the outer circumferential surface of the flange portion of the tool holder; master data based on normal state shape data measured in advance in a normal mounting state and measurement data which is actual operational shape data measured; and a phase-limited correlation function calculated by performing a discrete Fourier transform on the obtained phase difference signal; the amount of phase shift between the master data and the measurement data from the calculated phase-limited correlation function; the phase alignment of the master data and the measurement data based on the amount of phase shift; and the determination of a mounting abnormality based on the phase-aligned master data and the measurement data.

[0014] Furthermore, in the above-described method for detecting the mounting state of the tool holder, if there is a notch in the flange portion, it is desirable to perform a discrete Fourier transform on the master data with the notch interpolated and the measurement data.

[0015] Furthermore, in the above-described method for detecting the tool holder mounting state, it is desirable that the master data be the normal state shape data obtained by mounting the tool holder in a normal state beforehand and corresponding the measurement data for one rotation to the rotation angle.

[0016] Furthermore, in the above-described method for detecting the mounting state of the tool holder, it is desirable that the master data be created by obtaining multiple data points by measuring the outer circumferential surface of the flange portion with the sensor using the tool holder, and then making statistical corrections based on that data.

[0017] Furthermore, in the above-described method for detecting the mounting state of the tool holder, it is desirable to perform a discrete Fourier transform on the master data and the measurement data after phase alignment, and to determine if there is a mounting abnormality by calculating the amplitude and phase of each single peak component.

[0018] Further, in the tool holder mounting state detection device that determines abnormal mounting indicating whether the tool holder is normally mounted on the spindle, a sensor that measures the shape of the outer peripheral surface of the flange portion of the tool holder, master data based on the normal state shape data measured in a normal mounting state in advance, and measurement data that is the actually measured actual operation shape data are each subjected to discrete Fourier transform to extract only the phase component, and the obtained phase difference signal is subjected to inverse discrete Fourier transform to calculate a phase-limited correlation function by a phase-limited correlation function calculation means, waveform position detection means for obtaining the phase shift amount between the master data and the measurement data from the calculated phase-limited correlation function, waveform position alignment means for aligning the phases of the master data and the measurement data based on the phase shift amount, and mounting abnormality determination means for determining mounting abnormality based on the master data and the measurement data after phase alignment.

[0019] Furthermore, in the above tool holder mounting state detection device, when there is a notch in the flange portion, it is desirable that the phase-limited correlation function calculation means perform the discrete Fourier transform on the master data and the measurement data in which the notch is interpolated.

[0020] In addition, the present invention is a machine tool incorporated with a tool holder mounting state detection device that determines whether the tool holder is properly mounted on the spindle. The present invention includes a sensor that measures the shape of the outer peripheral surface of the flange portion of the tool holder, and a data processing device that detects the mounting state of the tool holder based on the shape data measured by the sensor. The data processing device performs discrete Fourier transform on master data based on normal state shape data measured in a normal mounting state in advance and measurement data that is actual operation shape data actually measured, extracts only the phase components, and calculates a phase-limited correlation function by performing inverse discrete Fourier transform on the obtained phase difference signal. The data processing device further includes a waveform position detection means for obtaining the phase shift amount between the master data and the measurement data from the calculated phase-limited correlation function, a waveform position alignment means for aligning the phases of the master data and the measurement data based on the phase shift amount, and a mounting abnormality determination means for determining mounting abnormality based on the master data and the measurement data after phase alignment.

Effect of the Invention

[0021] According to the present invention, in the tool holder mounting state detection method, discrete Fourier transform is performed on master data based on normal state shape data measured in a normal mounting state in advance and measurement data that is actual operation shape data actually measured, only the phase components are extracted, and a phase-limited correlation function is calculated by performing inverse discrete Fourier transform on the obtained phase difference signal. From the calculated phase-limited correlation function, the phase shift amount between the master data and the measurement data is obtained. Therefore, regardless of the notch of the flange portion of the tool holder and the accuracy of the flange portion, it is possible to easily and highly accurately align the phases of the normal state shape data (master data) and the actual operation shape data (measurement data), and improve the detection accuracy of the mounting abnormality (chuck error) of the tool holder on the spindle.

Brief Description of the Drawings

[0022] [Figure 1] Block Diagram of Tool Holder Mounting State Detection Device According to an Embodiment of the Present Invention [Figure 2] Figure 1 is a plan view showing the relationship between the tool holder and the data processing device. [Figure 3] A graph showing measurement data for one rotation in one embodiment. [Figure 4] Block diagram showing the phase-limited correlation function calculation means in one embodiment. [Figure 5] Graph showing an example of how the phase-limited correlation function is calculated. [Figure 6] A flowchart illustrating a method for detecting the mounting state of a tool holder in one embodiment. [Modes for carrying out the invention]

[0023] Embodiments of the present invention will be described in detail below with reference to the drawings. A machining center is a numerically controlled machine tool that performs various machining operations such as milling, drilling, boring, and tapping without changing the workpiece mounting. The tool magazine stores a large number of cutting tools, and the tools are automatically changed according to computer numerical control commands to perform machining. Therefore, because machining is the primary purpose, the environment in which a machining center is installed contains fine particles such as oil mist and dust, and furthermore, there is debris and chips around the workpiece and spindle.

[0024] Figure 1 is a block diagram showing one embodiment of a tool holder mounting state detection device 50 incorporated into a machine tool according to the present invention, with the tool holder 11 shown as a side view. The tool holder mounting state detection device 50 is a device that automatically detects chuck errors (abnormal mounting conditions) of the tool holder 11 mounted on the spindle head 26 by an ATC device, and mainly consists of a sensor 1 which is a measuring means and a data processing device 3 which is a mounting state determination means.

[0025] The probe 9, which is capable of measuring the dimensions of the workpiece, is configured to be automatically attached to and detached from the spindle head 26 by the ATC device when the control device 22 executes a program. In other words, the probe 9, which is the measuring part, can be integrally mounted to the tool holder 11. That is, the fitting part 11A of the tool holder 11 is pressed against the conical fitted part 26A of the spindle head 26. As a result, the fitting part 11A is tightly fitted (chucking) to the fitted part 26A.

[0026] Workpiece measurements are performed by the control device 22 during the machining process or when the initial machining process is completed, to confirm that the machining has been performed correctly. When the ATC device automatically measures the shape of the workpiece, etc., the control device 22 controls the tool holder 11 to be fitted with a probe 9, which is a measuring instrument for dimensional measurement.

[0027] Sensor 1 is mounted on the spindle head 26 via a bracket 10. This sensor 1 is an eddy current sensor and detects the distance d from the flange portion 11B of the tool holder 11 mounted on the spindle head 26 to the outer surface as an electrical displacement signal. Although an eddy current sensor is used as sensor 1, sensor 1 is not limited to an eddy current sensor; other sensors can be used as long as they can measure the distance d from a specific measurement point to the outer surface of the tool holder 11. In this case, sensor 1 is not limited to a non-contact sensor such as an eddy current sensor; a contact sensor can also be used.

[0028] The data processing device 3 detects the mounting status of the tool holder 11 based on the shape data measured by the sensor 1, and includes an A / D converter 4, a CPU 6, a memory 5, an input / output circuit 7, etc. The A / D converter 4 converts the electrical signal indicating the distance d output from the sensor 1 into a digital signal and outputs it to the CPU 6. Based on the shape data of the sensor 1 converted into this digital signal, the CPU 6 determines whether the tool holder 11 is mounted abnormally and calculates a correction value for the measurement value of the probe 9.

[0029] To determine if there is an abnormality in the mounting and to calculate a correction value, the CPU 6 first mounts the tool holder 11 in a normal mounting state and stores measurement data 32 for one rotation of distance d in memory 5, corresponding to the rotation angle of the tool holder 11 (this data is referred to as normal state shape data). The stored shape data (normal state shape data) also includes shape errors of the outer surface of the flange portion 11B. During actual operation, immediately after mounting the tool holder 11 with the probe 9 attached, the outer surface of the flange portion 11B is similarly measured with the sensor 1 and stored in memory 5, corresponding to the rotation angle (this data is referred to as actual operation shape data).

[0030] The CPU 6 compares the normal state shape data stored in memory 5 with the actual operating shape data in the normal mounting state to determine if there is an abnormality in the mounting state. The calculation processing by the CPU 6 is performed after receiving a command to start measurement from the control device 22 via the input / output circuit 7. This measurement may be averaged after rotating the tool holder 11 once or multiple times to improve accuracy.

[0031] Sensor 1 is an eddy current sensor that detects the distance d to the outer surface of the flange portion 11B as an electrical signal when mounted on the spindle head 26. The shape of the tool holder 11 is standardized and identical not only for the fitting portion 11A but also for the outer surface shape of the flange portion 11B. Therefore, the mounting state of the measuring instrument can be determined by measuring the outer surface shape of the flange portion 11B. The outer surface shape is determined by rotating the tool holder 11 once and measuring the distance d.

[0032] Figure 2 is a plan view showing the relationship between the tool holder 11 and the data processing device 3 in Figure 1. Generally, the flange portion 11B of the tool holder 11 often has two notches 11C for the chuck, as shown in Figure 2.

[0033] Figure 3 is a graph showing measurement data 32 for one rotation, with the horizontal axis representing phase (θ) and the vertical axis representing displacement (d). When there are two notches 11C in the flange portion 11B of the tool holder 11, the measurement data 32 for one rotation, when displayed graphically, will look like the one shown in Figure 3(a). As shown in the figure, the measurement data 32 changes abruptly at the two notches 11C.

[0034] Figure 3(b) shows a graph of the measurement data 32 after the notch 11C portion has been linearly interpolated by the notch correction means 33 (see Figure 6) in the CPU 6. The interpolation of the measurement data 32 may also be calculated based on interpolation methods such as spline interpolation, Lagrangian interpolation, polynomial interpolation, Newton interpolation, Neville interpolation, or continued fraction interpolation.

[0035] (a) of the fitting abnormality determination means 38 (see Figure 6) is performed as described in Patent Document 2 by performing FFT analysis on the interpolated normal state shape data and the actual operating shape data, and calculating the amplitude and phase of each peak component. Furthermore, the amplitude obtained from the normal state shape data is used as the basic eccentricity vector with the basic eccentricity amount and phase in the direction of the basic eccentricity, and the amplitude obtained from the actual operating shape data is used as the measured eccentricity vector with the measured eccentricity amount and phase in the direction of the measured eccentricity. The difference between the measured eccentricity vector and the basic eccentricity vector is calculated as the true eccentricity vector. Furthermore, the magnitude of the true eccentricity vector is used as the true eccentricity amount, and fitting abnormality is determined based on this.

[0036] Alternatively, (b) of the mounting abnormality determination means 38 (see Figure 6) compares the stored normal state shape data with the actual operating shape data, as described in Patent Document 1. The mounting abnormality determination means 38 then determines the centroid position G1 of the normal state shape data and the centroid position G2 of the actual operating shape data, and calculates the distance and angle between G1 and G2 as correction values ​​for correcting the actual operating shape data. Furthermore, if the corrected actual operating shape data is above a threshold value around the entire circumference, the data processing device 3 determines that there is a mounting abnormality. In the case of a mounting abnormality, the input / output circuit 7 instructs the control device 22 to reattach or to check the mounting.

[0037] In both cases of the mounting abnormality determination means 38(a) and (b), the phases of the normal state shape data and the actual operating shape data must be precisely aligned. Furthermore, if the tool holder 11 has a notch 11C, it is necessary to determine which of the two notches 11C is located at 0 degrees. However, measurements by the sensor 1 may be subject to fluctuations and timing discrepancies, and the phases may not always be precisely aligned. For example, if the phase (angle) at which the measurement of the actual operating shape data begins is shifted by 90 to 270 degrees relative to the normal state shape data, it may align with the opposite notch 11C of the two notches 11C.

[0038] Figure 4 is a block diagram showing the phase-only correlation function calculation means 30 in the CPU 6 according to one embodiment. In Figure 4, the normal state shape data is used as master data 31, and the actual operating shape data is used as measurement data 32. The normal state shape data has been previously described as measurement data 32 of the distance d in the normal mounting state, but the master data 31 is not limited to this. For example, the master data 31 is obtained by using a typical tool holder 11 to measure the outer circumferential surface of the flange portion 11B with a sensor 1 and obtaining multiple data points, then statistically processing and correcting them. The master data 31 is then created as the ideal normal state shape data and stored in memory 5. The measurement data 32 in Figure 4 is distinguished from the master data 31 and is the actual operating shape data that has been measured. The phase difference between the master data 31 and the measurement data 32 is determined by the Phase Only Correlation (POC) method.

[0039] The phase-only correlation method is a correlation method that focuses only on the phase component of a Fourier-transformed signal. The phase-only correlation function calculation means 30 specifies the acquisition regions of the master data 31 and the measurement data 32, performs a discrete Fourier transform (DFT) on each to extract only the phase components 31-2 and 32-2, and calculates the phase-only correlation function 35 by performing an inverse discrete Fourier transform (IDFT) on the obtained phase difference signal 34.

[0040] If the flange portion 11B of the tool holder 11 has two notches 11C, the master data 31 uses data with the notches 11C interpolated. Since the measurement data 32 is the actual operational shape data measured, similarly, data with the notches 11C interpolated is obtained. Let the signal waveform of the master data 31 at the N sampled points be f(n), and the signal waveform of the measurement data 32 be g(n). If the discrete Fourier transform of the signal waveform f(n) is F(k) and the discrete Fourier transform of the signal waveform g(n) is G(k), then equations (1) and (2) are obtained.

number

number

[0041]

number

number

[0042] The phase difference signal 34 between the phase component 31-2 of the master data 31 and the phase component 32-2 of the measurement data 32 is calculated as equation (3).

number

number

[0043] The phase - limited correlation function 35 can be obtained as the inverse discrete Fourier transform (IDFT) of R fg (n) as r FG (k) as shown in Equation (4). [Number]

[0044] FIG. 5 is a graph showing an example in which the phase - limited correlation function 35 is calculated. The horizontal axis represents sample points and corresponds to the measurement data 32 for one cycle. The vertical axis represents displacement in FIG. 5(a) and the phase - limited correlation function 35 calculated in FIG. 5(b). Note that FIG. 5 shows an example of calculating the phase - limited correlation function 35, and the illustrated master data 31 and measurement data 32 are not actual data. FIG. 5(a) is a waveform in which the phase of the measurement data 32 is shifted while the master data 31 and the measurement data 32 have the same waveform.

[0045] FIG. 6 is a flowchart showing a method for detecting the tool - holder mounting state. As described in FIG. 3, the notch correction means 33 generates data obtained by interpolating the data invalid periods of the two notches 11C for the data of the displacement of the outer peripheral surface in the flange portion 11B.

[0046] Since the master data 31 and the measurement data 32 are originally based on the data obtained by measuring the outer peripheral surface of the flange portion 11B with the sensor 1, they are originally approximate waveforms. Therefore, when the phase of the measurement data 32 is shifted as shown in FIG. 5(a) with respect to the master data 31 shown in FIG. 5(a), without performing more complex processing, there is only one sharp peak at the location where the correlation is strong as shown in FIG. 5(b). When the phase is shifted as in the measurement data 32 shown in FIG. 5(b) with respect to the master data 31 shown in FIG. 5(a), without performing more complex processing, there is only one sharp peak at the location where the correlation is strong as shown in FIG. 5(b).

[0047] This peak represents a phase shift, which can be detected as the phase shift amount M by the waveform position detection means 36 in the CPU 6 (see Figure 1) with simple processing. In other words, the waveform position detection means 36 can determine the amount of phase shift M between the master data 31 and the measured data 32 simply by detecting the peak position of the phase-limited correlation function 35 obtained by the phase-limited correlation function calculation means 30.

[0048] When master data 31 and measurement data 32 are used as actual data, the calculation of the phase-limited correlation function 35 accurately determines the phase shift amount M even if chips or other debris get caught in the mating part 11A and the mated part 26A (see Figure 1), causing vibration. When vibration occurs in the measurement data 32, the fundamental frequency component (first-order component) obtained by the Fourier transform changes significantly. However, since the calculation of the phase-limited correlation function 35 does not use amplitude information, it is not affected by the amount of vibration, and therefore an accurate phase shift amount M can be detected.

[0049] The waveform alignment means 37 in the CPU 6 can perform phase alignment between the master data 31 and the measured data 32 with high accuracy based on the phase shift amount M detected by the waveform position detection means 36. The fitting abnormality determination means 38 determines whether there is a fitting abnormality based on the phase-aligned master data 31 and the measured data 32. Specifically, as already mentioned, the fitting abnormality determination means 38 (a) performs a discrete Fourier transform (FFT analysis) on the phase-aligned master data 31 and the measured data 32. Then, to determine the fitting state, it calculates the amplitude and phase of each single peak component, finds the true eccentricity vector of the measured data 32, and takes its magnitude as the true eccentricity. Then, to determine whether there is a fitting abnormality, it either (b) compares the phase-aligned master data 31 and the measured data 32 based on the true eccentricity, and makes a fitting abnormality determination based on that. [Explanation of symbols]

[0050] 1...Sensor 3…Data Processing Unit 4…A / D converter 5…Memory 6…CPU 7…Input / Output Circuits 9…Probe 10…Bracket 11…Tool holder 11A...Mating part 11B...Flange section 11C... Notch 22...Control device 26... Main shaft head 26A…Mated part 30...Method for calculating phase-limited correlation function 31…Master Data 31-1...amplitude component, 31-2...phase component 32...Measurement data 32-1...amplitude component, 32-2...phase component 33...Notch correction means 34…Phase difference signal 35…Phase-limited correlation function 36... Waveform position detection means 37... Waveform alignment means 38. Means for determining abnormal fitting 50...Tool holder mounting status detection device

Claims

1. A tool holder mounting status detection method for determining whether a tool holder having a flange portion with multiple notches has been properly mounted on a spindle, and for determining mounting abnormalities, Master data based on normal state shape data obtained by interpolating the data portions corresponding to the multiple notches in first data obtained by pre-measuring the shape of the outer circumferential surface of the flange portion with a sensor in a normal mounting state, and measurement data which is actual operation shape data obtained by interpolating the data portions corresponding to the multiple notches in second data obtained by actually measuring the shape of the outer circumferential surface of the flange portion with the sensor during actual operation, are each subjected to a discrete Fourier transform to extract only the phase component, and the obtained phase difference signal is subjected to an inverse discrete Fourier transform to calculate a phase-limited correlation function which has a sharp peak only at one location where the correlation is strong when the phases of the master data and the measurement data are misaligned. By detecting the peak of the calculated phase-limited correlation function, the amount of phase shift between the master data and the measurement data is determined. A method for detecting the mounting state of a tool holder, characterized by performing phase alignment between the master data and the measurement data based on the phase shift amount, and determining whether there is a mounting abnormality based on the phase-aligned master data and the measurement data.

2. In a tool holder mounting status detection device that determines whether a tool holder having a flange portion with multiple notches is properly mounted on the spindle, A sensor that starts measuring the shape of the outer surface of the flange portion from one of the plurality of notches, A phase-limited correlation function calculation means calculates a phase-limited correlation function having a sharp peak at only one location where the correlation is strong when the phases of the master data and the measurement data are shifted, by performing a discrete Fourier transform on each of the following: master data based on normal state shape data obtained by interpolating the data portion corresponding to the plurality of notches in first data obtained by pre-measurement of the shape of the outer surface of the flange portion in a normal mounting state using the sensor, and measurement data which is actual operation shape data obtained by interpolating the data portion corresponding to the plurality of notches in second data obtained by actual measurement of the shape of the outer surface of the flange portion from one of the plurality of notches using the sensor during actual operation, to extract only the phase component, and by performing an inverse discrete Fourier transform on the obtained phase difference signal, A waveform position detection means that determines the amount of phase shift between the master data and the measurement data from the calculated phase-limited correlation function, A waveform alignment means that performs phase alignment between the master data and the measurement data based on the phase shift amount, A means for determining an abnormal fit, which determines an abnormal fit based on the master data and measurement data after the phase alignment has been performed, A tool holder mounting status detection device characterized by being equipped with [a specific feature].

Images