Vibration Measuring Device
The vibration measurement device for rotary tools transmits high-resolution waveforms with frequency information, addressing the limitations of averaging methods to enhance real-time analysis and detection of tool abnormalities, thereby improving machining accuracy and efficiency.
Patent Information
- Application Number
- JP2021539296
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2020-02-10
- Filing Date
- 2020-08-11
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2040-08-11
AI Technical Summary
Existing vibration measurement devices for rotary tools in machining and friction stir welding lack the ability to perform real-time evaluation of vibration signals with frequency information, relying on averaging methods that obscure detailed analysis and fail to detect critical machining phenomena.
A vibration measurement device with symmetrically mounted acceleration sensors transmits high-resolution waveforms at a high sampling rate, preserving frequency information and enabling advanced analysis techniques like FFT and autocorrelation processing to detect tool abnormalities and machining conditions in real time.
Enables detailed analysis of machining phenomena, detecting tool defects and optimizing machining conditions by preserving frequency information, allowing for real-time detection of issues like chatter and tool breakage, and improving machining accuracy and efficiency.
Smart Images

Figure 0007592934000002 
Figure 0007592934000003 
Figure 0007592934000004
Abstract
Description
[Technical field]
[0001] The present invention relates to a vibration measuring device that can detect in real time the averaged effective value of vibration acceleration detection data of a rotating tool used in a cutting device or the like, and a high-resolution waveform ("a waveform collected at a high sampling rate that does not undergo averaging processing using RMS calculation, and that retains frequency information"). [Background technology]
[0002] In machining equipment such as cutting equipment and friction stir welding equipment, when considering the product accuracy of the workpiece, manufacturing efficiency, and yield of the processed product, it is required to evaluate the condition of the tool during processing (tool evaluation), such as wear, fatigue, breakage, chatter, etc. Conventionally, tool evaluation has been performed based on evaluation criteria that are generalized by equipment manufacturers or tool manufacturers for each device or tool, or academically standardized evaluation criteria. However, real-time verification of the actual tool during processing has not been possible.
[0003] In response to this, the applicant has developed and provided various technologies for measuring the temperature of rotating tools during machining, and for predicting abnormalities based on the results of these measurements (Patent Document 1, Patent Document 2). On the other hand, it is also known that vibration, in addition to temperature, is an important factor in tool damage, etc. However, no specific method has been provided for evaluating this in real time during machining, and the current situation is that it is left to the five senses of skilled workers at each machining site (in addition to visual inspection, touch and hearing, etc.). Therefore, standardized machining conditions for tools and equipment and machining conditions based on the empirical rules of skilled workers have not been fully understood, and it cannot be said that high-speed, high-precision machining has necessarily been objectively achieved.
[0004] In response to this, the present inventors developed a vibration measuring device using a tool holder that detects abnormal vibrations of rotating tools used in rotary machining equipment in real time and transmits the detected vibrations wirelessly to an external PC or the like for analysis, making it possible to detect early signs of tool breakage, improve machining accuracy, and shorten machining time (see Patent Document 3). This tool holder vibration measuring device uses a pair of acceleration sensors mounted symmetrically about the rotation axis on a horizontal plane, and adopts a method in which, due to the limitations of the wireless transmission sampling rate, an averaging process is performed using an RMS (Root Means Square) calculation to evaluate the effective value of the output acceleration as the average vibration acceleration during machining.
[0005] However, the method of outputting the effective value of acceleration by averaging processing using RMS calculation is a good output method in terms of threshold control to avoid the limit of the wireless transmission sampling rate, but since frequency information is lost in the RMS calculation, there is insufficient information when analyzing more detailed machining phenomena. On the other hand, as described above, in the past, real-time evaluation of vibration acceleration in rotating tools was not performed in the first place, and it was only possible to evaluate it using the vibration measuring device using the tool holder provided by the present inventors, so the current situation was that it was not even clear whether there were machining phenomena that could not be analyzed by the effective value of acceleration averaged by RMS calculation. Therefore, there is a possibility that there are machining phenomena that have not been analyzed or evaluated, and it is considered that the potential need for analysis of these machining phenomena will be large or will become large in the future. [Prior art documents] [Patent documents]
[0006] [Patent Document 1] International Publication WO2015-022967 [Patent Document 2] International Publication WO2016-111336 [Patent Document 3] JP 2018-54611 A Summary of the Invention [Problem to be solved by the invention]
[0007] The present invention was created in consideration of the above circumstances, and has an object to provide a vibration measuring device that wirelessly transmits a collected waveform (high-resolution waveform) at a high sampling rate that retains frequency information, without performing averaging processing on the vibration signal from an acceleration sensor within the limits of the wireless transmission sampling rate in the sensor-side device, for the vibration acceleration of a rotating tool used in rotary processing equipment such as cutting equipment and friction stir welding equipment. [Means for solving the problem]
[0008] Specifically, the present invention relates to a vibration measuring device for a rotating tool held by a tool holder that is connected to a spindle and rotates in cooperation with the spindle, a pair of acceleration sensors mounted symmetrically to the rotation axis of a rotary tool on a horizontal plane relative to the rotation axis; an amplifier circuit that matches the impedance of acceleration information from the acceleration sensors and amplifies the voltage; a low-pass filter that removes a predetermined high frequency from an output signal of the amplifier circuit; a subtraction circuit and an addition circuit that output a parallel vibration signal in the XY directions and a vibration signal in the rotation direction from the output signal of the low-pass filter, respectively; a wireless transceiver on the acceleration sensor side (including a wireless microcomputer (sensor side) 22 and a wireless transceiver 23) that transmits the parallel vibration signals in the XY directions and the vibration signal in the rotation direction to the outside; and an external wireless transceiver (including a wireless transceiver 24, a wireless microcomputer (PC side) 25, a personal computer 27, and an audio input / output device 26) that receives the vibration signals transmitted from the wireless transceiver on the sensor side and displays and / or records them in real time; The wireless transceiver on the acceleration sensor side performs A / D conversion on the high-resolution waveform output signal including frequency information from the subtraction circuit and the addition circuit, processes it as transmission data, and transmits it as is to the outside.
[0009] Conventionally, as mentioned above, a method was adopted in which the vibration signal from the acceleration sensor was averaged by RMS calculation due to the limitations of the wireless transmission sampling rate of the sensor side device such as the tool holder, and the effective value of the acceleration was output and transmitted externally, but this vibration measuring device does not perform averaging by RMS calculation, and wirelessly transmits the collected waveform (high-resolution waveform) at a high sampling rate that still contains frequency information. This makes it possible to analyze machining phenomena that could not be analyzed using the effective value of acceleration averaged by RMS calculation in the past.
[0010] In addition, it is preferable that the external wireless transceiver device displays and / or records the high-resolution waveform output signal received from the wireless transceiver on the sensor side in real time, either directly or after fast Fourier transform.
[0011] The wireless transceiver external to this vibration measuring device (particularly the external PC 27) can display the high-resolution waveform, which is a collected waveform sent from the wireless transceiver on the sensor side at a high sampling rate, and the waveform obtained by performing a fast Fourier transform (FFT) on this waveform. This makes it possible to detect characteristic peaks and the like that cannot be detected by the high-resolution waveform alone, but which can be detected by performing an FFT on the frequency information contained in the output signal of the high-resolution waveform, enabling various advanced analyses. For example, it is possible to detect in real time so-called "chatter" that is based on the influence of the bending natural frequency of the rotating spindle, which cannot be detected by displaying the high-resolution waveform.
[0012] In addition, the external wireless transceiver may display and / or record the high-resolution waveform output signal received from the sensor-side wireless transceiver in real time, either directly or after autocorrelation processing.
[0013] In the above example, even if no difference is detected in the high-resolution waveform vibration signal alone, a difference is detected in the FFT converted signal data, and an example is proposed in which an abnormality can be detected. In this vibration measuring device, an example of autocorrelation processing is proposed. Autocorrelation processing is a process of calculating the correlation coefficient between the original signal data and the original signal data by shifting the time, and shows the relationship between the time shifted from the original signal data and the calculated correlation coefficient. For example, it is known that when one blade of a blade with multiple blades is damaged, there is a large difference between the correlation coefficient of the period of the damaged blade and the correlation coefficient of the period of the normal blade.
[0014] Also This vibration measuring device is a filter means for extracting only output signals of a specific frequency range from high-resolution waveform output signals collected at a high sampling rate while still including frequency information, without performing averaging processing on vibration signals output in real time during machining from an acceleration sensor attached to a rotating tool held in a tool holder that is connected to a spindle and rotates in cooperation with the spindle; The signal processing device may further include an averaging means for averaging and outputting the output signal extracted by the filter means.
[0015] As described above, when detecting a high-resolution waveform, which is a waveform collected at a high sampling rate that contains frequency information without performing averaging processing such as normal RMS processing, it is possible to detect output information that cannot be detected by RMS processing, and therefore to analyze the processing phenomenon in more detail. However, even in such a high-resolution waveform, unnecessary output signals such as noise are included in a certain frequency range, which may hinder the analysis of the processing phenomenon. This vibration measurement device has a filter means for removing frequency regions containing unnecessary output signals at the high-resolution waveform stage and extracting only the frequency region required for analyzing processing phenomena, and can output the filtered data through RMS processing. Under certain processing conditions, if a processing phenomenon to be detected, such as the presence of a defect, affects a specific frequency region, this can be detected in real time during processing, and various analyses can be performed according to the user's request.
[0016] Furthermore, it is preferable that the specific frequency range extracted by the filter means is a frequency range that is determined to be necessary from an output signal obtained by performing a fast Fourier transform on the output signal of the high-resolution waveform.
[0017] In the above vibration detection means example, it was found that a fast Fourier transform may be performed to find unnecessary frequency regions in the high-resolution waveform to be removed by the filter means for detailed analysis of the processing phenomenon, and the unnecessary regions may become clear. In this case, it is preferable to remove the unnecessary frequency regions detected by the fast Fourier transform, and then perform averaging processing such as RMS processing. Effect of the Invention
[0018] As described above, according to the vibration measuring device of the present invention, for the vibration acceleration of a rotating tool or the like, the vibration signal from the acceleration sensor is not averaged within the limits of the wireless transmission sampling rate in the device on the sensor side, but a high-resolution waveform signal is wirelessly transmitted as a collected waveform at a high sampling rate that still contains frequency information, and the output signal is displayed on an external computer or the like that receives it, allowing for advanced analysis of the machining state that was not possible in the past to be performed in real time. Furthermore, the output signal of the high-resolution waveform can be processed and calculated in other ways on the external computer or the like and displayed, making it possible to perform more advanced analysis in real time that could not be analyzed or detected even with the high-resolution waveform.
[0019] Furthermore, according to the vibration measuring device of the present invention, in cases where detailed analysis of processing phenomena is not possible simply by detecting the high-resolution waveform, which is the raw collected waveform, without performing averaging processing such as conventional RMS processing, it is possible to detect only the processing phenomena to be detected in real time by deleting frequency regions containing unnecessary output signals at the high-resolution waveform stage and extracting only the frequency regions necessary for analyzing the processing phenomena, and it is possible to perform various analyses according to the user's requests. [Brief description of the drawings]
[0020] [Figure 1](a) shows a photograph of a tool holder unit equipped with this vibration measuring device being gripped on the rotating spindle of a machining center, and (b) shows an example photograph of an external terminal that receives and analyzes data from the tool holder unit in (a). [Diagram 2] FIG. 2 shows a vertical cross-sectional view of a tool holder unit. [Diagram 3] FIG. 2 is a schematic diagram showing six different arrangement positions of the acceleration sensor. [Figure 4] FIG. 1A is a perspective view of a tool holder unit, and FIG. 1B shows the position of the acceleration sensor on a cross-sectional view taken along a line perpendicular to the axis at the position where the acceleration sensor is disposed in FIG. [Diagram 5] (a) shows the relationship between the load and acceleration generated when cutting a workpiece, and (b) shows the acceleration as seen from the tool. [Figure 6] A block diagram showing an example of vibration measurement in which a high-resolution waveform is output without RMS calculation (averaging processing) and transmitted externally is shown. [Figure 7] 7 is a graph showing a first example (embodiment 1) of application of the measurement results of vibration of a tool using the output signal shown in FIG. 6. FIG. [Figure 8] FIG. 1 is a schematic diagram of the setup for Experimental Example 2, and a photograph showing the relationship between the cutting edge of an end mill and a tool rotation coordinate system. [Figure 9] 13 shows the monitoring results of a high-resolution waveform of vibration acceleration in Experimental Example 2. [Figure 10] In Experimental Example 2, focusing on 1850 rpm, which is the rotational speed during stable machining, the results of FFT analysis of both physical quantities are shown. [Figure 11] Focusing on 1900 rpm, which is the rotational speed during unstable machining in Experimental Example 2, the results of FFT analysis of both physical quantities are shown. [Figure 12] FIG. 11 is a photograph showing X-ray transmission images under each condition in Experimental Example 3, showing how internal defects were observed at positions surrounded by dashed lines under all conditions. [Figure 13]FIG. 11 is a graph showing the change over time in vibration acceleration during welding at a tool rotation speed of 500 rpm in Experimental Example 3. [Figure 14] 13 is a diagram showing the relationship between the rotation speed and the amplitude of vibration acceleration during joining feed in Experimental Example 3. FIG. [Figure 15] The figure shows a high-resolution waveform and a waveform after autocorrelation processing when cutting with a throw-away end mill (two blades) as a cutting tool with multiple blades. [Figure 16] FIG. 11 is a schematic diagram of a throw-away drill having two blades, an inner blade and an outer blade, used in Example 5. [Figure 17] This is a detection result of vibration acceleration calculated by normal RMS calculation in the throw-away drill of FIG. 16. [Figure 18] The detection results of vibration acceleration of a high-resolution waveform without RMS processing are shown for the same translational direction (Acc.X) and rotational direction (Acc.Y) as in Figure 17. [Figure 19] The signal waveforms obtained by FFT-processing the high-resolution waveform signal for each time domain in FIG. 18 are shown. [Figure 20] This shows a signal in which only frequencies below 2000 Hz are extracted from the vibration signal of the high-resolution waveform in Figure 18. [Figure 21] This shows the frequency data in FIG. 20 subjected to RMS processing. BEST MODE FOR CARRYING OUT THEINVENTION
[0021] Example of device configuration Figure 1(a) shows a photograph of a tool holder unit 1 equipped with this vibration measuring device being held by the rotating spindle 2 of a machining center. This tool holder unit 1 is held by the rotating spindle 2 at its upper part like a normal tool holder, and holds a tool at its lower part, but unlike normal tool holders, it is formed as a unit with the function of detecting the condition in the vicinity of the tool during machining in real time.
[0022] Specifically, the vibration of the tool during machining is measured, the data is digitized and transmitted to an external device, where it is received and analyzed. Fig. 1(b) is an example photograph of an external terminal that receives and analyzes data from the tool holder unit 1 in Fig. 1(a). The receiver 31 receives the digital data from the tool holder unit 1 and transmits it to a personal computer 32. The personal computer 32 receives the data transmitted from the receiver 31, processes (or calculates) it using dedicated software inside, and displays it on a display.
[0023] <<Tool holder unit configuration and acceleration sensor mounting position>> FIG. 2 shows a vertical cross-sectional view of the tool holder unit 1. In FIG. 2, the upper side of the paper is the rotating spindle 2 side, and the lower side is the cutting tool side. The tool holder unit 1 is inserted and held in the rotating spindle 1 in a nested manner above the flange portion 3a, and rotates together with the rotating spindle 1. The inside is hollow, and a chuck 5 that holds a tool is fixed at the bottom. A gap 3c for arranging parts is provided above the chuck 5, and a battery 4 is disposed in this gap 3c. This battery 4 may be rechargeable. Data from various sensors is A / D converted by a control board 9 disposed in the gap 3c, and transmitted from a wireless transmission device 7 on the outer periphery side of the tool holder body 3 connected via a through hole. The data from the various sensors may be A / D converted on a control board on the outer periphery of the tool holder main body 3.
[0024] As for various sensors, examples of temperature measurement using thermocouples arranged inside the tool have already been developed, but an acceleration sensor is used to detect vibrations in cutting tools and the like, which is the subject of this invention. An example of the arrangement of this acceleration sensor is described below. In the example of Fig. 2, an example is shown in which the acceleration sensor is arranged at the lowest end position of the gap 3c in order to more accurately capture vibrations at the cutting point. Note that, as an example, a configuration in which a large-volume battery is accommodated on the side of the rotating main shaft 2 (upper side of the paper in Fig. 2) rather than the acceleration sensor 6 is given. An example of the arrangement position of the acceleration sensor 6 is described below.
[0025] FIG. 3 is a schematic diagram showing six types of examples of the arrangement of the acceleration sensor 6. FIG. 3(b) shows an example of arranging the acceleration sensor 6 near the rear end of the tool 11 in FIG. 2 described above. In addition, as an example similar to the method of measuring tool vibration, FIG. 3(a) shows the acceleration sensor 6 attached to the tip (lower end) of the cavity of the tool 11, and FIG. 3(c) shows the acceleration sensor 6 attached to the rear end (upper end) of the tool 11. FIG. 3(d) shows a cantilever 11a formed at the rear end (upper end) of the tool 11 for amplifying the vibration amplitude, and the acceleration sensor 6 attached to its tip (upper end). FIG. 3(e) shows the acceleration sensor 6 attached to the tip (upper end) of the chuck 5 of the tool. FIG. 3(f) shows an example of forming a cantilever 3e or a similar part for amplifying the vibration amplitude in the gap 3c in the tool holder body 3, and attaching the acceleration sensor 6 to its tip.
[0026] <Acceleration detection direction> Fig. 4(a) shows a perspective view of the tool holder unit 1, and (b) shows the arrangement of the acceleration sensor 6 on a cross section (position of reference number 6) taken along a line perpendicular to the axis at the arrangement position of the acceleration sensor 6 in (a). In Fig. 4, the X direction is any lateral direction perpendicular to the axial direction of the tool holder unit main body 3 as indicated by the arrow xx, and the Y direction is a direction rotated 90° around the axis from the arrow xx as indicated by the arrow yy. The acceleration sensors 6 shown here are piezoelectric acceleration sensors, and are arranged in pairs at opposing positions centered on the rotation axis O, with two pairs arranged perpendicular to each other for the X and Y directions.
[0027] Specifically, as shown in Fig. 4(b), at least two pairs (four) of acceleration sensors 6 are disposed on the same plane horizontal (perpendicular) to the rotation axis. A pair of acceleration sensors 6a are disposed facing each other in the X direction (arrow xx direction) at the same distance from the axial center O so that they have opposite sensitivities in the Y direction. Similarly, another pair of acceleration sensors 6b2, 6b2 are disposed facing each other in the Y direction (arrow yy direction) at the same distance from the center O so that they have opposite sensitivities in the X direction.
[0028] Here, the individual sensors of the two pairs of acceleration sensors 6 a and 6 b are Y1, Y2, X1, and X2, respectively, as shown in FIG. From the accelerations Ay1, Ay2, Ax1, and Ax2 (see arrows) in the Y and X directions of the sensors 6a (Y1), 6a (Y2), 6b (X1), and 6b (X2), the acceleration Ax in the X direction, the acceleration Ay in the Y direction, the acceleration Am' in the tangential direction, and the angular acceleration Am in the tangential direction of the chuck 5 at the arrangement position of the acceleration sensor 6 can be calculated as follows.
[0029] Ax = (Ax1-Ax2) / 2 Ay = (Ay1-Ay2) / 2 Am' = (Ax1+Ax2) / 2 Am [rev. / s 2 ] = Am' [m / s 2 ] / (Diameter [mm] × 1 0 -3 ×π) [m] Am [rad / s 2 ] = Am' [m / s 2 ] / (Diameter [mm] × 1 0 -3 / 2) [m] Therefore, if the acceleration sensor 6 is arranged symmetrically about the rotation axis O, it is possible to detect acceleration in both the horizontal direction (x, y) and the rotational direction.
[0030] <Relationship between processing and acceleration> In addition, Fig. 5 (a) shows the relationship between the load and acceleration generated when cutting the workpiece (work) 12, and (b) shows a diagram showing the acceleration as seen from the cutter (tool) 11. As shown in Fig. 5 (a), the tip 13 cuts and grinds the workpiece 12 by rotating the cutter 11 in the rotation direction W and moving it to the left of the paper. At this time, at the contact point (cutting point) between the tip 13a and the workpiece 12, a load acts on the rear of the cutter 11 (i.e., in front of the tangent) as shown in Fig. 5 (a), and the tangential component force is indicated as Fm and the normal component force is indicated as Fr. Furthermore, a kinetic friction force acts as MFm in the opposite direction to the component force Fm according to the component force Fr.
[0031] When such a force acts on the tip 13, an acceleration acts on the cutter 11, and ultimately on the tool holder unit 1 (chuck 5). Therefore, by providing the acceleration sensors 6a and 6b as described above in FIG. 4(b) and measuring the acceleration in the X and Y directions, it is possible to know how much acceleration is acting in the X direction, Y direction, and rotation direction. As a result, vibration measurement is also possible.
[0032] Vibration measurement signal flow (example of RMS calculation and output, and example of high-resolution waveform data output without RMS calculation) A block diagram showing how this vibration measuring device measures vibration and outputs high-resolution waveform data is shown in Figure 6. First, in the example of Figure 6, in the tool holder body 3, acceleration sensors 6a (y1), 6a (y2), 6b (x1), and 6a (x2) output analog signal vibration information, and parallel vibration signals and rotational vibration signals are output by amplifier circuits, low-pass filters, and subtraction (or addition) circuits (see reference numerals 15, 16, and 17).
[0033] Specifically, the analog signals of the acceleration sensors 6a(y1) and 6a(y2) are impedance-matched and voltage-amplified in the amplifier circuit of the circuit 17, and gain adjustment is performed according to the subsequent circuit. The output signal from the amplifier circuit is output after cutting high frequencies with a low-pass filter to avoid the influence of the resonance frequency of the acceleration sensors 6a(y1) and 6a(y2). The transmission signal from the low-pass filter is also amplified by an operational amplifier or the like. After that, a subtraction circuit (differential amplifier circuit) outputs a voltage that is a predetermined multiple of the voltage difference from the acceleration sensors 6a(y1) and 6a(y2), that is, a parallel vibration (y-translation) signal in the Y direction. Similarly, the analog signals of the acceleration sensors 6b(x1) and 6a(x2) are output as parallel vibration (x-translation) signals in the X direction through an amplifier circuit, a low-pass filter, and a subtraction circuit in the circuit 15.
[0034] Furthermore, in circuit 16, similar to circuit 15, the analog signals of acceleration sensors 6a(x1), 6a(x2) are output via an amplifier circuit and a low-pass filter, and then an adder circuit outputs a voltage that is the sum of predetermined multiples of the voltages from acceleration sensors 6a(x1), 6a(x2), that is, a vibration signal in the rotational direction (x rotations).
[0035] The x translation signal, x rotation signal, and y translation signal output from circuits 15 to 17 are not subjected to averaging processing such as RMS calculation, but are passed through a bypass filter and an A / D converter (not shown) to output digital signals at a high sampling rate.
[0036] The output digital signal is converted into transmission data by a wireless microcomputer 22 (sensor side) in the tool holder body 3 and transmitted to the outside by a wireless transceiver 23. The input terminal of the wireless microcomputer 22 has two channels, a left ch 22a and a right ch 22b, and the circuits 15, 16, and 17 are connected to the left ch 22a, the right ch 22b, and the right ch 22b via a changeover switch 28, respectively, and receive an x-translation signal, an x-rotation signal, and a y-translation signal, which are vibration signals (signals with a high-resolution waveform) at a high sampling rate. The vibration information received by the wireless microcomputer 22 (the x-translation signal, the x-rotation signal, and the y-translation signal as vibration data) is wirelessly transmitted to the outside by the wireless transceiver 23. In this example, the input terminal of the wireless microcomputer 22 has two channels, and two of the circuits 15, 16, and 17 are connected to the left and right channels 22a and 22b via the changeover switch 28. However, if the input terminal has three channels, it is also possible to connect the circuits 15, 16, and 17 to three channels without using a changeover switch, and have the wireless microcomputer 22 receive the x translation signal, x rotation signal, and y translation signal, which are vibration signals with high-resolution waveforms at a high sampling rate.
[0037] The wirelessly transmitted vibration information (x translation signal, x rotation signal, y translation signal) is received by an external wireless receiver 24, and processed by a personal computer 27 having dedicated software installed via a serial USB converter or the like of a wireless microcomputer 25 (PC side), and displayed on its display. The wireless microcomputer 25 also transmits the x translation signal, x rotation signal, and y translation signal, which are vibration data of a high-resolution waveform, to an audio input / output device (bedphone or speaker) 26, which reproduces the signal. The audio input / output device 26 then transmits audio data to the personal computer 27, which records the audio data.
[0038] Vibration measurement signal flow (example of command to start / stop sending waveform data from external computer to acceleration sensor) 6, the wireless microcomputer 25 on the PC 27 side receives a combination of two high-resolution waveforms from the x-translation signal, the x-rotation signal, and the y-translation signal because the wireless microcomputer 22 on the sensor side is 2ch. The PC 27 generates a switching command signal for the changeover switch 28 to switch this to another combination, and can wirelessly transmit the command signal to the wireless microcomputer 22 on the sensor side via the wireless microcomputer 25 on the PC side and the wireless transceivers 24 and 25. The wireless microcomputer 22 on the sensor side that receives the command signal outputs a digital voltage, drives the changeover switch 28 driven by the digital voltage, and switches between the left ch 22a and the right ch 22b that are connected to the circuits 15, 16, and 17.
[0039] Example 1 (Search for optimal processing conditions) This vibration measuring device can detect abnormal vibrations of the tool 11 during machining, and can also detect the occurrence of unstable cutting (so-called "chatter") during cutting, which is a serious sign of tool breakage. Conventionally, when the occurrence of this "chatter" was recognized by visual inspection, abnormal noise, or abnormal signals from the device, it was dealt with by reducing the cutting depth into the workpiece 12 or the rotation speed. In contrast, by utilizing this vibration measuring device, it is possible to detect areas where "chatter" does not occur even if the cutting depth or rotation speed is increased, rather than simply reducing the cutting depth or rotation speed.
[0040] Fig. 7 is a graph showing a first application example of the measurement results of the vibration of the tool 11 by the vibration measuring device. In Fig. 7, the vertical axis represents the cutting depth (mm) into the workpiece 12, and the horizontal axis represents the rotational speed (rpm) of the tool 11, and an unstable cutting area (white area A) and a stable cutting area (gray area B) are shown. The unstable cutting area (white area A) and the stable cutting area (gray area B) change depending on the conditions of the tool 11, the workpiece 12, etc., and although there are academic examples, they have not been specifically shown for each individual tool 11 that is the subject of actual vibration measurement.
[0041] However, for example, if chatter is actually detected at a rotation speed of 2500 rpm and a cutting depth of 1 mm, as shown by the black circle in Figure 7, it is clear that the tool has rotated 11 times in stable cutting region B (since chatter is actually detected). If the vibration measuring device is used to verify the occurrence of chatter while increasing the rotation speed from this state, as shown by arrow (1), it is possible to search for conditions under which chatter does not occur even at higher rotation speeds. In the example of Figure 7, it can be seen that chatter disappears once at around 2700 rpm, then occurs again, and then disappears again at around 5000 rpm.
[0042] Furthermore, from that point on, by increasing the amount of cut as shown by the arrow (2) and verifying whether or not chatter occurs with this vibration measuring device, it is possible to search for the limit point where chatter does not occur even under machining conditions with a larger amount of cut. In the example of Figure 7, the search will verify that the limit point where chatter does not occur is a cutting amount of about 3 mm (upper white circle). In this way, by exploratory measuring of vibration abnormalities at multiple points during machining with this vibration measuring device, it is possible to detect stable cutting area A under severe conditions that could not be verified in the past, and it is possible to easily achieve rapid machining while avoiding breakage of the tool 11.
[0043] Example 2: Cutting process example Settings and cutting conditions The experimental setup and the relationship between the end mill cutting edge and the tool rotation coordinate system are shown in Fig. 8. The spindle taper of the vertical machining center is BT40, and the end mill is held on the spindle by this holder system. The end mill used is a carbide end mill with φ10 mm, 2 blades, 20 mm blade length, and 30° helix angle. The end mill cutting edge was held so that it was parallel to the Xm axis of the tool rotation coordinate system. The workpiece material used was S50C, and it was fixed on a piezoelectric cutting dynamometer (Kistler 9255B) installed on the table of the machining center. The X, Y, and Z sensitivity directions of the cutting resistance are shown as Fx, Fy, and Fz in the top view overview in Fig. 9. Down cutting was performed with one side of the workpiece as one pass, and the cutting length of the tool center in the experiment was 200 mm in the X direction. The cutting conditions were a fixed feed rate of 0.06 mm / tooth per tooth, and the rotation speed was increased from 1500 rpm to 3000 rpm by 50 rpm for every 5 mm of tool center travel. The cutting depth was 0.4 mm in the radial direction, 10 mm in the axial direction, and the tool overhang was L / D = 3.3. No coolant was used, and the experiment was performed dry.
[0044] A method for monitoring high-resolution waveforms of vibration acceleration The vibration acceleration during machining was measured by arranging four acceleration sensors with sensitivity in one axis direction in the holder at equal intervals at a distance of 5 mm in the radial direction from the center of rotation (86 mm from the gauge line) so that they have sensitivity in the rotation direction on the Xm axis of the rotating coordinate system. The accelerations ax1 and ax2 of each sensor have a positive sensitivity in the direction of the arrow shown in the figure, and the Xm direction can be calculated by calculating the difference of the vibration acceleration with (ax1-ax2) / 2. In addition, the Rm direction, which indicates the rotation direction, can be calculated by adding (ay1+ay2) / 2. Therefore, it is possible to monitor the vibration acceleration in a total of two directions, the translation direction and the rotation direction of the tool. The natural frequency of the sensor used here is 20 kHz. In addition, an amplifier, an A / D converter, a microcontroller, and a wireless transmitter were installed. The microcontroller performs analog calculations to compress the information, then performs A / D conversion, and the results are transmitted wirelessly at a sampling rate of 44.1 kHz, allowing the measurement results to be displayed and recorded in real time by a PC connected to the receiver.
[0045] Experimental results and discussion The feed rate per tooth was kept constant at 0.06 mm / tooth, and the end mill speed was increased from 1500 to 3000 rpm in increments of 50 rpm (approximately 5 mm per increment of feed distance) during one pass (200 mm feed direction machining distance). The cutting resistance waveform and the high-resolution waveform of vibration acceleration monitored by this vibration measuring device are shown in Figures 9(a) and (b). It can be seen that the presence of stable and unstable rotation speeds can be monitored, as both physical quantities follow the regenerative chatter diagram. Here, focusing on 1850 rpm, which is the rotation speed during stable machining, the results of fast Fourier transform analysis (hereinafter also referred to as "FFT analysis") of both physical quantities are shown in Figures 10(a) and (b). The result of the FFT analysis of the cutting resistance waveform in Figure 10(a) showed a peak at around 61.6 Hz. This is equal to the cutting cycle when machining with a two-blade tool at a rotation speed of 1850 rpm, and can be said to be a stable machining state. On the other hand, the results of FFT analysis of the high-resolution waveform of vibration acceleration in Figure 10(b) show that a peak was observed in the Xm direction near 462 Hz, which is thought to be the bending natural frequency of the BT40 machining center spindle.1) As can be seen from Figure 10(a), no vibration of 462 Hz was generated at 1850 rpm, and the machining state was stable. The reason why a frequency peak appeared near the bending natural frequency of the spindle is thought to be due to the free vibration of the tool, which was excited by the load when one blade of the end mill was machining. In addition, slight vibration was also observed near 4 to 6 kHz. Here, the bending natural frequency can be expressed by the following equation when the holder end face is considered to be a beam with one end fixed and the other end free, the cross-sectional shape is round, and the vibration is considered to be in the first mode.
[0046]
number
[0047] where l is the tool overhang, d is the tool diameter, E is the modulus of elasticity, and ρ is the density. If we substitute the tool core thickness of φ6.5 mm, the overhang of 33 mm, the modulus of elasticity of the cemented carbide alloy of 550 GPa, and the density of 14250 kg / m3 into equation (1) and calculate, f1 becomes 5187 Hz. This also coincides with the 4 to 6 kHz that appeared in the FFT results in the Xm direction in Figure 4(b), and it is believed that the vibration in this frequency band is due to the bending natural frequency of the tool.
[0048] Furthermore, a peak was observed around 6 kHz in the Rm direction, but this peak was also confirmed at other rotation speeds, and since it occurred even when the spindle was idling, it is considered to be the effect of the dynamic characteristics in the rotation direction including the spindle servo system or the torsional natural vibration of the tool holder system. Next, focusing on 1900 rpm, which is the rotation speed during unstable machining, the results of FFT analysis of both physical quantities are shown in Figures 11(a) and (b). From the results of FFT of the cutting resistance waveform in Figure 11(a), a cutting cycle of 63.3 Hz and a peak of 428 Hz were observed due to chatter vibration in the bending direction of the spindle. On the other hand, from the results of FFT analysis of the high-resolution waveform of vibration acceleration in Figure 11(b), a peak of 460 Hz was observed in the Xm direction, and although there is a slight difference from the 428 Hz seen in the FFT of the cutting resistance waveform in Figure 11(a), it is considered to be a frequency caused by chatter vibration in the bending direction of the spindle. In addition, a peak was observed around 462 Hz in Fig. 11(b) during stable machining, but there was a large difference in the peak value between stable and unstable machining, and it is clear that it is also possible to monitor the amplitude during chatter vibration. In the Rm direction, the cutting cycle in which the peak appeared at 31.6 Hz, which is one tool rotation cycle, was 63.3 Hz, but it is believed that the occurrence of chatter vibration in the bending direction of the spindle caused the cutting depth of the two end mill blades to become unbalanced, resulting in cutting in which the load was significantly placed on one side. In other words, it was shown that the proposed monitoring method makes it possible to monitor in detail the frequency and amplitude information of the bending and twisting of the tool during end mill machining, and that it is possible to more accurately elucidate machining phenomena during stable and unstable machining.
[0049] Example 3: Friction stir welding example Experimental and study methods The tool shape was a taper shape with a shoulder diameter of 14 mm, a probe with a left-handed groove machined with M4M5 pitch of 0.8 mm, and a probe length of 4.5 mm. The tool material was heat-treated SKD61 HRC53, and the welded material was A6061 with a thickness of 5 mm. The weld distance was 80 mm. The vibration acceleration was detected by placing an acceleration sensor inside the tool holder. As described above, four acceleration sensors with sensitivity in one axis direction were placed at equal intervals on the Xm and Ym axes of the rotating coordinate system in the holder so that they had sensitivity in the rotation direction, and the vibration acceleration in the bending direction of the tool was measured by calculation. The vibration acceleration of Rm, which indicates the rotation direction, can also be calculated by calculation. Furthermore, the integral time was set to 0.1 s, and the RMS (Root Mean Square) was calculated, and the vibration acceleration during welding was detected with a wireless transmission frequency of 50 Hz. A vertical machining center was used as the processing machine. The joining conditions were tool rotation speed 500, 1000, 1500 rpm, tool insertion speed 30 mm / min, holding time 5 sec, joining speed 300 mm / min, tool plunge depth 4.8 mm, and advance angle 0 degrees.
[0050] Experimental and study results and considerations Figure 12 shows X-ray images for each condition. Under all conditions, internal defects were observed at the positions enclosed by the dashed lines. However, while defects were observed throughout the entire joint at a rotation speed of 500 rpm, at 1000 and 1500 rpm, the defects were relatively small in the initial joining portion, and defects of similar size were generated at 1000 and 1500 rpm. The main cause is presumed to be insufficient heat input at a rotation speed of 500 rpm, and at 1000 and 1500 rpm, the shoulder plunge was large, which caused excessive burr removal and prevented material from filling the groove created by the tool passing through.
[0051] Figure 13 shows the change in vibration acceleration over time during welding at a tool rotation speed of 500 rpm. In Figure 13, the probe comes into contact with the workpiece at a, the shoulder comes into contact at b, the workpiece is held in the range of cd, and the welding feed is performed in the range of de.
[0052] Figure 14 shows the relationship between the rotation speed and the amplitude of vibration acceleration during joining feed. In particular, the amplitude of vibration acceleration in the Rm direction was largest at a rotation speed of 500 rpm, and was smaller at 1000 and 1500 rpm than at 500 rpm, and was similar at 1000 and 1500 rpm. These magnitude relationships were similar to the magnitude relationship of the size of internal defects obtained from the observation of X-ray transmission images, suggesting that there is a correlation between the size of internal defects and the amplitude of vibration acceleration.
[0053] Example 4 - Comparison of high-resolution waveform and waveform after autocorrelation processing (autocorrelation coefficient) FIG. 15 shows a high-resolution waveform and a waveform obtained by performing autocorrelation processing when cutting with a throw-away end mill (two blades) as a blade having multiple blades. The top left row (a) of FIG. 15 displays a vibration signal of a high-resolution waveform of the translational velocity (x translation) of a normal throw-away end mill (two blades), the top middle row shows a photograph of one blade and the other blade of the throw-away end mill, and the top right row (c) displays a signal after autocorrelation processing of the vibration signal of the high-resolution waveform of (a). The bottom left row (b) displays a vibration signal of a high-resolution waveform of the translational velocity (x translation) of a normal throw-away end mill (two blades) with one blade broken, the top middle row shows a photograph of one blade (upper side) and the other broken blade (lower side) of the throw-away end mill, and the bottom right row (d) displays a signal after autocorrelation processing of the vibration signal of the high-resolution waveform of (b).
[0054] The autocorrelation processing here involves calculating the correlation coefficient with the original signal data by shifting the time for a throw-away end mill rotating at a rotational speed of 2000 rpm and a period of 30 ms (see the arrow indicating “one period” in (a)) as shown in (a) and (b). The horizontal axis represents the time shifted from the original signal data, and the vertical axis represents the calculated correlation coefficient.
[0055] As shown in Figure 15 (a) and (b), even if the blade of the throw-away end mill is broken, there is no significant difference in the high-resolution waveform itself that has not been subjected to autocorrelation processing. On the other hand, in the signal after autocorrelation processing in (c) and (d), in a normal throw-away end mill, there is no change in the waveform for each blade (waveform (i) and waveform (ii)), but when one of the blades is broken, it can be seen that the peak of the waveform for each blade (= each half cycle) is significantly different (the autocorrelation coefficient of the peak of the waveform (ii) is larger than that of the peak of the waveform (i)). Therefore, by looking at the autocorrelation coefficient of a blade with multiple blades rotating at a certain cycle, the presence or absence of abnormalities such as breakage becomes clear, and it can be seen that it can be used as an indicator for abnormality judgment.
[0056] Example 5 - Example of detecting defects inside material during drilling The tool shape is a throw-away drill with two blades, an inner blade and an outer blade, as shown in Fig. 16. Specifically, a Sandvik 880-D3200L40-03 was used. The material to be joined is S50C with a thickness of 40 mm, and a simulated defect with a depth of 10 mm and a width of 15 mm is provided. The vibration acceleration is detected by placing an acceleration sensor inside the tool holder. As described above, four acceleration sensors with sensitivity in one axis direction are placed at equal intervals on the X-axis and Y-axis of the rotating coordinate system in the holder so that they have sensitivity in the rotation direction, and the vibration acceleration in the bending direction of the tool is measured by calculation. Furthermore, the vibration acceleration of R, which indicates the rotation direction, can also be calculated by calculation. The cutting conditions were V = 60 m / min and f = 0.2 mm. The experiment was conducted using water-soluble cutting oil as the coolant. The processing machine used was a vertical machining center.
[0057] Figure 17 shows the results of detecting vibration acceleration during cutting using a normal RMS (Root Mean Square) calculation with an integral time of 0.1 s and a wireless transmission rate of 50 Hz. Figure 17 shows that there is almost no difference in vibration acceleration in the translational direction (Acc.X) and rotational direction (Acc.Y) between areas with and without simulated defects. This means that defects cannot be detected because the data is RMS processed with an integral time of 0.1 s.
[0058] Next, Figure 18 shows the detection results of vibration acceleration of high-resolution waveforms at a high sampling rate without RMS processing for the same translational direction (Acc.X) and rotational direction (Acc.Y) as Figure 17. In the figure, time domain a shows the state where the drill is idling, time domain b and time domain e show the state where the drill is cutting at a location where there is no artificial defect, and time domain c and time domain d show the state where the drill is cutting at a location where there is an artificial defect. It can be seen that even in the high-resolution waveform signal, which is the raw acceleration waveform, there is almost no difference between the locations where there is an artificial defect and the locations where there is no artificial defect.
[0059] Next, Figure 19 shows signal waveforms obtained by FFT-processing the high-resolution waveform signal for each of the time domains a to e in Figure 18. As indicated by the star, peaks at frequencies below 2000 Hz are detected only in the signal waveforms for time domains c and d, where cutting is performed at the location with the artificial defect, while as indicated by the *, peaks are detected at frequencies near 5000 Hz in all time domains b, c, d, and e, where cutting is performed regardless of whether or not there is an artificial defect. In other words, it is believed that the vibration signal near 5000 Hz is noise, and the effects of the artificial defect are detected in vibration signals below 2000 Hz.
[0060] Therefore, we decided to extract only frequencies below 2000 Hz, which are affected by the presence or absence of simulated defects, from the vibration signal of the high-resolution waveform, which is the raw acceleration waveform in Figure 18, and remove frequencies above 2000 Hz as noise. The resulting waveform is shown in Figure 20 (only the translation direction (Acc.X) is shown). In Figure 20, it can be seen that an acceleration signal that is clearly different from other time domains is detected in the time domain where cutting is being performed at the location where the simulated defect is present. In addition, Figure 21 shows the RMS processing of this frequency data above 2000 Hz with an integration time of 0.1 s. In Figure 21, it can be seen that the acceleration signal is also clearly detected at the location where the simulated defect is present. Therefore, it was verified that FFT processing can be used as a filter function to remove noise, and that by performing RMS processing on the raw acceleration waveform from which noise has been removed, it is possible to detect the presence or absence of simulated defects that would not be detected with conventional RMS processed data. [Explanation of symbols]
[0061] 1 Tool holder unit 2 Rotating Spindle 3 Tool holder body 3 a Flange 3 c void 4 batteries 5 Chuck 6 Acceleration Sensor 7 Radio transmitting devices (antennas) 9 Control Board 11 Tools (Cutter) 11a Cantilever beam 12 Workpiece 13 Tip (cutting edge) 15, 16, 17 Circuit 22 Wireless microcontroller (sensor side) 23,24 Radio transmitter / receiver 25 Wireless microcomputer (PC side) 26 Audio Input / Output Devices 27 Computer 28 Changeover switch 31 Receiver 32 PC
Claims
1. A vibration measuring device for a rotating tool held by a tool holder that is connected to a spindle and rotates in cooperation with the spindle, a pair of acceleration sensors mounted symmetrically to the rotation axis of a rotary tool on a horizontal plane relative to the rotation axis; an amplifier circuit that matches the impedance of acceleration information from the acceleration sensors and amplifies the voltage; a low-pass filter that removes a predetermined high frequency from an output signal of the amplifier circuit; a subtraction circuit and an addition circuit that output parallel vibration signals in the XY directions and a vibration signal in the rotation direction from the output signal of the low-pass filter; a wireless transceiver on the acceleration sensor side that transmits the parallel vibration signals in the XY directions and the vibration signal in the rotation direction to the outside; and an external wireless transceiver that receives the vibration signals transmitted from the wireless transceiver on the sensor side and displays and / or records them in real time, The wireless transceiver on the sensor side A / D converts the high-resolution waveform output signal including frequency information from the subtraction circuit and the addition circuit, processes it as transmission data, and transmits it as is to the outside, in this vibration measuring device.
2. The vibration measuring device according to claim 1 , wherein the external wireless transceiver device displays and / or records in real time the high-resolution waveform output signal received from the sensor-side wireless transceiver device either directly or after fast Fourier transformation.
3. The vibration measuring device according to claim 1 , wherein the external wireless transceiver device displays and / or records the high-resolution waveform output signal received from the sensor-side wireless transceiver device in real time either directly or after autocorrelation processing.
4. A filter means for extracting only output signals of a specific frequency range from high-resolution waveform output signals collected at a high sampling rate that still contains frequency information, without averaging the vibration signals output in real time during machining from an acceleration sensor attached to a rotating tool held in a tool holder that is connected to a spindle and rotates in cooperation with the spindle; 2. The vibration measuring device according to claim 1, further comprising: averaging means for averaging and outputting the output signal extracted by said filter means.
5. 5. The vibration measuring device according to claim 4, wherein the specific frequency range extracted by the filter means is a frequency range determined to be necessary from an output signal obtained by performing a fast Fourier transform on the output signal of the high-resolution waveform.
Citation Information
Patent Citations
Follow-up integrated intelligent monitoring system for cutting vibration
CN106625024A
Rotary cutter on-line monitoring system
CN202411967U
Cutter tip abrasion detecting method for cutting tool
JP1996085047A
Vibration measuring device
JP2018054611A
Temperature measurement method, and temperature measurement device
WO2015022967A1
Cited By
Intelligent detection equipment for automobile sensor processing and use method thereof
CN122283193A
