JP7312009B2 - Anomaly sign detection system - Google Patents

Description

TECHNICAL FIELD The present invention relates to a technology for detecting signs of abnormality in, for example, tools of machine tools, using information processing technology.

2. Description of the Related Art A machine having a rotating mechanism, for example, a machine tool having a rotating tool such as a drill performs processing such as cutting and grinding on a workpiece (in other words, an object, a workpiece, etc.) by rotating the tool. During machining, an abnormality such as damage to a tool or workpiece may occur. A machine tool or an information processing system including a machine tool is preferably capable of detecting anomalies in advance as anomaly signs.

As a prior art example of an anomaly portent detection system, there is a technology for determining an anomaly or an anomaly portent using parameter values such as the voltage of an AE signal using an AE (acoustic emission) sensor. As a general method, there is known a method of judging that the signal is abnormal when the parameter value exceeds a threshold value in response to the phenomenon that the parameter value of the signal increases exponentially in the event of an abnormality.

As a prior art example related to the abnormality sign detection, there is International Publication No. 2009/096551 (Patent Document 1). Patent Document 1 describes that a bearing diagnosis system includes a sensor attached to a bearing, a monitoring diagnostic device connected to the sensor, and the like, and the following. This diagnostic system equally divides the time for one rotation of a rotating shaft supported by bearings or the time for intermittent operation for one rotation in an intermittent operation into a plurality of intervals, compares the abnormality judgment reference level with the measurement data of each interval, and determines the interval. The presence or absence of an abnormality is determined for each

WO2009/096551

In the determination method of the abnormality portent detection system of the prior art example, for example, a threshold value is set for the amplitude of the AE signal, and when the amplitude exceeds the threshold value, it is determined as an abnormality or an abnormality portent. Such a method requires the amplitude of the AE signal to be large, and is difficult to detect, for example, when only a minute signal is generated in the case of low-load machining. Moreover, even if such a method can detect, it is often too late at the time of detection, and it is difficult to prevent abnormalities such as tool damage. SUMMARY OF THE INVENTION An object of the present invention is to provide a technique capable of detecting a sign of abnormality in a machine tool with high sensitivity.

A representative embodiment of the present invention has the following configuration. An abnormality sign detection system according to one embodiment is an abnormality sign detection system for detecting an abnormality sign in a machine tool having a rotating tool, comprising an AE sensor installed on the machine tool or a workpiece, and an AE sensor from the AE sensor. an AE event detection circuit for detecting an AE event from the AE signal; and an AE event rate for calculating an AE event rate, which is the number of AE events per unit time, using the AE event. a rate calculation circuit for determining whether or not a plurality of said index values arranged in time series have reached a first state in which they are consecutive for a first consecutive number of times or more within a first range; and an output control circuit that performs output control including an alert output indicating the anomaly sign or a machining operation stop when the anomaly sign is detected.

According to the representative embodiments of the present invention, an abnormality sign in a machine tool can be detected with high sensitivity.

BRIEF DESCRIPTION OF THE DRAWINGS It is a figure which shows the structure of the abnormality sign detection system of embodiment of this invention. It is a figure which shows the structural example, such as a tool, a workpiece|work, a table, and an AE sensor, in embodiment. 1 is a diagram showing a configuration example in which an AE sensor is installed on a work in an embodiment; FIG. 1 is a diagram showing a configuration example of an AE sensor in an embodiment; FIG. FIG. 4 is a diagram showing a propagation path of an AE signal in an embodiment; FIG. 1 is a diagram showing a configuration example of an AE signal detection circuit in an embodiment; FIG. FIG. 4 is a diagram showing a processing flow of an abnormality sign detection function in an embodiment; FIG. 4 is a diagram showing an example of an AE wave in an embodiment; FIG. FIG. 4 is a diagram showing another example of AE waves in an embodiment; FIG. 4 is a diagram showing a first method of AE event detection in an embodiment; FIG. 4 is a diagram showing a second scheme of AE event detection in an embodiment; FIG. 4 is a diagram showing calculation of an AE event rate in an embodiment; FIG. 10 is a diagram showing an example of AE waves according to changes in the state of the tool in the embodiment; FIG. 4 is a diagram showing a first example of AE waves and AE event rates in an embodiment; FIG. 4 is a diagram showing a second example of AE waves and AE event rates in the embodiment; FIG. 9 is a diagram showing a third example of AE waves and AE event rates in the embodiment; FIG. 4 is a diagram showing a first example of abnormality sign determination in an embodiment; It is a figure which shows the 2nd example of abnormality portent determination in embodiment. FIG. 10 is a diagram showing a third example of abnormality portent determination in the embodiment; FIG. 4 is a diagram showing an example of transition of AE amplitude and AE event rate over a long period of time, etc., in an embodiment; FIG. 4 is a diagram showing examples of AE waves during processing and non-processing in the embodiment; FIG. 4 is a diagram showing an example of a GUI screen in an embodiment; FIG. FIG. 4 is a diagram showing an example of the state of the cutting edge of the tool in the embodiment;

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In principle, the same parts are denoted by the same reference numerals in all the drawings for describing the embodiments, and repeated descriptions are omitted.

[Comparative example]
As a comparative example with respect to the embodiments, the abnormality predictor determination method of the prior art example compares parameter values such as voltage and current of a signal from a sensor such as an AE sensor with preset threshold values to determine an abnormality or an abnormality predictor. judge. For example, when the peak amplitude exceeds a threshold, it is determined to be abnormal. In the case of high-load machining, large values of voltage and current are obtained, so determination can be made from these values. Examples of high-load machining include roughing and semi-finishing. However, in the case of low-load machining, the obtained voltage and current values are small, so it is difficult to make judgments based on these values, and high-sensitivity detection cannot be achieved. In the case of low-load machining, for example, the current of the spindle does not easily reflect the machining state, so it is difficult to make a determination using the current. In the case of low-load machining, signal changes are often small before an abnormality occurs. Moreover, when expensive measuring instruments are used, anomalies or signs of anomalies can sometimes be detected from current and voltage values, but expensive measuring instruments are difficult to introduce into the field. Also in this case, since it is common to determine an abnormality when the parameter value increases and exceeds the threshold value, the tool abnormality cannot actually be detected until immediately before the occurrence of the tool abnormality. Such slow detection leads to anomalies such as tool breakage in most cases. In other words, it is difficult to detect an abnormality sign and prevent tool abnormality before the tool abnormality occurs.

(Embodiment)
An abnormality sign detection system and method according to an embodiment of the present invention will be described with reference to FIGS. 1 to 23. FIG. An abnormality sign detection system according to an embodiment is a system for detecting an abnormality sign for a tool of a machine tool or the like. An AE sensor or the like, which is an element necessary for detection, is installed in the machine tool. A part of the abnormality sign detection function may or may not be installed in the machine tool. The abnormality portent detection method of the embodiment is a method having steps executed in the abnormality portent detection system of the embodiment.

The abnormality sign detection system uses an AE sensor to perform monitoring for abnormality sign detection related to an abnormality in a tool or the like when a work is machined by a machine tool. The AE sensor can detect even minute signals during low-load machining. The abnormality sign detection system detects an abnormality sign with high precision by using the signal from the AE sensor and making a determination according to a predetermined method even when only a minute signal is obtained in the case of low-load machining. The predetermined method is a method of determination using an AE event rate (AE event rate). The anomaly sign detection system determines an anomaly sign from the state of the AE event rate in time series. The anomaly predictor detection system determines an anomaly predictor when the value of the AE event rate continuously appears within a predetermined range. The anomaly sign detection system performs output control such as alarm output and operation stop when an anomaly sign is detected. As a result, it is possible to take measures such as stopping the operation before an abnormality such as breakage occurs in the tool, in other words, before the degree of abnormality increases.

[Abnormality sign detection system]
FIG. 1 shows the configuration of an abnormality sign detection system according to an embodiment. The abnormality sign detection system is mainly composed of the machine tool 1 . The machine tool 1 includes a housing 11, a control device 10, a preamplifier 70, a signal processing circuit 71, an oscilloscope 72, and the like. A user operates the machine tool 1 . The user may be a worker who performs processing work, or a manager who monitors abnormalities.

The control device 10 is a device that performs main control including machining control of the machine tool 1, and is equipped with an abnormality sign detection function. A control device connected separately from the machine tool 1 may be provided, and the abnormality sign detection function may be implemented in the control device. A body portion of the machine tool 1 other than the control device 10 includes a housing 11, a spindle stage 31, a spindle 32, a tool 3, an AE sensor 2 (AE sensor unit 20), a table 6, a vise 5, a liquid feed pipe 7, and the like. . Components such as the spindle stage 31 and the table 6 are connected to the housing 11 . The spindle stage 31 drives the movement of the spindle 32 . The spindle 32 includes a holder that holds the tool 3 and drives the rotation of the tool 3 . The spindle 32 includes a rotating shaft 33 connected with the tool 3 .

The table 6 has a mechanism capable of moving and rotating in directions of three orthogonal axes (X, Y, Z) shown in the figure. A vise 5 is provided on the table 6 . A vise 5 fixes the workpiece 4 . The work 4 is an object or material to be processed such as cutting by the tool 3 .

The liquid feed pipe 7 feeds a cutting liquid 8 as a processing liquid to a portion where the work 4 is processed such as cutting by the tool 3 . In addition, the main shaft 32 may be provided with the liquid feeding pipe 7 .

The AE sensor 2 detects AE waves generated during processing. The AE sensor 2 is installed somewhere on the machine tool 1 . The position is a position where the AE wave from the tool 3 or the like propagates, and is not particularly limited, and a position where the AE wave can be detected as easily as possible may be selected. In installation example 9A in FIG. 1, the AE sensor 2 is installed so as to be fixed to the main shaft 32 . In this example, two AE sensors 2 are fixed to the main shaft 32 . The AE sensor 2 may be installed in the form of an AE sensor unit 20 such as the example shown in FIG. 3 which will be described later.

FIG. 1 also shows another installation example of the AE sensor 2 . Installation example 9B is an example in which the AE sensor 2 is installed on the workpiece 5 . Installation example 9C is an example in which the AE sensor 2 is installed in the vise 5 . Installation example 9D is an example in which the AE sensor 2 is installed on the table 6 . Installation example 9E is an example in which the AE sensor 2 is installed in a hollow position via a support fixture. This position is a position in the vicinity of a machined portion or a portion to which the cutting fluid 8 is supplied. Installation example 9F shows an example in which the AE sensor 2 is installed so as to be embedded inside the table 6 .

A preamplifier 70, a signal processing circuit 71, and an oscilloscope 72 are elements for acquiring an AE signal, and correspond to the configuration example of FIG. 6 described later. AE sensor 2 or AE sensor unit 20 is connected to preamplifier 70 via a signal cable. The preamplifier 70 is connected to a signal processing circuit 71 via a signal cable, and the signal processing circuit 71 is connected to an oscilloscope 72 via a signal cable. The oscilloscope 72 is connected to the connection interface section 103 of the control device 10 via a signal cable. Incidentally, in the subsequent stage from the signal processing circuit 71, data transmission may be performed not only by wired communication using a signal cable, but also by wireless communication.

Each part of the machine tool 1 is connected to a control device 10 via a signal cable or the like, and the drive is controlled by the control device 10 . The control device 10 controls the driving of the table 6, the spindle 32, the tool 3, the liquid feed pipe 7, and the like when machining the workpiece 4. FIG. The control device 10 acquires and inputs an AE signal from the AE sensor 2 during machining, and performs control including detection of anomaly sign relating to the state of the tool 3 and the like.

The control device 10 includes a processor 101, a memory 102, a connection interface section 103, an input device 104, a display device 105, a speaker 106, a lamp 107, etc. These are interconnected via a bus or the like. The processor 101 is composed of a CPU, ROM, RAM, etc., and constitutes a controller of the machine tool 1 . The processor 101 implements an abnormality sign detection function by executing processing according to the control program 111 .

The memory 102 is composed of a non-volatile storage device or the like, and stores various data and information handled by the processor 101 . The memory 102 stores control information 121, setting information 122, and the like. The control information 121 corresponds to NC data and is data and information for NC-controlling the machining of the workpiece 4 . The setting information 122 includes system setting information and user setting information related to the abnormality portent detection function.

The connection interface unit 103 is an interface unit that connects the input device 104 and the like, a communication interface device (not shown) and the like. The input device 104 is a device such as an operation panel, buttons, or a keyboard that receives user input operations. The display device 105 is a device that displays various information to the user on a display screen through a graphical user interface (GUI). The speaker 106 is an audio output device that outputs audio such as guides and alarms. A lamp 107 is a device such as an LED that emits light in response to a guide, an alarm, or the like. The control device 10 may be built in the main body of the machine tool 1 or may be externally connected. The control device 10 may be composed of a computer such as a PC or a server. A form in which another computer, a storage device, a communication device, or the like is connected to the control device 10 is also possible.

[Tables, tools and machining]
FIG. 2 shows a configuration example of the table 6, the vise 5, the workpiece 4, the tool 3, etc. of the machine tool 1, and an installation example 9B of the AE sensor 2. As shown in FIG. Installation example 9B is an example in which an AE sensor unit 20 is installed as the AE sensor 2 on the side surface of the work 4 . A signal cable 220 extends from the AE sensor unit 20 and is connected to the preamplifier 70 . Since noise is likely to be picked up between the AE sensor 2 and the preamplifier 70, it is preferable to use a signal cable 220 with a short cable length and low noise.

The workpiece 4 is, for example, SUS304 stainless steel and has a rectangular parallelepiped shape. For example, the upper surface of the workpiece 4 is the processing surface WS. A tool 3 is brought into contact with the processing surface WS to perform processing such as cutting. An example of the tool 3 is a drill. An example of machining is drilling in which a small-diameter hole is formed in the machining surface WS by cutting with a drill, which is the tool 3 . This drilling is, in particular, multiple hole drilling in which a plurality of holes are continuously formed at a plurality of parallel positions.

A side surface of the work 4 is fixed by a vise 5 . The state of the workpiece 4, such as its position and orientation, can be changed by moving the table 6 along three axes and rotating it. By driving the spindle 32, the state of the tool 3, such as its position and orientation, can be changed. The tool 3 protrudes downward from the spindle 32 in the vertical direction (Z direction). The tip of the tool 3 is a small diameter drill.

As an example of processing, milling using a drill, which is the tool 3, will be treated. During hole machining by milling, the machine tool 1 moves the table 6 in, for example, the X and Y directions so that a predetermined position on the machining surface WS of the workpiece 4 (that is, the hole formation target position) is aligned with the tip of the tool 3. to the bottom of the The machine tool 1 moves the tool 3 downward in the Z direction while rotating. As a result, the machine tool 1 digs a hole while the tool 3 is in contact with the machined surface WS. After forming the hole, the machine tool 1 moves the tool 3 upward in the Z direction. The machine tool 1 moves the spindle stage 31 or the table 6 fixing the workpiece 4 in the X and Y directions for machining the next hole. In the case of horizontal hole machining, the spindle stage 31, the spindle 32, and the tool 3 are rotated 90 degrees from the positions shown in FIGS. , Y and Z directions.

The tool 3 is, for example, a small-diameter drill made of high-speed steel. The tool 3 has a helical groove and blade at its tip as shown. The tool 3 has, for example, a tip side diameter (blade diameter) of 1 mm, a shank diameter of 3 mm, an overall length of 60 mm, a protrusion (length from the holder) of 30 mm, and a groove length of 12 mm.

As an example of processing, processing conditions for hole processing by cutting are as follows. The hole diameter is 1 mm and the hole depth is 1 mm. The rotation speed of the tool 3 is, for example, 4800 rpm, which is 12.5 milliseconds when converted to a rotation cycle (=one rotation time). The feed amount is 0.02 mm/rev. The contact time in machining one hole is 0.6 seconds. The number of holes in multi-hole machining ranges from several hundred to several hundred. Moreover, there is a case where the cutting fluid 8 is supplied using the liquid feed pipe 7 during machining (so-called wet method) and a case where it is not supplied (so-called dry method), and both are applicable. When the material of the tool 3 is high-speed steel, the tool 3 bends when a load is applied, so it is difficult to break. On the other hand, if the tool 3 is made of cemented carbide, it is likely to break when a large load is applied.

[Workpiece and AE sensor]
FIG. 3 shows a configuration example in which the AE sensor unit 20 including the AE sensor 2 is installed on one side of the workpiece 4 corresponding to installation example 9B. In addition, the outline of the installation location in the installation example 9C to the vise 5 is also shown. As for the AE sensor unit 20, the outer cover is illustrated in FIG. This cover is fixed to one side of the workpiece 4 . An AE sensor 2 (FIG. 4, which will be described later) is housed in this cover. A signal cable 220 protrudes from a hole in the cover. This cover is waterproof, oil-proof, and the like. Note that in other embodiments, a waterproof or oil-proof type AE sensor 2 may be applied. In that case, the mechanism such as the cover can be simplified. An acoustic coupler (FIGS. 4 and 5), which will be described later, is provided at a fixed position of the AE sensor 2 on the side surface of the work 4. As shown in FIG. This acoustic coupler is, for example, grease, wax or glue.

[AE sensor (1)]
FIG. 4 shows a cross-sectional structure of the AE sensor 2 installed on the installation surface 401 as a configuration example of the AE sensor 2 . There are various types of known AE sensors, and any of them can be applied. In the embodiment, the following types of AE sensors 2 are applied. This type is broadband and single-ended. The broadband type includes a flat band as frequency characteristics, and has a wide band of, for example, 100 kHz to 1 MHz. The AE sensor 2 includes a piezoelectric element 21, a receiving plate 22, a damper 23, a shield case 25, a lid 26, a connector 27, a signal cable 28, and the like. This type suppresses resonance by providing a damper 23 above the piezoelectric element 21 . The sensitivity of this AE sensor 2 is, for example, 40-55 dB. The sensitivity criterion is 0 dB = 1 V/m/s. The receiving plate 22 , particularly the wave receiving surface thereof, is fixed to the installation surface 401 via an acoustic coupler such as grease 402 . For fixing, various means such as adhesion and screwing can be applied. A signal cable 28 from the piezoelectric element 21 is connected to a connector 27 and connected to an external signal cable 220 through the connector 27 .

The installation surface 401 is the surface of the spindle 32 in the installation example 9A of FIG. 1, and is the surface of the workpiece 4 in the installation example 9B of FIG. AE waves from the installation surface 401 propagate to the receiving plate 22 via the acoustic coupler. It is desirable to use an acoustic coupler such as grease 402 to adhere to the interface between the members so that air does not intervene in the propagation path. The piezoelectric element 21 converts the AE wave propagating to the receiving plate 22 into an electric signal (that is, AE signal) by the piezoelectric effect, and outputs the electric signal from the signal cable 28 . Piezoelectric ceramics such as PZT (lead zirconate titanate) are used for the piezoelectric element 21 so that minute strain can be detected. The broadband AE sensor 2 is suitable for observing phenomena with various characteristics at different frequencies. The AE sensor 2 as shown in FIG. 4 is further accommodated in a cover as shown in FIG. 3 to constitute an AE sensor unit 20 . A cover of the AE sensor unit 20 has a structure for fixing to the installation surface 401 .

[AE sensor (2)]
The AE sensor 2 installed on the spindle 32 in the installation example 9A of FIG. 1 and the AE sensor 2 installed on the work 4 in the installation example 9B of FIG. AE waves are detected as AE signals. The AE wave reflects the state of the force that the tool 3 receives from the workpiece 4 as the tool 3 rotates. AE waves include frequencies from 100 kHz to 1 MHz, for example. As the sampling frequency of the AE signal in the AE sensor 2, a frequency of 2 MHz or higher, for example 4 MHz is used. Control device 10 acquires the AE signal from AE sensor 2 through preamplifier 70 , signal processing circuit 71 , and oscilloscope 72 . The control device 10 acquires various information based on the date, ID, and control information 111 together with the AE signal. IDs (identification information) include the IDs of the machine tool 1, the tool 3, the workpiece 4, the AE sensor 2, and the like.

Note that the machine tool 1 may be provided with not only the AE sensor 2 but also other types of sensors. Examples of other types of sensors are voltmeters, ammeters, force sensors, accelerometers, displacement sensors, gyro sensors, ultrasonic sensors, strain gauges, laser doppler vibrometers, temperature sensors, sound level meters, cameras, MEMS sensors, etc. is mentioned. A sound level meter measures environmental sounds. The AE sensor 2 may be installed so as to be embedded inside the table 6 or the like. The AE sensor 2 is preferably installed at a position where it is difficult to detect environmental sounds as much as possible. The machine tool 1 may be provided with a sound absorbing material for absorbing operating noise and the like.

[AE sensor (3)]
In installation example 9A of FIG. 1, two AE sensors 2 are installed with respect to main shaft 32 . When the AE sensor 2 is installed on the spindle 32, basically even one AE sensor 2 can provide a sufficient effect. Furthermore, when two AE sensors 2 are installed, the following effects are obtained. The two AE sensors 2 have different frequency bands, for example. For example, when the material of the work 4 (especially the sound velocity and density) changes, the frequency of the AE wave also changes. Therefore, by detecting using a plurality of AE sensors 2 with different frequency bands, it is possible to deal with materials having a wide range of acoustic characteristics.

Also, as in installation example 9D, when installing on the table 6, for example, two or more AE sensors 2 having the same frequency band may be installed. In this case, using two or more AE signals, the position of the AE source can be calculated from the arrival time difference of the AE waves. The speed of sound of the AE wave and the position of the AE sensor at that time are known information. Two AE sensors 2 are required for one-dimensional positioning. By installing two or more AE sensors 2 on the same plane of the table 6, the position of the AE generation source can be determined. That is, it is possible to specify in detail the position where the abnormality of the tool 3 or the like occurs. When the position is determined, the AE sensor 2 is preferably installed on the same plane as the upper surface (or the lower surface or inside the table 6) of the table 6, as in installation example 9D, rather than the side surface of the table 6. If three or more identical AE sensors 2 are used, two-dimensional localization is possible.

[AE wave propagation path]
FIG. 5 shows propagation paths of AE waves in various installation examples. (A) shows a propagation path corresponding to installation example 9A. A sound source for generating AE waves is a portion where the tool 3 such as a drill and the workpiece 4 come into contact with each other. AE waves generated during machining propagate from the tool 3 to the spindle 32 . The AE wave propagates from the main shaft 32 to the AE sensor 2 through an acoustic coupler 501 such as grease. (B) shows a propagation path corresponding to installation example 9B. AE waves generated during machining propagate from the tool 3 to the workpiece 4 . AE waves propagate from the workpiece 4 to the AE sensor 2 through the acoustic coupler 501 . (C) shows a propagation path corresponding to installation example 9C. AE waves generated during machining propagate from the tool 3 to the work 4 and from the work 4 to the vise 5 . AE waves propagate from the vice 5 to the AE sensor 2 through the acoustic coupler 501 . (D) shows a propagation path corresponding to installation example 9D. This propagation path includes two types of paths shown. AE waves generated in machining propagate from the tool 3 to the work 4 , from the work 4 to the table 6 , and from the table 6 to the AE sensor 2 through the acoustic coupler 501 . Further, AE waves generated during machining propagate from the tool 3 to the work 4, from the work 4 to the vise 5, from the vise 4 to the table 6, and from the table 6 to the AE sensor 2 through the acoustic coupler 501. do.

(E) shows the propagation path corresponding to installation example 9E. This propagation path includes three types of paths shown. On the first path, AE waves generated during machining propagate from the tool 3 to the AE sensor 2 via the cutting fluid 8 . The cutting fluid 8 acts as an acoustic coupler. On the second path, the AE wave propagates from the workpiece 4 to the AE sensor 2 via the cutting fluid 8 . In the third path, the AE wave propagates from the tool 3 to the workpiece 4 and propagates from the workpiece 4 through the cutting fluid 8 to the AE sensor 2 . The AE wave can be propagated through the cutting fluid 8 which is a liquid. Note that the AE wave is well transmitted at the solid-solid interface. The AE wave is slightly reflected at the solid-liquid interface but is transmitted through the interface. The AE wave is almost totally reflected at the interface between the solid and the gas. Therefore, it is preferable that the propagation path does not include air.

[Example of circuit configuration]
FIG. 6 shows a configuration example of a circuit or the like for detecting an AE signal from the AE sensor 2 and detecting a sign of abnormality. The AE signal from the AE sensor 2 is amplified by the preamplifier 70 and then input to the signal processing circuit 71 (which may be referred to as a first signal processing circuit for explanation). The signal processing circuit 71 is a part that performs the first signal processing of the AE signal. The signal processing circuit 71 may be a circuit connected to the connection interface section 103 within the control device 10 or may be an independent device or the like connected to the outside of the control device 10 . The signal processing circuit 71 has a BPF (band pass filter) 81 , a main amplifier 82 , an envelope detection circuit 83 and an AE event detection circuit 84 .

The BPF 81 passes AE signals in a predetermined frequency band. The BPF 81 removes unnecessary low-frequency signals, which are vibration components. A main amplifier 82 further amplifies the AE signal. In this example, the preamplifier 70 and the main amplifier 82 constitute a two-stage amplification configuration, but a single-stage amplification configuration may be used. In the case of machining where the signal level of the sound source is high, only the preamplifier 70 may be used and the gain of the main amplifier 82 may be 0 dB. Alternatively, a form in which only the preamplifier 70 is used and the main amplifier 82 is not connected may be used. An envelope detection circuit 83 performs envelope detection, which will be described later, from the AE signal. The AE event detection circuit 84 detects an AE event (AE event) from the AE signal by a method described later. Note that the BPF 81 may be provided in the front stage of the preamplifier 70 . Also, the BPF 81 may be provided after the main amplifier 82 . Further, a full-wave rectifier circuit or the like may be further provided after the main amplifier 82 .

An oscilloscope 72 is connected after the AE event detection circuit 84 of the signal processing circuit 71 . The oscilloscope 72 is a waveform measuring instrument and includes an ADC (analog-digital conversion circuit) 85 , a display 86 and a recording section 87 . Note that the oscilloscope 72 and the signal processing circuit 71 may be configured as an integrated device. The ADC 85 converts the AE signal, which is an analog signal, into a digital signal by sampling at a predetermined frequency. The display 86 is a waveform display and displays the state of the AE wave of the AE signal (amplitude, AE event, etc.) on the display screen. The recording unit 87 records the data of the AE signal as a log and enables output.

A signal processing circuit provided in the control device 10 (may be referred to as a second signal processing circuit for explanation) is connected to the rear stage of the ADC 85 of the oscilloscope 72 . This second signal processing circuit may be a circuit connected to the connection interface section 103 or may be realized by software program processing by the processor 101 . The second signal processing circuit includes an AE event rate calculation circuit 91, an abnormality portent determination circuit 92, an output control circuit 93, a threshold setting circuit 94, a threshold setting circuit 95, and the like. Note that the control device 10 includes a machining control function 90 as an existing function. The machining control function 90 is a function of generating control information 121 for machining and controlling the driving of the tool 3, the table 6, etc. based on the control information 121. FIG. Tool information 123 is also included in the control information 121 .

The AE event rate calculation circuit 91 receives the AE signal, which is a digital signal, from the ADC 85 and acquires the tool information 123 of the control information 121 . The tool information 123 includes information such as the number of rotations or rotation period of the tool 3 . The AE event rate calculation circuit 91 calculates the AE event rate as an index value ED from the input information. The AE event rate is the number of AE events per unit time.

The abnormality portent determination circuit 92 receives information of the AE event rate (index value ED) from the AE event rate calculation circuit 91, uses a threshold value to determine an abnormality portent related to the state of the machine tool 1, and outputs determination result information. to output The anomaly predictor determination circuit 92 compares a plurality of index values ED arranged in time series with a threshold to determine the presence or absence of an anomaly predictor. The threshold setting circuit 94 dynamically or statically sets a threshold value for abnormality portent determination to the abnormality portent determination circuit 92 . Thresholds, which are set values, include the number of consecutive times N, a first range H1, and the like, which will be described later.

The threshold setting circuit 95 sets a threshold for AE event detection (for example, a two-level voltage threshold described later) to the AE event detection circuit 84 .

The output control circuit 93 performs predetermined output control according to the determination result information from the abnormality portent determination circuit 92 . The output control includes alert output 93A, operation stop control 93B, graph display 93C, signal waveform display 93D, and the like. The alert output 93A is an alert that informs the user of a sign of abnormality, such as an alert voice output from the speaker 106, alert light emission from the lamp 107, alert information display on the display screen of the display device 105, and the like. The alert output 93A can also be an alert output of a plurality of stages (in other words, a plurality of levels regarding the degree of abnormality), which will be described later. The operation stop control 93B is control for immediately stopping the machining operation. The graph display 93C is a display of a graph of the AE event rate on the display screen of the display device 105 or the like, an abnormality sign determination result, or the like. The signal waveform display 93D is a display such as a graph of the AE wave on the display screen of the display device 105 or the like. From each output as described above, the user can recognize a sign of abnormality before an abnormality such as damage to the tool 3 of the machine tool 1 occurs, and can take countermeasures.

As a modified example of the configuration of FIG. 6, the envelope detection circuit 83 and the AE event detection circuit 84 of the first signal processing circuit 71 may be provided in the second signal processing circuit of the control device 10 . Various modifications such as addition or deletion of circuit portions, separation or merging of circuit portions, etc. are possible.

[AE measurement conditions]
AE measurement conditions in the embodiment are as follows. As for the amplifier gain, the gain of the preamplifier 70 is set to 20 dB, and the gain of the main amplifier 82 is set to 0 dB. The BPF 81 was subjected to high-pass filter processing for the purpose of noise removal of vibration components, and was measured at a sampling frequency of 4 MHz. The cutoff frequency of the HPF (high-pass filter) of the BPF 81 was set to 20 kHz in order to measure signals of 20 kHz or higher, which is a general ultrasonic range. The LPF (low-pass filter) of the BPF 81 was THRU (through; no filtering). Since the AE frequency band for plastic deformation of metal materials is generally 100 kHz to 1 MHz, the HPF cutoff frequency is set to about 100 kHz in order to specifically detect AE signals associated with plastic deformation of metal materials. You may

[Processing flow]
FIG. 7 shows the processing flow of the abnormality portent detection function in the abnormality portent detection system and method of the embodiment. FIG. 7 has steps S1 to S8, which will be described in order below. Step S1 is to set the machine tool 1 and the workpiece 4 and start machining. The setting includes installing the AE sensor 2 in advance at a predetermined position of the machine tool 1 as shown in FIG. 1 and the like. The setting includes setting of initial set values by the threshold setting circuit 94 and the threshold setting circuit 95 . A user inputs a machining start instruction to the machine tool 1 . The control device 10 generates control information 121 and starts processing control and abnormality sign detection control.

In step S2, the control device 10 acquires and inputs the AE signal and related information from the AE sensor 2 through the signal processing circuit 71 and the like. At step S3, the AE event detection circuit 84 of the signal processing circuit 71 of FIG. 6 detects an AE event from the AE signal (FIG. 10 or 11 to be described later). In step S4, the AE event rate calculation circuit 91 of FIG. 6 calculates the AE event rate as the index value ED from the AE signal (including AE event information).

In step S<b>5 , the threshold setting circuit 94 of the control device 10 sets a dynamic threshold, for example, in the abnormality portent determination circuit 92 . This threshold is the number of consecutive times N, the first range H1, and the like. A threshold setting circuit 94 uses information from the AE event rate calculation circuit 91 to determine the dynamic threshold.

In step S6, the abnormality portent determination circuit 92 arranges the index values ED of the AE event rates obtained in step S4 in time series, and uses the threshold value set in step S5 to perform abnormality portent determination for each time point. The abnormality portent determination is a determination as to whether or not a plurality of index values ED are in the first range H1 and have continued for the number of consecutive times N or more (referred to as the first state) in time series. If the determination result in step S6 indicates that there is a sign of abnormality (YES), the process proceeds to step S7, and if there is no sign of abnormality (NO), the process proceeds to step S8.

In step S7, the output control circuit 93 performs output control (alarm output 93A and operation stop control 93B described above) according to the abnormality sign determination result (first state) in step S6.

In step S8, the control device 10 confirms whether or not the machining is finished, for example, whether or not all of the multiple holes have been machined. End the flow.

[Tool rotation speed and machining time]
The control device 10 refers to information such as the number of revolutions of the tool 3 from the tool information 123 of the control information 121 to grasp the rotation cycle (=one rotation time) of the tool 3 . The number of rotations of the tool 3 is, for example, the number of rotations per minute ([rpm]). The number of revolutions is, for example, 4800 rpm (80 Hz in terms of frequency). A rotation period can be calculated from the number of rotations (or rotation speed, etc.). For example, from 4800 rpm, the rotation period is 12.5 milliseconds. The AE event rate calculation circuit 91 sets the rotation cycle of the tool 3 as a unit time TU (FIG. 12 described later), counts the number of AE events for each unit time TU, and obtains an AE event rate.

In the machining example according to the embodiment, abnormality sign detection is performed for a long period of time during machining of multiple holes. The machining time for one hole is, for example, one second. For example, when processing 1000 holes, the processing time takes at least 1000 seconds.

[AE waves and abnormalities]
AE is a phenomenon in which elastic energy accumulated inside a material is emitted as sound waves (called AE waves) when the material deforms or breaks. AE waves are ultrasonic waves. The AE wave is generated with the occurrence of deformation or the like before the material breaks down. A method of non-destructively evaluating material deformation and the like using AE waves is called an AE method. The AE method uses an AE sensor. The AE sensor detects elastic waves emitted starting from microscopic destruction of materials. Since the AE signal is weak, an amplifier is used. The AE method can detect signs of abnormalities such as destruction by observing the behavior of AE waves in real time.

FIG. 8 shows a sudden AE wave as an example of the AE wave. (A) shows a set of sudden AE waves. A part of (A) expanded in time is shown in (B). (B) shows a plurality of sudden AE waves. (C) shows a part of (B) expanded in time. (C) shows two sudden AE waves. One sudden AE wave corresponds to one AE event. Such a sudden AE wave may occur before an abnormality occurs in the tool 3 or work 4 . Such a sudden type AE wave, which will be described later, is generated when, for example, chips generated by cutting cannot be discharged out of the hole and remain between the hole and the drill, or chips adhere to the tip of the tool 3. , is considered to be caused by the influence of the chips.

An abnormal state is a state different from the defined normal state. Examples of abnormalities include damage or breakage in the tool 3, workpiece 4, spindle 32, or the like. In the case of a tool 3 such as a small-diameter drill, breakage may occur as an abnormality. Other anomalies include cracks, edge chipping, and the like. Abnormalities that are targets of abnormality sign detection are a general term including such damage, breakage, and the like of the tool 3 . In the case of breakage of the tool 3, the damage to the work 4 is also large, and the processing up to that point is wasted. In addition, in the case of breakage, the load on the machine tool 1 is large, which causes machine failure. In particular, in the case of finish machining, which is low-load machining, it is often the final process, and repair is difficult in the subsequent processes. The user wants to prevent breakage of the tool 3 as much as possible.

The abnormality portent detection system of the embodiment detects such a sudden AE wave as an AE event, calculates an AE event rate from a plurality of AE events, and determines an abnormality portent from the state of the AE event rate. In particular, the abnormality portent detection system detects, as an abnormality portent, a phenomenon in which the index value ED of the AE event rate continuously occurs within a certain range. This makes it possible to take action before the tool 3 breaks.

FIG. 9 shows variations of sudden AE waves. (A) shows a single burst AE wave with simple decay. The AE event detection circuit 84 counts such a sudden AE wave as one AE event. (B) shows a sudden-type AE wave with a complex shape that appears to have multiple peaks at its rising edge. The AE event detection circuit 84 also counts such a sudden AE wave as one AE event. (C) shows a complex shaped sudden AE wave that appears to have multiple peaks in attenuation. The AE event detection circuit 84 also counts such a sudden AE wave as one AE event.

As in the above example, there can be various shapes of AE waves. It is difficult for the determination method in the conventional technique to detect an abnormality sign with high precision for such various shapes of AE waves. On the other hand, since the anomaly portent detection system of the embodiment employs a judgment method for judging the phenomenon of the AE event rate, it is possible to detect an anomaly portent with high accuracy even for such various shapes of AE waves.

[AE event detection]
10 and 11 show examples of AE event detection schemes. AE event detection circuit 84 detects an AE event using the scheme of FIG. 10 or FIG. Which method to use can be set in advance. The AE event detection method is not limited to this example, and other methods may be applied.

FIG. 10 shows a scheme for determining and detecting an AE event using a 1-level voltage threshold V1 as the threshold. The vertical axis of the AE wave of the AE signal in FIG. 10 is the voltage value. (a) shows the peak amplitude of the voltage value of the AE wave. The AE event detection circuit 84 compares the amplitude of the AE wave with the voltage threshold V1, and measures the time when the amplitude becomes equal to or greater than the voltage threshold V1 as an AE count as shown in (c). (b) is the AE event duration associated with the AE count in (c). Time t1 is the start time of the AE event, and time t2 is the end time of the AE event. (d) shows the rise time of the AE event, which is the time from time t1 to the time of peak amplitude in (a). The AE event detection circuit 84 counts and detects durations such as (b) as one AE event based on the AE count of (c).

FIG. 11 shows a scheme for determining and detecting an AE event using two levels of high and low voltage thresholds {voltage thresholds V _H , V _L }. (A) shows an AE wave of an AE signal, which is an analog signal. The vertical axis indicates the voltage amplitude. (B) shows a waveform (envelope waveform) after performing full-wave rectification and envelope detection by the envelope detection circuit 83 of FIG. 6 on the AE signal of (A). The AE event detection circuit 84 compares this waveform with two-level voltage thresholds V _H and V _L to determine and detect AE events. An envelope is a line connecting points of peak voltage amplitude. The waveform (B) shows the waveform after full-wave rectification from the waveform (A). Full-wave rectification is rectification that converts a negative voltage portion into a positive voltage.

The AE event detection circuit 84 detects, as one AE event, the period from when the voltage value of the waveform (B) exceeds the high-side voltage threshold _VH to when it falls below the low-side voltage threshold _VL . (C) shows an AE event count pulse, in other words, an AE event detection signal based on the waveform of (B), and the portion detected as one AE event is in the ON state of the pulse.

[AE event rate calculation]
FIG. 12 shows the calculation of the AE event rate. FIG. 12 shows an example of an AE signal with the horizontal axis representing time. The AE event rate calculation circuit 91 counts the number of AE events per unit time TU on the time series of the AE signal, and sets the AE event rate, which is the number of AE events per unit time, as an index value ED. FIG. 12 describes the number of AE events per unit time TU.

An example of the AE signal in (A) shows a portion of machining of a certain hole. An example of the AE signal in (B) shows part of machining of another hole using the same tool 3 . In the example of (A), sudden AE waves are sporadically generated. In the unit times TU1 to TU8, the index value ED, which is the AE event rate, is {1, 1, 1, 1, 0, 0, 1, 1} in order.

At the time of (B), the tool 3 has already experienced machining of a larger number of holes than at the time of (A), and deterioration is progressing. Therefore, in the example of (B), a plurality of sudden side AE waves are continuously generated. In the unit times TU1 to TU8, the index value ED, which is the AE event rate, is {7, 5, 6, 5, 5, 5, 6, 5} in order.

Note that the AE event rate calculation circuit 91 appropriately counts the AE waves arranged over the unit time TU section in any unit time TU.

[AE waves according to changes in the state of tools, etc.]
FIG. 13 shows an example of transition of AE waves according to changes in the state of the tool 3 and the like. (A) shows AE waves when the tool 3 is in a normal state, and no sudden AE waves (corresponding AE events) exceeding the threshold are detected. (B) is a case within the normal range, in which several sudden-type AE waves appear sporadically. (C) corresponds to the case of an anomaly sign, in which a plurality of sudden AE waves appear continuously. In other words, a plurality of sudden AE waves are observed as continuous AE waves. That is, in this state, multiple AE events appear in succession. The AE event rate can be calculated from such continuous AE waves. If the AE event rate continues within a predetermined range, it is determined as an anomaly sign. (D) shows a continuous AE wave corresponding to the case of an abnormality immediately before the occurrence of the tool abnormality. (E) shows the case of abnormality when tool abnormality occurs. The anomaly sign detection system of the embodiment detects an anomaly sign at a time like (C).

[Change in AE event rate]
FIG. 14 shows an example of the time series transition of the amplitude of the AE signal and the corresponding AE event rate. FIG. 14 shows the transition in the time (assumed to be machining time TP) for machining one hole, where (A) shows the amplitude of the AE signal and (B) shows the index value ED which is the AE event rate. The example of FIG. 14 shows a case corresponding to within the normal range or an early sign of abnormality. In (A), the unit of amplitude is [mV]. The processing time TP is, for example, 1 second. In (B), one point indicates the AE event rate when the unit time TU is one rotation time of the tool. In the first normal period, the index value ED is zero. In the normal case, 0 continues as the index value ED, and even if a value other than 0 appears, it returns to 0 without appearing continuously. After that, for example, in period 141, 1 and 2 appear as the index value ED. If a value other than 0 appears continuously as the index value ED, there is a possibility of an abnormality.

FIG. 15 shows another example of the machining time TP, similarly showing a case corresponding to an abnormality sign. In the example of FIG. 15, the index value ED is larger than in the case of FIG. For example, in period 151, index values ED=2, 3, and 4 appear.

FIG. 16 shows another example of the machining time TP, similarly showing a case corresponding to an abnormality sign. The example of FIG. 16 shows the case immediately before an abnormality such as damage to the tool 3 occurs. In the example of FIG. 16, the index value ED is larger than in the case of FIG. For example, in period 161, index values ED=5, 6, and 7 appear. At time point 162, an abnormality such as damage to the tool 3 has occurred. In the period after time 162, the index value ED has decreased to a smaller value than before.

As in the examples of FIGS. 14, 15, and 16, a phenomenon in which multiple AE event rates saturate continuously within a certain range can be observed on the time series, and this phenomenon can be associated with an anomaly sign. can. In this phenomenon, a plurality of AE event rates are distributed within a certain range with some variation.

[Abnormal sign determination example (1)]
FIG. 17 shows an example of abnormality sign determination using the AE event rate. FIG. 17 shows an enlargement of a portion corresponding to period 141 of FIG. As an example of threshold setting, a setting example of the number of consecutive times N and the first range H1 is shown. For example, the consecutive number N is five. The first range H1 has an integer greater than or equal to 0. For the first range H1, the reference value D0=1 and the variation range is ±1. The reference value D0 is the reference for the first range H1. The variation range indicates the permissible variation of the index value ED. However, in this example, the first range H1 has a condition of being an integer of 1 or more without including 0. That is, the first range H1 is set as a range {1, 2} in which the upper limit value DH is 2 and the lower limit value DL is 1.

When the abnormality sign determination circuit 92 of the control device 10 enters the first state in which the plurality of index values ED are values within the first range H1 in time series and are consecutive for the number of consecutive times N or more, Judged as an anomaly sign. In the illustrated example, period 171 and period 172 are detected as an anomaly sign because the index value ED is 1 or 2 five times in succession.

As a modification regarding the above determination, the following may be performed. Conditions that allow 0 as the index value ED in the first range H1 (in other words, in the values that continue multiple times within the first range H1, some values are 1 or more, and some values include 0 condition that it is good). For example, the first range H1 is set as a range {0, 1, 2} in which the reference value D0 is 1, the variation range is ±1, and the upper limit value DH is 2 and the lower limit value DL is 0 under the condition that 0 is allowed. be done. In this case, for example, in the period 173, the value within the first range H1 continues five times, which can be detected as an anomaly sign.

FIG. 18 shows another determination example with another AE signal. FIG. 18 shows an enlargement of a portion corresponding to period 151 of FIG. As an example of threshold setting, the number of consecutive times N is set to 10. The first range H1 is set to a reference value D0=2, a variation range of ±2, and an integer of 1 or more. That is, the first range H1 is set as a range {1, 2, 3, 4} with an upper limit value DH=4 and a lower limit value DL=1. With this setting, the abnormality portent detection circuit 92 similarly performs abnormality portent determination. For example, period 181 is detected as an anomaly sign because values within the first range H1 appear ten times in succession.

The reference value D0=2 is set dynamically, for example. For example, it is assumed that the reference value D0=1 is initially set. Next, when the index value ED=2 appears to some extent, the reference value D0 is updated as 2 accordingly. After that, the reference value D0 is updated in accordance with the value of the index value ED that appears frequently to some extent.

By using the above determination, the abnormality sign detection system can detect an abnormality sign before an abnormality occurs in the tool 3 or the like, and take measures such as stopping the operation.

FIG. 19 shows another determination example with another AE signal. FIG. 19 shows an enlargement of a portion corresponding to period 161 of FIG. As an example of threshold setting, the number of consecutive times N is set to 5. The first range H1 is set to a reference value D0=6, a variation range of ±1, and an integer of 1 or more. That is, the first range H1 is set as a range {5, 6, 7} with an upper limit value DH=7 and a lower limit value DL=5. The reference value D0=6 is dynamically set. With this setting, the abnormality portent detection circuit 92 similarly performs abnormality portent determination. For example, period 191 is detected as an anomaly sign because the value within the first range H1 appears five times in succession.

As another threshold setting example, a condition may be added that the number of consecutive times N must be consecutive with the same value within the first range H1. For example, with the same value, the number of consecutive times N=5 or more. In this setting, for example, in period 192, the index value ED=6 continues five times, which is detected as an anomaly sign. A similar determination is possible even in the case of a setting in which no variation range is provided.

[Abnormal sign determination example (2)]
FIG. 20 shows an example of the transition of the amplitude of the AE wave and the AE event rate over a long period of time, and an example of two-step abnormality sign determination. (A) shows the transition of the amplitude of the AE signal. (B) shows changes in the AE event rate corresponding to (A) in time series. Since the time axis is long, the AE event rates of a plurality of unit times TU in (B) are temporally compressed and illustrated. For example, at time 201, damage or breakage occurs as a tool abnormality. In the case of the prior art example in which the sign of abnormality is determined using the amplitude as shown in (A), the amplitude increases during the period 202 immediately before the tool abnormality at the time 201, so detection is possible. However, even if an anomaly sign can be detected at that time, it is too late because it cannot be dealt with and the degree of destruction is large. In the prior art example, since the amplitude is small in the period before the period 202, the sign of abnormality cannot be detected.

On the other hand, in the case of determination using the AE event rate in the abnormality portent detection system of the embodiment as in (B), the abnormality portent can be detected long before the tool abnormality point 201 . For example, by performing the determination with the first threshold setting, the first stage of abnormality sign detection can be performed at any point in the period 203, and the first stage of output control can be realized accordingly. The first threshold setting is, for example, the number of consecutive times N=10 and the first range H1 is {1, 2, 3}. The output control of the first stage is, for example, the first alarm output. The first alarm is an alarm that informs that the life of the tool 3 is near.

Further, for example, by performing the second-stage determination with the second threshold setting, the second-stage abnormality sign detection can be performed at any point in the period 204, and the second-stage output control can be performed accordingly. realizable. The second threshold setting is, for example, the number of consecutive times N=10 and the first range H1 is {1, 2, 3, 4, 5}. The second-stage output control is, for example, a second alarm output or operation stop control. The second alarm is an alarm indicating that the degree of abnormality is greater than that of the first alarm. Similarly, three or more stages of determination can also be performed. The period 204 can be, for example, several seconds to several tens of seconds before the time point 201 of the tool failure, and there is sufficient time. Time period 203 can be an even earlier time.

[AE waveform during processing and non-processing]
FIG. 21 shows examples of AE waveforms during processing and non-processing in time series. The machining time TP is the machining time for one hole, and the non-machining time TN is the non-machining time between the machining times TP. Since the tool 3 is not in contact with the workpiece 4 during the non-machining time TN, no sudden AE wave or the like is generated. The control device 10 can grasp the machining time TP and the non-machining time TN from such a waveform of the AE signal, and can use them for machining control. Based on this understanding, the control device 10 may, for example, execute the abnormality sign determination only during the machining time TP.

[Unit time setting]
The reason for setting the unit time TU in the AE event rate as one rotation time (corresponding rotation period) of the tool 3 will be supplemented. As shown in FIG. 12 and the like, the control device 10 sets the unit time TU as one rotation time, and calculates the AE event rate for each unit time TU on the time series. As a result, as a phenomenon suitable for abnormality portent detection, the phenomenon of continuation or saturation of the AE event rate, such as the examples of FIGS. 14 to 19, appears.

One of the reasons why the sudden AE wave (corresponding AE event) shown in FIG. There is a case where machining is performed while biting the chips between. Chip clogging occurs in the grooves and blades at the tip of the tool 3 . In such a case, a destructive phenomenon occurs to the tool 3 and the like. It is considered that a sudden AE wave (corresponding AE event) is generated based on this destruction phenomenon. Based on the above consideration, the attachment and detachment of chips can be estimated from the transition of the appearance density of AE events per rotation time of the tool 3 . Moreover, it is considered that there is an upper limit to the chips that can adhere to the tool 3 . Therefore, it is considered that there is an upper limit to the frequency of the destruction phenomenon (corresponding AE event rate) that occurs during one rotation time due to adhesion of chips. From this, it is considered that phenomena of continuation and saturation related to the AE event rate as shown in FIG. 20 and the like appear.

FIG. 23 shows an example of a tool state corresponding to the occurrence of a sudden AE wave. A sudden type AE wave is generated when chips generated by cutting adhere to the cutting edge of a drill, which is a tool, and the chips accumulate in the groove. The state of FIG. 23 is the state of the cutting edge of the drill observed when the system according to the embodiment detects an anomaly sign and stops the operation at a certain threshold setting. This state is an abnormal state in which chips stick to the cutting edge. When the machining was continued in this state, it was confirmed that a sudden type AE wave was generated. In such a state, chips tend to clog the tool, and if machining continues, the tool tends to break or chip.

Depending on how the unit time TU is set, the phenomenon of continuation or saturation of the AE event rate as described above does not always occur, and in such cases, suitable anomaly sign detection cannot be performed. That is, it is necessary to select and set a suitable time for the unit time TU. For example, if the unit time TU is set to be sufficiently smaller than the one-rotation time of the tool 3, the values of the AE event rate will not continue. Moreover, when the unit time TU is set sufficiently longer than the one-rotation time of the tool 3 (for example, 1 second), the above phenomenon does not occur and the phenomenon increases rapidly. In this case, it becomes similar to the determination method of the prior art example, and suitable detection cannot be performed.

Also, when dividing the time by the unit time TU, if the appropriate unit time TU is not set, the machining time TP and the non-machining time TN (especially the time during which the tool 3 is not in contact with the workpiece 4) as shown in FIG. It divides time without distinguishing between . Therefore, for example, when the unit time TU is too long, the ratio of the machining time TP within the divided time period becomes short, deteriorating the accuracy of abnormality sign detection.

[Threshold setting]
The threshold value for AE event detection and the threshold value for abnormality sign determination may be either static set values or dynamic set values. The number of consecutive times N and the first range H1 may be set according to the type, material, size, and number of rotations of the tool 3, the type, material, and size of the workpiece 4, the type of machining, the load, and the like. As an example, the number of consecutive times N can be set to 3 times, 5 times, 10 times, 15 times, or the like, but it is possible without being limited to this.

Methods for dynamically setting a threshold for abnormality sign determination using an AE event rate are as follows. The control device 10 measures past data (for example, several unit time TUs) from the input AE signal, and once determines and sets the threshold from the measured data. For example, the threshold setting circuit 94 determines and sets the first range H1 and the number N of consecutive times. The abnormality portent determination circuit 92 uses the threshold value to compare it with the latest index value ED to determine an abnormality portent. After a certain amount of time has passed, the control device 10 similarly updates the threshold from past data.

Other methods include the following. The threshold setting circuit 94 of the control device 10 counts the number of appearances for each value of the index value ED that appears from the AE signal. In the example of FIG. 18, the index value ED=2 appears for the first time and continues three times. The threshold setting circuit 94 counts the number of consecutive times for the index value ED=2. The threshold setting circuit 94 updates the reference value D0 of the first range H1 to 2 when the index value ED=2 appears a certain number of times. After that, the index value ED=3 appears for the first time and continues three times. Similarly, the threshold setting circuit 94 counts the number of consecutive times for the index value ED=3, and updates the reference value D0 to 3 when the index value ED=3 appears a certain number of times. The threshold setting circuit 94 may set, for example, the value of the index value ED having the largest number of appearances as the reference value D0. In this way, the first range H1 can be dynamically changed. Further, the threshold setting circuit 94 may dynamically set the first range H1 according to the variation range of appearance of each value of the index value ED.

The abnormality portent detection system of the embodiment can perform abnormality portent detection without depending on a specific threshold value setting when a determination method with dynamic threshold setting is used. In the case of this determination method, for example, it is possible to suitably detect the machining conditions of the workpiece 4 made of an unknown material.

[GUI screen]
The control device 10 displays information such as the AE wave of the AE signal, the AE event, the AE event rate, and the threshold for determination through the GUI on the display screen of the display device 10 or the oscilloscope 72, etc., and allows the user to confirm and set the information. enable Information including graphs such as those shown in FIGS. 12, 14 to 16, 17 to 19, and 20 can be displayed on the screen.

FIG. 22 shows an example of a GUI screen. In this screen example, it is possible to confirm and set the number of consecutive times N and the first range H1 as a threshold value for abnormality sign determination based on the user's operation. Different settings for each tool 3 are also possible. As a modification, the user may set the unit time TU.

[Effects, etc.]
As described above, according to the abnormality sign detection system of the embodiment, an abnormality sign of the tool 3 or the like in the machine tool 1 can be detected with high sensitivity. According to the embodiment, it is possible to realize highly versatile abnormality sign detection that can cope with various processing conditions. According to the method of determining an abnormality sign using the AE event rate in the embodiment, compared to the method of determining using the voltage, current, duration, cutting force, acceleration, etc. of the AE wave in the example of the prior art, Even in the case of minute signals, it is possible to detect signs of anomalies with high accuracy. According to the embodiment, it is possible to detect and take measures before the tool 3 is broken or damaged. Even if the worker who performs the processing work is not a skilled worker, he or she can check the result of the abnormality sign detection and perform the work efficiently while preventing breakage or damage. According to the embodiment, since each individual tool 3 can be used until just before the end of its life, it is possible to improve machining efficiency and reduce costs. According to the embodiment, it is possible to detect even when the degree of abnormality is small, before the tool 3 breaks. Therefore, it is possible to prevent the broken part of the tool 3 from being buried in the work 4 and make it impossible to repair, and the possibility that the work 4 can be repaired at the stage of light damage is increased.

In the prior art example, when setting the threshold for determination using amplitude, it is necessary to set an appropriate threshold that differs according to the material, shape, sensor position, etc. of the tool or workpiece. Difficult to set. On the other hand, according to the embodiment, it is easy to set the threshold for determination, and it is applicable to processing conditions such as unknown materials.

Further, conventionally, for example, in the case of machining using a workpiece of an unknown material, the machining is performed on a trial basis under machining conditions that are not optimal. According to the embodiment, even in such a case, the abnormality sign detection function can be used to cope with the situation. That is, according to the embodiment, it is possible to start machining under the initial non-optimal machining conditions, detect an anomaly sign at an early stage, stop the machining operation with an alert, and then allow the user to adjust the machining conditions to more appropriate machining. Processing can be performed by modifying the conditions.

The abnormality sign detection function and method according to the embodiment are not limited to the examples of the tool 3 and machining conditions described above, but can be similarly applied to various machine tools 1, tools 3, workpieces 4, and machining conditions. As processing, it can be applied to reaming, grinding, turning, milling, and the like. The tool 3 can also be applied to a turning insert or the like. As another example of the tool 3, a high-speed steel drill having a diameter ranging from 0.5 mm to 6.0 mm was used for machining. In addition, processing was performed by dry and wet methods. In these cases, it was confirmed that it was possible to detect an anomaly sign based on similar AE event rate continuation and saturation phenomena.

In addition, the abnormality sign detection system of the embodiment can grasp the characteristics of abnormality to some extent from the results of observing and recording the AE event rate etc. when machining is performed under the conditions of a certain tool 3 or workpiece 4. be able to. For example, it can be seen that the appearance of AE events is not periodic, and that the AE event rate in a certain range of 1 or more occurs continuously. Also, it can be seen that the AE event rate does not rise sharply, and that the AE event rate has an upper limit (in other words, saturation). The abnormality portent detection system may set or update a threshold value for abnormality portent determination, etc. for processing of similar conditions based on the understanding of such characteristics. As an application, the abnormality portent detection system of the embodiment may perform abnormality portent determination using machine learning with data of the index value ED of the AE event rate as input. Machine learning updates a model including thresholds for anomaly sign determination.

The following is also possible as a modified example of the abnormality sign detection system of the embodiment. The method of determining an abnormality sign using the AE event rate may be as follows. In a modified example, the control device 10 does not use the above-described number of consecutive times N to determine the continuation, but uses the reference value D0 or the first range H1 as the threshold for determination. When the index value ED of the AE signal reaches the reference value D0 or falls within the first range H1, the controller 10 detects it as an anomaly sign. Although the present invention has been specifically described above based on the embodiments, the present invention is not limited to the above-described embodiments, and can be variously modified without departing from the scope of the invention.

DESCRIPTION OF SYMBOLS 1... Machine tool, 2... AE sensor, 3... Tool, 4... Work, 5... Vise, 6... Table, 7... Liquid feed pipe, 8... Cutting liquid, 10... Control device, 32... Spindle.

Claims (5)

An anomaly sign detection system for detecting an anomaly sign in a machine tool having a rotating tool,
an AE sensor installed on the machine tool or workpiece;
a signal processing circuit that acquires an AE signal from the AE sensor;
an AE event detection circuit that detects a sudden AE wave from the AE signal as an AE event;
an AE event rate calculation circuit that calculates an AE event rate, which is the number of AE events per unit time, using the AE events;
It is determined whether or not the plurality of AE event rates arranged in chronological order has entered a first state in which the AE event rates are consecutive for a first consecutive number of times or more within a first range, and if the first state has occurred, a determination circuit that detects the abnormality sign;
an output control circuit that performs output control including an alert output indicating the anomaly sign or a machining operation stop when the anomaly sign is detected;
with
The first range has an integer of 0 or more and is a range in which a predetermined variation is allowed with respect to the reference value,
The first range defined using the reference value and the predetermined variation, and the first number of consecutive times, are the type, material, and dimensions of the tool, the type, material, and dimensions of the work, the type of processing, and the It is a static setting value set by the user on the screen based on the phenomenon of the distribution of the AE event rate that appears in the result of actual machining according to the load.
Abnormal sign detection system. An anomaly sign detection system for detecting an anomaly sign in a machine tool having a rotating tool,
an AE sensor installed on the machine tool or workpiece;
a signal processing circuit that acquires an AE signal from the AE sensor;
an AE event detection circuit that detects a sudden AE wave from the AE signal as an AE event;
an AE event rate calculation circuit that calculates an AE event rate, which is the number of AE events per unit time, using the AE events;
It is determined whether or not the plurality of AE event rates arranged in chronological order has entered a first state in which the AE event rates are consecutive for a first consecutive number of times or more within a first range, and if the first state has occurred, a determination circuit that detects the abnormality sign;
an output control circuit that performs output control including an alert output indicating the anomaly sign or a machining operation stop when the anomaly sign is detected;
a threshold setting circuit that dynamically sets the first range based on the AE signal ;
The first range and the first number of consecutive times are determined by the AE event that appears in the actual machining result according to the type, material, and dimensions of the tool, the type, material, and dimensions of the workpiece, the type of machining, and the load. a setting value that is set based on a rate distribution phenomenon, in particular, the first range is a dynamic setting value;
The threshold setting circuit counts the number of appearances and the number of consecutive occurrences for each value of the AE event rate that appears based on the AE event rate data of the past time, and based on the count, the 1 update range,
Abnormal sign detection system. An anomaly sign detection system for detecting an anomaly sign in a machine tool having a rotating tool,
an AE sensor installed on the machine tool or workpiece;
a signal processing circuit that acquires an AE signal from the AE sensor;
an AE event detection circuit that detects a sudden AE wave from the AE signal as an AE event;
an AE event rate calculation circuit that calculates an AE event rate, which is the number of AE events per unit time, using the AE events;
It is determined whether or not the plurality of AE event rates arranged in chronological order has entered a first state in which the AE event rates are consecutive for a first consecutive number of times or more within a first range, and if the first state has occurred, a determination circuit that detects the abnormality sign;
an output control circuit that performs output control including an alert output indicating the anomaly sign or a machining operation stop when the anomaly sign is detected;
with
The AE event rate calculation circuit grasps the rotation period of the tool based on the control information of the machine tool, sets the rotation period to the unit time,
The first range and the first number of consecutive times are determined by the AE event that appears in the actual machining result according to the type, material, and dimensions of the tool, the type, material, and dimensions of the workpiece, the type of machining, and the load. A setting value that is set based on the phenomenon of rate distribution,
Abnormal sign detection system. An anomaly sign detection system for detecting an anomaly sign in a machine tool having a rotating tool,
an AE sensor installed on the machine tool or workpiece;
a signal processing circuit that acquires an AE signal from the AE sensor;
an AE event detection circuit that detects a sudden AE wave from the AE signal as an AE event;
an AE event rate calculation circuit that calculates an AE event rate, which is the number of AE events per unit time, using the AE events;
It is determined whether or not the plurality of AE event rates arranged in chronological order has entered a first state in which the AE event rates are consecutive for a first consecutive number of times or more within a first range, and if the first state has occurred, a determination circuit that detects the abnormality sign;
an output control circuit that performs output control including an alert output indicating the anomaly sign or a machining operation stop when the anomaly sign is detected;
with
The AE event detection circuit uses a first voltage threshold and a second voltage threshold, which are two-level voltage thresholds, to perform envelope detection on the AE signal. detecting a period from exceeding the voltage threshold to attenuating below the second voltage threshold as one AE event;
The first range and the first number of consecutive times are determined by the AE event that appears in the actual machining result according to the type, material, and dimensions of the tool, the type, material, and dimensions of the workpiece, the type of machining, and the load. A setting value that is set based on the phenomenon of rate distribution,
Abnormal sign detection system. An anomaly sign detection system for detecting an anomaly sign in a machine tool having a rotating tool,
an AE sensor installed on the machine tool or workpiece;
a signal processing circuit that acquires an AE signal from the AE sensor;
an AE event detection circuit that detects a sudden AE wave from the AE signal as an AE event;
an AE event rate calculation circuit that calculates an AE event rate, which is the number of AE events per unit time, using the AE events;
It is determined whether or not the plurality of AE event rates arranged in chronological order has entered a first state in which the AE event rates are consecutive for a first consecutive number of times or more within a first range, and if the first state has occurred, a determination circuit that detects the abnormality sign;
an output control circuit that performs output control including an alert output indicating the anomaly sign or a machining operation stop when the anomaly sign is detected;
with
The determination circuit uses a plurality of thresholds in multiple stages including a first threshold in a first stage and a second threshold in a second stage as thresholds including the first range and the first number of consecutive times, and the AE determining for the first state of event rate;
The output control circuit performs different output controls according to the stage at which the abnormality sign is determined,
The first range and the first number of consecutive times are determined by the AE event that appears in the actual machining result according to the type, material, and dimensions of the tool, the type, material, and dimensions of the workpiece, the type of machining, and the load. A setting value that is set based on the phenomenon of rate distribution,
Abnormal sign detection system.

Images