JP7417051B2 - Lathes and lathe systems - Google Patents

Description

The present invention relates to a lathe including a cooling device for cooling a spindle motor, and a lathe system.

2. Description of the Related Art NC (numerically controlled) lathes are known in which a main spindle is rotated at high speed by a built-in motor. If the temperature of the spindle increases too much, it will affect the machining accuracy of the workpiece, so a cooling device is used to suppress the temperature rise of the spindle. For example, the cooling device uses a pump to circulate cooling oil through a circulation path that passes near the built-in motor, and radiates heat from a radiator provided in the circulation path.

In the cooling device for a machine tool disclosed in Patent Document 1, a first temperature sensor is installed on a cooling oil discharge side pipe where cooling oil comes out from a cooling tank that performs heat exchange of the cooling oil, and A second temperature sensor is installed on the cooling oil return pipe. If the difference in temperature between both temperature sensors is 0.5° C. or less, the cooling device determines that cooling is insufficient, issues an alarm, and stops the machine tool.

Utility Model Publication No. 4-122438

The above-described cooling device requires the installation of two or more temperature sensors, which leads to an increase in the cost of the machine.
Incidentally, the above-mentioned problems exist in various lathes.

The present invention discloses a lathe and a lathe system that are capable of determining a failure of the cooling system without separately installing a sensor for detecting failure of the cooling system, such as a sensor that detects the temperature of cooling oil. be.

The lathe of the present invention is
a rotatable main shaft,
a main shaft motor having a built-in temperature sensor and rotating the main shaft;
a cooling device that cools the main shaft motor;
When the change in the detected temperature sequentially obtained by the built-in temperature sensor after starting continuous machining of the workpiece becomes below a predetermined threshold value , it is determined that the temperature has become steady; A failure determination unit that determines that the cooling device is malfunctioning is provided.
Further, the lathe of the present invention has
a rotatable main shaft,
a main shaft motor having a built-in temperature sensor and rotating the main shaft;
A cooling device that cools the main shaft motor, the cooling device having a cooling fluid circulation path that cools the main shaft motor, a pump that circulates the cooling fluid in the circulation path, and a fan that promotes heat dissipation in the circulation path;
A failure in which it is determined that the cooling device is malfunctioning when a change in the detected temperature exceeds an allowable range after it is determined that the detected temperatures sequentially obtained by the built-in temperature sensor have become steady after starting continuous processing of the workpiece. A determination section;
The failure determination unit determines that the pump is malfunctioning when a change in the detected temperature that exceeds the allowable range exceeds a predetermined discrimination criterion, and when the change in the detected temperature does not exceed the discrimination criterion. The present invention has an aspect in which it is determined that the fan is out of order.
Furthermore, the lathe of the present invention
a rotatable main shaft,
a main shaft motor having a built-in temperature sensor and rotating the main shaft;
a cooling device that cools the main shaft motor;
A failure in which it is determined that the cooling device is malfunctioning when a change in the detected temperature exceeds an allowable range after it is determined that the detected temperatures sequentially obtained by the built-in temperature sensor have become steady after starting continuous processing of the workpiece. A determination section;
a second temperature sensor that detects temperature influenced by outside temperature;
represents the detected temperature sequentially obtained by the built-in temperature sensor, the second detected temperature sequentially obtained by the second temperature sensor, and a malfunctioning section among the plurality of sections included in the cooling device. Through machine learning based on failure part information, a failure part in the cooling device is determined based on the detected temperature sequentially obtained by the built-in temperature sensor and the second detected temperature sequentially obtained by the second temperature sensor. and a machine learning unit that generates a learning model for determining.

Further, the lathe system of the present invention is a lathe system including a lathe and a computer connected to the lathe,
The lathe is
a rotatable main shaft,
a main shaft motor having a built-in temperature sensor and rotating the main shaft;
A cooling device that cools the main shaft motor,
The lathe system determines that the state has become steady when changes in the detected temperature sequentially obtained by the built-in temperature sensor after starting continuous machining of the workpiece become below a predetermined threshold , and then determines that the change in the detected temperature is permissible . The present invention has an aspect including a failure determination unit that determines that the cooling device is malfunctioning when the range is exceeded.
Furthermore, the lathe system of the present invention is a lathe system including a lathe and a computer connected to the lathe,
The lathe is
a rotatable main shaft,
a main shaft motor having a built-in temperature sensor and rotating the main shaft;
a cooling device that cools the main shaft motor;
A second temperature sensor that detects the temperature affected by the outside temperature,
The lathe system includes:
A failure in which it is determined that the cooling device is malfunctioning when a change in the detected temperature exceeds an allowable range after it is determined that the detected temperatures sequentially obtained by the built-in temperature sensor have become steady after starting continuous processing of the workpiece. A determination section;
represents the detected temperature sequentially obtained by the built-in temperature sensor, the second detected temperature sequentially obtained by the second temperature sensor, and a malfunctioning section among the plurality of sections included in the cooling device. Through machine learning based on failure part information, a failure part in the cooling device is determined based on the detected temperature sequentially obtained by the built-in temperature sensor and the second detected temperature sequentially obtained by the second temperature sensor. and a machine learning unit that generates a learning model for determining.

According to the present invention, it is possible to provide a lathe and a lathe system that can determine failure of the cooling device without separately installing a sensor for detecting failure of the cooling device.

FIG. 1 is a diagram schematically showing a configuration example of a lathe system. FIG. 2 is a diagram schematically showing a configuration example of a cooling device of a lathe together with main parts of a headstock. FIG. 2 is a block diagram schematically showing a configuration example of an electric circuit of a lathe. FIG. 4A is a diagram showing an example of the temperature change of the built-in temperature sensor when the pump fails during continuous workpiece machining, and FIG. 4B is a diagram showing an example of the temperature change of the built-in temperature sensor when the fan fails during continuous workpiece machining. It is a figure showing an example. FIG. 3 is a diagram schematically showing an example of determining the state of a lathe from discrete data of detected temperatures. FIG. 6 is a diagram schematically showing an example of a temperature change ΔT1(t) of a built-in temperature sensor depending on time. 2 is a flowchart illustrating an example of cooling device failure determination processing. FIG. 2 is a diagram schematically showing an example of a lathe including a spindle motor for driving a rotary tool and a cooling device for the spindle motor. 1 is a diagram schematically showing an example of a lathe system including a machine learning section. 3 is a flowchart illustrating an example of learning processing. 7 is a flowchart illustrating an example of failure part determination processing. FIG. 2 is a diagram schematically showing an example of a lathe including a machine learning section.

Embodiments of the present invention will be described below. Of course, the following embodiments are merely illustrative of the present invention, and not all of the features shown in the embodiments are essential to the solution of the invention.

(1) Overview of technology included in the present invention:
First, an overview of the technology included in the present invention will be explained with reference to examples shown in FIGS. 1 to 12. Note that the figures in this application are diagrams schematically showing examples, and the magnification in each direction shown in these figures may be different, and the figures may not be consistent. Of course, each element of the present technology is not limited to the specific examples indicated by the symbols.

[Aspect 1]
As illustrated in FIG. 1 etc., a lathe 1 according to an embodiment of the present technology includes a rotatable main shaft 11, a main shaft motor M1 that rotates the main shaft 11, a cooling device 40 that cools the main shaft motor M1, and A determination unit U1 is provided. The main shaft motor M1 has a built-in temperature sensor S1. As illustrated in FIGS. 5 and 7, the failure determination unit U1 determines that the detected temperature T1(t) sequentially obtained by the built-in temperature sensor S1 after starting continuous processing of the workpiece W1 has become steady. Later, when a change in the detected temperature T1(t) (for example, ΔT1(t)) exceeds a permissible range (for example, threshold TH2), it is determined that the cooling device 40 is out of order.

Spindle motors sometimes have a built-in temperature sensor to prevent seizure. This temperature sensor can be used as the above-mentioned built-in temperature sensor S1.
When the cooling device 40 is operating normally, the temperature of the spindle motor M1 starts to change when continuous machining of the workpiece W1 is started, and after a certain period of time, the temperature of the spindle motor M1 becomes steady. If the cooling device 40 fails during continuous machining of the work W1, the temperature of the spindle motor M1 changes again and exceeds the allowable range (TH2). Therefore, after the detected temperature T1(t) sequentially obtained by the built-in temperature sensor S1 after starting continuous machining of the workpiece W1 becomes steady, the change in the detected temperature T1(t) (ΔT1(t)) is within the allowable range ( TH2), it can be determined that the cooling device 40 is out of order. Therefore, the above aspect can provide a lathe that can determine failure of the cooling device without separately installing a sensor for detecting failure of the cooling device, and can suppress an increase in costs. By determining that the cooling device has failed, the lathe can, for example, stop its operation or output a warning.

Here, the spindle includes a spindle that acts directly or indirectly on the workpiece, such as a spindle that grips the workpiece or a tool spindle that rotates together with a rotary tool.
The main shaft motor may be a motor that is cooled by a cooling device, and may be a built-in motor or a motor externally attached to the main shaft.
The cooling system includes an oil-cooled cooling system that discharges heat from the main shaft motor by circulating cooling oil as a cooling medium, a water-cooled cooling system that discharges heat from the main shaft motor by supplying cooling water as a cooling medium, and a cooling medium. This includes air-cooled cooling devices, which remove heat from the main shaft motor by means of air flow.
A change in the detected temperature exceeding the allowable range is not limited to a change due to an increase in the detected temperature, but may also be a change due to a decrease in the detected temperature.
Note that the above-mentioned additional remarks are also applied to the following aspects.

[Aspect 2]
As illustrated in FIGS. 5 to 7, the failure determination unit U1 determines whether a plurality of parts (for example, the pump 43 and fan 44) that is malfunctioning. If the degree of contribution to cooling differs depending on the location, the temperature change of the main shaft motor will vary depending on the location. Therefore, this aspect can provide a lathe that can determine the failure part of the cooling device without separately installing a sensor for detecting failure of each part of the cooling device, and can suppress cost increase. can. By determining the failure part of the cooling device, the lathe can, for example, notify the failure part. This improves operator convenience and facilitates repair of the cooling device.

[Aspect 3]
As illustrated in FIG. 2, the cooling device 40 may have a circulation path 41 for a cooling fluid F1 that cools the main shaft motor M1. The plurality of parts may include a pump 43 that circulates the cooling fluid F1 in the circulation path 41, and a fan 44 that promotes heat radiation in the circulation path 41. As illustrated in FIGS. 5 to 7, the failure determination unit U1 determines that when a change (ΔT1(t)) in the detected temperature T1(t) exceeding the allowable range (TH2) exceeds a predetermined determination criterion, It is determined that the pump 43 is malfunctioning (for example, when ΔT1max>TH3), and when the change in the detected temperature T1(t) (ΔT1(t)) does not exceed the discrimination criterion, the fan 44 is It may be determined that there is a failure.

The test revealed that the change in detected temperature T1(t) when the pump 43 fails (ΔT1(t)) is the same as the change in detected temperature T1(t) when the fan 44 fails (ΔT1(t)). It turned out to be bigger than that. Therefore, if the change (ΔT1(t)) in the detected temperature T1(t) exceeding the allowable range (TH2) exceeds a predetermined criterion, it can be determined that the pump 43 is out of order, and the detected temperature T1 If the change in (t) (ΔT1(t)) does not exceed the determination criterion, it can be determined that the fan 44 is out of order. Therefore, the above embodiment can provide a lathe that can determine whether the pump or the fan is out of order without separately installing a sensor for detecting a failure in the pump or fan.
Here, the cooling fluid includes liquids such as cooling oil and cooling water, gases such as air, and the like. This additional statement also applies to the following aspects.

[Aspect 4]
As illustrated in FIGS. 1, 12, etc., the present lathe 1 may further include a second temperature sensor S2 that detects the temperature influenced by the outside air temperature. Further, the present lathe 1 may further include a machine learning section U2. The machine learning unit U2 detects the detected temperature T1(t) sequentially obtained by the built-in temperature sensor S1, the second detected temperature T2(t) sequentially obtained by the second temperature sensor S2, and the detected temperature T2(t) sequentially obtained by the second temperature sensor S2. The detected temperature T1(t) sequentially obtained by the built-in temperature sensor S1 and the first A learning model LM for determining a malfunctioning part of the cooling device 40 is generated based on the second detected temperatures T2(t) sequentially obtained by the two temperature sensors S2. By using this learning model LM, the detected temperature T1(t) sequentially obtained by the built-in temperature sensor S1 provided in the main shaft motor M1, and the second detected temperature T2(t) sequentially obtained as a temperature influenced by the outside temperature. t), it is possible to determine which part of the cooling device 40 is malfunctioning. Therefore, this aspect can provide a lathe that can determine the failure part of the cooling device without separately installing a sensor for detecting failure of each part of the cooling device, and can suppress cost increase. can.
Here, a value obtained from the detected temperature T1(t) such as a change in the detected temperature T1(t) (for example ΔT1(t)) may be used for machine learning, a change in the second detected temperature T2(t), etc. Using the value obtained from the second detected temperature T2(t) for machine learning is also included in the machine learning of the above aspect. Further, the value input to a known neural network or the like used in the learning model is not limited to the detected temperature itself or the second detected temperature itself. In the learning model of the above aspect, there is a case where a value obtained from the detected temperature T1(t), such as a change in the detected temperature T1(t) (for example, ΔT1(t)), is input into a neural network, etc., a second detected temperature T2 This also includes a case where a value obtained from the second detected temperature T2(t), such as a change in temperature T2(t), is input into a neural network or the like. These additional remarks also apply to the following aspects.

[Aspect 5]
The learning model is configured such that the detected temperature T1(t) sequentially obtained by the built-in temperature sensor S1 and the second detected temperature T2(t) sequentially obtained by the second temperature sensor S2 are input, The determination result of a malfunctioning part (for example, the pump 43 and the fan 44) among the plurality of parts included in the cooling device 40 may be output. According to this aspect, it is possible to provide a lathe that can determine the failure part of the cooling device without separately installing a sensor for detecting failure in each part of the cooling device, and it is possible to suppress an increase in cost.

[Aspect 6]
By the way, the lathe system SY1 according to one aspect of the present technology includes a lathe 1 and a computer 100 connected to the lathe 1. The lathe 1 includes a rotatable main shaft 11, a main shaft motor M1 that rotates the main shaft 11, and a cooling device 40 that cools the main shaft motor M1. The main shaft motor M1 has a built-in temperature sensor S1. This lathe system SY1 includes a failure determination section U1 in at least one of the lathe 1 and the computer 100. The failure determination unit U1 detects a change in the detected temperature T1(t) after determining that the detected temperature T1(t) sequentially obtained by the built-in temperature sensor S1 after starting continuous machining of the workpiece W1 has become steady. If ΔT1(t)) exceeds the allowable range (TH2), it is determined that the cooling device 40 is out of order. Therefore, this aspect can provide a lathe system that can determine failure of the cooling device without separately installing a sensor for detecting failure of the cooling device, and can suppress cost increase.

[Aspect 7]
As illustrated in FIG. 1 and the like, the lathe 1 may further include a second temperature sensor S2 that detects temperature influenced by outside air temperature. As illustrated in FIG. 9, the lathe system SY1 may further include a machine learning section U2. As illustrated in FIG. 10, the machine learning unit U2 uses the detected temperature T1(t) sequentially obtained by the built-in temperature sensor S1 and the second detected temperature sequentially obtained by the second temperature sensor S2. T2(t), and the detected temperature T1 sequentially obtained by the built-in temperature sensor S1 by machine learning based on failure part information IN1 indicating a failed part among the plurality of parts included in the cooling device 40. (t) and the second detected temperature T2(t) sequentially obtained by the second temperature sensor S2, a learning model LM for determining a malfunctioning part in the cooling device 40 is generated. Therefore, this aspect can provide a lathe system that can determine the failure part of the cooling device without separately installing a sensor for detecting failure of each part of the cooling device, and can suppress cost increase. I can do it.

(2) Specific example of lathe system configuration:
FIG. 1 schematically illustrates the configuration of a lathe system SY1 including a lathe 1 and a computer 100. The lathe 1 shown in FIG. 1 is an NC lathe equipped with an NC (numerical control) device 70 that numerically controls machining of a workpiece W1. It is equipped with an embedded temperature sensor 3, etc. The control axes of the lathe 1 shown in FIG. 1 include an X axis indicated by "X", a Y axis indicated by "Y", and a Z axis indicated by "Z". The Z-axis direction is a horizontal direction along the spindle center line AX1, which is the rotation center of the workpiece W1. The X-axis direction is a horizontal direction orthogonal to the Z-axis. The Y-axis direction is a vertical direction orthogonal to the Z-axis. Note that the Z-axis and the They do not need to be orthogonal. Further, the drawings referred to in this specification merely show examples for explaining the present technology, and do not limit the present technology. Moreover, the description of the positional relationship of each part is merely an example. Therefore, reversing the left and right sides, reversing the rotation direction, etc. is also included in the present technology. Furthermore, the sameness in direction, position, etc. is not limited to exact coincidence, but includes deviations from exact coincidence due to errors.

The headstock 10 shown in FIG. 1 collectively includes a headstock 10A, which is a front headstock, and a headstock 10B, which is a rear headstock. Therefore, the description of the headstock 10 applies to both the headstock 10A and the headstock 10B.
A main spindle 11 and a built-in motor 20 are incorporated into the main spindle 10 . Built-in motor 20 is an example of main shaft motor M1. The lathe 1 shown in FIG. 1 is a main-spindle movable lathe, and the headstock 10 is movable in the Z-axis direction. Of course, the lathe may be a fixed spindle type lathe in which the headstock 10A does not move, or the headstock 10A may move in the Z-axis direction without moving the headstock 10B.

The main shaft 11 has a gripping part 12 such as a collet, and the gripping part 12 releasably grips the workpiece W1. When the workpiece W1 to be processed is, for example, a cylindrical (bar-shaped) long material, the workpiece W1 may be supplied to the gripping section 12 from the rear end of the main shaft 11 of the headstock 10A. In this case, a guide bush that supports the workpiece W1 slidably in the Z-axis direction may be arranged on the front side of the main shaft 11 of the headstock 10A. When the workpiece W1 to be processed is a short material, the workpiece W1 may be supplied to the gripping section 12 from the front end of the main spindle 11 of the headstock 10A. The main spindle 11 holding the workpiece W1 is rotatable together with the workpiece W1 about the main spindle center line AX1. The workpiece W1 after front processing is transferred from the main spindle 11 of the headstock 10A to the main spindle 11 of the headstock 10B, and is turned into a product by back processing.

FIG. 2 schematically illustrates the configuration of the main parts of the headstock 10 including the built-in motor 20 and the cooling device 40 of the lathe 1. The built-in motor 20 shown in FIG. 2 has a non-rotating stator 21 attached to the headstock 10, a rotating rotor 22 attached to the main shaft 11, and a jacket 23 covering the outside of the stator 21. The rotor 22 can rotate together with the main shaft 11 around the main shaft center line AX1 inside the stator 21. The built-in motor 20 rotates the main shaft 11 at high speed by rotating the rotor 22 at a timing controlled by the NC device 70. Heat is generated in the built-in motor 20 due to the high speed rotation of the rotor 22. A circulation path 41 for cooling oil F2 as cooling fluid F1 is connected to the jacket 23. Therefore, the jacket 23 contains cooling oil F2.

As shown in FIG. 1, the built-in motor 20 has a built-in temperature sensor 25 that detects its own temperature in order to prevent seizure. The temperature sensor 25 is an example of a built-in temperature sensor S1 provided in the main shaft motor M1. Temperature sensor 25 is also called a thermistor. In this specific example, the temperature sensor 25, which is the built-in temperature sensor S1, is used to determine the failure of the cooling device 40.

The tool rest 28 shown in FIG. 1 is attached with a plurality of tools TO1 for processing the workpiece W1, and is movable in the Y-axis direction. Of course, the tool rest 28 may be moved in the X-axis direction or the Z-axis direction. The tool rest 28 may be a turret tool rest, a comb-shaped tool rest, or the like. The plurality of tools TO1 include cutting tools including cut-off tools, rotating tools such as drills and end mills, and the like.

The drive device 30 shown in FIG. 1 collectively refers to the drive device 30A for the headstock 10A, the drive device 30B for the headstock 10B, and the drive device 30C for the tool rest 28. Therefore, the description of the drive device 30 applies to any of the drive devices 30A, 30B, and 30C.
The drive device 30 includes a servo motor 31 that rotates under the control of the NC device 70, and converts the rotational force from the servo motor 31 into linear force using, for example, a ball screw mechanism. The servo motor 31 has a built-in temperature sensor 32 that detects its own temperature. The drive device 30A moves the headstock 10A together with the main spindle 11 and the built-in motor 20 in the Z-axis direction under the control of the NC device 70. The drive device 30B moves the headstock 10B in the Z-axis direction together with the main spindle 11 and the built-in motor 20 under the control of the NC device 70. The drive device 30C moves the tool rest 28 in the Y-axis direction under the control of the NC device 70.

The cooling device 40 shown in FIG. 2 is oil-cooled and includes a circulation path 41 in which a radiator 42 and a pump 43 are provided, and a fan 44 that promotes heat radiation from the radiator 42. The cooling oil F2 flows through the circulation path 41 and cools the main shaft motor M1. The radiator 42 is a structure that forms an expanded portion of the circulation path 41, and because the inner surface area of the circulation path 41 per unit volume of the cooling oil F2 is large, it is easy to release the heat of the cooling oil F2. The pump 43 circulates the cooling oil F2 in the circulation path 41. After coming out of the pump 43, the cooling oil F2 shown in FIG. exits from the bottom of the pump and returns to the pump 43. The fan 44 removes heat from the cooling oil F2 by blowing air onto the radiator 42. Therefore, the fan 44 promotes heat dissipation in the circulation path 41.

FIG. 3 schematically illustrates the configuration of the electric circuit of the lathe 1. In the lathe 1 shown in FIG. 3, an operating unit 80, a servo motor 31, a built-in motor 20 containing a temperature sensor 25, a temperature sensor 3 embedded in the exterior 2, etc. are connected to the NC device 70. . The NC device 70 includes a CPU (Central Processing Unit) 71 which is a processor, a ROM (Read Only Memory) 72 which is a semiconductor memory, a RAM (Random Access Memory) 73 which is a semiconductor memory, a timer circuit 74, and an I/F (interface). 75, etc. Therefore, the NC device 70 is a type of computer. In FIG. 3, I/Fs such as the operation unit 80, the servo motor 31, the built-in motor 20, the temperature sensor 3, and the external computer 100 are collectively shown as an I/F 75. A control program PR1 for interpreting and executing the machining program PR2 and determining a failure of the cooling device 40 is written in the ROM 72. The ROM 72 may be a rewritable semiconductor memory. A machining program PR2 created by an operator is stored in the RAM 73 in a rewritable manner. The machining program is also called an NC program. The CPU 71 uses the RAM 73 as a work area and executes the control program PR1 recorded in the ROM 72, thereby realizing the functions of the NC device 70. The NC device 70 that executes the control program PR1 is an example of the failure determination section U1. Of course, some or all of the functions realized by the control program PR1 may be realized by other means such as an ASIC (Application Specific Integrated Circuit).

The operation unit 80 includes an input unit 81 and a display unit 82, and functions as a user interface for the NC device 70. The input unit 81 includes, for example, buttons and a touch panel for receiving operation input from an operator. The display unit 82 includes, for example, a display that displays the contents of various settings received from the operator and various information regarding the lathe 1. The operator can store the machining program PR2 in the RAM 73 using the operation unit 80 or an external computer.
The servo motor 31 moves the headstock 10 and the tool rest 28 according to commands from the NC device 70. The built-in motor 20 rotates the main shaft 11 according to commands from the NC device 70.

As shown in FIG. 1, the external computer 100 includes a CPU 101, a ROM 102, a RAM 103, a storage device 104, an input device 105, a display device 106, an audio output device 107, an I/F 108, a clock circuit 109, and the like. A control program for the computer 100 is stored in the storage device 104, read into the RAM 103 by the CPU 101, and executed by the CPU 101. As the storage device 104, a semiconductor memory such as a flash memory, a magnetic recording medium such as a hard disk, etc. can be used. As the input device 105, a pointing device, a keyboard, a touch panel attached to the surface of the display device 106, or the like can be used. The I/F 108 is connected to the I/F 75 of the NC device 70 by wire or wirelessly, and receives data from the NC device 70 and transmits data to the NC device 70. The I/Fs 108 and 75 may be connected to a network such as the Internet or an intranet. The computer 100 includes a personal computer including a tablet terminal, a mobile phone such as a smartphone, and the like.

First, with reference to FIGS. 4A and 4B, a temperature change in the built-in motor 20 when the cooling device 40 fails during continuous processing of the workpiece W1 will be described.
FIG. 4A schematically illustrates a temperature change of the temperature sensor 25 of the built-in motor 20 when the pump 43 fails during continuous processing of the workpiece. FIG. 4B schematically illustrates a temperature change of the temperature sensor 25 of the built-in motor 20 when the fan 44 fails during continuous processing of the workpiece. In FIGS. 4A and 4B, the horizontal axis indicates the time t from the start of workpiece machining, and the vertical axis indicates the temperature T1 detected by the temperature sensor 25.

When continuous machining of the workpiece W1 starts at t=0, the temperature T1 detected by the temperature sensor 25 increases as shown in FIGS. 4A and 4B. This is because the temperature of the built-in motor 20 increases as the built-in motor 20, which had not been operating, operates frequently due to the start of continuous machining of the workpiece. On the other hand, since the cooling device 40 also operates during continuous processing of the workpiece W1, the heat of the built-in motor 20 is taken away by the cooling oil F2 circulated by the cooling device 40, and the temperature rise of the built-in motor 20 is stopped. Therefore, once the detected temperature T1 becomes steady at the timing of t=t1, the detected temperature T1 hardly changes thereafter.

After the detected temperature T1 becomes steady, as shown in FIG. 4A, if the pump 43 fails at timing t=t2, the detected temperature T1 rapidly increases. This is because the circulation of the cooling oil F2 is stopped when the pump 43 is stopped, and thereby the heat of the built-in motor 20 is no longer released. Although the detected temperature T1 increases gradually as time t passes, the detected temperature T1 continues to rise. Therefore, it is necessary to suppress the temperature rise of the built-in motor 20 by stopping continuous processing of the workpiece W1.

After the detected temperature T1 becomes steady, as shown in FIG. 4B, if the fan 44 fails at timing t=t2, the detected temperature T1 suddenly rises, although more slowly than when the pump 43 fails. I will do it. This is because, although some heat from the built-in motor 20 is released as the circulation of the cooling oil F2 continues, the release of heat from the cooling oil F2 flowing through the radiator 42 is suppressed by stopping the fan 44. Although the detected temperature T1 increases gradually as time t passes, the detected temperature T1 continues to rise. Therefore, it is necessary to suppress the temperature rise of the built-in motor 20 by stopping continuous processing of the workpiece W1.

The NC device 70 of this specific example determines that the detected temperature T1 sequentially obtained by the temperature sensor 25 of the built-in motor 20 after starting continuous machining of the workpiece W1 has become steady, and then the change in the detected temperature T1 falls within the allowable range. If it exceeds the threshold, it is determined that the cooling device 40 is out of order. Further, the NC device 70 determines which part is malfunctioning among the plurality of parts included in the cooling device 40 based on the change in the detected temperature T1. The NC device 70 of this specific example determines whether the pump 43 or the fan 44 has failed depending on whether the change in the detected temperature T1 exceeds a predetermined criterion.

FIG. 5 schematically shows an example in which the state of the lathe 1 is determined from discrete data of the temperature T1(t) detected by the temperature sensor 25. In FIG. 5 as well, the horizontal axis shows the time t from the start of workpiece machining, and the vertical axis shows the temperature T1 detected by the temperature sensor 25.
The NC device 70 periodically acquires the detected temperature T1(t) from the temperature sensor 25 of the built-in motor 20. The interval Δt for acquiring the detected temperature T1(t) may be approximately 1 minute, but may be as short as approximately 30 seconds or as long as approximately 5 minutes. The resolution of the detected temperature T1(t) may be about 1° C., but it may be as fine as about 0.1° C. in order to improve the accuracy of determination.

The change ΔT1(t) in the detected temperature T1(t) is defined as the difference between the maximum value of the detected temperature T1 and the minimum value of the detected temperature T1 in a predetermined period P1 before time t, in order to reduce errors. Hereinafter, the change ΔT1(t) in the detected temperature will also be referred to as the temperature change ΔT1(t). If the number of data of the detected temperature T1 included in the predetermined period P1 is n (n is an integer of 3 or more) and the variable i is an integer from 0 to n-1, then the detected temperature T1 (i) included in the predetermined period P1 is teeth,
T1(i)=T1(t),...,T1(t-i×Δt),...,T1(t-(n-1)×Δt)
It is expressed as Therefore, when the number n of data of the detected temperature T1 included in the predetermined period P1 is 6, the detected temperature T1(i) included in the predetermined period P1 is
T1(i)=T1(t-0×Δt), T1(t-1×Δt), T1(t-2×Δt), T1(t-3×Δt), T1(t-4×Δt), T1 (t-5×Δt)
It is expressed as If the maximum value of the detected temperature T1(i) is MAX(T1(i)) and the minimum value of the detected temperature T1(i) is MIN(T1(i)), the temperature change ΔT1(t) is t≧P1 In,
ΔT1(t)=MAX(T1(i))-MIN(T1(i))...(1)
It is expressed as
As shown in FIG. 5, when the number n of detected temperature T1 data included in the predetermined period P1 is 6, the temperature change ΔT1(t) is the detected temperature T1(i)=T1( It is the difference between the maximum value MAX (T1 (i)) and the minimum value MIN (T1 (i)) at t), ..., T1 (t-5×Δt).

When continuous machining of the workpiece W1 starts at t=0, as shown in FIGS. 4A and 4B, the detected temperature T1(t) rises for a while, and then reaches a steady timing t1. Therefore, the NC device 70 determines that the detected temperature T1(t) is steady if the temperature change ΔT1(t) exceeds the predetermined threshold TH1 (TH1>0) after the time t reaches the predetermined period P1. If the temperature change ΔT1(t) becomes equal to or less than the threshold value TH1, it is determined that the detected temperature T1(t) has become steady.
Note that if the threshold value obtained by adding a small amount to the threshold value TH1 is TH1+α, the above-mentioned determination can be replaced with a determination as to whether ΔT1(t)≧TH1+α. This case is also included in the determination of whether ΔT1(t) exceeds the threshold value TH1.

FIG. 6 schematically shows an example of temperature change ΔT1(t) according to time t. In FIG. 6, the horizontal axis indicates the time t from the start of workpiece machining, and the vertical axis indicates the temperature change ΔT1(t) of the temperature sensor 25.
As shown in FIG. 6, after the time t reaches the predetermined period P1, the temperature change ΔT1(t) decreases for a while, and eventually reaches a timing t1 when it becomes equal to or less than the threshold value TH1.

If the cooling device 40 fails after the detected temperature T1(t) becomes steady during continuous processing of the workpiece, the detected temperature T1(t) of the temperature sensor 25 suddenly changes as shown at timing t2 shown in FIGS. 4A and 4B. Rise. The NC device 70 determines that the cooling device 40 is malfunctioning when the temperature change ΔT1(t) exceeds the threshold value TH2 (TH2>TH1) after ΔT1(t)≦TH1 as shown in FIG. If the temperature change ΔT1(t) does not exceed the threshold value TH2, it is determined that there is no failure in the cooling device 40. The threshold value TH2 is an example of an allowable range of the temperature change ΔT1(t). As shown in FIG. 6, the temperature change ΔT1(t), which was below the threshold TH1, suddenly increases until it exceeds the threshold TH2 after timing t2 when the cooling device 40 fails. FIG. 6 shows timing t3 at which ΔT1(t)=TH2.
Note that if the threshold value obtained by adding a small amount to the threshold value TH2 is TH2+α, the above-mentioned judgment can be replaced with a judgment as to whether ΔT1(t)≧TH2+α. This case is also included in the determination of whether ΔT1(t) exceeds the threshold value TH2.

After ΔT1(t)>TH2 during continuous workpiece machining, if the pump 43 fails as shown in FIG. 4A, the detected temperature T1(t) will rise rapidly and the fan 44 will stop as shown in FIG. 4B. In the case of a failure, the detected temperature T1(t) increases less than in the case of a failure of the pump 43. In order to determine the failure part, the NC device 70 first searches for the local maximum value ΔT1max of the temperature change ΔT1(t) that is obtained sequentially after ΔT1(t)>TH2. The maximum value ΔT1max means the value at which the temperature change ΔT1(t) first becomes a peak in the graph in the range ΔT1(t)>TH2. Thereafter, the NC device 70 determines that the pump 43 is out of order when the local maximum value ΔT1max exceeds the threshold value TH3 (TH3>TH2), and determines that the fan 44 is malfunctioning when the local maximum value ΔT1max does not exceed the threshold value TH3. I am deciding that it is malfunctioning. Whether or not ΔT1max>threshold value TH3 is an example of a criterion for determining the temperature change ΔT1(t). In FIG. 6, the temperature change ΔT1(t) according to time t when the pump 43 fails is shown as a solid line, and the temperature change ΔT1(t) according to time t when the fan 44 fails is shown at two points. Indicated by a dashed line. It is shown that when the pump 43 fails, there is a timing t4 after the timing t3 where the temperature change ΔT1(t) exceeds the threshold value TH3, and then there is a timing t5 where the temperature change ΔT1(t) reaches the maximum value ΔT1max. There is. It is shown that when the fan 44 fails, there is a timing after timing t3 when the temperature change ΔT1(t) reaches a local maximum value ΔT1max that is less than or equal to the threshold value TH3.
Note that if the threshold value obtained by adding a small amount to the threshold value TH3 is TH3+α, the above-mentioned judgment can be replaced with a judgment as to whether ΔT1max≧TH3+α. This case is also included in the determination of whether ΔT1max exceeds the threshold value TH3.

(3) Specific example of cooling device failure determination processing:
FIG. 7 shows an example of a cooling device failure determination process that implements the failure determination method described above. This process is performed by the NC device 70 that executes the control program PR1, and starts when continuous machining of the workpiece W1 is started. The cooling device failure determination process shown in FIG. 7 will be described below with reference to FIGS. 1 to 6 as well.

When the cooling device failure determination process starts, the NC device 70 acquires detected temperatures T1(t) at constant intervals Δt from the temperature sensor 25 of the built-in motor 20 shown in FIGS. 1 to 3 (step S102). Hereinafter, the description of "step" will be omitted. The NC device 70 repeats the process of S102 until the time t from the start of workpiece machining reaches the predetermined period P1. After acquiring the detected temperature T1(t), the NC device 70 calculates a change ΔT1(t) in the detected temperature T1(t) during the predetermined period P1 according to the above-mentioned equation (1) (S104). After calculating the temperature change ΔT1(t), the NC device 70 branches the process depending on whether the temperature change ΔT1(t) has become equal to or less than the threshold value TH1 (S106). If ΔT1(t)>TH1, the NC device 70 determines that the detected temperature T1(t) is not steady, and performs the process of S102 to obtain the detected temperature T1(t), and the temperature change ΔT1(t). The process of calculating S104 and the judgment process of S106 are repeated. On the other hand, when ΔT1(t)≦TH1, the NC device 70 determines that the detected temperature T1(t) has become steady, and advances the process to S108.

In S108, the NC device 70 further acquires the detected temperature T1(t) from the temperature sensor 25 at a constant interval Δt. After obtaining the detected temperature T1(t), the NC device 70 calculates the change ΔT1(t) in the detected temperature T1(t) during the predetermined period P1 according to the above-mentioned equation (1) (S110). After calculating the temperature change ΔT1(t), the NC device 70 branches the process depending on whether the temperature change ΔT1(t) exceeds the threshold TH2 (S112). If ΔT1(t)≦TH2, the NC device 70 determines that there is no failure in the cooling device 40, and performs the process of S108 to obtain the detected temperature T1(t) and the process of S110 to calculate the temperature change ΔT1(t). The process and the judgment process of S112 are repeated. The repeated processing of S108 to S112 is continued until the continuous machining of the workpiece W1 is completed as long as the temperature change ΔT1(t) does not exceed the threshold value TH2. On the other hand, if ΔT1(t)>TH2, the NC device 70 determines that the cooling device 40 is out of order and advances the process to S114.

In S114, the NC device 70 notifies the operator of the failure of the cooling device 40. The notification process in S114 includes, for example, a process of displaying on the display unit 82 of the lathe 1 shown in FIGS. 106 to display that the cooling device 40 has failed, a process to cause the audio output device 107 of the computer 100 to output a warning sound or a voice indicating that the cooling device 40 has failed, a combination of at least a portion of these processes, etc. can do. If the computer 100 is a mobile terminal, it is possible to notify an operator who is away from the factory of the failure of the cooling device 40.
Further, along with the failure notification, the NC device 70 may display a list of detected temperatures T1(t) obtained so far on the display unit 82 of the lathe 1, the display device 106 of the computer 100, etc. Further, the NC device 70 may display a graph showing the detected temperature T1(t) and the temperature change ΔT1(t) with respect to time t on the display unit 82, the display device 106, etc. together with the notification of the failed part. As a result, the operator who sees the detected temperature T1(t) and the temperature change ΔT1(t) can predict the failure location before being notified of the failure location.

The NC device 70 may stop the processing of the workpiece W1 by issuing a workpiece processing stop command to stop the processing of the workpiece W1 when it is determined that the cooling device 40 has failed. Although this case is also included in the present technology, in this specific example, in order to determine which part is malfunctioning among the plurality of parts included in the cooling device 40, continuous processing of the workpiece W1 is continued for a little while longer.

After the notification processing in S114, the NC device 70 substitutes the current temperature change ΔT1(t) for the variable ΔT1max in order to search for the local maximum value ΔT1max (S116). After substituting ΔT1(t), the NC device 70 further acquires the detected temperature T1(t) from the temperature sensor 25 at a constant interval Δt (S118). After obtaining the detected temperature T1(t), the NC device 70 calculates the change ΔT1(t) in the detected temperature T1(t) during the predetermined period P1 according to the above-mentioned equation (1) (S120). After calculating the temperature change ΔT1(t), the NC device 70 branches the process depending on whether the temperature change ΔT1(t) is less than ΔT1max (S122). If ΔT1(t)≧ΔT1max, the NC device 70 determines that the maximum value of the temperature change ΔT1(t) is not determined, and performs the process of S116 of substituting the current temperature change ΔT1(t) for ΔT1max, and detects the detected temperature. The process of S118 to obtain T1(t), the process of S120 to calculate the temperature change ΔT1(t), and the determination process of S122 are repeated. On the other hand, if ΔT1(t)<ΔT1max, the NC device 70 determines that ΔT1max as a variable has reached the maximum value, and advances the process to S124.

In S124, the NC device 70 branches the process depending on whether the local maximum value ΔT1max exceeds the threshold value TH3. If ΔT1max>TH3, the NC device 70 determines that the pump 43 is out of order and notifies the operator of the outage of the pump 43 (S126). On the other hand, if ΔT1max≦TH3, the NC device 70 determines that the fan 44 is malfunctioning, and notifies the operator of the malfunction of the fan 44 (S128). The notification processes in S126 and S128 include, for example, the process of displaying information indicating the faulty part (pump 43 or fan 44) on the display unit 82 of the lathe 1 shown in FIGS. 1 and 3, and the process of displaying a warning light (not shown) corresponding to the faulty part. A process of causing the display device 106 of the computer 100 to display information on the failed part, a process of causing the audio output device 107 of the computer 100 to output the information of the failed part as a voice, and a combination of at least a portion of these processes. , etc. If the computer 100 is a mobile terminal, it is possible to notify an operator who is away from the factory of information on the location of the failure.
Further, along with the notification of the failure part, the NC device 70 may display a list of detected temperatures T1(t) obtained so far on the display unit 82 of the lathe 1, the display device 106 of the computer 100, etc. Further, the NC device 70 may display a graph showing the detected temperature T1(t) and the temperature change ΔT1(t) with respect to time t on the display unit 82, the display device 106, etc. together with the notification of the failed part. Thereby, the operator who sees the detected temperature T1(t) and the temperature change ΔT1(t) can judge whether the notification of the failure part is correct or not.

After the processing in S126 or S128, the NC device 70 stops the processing of the workpiece W1 by issuing a workpiece processing stop command to stop processing the workpiece W1 (S130), and ends the cooling device failure determination process shown in FIG. 7. . As a result, continuous processing of the workpiece W1 is stopped when the cooling device 40 fails, and a rise in temperature of the built-in motor 20 is suppressed.

As explained above, if the cooling device 40 breaks down after the temperature T1(t) detected by the temperature sensor 25 of the built-in motor 20 becomes steady during continuous processing of the workpiece W1, the change ΔT1 in the detected temperature T1(t) will eventually occur. (t) exceeds the threshold value TH2, which is an allowable range. By determining that ΔT1(t)>TH2 in the process of S112 described above, it can be determined that the cooling device 40 is out of order.
Furthermore, since the temperature change ΔT1(t) is larger when the pump 43 fails than when the fan 44 fails, when the pump 43 fails, the maximum value ΔT1max of the temperature change ΔT1 is the threshold value TH3 which is the criterion. Exceed. If the fan 44 fails, the local maximum value ΔT1max does not exceed the threshold value TH3. By determining whether the local maximum value ΔT1max exceeds the threshold value TH3 in the process of S124 described above, it is possible to determine whether the pump 43 or the fan 44 has failed.

Therefore, the lathe of this specific example can determine the failure of the cooling system without separately installing a sensor for detecting failure of the cooling system, and can determine the failure part of the cooling system, thereby reducing the cost increase. Can be suppressed. Additionally, since the location of the failure in the cooling device can be identified, the cooling device can be easily repaired, making this lathe convenient.

The above explanation was based on the assumption that the temperature of the built-in motor 20 suddenly rises from its normal temperature due to a failure of the cooling device 40. If it descends, it is also determined that the cooling device 40 is malfunctioning. This is because even if the detected temperature T1(t) suddenly falls after the detected temperature T1(t) becomes steady, ΔT1(t)>TH2. Therefore, this specific example can provide a lathe with high reliability.

Further, the cooling device failure determination process shown in FIG. 7 may be performed by the computer 100 shown in FIG. In this case, by storing the control program PR1 that executes the cooling device failure determination process in the storage device 104 of the computer 100, the computer 100 can read the control program PR1 from the storage device 104 to the RAM 103 and execute it. It is assumed that the lathe 1 acquires the detected temperature T1(t) from the temperature sensor 25 at regular intervals Δt, transmits it to the computer 100, and stops machining the workpiece W1 when receiving a workpiece machining stop command from the computer 100. . The computer 100 may perform processing to obtain the detected temperature T1(t) from the lathe 1 in S102, S108, and S118. By determining in S112 that the temperature change ΔT1(t) exceeds the threshold value TH2, the computer 100 can determine that the cooling device 40 is out of order. Further, the computer 100 can determine whether the pump 43 or the fan 44 has failed by determining whether the local maximum value ΔT1max exceeds the threshold value TH3 in S124. In S130, the computer 100 may send a command to the lathe 1 to stop processing the workpiece.
Furthermore, the cooling device failure determination process may be performed by the NC device 70 and the computer 100 in cooperation. For example, it is conceivable that the NC device 70 performs the processes of S102 to S114 up to the notification of the failure of the cooling device 40, and the computer 100 that receives the failure notification determines the location of the failure by the processes of S116 to S130. In this case, the failure detection unit U1 is realized by cooperation between the NC device 70 and the computer 100.

(4) Modification example:
Various modifications of the present invention are possible.
For example, the time interval of the detected temperature T1(t) is not limited to a fixed interval, and may vary.
The temperature change ΔT1(t) during the predetermined period P1 may be the slope of an approximate straight line obtained from the coordinates of each detected temperature T1(t) during the predetermined period P1 on the t-T1 plane, or may be the absolute value of the slope. Therefore, the lathe or computer may determine that the detected temperature T1(t) has become steady when the absolute value of the above-mentioned inclination has become below the threshold TH1, and the absolute value of the above-mentioned inclination may be determined to be the threshold TH2. If it exceeds , it may be determined that the cooling device 40 is out of order, and the location of the failure may be determined based on the maximum value of the absolute value of the above-mentioned slope.
The above-described processing can be changed as appropriate, such as by changing the order. For example, in the cooling device failure determination process of FIG. 7, the process of S130 for issuing a workpiece processing stop command can be performed before the determination process of S124.
The spindle rotated by the spindle motor is not limited to a spindle that grips a workpiece, but may be a tool spindle that rotates together with a rotary tool that processes the workpiece. Further, the main shaft motor is not limited to a built-in motor, but may be a motor externally attached to the main shaft. Furthermore, the cooling fluid is not limited to cooling oil, but may be cooling water, gas, or a refrigerant whose state changes between gas and liquid.

FIG. 8 schematically shows an example of a lathe 1 including a spindle motor M1 for driving a rotary tool and a cooling device 40B for the spindle motor M1. The cooling device 40B is included in the cooling device 40 of the present technology. The lathe 1 shown in FIG. 8 includes a tool rest 28B holding a plurality of rotary tools TO2 for machining a workpiece. The tool post 28B includes a tool spindle 11B that rotates together with each rotary tool TO2, and a spindle motor M1 that rotates the tool spindle 11B about the spindle center line AX1. The tool spindle 11B is included in the spindle 11 of the present technology. The main shaft motor M1 has a built-in temperature sensor S1 and is controlled by an NC device 70 that executes a control program PR1. The spindle motor M1 shown in FIG. 8 is externally attached to the tool spindle 11B and exposed outside the tool post 28B. The cooling device 40B is of an air cooling type and includes a fan 45 that sends air F3 to the main shaft motor M1. Air F3 is an example of cooling fluid F1. The fan 45 removes heat from the main shaft motor M1 by applying wind to the main shaft motor M1.

If the fan 45 fails during continuous workpiece machining, the temperature of the main spindle motor M1 changes as illustrated in FIG. 4B. Therefore, by performing the processes S102 to S114 and S130 shown in FIG. 7, the NC device 70 can determine the failure of the cooling device 40B and stop processing the workpiece. Of course, the processing of S102 to S114 and S130 shown in FIG. 7 may be performed by the computer 100.

Further, by using the temperature sensor 3 and the like provided on the exterior 2 shown in FIG. 1, it is possible to improve the accuracy of determining the failure part of the cooling device 40. Temperature sensor 3 is an example of second temperature sensor S2 that detects temperature influenced by outside temperature. Machine learning may be used to improve the accuracy of determining faulty parts.

FIG. 9 schematically shows an example of a lathe system SY1 including a machine learning unit U2 in the computer 100. In FIG. 9, descriptions and explanations of elements that partially overlap with those in FIGS. 1 and 3 are omitted. The lower part of FIG. 9 shows an example of the structure of the database DB.

The storage device 104 of the computer 100 shown in FIG. 9 stores a machine learning program PR3 corresponding to the machine learning section U2 and a failure part determination program PR4 corresponding to the failure determination section U1. These programs (PR3, PR4) are executed by being read into the RAM 103 by the CPU 101. The RAM 103 of the computer 100 stores a database DB and a learning model LM generated based on the database DB. The learning model LM indicates that there is a failure in the cooling device 40 based on the detected temperature T1(t) sequentially obtained by the built-in temperature sensor S1 and the second detected temperature T2(t) sequentially obtained by the second temperature sensor S2. This is a program for causing the computer 100 to function to discriminate parts. The generated learning model LM may be transmitted from the computer 100 to the NC device 70 and stored in the RAM 73 of the NC device 70. Thereby, the NC device 70 can determine the failure part of the cooling device 40 according to the learning model LM.

In the database DB, an identification number j, which is identification information for identifying a record, is linked to a detected temperature T1(t), a second detected temperature T2(t), and failure part information IN1 representing a failure part. stored in the state. The failure part information IN1 represents a part that has failed among the plurality of parts included in the cooling device 40. In FIG. 9, the detected temperature T1(t) linked to the identification number j is shown as T1-j(t), and the second detected temperature T2(t) linked to the identification number j is shown as T2 -j(t), and failure part information IN1 linked to identification number j is shown as part j.
Note that since the cooling device 40 does not break down frequently, the computer 100 may receive detected temperatures (T1(t), T2(t)) from a plurality of lathes and store them in the database DB.

FIG. 10 shows an example of a learning process for generating a learning model LM. This process is performed by the computer 100 that executes the machine learning program PR3.
When the learning process starts, the computer 100 detects the detected temperature T1(t) sequentially obtained by the built-in temperature sensor S1 (for example, temperature sensor 25) and the detected temperature T1(t) sequentially obtained by the second temperature sensor S2 (for example, temperature sensor 3). The second detected temperature T2(t) is obtained (S202). The computer 100 may acquire the detected temperatures (T1(t), T2(t)) from the lathe 1 during continuous machining of the workpiece at regular intervals Δt, or may acquire the detected temperatures from the lathe 1 all at once after continuous machining of the workpiece. (T1(t), T2(t)) may be obtained.

Next, the computer 100 receives an input of the failure part of the cooling device 40 (S204). For example, when the operator of the lathe 1 locates the faulty part when the cooling device 40 breaks down, he or she may input the faulty part into the computer 100 using the input device 105. Also, it is better to have the detected temperatures (T1(t), T2(t)) when the cooling device 40 is not out of order in the database DB, so if the cooling device 40 is not out of order, the operator can What is necessary is to perform an operation to input into the computer 100 that the cooling device 40 is not out of order. The computer 100 may perform a process of accepting an operation for inputting a failed part or that there is no failure using the input device 105.

Furthermore, the computer 100 stores the detected temperatures (T1(t), T2(t)) and failure part information IN1 indicating a failure part or no failure in the database DB in association with the identification number j of the record. (S206). Since it is better to have more records in the database DB, the processes of S202 to S206 are repeated.

After the information is stored in the database DB, the computer 100 generates the learning model LM in the RAM 103 by supervised machine learning based on the information stored in the database DB (S208). As the learning model LM, a neural network, a Bayesian network, a learning model in which at least one of these is used as a main part and a conversion formula is combined, etc. can be used. When the learning model LM includes a neural network, learning may be performed using a deep learning method. Note that the details of neural networks, Bayesian networks, deep learning, etc. are well known and will not be described here. The obtained learning model LM is constructed by inputting the detected temperature T1(t) sequentially obtained by the built-in temperature sensor S1 and the second detected temperature T2(t) sequentially obtained by the second temperature sensor S2. Outputs the determination result of the failed part among the plurality of parts included in 40. That is, the learning model LM is a learned model that determines the failure part of the cooling device 40 based on the sequentially obtained detected temperatures (T1(t), T2(t)).

After generating the learning model LM, the computer 100 stores the learning model LM (S210) and ends the learning process. When the lathe 1 uses the learning model LM, the computer 100 only needs to transmit the learning model LM to the lathe 1. The lathe 1 that has received the learning model LM can perform the process of determining the faulty part of the cooling device 40 by storing the learning model LM in the RAM 73.

FIG. 11 shows an example of a failure part determination process for determining a failure part in the cooling device 40. This process is performed, for example, by the computer 100 holding the learning model LM in the RAM 103, and starts when continuous machining of the workpiece W1 starts.
First, the computer 100 sequentially acquires the temperature T1(t) detected by the built-in temperature sensor S1 and the second temperature T2(t) detected by the second temperature sensor S2 from the lathe 1 (S302).

Next, the computer 100 inputs the sequentially obtained detected temperatures (T1(t), T2(t)) to the learning model LM, thereby providing the learning model LM with the determination result of the faulty part or the faulty part. The determination result that there is no such information is output (S304). The process in S304 may be performed when it is determined in S112 in FIG. 7 that the cooling device 40 is out of order. In this case, the computer 100 inputs the detected temperatures (T1(t), T2(t)) into the learning model LM in S304 to perform a process of determining the faulty part of the cooling device 40. Note that when the determination result that there is no failure is output in S304, no further processing is performed until the detected temperature (T1(t), T2(t)) is input to the learning model LM again. It is also possible to wait.

When the faulty part is determined, the computer 100 notifies the operator of the faulty part of the cooling device 40 (S306), and ends the faulty part determination process. The notification processing in S306 includes, for example, a process of displaying information indicating a faulty part on the display unit 82 of the lathe 1 shown in FIGS. A process of causing the display device 106 of the computer 100 to display the information on the failed part, a process of causing the audio output device 107 of the computer 100 to output the information of the failed part, a combination of at least a portion of these processes, etc.
Of course, the NC device 70 may perform the failure part determination process, or the NC device 70 and the computer 100 may cooperate to perform the failure part determination process.

In the examples shown in FIGS. 9 to 11, the failure location of the cooling device is determined by adding the influence of the outside temperature to the temperature detected by the built-in temperature sensor, so that the accuracy of determining the failure location is improved. Further, since there is no need to separately install a sensor for detecting a failure in each part of the cooling device, an increase in cost is suppressed.

Note that when performing machine learning in S208 of FIG. 10, the change ΔT1(t) in the detected temperature T1(t) may be used instead of the detected temperature T1(t), and the change ΔT1(t) in the second detected temperature T2(t) may be used instead of the detected temperature T1(t). Instead, a change in the second detected temperature T2(t) (referred to as ΔT2(t)) may be used. Here, the second detected temperature included in the predetermined period P1 is
T2(i)=T2(t),...,T2(t-i×Δt),...,T2(t-(n-1)×Δt)
It is expressed as If the maximum value of the second detected temperature T2(i) is MAX(T2(i)) and the minimum value of the second detected temperature T2(i) is MIN(T2(i)), the temperature change ΔT2(t) is , t≧P1,
ΔT2(t)=MAX(T2(i))-MIN(T2(i))...(2)
It is expressed as
When temperature changes (ΔT1(t), ΔT2(t)) are used in machine learning, a provisional learning model is created that determines the failure location based on the sequentially obtained temperature changes (ΔT1(t), ΔT2(t)). generated. Combining the conversion formula for calculating temperature changes (ΔT1(t), ΔT2(t)) from detected temperatures (T1(t), T2(t)) with a temporary learning model, the sequentially obtained detected temperatures (T1(t) ), T2(t)), a learning model LM for determining the failure location is obtained.

Further, the database DB may also hold a change ΔT1(t) in the detected temperature T1(t) instead of the detected temperature T1(t), and a second detected temperature instead of the second detected temperature T2(t). The change ΔT2(t) in T2(t) may be maintained.

Furthermore, the second detected temperature T2(t), which is influenced by the outside temperature, is not limited to the detected temperature of the temperature sensor 3 provided in the exterior 2 shown in FIG. The temperature detected by the temperature sensor 32 may be used. In this case, the temperature sensor 32 is an example of the second temperature sensor S2 that detects the temperature influenced by the outside air temperature. Further, the machine learning unit may use the detected temperatures of two or more temperature sensors among the plurality of temperature sensors including the temperature sensors 3 and 32 as the second detected temperature in order to generate the learning model LM. In addition, the machine learning unit may additionally use information other than the detected temperature, such as the size of the workpiece W1, in order to generate the learning model LM. Thereby, machine learning can be performed that takes into account information such as the size of the workpiece W1.
Of course, the time interval between the detected temperatures (T1(t), T2(t)) is not limited to a fixed interval, and may vary.

Furthermore, as illustrated in FIG. 12, the lathe 1 may generate the learning model LM by executing the machine learning program PR3. FIG. 12 schematically shows an example of a lathe 1 including a machine learning section U2. In FIG. 12, descriptions and explanations of elements that partially overlap with those in FIG. 3 are omitted. The lower part of FIG. 12 shows an example of the structure of the database DB. The database DB shown in FIG. 12 is the same as the database DB shown in FIG. 9, so a description thereof will be omitted.

A control program PR1 corresponding to the failure determination section U1 and a machine learning program PR3 corresponding to the machine learning section U2 are written in the ROM 72 of the NC device 70 shown in FIG. The RAM 73 of the NC device 70 stores a machining program PR2, a database DB, and a learning model LM. The learning model LM is a program for causing the NC device 70 to function so as to determine which part of the cooling device 40 is malfunctioning based on sequentially obtained detected temperatures (T1(t), T2(t)).

The NC device 70 can perform the learning process according to the flowchart shown in FIG.
When the learning process starts, the NC device 70 sequentially acquires the detected temperature T1(t) of the built-in temperature sensor S1 and the second detected temperature T2(t) of the second temperature sensor S2 (S202). After sequentially acquiring the detected temperatures (T1(t), T2(t)), the NC device 70 receives an input of the failure part of the cooling device 40 through the input unit 81 (S204). Furthermore, the NC device 70 stores the detected temperatures (T1(t), T2(t)) and failure part information IN1 indicating a failure part or no failure in the database DB in association with the identification number j. (S206). The processes of S202 to S206 are repeatedly performed. After the information is stored in the database DB, the NC device 70 generates the learning model LM in the RAM 73 by supervised machine learning based on the information stored in the database DB (S208). After generating the learning model LM, the NC device 70 stores the learning model LM as necessary (S210), and ends the learning process. The learning model LM may be stored in any of the ROM 72, the storage device (not shown) in the lathe 1, the storage device 104 of the computer 100, etc. Note that when the NC device 70 performs the above-mentioned failure part determination process (see FIG. 11), the failure part determination process is performed with the learning model LM stored in the RAM 73.

The example shown in FIG. 12 can provide a lathe that improves the accuracy of determining failure parts while suppressing cost increases.
The above-mentioned machine learning section U2 may be realized by the cooperation of the NC device 70 and the computer 100, and the above-mentioned failure detection section U1 may also be realized by the cooperation of the NC device 70 and the computer 100.

Note that it is also possible to use the second detected temperature T2(t) of the second temperature sensor S2 to correct the detected temperature T1(t) of the built-in temperature sensor S1 without using machine learning. For example, by using a correction coefficient a that satisfies 0<a<1 and setting the temperature obtained by subtracting a×T2(t) from the detected temperature T1(t) as the new detected temperature T1(t), as shown in FIG. Cooling device failure determination processing may also be performed. For example, the correction coefficient a may be determined from the relationship between the change ΔT1(t) in the detected temperature T1(t) of the built-in temperature sensor S1 and the change ΔT2(t) in the second detected temperature T2(t) of the second temperature sensor S2. Accordingly, it is possible to determine the failure of the cooling device 40 and to determine the location of the failure. In this case as well, there is no need to separately install a sensor for detecting a failure in each part of the cooling device, so cost increases are suppressed. Further, since the failure part of the cooling device is determined by taking into account the influence of the outside temperature on the temperature detected by the built-in temperature sensor, the accuracy of determining the failure part is improved.

(5) Conclusion:
As explained above, according to the present invention, a technique for a lathe, a lathe system, etc. that can determine a failure of a cooling device without separately installing a sensor for detecting a failure of the cooling device is provided in accordance with various aspects. can be provided. Of course, the above-mentioned basic operation and effect can be obtained even with a technique consisting only of the constituent elements according to the independent claim.
In addition, configurations in which the configurations disclosed in the above-mentioned examples are mutually replaced or the combinations are changed, and configurations in which the configurations disclosed in the publicly known technology and the above-mentioned examples are mutually replaced or the combinations are changed. It is also possible to implement such a configuration. The present invention also includes these configurations.

1... Lathe, 2... Exterior, 3... Temperature sensor,
10... Headstock, 11... Spindle, 11B... Tool spindle, 12... Gripping part,
20...Built-in motor, 23...Jacket, 25...Temperature sensor,
28, 28B...turret,
30... Drive device, 31... Servo motor, 32... Temperature sensor,
40...cooling device,
41... Circulation path, 42... Radiator, 43... Pump, 44, 45... Fan,
70...NC device,
100...computer,
F1...Cooling fluid, F2...Cooling oil, F3...Air,
IN1...Failure part information,
LM...Learning model,
M1...Main shaft motor,
S1...Built-in temperature sensor, S2...Second temperature sensor,
SY1...Lathe system,
TO1...tool, TO2...rotary tool,
U1...failure determination section, U2...machine learning section,
W1...Work.

Claims (7)

a rotatable main shaft,
a main shaft motor having a built-in temperature sensor and rotating the main shaft;
a cooling device that cools the main shaft motor;
When the change in the detected temperature sequentially obtained by the built-in temperature sensor after starting continuous machining of the workpiece becomes below a predetermined threshold value , it is determined that the temperature has become steady; A lathe comprising: a failure determination unit that determines that a cooling device is malfunctioning. The lathe according to claim 1, wherein the failure determination unit determines which part of a plurality of parts included in the cooling device is in failure based on a change in the detected temperature. a rotatable main shaft,
a main shaft motor having a built-in temperature sensor and rotating the main shaft;
A cooling device that cools the main shaft motor , the cooling device having a cooling fluid circulation path that cools the main shaft motor, a pump that circulates the cooling fluid in the circulation path, and a fan that promotes heat dissipation in the circulation path;
A failure in which it is determined that the cooling device is malfunctioning when a change in the detected temperature exceeds an allowable range after it is determined that the detected temperatures sequentially obtained by the built-in temperature sensor have become steady after starting continuous processing of the workpiece. A determination section;
The failure determination unit determines that the pump is malfunctioning when a change in the detected temperature that exceeds the allowable range exceeds a predetermined discrimination criterion, and when the change in the detected temperature does not exceed the discrimination criterion. A lathe in which it is determined that the fan is out of order. a rotatable main shaft,
a main shaft motor having a built-in temperature sensor and rotating the main shaft;
a cooling device that cools the main shaft motor;
A failure in which it is determined that the cooling device is malfunctioning when a change in the detected temperature exceeds an allowable range after it is determined that the detected temperatures sequentially obtained by the built-in temperature sensor have become steady after starting continuous processing of the workpiece. A determination section;
a second temperature sensor that detects temperature influenced by outside temperature;
represents the detected temperature sequentially obtained by the built-in temperature sensor, the second detected temperature sequentially obtained by the second temperature sensor, and a malfunctioning section among the plurality of sections included in the cooling device. Through machine learning based on failure part information, a failure part in the cooling device is determined based on the detected temperature sequentially obtained by the built-in temperature sensor and the second detected temperature sequentially obtained by the second temperature sensor. A lathe, comprising: a machine learning unit that generates a learning model for determining. The learning model is configured to learn about a plurality of parts included in the cooling device by inputting the detected temperatures sequentially obtained by the built-in temperature sensor and the second detected temperatures sequentially obtained by the second temperature sensor. 5. The lathe according to claim 4, wherein the lathe outputs a determination result of a malfunctioning part. A lathe system including a lathe and a computer connected to the lathe, the lathe system comprising:
The lathe is
a rotatable main shaft,
a main shaft motor having a built-in temperature sensor and rotating the main shaft;
A cooling device that cools the main shaft motor,
The lathe system determines that the state has become steady when changes in the detected temperature sequentially obtained by the built-in temperature sensor after starting continuous machining of the workpiece become below a predetermined threshold , and then determines that the change in the detected temperature is permissible . A lathe system comprising a failure determination unit that determines that the cooling device is malfunctioning when a range is exceeded. A lathe system including a lathe and a computer connected to the lathe, the lathe system comprising:
The lathe is
a rotatable main shaft,
a main shaft motor having a built-in temperature sensor and rotating the main shaft;
a cooling device that cools the main shaft motor;
A second temperature sensor that detects the temperature affected by the outside temperature ,
The lathe system includes:
A failure in which it is determined that the cooling device is malfunctioning when a change in the detected temperature exceeds an allowable range after it is determined that the detected temperatures sequentially obtained by the built-in temperature sensor have become steady after starting continuous processing of the workpiece. A determination section;
represents the detected temperature sequentially obtained by the built-in temperature sensor, the second detected temperature sequentially obtained by the second temperature sensor, and a malfunctioning section among the plurality of sections included in the cooling device. Through machine learning based on failure part information, a failure part in the cooling device is determined based on the detected temperature sequentially obtained by the built-in temperature sensor and the second detected temperature sequentially obtained by the second temperature sensor. A lathe system comprising: a machine learning unit that generates a learning model for determining

Images