JP2021053655A - Arithmetic device, arithmetic system, manufacturing system, and arithmetic method - Google Patents

Abstract

To properly set operation conditions that can suppress quality defects.SOLUTION: An arithmetic device 38 is used in a friction-stir-welding device for joining a first member T1 and a second member T2 by bringing a rotated tool 18 into contact with at least one of the first member T1 and the second member T2. The arithmetic device 38 includes: a detection part for detecting a member state which is a state of the first member T1, the second member T2, and at least a part of a joint T3 of the first member T1 and the second member T2, and a tool state which is a state of the tool 18; and a change information generating part for generating change information for changing operation conditions of the friction-stir-welding device from a first value to a second value, on the basis of the member state and the tool state detected by the detection part.SELECTED DRAWING: Figure 1

Description

The present invention relates to arithmetic units, arithmetic systems, manufacturing systems and arithmetic methods.

As a method of joining materials to be joined, friction stir welding (FSW: Friction Stir Welding) is used to soften the materials to be joined by the frictional heat generated by rotating a tool and agitate the part to join the materials to be joined. )It has been known.

U.S. Patent Application Publication No. 2017/02666754

According to the aspect of the present invention, the arithmetic unit is a friction stir welding that joins the first member and the second member by bringing a rotated tool into contact with at least one of the first member and the second member. An arithmetic unit used in the device, which is a member state which is a state of at least a part of the first member, the second member, and a joint portion between the first member and the second member, and a state of the tool. Based on the detection unit that detects the tool state, and the member state and the tool state detected by the detection unit, the operating conditions of the friction stir welding device are set to a first value to a second value. It is provided with a change information generation unit that generates change information for changing to.

According to the aspect of the present invention, the calculation method is a friction stir welding in which a rotated tool is brought into contact with at least one of the first member and the second member to join the first member and the second member. A calculation method used in an apparatus, which is a member state which is a state of at least a part of the first member, the second member, and a joint portion between the first member and the second member, and a state of the tool. The change for changing the operating condition of the friction stir welding device from the first value to the second value based on the detection of the tool state and the detected member state and the tool state. Includes generating information.

It is a schematic diagram of the manufacturing system which concerns on 1st Embodiment. It is a schematic diagram explaining the method of joining by a manufacturing apparatus. It is a schematic diagram explaining the method of joining by a manufacturing apparatus. It is a schematic diagram of the control device which concerns on 1st Embodiment. It is a schematic block diagram of the arithmetic unit which concerns on 1st Embodiment. It is a schematic diagram for demonstrating the imaging region in 1st Embodiment. It is a schematic diagram for demonstrating the imaging region in 1st Embodiment. It is a schematic diagram of the irradiation part and the 1st image pickup part which concerns on 1st Embodiment. It is a graph which shows the relationship between the wavelength of light and the reflectance. It is a flowchart explaining the abnormality determination method of the reflectance which concerns on 1st Embodiment. It is a schematic diagram of the 2nd imaging unit which concerns on 1st Embodiment. It is a schematic diagram for demonstrating an opening, a burr, and a processing path. It is a flowchart explaining the detection method of the opening. It is a flowchart explaining the method of detecting a burr. It is a schematic diagram explaining the insertion amount of a tool. It is a schematic diagram explaining the insertion amount of a tool. It is a schematic diagram explaining the detection of the shape of a tool. It is a flowchart explaining the processing flow of the arithmetic unit which concerns on 1st Embodiment. It is a flowchart explaining the generation of change information. It is a flowchart explaining the generation of change information. It is a flowchart explaining the generation of change information. It is a flowchart explaining the generation of change information. It is a schematic diagram explaining the modification of 1st Embodiment. It is a schematic diagram explaining the modification of 1st Embodiment. It is a schematic diagram explaining the modification of 1st Embodiment. It is a schematic diagram explaining the modification of 1st Embodiment. It is a schematic diagram explaining the modification of 1st Embodiment. It is a schematic diagram explaining the modification of 1st Embodiment. It is a schematic diagram of the manufacturing system which concerns on 2nd Embodiment. It is a schematic diagram explaining the irradiation part and the image pickup part which concerns on 2nd Embodiment. It is a schematic diagram of the irradiation part and the 1st image pickup part which concerns on 2nd Embodiment. It is a graph for demonstrating the intensity of light from a detection area. It is a graph for demonstrating the intensity of light from a detection area. It is a flowchart explaining the method of detecting a hole | hole which concerns on 2nd Embodiment. It is a schematic diagram explaining the heat transfer in a member. It is a graph for demonstrating the intensity of light from a detection area. It is a graph for demonstrating the intensity of light from a detection area. It is a flowchart explaining the method of detecting the thickness of the compound which concerns on 2nd Embodiment. It is a figure explaining the detection of light by the 1st imaging unit which concerns on 2nd Embodiment. It is a flowchart explaining the processing flow of the arithmetic unit which concerns on 2nd Embodiment. It is a flowchart explaining the generation of change information. It is a flowchart explaining the generation of change information. It is a schematic diagram of the manufacturing system which concerns on 3rd Embodiment. It is a schematic diagram explaining the method of joining by the manufacturing apparatus which concerns on 3rd Embodiment. It is a figure explaining the tool state which concerns on 3rd Embodiment. It is a flowchart explaining the processing flow of the arithmetic unit which concerns on 3rd Embodiment. It is a flowchart explaining the generation of change information which concerns on 3rd Embodiment. It is a flowchart explaining the generation of change information which concerns on 3rd Embodiment. It is a flowchart explaining the generation of change information which concerns on 3rd Embodiment. It is a schematic diagram which shows the structure of the system which has a manufacturing system. It is a block block diagram of a system. It is a flowchart which showed the flow of processing by a system.

Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. In addition, the components in the following embodiments include those that can be easily assumed by those skilled in the art, those that are substantially the same, that is, those in a so-called equal range. Further, the components disclosed in the following embodiments can be appropriately combined.

In the following description, the XYZ Cartesian coordinate system is set, and the positional relationship of each part will be described with reference to the XYZ Cartesian coordinate system. A predetermined direction in the horizontal plane is the X-axis direction, a direction orthogonal to the X-axis direction in the horizontal plane is the Y-axis direction, and a direction orthogonal to each of the X-axis direction and the Y-axis direction (that is, the vertical direction) is the Z-axis direction. Further, the rotation (tilt) directions around the X-axis, the Y-axis, and the Z-axis are set to the θX, θY, and θZ directions, respectively. Further, the direction upward in the vertical direction in the Z-axis direction is defined as the + Z-axis direction, and the direction downward in the vertical direction in the Z-axis direction is defined as the −Z-axis direction.

(First Embodiment)
FIG. 1 is a schematic view of a manufacturing system according to the first embodiment. The manufacturing system 1 according to the first embodiment includes a manufacturing apparatus 10 and a calculation system 12. The manufacturing apparatus 10 is an apparatus for manufacturing a bonded body of the first member T1 and the second member T2 by joining the first member T1 and the second member T2, which are materials to be joined. The manufacturing apparatus 10 is a friction stir welding apparatus that joins the first member T1 and the second member T2 by bringing a rotated tool 18 into contact with at least one of the first member T1 and the second member T2. .. Hereinafter, the joint portion between the first member T1 and the second member T2 by the manufacturing apparatus 10 will be referred to as a joint portion T3. When the first member T1, the second member T2, and the joint portion T3 are not distinguished, they are described as the member T. That is, the member T refers to at least one of the first member T1, the second member T2, and the joint portion T3. The joint portion T3 can be said to be a region where a part of the first member T1 and a part of the second member T2 are mixed by friction stir welding of the manufacturing apparatus 10, and the first member T1 and the second member T2 It can be said that the area where the tool 18 contacts and passes through at least one of them. If friction stir welding is inappropriate, defects such as openings and vacancies, which will be described later, will occur in the joint portion T3.

The first member T1 and the second member T2 are metal materials in the present embodiment. In the present embodiment, the first member T1 and the second member T2 are different materials from each other. In the example of this embodiment, the first member T1 is aluminum and the second member T2 is iron. However, the first member T1 and the second member T2 may be any material, and may not be, for example, a metal material, or may be the same material as each other.

(manufacturing device)
As shown in FIG. 1, the manufacturing apparatus 10 includes a base portion 13, a tool stage 14, a tool holding portion 16, a tool 18, a base portion 24, a first work stage 26, and a second work stage 28. , The control device 30 is provided. The base 13 houses various devices of the manufacturing apparatus 10. In the example of FIG. 1, the control device 30 and the arithmetic unit 38 described later are housed in the base 13, but the positions of the control device 30 and the arithmetic unit 38 are not limited to the base 13 and may be arbitrary. .. Further, in the present embodiment, the control device 30 and the arithmetic unit 38 are separate devices, but the control device 30 and the arithmetic unit 38 may be one device.

The tool stage 14 is connected to the base 13 via the slider 14A. The tool holding portion 16 is connected to the tool stage 14 and projects from the tool stage 14 in the −Z axis direction. The tool holding portion 16 has a tool 18 connected to an end portion in the −Z axis direction, and holds the tool 18 at the end portion in the −Z axis direction. The tool 18 includes a shoulder 20 and a probe 22. The shoulder 20 is connected to the end of the tool holding portion 16 in the −Z axis direction. The probe 22 is provided at the end portion 20a of the shoulder 20 on the opposite side (−Z axis direction side) of the tool holding portion 16. The outer diameter of the probe 22 is smaller than the outer diameter of the shoulder 20, and a groove (in other words, a thread-like unevenness) is formed on the outer peripheral surface. The tool 18, that is, the shoulder 20 and the probe 22, rotate about the central axis AX of the tool 18.

The base portion 24 is a pedestal on which the first work stage 26 and the second work stage 28 are placed. The base portion 24 has a fixed position with respect to the base portion 13 and is located on the −Z axis direction side of the tool 18. The first work stage 26 is provided on the upper part of the base portion 24, and the second work stage 28 is placed on the upper part. The second work stage 28 is a table on which the member T is placed. The member T is placed on the second work stage 28 so that the position is fixed with respect to the second work stage 28. The upper part here refers to the surface on the + Z axis direction side.

The tool stage 14, the first work stage 26, and the second work stage 28 are moving mechanisms that change the relative positions of the member T and the tool 18. The relative position is the position of the tool 18 with respect to the member T, in other words, the relative position of the member T and the tool 18. The relative position changes when the position (coordinates) of the tool 18 changes in at least one of the X-axis direction, the Y-axis direction, and the Z-axis direction with respect to the position (coordinates) of the member T.

In the present embodiment, the tool stage 14 is movably attached to the base 13 in the Z-axis direction, for example, via a slider 14A extending along the Z-axis direction. By moving the tool stage 14 in the Z-axis direction with respect to the base portion 13, the tool 18 attached via the tool holding portion 16 is moved in the Z-axis direction. The tool stage 14 changes the relative position of the member T and the tool 18 along the Z-axis direction by moving the tool 18 in the Z-axis direction. The first work stage 26 is movably attached to the base portion 24 in the Y-axis direction, for example, via a slider 24A extending along the Y-axis direction. By moving the first work stage 26 in the Y-axis direction with respect to the base portion 24, the member T attached via the second work stage 28 is moved in the Y-axis direction. The first work stage 26 changes the relative position of the member T and the tool 18 along the Y-axis direction by moving the member T in the Y-axis direction. The second work stage 28 is movably attached to the first work stage 26 in the X-axis direction, for example, via a slider 26A extending along the X-axis direction. The second work stage 28 moves the member T attached to the upper portion in the X-axis direction by moving with respect to the first work stage 26 in the X-axis direction. The second work stage 28 changes the relative position of the member T and the tool 18 along the X-axis direction by moving the member T in the X-axis direction.

As described above, in the present embodiment, the tool 18 is moved in the Z-axis direction by the tool stage 14, and the member T is moved in the X-axis direction and the Y-axis direction by the first work stage 26 and the second work stage 28. , The relative position of the member T and the tool 18 is changed. However, if the manufacturing apparatus 10 is configured so that the relative positions of the member T and the tool 18 can be changed, the tool 18 is moved in the Z-axis direction and the member T is moved in the X-axis direction and the Y-axis direction. Not limited to. For example, the manufacturing apparatus 10 moves the member T in the X-axis direction, the Y-axis direction, and the Z-axis direction, or moves the tool 18 in the X-axis direction, the Y-axis direction, and the Z-axis direction. The relative positions of the member T and the tool 18 may be changed.

Further, the manufacturing apparatus 10 changes the relative posture between the tool 18 and the member T. The relative posture is the relative angle of the tool 18 with respect to the member T, in other words, the relative posture (relative angle) between the tool 18 and the member T. The relative posture changes when the posture of the tool 18 changes with respect to the posture of the member T in at least one of the θX-axis direction, the θY-axis direction, and the θZ-axis direction. In the present embodiment, the manufacturing apparatus 10 changes the relative posture of the tool 18 and the member T in the θX axis direction. For example, the tool 18 changes its relative posture with respect to the tool holding portion 16, that is, with respect to the member T in the θX axis direction by rotating around the rotation axis AX1 parallel to the X-axis direction with respect to the tool holding portion 16. Let me. By rotating at least one of the first work stage 26 or the second work stage 28 about a rotation axis parallel to the X-axis direction, the relative postures of the tool 18 and the member T in the θX-axis direction are changed. May be good. Further, the tool 18 rotates about the central axis AX of the tool 18 as described above to change the relative posture with respect to the tool holding portion 16, that is, with respect to the member T in the θZ axis direction. Further, at least one of the tool 18 and the member T may be rotated in the θY axis direction to change the relative posture of the tool 18 and the member T in the θY axis direction.

The positions of the irradiation unit 32A and the imaging unit 34 of the arithmetic system 12, which will be described later, are fixed with respect to the tool holding unit 16. Therefore, the positions of the irradiation unit 32A and the imaging unit 34 are fixed relative to the tool 18. Therefore, the irradiation unit 32A and the imaging unit 34 move integrally with the tool 18 in the Z-axis direction, and the relative positions with respect to the member T move in the Z-axis direction. Further, when the member T moves in the X-axis direction and the Y-axis direction, the relative positions of the irradiation unit 32A and the image pickup unit 34 with respect to the member T move in the X-axis direction and the Y-axis direction. Further, the irradiation unit 32A and the imaging unit 34 are not fixed in relative posture with respect to the tool 18, and do not rotate in the θX direction integrally with the tool 18. Therefore, when the tool 18 rotates in the θX direction, the relative postures of the irradiation unit 32A and the imaging unit 34 and the tool 18 change. Further, in the present embodiment, the irradiation unit 32A and the imaging unit 34 do not rotate in the θZ direction integrally with the tool 18, but may be configured to rotate in the θZ axis direction integrally with the tool 18, for example. Further, the irradiation unit 32A may be movable so that at least one of the relative position and the relative posture of the irradiation unit 32A and the tool 18 can be changed, or the image pickup unit 34 may be movable to make the image pickup unit 34 movable. At least one of the relative position and the relative posture of the 34 and the tool 18 may be changed.

2 and 3 are schematic views illustrating a method of joining by a manufacturing apparatus. As shown in FIG. 2, when the first member T1 and the second member T2 are joined by the manufacturing apparatus 10, the surface T1A of the first member T1 and the surface T2A of the second member T2 are on the tool 18 side. The first member T1 and the second member T2 are placed on the second work stage 28 so that the side surface T1B of the first member T1 and the side surface T2B of the second member T2 abut each other. Deploy. The abutting here means that the side surface T1B and the side surface T2B face each other and come into contact with each other. The manufacturing apparatus 10 moves the tool 18 in the Z-axis direction while rotating the tool 18 in the θZ direction, and inserts the probe 22 into at least one of the first member T1 and the second member T2. The probe 22 has at least one of the first member T1 and the second member T2 until the end portion 20a of the shoulder 20 contacts at least one of the surface T1A of the first member T1 and the surface T2A of the second member T2. Will be inserted into. The probe 22 is inserted into at least one of the first member T1 and the second member T2 and rotates to transfer frictional heat due to rotation to the first member T1 and the second member T2. Further, at this time, the shoulder 20 rotates with the end portion 20a in contact with at least one of the surface T1A of the first member T1 and the surface T2A of the second member T2, so that the first member T1 and the second member Friction heat due to rotation is transmitted to T2. The first member T1 and the second member T2 are softened by frictional heat and agitated by the rotation of the probe 22 and the shoulder 20, so that they are mixed and joined to each other.

As shown in FIG. 3, in the manufacturing apparatus 10, the tool 18 is inserted into at least one of the first member T1 and the second member T2 while the probe 22 is rotating in the θZ direction. Is moved relative to the feed direction. In the present embodiment, since the first work stage 26 is moved in the Y-axis direction, the feed direction is the Y-axis direction. The probe 22 rotates in the θZ direction while being inserted into at least one of the first member T1 and the second member T2, and in the Y-axis direction with respect to at least one of the first member T1 and the second member T2. Move relative to each other. As a result, the probe 22 frictionally stirs the first member T1 and the second member T2 at each position along the feeding direction (Y-axis direction) to join the first member T1 and the second member T2. The shoulder 20 is integrally with the probe 22 in the feed direction while rotating in the θZ direction in a state where the end portion 20a is in contact with at least one of the surface T1A of the first member T1 and the surface T2A of the second member T2. Move relative to each other. The feed direction is not limited to the Y-axis direction and may be arbitrary, for example, the X-axis direction.

The first member T1 and the second member T2 are joined in this way. In the present embodiment, as shown in FIGS. 2 and 3, the joint portion T3 is a region of the member T in which the end portion 20a of the shoulder 20 and the probe 22 are in contact with the member T when viewed from the Z-axis direction. Further, a portion of the entire member T that overlaps the end portion 20a of the shoulder 20 and the region after the probe 22 has passed when viewed from the Z-axis direction can be said to be the joint portion T3. It can be said that the joint portion T3 is a region where the first member T1 and the second member T2 are mixed with each other by friction stir welding by the tool 18. In the following, the surface of the member T on the tool 18 side will be referred to as the surface TA. In the description of FIGS. 2 and 3, the probe 22 is inserted into both the first member T1 and the second member T2, but is inserted into at least one of the first member T1 and the second member T2. May be good.

In FIGS. 2 and 3, so-called butt joints in which the side surfaces T1B and T2B are butt-joined are taken as an example, and butt-joining will be described below as an example, but the manufacturing apparatus 10 has a butt joint. It is not limited to performing joining. For example, the manufacturing apparatus 10 may perform so-called superposition joining. In the lap joining, for example, the first member T1 and the second member T2 are overlapped by bringing the back surface T1C opposite to the surface T1A of the first member T1 into contact with the surface T2A of the second member T2. Then, friction stir welding is performed by inserting the probe 22 into a portion where the first member T1 and the second member T2 overlap while rotating the probe 22.

FIG. 4 is a schematic view of the control device according to the first embodiment. The control device 30 is a device that controls each part of the manufacturing device 10 to join the first member T1 and the second member T2 to the manufacturing device 10. The control device 30 includes a drive control unit 30A and a storage unit 30B. The drive control unit 30A is an arithmetic unit, that is, a CPU (Central Processing Unit). The storage unit 30B is a memory that stores the control contents of the drive control unit 30A, program information, and the like. For example, a RAM (Random Access Memory), a ROM (Read Only Memory), a flash memory (Flash Memory), and the like are used. Includes at least one of the external storage devices. The control device 30 may further include at least one input unit that receives an operation from the user and an output unit that outputs information. The input unit is, for example, a mouse, a keyboard, a touch panel, or the like, and the output unit is, for example, a display device.

The drive control unit 30A sets the operating conditions of the manufacturing apparatus 10, and drives the manufacturing apparatus 10 under the set operating conditions. Examples of the operating conditions include at least one of the rotation speed of the tool 18, the feed rate of the tool 18, the insertion amount of the tool 18, and the machining path in the member T by the tool 18. The rotation speed of the tool 18 is the rotation speed per unit time of rotation (in the θZ direction) about the central axis AX of the tool 18. The drive control unit 30A rotates the tool 18 in the θZ direction at the set rotation speed of the tool 18. The feed rate of the tool 18 is the relative moving speed of the tool 18 with respect to the member T in the feed direction (Y-axis direction in this embodiment). The drive control unit 30A changes the relative position of the tool 18 with respect to the member T in the feed direction (here, the Y-axis direction) at the set feed rate. The insertion amount of the tool 18 is the insertion amount of the tool 18 with respect to the member T, in other words, the length of the portion inserted into the member T of the tool 18. The drive control unit 30A adjusts the relative position of the tool 18 with respect to the member T in the Z-axis direction to set the insertion amount of the tool 18 as the insertion amount of the tool 18. The details of the insertion amount of the tool 18 will be described later. The machining path is a movement path of the tool 18 with respect to the member T in a state where the probe 22 is inserted into the member T while rotating, and more specifically, in a direction orthogonal to the feed direction (here, the X-axis direction). This is the relative position of the tool 18 with respect to the member T. In other words, the machining path is a machining path in the member T by the tool 18, and is a machining position with respect to the member T in a direction orthogonal to the feed direction. The drive control unit 30A changes the relative position of the tool 18 in the direction orthogonal to the feed direction with respect to the member T so as to obtain the set machining path. It is preferable that the drive control unit 30A sets all of the rotation speed of the tool 18, the feed rate of the tool 18, the insertion amount of the tool 18, and the machining path in the member T by the tool 18 as operating conditions. Further, the drive control unit 30A may set the relative position of the tool 18 and the member T in the θX axis direction as an operating condition.

The manufacturing apparatus 10 is driven under the operating conditions set by the control apparatus 30 to join the first member T1 and the second member T2. In such a manufacturing apparatus 10, poor joining of the first member T1 and the second member T2 may lead to poor quality of the member T. Quality defects may be suppressed by resetting the operating conditions of the manufacturing apparatus 10. In the present embodiment, the arithmetic system 12 determines the member state of the member T (at least one of the first member T1, the second member T2, and the joint T3) and the tool state of the tool 18. By detecting, change information indicating how to change the operating conditions in order to suppress quality defects is generated. The manufacturing apparatus 10 can appropriately suppress the quality defect of the member T by detecting the quality abnormality of the member T and resetting the operating conditions according to the change information. The poor quality of the member T means that the member T cannot be used as it is and the quality of the member T is deteriorated to the extent that it needs to be discarded. The quality abnormality of the member T refers to a state in which the degree of deterioration of the quality of the member T is slight to the extent that it can be used as it is without disposal. In other words, the quality abnormality of the member T is a state in which the quality defect has not been reached, but the quality defect tends to occur. Hereinafter, the arithmetic system 12 will be described.

(Computational system)
As shown in FIG. 1, the arithmetic system 12 includes an irradiation unit 32A, a first imaging unit 34A, a second imaging unit 34B, and an arithmetic unit 38. Hereinafter, when the first imaging unit 34A and the second imaging unit 34B are not distinguished, they are referred to as an imaging unit 34. Before explaining the irradiation unit 32A, the first imaging unit 34A, and the second imaging unit 34B, the arithmetic unit 38 will be described. FIG. 5 is a schematic block diagram of the arithmetic unit according to the first embodiment. As shown in FIG. 5, the arithmetic unit 38 includes an arithmetic unit 40 and a storage unit 42. The calculation unit 40 is a CPU (Central Processing Unit). The storage unit 42 is a memory that stores the calculation contents of the calculation unit 40, program information, and the like. For example, an external such as a RAM (Random Access Memory), a ROM (Read Only Memory), and a flash memory (Flash Memory). Includes at least one of the storage devices. The arithmetic unit 38 may further include at least one input unit that receives an operation from the user and an output unit that outputs information. The input unit is, for example, a mouse, a keyboard, a touch panel, or the like, and the output unit is, for example, a display device.

The calculation unit 40 includes a device control unit 50, a detection unit 52, a joint state determination unit 54, a change information generation unit 56, and a change information output unit 58. The device control unit 50, the detection unit 52, the junction state determination unit 54, the change information generation unit 56, and the change information output unit 58 are such that the calculation unit 40 reads out the software (program) stored in the storage unit 42. This is realized by executing the process described later.

The device control unit 50 controls the operations of the irradiation unit 32A, the first imaging unit 34A, and the second imaging unit 34B. The device control unit 50 irradiates the irradiation unit 32A with light, and causes the first imaging unit 34A and the second imaging unit 34B to take an image. Details of the irradiation of light by the irradiation unit 32A and the imaging by the first imaging unit 34A and the second imaging unit 34B will be described later.

The detection unit 52 detects the member state, which is the state of the member T, and the tool state, which is the state of the tool 18. The detection unit 52 detects the member state and the tool state based on the image captured by the image pickup unit 34. The device control unit 50 causes the image pickup unit 34 to take an image at predetermined time intervals during the joining between the first member T1 and the second member T2, and the detection unit 52 connects the first member T1 and the second member T2. During joining, each time the imaging unit 34 captures an image, the member state and the tool state are detected. As a member state, the detection unit 52 includes the light reflectance of the surface TA of the member T, the opening formed in the joint portion T3, the burr formed around the joint portion T3, and the processing in the member T by the tool 18. Detect the path. The machining path is included in the member state because it is the position where the member T is machined in the direction orthogonal to the feed direction. It can be said that the detection unit 52 according to the first embodiment detects the state of the outside (surface TA) of the member T as the member state. Further, the detection unit 52 detects the insertion amount of the tool 18 and the shape of the tool as the tool state. Hereinafter, each detection of the member state and the tool state will be described.

First, the detection of the light reflectance of the surface TA of the member T will be described. The detection unit 52 detects the light reflectance of the surface TA of the member T based on the image of the member T on which the light LAB projected from the irradiation unit 32A is captured by the first imaging unit 34A. Furthermore, the detection unit 52 is based on the image of the surface T3A of the joint portion T3 on which the light LAB projected from the irradiation unit 32A is imaged by the first imaging unit 34A, and the light of the surface T3A of the joint portion T3. Detect reflectance. Hereinafter, the irradiation unit 32A and the first imaging unit 34A will be described.

6 and 7 are schematic views for explaining the imaging region in the first embodiment. As shown in FIG. 6, the irradiation unit 32A irradiates the region R0 with light LAB. The region R0 refers to a range in which the irradiation unit 32A irradiates the light LAB. The region R0 overlaps the surface TA of the member T, and in the present embodiment, as shown in FIG. 7, overlaps at least a part of the surface T3A of the joint portion T3. That is, the irradiation unit 32A projects the light LAB as the projected light on at least a part of the surface T3A of the joint portion T3. However, the region R0 is not limited to the example of FIG. 7, and for example, at least a part of the surface T1A of the first member T1, at least a part of the surface T2A of the second member T2, and the joint portion T3. It may overlap at least a part of the surface T3A and the entire surface area of the tool 18. Hereinafter, the region of the surface TA of the member T that overlaps with the region R0 will be referred to as an irradiation region AR0. It can be said that the irradiation unit 32A projects the light LAB as the projected light on the irradiation region AR0.

As shown in FIG. 6, the first imaging unit 34A takes an image within the range of the imaging region R1. The imaging region R1 is an imaging region of the first imaging unit 34A, in other words, a range imaged by the first imaging unit 34A. The imaging region R1 overlaps the surface TA of the member T, and in the present embodiment, as shown in FIG. 7, overlaps at least a part of the surface T3A of the joint portion T3. When the region overlapping the imaging region R1 on the surface TA of the member T is described as the detection region AR1, it can be said that the first imaging unit 34A captures an image of the detection region AR1 of the member T on which the optical LAB is projected. The detection area AR1 and the irradiation area AR0 overlap at least in a part. Therefore, it can be said that the first imaging unit 34A captures an image of the detection region AR1 that overlaps the irradiation region AR0 on which the light LAB projected by the irradiation unit 32A is projected, and in the present embodiment, the first imaging unit 34A It can be said that 34A captures the surface T3A image of the joint portion T3 on which the light LAB projected from the irradiation portion 32A is projected. It is sufficient that at least a part of the irradiation area AR0 and the imaging area R1 overlap, and at least a part of the imaging area R1 and the detection area AR1 may overlap. That is, the detection region AR1 and the irradiation region AR0, the irradiation region AR0 and the imaging region R1, and the imaging region R1 and the detection region AR1 may be completely the same region, or only a part thereof may overlap.

FIG. 8 is a schematic view of the irradiation unit and the first imaging unit according to the first embodiment. As shown in FIG. 8, the irradiation unit 32A includes light source units 60A and 60B, lenses 62A and 62B, a photosynthesis unit 64, a light diffusion unit 66, and a lens 68. The light source unit 60A is a light source that irradiates light LA, and the light source unit 60B is a light source that irradiates light LB. The light source units 60A and 60B are, for example, LEDs (Light Emitting Diodes). The light LA is the light of the first wavelength, the light LB is the light of the second wavelength, and the first wavelength and the second wavelength are different from each other (do not overlap). That is, the light LA and the light LB are light having different wavelengths from each other. In the present embodiment, the light LA and the light LB are light having a wavelength of visible light. In the example of this embodiment, the first wavelength of the light LA is, for example, 600 nm. The second wavelength of the light LB is, for example, 400 nm. However, the optical LA and the optical LB are not limited to the above wavelengths as long as they are light having different wavelengths (wavelengths do not overlap). Further, the light LA and the light LB are not limited to light having a wavelength of visible light. The light source units 60A and 60B do not have to be LEDs, and other existing light sources may be used. Further, it is not necessary to use a plurality of light source units (60A, 60B), and one light source that emits light in a wavelength band including the first wavelength and the second wavelength may be used as the light source unit. In this case, a light source that emits wideband light such as a lamp (mercury lamp or xenon lamp) or SLD (super luminescence diode) is used, and an existing bandpass filter is used to generate light emitted from the mercury lamp or xenon lamp. Light of the first wavelength and light of the second wavelength (that is, light LAB described later) may be selected. When one light source is used as the light source unit, the photosynthesis unit 64 is unnecessary, and a lens for collimating the light emitted from the light source may be provided.

The light LA emitted from the light source unit 60A is incident on the lens 62A, collimated by the lens 62A, and is incident on the photosynthesis unit 64 from the lens 62A. The light LB emitted from the light source unit 60B is incident on the lens 62B, collimated by the lens 62B, and is incident on the photosynthesis unit 64 from the lens 62B.

In the photosynthesis unit 64, the light LA that has passed through the lens 62A from the light source unit 60A and the light LB that has passed through the lens 62B from the light source unit 60B are incident. The photosynthesis unit 64 is an optical element that aligns the traveling directions of the incident light LA and the light LB and superimposes the light LA and the light LB. The light LA and the light LB incident on the photosynthesis unit 64 are superposed in the photosynthesis unit 64 and emitted from the photosynthesis unit 64 as the photosynthesis unit 64. That is, the light LAB is light in which light LA and light LB are superposed, and is light including a first wavelength and a second wavelength. In the present embodiment, the photosynthetic unit 64 is a dichroic mirror, which reflects light of one wavelength of light LA and light LB and transmits light of the other wavelength. In the example of this embodiment, the photosynthetic unit 64 reflects the light LB of the second wavelength and transmits the light LA of the first wavelength. In the photosynthesis unit 64, light LA is incident from one surface 64a, passes through the incident light LA, and is emitted from the other surface 64b. Further, in the photosynthesis unit 64, the light LB is incident on the surface 64b and the light LB is reflected on the surface 64b. The light LB reflected by the surface 64b is superimposed on the light LA emitted from the surface 64b and emitted from the photosynthesis unit 64 as an optical LAB. The photosynthesis unit 64 is not limited to a dichroic mirror as long as it is an optical element in which an optical LA and an optical LB are superposed to form an optical LAB, and may be, for example, a half mirror. In this case, for example, the photosynthesis unit 64 transmits a part of the light LA from the light source unit 60A, reflects a part of the light LA from the light source unit 60B, and reflects the transmitted light LA. The light that is superimposed on the light LB is emitted as an optical LAB.

The light diffusing unit 66 is arranged on the traveling direction side of the light LAB emitted from the photosynthetic unit 64 with respect to the photosynthetic unit 64. The light diffusing unit 66 is an optical element that equalizes the illuminance of the light LAB in the plane of the irradiation region AR0. In the present embodiment, the light diffusing unit 66 is a fly-eye lens in which a plurality of lenses are arranged on the surface. In the light diffusing unit 66, light LAB is incident on each of the plurality of lenses, and light is emitted from each of the plurality of lenses. The lens 68 collimates and emits the light LAB from the light diffusing unit 66. The light LAB emitted from the lens 68 irradiates the irradiation region AR0 of the surface T3A of the junction T3. More specifically, the light LAB emitted from each of the plurality of lenses of the light diffusing unit 66 is incident on the lens 68, and the respective light LABs are collimated. The respective light LABs collimated by the lens 68 are superposed on each other and irradiated to the irradiation region AR0. That is, the light LAB emitted from the lens 68 is projected as projected light on the irradiation region AR0. The light diffusing unit 66 is not an essential configuration.

The first image pickup unit 34A includes a lens 70, a first light separation unit 72, a second light separation unit 74, a first light reflection unit 76, a second light reflection unit 78, a lens 79, and an image sensor 80. And. The light LAB emitted from the lens 68 of the irradiation unit 32A and irradiated to the irradiation region AR0 is reflected in the irradiation region AR0, and at least a part thereof is incident on the lens 70 as the light L1AB. Specifically, among the light LABs emitted from the lens 68 of the irradiation unit 32A and irradiated to the irradiation region AR0, the light LAB reflected by the detection region AR1 is incident on the lens 70 as the light L1AB. The wavelength of the light L1AB is the same as that of the light LAB, and the light L1AB includes light having a first wavelength and light having a second wavelength.

The first light separation unit 72 is an optical element that separates the light L1AB emitted from the lens 70 into the light L1A and the light L1B. The light L1A is the light of the first wavelength, and the light L1B is the light of the second wavelength. In the present embodiment, the first light separation unit 72 is a dichroic mirror, and reflects light of one wavelength of the light L1A of the first wavelength and the light L1B of the second wavelength contained in the light L1AB and reflects the light of the other. Transmits light of wavelength. In the example of the present embodiment, the first light separation unit 72 reflects the light L1A of the first wavelength and transmits the light L1B of the second wavelength. The first light separation unit 72 receives the light L1AB from the surface 72a, transmits the light L1B contained in the incident light L1AB, emits the light L1B from the other surface 72b, and reflects the light L1A contained in the incident light L1AB. To do.

The light L1B emitted from the first light separation unit 72 is incident on the surface 74a of the second light separation unit 74.

The light L1A reflected by the first light separation unit 72 is incident on the first light reflection unit 76 and is reflected by the first light reflection unit 76. The light L1A reflected by the first light reflecting unit 76 enters the second light reflecting unit 78 and is reflected by the second light reflecting unit 78. The light L1A reflected by the second light reflecting unit 78 is incident on the surface 74b of the second light separating unit 74.

The second light separation unit 74 emits the incident light L1A and the light L1B in different directions. In the present embodiment, the second light separation unit 74 is a dichroic mirror, and reflects one of the light L1A of the first wavelength and the light L1B of the second wavelength and transmits the other. In the example of the present embodiment, the second light separation unit 74 reflects the light L1A and transmits the light L1B. In the second light separation unit 74, the light L1B transmitted through the first light separation unit 72 is incident on the surface 74a and emitted from the surface 74b opposite to the surface 74a. Further, in the second light separation unit 74, the light L1A reflected from the second light reflection unit 78 is reflected on the surface 74b in a direction different from that of the light L1B. In the present embodiment, the surface (reflection surface) of the second light reflecting portion 78 and the surface 74a of the second light separating portion 74 are set to be non-parallel. Specifically, the surface (reflecting surface) of the second light reflecting portion 78 rotates counterclockwise on the Y axis with respect to the surface 74a of the second light separating portion 74. Further, the surface 72a of the first light separating portion 72 and the surface (reflecting surface) of the first light reflecting portion 76 are set in parallel. Therefore, the light L1A reflected by the surface 74b of the second light separation unit 74 and the light L1B transmitted through the second light separation unit 74 travel in different directions. However, a method of advancing the light L1A reflected by the surface 74b of the second light separating portion 74 and the light L1B transmitted through the second light separating portion 74 in different directions is the surface (reflecting surface) of the second light reflecting portion 78. And the surface 74a of the second light separation unit 74 are not limited to being set non-parallel, and are optional. For example, the surface 72a of the first light separating portion 72 and the surface (reflecting surface) of the first light reflecting portion 76 may be set to be non-parallel.

The light L1A and the light L1B emitted from the second light separation unit 74 are focused by the lens 79 and incident on the image sensor 80. The lens that collects the light L1A and the lens that collects the light L1B may be separated. In the image pickup device 80, a plurality of pixels are arranged in a matrix on the light receiving surface. The pixel of the image sensor 80 is a light receiving element such as a photodiode. Specifically, the image sensor 80 is a CCD (Charged Coupled Device), a CMOS (Complementary Metal Oxide Semiconductor), or the like. The light receiving surface of the image sensor 80 has a conjugate relationship with the surface T3A (detection region AR1) of the junction T3 on which the light LAB emitted from the irradiation unit 32A is projected. Here, since the light L1A and the light L1B from the second light separation unit 74 have different traveling directions, they are incident on the image sensor 80 in a state of being separated from each other without being overlapped. In the example of FIG. 8, the light L1A is incident on the light receiving surface 80A which is a part of the light receiving surface of the image sensor 80, and the light L1B is the light receiving surface 80A of the light receiving surface of the image sensor 80. It is incident on the light receiving surface 80B, which is a different region. The image sensor 80 receives an image of the surface T3A (detection region AR1) of the junction T3 by the light L1A of the first wavelength by receiving the incident light L1A at the pixels provided on the light receiving surface 80A. Further, the image sensor 80 captures an image of the surface T3A (detection region AR1) of the junction T3 by the light L1B of the second wavelength by receiving the incident light L1B at the pixels provided on the light receiving surface 80B. .. The first light separation unit 72, the second light separation unit 74, the first light reflection unit 76, and the second light reflection unit 78 are light L1AB containing light of the first wavelength and light of the second wavelength. Since the light L1A having the first wavelength and the light L1B having the second wavelength are separated from each other, it can be said to be a wavelength separation unit. This wavelength separation unit spatially separates the light L1A of the first wavelength and the light L1B of the second wavelength into the light receiving surface 80A and the light receiving surface 80B of the image sensor 80.

The light L1A incident on the light receiving surface 80A is the light of the first wavelength among the light LABs irradiated from the irradiation unit 32A and projected onto the detection region AR1. On the other hand, the light L1B incident on the light receiving surface 80A is the light of the second wavelength among the light LABs irradiated from the irradiation unit 32A and projected onto the detection region AR1. That is, the image sensor 80 images the image of the detection region AR1 by the light of the first wavelength reflected by the detection region AR1 on the light receiving surface 80A, and the image of the detection region AR1 by the light of the second wavelength reflected by the detection region AR1. Is imaged on the light receiving surface 80B. Therefore, it can be said that the image sensor 80 captures one image of light reflected in the same region (detection region AR1) and has different wavelengths on the light receiving surface 80A and the other on the light receiving surface 80B. In the present embodiment, the light L1A and the light L1B are incident on different regions on the light receiving surface of one image sensor 80, but a plurality of image sensors are provided and the light L1A is applied to one image sensor. The light L1B may be incident (imaging the image of the detection region AR1 by the light L1A) and the light L1B may be incident (imaging the image of the detection region AR1 by the light L1B) on another image sensor.

The detection unit 52 includes an image data PA which is data of an image of the detection region AR1 by the light L1A imaged by the image sensor 80 on the light receiving surface 80A and an image of the detection area AR1 by the light L1B imaged by the image sensor 80 on the light receiving surface 80B. The image data PB, which is the data of the above, is acquired. The image data here refers to the data of the image captured by the image sensor 80. Specifically, the image data PA refers to the data of the brightness value for each pixel on the light receiving surface 80A, and can be said to be the data of the intensity value of the reflected light L1A for each position (coordinates) in the detection area AR1. .. The detection unit 52 calculates the reflectance of the light of the first wavelength for each position in the detection region AR1 based on the image data PA, that is, the intensity of the light L1A for each position in the detection region AR1. For example, the detection unit 52 calculates in advance the relationship between the luminance value and the reflectance of the pixels of the image sensor 80, and substitutes the luminance value for each pixel of the image data PA into the relationship between the luminance value and the reflectance. By doing so, the reflectance of the light of the first wavelength for each position in the detection region AR1 is calculated. The relationship between the luminance value of the pixel of the image sensor 80 and the reflectance is, for example, the luminance value of the pixel when the light is irradiated from the irradiation unit 32A to the mirror in advance and the light reflected by the mirror is detected by the image sensor 80. The reflectance for each pixel value may be calculated and obtained with reference to. Further, the image data PB refers to the data of the brightness value for each pixel on the light receiving surface 80B, and can be said to be the data of the intensity value of the reflected light L1B for each position (coordinates) in the detection area AR1. The detection unit 52 calculates the reflectance of the light of the second wavelength for each position in the detection region AR1 based on the image data PB, that is, the intensity of the light L1B for each position in the detection region AR1. For example, the detection unit 52 reflects the light of the second wavelength for each position in the detection region AR1 by substituting the brightness value for each pixel of the image data PB into the relationship between the brightness value calculated in advance and the reflectance. Calculate the rate.

The detection unit 52 determines the reflectance of light on the surface T3A (detection region AR1) of the junction T3 based on the reflectance of light of the first wavelength and the second wavelength in the detection region AR1 calculated based on the image data PA and PB. Is detected. The detection unit 52 determines whether the reflectance of light on the surface T3A of the joint portion T3 is abnormal based on the reflectance of light on the surface T3A of the joint portion T3. That is, the detection unit 52 detects an abnormality in the light reflectance on the surface T3A of the joint portion T3 as a quality abnormality of the member T. In other words, the arithmetic unit 38 can detect a quality abnormality between the first member T1 and the second member T2 based on the reflectance of light on the surface T3A of the joint portion T3. Here, the joint portion T3 is formed by mixing the first member T1 and the second member T2 by stirring with the tool 18. If the material of the first member T1 and the material of the second member T2 are not properly mixed in the joint portion T3, the strength of the joint portion T3 is lowered, for example. From the reflectance of light on the surface T3A of the joint portion T3, the detection unit 52 determines whether the joint portion T3 is in a state where the material of the first member T1 and the material of the second member T2 are not properly mixed. ,To detect.

The cause of the abnormal light reflectance on the surface T3A, in other words, the cause of the first member T1 and the second member T2 not being properly mixed at the joint portion T3 is the first member T1 and the first member T1. Examples include poor mixing with the two members T2 and insufficient heat input to the first member T1 and the second member T2. The first member T1 and the second member T2 are agitated by the rotating tool 18 and mixed with each other to form a joint portion T3 in which both the first member T1 and the second member T2 are mixed. However, for example, when the rotation of the tool 18 is inappropriate (for example, when the rotation speed of the tool 18 is insufficient), the first member T1 and the second member T2 are not sufficiently mixed, or mixed. Defects occur. When a mixing failure occurs, the first member T1 and the second member T2 are not properly mixed. Further, when the frictional heat of the tool 18 is not sufficient (for example, when the feeding speed of the tool 18 is too fast), insufficient heat input to the first member T1 and the second member T2 occurs. When the heat input is insufficient, the first member T1 and the second member T2 are not sufficiently softened by the heat, so that the stirring by the tool 18 is hindered and the first member T1 and the second member T2 are appropriately mixed. It will be in a state where it is not. In the following, the state in which the first member T1 and the second member T2 are appropriately mixed in the joint portion T3 is defined as a normal mixed state, and the first member T1 and the second member T2 are appropriately mixed in the joint portion T3. The unmixed state is defined as an abnormally mixed state. The abnormally mixed state means, for example, a case where the amount of the first member T1 contained in the joint portion T3 is too small or the amount of the second member T2 contained is too small due to improper mixing. .. The joint portion T3 may include both a portion that is in a normal mixed state and a portion that is in an abnormally mixed state.

The joint portion T3 in the abnormally mixed state and the joint portion T3 in the normally mixed state are different from each other in the mixed state of the first member T1 and the second member T2. Therefore, the reflectance of the surface T3A of the joint portion T3 in the abnormally mixed state and the reflectance of the surface T3A of the joint portion T3 in the normally mixed state are different. That is, in the case of poor mixing or insufficient heat input, in other words, in an abnormally mixed state, the reflectance of the surface T3A of the joint portion T3 is the same as the reflectance of the surface T3A of the joint portion T3 in the normally mixed state. ,different. Therefore, the detection unit 52 compares the reflectance of the detected joint portion T3 with the reflectance which is a reference when the mixture and heat input are properly mixed and is in a normal mixing state, so that the mixing is defective. It is possible to detect whether there is insufficient heat input or, in other words, an abnormally mixed state.

In the present embodiment, the detection unit 52 determines the reflectance of the light L1A of the first wavelength and the reflectance of the light L1B of the second wavelength calculated for each pixel of the image sensor 80 as the reflectance of the light of the junction T3. By calculating the ratio with, the reflectance ratio at each position of the junction T3 corresponding to each pixel of the image sensor 80 is calculated. The reflectance is the reflectance of light L1A of the first wavelength in the position (unit region described later) of the detection region AR1 (surface T3A of the junction T3) and the second in the same position (same unit region) of the detection region AR1. It is a ratio with the reflectance of light L1B of a wavelength. The reflection ratio of the detection region AR1 in the normal mixed state and the reflection ratio of the detection region AR1 in the abnormally mixed state are different values. Therefore, the detection unit 52 is based on the calculated reflection ratio and a numerical range of the reflection ratio (hereinafter referred to as a reference reflection ratio range) set in advance with reference to the numerical value of the reflection ratio of the joint portion T3 in the normally mixed state. , It is possible to detect whether there is an abnormality in the reflectance, that is, whether the joint portion T3 is in an abnormally mixed state. The reference reflection ratio range may be arbitrarily set, but is set in advance based on, for example, the numerical value of the reflection ratio of the joint portion T3 in the normally mixed state.

FIG. 9 is a graph showing the relationship between the wavelength of light and the reflectance. FIG. 9 is a graph showing the reflectance of light for each material for each wavelength of light. The line segment C1 and the line segment C2 show the reflectances of different materials. For example, the line segment C1 is the reflectance of the material (aluminum in this case) of the first member T1, and the line segment C2 is the second. It is the reflectance of the material (here, iron) of the member T2. Further, the line segment C3 is the reflectance of the joint portion T3 in the normally mixed state, that is, the member in which the material of the first member T1 and the material of the second member T2 are appropriately mixed. As shown in the line segment C1 and the line segment C2, the reflectance of the material changes according to the wavelength of light, and the tendency of the change in the reflectance according to the wavelength of light varies from material to material, that is, the first member T1. And the second member T2 are different. Therefore, the reflectance, which is the ratio of the reflectance of the light of the first wavelength to the reflectance of the light of the second wavelength, differs for each material, that is, for the first member T1 and the second member T2. Further, since the joint portion T3 in the normally mixed state is a member in which the material of the first member T1 and the material of the second member T2 are mixed, light is emitted from the first member T1 unit and the second member T2 unit. The tendency of the reflectance to change according to the wavelength and the reflectance ratio are different. As described above, the reflection ratios of the first member T1 and the second member T2 and the joint portion T3 in the normally mixed state are also different from each other. For example, the ratio of the reflectance of the light having the wavelength λA in the first member T1 shown by the line C1 in FIG. 9 to the light having the wavelength λB and the light having the wavelength λA in the second member T2 shown by the line C2 in FIG. The reflection ratio of the light of the wavelength λB to the light of the wavelength λB and the reflection ratio of the light of the wavelength λA and the light of the wavelength λB at the junction T3 in the normally mixed state shown by the line segment C3 in FIG. 9 are different from each other.

Since the joint portion T3 in the abnormally mixed state is in a state in which the first member T1 and the second member T2 are not properly mixed, a region containing a large amount of the first member T1 or a region containing a large amount of the second member T2. There are many (wide), and the region where the first member T1 and the second member T2 are appropriately mixed (the region in which the normal mixing state is obtained) is small. Therefore, when the surface T3A of the joint portion T3 has a small region where the reflection ratio of the joint portion T3 is in the normal mixed state, the detection unit 52 has an abnormal reflectance of the joint portion T3 and is in an abnormally mixed state. It can be judged that it has become.

The reason for calculating the reflectance ratio as the reflectance in the present embodiment is to appropriately detect the reflectance of light. Since the reflectance of light changes according to the roughness of the surface on which light is reflected, even if the members are made of the same material, if the surface roughness (roughness of the surface on which light is reflected) is different, the wavelength is the same. The reflectance of light (for example, light of the first wavelength) is different. Therefore, when the reflectance abnormality is determined based on the reflectance of light of one type of wavelength (for example, light of the first wavelength), the influence of the surface roughness is added to the reflectance in addition to the influence of the material. , There is a risk that the detection accuracy of whether it is in an abnormally mixed state (in other words, a normal mixed state) will decrease. On the other hand, members of the same material have a ratio of light reflectances of different wavelengths (that is, light reflectance of the first wavelength and light reflectance of the second wavelength) even if the members have different surface roughness. (Reflection ratio with) is constant. Therefore, by using the reflection ratios of light having different wavelengths as the reflectance as in the present embodiment, it is possible to remove the influence of the surface roughness from the reflectance and appropriately extract the influence of the material. Therefore, it can be appropriately determined whether or not the state is abnormally mixed. However, it is not essential to calculate the reflectance as the reflectance, and as will be described in a modified example described later, for example, an abnormality in the reflectance may be detected based on the reflectance of light of one type of wavelength. ..

The flow of determining the abnormality of the light reflectance in the detection region AR1 will be described below. FIG. 10 is a flowchart illustrating a method for determining an abnormality in reflectance according to the first embodiment. As shown in FIG. 10, the detection unit 52 includes an image data PA which is data of an image of the detection region AR1 by the light L1A (light of the first wavelength) imaged by the image sensor 80 on the light receiving surface 80A, and the image sensor 80. The image data PB, which is the data of the image of the detection region AR1 by the light L1B (light of the second wavelength) captured by the light receiving surface 80B, is acquired (step S10). Then, the detection unit 52 detects the reflection ratio of the detection region AR1 for each pixel of the image sensor 80 based on the image data PA and PB (step S12). Here, one region when the detection region AR1 is divided into a matrix is used as a unit region of the detection region AR1. The number and arrangement of the unit regions of the detection region AR1 match the number and arrangement of pixels on the light receiving surface 80A and the light receiving surface 80B of the image sensor 80. That is, the positions (coordinates) of the pixels on the light receiving surface 80A and the positions (coordinates) of the pixels on the light receiving surface 80B correspond to the positions (coordinates) of the unit area of the detection area AR1, respectively. Based on the image data PA, that is, the brightness value of each pixel of the light receiving surface 80A, the detection unit 52 determines the reflectance of the light of the first wavelength for each pixel of the light receiving surface 80A, that is, for each unit area of the detection area AR1. calculate. Similarly, the detection unit 52 reflects light of the second wavelength for each pixel of the light receiving surface 80B, that is, for each unit area of the detection region AR1, based on the image data PB, that is, the brightness value for each pixel of the light receiving surface 80B. Calculate the rate. Then, the detection unit 52 calculates the ratio of the reflectance of the light of the first wavelength and the reflectance of the light of the second wavelength in the pixels (unit regions) at the same positions as the reflection ratio. The detection unit 52 calculates the reflection ratio for each pixel of the image sensor 80, that is, for each unit area of the detection area AR1.

Then, the detection unit 52 determines whether or not each of the reflection ratios calculated for each pixel is outside the range of the preset reference reflection ratio range. Specifically, the detection unit 52 compares the calculated reflection ratio with the reference reflection ratio range, the first member reflection ratio range, and the second member reflection ratio range. Here, the first member reflection ratio range is a numerical range of the reflection ratio preset with reference to the numerical value of the reflection ratio of the first member T1, and in the present embodiment, the numerical range having a value larger than the reference reflection ratio range. It becomes. The second member reflection ratio range is a numerical range of the reflection ratio preset with reference to the numerical value of the reflection ratio of the second member T2, and in the present embodiment, the value is smaller than the reference reflection ratio range. The detection unit 52 determines whether the reflection ratio calculated for the pixel (unit area of the detection area AR1) is within the reference reflection ratio range, the first member reflection ratio range, or the second member reflection ratio range. To do. When the calculated reflection ratio is within the range of the reference reflection ratio range, the detection unit 52 indicates that the unit region corresponding to the pixel is the region in the normal mixed state, in other words, the first member T1 and the second member. It is determined that T2 is an appropriately mixed region. When the calculated reflection ratio is outside the range of the reference reflection ratio range and is within the range of the first member reflection ratio range (here, the calculated reflection ratio is a value larger than the reference reflection ratio range). In the case), it is determined that the unit region corresponding to the pixel is the region in the abnormally mixed state and is the region in which the first member T1 is abundant. When the calculated reflection ratio is outside the range of the reference reflection ratio range and is within the range of the second member reflection ratio range (here, the calculated reflection ratio is smaller than the reference reflection ratio range). In the case), it is determined that the unit region corresponding to the pixel is the region in the abnormally mixed state and is the region in which the second member T2 is abundant.

In this way, the detection unit 52 determines for each pixel (unit area) whether it is a region in a normal mixed state, a region having a large number of first member T1, or a region having a large number of second member T2. .. However, the detection unit 52 does not have to determine whether it is a region having a large number of the first member T1 or a region having a large number of the second member T2, and whether it is a region in a normal mixed state or a region in an abnormal mixed state. May be determined for each pixel. In this case, when the calculated reflection ratio is within the range of the reference reflection ratio range, the detection unit 52 determines that the unit region corresponding to the pixel is the region in the normal mixed state, and the calculated reflection ratio is calculated. If it is outside the range of the reference reflection ratio range, it is determined that the unit region corresponding to the pixel is the region in the abnormally mixed state.

In the detection unit 52, the number of pixels (unit area) whose calculated reflectance ratio is outside the range of the reference reflectance ratio range, in other words, the number of pixels determined to be in an abnormally mixed state, is a predetermined threshold value (hereinafter). , It is determined whether it is equal to or higher than the first reflectance threshold (referred to as the first reflectance threshold value) (step S14). The first reflectance threshold value here may be set arbitrarily, but for example, when the number of pixels whose calculated reflectance ratio is within the range of the reference reflectance ratio range is equal to or greater than the first reflectance threshold value. It is set based on the fact that the region in the abnormally mixed state in the detection region AR1 becomes large (wide) and there is a high possibility that the first member T1 and the second member T2 are not properly mixed. You can.

When the number of pixels (unit area) whose calculated reflectance ratio is outside the range of the reference reflectance ratio range is equal to or greater than the first reflectance threshold value (step S14; Yes), the detection unit 52 has a quality abnormality of the member T. More specifically, it is determined that the reflectance in the detection region AR1, that is, the reflectance of the surface T3A of the joint portion T3, is abnormal (step S16). On the other hand, when the number of pixels (unit regions) whose calculated reflectance ratio is outside the range of the reference reflectance ratio range is not equal to or greater than the first reflectance threshold value (step S14; No), the detection unit 52 reflects in the detection region AR1. It is determined that the rate, that is, the reflectance of the surface T3A of the joint portion T3 is normal (not abnormal) (step S18). In this way, the detection unit 52 detects the reflectance of light in the detection region AR1 and detects an abnormality in the reflectance.

In addition, even if the detection unit 52 determines in step S12 whether the coordinates of the pixels (unit area) whose calculated reflection ratio is within the range of the reference reflection ratio range are continuous (whether the pixels are adjacent to each other). Good. Then, in step S14, the detection unit 52 determines that the number of pixels whose reflectance is within the range of the reference reflectance range and whose coordinates are continuous is equal to or greater than the first reflectance threshold (the reflectance is the reference reflectance range). The reflectance of the detection region AR1 may be determined to be abnormal when the pixels outside the range of 1 are continuous for the first reflectance threshold value or more). In other words, the detection unit 52 may determine that the reflectance of the detection area AR1 is abnormal when the area determined to be in the abnormally mixed state in the detection area AR1 is wider than a predetermined area.

Further, in the present embodiment, the reflectance ratio of the detection region AR1 (joint portion T3) is calculated, and an abnormality in the reflectance of the detection region AR1 is detected based on the reflection ratio of the detection region AR1, but the present invention is not limited to this. .. For example, the detection unit 52 is based on the ratio of the luminance values of the pixels in the image data PA and PB, that is, when the luminance value of the pixel when the light L1A of the first wavelength is received and the light L1B of the second wavelength are received. An abnormality in the reflectance of the detection area AR1 may be detected based on the ratio to the brightness value of the pixel. In this case, the ratio of the brightness value of the pixel when receiving the light L1A of the first wavelength to the brightness value of the pixel when receiving the light L1B of the second wavelength (hereinafter referred to as the brightness ratio), and the type of the member. Are associated in advance. That is, the detection unit 52 sets the numerical range of the luminance ratio in the region in the normal mixed state, the numerical range of the luminance ratio in the region where the first member T1 is abundant, and the numerical range of the luminance ratio in the region where the second member T2 is abundant. , Get it in advance. Then, when the calculated luminance ratio is within the range of the numerical range of the luminance ratio in the region of the normal mixed state, it is determined that the unit region corresponding to the pixel is the region of the normal mixed state. If the calculated luminance ratio is outside the numerical range of the luminance ratio in the region of the normal mixed state, it is determined that the unit region corresponding to the pixel is the region of the abnormally mixed state. Furthermore, when the calculated luminance ratio is within the numerical range of the luminance ratio in the region where the first member T1 is abundant, the unit region corresponding to the pixel is the region where the first member T1 is abundant. to decide. Furthermore, when the calculated luminance ratio is within the numerical range of the luminance ratio in the region where the second member T2 is abundant, the unit region corresponding to the pixel is the region where the second member T2 is abundant. to decide.

Next, the detection of the opening, the burr, and the processing path by the detection unit 52 will be described. The detection unit 52 detects openings, burrs, and processing paths based on the image data P captured by the second imaging unit 34B. As shown in FIGS. 6 and 7, the second imaging unit 34B images at least a part of the surface TA of the member T and at least a part of the tool 18 within the range of the imaging region R2. The image pickup area R2 is an image pickup area of the second image pickup unit 34B, in other words, a range of images taken by the second image pickup unit 34B. The imaging region R2 overlaps at least a part of the surface TA of the member T and at least a part of the surface of the tool 18. In the present embodiment, the imaging region R2 is at least a part of the surface T1A of the first member T1, at least a part of the surface T2A of the second member T2, and at least a part of the surface T3A of the joint portion T3. Overlaps the area of the tool 18 with at least a portion of the surface of the tool 18.

Hereinafter, the region overlapping the surface TA of the member T and the imaging region R2 on the surface of the tool 18 will be referred to as a detection region AR2. The second imaging unit 34B captures an image of the detection region AR2. Here, as an example, the second imaging unit 34B images an image of the detection region AR2 with natural light emitted from the sun, a fluorescent lamp (for example, a fluorescent lamp in a factory), or the like. For example, as shown in a modified example described later, a light source unit (irradiation unit) that projects light such as visible light toward the detection region AR2 may be provided. Here, in the present embodiment, as shown in FIG. 7, the detection region AR2 partially overlaps the detection region AR1 of the first imaging unit 34A and the irradiation region AR0 of the irradiation unit 32A. Therefore, the light L1AB reflected by the irradiation region AR0 due to the irradiation of the light of the irradiation unit 32A is also incident on the second imaging unit 34B. On the other hand, the second image pickup unit 34B according to the present embodiment blocks the incident light L1AB to cause the image pickup device to take an image of the detection region AR2 excluding the light L1AB. Hereinafter, a specific description will be given.

FIG. 11 is a schematic view of a second imaging unit according to the first embodiment. As shown in FIG. 11, the second image pickup unit 34B includes a lens 90, a filter 92, a lens 94, and an image pickup element 96. Light L2AB from the detection region AR2 is incident on the lens 90. In the present embodiment, the light L2AB is light obtained by superimposing the light from the detection region AR2 (light such as natural light or fluorescent lamp) and the light L1AB irradiated from the irradiation unit 32A and reflected in the detection region AR2. is there. The lens 90 collimates the incident light L2AB and emits it. The filter 92 is a barrier filter that absorbs the light of the component of the light L1AB, that is, the light of the first wavelength and the light of the second wavelength, from the light L2AB emitted from the lens 90. For example, the filter 92 absorbs or reflects light having a wavelength of light L1AB (light having a first wavelength and a second wavelength) among light L2AB, and transmits light having a wavelength other than light L1AB as light L2. That is, the light L2 is the light obtained by excluding the light having the wavelength of the light L1AB from the light L2AB reflected by the detection region AR2 and incident on the lens 90. Light L2 from the filter 92 is incident on the lens 94. The lens 94 collects and emits the light L2 from the filter 92. The image sensor 96 receives the light L2 emitted from the lens 94 and captures an image of the detection region AR2 by the light L2.

The second imaging unit 34B images the detection region AR2 in this way. The detection unit 52 detects openings, burrs, and processing paths from the image of the detection region AR2 captured by the second imaging unit 34B, and more specifically from the image data P captured by the second imaging unit 34B. The image data P refers to the data of the brightness value for each pixel in the image sensor 96, and can be said to be the data of the intensity value of the reflected light L2 for each unit area of the detection area AR2. The unit area of the detection area AR2 refers to one area when the detection area AR2 is divided into a matrix, and the number and arrangement of the unit areas match the number and arrangement of pixels in the image sensor 96.

FIG. 12 is a schematic view for explaining an opening, a burr, and a processing path. When the first member T1 and the second member T2 are not properly joined, an opening F1 may be formed on the surface T3A of the joint portion T3, for example, as shown in FIG. It can be said that the formation of the opening F1 is a quality abnormality because the formation of the opening F1 may cause a decrease in strength. The opening F1 is an opening hole that opens in the surface T3A of the joint portion T3. The opening F1 may be formed in a long groove shape along the feed direction (Y-axis direction) as a groove-like defect. Such an opening F1 is formed due to insufficient heat input to the first member T1 and the second member T2. When the heat input is insufficient, the first member T1 and the second member T2 are not sufficiently softened by the heat, so that the stirring by the tool 18 is hindered and the first member T1 and the space formed by the insertion of the probe 22 are filled with the first member T1. The opening F1 may be formed because the second member T2 does not enter and remains as a space.

Based on the image data P of the second imaging unit 34B, the detection unit 52 detects whether the opening F1 is formed in the region overlapping the detection region AR2 of the surface T3A of the joint portion T3. The opening F1 is formed on the surface T3A of the joint portion T3, and the bottom surface of the opening F1 is located on the inner side (−Z axis direction side) of the joint portion T3 with respect to the surface T3A of the joint portion T3. Therefore, the light to the joint portion T3 (natural light, light from a fluorescent lamp, etc.) is blocked by the surface T3A of the joint portion T3 around the opening F1 and does not reach the bottom surface of the opening F1 or reaches the bottom surface of the opening F1. Even if this is done, the reflected light from the bottom surface may be blocked by the side surface of the opening F1 or the like. Therefore, the intensity of the reflected light from the region where the opening F1 is formed on the surface T3A of the joint portion T3 toward the second imaging unit 34B becomes low, and the pixel at the position corresponding to the opening F1 in the image data P is the opening F1. The brightness value is lower than that of the pixel corresponding to the position where is not formed. Therefore, the detection unit 52 can detect the presence or absence of the opening F1 by detecting the pixels having a low brightness value among the pixels at the positions corresponding to the surface T3A of the joint portion T3.

The method of detecting the opening F1 by the detection unit 52 will be described in more detail. FIG. 13 is a flowchart illustrating a method of detecting an opening. As shown in FIG. 13, the detection unit 52 acquires the image data P imaged by the second image pickup unit 34B, and acquires the brightness value for each pixel of the image pickup element 96 (step S20). The detection unit 52 detects low-luminance pixels whose brightness value is equal to or less than a predetermined value among the pixels of the image sensor 96 (step S22). The predetermined value here may be set arbitrarily, but for example, it may be set as a low brightness value to the extent that an opening is formed. In this case, a predetermined value is obtained in advance based on the result (brightness value of the pixel) of imaging the image of the aperture that is the target of the quality abnormality.

When the low-luminance pixel is detected, the detection unit 52 includes the pixels at the position corresponding to the surface T3A of the joint portion T3, which are consecutively adjacent to each other by a threshold value (hereinafter referred to as a first aperture threshold value) or more. (Step S24). The detection unit 52 extracts low-luminance pixels at positions corresponding to the surface T3A of the joint portion T3. Then, it is determined whether the coordinates of the low-luminance pixels at the positions corresponding to the surface T3A of the joint portion T3 are adjacent to each other by the first opening threshold value or more. The first opening threshold value may be set arbitrarily, but for example, it may be set as a number at which it can be determined that the opening F1 is formed. In this case, the first aperture threshold is determined in advance based on the result of imaging the image of the aperture that is the target of the quality abnormality so that the image consisting of continuously adjacent low-luminance pixels has an elongated shape representing the aperture. deep. If the low-luminance pixels at the positions corresponding to the surface T3A of the joint portion T3 are adjacent to each other by the first opening threshold value or more, the detection unit 52 is continuously connected to the pixels at the positions corresponding to the surface T3A of the joint portion T3. It is determined that low-luminance pixels adjacent to each other by one aperture threshold value or more are included. When the detection unit 52 determines that the pixels at the positions corresponding to the surface T3A of the joint portion T3 include low-luminance pixels that are consecutively adjacent to each other by the first opening threshold value or more (step S24; Yes), the quality abnormality of the member T is abnormal. More specifically, it is determined that the opening F1 is formed in the joint portion T3 (step S26). Further, when the detection unit 52 determines that the pixels at the positions corresponding to the surface T3A of the joint portion T3 do not continuously include the low-luminance pixels adjacent to each other by the first opening threshold value or more (step S24; No), the joint portion 52 It is determined that the opening F1 is not formed in T3 (step S28). By determining that low-luminance pixels that are continuously adjacent to each other at the first opening threshold value or higher are the aperture F1, when the brightness of pixels less than the first aperture threshold value is low due to, for example, a small foreign object or shot noise, the opening F1 is used. It is possible to suppress false detection if there is.

The detection unit 52 presets information on which of the pixels included in the image sensor 96 is the pixel corresponding to the position of the surface T3A of the junction T3. For example, at the time of joining, the distance between the second imaging unit 34B and the surface TA of the member T in the imaging direction of the second imaging unit 34B is kept constant. Therefore, the relationship between the position (coordinates) of the unit region on the surface TA of the member T at the time of joining and the position (coordinates) of the pixels of the second imaging unit 34B is fixed. From the image data of the member T, which pixel corresponds to the position of the surface T3A of the joint portion T3 can be set in advance. Further, the detection unit 52 analyzes the image data P captured by the second imaging unit 34B without setting in advance which pixel corresponds to the position of the surface T3A, and which pixel is the surface T3A. It may be detected whether the pixel corresponds to the position.

Next, the detection of the burr F2 will be described. The detection unit 52 detects the burr F2 as a quality abnormality of the member T. Burr is caused by excessive heat input to the first member T1 and the second member T2. Excessive heat input occurs when the frictional heat of the tool 18 is too high. When excessive heat input occurs, for example, the first member T1 and the second member T2 are heated to near the melting point, and the metal structure of the member T, more specifically, the joint portion T3, grows and becomes coarse. When the metal structure becomes coarse, it is likely to be damaged (that is, breakage at the joint interface between the first member T1 and the second member T2) starting from the interface of the metal structure.

Here, when excessive heat input occurs, the member T is softened more than expected, for example, as shown in FIG. 12, the member T pushed away by the tool 18 (specifically, the softened member T). A part) may protrude as a burr F2 from the boundary portion T3B of the joint portion T3 with the first member T1 or the second member T2. By detecting the burr F2, the detection unit 52 can detect whether there is a risk of excessive heat input, in other words, damage due to coarsening of the metal structure.

Based on the image data P of the second imaging unit 34B, the detection unit 52 detects whether the burr F2 is formed in the region overlapping the detection region AR2 around the surface T3A of the joint portion T3 (boundary portion T3B). The burr F2 is formed around the joint portion T3, more specifically, at the boundary portion T3B of the joint portion T3 with the first member T1 or the second member T2. Further, since the burr F2 has a large number of protrusions in various directions, a part of the light incident on the burr F2 is specularly reflected and incident on the second imaging unit 34B. The specularly reflected light incident on the second imaging unit 34B has a higher intensity than the light reflected by a portion other than the burr F2 and incident on the second imaging unit 34B (more specifically, diffusely reflected light). Therefore, the pixel at the position corresponding to the burr F2 in the image data P has a higher luminance value than the pixel corresponding to the position where the burr F2 is not formed. The detection unit 52 detects the presence or absence of the burr F2 by detecting the pixel having a high luminance value among the pixels at the position corresponding to the boundary portion T3B.

The method of detecting the burr F2 by the detection unit 52 will be described in more detail. FIG. 14 is a flowchart illustrating a method for detecting burrs. As shown in FIG. 14, the detection unit 52 acquires the image data P imaged by the second image pickup unit 34B, and acquires the brightness value for each pixel of the image pickup element 96 (step S30). The detection unit 52 detects high-luminance pixels whose brightness value is equal to or higher than a predetermined value among the pixels of the image sensor 96 (step S32). In other words, the detection unit 52 extracts the pixels whose brightness value is equal to or higher than the predetermined value among the pixels of the image sensor 96. The predetermined value here may be set arbitrarily, but for example, it may be set as a brightness value high enough to form the burr F2. In this case, a predetermined value is obtained in advance based on the result (brightness value of the pixel) of imaging the image of the burr that is the target indicating the quality abnormality.

When the high-luminance pixel is detected, the detection unit 52 continuously adjacent the high-luminance pixel around the joint portion T3, that is, the pixel at the position corresponding to the boundary portion T3B, by a threshold value (hereinafter referred to as a first burr threshold value) or more. Is included (step S34). Then, the detection unit 52 determines whether the high-luminance pixels at the positions corresponding to the boundary portion T3B are adjacent to each other by the first burr threshold value or more. The first burr threshold value may be set arbitrarily, but may be set as a number that can be determined to form, for example, the burr F2. In this case, the first burr threshold value is obtained in advance based on the result of capturing an image of a burr representing a quality abnormality so that an image composed of continuously adjacent high-luminance pixels has a shape representing a burr. If the high-luminance pixels at the positions corresponding to the boundary portion T3B are adjacent to each other by the first burr threshold value or more, the detection unit 52 is continuously adjacent to the pixels at the position corresponding to the boundary portion T3B by the first burr threshold value or more. It is determined that high-luminance pixels are included. When the detection unit 52 determines that the pixels at the positions corresponding to the boundary portion T3B continuously include high-luminance pixels adjacent to each other by the first burr threshold value or more (step S34; No), a quality abnormality of the member T occurs. More specifically, it is determined that the burr F2 is formed at the boundary portion T3B (step S36). Further, when the detection unit 52 determines that the pixels at the positions corresponding to the boundary portion T3B do not continuously include high-brightness pixels adjacent to each other by the first burr threshold value or more (step S34; No), the detection unit 52 burrs the boundary portion T3B. It is determined that F2 is not formed (step S38). By determining that high-brightness pixels that are consecutively adjacent to each other above the first burr threshold value are burr F2, for example, when the brightness value of pixels smaller than the first burr threshold value is high due to small foreign matter or shot noise, the burr F2 It is possible to prevent erroneous detection.

The detection unit 52 presets information on which of the pixels included in the image sensor 96 is the pixel corresponding to the boundary portion T3B. When joining the first member T1 and the second member T2, the distance between the second imaging unit 34B and the surface TA of the member T in the imaging direction of the second imaging unit 34B is kept constant. .. Therefore, when joining the first member T1 and the second member T2, the position (coordinates) of the unit region on the surface TA of the member T and the position (coordinates) of the pixels of the image sensor 96 of the second image pickup unit 34B. Since the relationship with is fixed, for example, in calibration, it is possible to preset which pixel corresponds to the position of the boundary portion T3B from the image data of the appropriately joined member T. Further, the detection unit 52 analyzes the image data P captured by the second imaging unit 34B without setting in advance which pixel corresponds to the position of the boundary portion T3B, and which pixel is the boundary portion. It may be detected whether the pixel corresponds to the position of T3B. The detection unit 52 does not have to detect the burr F2 based on the high-luminance pixels among the pixels of the image sensor 96. For example, the detection unit 52 may detect the burr F2 based on the contrast of the image (image data P) captured by the second imaging unit 34B. In this case, even if the detection unit 52 identifies the contour of the characteristic shape of the burr F2 (for example, the shape of the protrusion of the burr F2) based on the contrast of the image (image data P) and detects the burr F2. Good.

Next, the processing path will be described. The detection unit 52 detects an abnormality in the machining path as a quality abnormality of the member T. If the processing path is abnormal, the thickness of the compound between the first member T1 and the second member T2 may increase. The compound of the first member T1 and the second member T2 is, for example, an intermetallic compound (FeAl or the like in this embodiment) and has high brittleness. Therefore, if the thickness of the intermetallic compound between the first member T1 and the second member T2 becomes thicker, the risk of damage starting from the intermetallic compound increases. The thickness of the intermetallic compound between the first member T1 and the second member T2 here is the first member T1 and the first member T1 in a direction orthogonal to the relative feeding direction of the tool 18 (here, the X-axis direction). 2 Refers to the length of the intermetallic compound with the member T2. The machining path is the tool 18 (probe 22) with respect to the member T in the direction orthogonal to the feed direction (here, the X-axis direction) in the state where the probe 22 is inserted into the member T while rotating (the state during joining). It can also be said to be the position of. The machining path may be shifted to the direction orthogonal to the feed direction with respect to an appropriate position, or the preset position may be inappropriate. When the machining path is abnormal in this way, the heat input positions to the first member T1 and the second member T2 become abnormal. When the heat input position becomes abnormal, the balance between the amount of heat given to the first member T1 by the friction stir welding of the tool 18 and the amount of heat given to the second member T2 by the friction stir welding of the tool 18 becomes improper, and the first The thickness of the intermetallic compound between the member T1 and the second member T2 becomes thicker. Since the intermetallic compound is formed inside the member T, it is difficult to detect the thickness of the intermetallic compound from the outside, especially during the joining between the first member T1 and the second member T2. On the other hand, the detection unit 52 according to the first embodiment detects whether the heat input position is abnormal, that is, whether the thickness of the intermetallic compound is thick, by determining whether the processing path is abnormal. Can be done.

The detection unit 52 detects the processing path based on the image data P of the second imaging unit 34B. For example, the detection unit 52 detects the position of the boundary portion T3B in the direction orthogonal to the feed direction (X-axis direction) from the image captured by the second imaging unit 34B as a processing path. For example, the detection unit 52 extracts and extracts pixel groups in which the brightness values of adjacent pixels in the direction orthogonal to the feed direction (X-axis direction) differ by a predetermined value or more from the image data P of the second imaging unit 34B. When the pixel group is continuous in the feed direction (Y-axis direction), the continuous pixel group is defined as a pixel corresponding to the position of the boundary portion T3B. In this case, a predetermined value is obtained in advance based on the result of imaging the image including the boundary portion T3B on the surface TA of the member T (the brightness value of the pixel). When the first member T1 and the second member T2 are joined, the distance between the second image pickup unit 34B and the surface TA of the member T in the imaging direction of the second image pickup unit 34B is kept constant. Since the relationship between the position (coordinates) of the unit region on the surface TA (detection region AR2) of the member T and the position (coordinates) of the pixels of the image sensor 96 of the second image pickup unit 34B is fixed, the second image pickup is performed. The pixels in the image captured by the unit 34B can be associated with the position of the boundary portion T3B on the surface TA of the member T in advance. Therefore, the detection unit 52 can detect the processing path based on the image data P of the second imaging unit 34B.

The detection unit 52 presets a reference machining path which is the position of the boundary portion T3B in the direction orthogonal to the feed direction (X-axis direction) when the tool 18 is in an appropriate position. The reference machining path may be set arbitrarily, but it may be set as the position of the boundary portion T3B when the heat input positions to the first member T1 and the second member T2 are appropriate. The detection unit 52 calculates the difference between the detected machining path and the reference machining path, and when the difference between the detected machining path and the reference machining path is equal to or greater than the threshold value (hereinafter referred to as the first pass threshold value), the member It is determined that the processing path is abnormal as the quality abnormality of T. The first pass threshold value may be arbitrarily set, but may be set as a threshold value when the thickness of the intermetallic compound between the first member T1 and the second member T2 becomes thick enough to cause quality abnormality. Then, the detection unit 52 determines that the machining path is normal (not abnormal) when the difference between the detected machining path and the reference machining path is smaller than the first pass threshold value. The detection unit 52 sets an appropriate position of the tool 18 in a direction orthogonal to the feed direction as a reference processing path, and sets a direction orthogonal to the feed direction from the image data P imaged by the second image pickup unit 34B. The position of the tool 18 in the X-axis direction) may be detected as a machining path. In this case, the detection unit 52 detects the machining path as a tool state indicating the state of the tool 18.

As described above, the detection unit 52 detects the light reflectance of the surface TA of the member T, the opening F1, the burr F2, and the processing path as the member state. Then, the detection unit 52 determines whether the light reflectance on the surface of the member T, the opening F1, the burr F2, and the processing path are abnormal, that is, the member state is abnormal (the member T determines. However, the detection unit 52 may detect at least one of the light reflectance on the surface of the member T, the opening F1, the burr F2, and the processing path as the member state, and reflects the light on the surface of the member T. It may be determined whether at least one of the rate, the opening F1, the burr F2, and the machining path is abnormal.

Next, the detection of the tool state by the detection unit 52 will be described. First, the detection of the insertion amount of the tool 18 as the tool state will be described. 15A and 15B are schematic views for explaining the insertion amount of the tool. The detection unit 52 detects the insertion amount of the tool 18 based on the image captured by the second imaging unit 34B. The insertion amount of the tool 18 is the length of the portion inserted into the member T of the tool 18 as described above. In the present embodiment, as shown in FIG. 15A (A), the detection unit 52 is the end portion 20A of the shoulder 20 and the surface of the member T at the timing when the tip portion 22A of the probe 22 comes into contact with the surface TA of the member T. The distance D0a between the TA and the TA is detected based on the image captured by the second imaging unit 34B at that timing. Then, as shown in FIG. 15A (B), the detection unit 52 imaged the distance D0 between the end portion 20A of the shoulder 20 and the surface TA of the member T during joining by the second imaging unit 34B. Detect based on the image. Then, the detection unit 52 calculates the amount obtained by subtracting the distance D0 from the distance D0a as the insertion amount of the tool 18. Since the position of the second imaging unit 34B with respect to the tool 18 is fixed at the time of joining, the second imaging unit 34B is between the second imaging unit 34B and the tool 18 in the imaging direction of the second imaging unit 34B at the time of joining. The distance is kept constant. Therefore, since the relationship between the position (coordinates) of the unit region on the surface of the tool 18 at the time of joining and the position (coordinates) of the pixels of the image sensor 96 of the second image sensor 34B is fixed, the second image sensor 34B The distance between the pixels in the image captured by the shoulder 20 can be associated with the distance between the end portion 20A of the shoulder 20 and the surface TA of the member T in advance. Therefore, the detection unit 52 can calculate the distance D0a and the distance D0 based on the image captured by the second imaging unit 34B. As an example, the distance D0a is detected in advance before joining the members.

The detection unit 52, for example, is based on the measurement result of the rotational torque of the tool 18 in the θZ direction measured by the torque sensor 36 (see FIG. 1) at the timing when the tip portion 22A of the probe 22 comes into contact with the surface TA of the member T. ,To detect. The torque sensor 36 is a sensor provided in the manufacturing apparatus 10 and measures the torque of rotation of the tool 18 in the θZ direction. The control device 30 of the manufacturing device 10 causes the torque sensor 36 to measure the torque of the tool 18 in the θZ direction. The detection unit 52 of the arithmetic unit 38 receives the torque measurement result by the torque sensor 36. FIG. 15B is an example of a graph showing the value of the torque of rotation of the tool 18 in the θZ direction for each time, which is detected by the torque sensor 36. In FIG. 15B, at time t0, the tool 18 is rotated with the tip 22A of the probe 22 of the tool 18 separated from the surface TA of the member T, and the tip 22A of the probe 22 is moved to the member T with the passage of time. An example of the torque of rotation of the tool 18 in the θZ direction when approaching the surface TA is shown. As shown in the torque from time t0 to time t1 in FIG. 15B, the torque value is constant until the tip portion 22A of the probe 22 comes into contact with the surface TA of the member T. Then, from the time t1 when the tip 22A of the probe 22 of the tool 18 comes into contact with the surface TA of the member T, the torque increases due to the resistance of the member T to the rotation of the tool 18 in the θZ direction. The detection unit 52 sequentially acquires the torque of rotation of the tool 18 in the θZ direction measured by the torque sensor 36, and detects the timing at which the torque of rotation of the tool 18 in the θZ direction starts to increase. Then, the detection unit 52 sets the timing at which the torque of rotation of the tool 18 in the θZ direction starts to increase (that is, time t1) as the timing at which the tip portion 22A of the probe 22 comes into contact with the surface TA of the member T. The distance between the end portion 20A of the shoulder 20 and the surface TA of the member T is calculated as the distance D0a. The distance D0a does not have to be calculated using the torque value of the torque sensor 36. For example, the distance D0a may be calculated by visually recognizing the contact of the tip end 22A of the probe 22 with the surface TA of the member T with the image captured by the second imaging unit 34B or by the user. The detection unit 52 does not have to obtain the insertion amount of the tool 18 based on the image captured by the second imaging unit 34B. For example, the detection unit 52 may detect the amount of movement of the tool stage 14 from the time t1 until the tool 18 moved in the −Z axis direction stops as the insertion amount of the tool 18. In this case, the arithmetic unit 38 (detection unit 52) acquires information on the amount of movement of the tool stage 14 from the manufacturing device 10.

The detection unit 52 determines whether or not the insertion amount of the tool 18 detected in this way is abnormal. The detection unit 52 presets a reference insertion amount, which is an insertion amount of the tool 18 when the insertion amount of the tool 18 is appropriate. The reference insertion amount may be arbitrarily set, but may be set as the insertion amount of the tool 18 when the joining between the first member T1 and the second member T2 is appropriate. Then, the detection unit 52 calculates the difference between the insertion amount of the tool 18 calculated from the image captured by the second imaging unit 34B and the reference insertion amount, and the calculated difference between the insertion amount of the tool 18 and the reference insertion amount. When is equal to or greater than the threshold value (hereinafter referred to as the first insertion threshold value), it is determined that the insertion amount of the tool 18 is abnormal. The first insertion threshold value may be arbitrarily set, but may be set as a threshold value when the insertion amount of the tool 18 becomes abnormal to the extent that the quality becomes abnormal. Then, when the difference between the calculated insertion amount of the tool 18 and the reference insertion amount is smaller than the first insertion threshold value, the detection unit 52 determines that the insertion amount of the tool 18 is normal (not abnormal). More specifically, in the present embodiment, when the calculated insertion amount of the tool 18 is larger than the reference insertion threshold by the first insertion threshold value or more, the insertion amount of the tool 18 is abnormal. Judge. Then, the detection unit 52 determines that the insertion amount of the tool 18 is normal when the calculated insertion amount of the tool 18 is not larger than the first insertion threshold value with respect to the reference insertion amount.

Next, the detection of the shape of the tool 18 as the tool state will be described. FIG. 16 is a schematic view illustrating the detection of the shape of the tool 18. For convenience of explanation, the outer shape of the probe 22 in FIG. 16 is expressed in more detail than in other drawings. The detection unit 52 detects the shape of the tool 18 based on the image of the tool 18 captured by the second imaging unit 34B. In the present embodiment, the detection unit 52 detects the outer diameter and groove dimensions of the tool 18 as the shape of the tool 18, but may detect at least one of the outer diameter and groove dimensions of the tool 18. The groove size can be rephrased as the uneven size. When detecting the shape of the tool 18, the arithmetic unit 38 causes the manufacturing device 10 to stop joining the first member T1 and the second member T2. Specifically, as shown in FIG. 16, the tool 18 is moved in the + Z axis direction from the state where the probe 22 is inserted into the member T during the joining between the first member T1 and the second member T2. The probe 22 is exposed to the outside of the member T. Then, the arithmetic unit 38 causes the second imaging unit 34B to take an image in the range of the imaging region R2 with the shoulder 20 and the probe 22 exposed to the outside of the member T, so that the second imaging unit 34B can perform the shoulder 20 and the probe 22. The probe 22 can be imaged. When the second imaging unit 34B is movable (for example, the second imaging unit 34B can be moved relative to the tool 18), the second imaging unit 34B is measured when measuring the outer diameter of the tool 18 and the groove size. The 34B may be moved with respect to the tool 18 to change the relative position between the second imaging unit 34B and the tool 18. In this case, the second imaging unit 34B may be moved to a predetermined position where the shoulder 20 and the probe 22 can be appropriately imaged, and the second imaging unit 34B may image the shoulder 20 and the probe 22.

The detection unit 52 detects the shape of the tool 18, here the outer diameter of the tool 18 and the dimensions of the groove, based on the image taken by the second imaging unit 34B of the tool 18 (shoulder 20 and probe 22). As shown in FIG. 16, the detection unit 52 detects the outer diameter D1 of the shoulder 20 as the outer diameter of the tool 18 based on the image captured by the second imaging unit 34B. Further, the detection unit 52 has an outer diameter D2 of a convex portion (male thread portion) on the outer peripheral surface of the probe 22 and a concave portion (male screw) on the outer peripheral surface of the probe 22 based on the image captured by the second imaging unit 34B. The outer diameter D3 of) is detected. The detection unit 52 detects the difference between the outer diameter D2 and the outer diameter D3 as the size of the groove of the tool 18 (probe 22).

The detection unit 52 determines whether the shape of the tool 18 calculated in this way is abnormal. First, the abnormality determination of the outer diameter of the tool 18 will be described. The detection unit 52 presets a reference outer diameter, which is the outer diameter of the tool 18 when the tool 18 is not worn. The detection unit 52 is a tool when the outer diameter D1 of the shoulder 20 detected from the image captured by the second imaging unit 34B is smaller than the threshold value (hereinafter referred to as the first outer diameter threshold value) with respect to the reference outer diameter. It is determined that the outer diameter of 18 (outer diameter D1 of the shoulder 20) is abnormal. The first outer diameter threshold value may be arbitrarily set, but may be set as a threshold value when it is determined that the outer diameter of the tool 18 is abnormal to the extent that the quality is abnormal. Then, when the outer diameter D1 of the shoulder 20 is not smaller than the first outer diameter threshold value by the detection unit 52, the outer diameter of the tool 18 (outer diameter D1 of the shoulder 20) is normal. Judge as (not abnormal).

Next, the abnormality determination of the groove dimension of the tool 18 will be described. Further, the detection unit 52 presets a reference groove dimension, which is a groove dimension of the tool 18 when the tool 18 is not worn. Then, when the groove size of the tool 18 calculated from the image captured by the second imaging unit 34B is smaller than the threshold value (hereinafter referred to as the first groove threshold value) with respect to the reference groove size, the detection unit 52 may use the detection unit 52. It is determined that the groove size of the tool 18 is abnormal. The first groove threshold value may be arbitrarily set, but it may be set as a threshold value when the groove of the tool 18 becomes abnormal to the extent that the quality becomes abnormal. Then, the detection unit 52 determines that the groove size of the tool 18 is normal (not abnormal) when the calculated groove size of the tool 18 is not smaller than the first groove threshold value with respect to the reference groove size. To do.

As described above, the detection unit 52 detects the insertion amount of the tool 18 and the shape of the tool 18 as the tool state, and whether the insertion amount of the tool 18 and the shape of the tool 18 are abnormal, that is, the tool. Determine if the condition is abnormal. However, the detection unit 52 may detect at least one of the insertion amount of the tool 18 and the shape of the tool 18 as a tool state, and at least one of the insertion amount of the tool 18 and the shape of the tool 18 is abnormal. You just have to judge.

The detection unit 52 detects the member state and the tool state as described above. The joint state determination unit 54 shown in FIG. 5 determines whether the joint state between the first member T1 and the second member T2 is defective based on the member state and the tool state determined by the detection unit 52. The joined state refers to the state of the member T when the first member T1 and the second member T2 are joined. Examples of the joining state include a mixed state of the first member T1 and the second member T2, an amount of heat input to the member T, a heat input position to the member T, and the like. The mixed state of the first member T1 and the second member T2 indicates the degree of agitation between the first member T1 and the second member T2 by the probe 22. The amount of heat input to the member T refers to the amount of heat transfer to the member T of the frictional heat generated by the friction stir welding of the tool 18. Further, the heat input position to the member T refers to a position on the member T to which the frictional heat of the tool 18 is applied. In the case where such a joining state becomes defective (hereinafter referred to as a defective joining state mode), the mixing state of the first member T1 and the second member T2 becomes defective to the extent that the quality becomes abnormal. To the extent that the quality is abnormal, the amount of heat input to the member T is excessive to the extent that the quality is abnormal, the amount of heat input to the member T is insufficient to the extent that the quality is abnormal, and the amount of heat is insufficient to the extent that the quality is abnormal. Examples thereof include a defective heat input position in which the heat input position to the member T is defective. The joint state determination unit 54 determines which joint state defect mode the member T corresponds to among the plurality of joint state defect modes mentioned above based on the member state and tool state determined by the detection unit 52. To judge. The joining state may be at least one of a mixed state of the first member T1 and the second member T2, an amount of heat input to the member T, and a heat input position to the member T. In other words, the joint state determination unit 54 may determine whether at least one of the mixed state, the amount of heat input to the member T, and the heat input position to the member T is defective.

The change information generation unit 56 generates change information based on the member state and the tool state detected by the detection unit 52. Furthermore, the change information generation unit 56 generates change information based on the joint state determined by the joint state determination unit 54 to be defective. The change information is information for changing the operating conditions of the manufacturing apparatus 10, but the details will be described later. The change information output unit 58 outputs the change information generated by the change information generation unit 56 to the control device 30. The control device 30 resets the operating conditions based on the change information output from the change information generation unit 56, controls each part of the manufacturing apparatus 10 under the reset operating conditions, and controls the first member T1 and the first member T1. 2 Join the member T2.

The arithmetic unit 38 has the above configuration. Hereinafter, the processing flow by the arithmetic unit 38 will be described. FIG. 17 is a flowchart illustrating a processing flow of the arithmetic unit according to the first embodiment. As shown in FIG. 17, in the arithmetic unit 38, the device control unit 50 triggers the joining of the first member T1 and the second member T2 as a trigger, and the arithmetic unit 38 uses the device control unit 50 to perform the imaging unit 34 at a predetermined frame rate. Is started to capture an image (step S40). When the manufacturing apparatus 10 starts joining the first member T1 and the second member T2, here, the tool 18 is moved in the + Z axis direction while rotating in the θZ direction to start approaching the member T. However, it may indicate that the probe 22 is inserted into the member T while the tool 18 rotates in the θZ direction. When the manufacturing apparatus 10 starts joining the first member T1 and the second member T2, the device control unit 50 causes the irradiation unit 32A to irradiate the region R0 with light LAB, and causes the first imaging unit 34A to range the imaging region R1. The image inside is imaged, and the second image pickup unit 34B is made to image the image within the range of the image pickup region R2. The device control unit 50 keeps the irradiation unit 32A irradiating the region R0 with the optical LAB until the manufacturing apparatus 10 stops joining the first member T1 and the second member T2, and at a predetermined frame rate. The first imaging unit 34A is made to capture an image within the range of the imaging region R1, and the second imaging unit 34B is made to capture an image within the range of the imaging region R2. The device control unit 50 causes the second image pickup unit 34B to take an image at the timing when the tip portion 22A of the probe 22 comes into contact with the surface TA of the member T, and the detection unit 52 causes the probe 22 to take an image based on the image. The distance D0a between the end portion 20A of the shoulder 20 and the surface TA of the member T at the timing when the tip portion 22A of the member T comes into contact with the surface TA of the member T is calculated.

When the imaging unit 34 captures an image, the arithmetic unit 38 acquires the image data generated by the imaging unit 34 by the detecting unit 52, and as a member state, the reflectance of the surface T3A of the joint portion T3 is abnormal. The presence / absence, the opening F1, the burr F2, and the presence / absence of an abnormality in the machining path are detected (step S42). The detection unit 52 detects the member state during the joining by the manufacturing apparatus 10. Each time the image pickup unit 34 captures an image, the detection unit 52 acquires the image data generated by the image pickup unit 34 and detects the member state. The detection unit 52 detects the reflectance of the surface T3A of the joint portion T3 as a member state from the image data PA and PB generated by the first imaging unit 34A. Further, the detection unit 52 detects the opening F1 as a member state from the image data P generated by the second imaging unit 34B. Further, the detection unit 52 detects the burr F2 as a member state from the image data P generated by the second imaging unit 34B. Further, the detection unit 52 detects the processing path as a member state from the image data P generated by the second imaging unit 34B. The detection unit 52 determines whether the detected member state is abnormal.

When the arithmetic device 38 has an abnormality in the detection result of the member state (step S44; Yes), more specifically, in step S42, the reflectance of the surface T3A of the joint portion T3 is abnormal (that is, the reflectance is abnormal). ), The opening F1, the burr F2, the processing path abnormality (that is, there is a processing path abnormality), the reflectance abnormality of the surface T3A of the joint portion T3, the opening F1, the burr F2, and the processing path. When at least one of the abnormalities is detected, in other words, when a quality abnormality of the member T is detected, it is determined whether the reflectance of the surface T3A of the joint portion T3 is abnormal and the burr F2 is detected ( Step S46).

When both the abnormality of the reflectance of the surface T3A of the joint portion T3 and the burr F2 are detected in the arithmetic unit 38 (step S46; Yes), the detection unit 52 determines the shape of the tool 18 and the insertion amount of the tool 18. Is detected as a tool state (step S48). The arithmetic unit 38 causes the manufacturing apparatus 10 to stop joining to expose the probe 22 to the outside of the member T, and causes the second imaging unit 34B to capture an image of the tool 18. The arithmetic unit 38 detects the shape of the tool 18 from the image of the tool 18 captured by the second imaging unit 34B. Specifically, when the reflectance of the surface T3A is abnormal and the burr F2 is detected, the arithmetic unit 38 generates a joining stop signal and transmits it to the control device 30 of the manufacturing apparatus 10. The joining stop signal is a signal instructing the manufacturing apparatus 10 to stop joining the first member T1 and the second member T2. When the control device 30 receives the joining stop signal from the control device 30, the control device 30 stops the joining between the first member T1 and the second member T2, moves the tool 18 in the + Z axis direction, and moves the end portion 20A of the shoulder 20 to the shoulder 20. By moving the probe 22 from the inside of the member T to the outside away from the surface TA of the member T, the shoulder 20 and the probe 22 are exposed from the member T. After the shoulder 20 and the probe 22 are exposed from the member T, the control device 30 transmits a response signal indicating that the shoulder 20 and the probe 22 are exposed from the member T to the arithmetic unit 38. When the arithmetic unit 38 receives the response signal from the control device 30, the arithmetic unit 38 causes the second imaging unit 34B to capture an image of the tool 18. The arithmetic unit 38 detects the shape of the tool 18 as a tool state based on the image of the tool 18 captured by the second imaging unit 34B by the detection unit 52. The arithmetic unit 38 determines whether the detected tool 18 has an abnormal shape.

Further, the arithmetic unit 38 detects the insertion amount of the tool 18 from the image data P generated by the second imaging unit 34B, that is, from the image captured by the second imaging unit 34B during joining. Specifically, the arithmetic unit 38 detects as the insertion amount of the tool 18 the amount obtained by subtracting the distance D0 detected based on the image data P generated by the second imaging unit 34B from the distance D0a detected in advance. .. The arithmetic unit 38 determines whether the detected insertion amount of the tool 18 is abnormal.

When the shape of the tool 18 and the insertion amount of the tool 18 are detected as the tool state, the arithmetic unit 38 is joining the manufacturing apparatus 10 based on the member state and the tool state detected by the detection unit 52 by the change information generation unit 56. Change information for changing the operating conditions is generated, and the generated change information is transmitted to the control device 30 (step S58). The method of generating change information will be described later.

In the arithmetic unit 38, when the reflectance of the surface T3A is abnormal and the condition that the burr F2 is detected is not satisfied (step S46; No), that is, at least one of the abnormal reflectance and the burr F2 is If it is not detected, it is determined whether or not the reflectance is abnormal. When the reflectance is abnormal (step S50; Yes), the arithmetic unit 38 detects the shape of the tool 18 as the tool state (step S52). That is, the arithmetic unit 38 detects the shape of the tool 18 as the tool state when the burr F2 is not detected but the reflectance of the surface T3A is abnormal. The method of detecting the shape of the tool 18 is the same as the method of detecting the shape of the tool 18 in step S48. After detecting the shape of the tool 18 as the tool state, the arithmetic unit 38 changes the operating condition during joining of the manufacturing apparatus 10 based on the member state and the tool state detected by the detection unit 52 by the change information generation unit 56. The change information of the above is generated, and the generated change information is transmitted to the control device 30 (step S58). The method of generating change information will be described later.

When the abnormality of the reflectance is not detected (step S50; No), the arithmetic unit 38 determines whether or not the burr F2 is detected. When the burr F2 is detected (step S54; Yes), the arithmetic unit 38 detects the insertion amount of the tool 18 as the tool state (step S56). The method of detecting the insertion amount of the tool 18 is the same as the method of detecting the insertion amount of the tool 18 in step S52. When the insertion amount of the tool 18 is detected as the tool state, the arithmetic unit 38 is changed by the change information generation unit 56 to change the operating conditions during joining of the manufacturing device 10 based on the tool state detected by the detection unit 52. Information is generated, and the generated change information is transmitted to the control device 30 (step S58). The method of generating change information will be described later. Since the shape of the tool 18 is not detected in step S56, the control for stopping the joining between the first member T1 and the second member T2 in order to detect the shape of the tool 18 is not executed (in the manufacturing apparatus 10). The joining between the first member T1 and the second member T2 is not stopped).

When the burr F2 is not detected (step S54; No), that is, when neither the reflectance abnormality nor the burr F2 is detected, the arithmetic unit 38 does not detect the tool state, but based on the detected member state, the arithmetic unit 38 of the manufacturing device 10. Change information for changing the operating conditions during joining is generated, and the generated change information is transmitted to the control device 30 (step S58). In other words, when the abnormality of the member state is other than the abnormality of the reflectance and the detection of the burr F2, here, when the abnormality of the opening F1 or the processing path is detected and the abnormality of the reflectance and the burr F2 are not detected, The arithmetic unit 38 generates change information based on the member state without detecting the tool state. The method of generating change information will be described later. Even when the abnormality of the reflectance is not detected and the burr F2 is not detected, the shape of the tool 18 is not detected. Therefore, the joining between the first member T1 and the second member T2 is stopped in order to detect the shape of the tool 18. The control is not executed (the manufacturing apparatus 10 does not stop the joining of the first member T1 and the second member T2).

The control device 30 of the manufacturing device 10 changes the operating conditions based on the change information transmitted from the arithmetic unit 38. When the joining between the first member T1 and the second member T2 is completed after the change information is generated and transmitted (step S60; Yes), this process is terminated. When the joining between the first member T1 and the second member T2 is not completed (step S60; No), the control device 30 of the manufacturing apparatus 10 has the first member T1 and the second member T1 and the second member under the operating conditions changed based on the change information. Continuing the bonding with T2, the arithmetic unit 38 returns to step S40, and continues the irradiation of the optical LAB by the irradiation unit 32A and the imaging by the imaging unit 34.

In this way, the arithmetic unit 38 detects the shape of the tool 18 when an abnormality in the reflectance is detected. The arithmetic unit 38 determines whether the detected tool 18 has an abnormal shape. Further, the arithmetic unit 38 detects the insertion amount of the tool 18 when the burr F2 is detected. The arithmetic unit 38 determines whether the detected insertion amount of the tool 18 is abnormal. The arithmetic unit 38 does not have to detect the insertion amount of the tool 18 triggered by the detection of the burr F2, and each time the second imaging unit 34B images at a predetermined frame rate, the second imaging unit 38 The insertion amount of the tool 18 may be sequentially detected based on the image data P of 34B.

In step S44, when there is no abnormality in the detection result of the member state (step S44; No), that is, when none of the abnormality of the reflectance, the opening F1, the burr F2, and the abnormality of the machining path is detected (step S44). When there is no abnormality in reflectance and abnormality in the processing path, and the opening F1 and the burr F2 are not detected), that is, when the quality abnormality of the member T is not detected, the process proceeds to step S60 and the first member T1 When the joining between the first member T2 and the second member T2 is completed (step S60; Yes), this process is completed and the joining between the first member T1 and the second member T2 is not completed (step S60; No). ), Returning to step S40, the irradiation of the optical LAB by the irradiation unit 32A and the imaging by the imaging unit 34 are continued. That is, when the abnormality of the member state is not detected, the arithmetic unit 38 does not generate the change information, and the manufacturing device 10 continues the joining without changing the operating conditions based on the change information.

When the reflectance is abnormal, the burr F2 is detected, the shape and insertion amount of the tool 18 are detected, and change information is generated, that is, when step S48 is reached and step S58 is reached, as described above. The joining between the first member T1 and the second member T2 is stopped. Therefore, when the step S58 is reached through the step S48 and the joining is not completed (step S60; No), the arithmetic unit 38 connects the manufacturing apparatus 10 with the first member T1 and the second member T2. Resume joining with. Specifically, the arithmetic unit 38 transmits a joining restart signal instructing the manufacturing device 10 to restart the joining of the first member T1 and the second member T2 to the control device 30. When the control device 30 receives the joining restart signal from the arithmetic unit 38 and the joining between the first member T1 and the second member T2 is not completed, the control device 30 rotates the tool 18 in the θZ direction and moves in the + Z axis direction. Move and resume joining under the operating conditions changed based on the change information. Similarly, when the shape of the tool 18 is detected because the burr F2 is not detected but the reflectance is abnormal, that is, when the step S58 is reached through the step S52, the first member T1 and the second member The connection with T2 is stopped. Therefore, even when the step S58 is reached through the step S52 and the joining is not completed (step S60; No), the arithmetic unit 38 has the first member T1 and the second member T1 and the second member T1 in the manufacturing apparatus 10. The joining with the member T2 is restarted. On the other hand, when the abnormality of the reflectance is not detected, the shape of the tool 18 is not detected, so that the joining is not stopped. In this case, the control device 30 changes the operating conditions based on the change information received from the arithmetic unit 38 during the joining, and causes the control device 30 to perform the joining under the changed operating conditions. As described above, in the present embodiment, the arithmetic unit 38 determines whether the reflectance detected during joining is abnormal, and if it is abnormal, stops the joining and detects the shape of the tool 18. On the other hand, the arithmetic unit 38 does not stop the joining and does not detect the shape of the tool 18 even if it determines that the joining conditions other than the reflectance are abnormal. However, the arithmetic unit 38 may stop the joining and detect the shape of the tool 18 even when it is determined that the joining conditions other than the reflectance are abnormal. Further, in the present embodiment, in the detection of the tool state, the insertion amount of the tool 18 is detected during joining, and the shape of the tool 18 is detected after the joining is stopped. For example, the tool 18 is detected after the joining is stopped. The insertion amount of may be detected. Summarizing the above, the arithmetic unit 38 may determine whether the member state during joining is abnormal, and detect the tool state when the member state is abnormal. The member state detected by the detection unit 52 may be at least one of the reflectance, the opening F1, the burr F2, and the machining path, and the tool state is at least one of the insertion amount of the tool 18 and the shape of the tool 18. It may be one. In the example of FIG. 17, it is determined that the reflectance abnormality and the burr F2 are detected (step S46), the reflectance abnormality is determined (step S48), and the burr F2 detection determination (step S50) is performed in this order. Although it is performed, the order of detecting the tool state is not limited to this and is arbitrary. That is, the order of determining the detection of the abnormality of the reflectance, the opening F1, the burr F2, and the abnormality of the processing path may be arbitrary.

Next, the generation of change information will be described. The change information generation unit 56 changes the type of change information to be generated according to the member state and the tool state that the detection unit 52 determines to be abnormal. Further, it can be said that the change information generation unit 56 changes the type of change information to be generated according to the defective mode of the joint state determined by the joint state determination unit 54. The change information is information for changing the operating conditions of the manufacturing apparatus 10, in other words, information including the changed contents of the operating conditions. Since the operating conditions are the rotation speed of the tool 18, the feed rate of the tool 18, the insertion amount of the tool 18, and the machining path of the tool 18, the change information includes information for changing the rotation speed of the tool 18. At least one of information for changing the feed rate of the tool 18, information for changing the insertion amount of the tool 18, and information for changing the machining path of the tool 18 can be mentioned. The change information may be information that specifies the amount of change in the operating conditions (for example, information that increases the rotation speed of the tool 18 by 100 rpm), or information that specifies the changed operating conditions (for example, the rotation of the tool 18). Information to change the number to 1000 rpm) may be used. That is, the change information may be information for changing the operating condition from the currently set first value to the second value.

Here, when the quality abnormality of the member T occurs, it may be possible to suppress the quality failure by changing the operating conditions of the manufacturing apparatus 10. However, it may be difficult to determine which operating condition is to be changed in order to suppress quality defects. Taking as an example the case where an abnormality in the reflectance of the surface T3A of the joint portion T3 is detected as a quality abnormality, the abnormality in the reflectance may be formed due to a poor mode of the joint state called poor mixing. , It may be formed due to a defective mode of joint state such as poor heat input. Therefore, even if the operating conditions that can eliminate the insufficient heat input are changed for the reflectance abnormality caused by the mixing failure, the mixing failure cannot be solved and the reflectance abnormality may not be solved. There is. On the other hand, the arithmetic unit 38 according to the present embodiment is an operation capable of identifying the defective mode of the joining state and eliminating the defective mode of the joining state by detecting the member state and the tool state by the detection unit 52. Generate change information to change the condition. Therefore, according to the arithmetic unit 38 according to the present embodiment, it is possible to suppress the quality defect of the member T and improve the quality of the member T. Hereinafter, the generation of change information will be specifically described.

18 to 21 are flowcharts illustrating the generation of change information. FIG. 18 shows a method of generating change information when an abnormality in the reflectance of the surface T3A of the joint portion T3 is detected as an abnormality in the member state. That is, FIG. 18 illustrates the details of the method of generating change information when an abnormality in reflectance is detected in step S58 of FIG. When the detection unit 52 determines that the reflectance of the surface T3A of the joint portion T3 is abnormal, the detection unit 52 detects both the groove size of the tool 18 and the outer diameter of the tool 18 as shown in FIG. When it is determined that is abnormal (step S62; Yes), the joint state determination unit 54 determines that a mixing abnormality and insufficient heat input are occurring among the defective modes of the joint state (step S64). .. When the change information generation unit 56 determines that the joining state determination unit 54 has caused a mixing abnormality and insufficient heat input, the change information generation unit 56 increases the rotation speed of the tool 18 and reduces the feed rate of the tool 18. The change information to the effect is generated (step S66). In other words, the change information generation unit 56 determines that the detection unit 52 has an abnormality in the reflectance of the surface T3A of the joint portion T3, and both the groove size of the tool 18 and the outer diameter of the tool 18 are abnormal. When it is determined that the above is true, change information for increasing the rotation speed of the tool 18 and change information for decreasing the feed rate of the tool 18 are generated.

Further, when the detection unit 52 determines that both the groove size of the tool 18 and the outer diameter of the tool 18 are not abnormal (step S62; No) (the groove size of the tool 18 is abnormal). And when the condition that the outer diameter of the tool 18 is abnormal is not satisfied), when the detection unit 52 determines that the groove size of the tool 18 is abnormal (step S68; Yes), the joint state determination unit In 54, it is determined that a mixing abnormality has occurred in the defective mode of the joining state (step S70). That is, when the joint state determination unit 54 determines that the reflectance of the surface T3A of the joint portion T3 is abnormal and the groove size of the tool 18 is abnormal, the joint state determination unit 54 mixes the mixture. Judge that a misalignment has occurred. When the joint state determination unit 54 determines that a mixing abnormality has occurred, the change information generation unit 56 generates change information to the effect that the rotation speed of the tool 18 is increased (step S72). In other words, when the change information generation unit 56 determines that the reflectance of the surface T3A of the joint portion T3 is abnormal and the detection unit 52 determines that the groove size of the tool 18 is abnormal, the change information generation unit 56 determines that the reflectance is abnormal. Generates change information to increase the rotation speed of the tool 18.

Further, when the detection unit 52 determines that the groove size of the tool 18 is not abnormal (step S68; No), and the detection unit 52 determines that the outer diameter of the tool 18 is abnormal (step S74). Yes), the joint state determination unit 54 determines that insufficient heat input has occurred in the defective mode of the joint state (step S76). That is, when the joint state determination unit 54 determines that the reflectance of the surface T3A of the joint portion T3 is abnormal and the outer diameter of the tool 18 is abnormal, the heat input unit 54 receives heat. Judge that there is a shortage. When the joint state determination unit 54 determines that the heat input is insufficient, the change information generation unit 56 generates change information to the effect that the feed rate of the tool 18 is reduced (step S72). In other words, when the detection unit 52 determines that the reflectance of the surface T3A of the joint portion T3 is abnormal and the outer diameter of the tool 18 is abnormal, the change information generation unit 56 determines that the tool is abnormal. Generates change information to the effect that the feed rate of 18 is reduced. When the arithmetic unit 38 determines that the groove size of the tool 18 is not abnormal and the outer diameter of the tool 18 is not abnormal (step S74; No), this process ends.

In this way, the arithmetic unit 38 determines whether the reflectance of the surface T3A of the joint portion T3 is abnormal by the detection unit 52. Then, the arithmetic unit 38 determines whether or not an abnormality has occurred in the tool state, here, the dimension of the groove of the tool 18 and the outer diameter of the tool 18. If the groove size of the tool 18 is abnormal, the outer peripheral surface of the probe 22 is worn and the groove is shallow (the length between the convex portion and the concave portion is short), so that the mixture is mixed. Anomalies may occur and anomalies in the reflectance of the surface T3A may occur. More specifically, if the groove of the tool is shallow, the probe 22 may insufficiently agitate the first member T1 and the second member T2, resulting in a mixing abnormality. Therefore, when the size of the groove of the tool 18 is abnormal, the arithmetic unit 38 determines that a mixing abnormality has occurred, and increases the rotation speed of the tool 18 in order to suppress the mixing abnormality. The change information is generated and transmitted to the control device 30. The control device 30 increases the rotation speed of the tool 18 according to the change information. By increasing the rotation speed of the tool 18, the control device 30 sufficiently agitates the first member T1 and the second member T2 by the probe 22, suppresses the mixing abnormality, and enters an abnormal mixing state. , Can be suppressed appropriately.

Further, when the outer diameter of the tool 18 is abnormal, there is a possibility that the reflectance is abnormal due to insufficient heat input due to the tool 18 being worn and the outer diameter being reduced. .. When the outer diameter of the tool 18, here the outer diameter of the shoulder 20 is short, the contact area between the shoulder 20 and the member T becomes smaller, and the frictional heat to the member T (here, the member into which the rotated tool 18 is inserted) is generated. It is judged that the amount of heat transfer has decreased and the heat input is insufficient. Therefore, when the outer diameter of the tool 18 is abnormal, the arithmetic unit 38 determines that insufficient heat input has occurred, and reduces the feed rate of the tool 18 in order to suppress the insufficient heat input. Information is generated and transmitted to the control device 30. The control device 30 reduces the feed rate of the tool 18 according to the change information. By reducing the feed rate of the tool 18, the control device 30 can prolong the contact time between the tool 18 and the member T (that is, the time for friction stir welding) to increase the amount of heat input, thereby reducing the heat input shortage. It can be suppressed to appropriately suppress the abnormal mixing state.

FIG. 19 shows a method of generating change information when the opening F1 is detected as an abnormality in the member state. That is, FIG. 19 illustrates the details of the method of generating change information when the opening F1 is detected in step S58 of FIG. As shown in FIG. 19, when the detection unit 52 detects the opening F1, the joint state determination unit 54 determines that insufficient heat input has occurred in the defective mode of the joint state (step S82). When the joint state determination unit 54 determines that the heat input is insufficient, the change information generation unit 56 generates change information to the effect that the feed rate of the tool 18 is reduced (step S84). In other words, the change information generation unit 56 generates change information to the effect that the feed rate of the tool 18 is reduced when the detection unit 52 detects the opening F1.

In this way, the arithmetic unit 38 detects whether the opening F1 is formed on the surface T3A of the joint portion T3 by the detection unit 52. When the opening F1 is formed, the arithmetic unit 38 determines that insufficient heat input has occurred, and generates change information that the feed rate of the tool 18 is reduced in order to suppress the insufficient heat input. , Is transmitted to the control device 30. The control device 30 reduces the feed rate of the tool 18 according to the change information. By reducing the feed rate of the tool 18, the control device 30 can prolong the contact time between the tool 18 and the member T to increase the amount of heat input, suppress the insufficient heat input, and open the opening F1. Formation can be suppressed appropriately.

FIG. 20 shows a method of generating change information when the burr F2 is detected as an abnormality in the member state. That is, FIG. 20 illustrates the details of the method of generating change information when the burr F2 is detected in step S58 of FIG. When the detection unit 52 detects the burr F2, as shown in FIG. 20, when the detection unit 52 determines that the insertion amount of the tool 18 is too large and abnormal (step S92; Yes), the change information The generation unit 56 determines that excessive heat input has occurred in the defective mode of the joint state (step S94). When it is determined that the insertion amount of the tool 18 is abnormal and excessive heat input occurs, the change information generation unit 56 generates change information to the effect that the insertion amount of the tool 18 is reduced (step S96). In other words, the change information generation unit 56 generates change information to the effect that the insertion amount of the tool 18 is reduced when the detection unit 52 detects the burr F2 and determines that the insertion amount of the tool 18 is abnormal. To do.

Further, when the detection unit 52 determines that the insertion amount of the tool 18 is not abnormal (step S92; No), the joint state determination unit 54 determines that excessive heat input has occurred in the defective mode of the joint state. Determine (step S98). The change information generation unit 56 generates change information to the effect that the feed rate of the tool 18 is increased when it is determined that the insertion amount of the tool 18 is normal although the heat input is excessive (step S100). In other words, the change information generation unit 56 generates change information to increase the feed rate of the tool 18 when the detection unit 52 detects the burr F2 and determines that the insertion amount of the tool 18 is normal. To do.

In this way, the arithmetic unit 38 determines whether the burr F2 is formed by the detection unit 52, and detects that the metal structure may be coarsened due to excessive heat input. Then, the arithmetic unit 38 determines whether the tool state, here, the insertion amount of the tool 18 is abnormal, or more specifically, whether the insertion amount of the tool 18 is too large. When the insertion amount of the tool 18 is too large, the contact area between the tool 18 and the member T becomes large due to the insertion amount of the tool 18 being too large (in other words, the pressing pressure of the tool 18 against the member T becomes high). There is a possibility that excessive heat input has occurred and the metal structure has become coarse. Therefore, when the insertion amount of the tool 18 is abnormal, the arithmetic unit 38 generates change information that the insertion amount of the tool 18 is reduced in order to suppress excessive heat input, and transmits the change information to the control device 30. .. The control device 30 reduces the insertion amount of the tool 18 according to the change information. By reducing the insertion amount of the tool 18, the control device 30 can reduce the contact area between the tool 18 and the member T to reduce the amount of heat input, suppress the excessive heat input, and suppress the excessive heat input of the metal structure. Coarseness can be appropriately suppressed. Further, when the burr F2 is formed but the insertion amount of the tool 18 is normal, the arithmetic unit 38 increases the feed rate of the tool 18 in order to eliminate the excessive heat input, and the heat input amount. Judge that it is necessary to reduce. Therefore, when the insertion amount of the tool 18 is normal, the arithmetic unit 38 generates change information that increases the feed rate of the tool 18 in order to suppress excessive heat input, and transmits the change information to the control device 30. To do. The control device 30 increases the feed rate of the tool 18 according to the change information. By increasing the feed rate of the tool 18, the control device 30 can shorten the contact time between the tool 18 and the member T to reduce the amount of heat input, suppress the excessive heat input, and suppress the excessive heat input of the metal structure. Coarseness can be appropriately suppressed.

FIG. 21 shows a method of generating change information when an abnormality in the machining path is detected as an abnormality in the member state. That is, FIG. 21 illustrates the details of the method of generating change information when an abnormality in the machining path is detected in step S58 of FIG. As shown in FIG. 21, when the detection unit 52 detects an abnormality in the machining path, the joint state determination unit 54 determines that a heat input position defect has occurred in the defective mode of the joint state (step S112). .. When the joint state determination unit 54 determines that the heat input position defect has occurred, the change information generation unit 56 generates change information to correct the machining path of the tool 18 (step S114). In other words, the change information generation unit 56 generates change information to correct the machining path of the tool 18 when the detection unit 52 detects an abnormality in the machining path. The change information generation unit 56 is changed to move the machining path of the tool 18 in the −X axis direction, for example, when the machining path detected by the detection unit 52 deviates from the reference machining path in the + X axis direction. Information is generated, and when the machining path detected by the detection unit 52 deviates from the reference machining path in the −X axis direction, change information indicating that the machining path of the tool 18 is moved in the + X axis direction is generated. .. That is, the change information generation unit 56 generates change information so that the processing path detected by the detection unit 52 moves in the direction opposite to the direction deviating from the reference processing path.

As described above, the arithmetic unit 38 may determine whether the machining path is abnormal by the detection unit 52, so that the thickness of the intermetallic compound between the first member T1 and the second member T2 may be increased. Is detected. The arithmetic unit 38 determines that the cause of the thickening of the intermetallic compound is caused by the defective heat input position, and corrects the processing path in order to eliminate the defective heat input position. The change information is generated and transmitted to the control device 30. The control device 30 corrects the machining path of the tool 18 according to the change information. By modifying the machining path, the control device 30 can modify the heat input position and reduce the thickness of the intermetallic compound between the first member T1 and the second member T2.

In the above description, the arithmetic unit 38 extracts the defective mode of the joint state by the joint state determination unit 54 based on the detection result of the member state and the tool state by the detection unit 52, and changes information generation unit 56. The change information is generated by. However, the process of extracting the defective mode of the joint state by the joint state determination unit 54 is not indispensable, and the arithmetic unit 38 detects the member state and the tool state by the detection unit 52 without extracting the defective mode of the joint state. Based on the result, the change information generation unit 56 may generate the change information. In this case, the joint state determination unit 54 of the arithmetic unit 38 becomes unnecessary.

Further, in the description of FIG. 17, if there is no abnormality in the detection result of the member state, the change information is not generated. However, even if there is no abnormality in the detection result of the member state, the arithmetic unit 38 operates at the next joining based on the member state detected during the joining between the first member T1 and the second member T2. Change information may be generated to change the condition. The next joining here refers to the joining of the next members by the manufacturing apparatus 10 after the joining of the first member T1 and the second member T2 when the detection unit 52 detects the member state is completed. .. In the next joining, members different from the first member T1 and the second member T2 when the detection unit 52 detects the member state may be joined.

The arithmetic unit 38 determines whether it is necessary to change the operating conditions in the next joining based on the member state detected during the joining of the first member T1 and the second member T2. Even when the joining is completed without determining that the member state detected during the joining is abnormal, the arithmetic unit 38 needs to change the operating conditions in the next joining for the member state detected during the joining. If it is defective to some extent, it is judged that it is necessary to change the operating conditions at the next joining. When the arithmetic unit 38 determines that it is necessary to change the operating conditions in the next joining, it generates change information to the effect that the operating conditions in the next joining are changed based on the member state and the tool state. , Is transmitted to the control device 30. The control device 30 changes the operating conditions at the time of the next joining based on the change information transmitted from the arithmetic unit 38.

For example, the arithmetic unit 38 has a unit region in which the difference between the reflection ratio detected by the detection unit 52 and the reference reflection ratio is equal to or greater than a predetermined value even when the joining is completed without determining that the reflectance is abnormal. When the number is equal to or more than a predetermined threshold value (hereinafter referred to as the second reflectance threshold value), it is determined that it is necessary to change the operating conditions at the time of the next joining. The second reflectance threshold here is a numerical value smaller than the first reflectance threshold when the abnormality of the reflectance is judged (that is, the number of pixels in which the difference between the reflectance ratio and the reference reflectance ratio is equal to or more than a predetermined value. Is preset). When the arithmetic unit 38 determines that it is necessary to change the operating conditions at the time of the next joining, the second imaging unit 34B is made to image the tool 18, and the detection unit 52 is used to image the tool 18 by the second imaging unit 34B. The shape of the tool 18 is detected based on the image of. The arithmetic unit 38 states that if the outer diameter of the tool 18 detected by the detection unit 52 is equal to or greater than a predetermined threshold value (hereinafter referred to as the second outer diameter threshold value), insufficient heat input may occur. Judging, the change information that the feed speed of the tool 18 is reduced at the next joining is generated. The second outer diameter threshold value here is set in advance as a numerical value larger than the first outer diameter threshold value when the abnormality of the outer diameter of the tool 18 is determined (that is, on the assumption that the outer diameter of the tool 18 is larger). Set. Further, the arithmetic unit 38 states that if the groove of the tool 18 detected by the detection unit 52 is equal to or larger than a predetermined threshold value (hereinafter referred to as a second groove threshold value), a mixing abnormality may occur. Judging, the change information that the rotation speed of the tool 18 is increased at the time of the next joining is generated. The second groove threshold value here is set in advance as a numerical value larger than the first groove threshold value when the abnormality of the groove of the tool 18 is determined (that is, assuming that the groove depth of the tool 18 is deeper). Will be done.

Further, for example, even when the arithmetic unit 38 completes the joining without determining that the opening F1 is formed, the low-luminance pixel detected by the detection unit 52 has a predetermined threshold value (hereinafter, the second opening threshold value). If they are adjacent to each other continuously, it is judged that it is necessary to change the operating conditions at the time of the next joining. The second aperture threshold value here is set in advance as a numerical value smaller than the first aperture threshold value when the aperture F1 is detected (that is, on the assumption that the number of consecutively adjacent low-luminance pixels is smaller). To. If the arithmetic unit 38 determines that it is necessary to change the operating conditions at the time of the next joining, it determines that insufficient heat input may occur, and determines that the feed rate of the tool 18 is set at the time of the next joining. Generate change information to the effect that it will decrease.

Further, for example, even when the arithmetic unit 38 completes the joining without determining that the burr F2 is formed, the high-luminance pixel detected by the detection unit 52 has a predetermined threshold value (hereinafter, the second burr threshold value). If they are adjacent to each other continuously, it is judged that it is necessary to change the operating conditions at the time of the next joining. The second burr threshold value here is set in advance as a numerical value smaller than the first burr threshold value when the burr F2 is detected (that is, on the assumption that the number of consecutively adjacent high-luminance pixels is smaller). To. The arithmetic unit 38 determines that excessive heat input may occur when the insertion amount of the tool 18 detected by the detection unit 52 is equal to or greater than a predetermined threshold value (hereinafter referred to as the second insertion threshold value). Then, change information is generated to the effect that the insertion amount of the tool 18 is reduced at the next joining. The second insertion threshold value here is set in advance as a numerical value larger than the first insertion threshold value when the abnormality of the insertion amount of the tool 18 is determined (that is, on the assumption that the insertion amount of the tool 18 is larger). To. Further, the arithmetic unit 38 determines that if the insertion amount of the tool 18 detected by the detection unit 52 is equal to or greater than the second insertion threshold value, a mixing abnormality may occur, and at the time of the next joining. Generates change information to the effect that the feed rate of the tool 18 is increased.

Further, for example, in the arithmetic unit 38, even when the joining is completed without detecting the abnormality of the machining path, the difference between the machining path detected by the detection unit 52 and the reference machining path is a predetermined threshold value (hereinafter, the second If it exceeds the 2-pass threshold value), it is judged that it is necessary to correct the machining path at the time of the next joining. The second pass threshold value here is set in advance as a numerical value smaller than the first pass threshold value when the machining pass is detected (that is, on the assumption that the difference between the machining pass and the reference machining path is smaller). .. If the arithmetic unit 38 determines that it is necessary to change the operating conditions at the time of the next joining, it determines that there is a possibility that a heat input position defect may occur, and corrects the machining path at the time of the next joining. Generate change information to that effect.

Further, when the arithmetic unit 38 detects that the quality of the member T is so bad that it exceeds the degree of quality abnormality and causes a quality defect, the joining by the manufacturing apparatus 10 may be terminated. For example, the arithmetic unit 38 determines that the member T is of poor quality when the detection result of the member state by the detection unit 52 is outside the range of a threshold value stricter than the threshold value for determining the abnormality of the member state. Then, the joining by the manufacturing apparatus 10 is completed. That is, for example, when the number of unit regions in which the difference between the reflection ratio detected by the detection unit 52 and the reference reflection ratio is equal to or greater than a predetermined value is equal to or greater than a threshold value (hereinafter referred to as a third reflectance threshold value), the member It may be determined that T is of poor quality and the joining may be terminated. The third reflectance threshold is a numerical value larger than the first reflectance threshold when the abnormality of the reflectance is judged (that is, the number of pixels in which the difference between the reflectance ratio and the reference reflectance ratio is equal to or more than a predetermined value is larger. (Assuming), it is set in advance. Further, for example, when the low-luminance pixels detected by the detection unit 52 are continuously adjacent to each other by a predetermined threshold value (hereinafter referred to as a third aperture threshold value) or more, it is determined that the member T is of poor quality. The joining may be terminated. The third aperture threshold is set in advance as a numerical value larger than the first aperture threshold when the aperture F1 is detected (that is, on the assumption that the number of consecutively adjacent low-luminance pixels is larger). Further, for example, when the high-luminance pixels detected by the detection unit 52 are continuously adjacent to each other by a predetermined threshold value (hereinafter referred to as a third burr threshold value) or more, it is determined that the member T is of poor quality. The joining may be terminated. The third burr threshold value is set in advance as a numerical value larger than the first burr threshold value when the burr F2 is detected (that is, on the assumption that the number of consecutively adjacent high-luminance pixels is larger). Further, for example, when the difference between the machining path detected by the detection unit 52 and the reference machining path is equal to or greater than a predetermined threshold value (hereinafter referred to as a third pass threshold value), it is determined that the member T is of poor quality. Then, the joining may be completed. The third pass threshold value is set in advance as a numerical value larger than the first pass threshold value when the machining pass is detected (that is, on the assumption that the difference between the machining pass and the reference machining path is larger).

As described above, the arithmetic unit 38 according to the first embodiment is used in the manufacturing apparatus 10 (friction stir welding apparatus), and includes a detection unit 52 and a change information generation unit 56. The detection unit 52 detects a member state that is at least a part of the first member T1, the second member T2, and the joint T3, and a tool state that is the state of the tool 18. The change information generation unit 56 generates change information for changing the operating condition of the manufacturing apparatus 10 from the first value to the second value based on the member state and the tool state detected by the detection unit 52. ..

In the manufacturing apparatus 10, poor quality of the member T may be caused by a poor joint state between the first member T1 and the second member T2. On the other hand, since quality may be improved by resetting the operating conditions of the manufacturing apparatus 10, it is required to reset the operating conditions so as to suppress the occurrence of quality defects. According to the arithmetic unit 38 according to the present embodiment, by detecting the member state and the tool state, it is possible to detect what kind of defective joint state is occurring, that is, what kind of quality abnormality is occurring. Therefore, by generating change information based on the member state and the tool state, the manufacturing apparatus 10 can appropriately set operating conditions that can suppress the occurrence of quality defects. Moreover, even if the quality abnormality is the same, it may be caused by different causes. On the other hand, according to the arithmetic unit 38 according to the present embodiment, by detecting the member state and the tool state, it is possible to recognize the cause of the poor joining state, that is, the quality abnormality with high accuracy, and the quality is poor. It can contribute appropriately by suppressing.

Further, the detection unit 52 detects the member state during the joining by the manufacturing apparatus 10. The change information generation unit 56 generates change information for changing the operating conditions during joining based on the member state during joining detected by the detection unit 52. The quality abnormality may be caused by the internal structure of the member T. In this case, it is possible to detect a quality abnormality by analyzing the internal structure after the joining is completed and to re-establish the operating conditions. However, it takes time to determine the change in the operating conditions, and the yield of joining deteriorates. On the other hand, the arithmetic unit 38 according to the present embodiment detects the member state, that is, the quality abnormality of the member T during the joining between the first member T1 and the second member T2, and changes the operating conditions during the joining. Generate change information for. Since the control device 30 can change the operating conditions during joining by using this change information, it is possible to shorten the time until the operating conditions are changed, and it is possible to improve the yield of joining.

Further, the change information generation unit 56 changes the operating conditions when the joining is completed when the detection unit 52 detects the member state and the next joining is performed, based on the member state during joining detected by the detection unit 52. Generate change information to do. By generating the change information for changing the operating condition of the next joining in this way, the arithmetic unit 38 according to the present embodiment does not change the operating condition without being judged as a quality abnormality during joining, for example. Even in this case, it is possible to detect a sign that a quality defect occurs and suppress the occurrence of a quality defect in the next joining.

Further, the detection unit 52 determines whether the detection result of the member state during joining by the manufacturing apparatus 10 is abnormal, and if it is abnormal, detects the tool state. The arithmetic unit 38 according to the present embodiment can more accurately identify the cause of the quality abnormality by detecting the tool state when the detection result of the member state is abnormal.

Further, the joint state determination unit 54 determines whether the joint state between the first member T1 and the second member T2 is defective based on the detection result of the member state and the tool state by the detection unit 52. By determining the defect in the joining state, the arithmetic unit 38 according to the present embodiment can recognize the cause of the quality defect with high accuracy even if the same quality defect is caused by different causes. It becomes possible and can contribute appropriately by improving quality.

The joint state determination unit 54 determines at least one of the mixed state of the first member T1 and the second member T2, the amount of heat input to the member T, and the heat input position to the member T as the joint state. .. The arithmetic unit 38 according to the present embodiment can recognize the cause of the quality abnormality with high accuracy by determining the mixed state, the amount of heat input, the heat input position, and the like as the joint state, and is more appropriate for quality improvement. Can contribute to.

The change information generation unit 56 operates at least one of the rotation speed of the tool 18, the feed rate of the tool 18, the amount of the tool 18 inserted into the member T, and the machining path which is the movement path of the tool with respect to the member T. To generate change information. The arithmetic unit 38 according to the present embodiment appropriately suppresses quality defects by generating change conditions using the rotation speed of the tool 18, the feed rate of the tool 18, the insertion amount of the tool 18, the machining path, and the like as operating conditions. Can be assisted.

The arithmetic system 12 according to the present embodiment has an imaging unit 34 (second imaging unit 34B in the first embodiment) that images the member T and the tool 18. The detection unit 52 detects the member state and the tool state based on the captured image of the imaging unit 34. Since the calculation system 12 detects the member state and the tool state based on the image captured by the imaging unit 34, the member state and the tool state can be detected with high accuracy, and the suppression of quality defects can be appropriately assisted.

The imaging unit 34 captures an image of the detection region AR2 (the surface TA of the member T and the surface of the tool 18). The detection unit 52 detects the member state and the tool state based on the image of the detection area AR2. The detection unit 52 can detect the member state and the tool state with high accuracy by detecting the member state and the tool state based on the image of the detection area AR2, and can appropriately assist in suppressing the quality defect.

The detection unit 52 detects whether the light reflectance of the surface T3A of the joint portion T3 is abnormal as a member state, and when it detects that the reflectance is abnormal, it determines the shape of the tool 18 as a tool state. To detect. Since the calculation system 12 detects the shape of the tool 18 when the light reflectance of the joint portion T3 is abnormal, the cause of the quality abnormality is determined based on the light reflectance of the joint portion T3 and the shape of the tool 18. It can be detected with high accuracy.

The detection unit 52 detects whether the dimension of the groove of the probe 22 of the tool 18 and the outer diameter of the tool 18 are abnormal as the tool state. The change information generation unit 56 generates information to increase the rotation speed of the tool 18 as change information when the detection unit 52 detects that the groove size is abnormal. The change information generation unit 56 generates information to the effect that the feed rate of the tool 18 is reduced when the detection unit 52 detects that the outer diameter of the tool 18 is abnormal. The arithmetic system 12 generates change information to increase the rotation speed of the tool 18 when the light reflectance of the joint portion T3 is abnormal and the groove size is abnormal, so that the manufacturing apparatus 10 It is possible to appropriately suppress abnormalities in mixing with light and appropriately assist in suppressing quality defects.

The detection unit 52 detects the burr F2 around the joint portion T3 as a member state, and when the burr F2 is detected, detects the amount of the tool 18 inserted into the member T as a tool state. Since the calculation system 12 detects the insertion amount of the tool 18 when the burr F2 is detected, the operating conditions to be changed can be appropriately selected based on the insertion amount of the burr F2 and the tool 18.

The detection unit 52 detects whether the insertion amount of the tool 18 is abnormal. The change information generation unit 56 generates information to the effect that the insertion amount of the tool 18 is reduced when the detection unit 52 detects that the insertion amount of the tool 18 is abnormal. When the detection unit 52 detects that the insertion amount of the tool 18 is not abnormal, the change information generation unit 56 generates information to increase the feed rate of the tool 18 as change information. When the burr F2 is detected and the insertion amount of the tool 18 is abnormal, the calculation system 12 generates change information to reduce the insertion amount of the tool 18, thereby causing excessive heat input to the manufacturing apparatus 10. It can be appropriately suppressed to appropriately assist the suppression of quality defects. When the burr F2 is detected and the insertion amount of the tool 18 is normal, the arithmetic system 12 generates change information to increase the feed rate of the tool 18, thereby causing excessive heat input to the manufacturing apparatus 10. It can be appropriately suppressed to appropriately assist the suppression of quality defects.

The detection unit 52 detects the opening F1 formed in the joint portion T3 as a member state. The change information generation unit 56 generates information to the effect that the feed rate of the tool 18 is reduced when the detection unit 52 detects the opening F1 as change information. When the opening F1 is detected, the arithmetic system 12 appropriately suppresses the heat input shortage in the manufacturing apparatus 10 by generating change information to the effect that the feed rate of the tool 18 is lowered, and suppresses quality defects. Can be properly assisted.

The detection unit 52 detects whether the machining path in the member T is abnormal as the member state. When the detection unit 52 detects that the machining path is abnormal, the change information generation unit 56 generates information to the effect that the machining path is changed as change information. When the change information generation unit 56 detects an abnormality in the machining path, it generates change information to the effect that the machining path is changed, thereby appropriately suppressing the heat input position defect in the manufacturing apparatus 10 and causing a quality defect. Suppression can be appropriately assisted.

(Modification example)
Next, a modified example of the first embodiment will be described. 22 to 26 are schematic views illustrating a modified example of the first embodiment. First, an example of FIG. 22 will be described. In the first embodiment, the first imaging unit 34A separates the light LAB projected by the irradiation unit 32A onto the surface T3A of the junction T3 into the first wavelength light L1A and the second wavelength light L1B. Images of light L1A and L1B were imaged, respectively. However, the irradiation unit 32A is not an indispensable configuration. For example, as shown in the example of FIG. 22, the irradiation unit 32A may not be provided, and the first imaging unit 34A may image an image of the surface T3A of the joint portion T3 by natural light, light from a fluorescent lamp, or the like. In this case, as shown in FIG. 22, light (natural light, light from a fluorescent lamp, or the like) reflected by the surface T3A of the joint portion T3 is incident on the first imaging unit 34A as light L1. The first imaging unit 34A is provided with, for example, a filter 82 on the junction T3 side of the lens 70. The filter 82 is of the light of the first wavelength and the light of the second wavelength of the light of the wavelength band included in the light L1. It is a bandpass filter that transmits only light. Therefore, the filter 82 transmits the light L1AB including the light of the first wavelength and the light of the second wavelength among the incident light L1. The light transmitted through the filter 82. The L1AB proceeds in the same path as that described in the first embodiment (see FIG. 8), is separated into the light L1A of the first wavelength and the light L1B of the second wavelength, and is incident on the image pickup element 80. The filter 82 may be arranged on the traveling direction side of the light L1 with respect to the lens 70 with respect to the lens 70. The irradiation unit 32A may be provided with respect to the configuration of FIG. 22 (FIG. 8). In this case, the irradiation unit 32A uses the light in the wavelength band including the first wavelength and the second wavelength instead of the light source unit 60A that irradiates the light of the first wavelength and the light source unit 60B that irradiates the light of the second wavelength. A light source (for example, one mercury lamp or a xenon lamp) that irradiates the light may be used, and only the light of the first wavelength and the light of the second wavelength can be selectively extracted by the filter 82. When one light source that irradiates light in a wavelength band including a wavelength and a second wavelength is used, the photosynthesis unit 64 is unnecessary.

Next, an example of FIG. 23 will be described. In the first embodiment and the example of FIG. 22, the first image pickup unit 34A has a wavelength including the first light separation unit 72, the second light separation unit 74, the first light reflection unit 76, and the second light reflection unit 78. The light L1A and L1B were spatially separated by the separation portion and incident on the image sensor 80. However, as shown in the example of FIG. 23, the first imaging unit 34A does not include the first light separating unit 72, the second light separating unit 74, the first light reflecting unit 76, and the second light reflecting unit 78. May be good. In this case, as shown in FIG. 23, the first image pickup unit 34A may include, for example, filters 82A and 82B, filter turrets (not shown), lenses 70 and 79, and an image pickup device 80. The filter 82A is a bandpass filter that transmits only the light of the first wavelength among the light L1 incident on the first imaging unit 34A, and the filter 82B is the first of the light L1 incident on the first imaging unit 34A. It is a bandpass filter that transmits only two wavelengths of light. In the first imaging unit 34A, the filters 82A and 82B are attached to a filter turret (not shown), and one of the filters 82A and 82B can be selectively arranged on the optical path of the light L1. The first imaging unit 34A switches between a first filter mode in which the filter 82A is arranged on the optical path of the light L1 and a second filter mode in which the filter 82B is arranged on the optical path of the light L1 in a time division manner. That is, the first imaging unit 34A switches between the first filter mode and the second filter mode at predetermined time intervals. In the first filter mode, the light L1 incident on the first image pickup unit 34A is selected (transmitted) by the filter 82A from the light L1A having the first wavelength and is incident on the image pickup element 80. Then, in the second filter mode, the light L1 incident on the first image pickup unit 34A is selected (transmitted) by the filter 82B from the light L1B having a second wavelength, and is incident on the image pickup element 80. Therefore, the first imaging unit 34A can image the image of the light L1A and the image of the light L1B, and can appropriately calculate the reflection ratio. The filter 82A, the filter 82B, and the filter turret (not shown) separate the light L1A of the first wavelength and the light L1B of the second wavelength from the light L1 containing the light of the first wavelength and the light of the second wavelength. Therefore, it can be said that it is a wavelength separator. This wavelength separation unit temporally separates the wavelength of light incident on the light receiving surface of the image sensor 80 into a first wavelength and a second wavelength.

Next, an example of FIG. 24A will be described. In the example of FIG. 24A, the first imaging unit 34A is not provided with the first light separating unit 72, the second light separating unit 74, the first light reflecting unit 76, and the second light reflecting unit 78, and is further provided with the filter 82. Absent. That is, in the example of FIG. 24A, the first image pickup unit 34A includes lenses 70 and 79 and an image pickup element 80. On the other hand, in the example of FIG. 24A, the irradiation unit 32A includes a light source unit 60, filters 84A and 84B, a filter turret (not shown), a lens 62, a light diffusing unit 66, and a lens 68. The light source unit 60 emits light L in a wavelength band including the first wavelength and the second wavelength. For example, the light source unit 60 is one light source that emits wide-band light such as an existing lamp (mercury lamp or xenon lamp) or SLD (superluminescence diode). The filter 84A is a bandpass filter that transmits only the light of the first wavelength of the light L emitted from the light source unit 60, and the filter 84B is the second wavelength of the light L emitted from the light source unit 60. It is a bandpass filter that transmits only the light of. In the irradiation unit 32A, the filters 84A and 84B are attached to a filter turret (not shown) and can be selectively arranged on the optical path of the light L. The irradiation unit 32A switches between a first filter mode in which the filter 84A is arranged on the optical path of the light L1 and a second filter mode in which the filter 84B is arranged on the optical path of the light L1 in a time division manner. That is, the irradiation unit 32A switches between the first filter mode and the second filter mode at predetermined time intervals. In the first filter mode, the light L emitted from the light source unit 60 is joined through the light diffusing unit 66 and the lens 68 after the light LA having the first wavelength is selected (transmitted) by the filter 84A. It is projected onto the surface T3A of the portion T3. In the first filter mode, the first image pickup unit 34A is incident with the light L1A which is the light LA projected on the surface T3A of the junction portion T3, and the image pickup element 80 images the image by the light L1A. In the second filter mode, the light L emitted from the light source unit 60 is incident on the filter 84B, the light LB of the second wavelength is transmitted through the light diffusing unit 66 and the lens 68, and the junction portion. It is projected onto the surface T3A of T3. In the second filter mode, the first image pickup unit 34A is incident with the light L1B, which is the light LB projected on the surface T3A of the junction portion T3, and the image pickup element 80 images the image by the light L1B. Therefore, the first imaging unit 34A can image the image of the light L1A and the image of the light L1B, and can appropriately calculate the reflection ratio. The filter 84A, the filter 84B, and the filter turret (not shown) separate the light LA of the first wavelength and the light LB of the second wavelength from the light L containing the light of the first wavelength and the light of the second wavelength. Therefore, it can be said that it is a wavelength separator. This wavelength separation unit temporally separates the wavelength of light projected on the surface T3A of the junction T3 (in other words, the wavelength of light incident on the light receiving surface of the image sensor 80) into a first wavelength and a second wavelength. ..

Next, an example of FIG. 24B will be described. In the first embodiment, the first imaging unit 34A images images of light having different wavelengths (an image of light L1A and an image of light L1B) and calculates the reflection ratio. As shown in FIG. 24B. In addition, the first imaging unit 34A may capture an image of light having one type of wavelength. In the example of FIG. 24B, the irradiation unit 32A includes a light source unit 60, a filter 84A, a lens 62, a light diffusing unit 66, and a lens 68. The light L emitted from the light source unit 60 is selected (transmitted) by the filter 84A from the light LA having the first wavelength, passes through the light diffusing unit 66 and the lens 68, and reaches the surface T3A of the junction portion T3. Be projected. Further, in the example of FIG. 24B, the first image pickup unit 34A includes lenses 70 and 79 and an image pickup element 80. In the first image pickup unit 34A, the light L1A, which is the light LA projected on the surface T3A of the joint portion T3, is incident, and the image pickup element 80 images the image by the light L1A. From the image data PA of the image of the light L1A captured by the first imaging unit 34A, the detection unit 52 uses the reflectance of the light of the first wavelength on the surface T3A of the junction T3 as the reflectance of the surface T3A of the junction T3. To detect. Then, the detection unit 52 calculates the difference between the detected reflectance and the reference reflectance, and the number of unit regions in which the difference between the detected reflectance and the reference reflectance is equal to or greater than a predetermined value is the first reflectance threshold value. If it is the above, it is judged that the reflectance is abnormal. The reference reflectance may be set arbitrarily, but for example, it may be set as a threshold value at which an abnormality in the reflectance, which is a target of quality abnormality, is likely to be detected. The irradiation unit 32A may project light having an arbitrary wavelength other than the first wavelength, and the first imaging unit 34A may image an image with light having an arbitrary wavelength.

Further, in the first embodiment, as shown in FIG. 7, at least a part of the detection region AR2 of the second imaging unit 34B is in the irradiation region AR0 of the irradiation unit 32A and the detection region AR1 of the first imaging unit 34A. It is superimposed. However, the detection region AR2 of the second imaging unit 34B does not have to overlap with the irradiation region AR0 of the irradiation unit 32A and the detection region AR1 of the first imaging unit 34A. By not superimposing the detection region AR2 on the irradiation region AR0, the light L1AB from the irradiation region AR0 is hardly incident on the second imaging unit 34B. In this case, the second imaging unit 34B does not have to provide the filter 92 (see FIG. 11) that absorbs the light (light of the first wavelength and the second wavelength) of the component of the light L1AB.

Further, in the first embodiment, the detection region AR2 imaged by the second imaging unit 34B is not provided with an irradiation unit that irradiates visible light, and the second imaging unit 34B detects by natural light or light such as a fluorescent lamp. The image of the region AR2 was imaged. However, an irradiation unit 32B that projects light L2a such as visible light may be provided in the detection region AR2. The irradiation unit 32B irradiates the light L2a toward the detection region AR2. In this case, the wavelength of the light emitted by the irradiation unit 32B is different from the first wavelength and the second wavelength. The second imaging unit 34B superimposes the light L2 irradiated from the irradiation unit 32B and reflected in the detection region AR2 and the light irradiated from the irradiation unit 32A and reflected in the detection region AR2 (that is, light L1A and light L1B). The light L2AB, which is the emitted light, is received. The second imaging unit 34B absorbs the components of the light L1A and the light L1B from the light L2AB by, for example, a filter, and images an image by the light L2 reflected by the detection region AR2. When the irradiation unit 32A and 32B are not distinguished, it can be said that the arithmetic system 12 includes an irradiation unit 32 that projects light (projected light) on the surface TA of the member T and the surface of the tool 18. .. The imaging unit 34 captures an image of the surface TA of the member T on which light is projected by the irradiation unit 32 and the surface of the tool 18.

Next, the examples of FIGS. 25 and 26 will be described. In the first embodiment, a plurality of image pickup units (here, two image pickup units 34A and a second image pickup unit 34B) are provided, but as shown in FIG. 25, for example, only one image pickup unit 34 is provided. May be provided. The imaging unit 34 in FIG. 25 performs imaging in the range of the imaging region R2, that is, the detection region AR2. The irradiation region AR0 of the irradiation unit 32A (region R0 of the irradiation unit 32A) and the detection region AR2 (imaging region R2 of the imaging unit 34) overlap each other. As shown in FIG. 26, the image pickup unit 34 of FIG. 25 includes the lens 70, the first light separation unit 72, the second light separation unit 74, and the first image pickup unit 34A of the first embodiment. It includes a 1-light reflecting unit 76, a second light-reflecting unit 78, a lens 79, and an image sensor 80. The light LAB emitted from the irradiation unit 32A is projected onto the irradiation region AR0 of the member T, that is, the detection region AR2 of the member T. The light L1AB reflected by the detection region AR2 is incident on the image pickup unit 34. The light L1B having the second wavelength of the light L1AB passes through the first light separation unit 72 and is incident on the light receiving surface 80B of the image sensor 80. On the other hand, the light L1A having the first wavelength of the light L1AB is reflected by the first light separation unit 72 and is incident on the light receiving surface 80A of the image pickup device 80. The image sensor 80 captures an image of the light L1A incident on the light receiving surface 80A and an image of the light L1B incident on the light receiving surface 80B. The detection unit 52 detects an abnormality in the reflectance in the detection region AR1 from the image taken by the light L1A and the image taken by the light L1B captured by the image pickup unit 34.

Further, the detection unit 52 is based on at least one of the images of the light L1A and the light L1B captured by the imaging unit 34, that is, the image data which is the image data of the detection region AR2 by the light L1A and the detection by the light L1B. Based on at least one of the image data which is the image data of the region AR2, the member state other than the abnormality of the reflectance, here, the opening F1, the burr F2, and the processing path are detected.

(Second Embodiment)
Next, the second embodiment will be described. The calculation system 12a of the manufacturing system 1a according to the second embodiment is different from the first embodiment in that the detection unit 52 detects the internal state of the member T as at least a part of the member states. In the second embodiment, the description of the parts having the same configuration as that of the first embodiment will be omitted.

FIG. 27 is a schematic view of the manufacturing system according to the second embodiment. The arithmetic system 12a according to the second embodiment includes an irradiation unit 32Aa, a first imaging unit 34Aa and a second imaging unit 34Ba as an imaging unit 34, and an arithmetic unit 38. In the calculation system 12a according to the second embodiment, the member T is irradiated with light La by the irradiation unit 32Aa to heat the member T, and an image by the light L1a which is infrared rays emitted from the member T is obtained by the first imaging unit 34Aa. To image. The detection unit 52 according to the second embodiment detects the internal state of the member T as the member state based on the image of the light L1a imaged by the first imaging unit 34Aa. It can be said that the arithmetic system 12a according to the second embodiment detects the internal state of the member T by active thermography. Hereinafter, a specific description will be given.

FIG. 28 is a schematic view illustrating an irradiation unit and an imaging unit according to the second embodiment. As shown in FIG. 28, the irradiation unit 32Aa irradiates the irradiation region AR0 (surface TA) of the member T with light La. The irradiation unit 32Aa irradiates the irradiation region AR0 with the light La so that the intensity of the light La changes periodically with the passage of time. The surface TA of the member T is heated by being irradiated with light La, and heat is transferred from the surface TA to the inside of the member T. Light L1a, which is infrared rays, is emitted from the heated surface TA. The intensity of the light L1a changes according to the temperature of the surface TA of the member T. Therefore, the intensity of the light L1a from the member T changes according to the intensity of the light La from the irradiation unit 32Aa. The light La is a laser light in the present embodiment, but is not limited to the laser light and may be any light such as a halogen lamp. Further, the irradiation unit 32Aa is not limited to irradiating the member T with light as long as the member T can be heated. For example, the member T may be irradiated with an electromagnetic wave other than light, or the member T is irradiated with ultrasonic waves. Alternatively, an eddy current may be applied to the member T. That is, it can be said that the irradiation unit 32Aa may be a heating unit that heats the member T by an arbitrary method.

The first image pickup unit 34Aa is a device capable of detecting the light La1 and taking an image by the light La1, and is an infrared image pickup device in the present embodiment. The imaging region R1a, which is the imaging range of the first imaging unit 34Aa, overlaps with at least a part of the surface TA of the member T, and more specifically, with at least a part of the surface T1A of the first member T1. It overlaps at least a part of the surface T2A of the second member T2 and at least a part of the surface T3A of the joint T3. Assuming that the region of the surface TA of the member T that overlaps with the imaging region R1a is the detection region AR1a, the first imaging unit 34Aa images the image of the detection region AR1a. The detection region AR1a is superimposed on the irradiation region AR0 on which the light La is irradiated by the irradiation unit 32A. Therefore, light L1a, which is infrared rays, is emitted from the detection region AR1a. It can be said that the first imaging unit 34Aa captures an image of the light L1a emitted from the detection region AR1a.

Further, the imaging region R2a, which is the imaging range by the second imaging unit 34Ba, overlaps with at least a part of the surface of the tool 18. Assuming that the region overlapping the imaging region R2a on the surface of the tool 18 is the detection region AR2a, the second imaging unit 34Ba captures an image of the detection region AR2a.

An example of the configuration of the irradiation unit 32Aa and the first imaging unit 34Aa will be described. FIG. 29 is a schematic view of the irradiation unit and the first imaging unit according to the second embodiment. As shown in FIG. 29, the irradiation unit 32Aa includes a light source unit 60a and lenses 62a and 68a. The light source unit 60a is a light source that irradiates light La. The light source unit 60a irradiates light La so that the intensity changes periodically with the passage of time. In the light La emitted from the light source unit 60a, the luminous flux diameter of the light La (that is, the luminous flux diameter of the laser beam) is enlarged in the lens 62a. The light La from the lens 62a is collimated in the lens 68a and irradiated to the irradiation region AR0 of the member T, that is, the detection region AR1a.

The first image pickup unit 34Aa includes a lens 79a and an image pickup element 80a. The light L1a emitted from the detection region AR1a is focused by the lens 79a and incident on the light receiving surface of the image sensor 80a. The light receiving surface of the image sensor 80a has a conjugate relationship with the detection region AR1a. The image sensor 80a is an infrared image sensor in this embodiment, and pixels that receive light L1a and detect the intensity of the light L1a are arranged in a matrix on the light receiving surface of the image sensor 80a. Therefore, the image sensor 80a can image an image by the light L1a emitted from the detection region AR1a.

Next, the detection unit 52 according to the second embodiment will be described. The detection unit 52 according to the second embodiment has, as a member state, a hole inside the member T, an opening F1 formed in the joint portion T3, a burr F2 formed around the joint portion T3, and a first. The thickness of the intermetallic compound between the member T1 and the second member T2 is detected.

The detection of the vacancies inside the member T as the internal state of the member T will be described. The vacancy here refers to a closed vacancy that is not open on the surface, and may be rephrased as a void. When a hole is formed, the member T breaks from the hole as a starting point, and the joint strength decreases. The detection unit 52 is based on the image taken by the first imaging unit 34Aa by the light L1a from the detection region AR1a, in other words, based on the image data Pa of the first imaging unit 34Aa, the vacancies inside the member T. More specifically, the vacancies inside the joint portion T3 are detected. The image data Pa refers to the data of the brightness value for each pixel of the image sensor 80a, and can be said to be the data of the intensity value of the light L1a for each unit area in the detection area AR1a.

Here, the intensity of the light L1a from the detection region AR1a will be described. 30 and 31 are graphs for explaining the intensity of light from the detection region. FIG. 30 describes an example in which no holes are formed inside the member T. The horizontal axis of FIG. 30A is time, and the vertical axis is the intensity of light La emitted from the irradiation unit 32Aa or the intensity of light L1a emitted from the surface TA (detection region AR1a) of the member T. .. As shown in FIG. 30A, assuming that the surface TA of the member T is irradiated with the light La of the intensity InA at the time tA, the intensity of the light L1a emitted from the detection region AR1a increases with the passage of time. , Waveform IA changes as shown. Specifically, the temperature of the surface TA rises significantly immediately after the time tA due to the irradiation of the light La of the intensity InA at the time tA. Then, the heat of the surface TA is gradually transferred to the inside of the member T, so that the temperature HA of the surface TA gradually decreases with the passage of time. When the heat of the surface TA is transferred to the bottom surface, which is the surface opposite to the surface TA of the member T, the heat transferred to the bottom surface is reflected by the bottom surface, and the heat from the bottom surface is transferred toward the surface TA. Therefore, the heat transfer from the bottom surface reaches the surface TA, and the temperature HA of the surface TA rises once. After that, the temperature HA of the surface TA decreases again. The intensity of the light L1a emitted from the detection region AR1a is proportional to the temperature of the detection region AR1a (the surface TA of the member T) (that is, the intensity of the light L1a indicates the temperature in the detection region AR1a). Therefore, when the light La of the intensity InA is irradiated at the time tA, the intensity of the light L1a from the detection region AR1a proportional to the temperature of the detection region AR1a (surface TA) is immediately after the time tA as shown in the waveform IA. The waveform becomes such that it rises significantly, gradually falls with the passage of time, rises once when the heat transfer from the bottom reaches the surface TA, and then falls again. That is, the waveform IA of the intensity of the light L1a from the detection region AR1a is the peak Q1 due to the temperature rise immediately after the irradiation of the light La of the intensity InA and the peak due to the heat transfer from the bottom surface at a time later than the peak Q1. It changes with the passage of time so as to include Q2.

In the present embodiment, as shown in FIG. 30A, the light La of the intensity InA is not irradiated once at the time tA, but the intensity of the surface TA of the member T changes periodically with the passage of time. Light La is continuously irradiated so as to be performed. As shown in FIG. 30 (B), when the intensity of the light La changes as shown in the waveform In, the temperature of the surface TA also changes periodically according to the change in the intensity of the light La with the passage of time. To do. Then, the waveform I of the intensity of the light L1a from the surface TA according to the passage of time is also cyclically according to the change of the intensity of the surface TA with the passage of time, that is, according to the change of the intensity of the light La with the passage of time. Change. The waveform I of the intensity of the light L1a from the surface TA is the waveform I1 (waveform corresponding to the peak Q1) indicating the temperature change due to the temperature rise immediately after the irradiation of the light La, and the temperature caused by the heat transfer from the bottom surface to the surface TA. A waveform obtained by synthesizing the waveform I2 (waveform corresponding to the peak Q2) indicating the change is obtained. Since the intensity of the light La changes as shown in the waveform In, the waveform I1 becomes a waveform corresponding to the waveform In (a sine wave in the example of FIG. 30). Therefore, the waveform I1 has the same period as the waveform In. Further, the waveform I2 is a temperature change caused by heat transferred from the surface TA of the member T to the bottom surface of the member T and returned to the surface TA from the bottom surface of the member T. Therefore, the waveform I2 is a waveform whose phase is delayed with respect to the waveform I1 by the time required for the heat of the surface TA to be transferred back and forth from the surface TA to the bottom surface of the member T, and the periods are substantially the same. In reality, the amount of heat transferred from the bottom surface of the member T tends to be small, so that the waveform I may be substantially the same as the waveform I1.

FIG. 31 describes an example in which a hole is formed inside the member T. As shown in FIG. 31 (A), assuming that the surface TA of the member T is irradiated with the light La of the intensity InA at the time tA, the intensity of the light L1a from the surface TA changes with the passage of time in the waveform IB. It changes as shown in. Specifically, the temperature of the surface TA rises significantly immediately after the time tA due to the irradiation of the light La of the intensity InA at the time tA. Then, the heat of the surface TA is gradually transferred to the inside of the member T, so that the temperature of the surface TA gradually decreases with the passage of time. Then, when the heat transferred from the surface TA to the inside of the member T reaches the vacancy, the heat transferred to the vacancy is reflected at the interface between the vacancy and the member T, and the heat from the interface between the vacancy and the member T is reflected. However, it is transmitted toward the surface TA. Therefore, the heat from the interface between the vacancies and the member T reaches the surface TA, and the temperature of the surface TA rises once. After that, the temperature of the surface TA decreases again, and then the temperature rises again due to heat transfer from the bottom surface of the member T, and then decreases. Therefore, the waveform IB of the intensity of the light L1a from the surface TA (that is, the intensity of the light L1a indicating the temperature of the surface TA), which is proportional to the temperature of the surface TA, is the peak Q1 due to the temperature rise immediately after the irradiation of the light La of the intensity InA. And peak Q3 due to heat transfer from the interface between the vacancies and the member T at a time later than peak Q1, and peak Q2 due to heat transfer from the bottom surface at a time later than peak Q3. It changes with the passage of time so as to include.

As shown in FIG. 31 (B), when a hole is formed inside the member T and the intensity of the light La changes as shown in the waveform In, the surface TA is released over time. The intensity of the waveform Ia of the intensity change of the light L1a also changes periodically with the passage of time. The waveform Ia of the intensity change of the light L1a from the surface TA is the waveform I1 (waveform corresponding to the peak Q1) indicating the temperature change due to the temperature rise immediately after the irradiation of the light La, and the reciprocating waveform from the surface TA to the bottom surface of the member T. Waveform I2 (waveform corresponding to peak Q2) indicating temperature change due to heat transfer and waveform I3 (peak Q3) indicating temperature change due to reciprocating heat transfer from the surface TA to the interface between the pore and the member T. The waveform corresponding to) is combined with the waveform. The waveform I3 is a temperature change caused by heat transferred from the surface TA to the interface between the vacancies and the member T and returned to the surface TA from the interface between the vacancies and the member T. Therefore, the waveform I3 has a waveform whose phase is delayed with respect to the waveform I1 by the amount of time it takes for the heat of the surface TA to transfer heat back and forth from the surface TA to the interface between the vacancies and the member T, and the period is almost the same. It will be the same. The waveform I3 has a phase lag and a smaller amplitude than the waveform I1. On the other hand, the waveform I3 has a larger phase and a larger amplitude than the waveform I2.

As described above, when the member T has a hole, the waveform Ia of the intensity of the light L1a in the detection region AR1a (the surface TA of the member T) has a hole in the member T due to the synthesis of the waveform I3. Is different from the waveform I of the intensity of the light L1a in the detection region AR1a when is not formed. For example, the waveform Ia is different in at least one of the phase and the amplitude from the waveform I, and more specifically, the phase is delayed with respect to the waveform I and the amplitude is large. Therefore, the detection unit 52 detects the intensity of the light L1a in the detection area AR1a every time, and calculates the waveform of the intensity of the light L1a in the detection area AR1a. The detection unit 52 compares the calculated waveform of the intensity of the light L1a in the detection region AR1a with, for example, the waveform I when the member T does not have a hole, so that the member T has a hole. Is detected.

A specific method for detecting vacancies by the detection unit 52 will be described below. FIG. 32 is a flowchart illustrating a method for detecting vacancies according to the second embodiment. As shown in FIG. 32, the detection unit 52 acquires the image data Pa1 which is the intensity data of the light L1a from the detection region AR1a detected by the first imaging unit 34Aa (step S120). The image data Pa1 is data on the intensity of the light L1a from the region that overlaps with the junction T3 of the detection region AR1a. More specifically, the light detected by the pixels corresponding to the region that overlaps with the junction T3 of the detection region AR1a. It is the data of the intensity of L1a. Since the holes are formed inside the joint portion T3, the detection unit 52 obtains image data of the intensity of the light L1a from the region overlapping the joint portion T3 in the image data Pa of the detection region AR1a. It can be said that it is acquired as Pa1. The irradiation unit 32Aa irradiates the light La so as to periodically change with the passage of time, and the first imaging unit 34Aa acquires the image data Pa at predetermined time intervals. The detection unit 52 acquires the image data Pa1 generated at predetermined time intervals from the first imaging unit 34Aa, and pixels corresponding to the region in which the waveform information of the intensity of the light L1a is superimposed on the junction portion T3 of the detection region AR1a. It is detected every time (step S122). Specifically, the intensity of the light L1a from the detection area AR1 is plotted for each time, and the waveform of the intensity of the light L1a obtained by fitting is superimposed on the junction T3 of the detection area AR1a for each pixel. The waveform information is detected as the waveform information of the intensity of the light L1a. The detection unit 52 may use at least one of the amplitude of the waveform of the intensity of the light L1a and the phase of the waveform of the intensity of the light L1a as waveform information. The waveform information is not limited to the amplitude and phase of the waveform of the intensity of the light L1a, and may be the intensity of the light L1a for each time.

After detecting the waveform information, the detection unit 52 determines for each pixel whether the waveform information is outside the preset reference waveform range. The value of the reference waveform range may be arbitrarily set, but may be set in advance as a numerical range of waveform information when, for example, no pores causing quality abnormality are not formed. The detection unit 52 determines whether the number of pixels whose waveform information is out of the reference waveform range is equal to or greater than a predetermined threshold value (hereinafter referred to as the first waveform threshold value) (step S124). The first waveform threshold value here may be set arbitrarily, but for example, when the number of pixels whose waveform information is outside the range of the reference waveform range is equal to or greater than the first waveform threshold value, vacancies are formed. It may be set based on the high possibility.

When the number of pixels whose waveform information is outside the reference waveform range is equal to or greater than the first waveform threshold value (step S124; Yes), the detection unit 52 is formed with holes in the detection region AR1a (joint portion T3). (Step S126). On the other hand, when the number of pixels whose waveform information is out of the reference waveform range is not equal to or greater than the first waveform threshold value (step S124; No), the detection unit 52 has a hole in the detection region AR1a (here, the joint portion T3). Is not formed (step S128). In this embodiment, the vacancies are detected based on the intensity data (image data Pa1) of the light L1a from the region that overlaps with the joint portion T3 of the detection region AR1a. However, the detection unit 52 may detect the vacancies using the data of the intensity of the light L1a from the region where the light L1a can be detected, and the intensity of the light L1a from the region overlapping with the junction T3 of the detection region AR1a. It is not limited to using data.

Next, the detection of the thickness of the intermetallic compound between the first member T1 and the second member T2 as the internal state of the member T will be described. Hereinafter, the intermetallic compound of the first member T1 and the second member T2 will be referred to as compound F3. The detection unit 52 detects the compound F3 based on the image taken by the first imaging unit 34Aa by the light L1a from the detection region AR1a, in other words, based on the image data Pa of the first imaging unit 34Aa.

FIG. 33 is a schematic view illustrating heat transfer in the member. The first member T1 and the second member T2 may have different heat capacities. Hereinafter, as an example, the heat capacity of the first member T1 is smaller than the heat capacity of the second member T2. Here, the heat capacity is the product of the specific heat and the mass. Hereinafter, assuming that the volumes of the first member T1 and the second member T2 are equal, the first member T1 is made of aluminum and the second member T2 is made of iron as described above. In this case, since aluminum has a specific heat about twice that of iron and aluminum has a mass about one-third that of iron, the heat capacity of the first member T1 is smaller than that of the second member T2. Since the first member T1 has a smaller heat capacity than the second member T2, the degree of temperature rise due to the irradiation of light La is larger than the degree of temperature rise of the second member T2 (that is, the same amount of heat is imparted by the light La). If so, the first member T1 is easier to warm up). Therefore, when the light La is irradiated, the temperature of the first member T1 is higher than that of the second member T2, and the heat of the first member T1 passes through the joint portion T3 and is lower than that of the first member T1. It is transmitted to. Here, when the compound F3 is formed in the joint portion T3, the heat of the first member T1 is transferred to the second member T2 via the compound F3. The thermal conductivity of the compound F3 is different from the thermal conductivity of the first member T1, the second member T2, and the joint portion T3 where the compound F3 is not formed. Therefore, as shown in FIGS. 33A and 33B, the heat transfer rate (transfer amount) transferred from the first member T1 to the second member T2 differs depending on the thickness of the compound F3. Here, as an example, the thermal conductivity of the compound F3 is smaller than the thermal conductivity of the first member T1, the second member T2, and the joint portion T3 where the compound F3 is not formed. In FIG. 33, the thicker the compound F3, the higher the heat transfer rate (transfer amount) from the first member T1 to the second member T2. However, when the thermal conductivity of the compound F3 is larger than the thermal conductivity of the first member T1, the second member T2, and the joint portion T3 where the compound F3 is not formed, the thinner the compound F3, the better. At least one of the heat transfer rate and the heat transfer amount transferred from the first member T1 to the second member T2 increases.

34A and 34B are graphs for explaining the intensity of light from the detection region. 34A and 34B show the second member T2 (the member having the smaller heat capacity of the first member T1 and the second member T2) when the irradiation region AR0 of the member T is irradiated with light La from the irradiation unit 32Aa. The temperature of the surface T2A and the intensity of the light L1a from the surface T2A are shown. FIG. 34A shows an example in which the thickness of the compound F3 is thinner than that of the example of FIG. 34B described later. As shown in FIG. 34A (A), assuming that the irradiation region AR0 is irradiated with the light La of the intensity InA at the time tA, the intensity of the light L1a emitted from the surface T2A changes with the passage of time. It changes as shown in. Specifically, the temperature of the surface T2A rises significantly immediately after the time tA due to the irradiation of the light La of the intensity InA at the time tA. Then, the heat of the surface T2A is gradually transferred to the inside of the second member T2, so that the temperature of the surface T2A gradually decreases with the passage of time. Further, the heat from the first member T1 is transmitted to the second member T2 via the compound F3. The heat from the first member T1 reaches the second member T2 after the time tA by the amount of the heat transfer time from the first member T1 to the second member T2 via the compound F3. Therefore, the temperature of the surface T2A rises once due to the heat from the first member T1 and then falls again. After that, the heat transfer from the bottom surface of the second member T2 reaches the surface T2A, and the temperature of the surface T2A rises once and then falls again. Therefore, the intensity of the light L1a emitted from the surface T2A, which is proportional to the temperature of the surface T2A (that is, the intensity of the light L1a indicating the temperature of the surface T2A) is the peak Q1 due to the temperature rise immediately after the irradiation of the light La of the intensity InA. Includes a peak Q4 due to heat transfer from the first member T1 through the compound F3 at a time later than the peak Q1 and a peak Q2 due to heat transfer from the bottom surface at a time later than the peak Q4. As such, it changes with the passage of time. Although it is described here that the heat from the bottom surface of the second member T2 is transferred slower than the heat from the first member T1, the heat from the bottom surface of the second member T2 is not limited to this, and the heat from the bottom surface of the second member T2 is transferred to the first member. The heat may be transferred earlier than the heat from T1, or may be transferred at the same timing as the heat from the first member T1.

In the present embodiment, the irradiation region AR0 of the member T is irradiated with light La so that the intensity changes periodically with the passage of time. As shown in FIG. 34A (B), when the intensity of the light La changes as shown in the waveform In, the intensity of the waveform Ic according to the passage of time of the light L1a from the surface T2A also has a period according to the passage of time. Change. The waveforms Ic are a waveform I1 (waveform corresponding to the peak Q1) indicating a temperature change due to a temperature rise immediately after irradiation with light La, and a waveform I4 (waveform corresponding to a peak Q1) indicating a temperature change due to heat transfer from the first member T1 through the compound F3. The waveform corresponding to the peak Q4) and the waveform I2 (waveform corresponding to the peak Q2) showing the temperature change due to the heat transfer from the bottom surface to the surface T2A are combined. The waveform I4 is an intensity waveform of the light L1a caused by the heat generated by the heat from the first member T1 and reaching the second member T2 via the compound F3. Therefore, the waveform I4 is a waveform whose phase is delayed by the amount of the heat transfer time from the first member T1 with respect to the waveform I1 indicating the temperature change due to the temperature rise immediately after the irradiation of the light La, and the periods are almost the same. .. Here, the waveform I4 has a phase lag and a smaller amplitude than the waveform I1. On the other hand, the waveform I4 has a phase advanced and a large amplitude with respect to the waveform I2, but is not limited to this, and the phase may be the same or lagging with respect to the waveform I2, and the amplitude may be the same or smaller.

FIG. 34B shows an example in which the compound F3 is thicker than the example of FIG. 34A. When the thickness of the compound F3 is large, in the example of this description, the heat transfer rate and the heat transfer amount transferred from the first member T1 to the second member T2 become smaller. Therefore, the heat from the first member T1 reaches the second member T2 later than the example of FIG. 34A. Therefore, the magnitude of the intensity peak Q4a of the light L1a from the surface T2A due to the heat transfer from the first member T1 shown in FIG. 34B (A) through the compound F3 is the peak Q4 shown in FIG. 34A (A). Is smaller than the size of. The time between the time tA in FIG. 34B (A) and the observation of the peak Q4a is larger than the time between the time tA in FIG. 34A (A) and the observation of the peak Q4. ,become longer.

In the present embodiment, the irradiation region AR0 of the member T is irradiated with light La so that the intensity changes periodically with the passage of time. As shown in FIG. 34B (B), when the intensity of the light La changes as shown in the waveform In, the intensity of the waveform Ica of the intensity of the light L1a from the surface T2A with the passage of time also depends on the passage of time. It changes periodically. The waveform Ica is a waveform I1 (waveform corresponding to the peak Q1) indicating a temperature change due to a temperature rise immediately after irradiation with light La, and a waveform I4a (waveform I4a) indicating a temperature change due to heat transfer from the first member T1 through the compound F3. The waveform is a combination of the waveform corresponding to the peak Q4a) and the waveform I2 (waveform corresponding to the peak Q2) indicating the temperature change caused by the heat transfer from the bottom surface to the surface T2A. Since the compound F3 is thicker in FIG. 34B than in the example of FIG. 34A, the waveform I4a is out of phase with the waveform I4 in FIG. 34A and has a smaller amplitude. Therefore, the waveform Ica of the surface T2A is out of phase with the waveform Ic of FIG. 34A, and the amplitude becomes smaller.

That is, the waveform of the intensity of the light L1a from the surface T2A differs depending on the thickness of the compound F3, and for example, at least one of the phase and the amplitude differs depending on the thickness of the compound F3. Therefore, the detection unit 52 detects the intensity of the light L1a of the surface T2A of the second member T2 every time, calculates the waveform of the intensity of the light L1a of the surface T2A of the second member T2, and calculates the surface T2A. By comparing the waveform of the intensity of the light L1a of No. 1 with, for example, the waveform Ic when the thickness of the compound F3 is thin, it is possible to detect whether the thickness of the compound F3 is thick. The detection unit 52 detected the intensity of the light L1a from the surface T2A and detected the thickness of the compound F3, but it does not have to be the surface T2A, and the intensity of the light L1a from the surface T1A of the first member T1 is determined. It may be detected and the thickness of the compound F3 may be detected, or the intensity of the light L1a from both the surface T1A and the surface T2A may be detected to detect the thickness of the compound F3. Even when the thermal conductivity of the compound F3 is larger than the thermal conductivity of the first member T1, the second member T2, and the joint portion T3 where the compound F3 is not formed, the thickness of the compound F3 is the same as described above. Can be detected.

A specific method for detecting the thickness of compound F3 by the detection unit 52 will be described below. FIG. 35 is a flowchart illustrating a method for detecting the thickness of the compound according to the second embodiment. As shown in FIG. 35, the detection unit 52 acquires the image data Pa2 which is the data of the intensity of the light L1a from the detection region AR1a detected by the first imaging unit 34Aa, and based on the image data Pa2, the first imaging The brightness is acquired for each pixel of the unit 34Aa. (Step S130). The image data Pa2 is data on the intensity of the light L1a from a region other than the junction T3 of the detection region AR1a. More specifically, the image data Pa2 is the data of the light L1a detected by the pixels corresponding to the region other than the junction T3 of the detection region AR1a. Intensity data. That is, the image data Pa2 is data on the intensity of the light L1a from the region of the detection region AR1a that overlaps at least one of the first member T1 and the second member T2. In other words, the detection unit 52 is the data of the intensity of the light L1a from the region other than the joint portion T3 (the region overlapping at least one of the first member T1 and the second member T2) in the image data Pa of the detection region AR1a. Can be said to be acquired as image data Pa2. Based on the image data Pa2, the detection unit 52 calculates the brightness for each pixel corresponding to the region other than the junction portion T3 of the detection region AR1a. As described above, the detection unit 52 detects the vacancies using the image data P1a of the region overlapping the junction T3 of the detection region AR1a, and uses the image data P1b of the region other than the junction T3 of the detection region AR1a to compound. The thickness of F3 is detected.

The irradiation unit 32Aa irradiates the light La so as to periodically change with the passage of time, and the first imaging unit 34Aa acquires the image data Pa at predetermined time intervals. The detection unit 52 acquires the image data Pa2 generated at predetermined time intervals from the first imaging unit 34Aa, and outputs the waveform information of the intensity of the light L1a to the second member T2 (the first member T1 and the second member T2). Of these, the member having the smaller heat capacity) is detected for each pixel corresponding to the unit region (step S132). Specifically, the detection unit 52 converts the brightness of the pixel into the intensity of the light L1a, plots the converted intensity of the light L1a for each time, and joins the waveform of the intensity of the light L1a obtained by fitting. It is generated for each pixel corresponding to a region other than the portion T3 (a region overlapping at least one of the first member T1 and the second member T2), and the waveform information thereof is detected as the waveform information of the intensity of the light L1a. The detection unit 52 may use at least one of the amplitude of the waveform of the intensity of the light L1a and the phase of the waveform of the intensity of the light L1a as waveform information. The waveform information is not limited to the amplitude and phase of the waveform of the intensity of the light L1a, and may be the intensity of the light L1a for each time.

After detecting the waveform information, the detection unit 52 determines for each pixel whether the waveform information is outside the preset reference waveform range. The value of the reference waveform range may be arbitrarily set, but may be set in advance as waveform information when, for example, the thickness of the compound F3 is thin enough not to cause quality abnormality. The detection unit 52 determines whether the number of pixels whose waveform information is out of the reference waveform range is equal to or greater than a predetermined threshold value (hereinafter referred to as a second waveform threshold value) (step S134). The second waveform threshold value here may be set arbitrarily, but for example, when the number of pixels whose waveform information is outside the range of the reference waveform range is equal to or greater than the second waveform threshold value, the thickness of the compound F3 becomes thicker. It may be set based on the fact that the possibility of quality abnormality increases.

When the number of pixels whose waveform information is out of the reference waveform range is equal to or greater than the second waveform threshold value (step S134; Yes), the detection unit 52 has an abnormal thickness of the compound F3, that is, the thickness of the compound F3 is thick. It is determined that the quality is abnormal (step S136). On the other hand, when the number of pixels whose waveform information is outside the reference waveform range is not equal to or greater than the second waveform threshold value (step S134; No), the detection unit 52 determines that the thickness of the compound F3 is normal (not abnormal), that is, the compound F3. It is determined that the thickness is not thick (sufficiently thin) (step S138).

As described above, the detection unit 52 detects the vacancies and the thickness of the compound F3 as member information indicating the internal state of the member T. The detection unit 52 detects the openings F1 and the burr F2 as member information indicating a state other than the internal state of the member T, that is, member information indicating the state of the outside (surface TA) of the member T.

The detection unit 52 according to the second embodiment detects the opening F1 and the burr F2 based on the image data Pa of the first imaging unit 34Aa. The opening F1 is formed on the surface T3A of the joint portion T3, and the bottom surface of the opening F1 is located on the inner side (−Z axis direction side) of the joint portion T3 with respect to the surface T3A of the joint portion T3. Therefore, even if the light La emitted from the irradiation unit 32Aa is blocked by the surface T3A of the joint portion T3 around the opening F1 and does not reach the bottom surface of the opening F1, or reaches the bottom surface of the opening F1, the opening F1 The light L1a from the bottom surface of the opening F1 is blocked by the side surface of the opening F1 or the like. When the light La does not reach the bottom surface of the opening F1, the bottom surface of the opening F1 is heated not by the irradiation of the light La but by the heat transfer from the surroundings. Therefore, as for the waveform of the temperature (intensity of the light L1a) at the bottom surface of the opening F1, the waveform I1 indicating the temperature change due to the temperature rise immediately after the irradiation of the light La is not observed, and the waveform is caused by the heat transfer from the periphery of the opening F1. It becomes. Therefore, the waveform of the intensity of the light L1a on the bottom surface of the opening F1 is different in at least one of the phase and the amplitude from the waveform of the intensity of the light L1a on the surface T3A where the opening F1 is not formed, for example, the phase is delayed. The amplitude becomes smaller. Therefore, the detection unit 52 detects the waveform information for each pixel corresponding to the region overlapping the surface T3A of the detection region AR1a, and the pixels whose waveform information is outside the predetermined reference waveform range are equal to or higher than a predetermined threshold value. It may be determined whether or not the pixels are continuously adjacent to each other, and it may be detected that the opening F1 is formed when the pixels whose waveform information is outside the reference waveform range are continuously adjacent to each other by a predetermined threshold number or more. The waveform information is information based on the intensity waveform of the light L1a in which the intensity of the light L1a from the surface T3A is plotted for each time, and may be at least one of the waveform and the phase of the intensity of the light L1a. .. The detection unit 52 does not have to determine whether the waveform information is outside the predetermined reference waveform range. For example, when the detection unit 52 detects pixels (low-luminance pixels) having a brightness value lower than a predetermined value in the image data Pa of the first imaging unit 34Aa and the number of the images is equal to or more than a predetermined threshold value. , It may be determined that the opening F1 is formed.

Further, in the burr F2, the heat given by the irradiation of the light La tends to be difficult to escape from the burr F2 to the member T (inside the member T), so that the intensity of the light L1a from the burr F2 is the place where the burr F2 does not exist. Higher than the intensity of light L1a from. Therefore, the detection unit 52 detects the high-luminance pixels whose brightness value is equal to or higher than the predetermined value from the image data Pa of the first imaging unit 34Aa, and the high-luminance pixels at the positions corresponding to the boundary portion T3B are continuously generated. When they are adjacent to each other by a predetermined threshold value or more, it is determined that the burr F2 is formed.

Further, the detection unit 52 according to the second embodiment detects the insertion amount of the tool 18 and the shape of the tool 18 as the tool state. The detection unit 52 according to the second embodiment detects the insertion amount of the tool 18 and the shape of the tool 18 based on the image captured by the second imaging unit 34Ba. The method of detecting the insertion amount of the tool 18 based on the image captured by the second imaging unit 34Ba is the same as the method of detecting the insertion amount of the tool 18 in the first embodiment. Similarly, the method of detecting the shape of the tool 18 based on the image captured by the second imaging unit 34Ba is the same as the method of detecting the shape of the tool 18 in the first embodiment.

As described above, in the second embodiment, the first imaging unit 34Aa is made to image the surface TA of the member T, and the second imaging unit 34Ba is made to image the surface of the tool 18. Then, the detection unit 52 detects the member state, that is, the pores, the thickness of the compound F3, the opening F1 and the burr F2 based on the image captured by the first imaging unit 34Aa, and the image captured by the second imaging unit 34Ba. Based on, the tool state, that is, the insertion amount of the tool 18 and the shape of the tool 18 are detected. However, the detection unit 52 may detect the aperture F1 and the burr F2 based on the image of the second imaging unit 34Ba as in the first embodiment. In this case, the imaging region R2a of the second imaging unit 34Ba is overlapped with the surface TA of the member T in addition to the surface of the tool 18. That is, the detection region AR2a in this case includes the surface TA of the member T and the surface of the tool 18. The detection unit 52 detects the opening F1 and the burr F2 in the same manner as in the first embodiment based on the image of the surface TA of the member T among the images captured by the second imaging unit 34Ba. Further, the tool state may be detected based on the image captured by the first imaging unit 34Aa without providing the second imaging unit 34Ba. In this case, the imaging region R1a of the first imaging unit 34Aa is overlapped with the surface of the tool 18 in addition to the surface TA of the member T. That is, the detection region AR1a in this case includes the surface TA of the member T and the surface of the tool 18. The detection unit 52 detects the tool state in the same manner as in the first embodiment based on the image of the surface of the tool 18 among the images captured by the first imaging unit 34Aa.

FIG. 36 is a diagram illustrating light detection by the first imaging unit according to the second embodiment. As shown in FIG. 36, the manufacturing system 1 moves the tool 18 relative to the member T in the feed direction (here, the Y-axis direction) to join the first member T1 and the second member T2. The first imaging unit 34Aa is made to detect the intensity of the light L1a from the detection region AR1a at each predetermined frame rate. Therefore, as the tool 18 moves relative to the member T in the feed direction, the irradiation unit 32Aa and the first imaging unit 34Aa also move relative to the member T in the feed direction, and the surface TA of the member T in the irradiation region AR0 The position on the top and the position on the surface TA of the member T of the detection region AR1a move. However, as shown in FIG. 36, even if the irradiation unit 32Aa and the first imaging unit 34Aa move relative to each other in the feeding direction, the unit region ARa of the surface TA is within the range of the irradiation region AR0 and the detection region AR1a for a predetermined time. Will fit in. Therefore, it is possible to irradiate the unit region ARa with light La whose intensity changes with the passage of time, detect the light L1a from the unit region ARa every hour, and detect the opening F1 or the sky in the unit region ARa. Pore and compound F3 can be detected. Since it is not necessary for the detection unit 52 to use the intensity waveform of the light L1a when detecting the burr F2, the intensity data of the light L1a detected every time (imaged at a predetermined frame rate). The burr F2 is detected using one of the image data of the first imaging unit 34Aa). However, the detection unit 52 detects the burr F2 by using a plurality of the intensity data of the light L1a detected for each time (image data of the first imaging unit 34Aa captured at a predetermined frame rate). You may.

Next, the processing flow by the arithmetic unit 38 according to the second embodiment will be described. FIG. 37 is a flowchart illustrating a processing flow of the arithmetic unit according to the second embodiment. As shown in FIG. 37, the arithmetic unit 38 according to the first embodiment is set by the device control unit 50 with the trigger that the manufacturing device 10 starts joining the first member T1 and the second member T2. The imaging unit 34 starts capturing an image at the frame rate (step S140). When the manufacturing apparatus 10 starts joining the first member T1 and the second member T2, the device control unit 50 causes the irradiation unit 32Aa to irradiate the region R0 with light La, and causes the first imaging unit 34Aa to range the imaging region R1a. The image inside is imaged. The device control unit 50 causes the irradiation unit 32Aa to irradiate the irradiation unit 32Aa with light La so that the intensity periodically changes with the passage of time until the manufacturing apparatus 10 stops joining the first member T1 and the second member T2. At the same time, the first imaging unit 34Aa is made to image an image within the range of the imaging region R1a at a predetermined frame rate. Further, the device control unit 50 causes the second image pickup unit 34Ba to take an image at the timing when the tip portion 22A of the probe 22 comes into contact with the surface TA of the member T, and the detection unit 52 causes the probe 22 to take an image based on the image. The distance D0a between the end portion 20A of the shoulder 20 and the surface TA of the member T at the timing when the tip portion 22A of the member T comes into contact with the surface TA of the member T is calculated.

When the image pickup unit 34 captures an image, the arithmetic unit 38 acquires the image data generated by the image pickup unit 34 by the detection unit 52, and sets the member state as a hole, an abnormality in the thickness of the compound F3, and the like. The opening F1 and the burr F2 are detected (step S142). The detection unit 52 detects the member state during the joining by the manufacturing apparatus 10. Each time the image pickup unit 34 captures an image, the detection unit 52 acquires the image data generated by the image pickup unit 34 and detects the member state. From the image data Pa generated by the first imaging unit 34Aa, the detection unit 52 includes holes formed inside the member T, the thickness of the compound F3, and the opening F1 formed on the surface T3A of the joint portion T3. , The burr F2 formed around the surface T3A of the joint portion T3 is detected as a member state. The detection unit 52 determines whether the detected member state is abnormal.

The arithmetic unit 38 detects when there is an abnormality in the detection result of the member state (step S144; Yes), that is, when at least one of a hole, an abnormality in the thickness of the compound F3, an opening F1 and a burr F2 is detected. , It is determined whether a hole is detected inside the member T and whether a burr F2 is detected (step S146). When the vacancy inside the member T and the burr F2 are detected (step S146; Yes), the arithmetic unit 38 detects the shape of the tool 18 and the insertion amount of the tool 18 as the tool state by the detection unit 52. (Step S148). The arithmetic unit 38 determines whether the detected shape of the tool 18 and the insertion amount of the tool 18 are abnormal. The method of detecting the shape of the tool 18 and the insertion amount of the tool 18 is the same as that of the first embodiment.

When the shape of the tool 18 and the insertion amount of the tool 18 are detected as the tool state, the arithmetic unit 38 is joining the manufacturing apparatus 10 based on the member state and the tool state detected by the detection unit 52 by the change information generation unit 56. Change information for changing the operating conditions is generated, and the generated change information is transmitted to the control device 30 (step S158). The method of generating change information will be described later.

The arithmetic unit 38 determines whether or not the vacancy is detected when both the vacancy and the burr F2 are not detected (step S146; No), that is, when at least one of the vacancy and the burr F2 is not detected. to decide. When a hole is detected (step S150; Yes), the arithmetic unit 38 detects the shape of the tool 18 as the tool state (step S152). That is, the arithmetic unit 38 detects the shape of the tool 18 as the tool state when the burr F2 is not detected but the vacancies are detected. The method for detecting the shape of the tool 18 is the same as the method for detecting the shape of the tool 18 in step S148. After detecting the shape of the tool 18 as the tool state, the arithmetic unit 38 changes the operating condition during joining of the manufacturing apparatus 10 based on the member state and the tool state detected by the detection unit 52 by the change information generation unit 56. The change information of the above is generated, and the generated change information is transmitted to the control device 30 (step S158). The method of generating change information will be described later.

When the vacancy is not detected (step S150; No), the arithmetic unit 38 determines whether or not the burr F2 is detected. When the burr F2 is detected (step S154; Yes), the arithmetic unit 38 detects the insertion amount of the tool 18 as the tool state (step S156). The method of detecting the insertion amount of the tool 18 is the same as the method of detecting the insertion amount of the tool 18 in step S152. When the insertion amount of the tool 18 is detected as the tool state, the arithmetic unit 38 changes the operating conditions during joining of the manufacturing apparatus 10 based on the member state and the tool state detected by the detection unit 52 by the change information generation unit 56. The change information for the purpose is generated, and the generated change information is transmitted to the control device 30 (step S158). The method of generating change information will be described later. Since the shape of the tool 18 is not detected in step S156, the control for stopping the joining between the first member T1 and the second member T2 in order to detect the shape of the tool 18 is not executed (to the manufacturing apparatus 10). The joining between the first member T1 and the second member T2 is not stopped).

When the burr F2 is not detected (step S154; No), that is, when the vacancies are not detected and the burr F2 is not detected, the arithmetic unit 38 does not detect the tool state and is based on the detected member state. Change information for changing the operating conditions during joining is generated, and the generated change information is transmitted to the control device 30 (step S158). In other words, when the abnormality of the member state is other than the detection of the vacancies and burrs F2, here, when the abnormality of the thickness of the compound F3 or the opening F1 is detected and the vacancies and burrs F2 are not detected, the arithmetic unit. 38 generates change information based on the member state without detecting the tool state. The method of generating change information will be described later. Since the shape of the tool 18 is not detected even when the vacancies are not detected and the burr F2 is not detected, the control for stopping the joining between the first member T1 and the second member T2 in order to detect the shape of the tool 18 is performed. , Do not execute (the manufacturing apparatus 10 does not stop the joining of the first member T1 and the second member T2).

The control device 30 of the manufacturing device 10 changes the operating conditions based on the change information transmitted from the arithmetic unit 38. When the joining between the first member T1 and the second member T2 is completed after the change information is generated and transmitted (step S160; Yes), this process is terminated. When the joining between the first member T1 and the second member T2 is not completed (step S160; No), the control device 30 of the manufacturing apparatus 10 has the first member T1 and the second member T1 and the second member under the operating conditions changed based on the change information. Continuing the bonding with T2, the arithmetic unit 38 returns to step S140 and continues the irradiation of the optical LAB by the irradiation unit 32A and the imaging by the imaging unit 34.

In step S144, when there is no abnormality in the detection result of the member state (step S144; No), that is, all of the pores, the abnormality of the thickness of the compound F3, the opening F1 and the burr F2 are detected. If not, the process proceeds to step S160, and when the joining between the first member T1 and the second member T2 is completed (step S160; Yes), this process is terminated and the first member T1 and the second member T2 are combined. If the joining is not completed (step S160; No), the process returns to step S140, and irradiation of light La by the irradiation unit 32Aa and imaging by the imaging unit 34 are continued. The order of detection of the pores, the abnormality in the thickness of the compound F3, the opening F1 and the burr F2 may be arbitrary.

In the description of FIG. 37, if there is no abnormality in the detection result of the member state, the change information is not generated. However, as in the first embodiment, the arithmetic unit 38 is based on the member state detected during the joining between the first member T1 and the second member T2 even when there is no abnormality in the member state detection result. , Generate change information for changing the operation condition of the next joining after the joining between the first member T1 and the second member T2 (joining when the detection unit 52 detects the member state) is completed. May be good.

Next, the generation of change information according to the second embodiment will be described. 38 and 39 are flowcharts illustrating the generation of change information. FIG. 38 shows a method of generating change information when a hole is formed inside the member T. That is, FIG. 38 describes in detail the method of generating change information when an abnormality in reflectance is detected in step S158 of FIG. 37. The causes of the formation of the vacancies include improper mixing of the first member T1 and the second member T2, insufficient heat input to the first member T1 and the second member T2, and the like. The first member T1 and the second member T2 are agitated by the rotating probe 22 and mixed with each other to form a joint portion T3 including the first member T1 and the second member T2. However, for example, when the rotation of the tool 18 is inappropriate (for example, when the rotation speed of the tool 18 is insufficient), the first member T1 and the second member T2 are not sufficiently mixed, or mixed. Defects occur. When a mixing failure occurs, for example, the first member T1 and the second member T2 may not enter the space formed by inserting the probe 22 and remain as a space, and a hole may be formed. Further, when the frictional heat of the tool 18 is not sufficient (for example, when the feeding speed of the tool 18 is too fast), insufficient heat input to the first member T1 and the second member T2 occurs. When the heat input is insufficient, the first member T1 and the second member T2 are not sufficiently softened by heat, so that stirring by the probe 22 is hindered. For example, the first member T1 is formed in the space formed by inserting the probe 22. And the second member T2 do not enter and remain as a space, and a hole may be formed. The arithmetic unit 38 according to the second embodiment determines the cause of the formation of vacancies, that is, the defective mode of the joint state, based on the member state and the tool state. Specifically, as shown in FIG. 38, when the detection unit 52 detects a hole inside the member T, it is determined that both the groove size of the tool 18 and the outer diameter of the tool 18 are abnormal. If this is the case (step S162; Yes), the joint state determination unit 54 determines that, among the defective modes of the joint state, a mixing abnormality and insufficient heat input have occurred (step S164). When the change information generation unit 56 determines that the joining state determination unit 54 has caused a mixing abnormality and insufficient heat input, the change information generation unit 56 increases the rotation speed of the tool 18 and reduces the feed rate of the tool 18. The change information to the effect is generated (step S166). In other words, the change information generation unit 56 determines that the detection unit 52 has a hole formed inside the member T, and both the groove size of the tool 18 and the outer diameter of the tool 18 are abnormal. When it is determined that there is, change information for increasing the rotation speed of the tool 18 and change information for decreasing the feed rate of the tool 18 are generated.

Further, when the detection unit 52 determines that both the groove size of the tool 18 and the outer diameter of the tool 18 are not abnormal (step S62; No) (the groove size of the tool 18 is abnormal). And when the condition that the outer diameter of the tool 18 is abnormal is not satisfied), when the detection unit 52 determines that the groove size of the tool 18 is abnormal (step S168; Yes), the joint state determination unit 54 determines that a mixing abnormality has occurred in the defective mode of the joining state (step S170). That is, when the joint state determination unit 54 determines that the hole is formed inside the member T and the detection unit 52 determines that the groove size of the tool 18 is abnormal, the mixing is performed. Judge that something is wrong. When the joint state determination unit 54 determines that a mixing abnormality has occurred, the change information generation unit 56 generates change information to the effect that the rotation speed of the tool 18 is increased (step S172). In other words, when the detection unit 52 determines that a hole is formed inside the member T and the change information generation unit 56 determines that the groove size of the tool 18 is abnormal, the tool Generates change information to increase the number of revolutions of 18.

Further, when the detection unit 52 determines that the groove size of the tool 18 is not abnormal (step S168; No), and the detection unit 52 determines that the outer diameter of the tool 18 is abnormal (step S174). Yes), the joint state determination unit 54 determines that insufficient heat input has occurred in the defective mode of the joint state (step S176). That is, when the joint state determination unit 54 determines that the detection unit 52 has a hole formed inside the member T and determines that the outer diameter of the tool 18 is abnormal, the heat input is insufficient. Judge that is happening. When the joint state determination unit 54 determines that the heat input is insufficient, the change information generation unit 56 generates change information to the effect that the feed rate of the tool 18 is reduced (step S172). In other words, when the detection unit 52 determines that a hole is formed inside the member T and the outer diameter of the tool 18 is abnormal, the change information generation unit 56 determines that the tool 18 is abnormal. Generate change information to the effect that the feed rate of the When the detection unit 52 determines that the groove size of the tool 18 is not abnormal and the outer diameter of the tool 18 is not abnormal (step S174; No), the arithmetic unit 38 ends this process.

FIG. 39 shows a method of generating change information when an abnormality in the thickness of compound F3 is detected as an abnormality in the member state. As shown in FIG. 39, when the detection unit 52 determines that the thickness of the compound F3 is abnormal, that is, the thickness of the compound F3 is too thick, the bonding state determination unit 54 receives heat in the defective mode of the bonding state. It is determined that a misalignment has occurred (step S182). When the joint state determination unit 54 determines that the heat input position defect has occurred, the change information generation unit 56 generates change information to correct the machining path of the tool 18 (step S184). In other words, the change information generation unit 56 generates change information to correct the machining path of the tool 18 when the detection unit 52 detects an abnormality in the thickness of the compound F3. When the detection unit 52 determines that the thickness of the compound F3 is not abnormal, that is, the thickness of the compound F3 is sufficiently thin (step S180; No), the arithmetic unit 38 ends this process.

Since the method of generating change information when the opening F1 and the burr F2 are detected is the same as that of the first embodiment, the description thereof will be omitted.

As described above, the first imaging unit 34Aa according to the second embodiment receives the light L1a (infrared light) from the member T, and the detection unit 52 receives the light L1a received by the first imaging unit 34Aa. Based on this, the member state is detected. Since the detection unit 52 according to the second embodiment detects the member state based on the light L1a from the member T, it can appropriately contribute to quality improvement.

Further, the detection unit 52 according to the second embodiment detects a vacancy inside the joint portion T3 as a member state, and when the vacancy is detected, detects the shape of the tool 18 as a tool state. Since the calculation system 12 according to the second embodiment detects the shape of the tool 18 when the vacancy is detected, the cause of the quality abnormality can be detected with high accuracy based on the vacancy and the shape of the tool 18.

The detection unit 52 according to the second embodiment detects whether the dimension of the groove of the probe 22 at the tip of the tool 18 and the outer diameter of the tool 18 are abnormal as the tool state. The change information generation unit 56 according to the second embodiment generates information to increase the rotation speed of the tool 18 as change information when the detection unit 52 detects that the groove size is abnormal. The change information generation unit 56 generates information to the effect that the feed rate of the tool 18 is reduced when the detection unit 52 detects that the outer diameter of the tool 18 is abnormal. By generating the change information in this way, the arithmetic system 12 can appropriately suppress the mixing abnormality and the insufficient heat input in the manufacturing apparatus 10, and can appropriately assist the suppression of quality defects.

The detection unit 52 according to the second embodiment detects the burr F2 around the joint portion T3, and when the burr F2 is detected, the detection unit 52 may detect the insertion amount of the tool 18 as a tool state. Since the calculation system 12 detects the insertion amount of the tool 18 when the burr F2 is detected, the operating conditions to be changed can be appropriately selected based on the insertion amount of the burr F2 and the tool 18.

When the detection unit 52 according to the second embodiment detects whether the insertion amount of the tool 18 is abnormal, and the change information generation unit 56 detects that the insertion amount of the tool 18 is abnormal. In addition, information to the effect that the insertion amount of the tool 18 is reduced is generated as change information. The change information generation unit 56 generates information to increase the feed rate of the tool 18 as change information when the detection unit 52 detects that the insertion amount of the tool 18 is not abnormal. By generating the change information in this way, the arithmetic system 12 can appropriately suppress excessive heat input to the manufacturing apparatus 10 and appropriately assist in suppressing quality defects.

The detection unit 52 according to the second embodiment detects the opening F1 formed in the joint portion T3, and the change information generation unit 56 reduces the feed rate of the tool 18 when the detection unit 52 detects the opening F1. Information to the effect is generated as change information. When the opening F1 is detected, the arithmetic system 12 appropriately suppresses the heat input shortage in the manufacturing apparatus 10 by generating change information to the effect that the feed rate of the tool 18 is lowered, and suppresses quality defects. Can be properly assisted.

The detection unit 52 according to the second embodiment detects whether the thickness of the compound F3 between the first member T1 and the second member T2 is abnormal as the member state. The change information generation unit 56 generates information to the effect that the processing path is changed when the detection unit 52 detects that the thickness of the compound F3 is abnormal. When the change information generation unit 56 detects an abnormality in the thickness of the compound F3, the change information generation unit 56 appropriately suppresses the heat input position defect in the manufacturing apparatus 10 by generating the change information to the effect that the processing path is changed, and the quality is improved. It can appropriately assist the suppression of defects.

Further, the irradiation unit 32Aa according to the second embodiment heats the member T as a heating unit. That is, the irradiation unit 32Aa heats at least one of the surface of the first member T1, the second member T2, the joint portion T3, and the tool 18 as a heating unit. The first imaging unit 34Aa as the imaging unit 34 receives light L1a, which is infrared light, from the surfaces of the first member T1, the second member T2, the joint portion T3, and the tool 18 heated by the heating unit. The detection unit 52 detects at least a part of the member state and the tool state based on the light L1a (infrared light) received by the image pickup unit 34. The arithmetic system 12a according to the second embodiment heats the member T and captures an image of light L1a which is infrared light emitted from the member T, that is, by active thermography, the internal state of the member T. Can be detected properly. However, the arithmetic system 12a is not limited to detecting the internal state of the member T by using active thermography, and detects the internal state of the member T by using any other method such as an ultrasonic flaw detection method. It's okay.

(Third Embodiment)
Next, the third embodiment will be described. The manufacturing system 1b according to the third embodiment is different from the second embodiment in that the manufacturing apparatus 10b is not a friction stir welder but an apparatus for joining the first member T1 and the second member T2 by laser light. .. In the third embodiment, the description of the parts having the same configuration as that of the second embodiment will be omitted.

FIG. 40 is a schematic view of the manufacturing system according to the third embodiment. FIG. 41 is a schematic view illustrating a method of joining by the manufacturing apparatus according to the third embodiment. As shown in FIG. 40, the manufacturing system 1b according to the third embodiment includes a manufacturing apparatus 10b and a calculation system 12b. The manufacturing apparatus 10b includes a light irradiation unit 18b that irradiates a laser beam for joining the first member T1 and the second member T2. The light irradiation unit 18b can be said to be a tool because it irradiates a laser beam to process the first member T1 and the second member T2. The light irradiation unit 18b is attached to the tool holding unit 16 in the same manner as the tool 18 of the second embodiment. That is, the light irradiation unit 18b can change its relative position with respect to the member T by moving in the Z-axis direction. Further, the light irradiation unit 18b changes its posture with respect to the tool holding unit 16, that is, with respect to the member T in the θX axis direction by rotating around the rotation axis AX1 parallel to the X-axis direction. However, in the third embodiment, the light irradiation unit 18b does not have to rotate about the central axis AXb.

The control device 30 of the manufacturing apparatus 10b emits the laser beam Lb from the end 18bA toward the member T while keeping the distance between the end 18bA and the member T on the −Z axis direction side of the light irradiation unit 18b constant. Irradiate. Further, the control device 30 relatively moves the light irradiation unit 18b in the feed direction (Y-axis direction) with respect to the member T while irradiating the member T with the laser beam Lb from the end 18bA of the light irradiation unit 18b. Let me. The light irradiation unit 18b melts at least one of the first member T1 and the second member T2 by irradiating at least one of the surface T1A of the first member T1 and the surface T2A of the second member T2 with the laser beam Lb. At least one of the first member T1 and the second member T2 is melted to form a molten pool. Then, the light irradiation unit 18b moves relative to the member T in the feed direction, and the molten pool outside the region irradiated with the laser light Lb is cooled and solidified. By solidifying the molten pool, the first member T1 and the second member T2 are joined. Therefore, as shown in FIG. 41, the joint portion T3 between the first member T1 and the second member T2 joined by the manufacturing apparatus 10b is a molten pool in which at least one of the first member T1 and the second member T2 is melted. However, it can be said that it is a solidified part.

FIG. 42 is a diagram illustrating a tool state according to the third embodiment. As shown in FIG. 41, the R0 (irradiation region AR0) in which the irradiation unit 32Aa in the third embodiment irradiates light is the same as in the second embodiment. Further, it is the same as the imaging region R1a (detection region AR1a) imaged by the first imaging unit 34Aa in the third embodiment. On the other hand, in the third embodiment, the imaging region R2a (detection region AR2a) of the second imaging unit 34B is at least a part of the light irradiation unit 18b and the end portion 18bA of the light irradiation unit 18b, as shown in FIG. It overlaps the region between the member T and the surface TA of the member T. Therefore, the image captured by the second imaging unit 34B of the light irradiation unit 18b of the third embodiment captures the region between the light irradiation unit 18b, the end portion 18bA of the light irradiation unit 18b, and the surface TA of the member T. , Including. In FIG. 42, for convenience of explanation, the second image pickup unit 34B is a picture taken from the Y-axis direction side (feed direction side) of the light irradiation unit 18b, but in reality, the second image pickup is performed. As shown in FIGS. 40 and 41, the unit 34B is preferably imaged from the X-axis direction side of the light irradiation unit 18b. Further, in FIG. 42, for convenience of explanation, the description of the irradiation unit 32Aa and the first imaging unit 34Aa is omitted.

The detection unit 52 of the arithmetic system 12b according to the third embodiment has, as a member state, a hole inside the member T, an opening F1 formed in the joint portion T3, and the first member T1 and the second member T2. The thickness of compound F3 is detected. The method of detecting the pores inside the member T, the opening F1, and the thickness of the compound F3 by the detection unit 52 according to the third embodiment is the same as that of the second embodiment. When joining with the laser beam Lb as in the third embodiment, the pores are also called voids or porosity, and the opening F1 is also called underfill.

The detection unit 52 according to the third embodiment detects the out-of-focus distance D4 of the laser beam Lb irradiated by the light irradiation unit 18b and the inclination angle θ of the light irradiation unit 18b as a tool state. As shown in FIG. 42, the defocusing distance D4 of the laser beam Lb is the position of the focusing point SP of the laser beam Lb (in other words, the focal point of the optical system (not shown) in the light irradiation unit 18b) and the surface of the member T. The distance to the TA. The condensing point SP of the laser light Lb refers to a place where the luminous flux diameter of the laser light Lb is minimized.

The detection unit 52 detects the out-of-focus distance D4 of the laser light Lb based on the image captured by the second imaging unit 34B while the light irradiation unit 18b is irradiating the laser light Lb. The detection unit 52 is an image of the laser beam Lb from the image captured by the second imaging unit 34B, which corresponds to the region between the end portion 18bA of the light irradiation unit 18b and the surface TA of the member T. Extract the laser beam image. For example, the detection unit 52 is a pixel (that is, a pixel) having a brightness value of continuously adjacent to each other of a predetermined value or more in the image corresponding to the region between the end portion 18bA of the light irradiation unit 18b and the surface TA of the member T. An image composed of a set of luminance values (pixels having a brightness value corresponding to the laser light Lb) is extracted as a laser light image. The detection unit 52 uses the portion of the extracted laser beam image having the smallest width of the laser beam Lb as the focusing point SP of the laser beam Lb. The width here refers to the length of the laser beam image in the direction orthogonal to the direction from the end portion 18bA of the light irradiation portion 18b to the surface TA of the member T. The detection unit 52 determines the position of the focusing point SP of the laser beam Lb and the surface of the member T from the position of the focusing point SP of the laser beam Lb in the laser beam image and the position of the surface TA of the member T in the laser beam image. The distance to the TA is detected as the out-of-focus distance D4 of the laser beam Lb. The detection unit 52 does not have to detect the out-of-focus distance D4 based on the laser beam image. For example, the position of the focusing point SP of the laser beam Lb is specified in advance with an existing beam profiler, and the distance from the end 18bA of the light irradiation unit 18b to the focusing point SP is obtained. The detection unit 52 calculates the distance between the end 18bA of the light irradiation unit 18b and the surface TA of the member T from the image captured by the second imaging unit 34B, and collects light from the end 18bA of the light irradiation unit 18b from that distance. The out-of-focus distance D4 may be detected by subtracting the distance to the point SP.

The detection unit 52 determines whether the out-of-focus distance D4 of the laser beam Lb detected in this way is abnormal. When the detection value of the defocusing distance D4 of the laser beam Lb is smaller than a preset threshold value (hereinafter, referred to as a distance threshold value), the detection unit 52 determines that the defocusing distance D4 of the laser beam Lb is abnormal. To do. When the detected value of the out-of-focus distance D4 of the laser beam Lb is equal to or greater than the distance threshold value, the detection unit 52 determines that the position information of the focusing point SP of the laser beam Lb is normal (not abnormal). The distance threshold value here may be set arbitrarily, but if the out-of-focus distance D4 of the laser beam Lb is smaller than the distance threshold value, there is a high possibility that pores having abnormal quality will be formed. May be set.

Further, the detection unit 52 detects the inclination angle θ of the light irradiation unit 18b based on the image captured by the second imaging unit 34B while the light irradiation unit 18b is irradiating the laser light Lb. The inclination angle θ of the light irradiation unit 18b is an angle of the central axis AXb of the light irradiation unit 18b with respect to the Z-axis direction, and the straight line AZ is a straight line along the central axis AXb and the Z-axis direction of the light irradiation unit 18b. It can be said that it is an angle to make. More specifically, as shown in FIG. 42, a straight line along the Z-axis direction is defined as a straight line AZ, and the end portion of the light irradiation portion 18b opposite to the end portion 18bA is designated as an end portion 18bB. In this case, the inclination angle θ becomes a positive value when the light irradiation unit 18b is inclined toward the feed direction (Y-axis direction) of the light irradiation unit 18b from the end portion 18bB toward the end portion 18bA. When the irradiation unit 18b is inclined in the direction opposite to the feeding direction of the light irradiation unit 18b from the end portion 18bB toward the end portion 18bA, it becomes a negative value. That is, the inclination angle θ of the light irradiation unit 18b can be said to be a forward angle indicating how much the end portion 18bA of the light irradiation unit 18b is inclined toward the feed direction side. For example, the detection unit 52 extracts the image of the light irradiation unit 18b from the image captured by the second imaging unit 34B, and detects the central axis of the image of the light irradiation unit 18b. The detection unit 52 detects the angle between the central axis of the image of the light irradiation unit 18b and the reference axis in the image captured by the second imaging unit 34B as the tilt angle θ. The reference axis here is an image that has been preset as an axis along the Z-axis direction in the image captured by the second imaging unit 34B. However, the method for detecting the tilt angle θ is not limited to this. For example, the tilt angle θ is detected by the arithmetic unit 38 receiving the value of a sensor such as an encoder measuring the tilt of the light irradiation unit 18b in the manufacturing apparatus 10b. You may.

The detection unit 52 determines whether the inclination angle θ of the light irradiation unit 18b detected in this way is abnormal. When the detection value of the tilt angle θ of the light irradiation unit 18b is smaller than a preset threshold value (hereinafter referred to as an angle threshold value), the detection unit 52 determines that the tilt angle θ of the light irradiation unit 18b is abnormal. .. When the detection value of the tilt angle θ of the light irradiation unit 18b is equal to or greater than the angle threshold value, the detection unit 52 determines that the tilt angle θ of the light irradiation unit 18b is normal (not abnormal). The angle threshold value here may be set arbitrarily, but it may be set based on the fact that if the inclination angle θ is smaller than the angle threshold value, the possibility that the opening F1 that causes quality abnormality is formed increases.

Next, the processing flow by the arithmetic unit 38 according to the third embodiment will be described. FIG. 43 is a flowchart illustrating a processing flow of the arithmetic unit according to the third embodiment. As shown in FIG. 43, the arithmetic unit 38 according to the third embodiment is set by the device control unit 50 with the trigger that the manufacturing device 10 starts joining the first member T1 and the second member T2. The imaging unit 34 starts capturing an image at the frame rate (step S190). When the manufacturing apparatus 10 starts joining the first member T1 and the second member T2, the device control unit 50 causes the irradiation unit 32Aa to irradiate the region R0a (see FIG. 41) with light La, and causes the first imaging unit 34Aa to irradiate the region R0a (see FIG. 41) with light La. An image within the range of the imaging region R1a (see FIG. 41) is imaged, and the second imaging unit 34B is made to image an image within the range of the imaging region R2a (see FIGS. 41 and 42). The device control unit 50 causes the irradiation unit 32Aa to irradiate the irradiation unit 32Aa with light La so that the intensity periodically changes with the passage of time until the manufacturing apparatus 10 stops joining the first member T1 and the second member T2. At a predetermined frame rate, the first imaging unit 34Aa is made to image an image within the range of the imaging region R1a, and the second imaging unit 34B is made to image an image within the range of the imaging region R2a. In the third embodiment, the timing at which the light irradiation unit 18b starts irradiating the member T with the laser beam Lb may be the timing at which the joining between the first member T1 and the second member T2 is started.

When the imaging unit 34 captures an image, the arithmetic unit 38 acquires the image data generated by the imaging unit 34 by the detecting unit 52, and sets the member states of the holes, the openings, and the thickness of the compound F3. An abnormality is detected (step S192). The detection unit 52 detects the member state during the joining by the manufacturing apparatus 10. Each time the image pickup unit 34 captures an image, the detection unit 52 acquires the image data generated by the image pickup unit 34 and detects the member state. From the image data Pa generated by the first imaging unit 34Aa, the detection unit 52 describes the pores formed inside the member T, the thickness of the compound F3, and the opening F1 formed on the surface T3A of the joint portion T3. Is detected as a member state. As in the first embodiment, the detection unit 52 may detect the opening F1 formed on the surface T3A of the joint portion T3 from the image data P generated by the second imaging unit 34B. The detection unit 52 determines whether the detected member state is abnormal.

The arithmetic unit 38 is inside the member T when there is an abnormality in the detection result of the member state (step S194; Yes), that is, when at least one of a hole, an opening, and an abnormality in the thickness of the compound F3 is detected. It is determined whether the vacancies are detected and whether the opening F1 is detected on the surface T3A of the joint portion T3 (step S196). When the hole and the opening F1 are detected in the arithmetic unit 38 (step S196; Yes), the detection unit 52 sets the defocus distance D4 of the laser beam Lb and the defocus distance of the light irradiation unit 18b as a tool state. Detect as D4. The arithmetic unit 38 causes the second imaging unit 34B to capture an image in a state where the joining by the manufacturing apparatus 10 is performed, that is, in a state where the laser beam Lb is irradiated, and the image captured by the second imaging unit 34B. Based on the above, the out-of-focus distance D4 of the laser beam Lb and the inclination angle θ of the light irradiation unit 18b are detected.

When the out-of-focus distance D4 and the inclination angle θ are detected as the tool state, the arithmetic unit 38 operates during joining of the manufacturing device 10 based on the member state and the tool state detected by the detection unit 52 by the change information generation unit 56. Change information for changing the condition is generated, and the generated change information is transmitted to the control device 30 (step S208). The method of generating change information will be described later.

Whether or not the vacancy is detected by the arithmetic unit 38 when both the vacancy and the opening F1 are not detected (step S196; No), that is, when at least one of the vacancy and the opening F1 is not detected. To judge. When a vacancy is detected (step S200; Yes), the arithmetic unit 38 detects the out-of-focus distance D4 as a tool state (step S202). That is, when the opening F1 is not detected but the vacancy is detected, the arithmetic unit 38 detects the out-of-focus distance D4 as a tool state. When the out-of-focus distance D4 is detected as the tool state, the arithmetic unit 38 changes the operating condition during joining of the manufacturing device 10 based on the member state and the tool state detected by the detection unit 52 by the change information generation unit 56. The change information of the above is generated, and the generated change information is transmitted to the control device 30 (step S158). The method of generating change information will be described later.

The arithmetic unit 38 determines whether or not the opening F1 is detected when the vacancy is not detected (step S200; No). When the opening F1 is detected (step S204; Yes), the arithmetic unit 38 detects the inclination angle θ as a tool state (step S206). When the inclination angle θ is detected as the tool state, the arithmetic unit 38 changes the operating condition during joining of the manufacturing apparatus 10 based on the member state and the tool state detected by the detection unit 52 by the change information generation unit 56. The change information is generated, and the generated change information is transmitted to the control device 30 (step S208). The method of generating change information will be described later.

When the opening F1 is not detected (step S204; No), that is, when no vacancies are detected and the opening F1 is not detected, the arithmetic unit 38 does not detect the tool state, but based on the detected member state, the manufacturing device 10 Change information for changing the operating conditions during joining is generated, and the generated change information is transmitted to the control device 30 (step S158). In other words, if the abnormality in the member state is other than the detection of the vacancies and the opening F1, here, if the abnormality in the thickness of the compound F3 is detected and the vacancies and the opening F1 are not detected, the arithmetic unit 38 will perform the arithmetic unit 38. Change information is generated based on the member state without detecting the tool state. The method of generating change information will be described later.

The control device 30 of the manufacturing device 10 changes the operating conditions based on the change information transmitted from the arithmetic unit 38. When the joining between the first member T1 and the second member T2 is completed after the change information is generated and transmitted (step S210; Yes), this process is terminated. When the joining between the first member T1 and the second member T2 is not completed (step S210; No), the control device 30 of the manufacturing apparatus 10 has the first member T1 and the second member T1 and the second member under the operating conditions changed based on the change information. Continuing the bonding with T2, the arithmetic unit 38 returns to step S190 and continues imaging by the imaging unit 34.

In step S194, if there is no abnormality in the detection result of the member state (step S194; No), that is, if none of the pores and openings and the abnormality in the thickness of the compound F3 are detected, the step S210 is performed. When moving and ending the joining between the first member T1 and the second member T2 (step S210; Yes), when the present process is finished and the joining between the first member T1 and the second member T2 is not finished (step). S210; No), the process returns to step S140, and irradiation of light La by the irradiation unit 32Aa and imaging by the imaging unit 34 are continued. The order of detecting the vacancies, openings, and abnormal thickness of compound F3 may be arbitrary.

In the description of FIG. 43, if there is no abnormality in the detection result of the member state, the change information is not generated. However, as in the second embodiment, the arithmetic unit 38 is based on the member state detected during the joining between the first member T1 and the second member T2 even when there is no abnormality in the detection result of the member state. , Generate change information for changing the operating conditions for the next joining after the joining between the first member T1 and the second member T2 (joining when the detection unit 52 detects the member state) is completed. You may. Further, the arithmetic unit 38 detects the tool state (here, the out-of-focus distance D4 and the inclination angle θ) by using the detection of the hole or the opening F1 as a trigger, but the present invention is not limited to this, and for example, the second Every time the image pickup unit 34B takes an image at a predetermined frame rate, the tool state may be sequentially detected based on the image data of the second image pickup unit 34B.

Next, the generation of change information according to the third embodiment will be described. In the third embodiment, the joint state determination unit 54 determines the keyhole formed in the member T, the amount of heat input to the member T, and the amount of heat input to the member T as the joint state based on the member state and the tool state detected by the detection unit 52. It is determined whether the heat input position to the member T and the solidified state of the molten pool are defective. However, the joint state determination unit 54 is at least one of the keyhole formed in the member T, the amount of heat input to the member T, the heat input position to the member T, and the solidified state of the molten pool as the joint state. It may be determined whether one is defective. The keyhole refers to a hole temporarily formed by evaporating the member T by the laser beam Lb when the member T is irradiated with the laser beam Lb. The solidified state of the molten pool refers to how the molten pool is solidified, for example, the position where the molten pool is solidified. The defective modes of such a bonding state include a keyhole defect in which the formed keyhole is too deep, an excessive heat input in which the amount of heat input to the member T is excessive, and a defective heat input position in the member T. At least one of a poor heat input position and a poor solidification state of the molten pool can be mentioned.

Further, in the third embodiment, the change information generation unit 56 is based on the member state and the tool state detected by the detection unit 52, and more specifically, based on the joint state determined by the joint state determination unit 54 to be defective. Generate change information to change the operating conditions. The operating conditions in the third embodiment include the defocusing distance D4 of the laser beam Lb, the feed rate of the light irradiation unit 18b, the laser path, the inclination angle θ, and the output of the laser beam Lb. However, the operating condition in the third embodiment may be at least one of the out-of-focus distance, the feed rate, the laser path, the inclination angle θ, and the output of the laser beam Lb. The feed rate is the relative moving speed of the light irradiation unit 18b with respect to the member T in the feed direction (Y-axis direction in this embodiment). The laser path is a movement path of the light irradiation unit 18b with respect to the member T in a state of irradiating the laser light Lb, and more specifically, light with respect to the member T in a direction orthogonal to the feed direction (here, the X-axis direction). This is the relative position of the irradiation unit 18b. The output of the laser beam Lb refers to the power density of the laser beam Lb.

44 to 46 are flowcharts illustrating the generation of change information according to the third embodiment. FIG. 44 shows a method of generating change information when a hole is formed inside the member T. That is, FIG. 44 illustrates the details of the method of generating change information when a vacancy is detected in step S208 of FIG. 43. As shown in FIG. 44, when the detection unit 52 determines that a hole is formed inside the member T, the joint state determination unit 54 determines that the defocusing distance D4 of the laser beam Lb is set by the detection unit 52. When it is determined that the abnormality is occurring (step S222; Yes), it is determined that a keyhole defect has occurred in the defective mode of the joining state (step S224). When the change information generation unit 56 determines that a keyhole defect has occurred, it generates change information to lengthen the out-of-focus distance D4 of the laser beam Lb (step S226). In other words, when the change information generation unit 56 determines that the detection unit 52 has a hole formed inside the member T and determines that the out-of-focus distance D4 of the laser beam Lb is abnormal. , Generates change information to lengthen the out-of-focus distance D4 of the laser beam Lb.

Further, when the detection unit 52 determines that the out-of-focus distance D4 of the laser beam Lb is not abnormal (step S222; No), the joint state determination unit 54 causes excessive heat input among the defective modes of the joint state. (Step S228). When the change information generation unit 56 determines that excessive heat input has occurred, it generates change information to the effect that the feed rate of the light irradiation unit 18b is increased (step S230). That is, when the change information generation unit 56 determines that the detection unit 52 has a hole formed inside the member T and determines that the out-of-focus distance D4 of the laser beam Lb is not abnormal, the light is emitted. Change information to increase the feed rate of the irradiation unit 18b is generated.

Here, as the cause of forming the vacancies, there are a plurality of causes such as the keyhole being too deep and excessive heat input to the member T. If the keyhole is too deep, at least a part of the keyhole may not be filled with the molten pool, and at least a part of the keyhole may remain as a hole to form a hole. Further, if the amount of heat input to the member T is too large, the space where the member T evaporates becomes large, and the evaporated portion remains as a hole, which may form a hole. On the other hand, the arithmetic unit 38 according to the present embodiment sets a change condition capable of suppressing the formation of the vacancy by identifying the cause of the vacancy based on the tool state as well as the member state of whether the vacancy is formed. Generate. That is, when the arithmetic unit 38 detects the vacancy and the defocus distance D4 is smaller than the distance threshold value, the focus SP of the laser beam Lb is too close to the member T, so that the keyhole becomes deep and empty. It is judged that a hole is formed. Then, in this case, the arithmetic unit 38 generates the change information to lengthen the defocus distance D4 of the laser beam Lb, thereby lengthening the defocus distance D4 in the manufacturing device 10b and shortening the keyhole. It suppresses the formation of vacancies. Further, when the arithmetic unit 38 detects the vacancy and the out-of-focus distance D4 is not smaller than the distance threshold value, it is determined that the vacancy is formed due to excessive heat input, not due to the keyhole defect. To do. Then, in this case, the arithmetic unit 38 generates change information to the effect that the feed rate of the light irradiation unit 18b is increased. By increasing the feed rate of the light irradiation unit 18b, the manufacturing apparatus 10b can shorten the irradiation time of the laser light Lb, reduce the amount of heat input, and suppress the formation of pores. Further, the arithmetic unit 38 may generate change information to reduce the output of the laser beam Lb when the vacancy is detected and the out-of-focus distance D4 is not abnormal (not smaller than the distance threshold value). Good. Even if the output of the laser light Lb is reduced, the amount of heat input by the laser light Lb can be reduced, so that the formation of pores can be suppressed.

FIG. 45 shows a method of generating change information when the opening F1 is formed in the member T. That is, FIG. 45 illustrates the details of the method of generating change information when the opening F1 is detected in step S208 of FIG. 43. As shown in FIG. 45, when the detection unit 52 determines that the opening F1 is formed in the member T, the joint state determination unit 54 causes the detection unit 52 to have an abnormal inclination angle θ of the light irradiation unit 18b. When it is determined that (step S232; Yes), it is determined that the molten pool is poorly solidified among the defective modes of the joint state (step S234). When the change information generation unit 56 determines that the molten pool is poorly solidified, it generates change information to increase the inclination angle θ of the light irradiation unit 18b (step S236). In other words, when the detection unit 52 determines that the opening F1 is formed in the member T and the change information generation unit 56 determines that the inclination angle θ of the light irradiation unit 18b is abnormal, the light is emitted. Change information for increasing the inclination angle θ of the irradiation unit 18b is generated.

Further, when the detection unit 52 determines that the inclination angle θ of the light irradiation unit 18b is not abnormal (step S232; No), the joint state determination unit 54 has a poor solidification of the molten pool in the poor mode of the joint state. Is determined to be occurring (step S238). When the change information generation unit 56 determines that the molten pool is poorly solidified, it generates change information to the effect that the output of the laser beam Lb is reduced (step S240). That is, when the detection unit 52 determines that the opening F1 is formed in the member T and the change information generation unit 56 determines that the inclination angle θ of the light irradiation unit 18b is not abnormal, the laser light Lb Generate change information to the effect that the output of is reduced.

Here, the cause of the formation of the opening F1 is poor solidification of the molten pool. For example, a molten pool is formed on the surface TA of the member T, but if at least a part of the melt in the molten pool jumps out of the molten pool to another position due to poor solidification of the molten pool, the melt in the molten pool is insufficient. Then, the solidified portion of the molten pool may become a dent and become an opening F1. The melt in the molten pool may jump out of the molten pool when the inclination angle θ of the light irradiation unit 18b is small or when the output of the laser beam Lb is high. When the arithmetic unit 38 according to the present embodiment detects the opening F1 and the inclination angle θ is smaller than the angle threshold value, it is determined that the molten pool is poorly solidified because the inclination angle θ is small. Then, in this case, the arithmetic unit 38 increases the inclination angle θ in the manufacturing apparatus 10b by generating the change information to increase the inclination angle θ, suppresses the solidification failure of the molten pool, and opens the opening F1. The formation of can be suppressed. Further, the arithmetic unit 38 detects the opening F1 and generates change information to the effect that the output of the laser beam Lb is reduced when the inclination angle θ is not smaller than the angle threshold value, so that the manufacturing apparatus 10b It is possible to reduce the output of the laser beam Lb, suppress the solidification failure of the molten pool, and suppress the formation of the opening F1.

FIG. 46 shows a method of generating change information when the thickness of compound F3 is abnormal. That is, FIG. 46 describes the details of the method of generating change information when the thickness abnormality of the compound F3 is detected in step S208 of FIG. 43. As shown in FIG. 46, when the detection unit 52 determines that the thickness of the compound F3 is abnormal, the bonding state determination unit 54 determines that the heat input position is defective in the bonding state defect mode. Determine (step S242). When the change information generation unit 56 determines that the heat input position is defective, it generates change information to correct the laser path (step S244). In other words, the change information generation unit 56 generates change information to correct the laser path when the detection unit 52 determines that the thickness of the compound F3 is abnormal. When the thickness of the compound F3 is abnormal, that is, when the compound F3 is too thick, there is a defect in the heat input position. When the arithmetic unit 38 according to the present embodiment detects an abnormality in the thickness of the compound F3, the arithmetic unit 38 adjusts the heat input position by generating change information to correct the laser path, so that the thickness of the compound F3 becomes thicker. It can be suppressed.

If the arithmetic unit 38 according to the present embodiment has an abnormal thickness of the compound F3 and pores are formed, the processing after step S222 in FIG. 44 may be performed. That is, when the arithmetic unit 38 determines that the thickness of the compound F3 is abnormal and the pores are formed and the out-of-focus distance D4 of the laser beam Lb is abnormal, the keyhole It is possible to determine that a defect has occurred and generate change information to lengthen the out-of-focus distance D4 of the laser beam Lb. Then, when the arithmetic unit 38 determines that the thickness of the compound F3 is abnormal and the pores are formed and the out-of-focus distance D4 of the laser beam Lb is not abnormal, the heat input is excessive. It is possible to generate change information to the effect that the feed rate of the light irradiation unit 18b is increased or the output of the laser light Lb is to be decreased.

Further, when the thickness of the compound F3 is abnormal and the opening F1 is formed, the arithmetic unit 38 may perform the processes after step S232 in FIG. 45. That is, when the arithmetic unit 38 determines that the thickness of the compound F3 is abnormal and the opening F1 is formed and the inclination angle θ of the light irradiation unit 18b is abnormal, the molten pool It is possible to determine that a solidification defect has occurred and generate change information to increase the inclination angle θ of the light irradiation unit 18b. Further, when the arithmetic unit 38 determines that the thickness of the compound F3 is abnormal and the opening F1 is formed and the inclination angle θ of the light irradiation unit 18b is not abnormal, the molten pool is solidified. It is possible to determine that a defect has occurred and generate change information to the effect that the output of the laser beam Lb is reduced.

Then, the arithmetic unit 38 may execute the process of FIG. 46 when the holes are not formed and the opening F1 is not formed. That is, when the thickness of the compound F3 is abnormal and the pores are not formed and the opening F1 is not formed, the arithmetic unit 38 determines that the heat input position is defective, and determines that the laser path is defective. Generate change information to correct.

As described above, even when the first member T1 and the second member T2 are joined by a method other than friction stir welding as in the third embodiment, the arithmetic unit 38 uses the member information detected by the detection unit 52. Since change information is generated based on the tool information and the tool information, it can appropriately contribute to quality improvement.

The optical elements such as lenses and filters described in the embodiments and modifications so far (for example, the lenses 62A and 62B, the photosynthesis unit 64, the light diffusing unit 66, and the lens 68 described in the first embodiment) are , Each is not limited to one optical element (for example, a lens or a filter), and may be a combination of existing optical elements (for example, a lens or a filter).

Although the manufacturing system 1 of the above embodiment is processed by one device, a plurality of them may be combined. FIG. 47 is a schematic diagram showing the configuration of a system having a manufacturing system. Next, the system 300 having the manufacturing system 1 will be described with reference to FIG. 47. The system 300 includes a plurality of manufacturing systems 1 (three in FIG. 47) and a program creation device 302. The manufacturing system 1 and the program creation device 302 are connected by a wired or wireless communication line. The program creation device 302 creates various settings and programs created by the control device of the manufacturing system 1 described above. The program creation device 302 outputs the created program and data to the manufacturing system 1. The manufacturing system 1 acquires various programs from the program creation device 302, and performs processing using the acquired data and programs. The system 300 can effectively utilize the created data and the program by executing the manufacturing in the manufacturing system 1 by using the data and the program created by the manufacturing system 1 and the program creating apparatus 302.

Next, a system including the above-mentioned manufacturing system will be described with reference to FIG. 48. FIG. 48 is a block configuration diagram of the system. The system 200 of this embodiment includes an inspection device 201, a design device 202, a manufacturing system 203 which is a manufacturing system 1 as described in the above embodiment, a control device (inspection device) 204, and a repair device 205. To be equipped. The control device 204 includes an internal structure storage unit 210 and a determination unit 211.

The design device 202 creates design information regarding the shape and composition of the member T, and transmits the created design information to the manufacturing system 203. Further, the design device 202 stores the created design information in the internal structure storage unit 210 of the control device 204.

The manufacturing system 203 joins the members to each other to create the member T based on the design information input from the design device 202. The inspection device 201 inspects the created member T and transmits the inspection result (for example, image data) to the control device 204.

The internal structure storage unit 210 of the control device 204 stores the design information. The determination unit 211 of the control device 204 reads the design information from the internal structure storage unit 210. The determination unit 211 compares the measurement result of the shape of the member T received from the inspection device 201 with the design information read from the internal structure storage unit 210. The determination unit 211 determines whether or not the member T is molded according to the design information based on the comparison result. In other words, the determination unit 211 determines whether or not the created member T is a non-defective product. The determination unit 211 determines whether or not the member T can be repaired when the member T is not formed according to the design information. When the member T can be repaired, the determination unit 211 calculates the defective portion and the repair content based on the comparison result, and transmits the information indicating the defective portion and the information indicating the repair content to the repair device 205.

The repair device 205 repairs the defective portion of the structure based on the information indicating the defective portion received from the control device 204 and the information indicating the repair content.

FIG. 49 is a flowchart showing the flow of processing by the system. In the system 200, first, the design device 202 creates design information regarding the member T (step S1010). Next, the manufacturing system 203 creates the member T based on the design information (step S1020). Next, the inspection device 201 inspects (measures) the shape of the created member T (step S1030). Next, the determination unit 211 of the control device 204 inspects whether or not the member T is created according to the design information by comparing the inspection result obtained by the inspection device 201 with the above design information (step S1040). ).

Next, the determination unit 211 of the control device 204 determines whether or not the created member T is a non-defective product (step S1050). When the determination unit 211 determines that the created member T is a non-defective product (Yes in step S1050), the system 200 ends the process. Further, when the determination unit 211 determines that the created member T is not a non-defective product (No in step S1050), the determination unit 211 determines whether or not the created member T can be repaired (step S1060).

When the determination unit 211 determines that the created member T can be repaired (Yes in step S1060), the repair device 205 repairs the member T (step S1070), and returns to the process of step S1030. When the determination unit 211 determines that the created member T cannot be repaired (No in step S1060), the system 200 ends the process and collects the defective product. With the above, the system 200 ends the processing of the flowchart shown in FIG. 49.

In the system 200 of the present embodiment, since the manufacturing system 203 can manufacture the internal structure of the member T with high accuracy, it is possible to determine whether or not the created member T is a non-defective product. Further, the system 200 can repair the member T when the member T is not a non-defective product.

The repair process executed by the repair device 205 in the present embodiment may be replaced with a process in which the manufacturing system 203 re-executes the manufacturing process. At that time, if the determination unit 211 of the control device 204 determines that the repair can be performed, the manufacturing system 203 re-executes the manufacturing process.

Although suitable embodiments and modifications according to the present invention have been described above with reference to the accompanying drawings, it goes without saying that the present invention is not limited to such examples. The various shapes and combinations of the constituent members shown in the above-mentioned examples are examples, and can be variously changed based on design requirements and the like within a range not deviating from the gist of the present invention. Each component of the above-described embodiment can be combined as appropriate. In addition, some components may not be used. In addition, to the extent permitted by law, the disclosures of all published gazettes and US patents relating to the shape measuring devices cited in each of the above-described embodiments and modifications are incorporated as part of the description of the main text. All other embodiments, operational techniques, and the like made by those skilled in the art based on the above-described embodiments are included in the scope of this embodiment.

1 Manufacturing system 10 Manufacturing equipment (friction stir welding equipment)
12 Arithmetic system 18 Tool 30 Control device 32 Irradiation unit 34 Imaging unit 38 Arithmetic unit 52 Detection unit 54 Joining state judgment unit 56 Change information generation unit AR0 Irradiation area AR1, AR2 Detection area T member T1 First member T2 Second member T3 Joining Department

Claims (21)

An arithmetic unit used in a friction stir welding device that joins a first member and a second member by bringing a rotated tool into contact with at least one of the first member and the second member.
Detection for detecting a member state which is a state of at least a part of the first member, the second member, and a joint portion between the first member and the second member, and a tool state which is a state of the tool. Department and
Change information generation for generating change information for changing the operating condition of the friction stir welding device from the first value to the second value based on the member state and the tool state detected by the detection unit. Department and
An arithmetic unit. The detection unit detects the member state during joining by the friction stir welding device, and detects the member state.
The arithmetic unit according to claim 1, wherein the change information generation unit generates the change information for changing the operating conditions during joining based on the member state during joining detected by the detection unit. The detection unit detects the member state during joining by the friction stir welding device, and detects the member state.
Based on the member state during joining detected by the detection unit, the change information generation unit completes the joining between the first member and the second member when the detection unit detects the member state. The arithmetic unit according to claim 1 or 2, wherein the change information for changing the operating conditions when joining the following members is generated. The detection unit determines whether the detection result of the member state during joining by the friction stir welding device is abnormal, and if it is abnormal, detects the tool state. The arithmetic unit according to any one of the following items. The detection unit detects whether the reflectance of light on the surface of the joint portion is abnormal as the member state, and when it detects that the reflectance is abnormal, the detection unit determines the tool state of the tool. The arithmetic unit according to any one of claims 1 to 4, which detects the shape. The detection unit detects whether the dimension of the groove of the probe at the tip of the tool and the outer diameter of the tool are abnormal as the tool state.
The change information generation unit
When the detection unit detects that the size of the groove is abnormal, it generates information to increase the rotation speed of the tool as the change information.
The arithmetic unit according to claim 5, wherein when the detection unit detects that the outer diameter of the tool is abnormal, information to the effect that the feed rate of the tool is reduced is generated as the change information. The detection unit detects burrs around the joint as the member state, and when the burr is detected, at least the first member, the second member, and the joint as the tool state. The arithmetic unit according to any one of claims 1 to 6, which detects the amount of the tool inserted into a part thereof. The detection unit detects whether the insertion amount of the tool is abnormal, and detects whether or not the insertion amount of the tool is abnormal.
The change information generation unit
When the detection unit detects that the insertion amount of the tool is abnormal, information to reduce the insertion amount of the tool is generated as the change information.
The arithmetic unit according to claim 7, wherein when the detection unit detects that the insertion amount of the tool is not abnormal, information to increase the feed rate of the tool is generated as the change information. The detection unit detects an opening formed in the joint portion as the member state, and detects the opening.
Any one of claims 1 to 8, wherein the change information generation unit generates information to the effect that the feed speed of the tool is reduced when the detection unit detects the opening as the change information. The arithmetic unit described in the section. The detection unit detects whether or not the machining path, which is the machining path by the tool for at least a part of the first member, the second member, and the joint portion, is abnormal as the member state.
The change information generation unit generates information to change the processing path as the change information when the detection unit detects that the processing path is abnormal, claims 1 to 9. The arithmetic unit according to any one of the above items. Claims 1 to claim 1, wherein the detection unit detects a vacancy inside the joint portion as the member state, and when the vacancy is detected, detects the shape of the tool as the tool state. The arithmetic unit according to any one of 10. The detection unit detects whether the dimension of the groove of the probe at the tip of the tool and the outer diameter of the tool are abnormal as the tool state.
The change information generation unit
When the detection unit detects that the size of the groove is abnormal, it generates information to increase the rotation speed of the tool as the change information.
The arithmetic unit according to claim 11, wherein when the detection unit detects that the outer diameter of the tool is abnormal, information to the effect that the feed rate of the tool is reduced is generated as the change information. The detection unit detects whether the thickness of the compound between the first member and the second member is abnormal as the member state.
When the detection unit detects that the thickness of the compound between the first member and the second member is abnormal, the change information generation unit has the first member, the second member, and the joint portion. The arithmetic unit according to any one of claims 1 to 12, which generates information for changing a machining path which is a machining path by the tool for at least a part of the above as the change information. The first aspect of the present invention further comprises a joint state determination unit for determining whether the joint state between the first member and the second member is defective based on the detection result of the member state and the tool state by the detection unit. Item 3. The arithmetic unit according to any one of Item 13. The joint state determination unit includes a mixed state of the first member and the second member, an amount of heat input to at least a part of the first member, the second member, and the joint portion, and the first member. The arithmetic unit according to claim 14, wherein at least one of a member, the second member, and a heat input position to at least a part of the joint portion is determined as the joint state. The change information generation unit includes the rotation speed of the tool, the feed rate of the tool, the insertion amount of the tool into at least a part of the first member, the second member, and the joint portion, and the first member. Claim 1 to claim that the change information is generated by using at least one of the machining paths, which are machining paths by the tool, for at least a part of the member, the second member, and the joint portion as the operating conditions. The arithmetic unit according to any one of 15. The arithmetic unit according to any one of claims 1 to 16.
It has an image pickup unit that images at least one of the first member, the second member, and the joint portion, and the tool.
The detection unit detects the member state and the tool state based on the captured image of the imaging unit.
Computational system. An irradiation unit that projects light is further provided on the surface of the first member, the second member, the joint portion, and the tool.
17. The image according to claim 17, wherein the imaging unit captures an image of at least one of the first member, the second member, and the joint portion and the surface of the tool on which light is projected by the irradiation unit. Computational system. The first member, the first member, said, further comprising a heating portion for heating at least one of the first member, the second member, the joint portion, and the surface of the tool. Receives infrared light from at least one of the second member, the joint, and the surface of the tool.
The calculation system according to claim 17, wherein the detection unit detects at least a part of the member state and the tool state based on the infrared light received by the imaging unit. The arithmetic system according to any one of claims 17 to 19.
A manufacturing system comprising the friction stir welding device. A calculation method used in a friction stir welding apparatus for joining a first member and a second member by bringing a rotated tool into contact with at least one of the first member and the second member.
To detect a member state which is a state of at least a part of the first member, the second member, and a joint portion between the first member and the second member, and a tool state which is a state of the tool. When,
Based on the detected member state and the tool state, change information for changing the operating condition of the friction stir welding device from the first value to the second value is generated.
Calculation method including.

Images