JP3137253B2 - Automatic welding method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic welding method for automatically repairing a defect such as a crack in a combustor, which is an aircraft engine component, when the defect occurs. .

[0002]

2. Description of the Related Art In a combustor (combustion cylinder) which is an aircraft engine part, a combustion chamber of a gas turbine, a heat exchanger of a nuclear reactor, a refrigerator, and the like, a large number of holes are regularly arranged in a portion where a thin tube passes. The provided panel is used. Such panels are often used under high temperature, high pressure, cryogenic conditions, or under severe conditions where there is a large temperature difference, so that cracks may form around the holes through which the tubules pass. .
Leaving such cracks may lead to serious accidents. For this reason, when a crack occurs, an operation of welding and repairing the crack is performed. Such cracks are very thin and are often invisible to the naked eye.Therefore, after cleaning the panel, immerse it in a special liquid to make it easier to see the cracks.Furthermore, a groove for welding with a grinder etc. Form. Conventionally, after forming such a groove, a human manually welds each groove.

[0003]

As described above, conventionally,
Since humans manually welded each groove, there was a problem that work efficiency was poor, work quality could not be made uniform, and recurrence of cracks could not be effectively prevented. For this reason, it is desired to realize an automatic welding system that allows a welding robot to automatically perform such a welding operation. However, in the case of repair welding, cracks are generated in various shapes at various positions, so the location and shape etc. of each groove are different, and if repair welding is repeated many times in the past However, since the surface of the panel may be deformed, it is difficult to teach the welding robot the welding conditions for the groove and the distance to the groove.

The present invention has been made based on the above circumstances, and provides an automatic welding method capable of automatically welding a defective portion generated in a heat exchanger such as a combustor according to the shape of the defective portion. It is intended to provide.

[0005]

According to a first aspect of the present invention, there is provided an automatic welding method comprising forming a groove in a defective portion generated in a heat exchanger and performing repair welding using a welding robot. A welding method, a step of obtaining distance information from the welding torch attached to the welding robot to the groove, and performing a predetermined process on an image obtained by imaging the heat exchanger. Obtaining the image information about the groove, obtaining welding position information that is information for moving the welding torch to a welding portion based on the image information of the groove, A step of recognizing which of the plurality of pre-classified groove patterns the groove portion corresponds to based on the image information, and the distance information to the groove portion, the welding position information, and the recognized The welding based on the groove pattern Bot is operated, and is characterized in that it comprises the the steps of welding the welding groove.

An automatic welding method according to a second aspect of the present invention is the automatic welding method according to the first aspect, further comprising a step of determining whether or not the interval between the groove portions is equal to or greater than a predetermined value. It is a feature.

According to a third aspect of the present invention, in the automatic welding method according to the first or second aspect, when information on the distance to the groove is obtained, an arbitrary plurality of positions on the heat exchanger are determined. The distance to the groove is calculated by measuring and interpolating between the plurality of positions.

According to a fourth aspect of the present invention, in the automatic welding method according to the first, second or third aspect, the welding robot has a positioner for tilting and rotating the heat exchanger. It is assumed that.

According to a fifth aspect of the invention, there is provided an automatic welding method according to the first, second, third or fourth aspect, wherein the heat exchanger is a combustor which is an aircraft engine part. Things.

[0010]

According to the first aspect of the present invention, a plurality of groove patterns classified in advance for the groove portion are prepared according to the above-mentioned configuration, and the groove portion is formed based on an image obtained by imaging the heat exchanger. By recognizing which of the groove patterns corresponds to, it is possible to easily determine welding conditions for the groove based on the recognized groove pattern. In addition, by operating the welding robot based on the distance information to the groove and the welding position information and the recognized groove pattern, the groove can be automatically and reliably welded, and the predetermined groove can be welded. With regard to the groove portion corresponding to the leading pattern, welding can always be performed under constant conditions, and the welding quality can be kept constant.

According to the second aspect of the present invention, the above structure further comprises a step of judging whether or not the interval between the groove portions is equal to or larger than a predetermined value. Considering the positional relationship with some other groove,
It can be determined whether the welding robot can actually weld the groove.

According to the third aspect of the present invention, when the distance information to the groove portion is obtained by the above-described configuration, by measuring arbitrary plural positions on the heat exchanger and interpolating between the plural positions. , The distance to the groove can be easily calculated.

According to the fourth aspect of the present invention, the welding robot has a positioner for tilting and rotating the heat exchanger, so that the groove is welded by tilting and rotating the heat exchanger. It can be moved to a position that is easy to do.

According to the fifth aspect of the present invention, the heat exchanger is a combustor, which is an aircraft engine part, so that a defective portion can be welded with a constant high quality, and the recurrence of the defect can be effectively performed. Therefore, occurrence of an aircraft accident can be effectively prevented.

[0015]

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a schematic configuration diagram of an automatic welding system to which an automatic welding method according to one embodiment of the present invention is applied.

The automatic welding system shown in FIG. 1 includes a welding robot 10, a robot controller 40, an image processing device 50, a central control unit 60, a storage unit 70, and a CRT display device 80.

In this embodiment, a case where a combustor, which is an aircraft engine part, is used as an object to be welded will be considered. FIG. 2 is a schematic plan view of a part of the combiner, and FIG. 3 is a schematic sectional view of the combiner in the direction of arrows AA. As shown in FIGS. 2 and 3, the combiner 2 has, for example, six cylindrical panels 4a, 4b,.
, 4f, and a large number of holes 6 are formed in each of the panels 4a, 4b,. Here, the side of the panel 4a is “upper”, and the side of the panel 4f is “lower”.
I will call it. Adjacent panels are partially overlapped and joined so that the end of the lower panel is located outside the end of the upper panel. Then, each panel 4a,
, 4f are arranged. In one panel, the angle between the centers of adjacent holes 6 with respect to a point on its central axis is 12 °. That is, one panel has holes 6
Are formed. Further, there are two types of shapes of the holes 6, a circle and an ellipse. For example, the holes 6 are formed in a circular shape from the panels 4a to 4d, and the holes 6 are formed in an elliptical shape in the panel 4e. In addition, each panel 4a,
The width of 4b,..., 4e is about 50 mm, and the width of panel 4f is about 30 mm.

By the way, when the combiner 2 is used under conditions such as high temperature and high pressure, as shown in FIG. 2, mainly the periphery of the hole 6 and each panel 4a, 4b,
.., 4f has a crack 8 at the lower edge portion. On the other hand, such a crack 8 rarely occurs alone in a portion having no hole 6 and rarely reaches the upper edge of the panel. If such a crack 8 occurs, leaving the crack 8 as it is may lead to a serious accident. Therefore, the portion where the crack 8 has occurred is repaired by welding. Although the crack 8 is exaggeratedly shown in FIG. 2, the actual crack 8 is very thin and cannot be easily seen by the naked eye. For this reason, when repairing by welding, first, a groove is formed in a portion where the crack 8 has occurred using a grinder. The groove is formed with a certain width, for example, about 3 mm. At this time, since a special grinder is used, if the crack 8 is long, a penetrating portion penetrating to the back surface of the panel is formed near the portion where the grinder was first applied at the groove. And, as the crack 8 becomes longer, the penetrating portion formed in the groove becomes larger. On the other hand, if the crack 8 is short, a penetration portion cannot be formed at the groove. FIG. 4 is a diagram illustrating an example of a groove portion formed in a portion where a crack 8 has occurred. FIG. 3A is a schematic plan view of the groove portion, and FIG. 3B is a schematic cross-sectional view of the groove portion in the direction of arrows BB. The groove 110 is formed by scraping a crack generated at a lower edge portion of one panel 4. On the left side of the groove 110, there is an inclined portion 112 caused by the curved shape of the tip of the grinder. On the right side of the groove 110, there is a penetrating portion 114 penetrating to the back surface of the panel 4. After forming the groove in this way,
The groove is welded from above.

In particular, in this embodiment, a case is considered in which a combustor that has been repaired and welded many times in the past is repaired again by welding. In such a case, if the formation of the groove and welding are repeated many times, the surface of the panel will naturally have irregularities as shown in FIG. For this reason, during welding, unless the distance between the welding torch and the base material is determined according to the distance to the groove, reliable welding cannot be performed.
Moreover, since cracks occur in various parts, it is necessary to specify the position of the groove, its shape, etc. every time repair welding is performed, and to determine the welding conditions for each groove based on that. Due to such circumstances, it is difficult to perform repair welding.

As shown in FIG. 1, the welding robot 10
It has a robot body 12 and a positioner 32. The positioner 32 holds the combiner 2,
The combiner 2 can be tilted and the combiner 2 can be rotated around its central axis. A welding torch 22 and a distance sensor 2 are provided on the arm of the robot body 12.
4. CCD camera 26 as image pickup means, illumination device 28
Is provided. The welding torch 22 is, for example, TIG
This is for performing overlay welding. Distance sensor 24
In order to keep the distance between the welding torch 22 and the combuster 2 at a constant distance of, for example, about 0.5 mm during welding,
Distance to Combustor 2 (hereinafter also referred to as height)
Is to measure. The CCD camera 26 is for imaging the combiner 2. The lighting device 28
The light source irradiates the combiner 2, and the combiner 2 is imaged under this illumination. The image obtained by the CCD camera 26 is sent to the image processing device 40, where predetermined image processing is performed. By rotating the arm of the robot body 12, the welding torch 22, the distance sensor 2
4. The CCD camera 26 can be separately moved to the position facing the converter 2 as needed. Here, the operations of height detection, image recognition, and welding are sequentially performed for each panel.

When picking up an image of each panel, the positioner 32 rotates the combiner 2 by about 6 °, and picks up an image in an image field of, for example, 50 mm × 50 mm. FIG. 6 shows an example of a captured image. With such a visual field size, an image in which one hole 6 entirely enters the image visual field and an image in which the hole 6 entirely does not enter the image visual field are obtained in order. When measuring the height of each panel, the positioner 32 rotates the combiner 2 by 2 °, and performs measurement at two measurement points above and below the panel 4 as shown in FIG. That is, the height is measured at a total of six measurement points, three in each of the upper and lower points, in one image visual field. The data on the height obtained here is sent to the central control unit 60.

The image processing device 50 includes a groove position detector 52.
And a groove pattern recognition unit 54. The groove position detection unit 52 performs a predetermined process on the input image obtained by imaging with the CCD camera 26, and detects the position and the shape of each groove. For example, specifically, FIG.
As shown in (a), the upper left coordinates (x ₁ , y ₁ ) of the circumscribed rectangle at the groove, and the lower right coordinates (x ₂ , y) of the circumscribed rectangle
₂ ) Measure the area of the groove, the coordinates of the center of gravity of the groove, and the main axis of the groove. Here, the x-axis is a horizontal direction, and the y-axis is a direction perpendicular to the x-axis. Further, as shown in FIG. 7B, from the feature amounts of these groove portions, the points P _S and P _E, which are the intersections of the boundary line of the circumscribed rectangle and the straight line of the main axis (here,
_{Let the} point on the upper side of the image be the point PS. A), P _S point and P
And P _M points is a midpoint between the point _E, the angle θ of the groove portion calculates the (spindle angle formed by the y-axis).

In the case of repair welding of a combiner, welding conditions for welding the groove differ depending on the position and shape of the groove, the presence or absence of a penetrating portion in the groove, and the like. For this reason, in the present embodiment, a plurality of groove patterns are prepared in advance for grooves having the same welding conditions for groove portions that occur at a considerable frequency. The groove pattern recognition unit 54 recognizes which groove pattern each groove corresponds to.

FIG. 8 is a diagram for explaining a groove pattern. In this embodiment, nine groove patterns are used. The groove patterns 1 and 2 have a groove portion extending from the upper part of the hole 6 and have a groove portion of ± 45 ° with respect to the y-axis.
In the angle range of The groove pattern 1 is a case of a circular hole, and the groove pattern 2 is a case of an oval hole.
The groove patterns 3 and 4 have a groove portion extending from the lower part of the hole and not reaching the lower edge of the panel, and the groove portion has an angle range of ± 45 ° with respect to the y-axis. It is in. The groove pattern 3 is a case of a circular hole, and the groove pattern 4 is a case of an oval hole. Groove patterns 5, 6
The groove portion extends from the lower portion of the hole and reaches the lower edge of the panel, and the groove portion has an angle range of ± 5 ° with respect to the y-axis. Groove pattern 5
Is a circular hole, and the groove pattern 6 is an elliptical hole. In the groove patterns 7, 8, and 9, the groove portions extend from the lower edge of the panel and do not reach the holes, and the groove portions are ± 5 ° with respect to the y-axis. It is in the angle range. Groove pattern 7, 8, 9 divided by the distance between P _S point and P _E point. Ie, groove pattern 7, as the distance between P _S point and P _E point is 5 to 10 mm, groove pattern 8, the distance between P _S point and P _E point is 10～13Mm, The groove pattern 9 is P _S
The distance between the point and the P _E point is one that is 13~25mm. Since the penetrating portion usually occurs at a relatively long groove, the presence or absence of the penetrating portion can be determined by the length of the groove in the above-described classification of the groove pattern. That is, among these groove patterns, in the case of groove patterns 5 and 6 and groove pattern 9, there is a penetrating portion at the groove portion. In the case of other groove patterns, there is no penetrating portion.

In addition, when the gap between the groove portions is short, or when the gap between the upper end portion and the hole of the groove portion classified in the groove pattern 9 is short, the groove portion may not be welded. Therefore, the groove pattern recognition unit 54 confirms that these intervals are each equal to or larger than a predetermined value, and finally classifies each groove portion into one of nine groove patterns. When it is determined that the groove portion does not correspond to any of the nine groove patterns described above, a human will perform welding manually.

The storage unit 70 stores welding conditions corresponding to each groove pattern. The central control unit 60 controls each unit. Specifically, based on height data at each measurement point sent from the distance sensor 24, the height of each groove portion of the panel is specified by interpolating between each measurement point. The height data of the panel obtained in this manner becomes position information (height information) in the height direction (z direction) to be taught to the welding torch 22 when welding is performed. Further, a groove position detection unit 52
From the information on the position and shape of the groove portion sent from the welding position information (welding start point, welding end point, and the like) for moving the welding torch 22 on the welding plane (xy plane) as the welding portion. The welding conditions (current, voltage, speed, etc.) for each groove are determined from the groove pattern information sent from the groove pattern recognition unit 54. Then, information on the height information, welding position information, and welding conditions is output to the robot controller 40. The robot controller 40 controls the operation of the welding robot 10 in accordance with a command from the central control unit 60. The CRT display device 80 displays an image processing result as needed.

Next, the operation of the automatic welding system to which the automatic welding method of this embodiment is applied will be described. FIG. 9 is a diagram for explaining the overall operation of the automatic welding system. First, as shown in FIG. 9, the combiner 2 is attached to the positioner 32 (step 2), and then the positioner 32 is controlled to tilt the combiner 2 to a position where welding can be performed (step 4). Next, the combiner 2 is rotated by 2 ° with respect to one panel 4, and the distance sensor 2 is rotated.
4 to measure the height up to the panel 4 (step
6). The height data obtained by the distance sensor 24 is output to the central control unit 60, and the central control unit 60 calculates the height of each groove of the panel 4 based on the height data.
Next, the positioner 32 is controlled to rotate the combiner 2 by about 6 °, and 60 images of the panel 4 are captured by the CCD camera 26. Then, the image processing device 50 performs predetermined image processing on the input image to detect the position of the groove portion and specify the groove pattern (step 8).

Next, the operation of the groove position detector 52 of the image processing device 50 will be described. FIG. 10 is a diagram for explaining the operation of the groove position detection unit 52. First, the original image is captured and sent to the groove position detecting section 52 by the CCD camera 26 at, for example, 256 gradations (step 22). FIG. 11A shows an example of this original image.
In this example, there are three groove portions, which are the case of the groove patterns 1, 5, and 9. In the groove position detection unit 52,
First, a projection histogram is calculated (step 24). From this, if there is a hole, the position of the hole and the positions of the two edges of the panel are detected. If there is no hole, the positions of the two edges of the panel are detected. (Step26). The information about the positions of the holes and the positions of the edges detected here is also output to the groove pattern recognition unit 54.

When the position of the edge of the panel or the position of the hole is detected, the positions are corrected, and then window processing is performed (step 28). The window processing is a mask processing for applying a window to a hole portion and another panel portion in order to separate a groove portion from a hole or an edge. As a result, for example, in the case of the original image shown in FIG. 11A, a shaded window is displayed as shown in FIG. 11B. Then, as shown in FIG. 3C, only three groove portions are recognized as having no holes and no edges.

The reason why the position is corrected before the window processing is that each panel expands and contracts due to the influence of heat at the time of past welding. Therefore, for example, the position of the hole is shifted or the position of the hole is changed. This is because the shape is deformed. In the window processing, the correction amount of the position is measured, and the input image is moved so that the input image fits in a predetermined window. This has the advantage that the same window can always be used.
The size of the window is set slightly larger in consideration of the deformation of the hole. However, since the holes tend to shift along the circumferential direction of the panel, for example, for a circular hole, an elliptical window slightly bulging in the circumferential direction of the panel is used. Since the groove is often formed in the vertical direction, the portion of the groove that is cut off by the window can be reduced as much as possible. Further, as shown in FIG. 11 (b), the window is set to be slightly larger especially at the lower edge portion. This is because the groove is mainly generated at the lower edge of the panel, and the position of the lower edge of the panel may fluctuate greatly due to the influence of the groove.

Next, the image from which only the groove portion has been extracted by the window processing is binarized (step 32). FIG. 12 is a diagram for explaining this binarization processing. FIG. 12 shows the groove shown in FIG. 4 as an example.

FIG. 12A shows the original image obtained by photographing the groove 110 shown in FIG. 4 in, for example, 256 gradations. In this case, the penetrating portion 11 of the groove 110
The image No. 4 has the lowest gradation and appears almost black. This is because there is no reflection of light from the through portion 114. On the other hand, since the illumination light is reflected from the flat portion 116 at the bottom of the groove 110, the obtained image has the highest gradation and looks almost white. The ground portion of the panel 4 shown in satin in FIG. 12A has a state different from the flat portion 116 cut off by the grinder, and is substantially an intermediate gradation of 256 gradations, that is, gray. When the groove 110 is formed by a grinder, an inclined portion 112 having a shape corresponding to the shape of the tip of the grinder is formed at the left end of the groove 110 as shown in FIG. 4B. . In the inclined portion 112, the manner of reflecting light changes depending on the angle of inclination. When this portion is viewed from above, as shown in FIG. 12 (a), the leftmost side is black, and gradually goes to the right. It is connected to the flat portion 116 while being bright.

In the binarization process for specifying the position and the shape of the groove 110, an ordinary method is used, that is, one threshold is set, and an image having a gradation higher than this threshold is white or If a method is used in which a black image having a gradation lower than the threshold is set to the opposite color, an image of the ground portion is formed by the groove 11.
0 is displayed in the same color as either the flat portion 116 or the penetrating portion 114. In addition, in the inclined portion 112, the boundary between the portion appearing black and the portion appearing white fluctuates left and right depending on how to set a threshold value. Therefore, in the method of obtaining a binary image by setting one threshold value as described above, the shape of the groove 110 becomes unclear, and the position of the groove 110 on the panel 4 and its shape are recognized. Can not do it.

Therefore, as shown in FIG. 13, the original image is binarized for each pixel using two threshold values. FIG.
In the graph, the horizontal axis represents the gradation of the input original image, and the vertical axis represents the output, and the input pixels are displayed in black or white depending on the gradation. Here, the first threshold is set to 10
0 and the second threshold value is 165. That is, in the input original image, the pixels of the part where the gradation is smaller than 100 and the part where the gradation is larger than 165 are displayed as white,
Pixels having a gray scale of 100 or more and 165 or less are displayed as black. The two thresholds 100 and 165 are such that the gray of the ground portion of panel 4 is included between the two thresholds, and the white of the flat portion 116 is smaller than the second threshold. It is selected so that it becomes larger and the black of the penetrating portion 114 becomes smaller than the first threshold value. It is needless to say that when binarizing, the white part and the black part can be displayed oppositely.

FIG. 12B is a binary image obtained by binarizing the original image of FIG. 12A with the characteristics shown in FIG. As shown in FIG. 12B, the penetrating part 114, the flat part 116, and the inclined part 11 of the groove 110 in FIG.
The left end of 2 is displayed in white. On the other hand, the ground portion of the panel 4 is displayed in black as indicated by hatching in FIG. Thereby, when the binarization, the groove portion 1
10 shapes can be clearly recognized. In order to obtain an image as shown in FIG.
In addition to obtaining a binary image in which the above-mentioned gradation is black, a binary image in which the gradation of 165 or less is black and a gradation higher than 165 is white is obtained, and both images are added later. The same result can be obtained even if the method of However, when such a method is adopted, it takes longer to perform the processing than when binarization is performed with the characteristics shown in FIG.

Regarding the inclined portion 114 of the groove portion 110,
Since the gradation gradually changes, the gradation is 100 or more and 16 or more.
Portions in the range of 5 or less are displayed in black when binarized. Therefore, such a portion of the inclined portion 114 remains in the image as a stripe-shaped portion 118 shown in FIG. However, since the shape of the grinder used to form the groove 110 is known in advance, as long as a predetermined grinder is used, the width of the stripe-shaped portion 118 in FIG. 12B is substantially constant. . For this reason, the known expansion
By performing the contraction process, the streak-like portion can be removed from the image. As a result of performing such a process, finally, as shown in FIG. 12C, a binary image in which only the groove portion having no stripe portion is displayed is obtained.

The groove 110 as shown in FIG.
When an image in which only the image is displayed is obtained, then each groove 11
Labeling is performed to distinguish 0 (step 34). After that, for each of the labeled groove portions 110, characteristic quantities such as a circumscribed rectangle of the groove portion, an area of the groove portion, a barycentric coordinate of the groove portion, and a main axis of the groove portion are measured. Further, from the feature quantity of the groove portion, calculates P _S point, P _M points, P _E point, the angle θ of the groove portion (STEP 36). The groove data thus obtained is output to the groove pattern recognition unit 54 and the central control unit 60.

Next, the operation of the groove pattern recognition section 54 will be described with reference to FIGS. 14 to 17 are diagrams for explaining the operation of the groove pattern recognition unit 54. First, when groove data for a groove portion is input from the groove position detecting section 52 (step 42), it is determined whether or not a hole exists (step 44). If it is determined that there is no hole, this groove portion is formed by groove pattern 7,
There is a possibility that it corresponds to either of 8 and 9. Then, check if the groove reaches the lower edge of the panel (step 4
6) It is determined whether the angle θ of the groove is within an angle range of ± 5 ° with respect to the y-axis (step 48). No either one
When it is determined that the groove portion does not correspond to the nine groove patterns, that is, it is determined to be NG, data on the groove portion is not transferred. On the other hand, step46, step48
In both when it is determined that Yes, calculates the length of the groove portion, i.e., the distance between P _S point and P _E point (step5
2) Based on the distance, the groove portion is grooved pattern 7,
It is classified into one of 8 and 9 (step 54).

If it is determined in step 44 that a hole exists, it is determined whether the groove is above the center of the hole (step 56). When it is determined that the groove portion is above the groove portion, the groove portion may correspond to groove pattern 1 or 2. In this case, first, it is confirmed that the groove is not in the hole (step 58). Since the window is hung, the groove does not actually go into the hole, but, for example, if the window is set to be shifted, the groove and a part of the hole are integrated. Therefore, the integrated part may be recognized as a groove, which may cause a malfunction. Therefore, in step 58, an error when the window is set is checked. Next, it is determined whether or not the angle θ of the groove portion is within an angle range of ± 45 ° with respect to the y-axis (step 62). If the angle θ of the groove is not in this angle range, it is determined as NG. If the angle θ is in this angle range, the groove is classified into groove patterns 1 and 2 according to the shape of the hole (step 64). .

If it is determined in step 56 that the groove is not above the center of the hole, the groove is formed in groove pattern 3,
4, 5, 6, 7, 8, and 9. Also in this case, first, it is determined whether after the groove portions and it was confirmed that not in the hole (step66), P _S point based on information about the position of the hole is in contact with the hole (step68) . When P _S point is determined to be in contact with the bore,
The groove may correspond to any of the groove patterns 3, 4, 5, and 6. Then, it is determined whether or not the groove has reached the edge of the lower end of the panel (step 72). If the groove does not reach the edge, the angle θ of the groove
Is determined to be within an angle range of ± 45 ° with respect to the y-axis (step 74). If the angle θ of the groove is not within the angle range, it is determined to be NG. The groove portion is classified into groove patterns 3 or 4 according to the shape of the hole (step 76). In step 72, when the groove reaches the edge of the lower end of the panel, the angle θ of the groove
Is in the range of ± 5 ° (step 7).
8) If the angle θ of the groove is not within the angle range, NG
On the other hand, if it is within the angle range, the groove portion is classified into groove patterns 5 or 6 according to the shape of the hole (step 8).
2).

Further, in Step68, when P _S point is determined not in contact with the hole, groove portion groove pattern 7,
There is a possibility that it corresponds to either of 8 and 9. Then, check whether the groove reaches the lower edge of the panel (step 8
4) It is determined whether or not the angle θ of the groove is within an angle range of ± 5 ° with respect to the y-axis (step 86). No either one
Is determined to be NG, on the other hand, if both are determined to be Yes in step 84 and step 86, the length of the groove portion,
That is, the distance between the points P _S and P _E is calculated (step 88),
On the basis of the distance, the groove portion is formed by groove patterns 7, 8, 9
(Step 92).

Next, for each groove portion classified into a predetermined groove pattern based on the shape and position of each groove portion as described above, the mutual positional relationship and the positional relationship with the hole are examined. First, for each groove portion classified into each groove pattern in step 54, step 64, step 76, step 82, and step 92, between each P _S point and P _S point, P _S point and P _S
Between the point _M, between P _S point and P _E point between the P _M point and the P _M point between the P _M point and P _E point between P _E point and the P _E point Calculate the distance and find the shortest distance (step9
Four). In particular, in Step92, for welding groove classified into groove pattern 9 obtains its P _S point, P _M points, even the shortest distance between P _E point and the hole (step96). Then, it is checked whether or not those shortest distances are equal to or more than a predetermined value. If the shortest distance is smaller than that value, welding is difficult, so the groove portion is NG, while the shortest distance is If the value is equal to or more than that value, the classification of the groove pattern for the groove portion is determined as it is (step 98).

Next, the central controller 60 controls the distance sensor 24
To the robot controller 40, height information calculated based on the height data sent from the apparatus, welding position information calculated based on various data output from the image processing device 50, and information on welding conditions. I do. Then, the robot controller 40 controls the welding robot 10 based on the information to start welding (step 12).

In this way, when the welding is completed for one panel, it is determined whether or not to end the welding (step 14).
If there is a next panel to be welded, proceed to step 6 and perform the same processing. When the welding is completed for all the panels, the positioner 32 is returned to the original position (step 16), the combasterer 2 is removed, and the processing is completed.

In the automatic welding method according to this embodiment, a plurality of groove patterns classified in advance with respect to the groove portion are prepared.
By recognizing which of the groove patterns the groove portion corresponds to based on an image obtained by imaging the combiner, welding conditions for the groove portion are determined based on the recognized groove pattern. Can be easily determined.
In addition, by operating the welding robot based on the height information up to the groove portion and the welding position information and the recognized groove pattern, the groove portion can be automatically and reliably welded, and the predetermined With regard to the groove portion corresponding to the groove pattern, welding can always be performed under constant conditions, and the welding quality can be kept constant.

The present invention is not limited to the above-described embodiment, and various modifications can be made within the scope of the invention. For example, in the above embodiment, the case where a combustor, which is an aircraft engine part, is used as an object to be welded, but a heat exchanger such as a refrigerator, a thin tube portion of a nuclear reactor, a heat exchange portion of a turbine, etc. Therefore, the repair welding can be surely performed by applying the automatic welding method of the present invention even to the one that requires repair of a crack.

[0047]

As described above, according to the first aspect of the present invention, a plurality of groove patterns classified in advance with respect to the groove portion are prepared, and an image obtained by imaging the heat exchanger is obtained. By recognizing which of the groove patterns the groove portion corresponds to, the welding conditions for the groove portion can be easily determined based on the recognized groove pattern, and By operating the welding robot based on the distance information to the groove and the welding position information and the recognized groove pattern, the groove can be automatically and reliably welded, and the predetermined groove pattern can be obtained. With regard to the groove portion corresponding to the above, welding can always be performed under constant conditions, and an automatic welding method capable of maintaining constant welding quality can be provided.

According to the second aspect of the present invention, a step of determining whether or not the interval between the groove portions is equal to or more than a predetermined value is provided. In consideration of the positional relationship with the groove, the welding robot can determine whether or not the groove can be actually welded.

According to the third aspect of the present invention, when obtaining information on the distance to the groove, a plurality of arbitrary positions on the heat exchanger are measured, and the distance between the plurality of positions is interpolated. An automatic welding method that can easily calculate the distance to the tip can be provided.

According to the fourth aspect of the present invention, since the welding robot has the positioner that tilts and rotates the heat exchanger, the welding robot tilts and rotates the heat exchanger.
An automatic welding method capable of moving a groove to a position where welding is easy can be provided.

According to the fifth aspect of the present invention, since the heat exchanger is a combustor which is an aircraft engine part, the defective portion is welded with a constant high quality, and the recurrence of the defective is effectively prevented. Therefore, it is possible to provide an automatic welding method capable of effectively preventing occurrence of an aircraft accident.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram of an automatic welding system to which an automatic welding method according to an embodiment of the present invention is applied.

FIG. 2 is a schematic plan view of a part of a combiner which is an object to be welded.

FIG. 3 is a schematic sectional view of the combiner in the direction of arrows AA.

FIG. 4 is a diagram for explaining a groove portion formed in the combiner.

FIG. 5 is a diagram for explaining an uneven state generated on the surface of the combiner.

FIG. 6 is a diagram for explaining measurement points of the imaging field of view and the height of the combiner.

7 is a diagram for explaining a feature amount of a groove measured by a groove position detection unit of the automatic welding system shown in FIG. 1;

FIG. 8 is a diagram for explaining a groove pattern.

FIG. 9 is a diagram for explaining the operation of the automatic welding system shown in FIG.

FIG. 10 is a diagram for explaining the operation of a groove position detection unit of the automatic welding system.

FIG. 11 is a diagram for explaining window processing in the groove position detection unit.

FIG. 12 is a diagram for explaining a binarization process in the groove position detection unit.

FIG. 13 is a diagram for explaining how to set a threshold value when performing the binarization processing.

FIG. 14 is a diagram for explaining an operation of a groove pattern recognition unit of the automatic welding system shown in FIG.

15 is a diagram for explaining the operation of a groove pattern recognition unit of the automatic welding system shown in FIG.

FIG. 16 is a diagram for explaining an operation of a groove pattern recognition unit of the automatic welding system shown in FIG.

FIG. 17 is a view for explaining the operation of the groove pattern recognition unit of the automatic welding system shown in FIG. 1;

[Explanation of symbols]

 2 Convaster 4 Panel 6 Hole 8 Crack 10 Welding Robot 12 Robot Body 22 Welding Torch 24 Distance Sensor 26 CCD Camera 28 Illumination Device 32 Positioner 40 Robot Controller 50 Image Processing Device 52 Groove Position Detector 54 Groove Pattern Recognition Unit 60 Center Control section 70 Storage section 80 CRT display device 110 Groove section 112 Slope section 114 Penetration section 116 Flat section 118 Stripe section

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁷ Identification code FI B23K 9/127 508 B23K 9/127 508B 37/047 501 37/047 501D F02C 7/00 F02C 7/00 D F23R 3/42 F23R JP-A-5-209502 (JP, A) JP-A-2-104476 (JP, A) JP-A-4-22579 (JP, A) JP-A-5-340881 (JP) , A) (58) Fields investigated (Int. Cl. ⁷ , DB name) B23K 9/12 B23K 9/04 B23K 9/127 B23K 37/047 F02C 7/00 F23R 3/42

Claims (6)

(57) [Claims] 1. An automatic welding method for forming a groove on a defective portion generated in a heat exchanger and performing repair welding using a welding robot, wherein the groove is formed by a welding torch attached to the welding robot. Obtaining information on the distance to the groove, applying predetermined processing to an image obtained by imaging the heat exchanger, obtaining image information on the groove, and an image of the groove. A step of obtaining welding position information that is information for moving the welding torch to a welding portion based on information; and a plurality of groove patterns in which the groove portion is pre-classified based on image information of the groove portion. Recognizing which of the following, and operating the welding robot based on the distance information to the groove portion, the welding position information and the recognized groove pattern, welding the groove portion And a step of performing And an automatic welding method. 2. The automatic welding method according to claim 1, further comprising the step of judging whether or not an interval between the groove portions is equal to or greater than a predetermined value. 3. When obtaining distance information to the groove,
The automatic measuring apparatus according to claim 1, wherein a plurality of arbitrary positions on the heat exchanger are measured, and a distance to the groove portion is calculated by interpolating between the plurality of positions. Welding method. 4. The automatic welding method according to claim 1, wherein the welding robot has a positioner for tilting and rotating the heat exchanger. 5. The air conditioner according to claim 1, wherein the heat exchanger is a combustor that is an aircraft engine part.
The automatic welding method according to 2, 3, or 4. 6. A groove portion is formed at a defective portion generated in the heat exchanger.
Automatic welding equipment for forming and repair welding using a welding robot
A location, the groove from the welding torch to be attached to the welding robot
Means for obtaining information on the distance to the heat exchanger, and applying predetermined processing to an image obtained by imaging the heat exchanger.
By Succoth hand to obtain image information about the open tip portion
A step and the welding toe at the welding portion based on the image information of the groove.
For obtaining welding position information, which is information for moving the switch
And the groove is classified in advance based on the image information of the groove.
Which groove pattern corresponds
Means for determining the distance to the groove , the welding position information,
The welding robot based on the recognized groove pattern
Means for operating and welding the groove .