JPH01113175A - Welding abnormality detection method - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field 1] The present invention relates to a welding abnormality detection method for detecting welding abnormalities during automatic welding, particularly welding abnormalities such as "burn-through."

[Conventional Technique #i] An abnormal phenomenon that occurs during welding is a phenomenon called "burn-through" in which the welded part melts to the back surface and the molten metal flows out and scatters. Once this "burn-through" phenomenon occurs, it progresses continuously, causing serious damage to the quality of the welded part. Therefore, when this welding abnormality called "burn-through" occurs, it is necessary to detect it early and take measures such as stopping the welding operation.

The conventional method for detecting this type of welding abnormality is as follows:
There are two methods: (1) detecting from the relationship between welding current and arc voltage as shown in FIG. 8; (2) detecting from fluctuations in arc sound as shown in FIG. 9. In addition, as a method similar to these, as disclosed in Japanese Patent Application Laid-Open No. 59-70475, a light intensity detector is installed so as to be in direct contact with the back surface of the welded part, and the arc light penetrating the welded part or the melted There is an optical method in which welding lb11m is performed by detecting radiant light from a pond.

[Problems to be Solved by the Invention] Method (2) shown in Figure 8, that is, a method of detecting from the relationship between welding current and arc voltage, is Welding 1! The current is detected by a current detector 31, the arc pressure is detected by a voltage detector 32,
This method uses an analyzer 33 to analyze and determine whether or not these are within the appropriate range to detect an abnormality, but in this method, if a momentary fluctuation occurs even during proper welding, this is detected as an abnormality. Therefore, there is a drawback that welding abnormalities cannot be accurately determined. In addition, 1 in Fig. 8 is a welding torch, 2
is a wire, 3 is a wire feeding roller, 4 is a flux, 5
3 is a base material (rA tube), 30 is a welding power source, 34 is an abnormality alarm device, and 35 is a display device.

In the method (■) shown in FIG. 9, that is, the method of detecting from the sound pressure level fluctuation of the welding arc sound, the sound pressure level or frequency detected by the detector 41° signal detection circuit 42 is analyzed by the analyzer 43, and an abnormality is detected. Although this method can be applied to monitoring the seal condition of gas seal welding, it has the disadvantage that it is difficult to distinguish it from ambient noise when the arc noise is small, such as submerged arc welding. be. In FIG. 9, 1 to 3 and 5 are the same as those in FIG. 8, 44 is an abnormality alarm, and 45 is a display device.

In addition, the above-mentioned method (2), which is the method of placing a photodetector on the back side of the welding part, distinguishes between molten metal light or penetrating arc light caused by burn-through and radiation light caused by burnt base metal during normal welding to determine abnormal conditions. It is not something that is done. Moreover, since the detector is in direct contact with the back surface of the welded part, the heat resistance of the detector becomes a problem when burn-through occurs.

As described above, the conventional welding abnormality detection methods have problems in industrial application, such as the applicable range, detection accuracy, and durability of the detector.

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide a welding abnormality detection method that can reliably detect welding abnormalities, particularly the "burn-through" phenomenon, and can stabilize quality.

[Means for Solving the Problems] In order to solve the above problems and achieve the objects, the present invention takes the following measures. That is, only light in a specific wavelength range is selectively received from the light emitted from the back side of the part to be welded (molten metal light, welding arc light, etc.), and is captured two-dimensionally into an image sensor continuously over time. The captured two-dimensional image is binarized and the areas where the brightness is higher than the preset standard brightness are determined as bright areas.The area of this bright area is compared with the preset standard area, and the area of this bright area is determined. When a state in which M is larger than a preset M quasi-area continues for a longer time than a preset reference time, it is determined that a welding abnormality called "burn-through" has occurred.

[Effects] By taking such measures, the following effects are achieved. For example, the welding side and the side opposite to this welding side (back side)
A monitoring device (filter camera, etc.) installed at an appropriate location and moving in synchronization with the welding selectively collects (short wavelength side only) the radiant energy (light) generated from the welding part during abnormal welding, and detects the signal. Since welding abnormalities are detected by processing, it is possible to reliably detect welding abnormalities, especially the burn-through phenomenon of weld metal, and it is possible to stabilize the quality of welding.

[Actual Droplet Example] FIG. 1 is a diagram showing an example in which the method of the present invention is applied to automatic welding of the inner surface of a longitudinal joint of steel pipes. In the figure, 1 is a welding torch and 2 is a wire. 3 is the above wire 2
4 is a wire feeding roller that feeds 7 lux for smooth welding. 5 is a base material to be welded, 6 is a backing material (fA plate), and 7 is a support for the backing material 6. 8 is a support arm for the welding torch 1, etc., and 9 is a welding cable, which is capable of automatically welding inside a long steel pipe. 10 is a support for the base material (II pipe). There is an ND filter in the appropriate place on the back side of this welding part.
A monitoring device 11 consisting of a combination of an interference filter, an i! optical filter, a small COD camera, etc. is installed to monitor the welding condition. The image processing device 14 is a short-wavelength projector (including lenses) that emits light. The system control ill device 15 is the floodlight 1.
20 Image processing device 14. An activation signal is given to the monitoring device 11, etc., an indicator light 16 and a buzzer 17 are operated 0NWh based on the processing data of the image processing device 14, and the welding power source 18. It controls the traveling device 19 (including peripheral equipment such as turning rollers and positioners).

Figure 2 is an enlarged view of the welded part in Figure 1.
) is a diagram showing a case in which welding is normal, and FIG. 3(b) is a diagram showing a case in which welding is abnormal.

As shown in FIG. 2(a), under normal conditions, the welding arc 20 properly forms the welding part gA21. However, as shown in FIG. 2(b), under abnormal conditions, the base metal 5
Welding abnormality (burn-through) occurs because the groove (groove) is not in the proper shape or the welding conditions are not appropriate, and as a result, welding arc 20 and weld metal 21 are transferred to the backing material 6 from the welding part. Formed, radiant energy (light) 2
2 is leaking from the gap between the base material 5 and the backing material 6.

FIG. 3 shows a spectrum of blackbody radiation obtained using a blank radiant in order to understand the nature of the light emitted from the welded part during the above-mentioned abnormality. The vertical axis shows relative intensity, the horizontal axis shows wavelength, and 100
Spectra at 0°C, 1540°C, and 2750°C are analytically illustrated.

Since iron-based metals melt at around 1500°C, it is only necessary to be able to receive light at a temperature higher than this temperature. Based on these facts, the method of the present invention can be used within the sensitivity wavelength range of the camera and 15
A combination of an ND filter, an interference filter, and a polarizing filter is used to selectively receive wavelengths higher than 40°C.

FIG. 4 is a diagram that does not show the operation of the monitoring device 11.

In the figure, the upper column shows normal welding, and the lower column shows abnormal welding. This shows an application situation for thin plates that can lead to erroneous judgments, especially when there is no regular filter in the welding situation. In other words, even in normal conditions, the arc heat causes the surrounding area of a thin plate to be in a burnt state at a fairly high temperature, causing light to be radiated, so we purposely show an example of its application to a thin plate. Therefore, if the welded part is imaged without a filter or with the use of a normal filter, burnt parts due to arc heat will be captured even during normal welding, and it will be difficult to distinguish from the abnormal state based on the image brightness distribution.

On the other hand, when this device is equipped with an appropriate filter, for example, a combination of an ND filter, an interference filter, and a polarizing filter, radiation from molten metal (including welding arc light) of <1540°C or higher, as explained in Fig. 3, can be used. Since light is selectively received, burnt during normal conditions is not input, and bright areas appear on the screen only when welding is abnormal. Therefore, welding abnormalities can be easily determined.

FIG. 5 is a diagram showing a welding abnormality detection flowchart according to the method of this embodiment, and the procedure of the abnormality detection method will be explained with reference to this. First, when welding is started in step A, an image is input by the monitoring device e111. Even if welding is not performed, if you start the visual field confirmation process,
You can input images.

Next, in step B, the input image is processed into binary values of white and black using a preset threshold. Furthermore, in step C, it is determined whether or not a spot light for confirming the field of view has been projected, and if a spot light has been projected (
If YES), the center of gravity of the bright image (spot light) is detected in step D, and it is determined in step E whether the center of gravity is within a predetermined range within the screen. If the center of gravity position is not within the predetermined range (No), an alarm indicating that the setting condition of the monitoring device @11 is defective is outputted in step F), and the position is corrected to the normal position.

If the center of gravity is within the specified range of the screen (YES
), it is determined that the field of view is OK and the process proceeds to the next abnormality detection process.

Note that if the spotlight is not being projected (No), the visual field confirmation process is skipped and the process proceeds to the next abnormality detection process (the visual field confirmation process can be intentionally omitted).

In the abnormality detection process, the number of bright pixels (Q) is measured from the binarized image (step G), and the number of bright pixels (Q) is calculated in step H).
) is compared with a preset threshold value S, and if QC8 (No), the process returns to step A, that is, the initial image input. If Q≧S, the number of screens (P) for which Q≧S is measured in step J, and in step K, P is compared with a preset number of screens (R) to evaluate the continuity of the abnormal state. If P<R (No), return to the initial image input, and if P≧R (YES), it means that a welding abnormality has occurred, so step and output an abnormality alarm. At the same time, a welding stop process is performed in step M. Thereafter, in step N, the operator corrects the defect and performs processing such as rewelding.

Through the above-described flow, welding abnormalities during automatic welding can be detected at an early stage, and damage to the base material, jigs, tools, and equipment can be prevented to a minimum. As a result, there is no need for a worker to monitor abnormalities during automatic welding, contributing to cost reduction and stabilizing quality.

FIG. 6 is a diagram showing a modification of the shape of the backing material 6. The backing material 6' shown on the left side of FIG. 6 has a large number of parallel cut grooves S on the surface that is in contact with the base material 5. According to the backing material 6' having such a shape, light leakage can be detected more reliably in the event of a welding abnormality, compared to a backing material 6' which has a substantially flat shape like the backing material 6 shown on the right side of Fig. 6. It becomes possible.

Fig. 7(a>(b) is a diagram showing an example in which the method of the present invention is applied to flat joint welding (gas shield). Fig. 7(a)
) is a diagram showing the case where the backing material 6 is present, and (b) is a diagram showing the case where the backing material 6 is not present. In either case, welding abnormalities can be detected by installing a monitoring device 11 at an appropriate location behind the weld.

Note that the present invention is not limited to the embodiments described above, and it goes without saying that various modifications can be made without departing from the gist of the present invention.

[Effects of the Invention] According to the present invention, only light in a specific wavelength range is selectively received from among the light emitted from the back surface of the part to be welded (molten metal light, welding arc light, etc.), and the light is continuously and temporally received. The image sensor captures the captured two-dimensional image two-dimensionally, binarizes the captured two-dimensional image, processes the captured two-dimensional image 118, determines the area where the brightness is higher than a preset standard brightness as a bright area, and uses the area of this bright area as a preset standard. If the area of this bright area is larger than the preset reference area and continues for a longer period of time than the preset reference time, it is determined that a welding abnormality called "burn-through" has occurred. As a result, the radiant energy (light) emitted from the welding part during welding abnormalities is selectively collected and signal processed, making it possible to reliably detect welding abnormalities, especially the "burn-through" phenomenon, and stabilize quality. A welding abnormality detection method can be provided.

[Brief explanation of the drawing]

Figures 1 to 5 are diagrams showing one embodiment of the method of the present invention. Figure 1 is a diagram showing an example in which the method of the present invention is applied to longitudinal joint welding of steel pipes, and Figures 2 (a) and (b) are Figure 1 is an enlarged view of the welded part, Figure 3 is a diagram showing the blackbody radiation spectrum, Figure 4 is a diagram to explain the operation of the monitoring device, and Figure 5 is a diagram to explain welding abnormality detection. FIG. 6th
The figure shows a modification of the backing material, and FIGS. 7(a) and 7(b) show an example in which the method of the present invention is applied to flat plate joint welding. Figures 8 and 9 are diagrams showing the conventional welding method.
9 is a diagram showing an example of welding monitoring using welding current and arc voltage, and FIG. 9 is a diagram showing an example of shield monitoring using arc sound. DESCRIPTION OF SYMBOLS 1... Welding torch, 2... Wire, 3... Wire feed roller, 4... Flux, 5... Base material, 6...
... Backing material, 1... Supporting stand for backing material, 8... Support arm, 9... Welding cable, 10... Support stand, 11
... Monitoring device, 12... Floodlight, 13... Spot, 14... Image processing device, 15... System υ,
I III device, 16... indicator light, 17... buzzer, 18... welding power source, 19... traveling device, 20...
- Welding arc, 21... Welding metal, 22... Synchrotron radiation, 31... Current detector, 32... Voltage detector, 33
...Analyzer, 41...Detector, 43...Analyzer,
44... Abnormal) alarm device, 45... Display device. Applicant's representative Patent attorney Takehiko Suzue Figure 1 Figure 2 Kakucho (nm) - Figure 3 Figure 5 Figure 6 (a) Figure 7 Figure 8

Claims (1)

[Claims] In various welding methods, light in a specific wavelength range is selectively received from the light emitted from the back side of the workpiece (molten metal light, welding arc light, etc.), and the image sensor continuously displays two-dimensional images over time. The captured two-dimensional image is binarized and the areas where the brightness is higher than the preset standard brightness are determined as bright areas.The area of this bright area is compared with the preset standard area. If the area of the part continues to be larger than a preset reference area for a longer period of time than a preset reference time, it is determined that a welding abnormality called "burn-through" has occurred. Welding abnormality detection method.