JPS6117366A - Method and device for automatic welding - Google Patents

Abstract

PURPOSE:To form a welded joint part of high quality automatically in automatic welding of a base metal, by detecting the shape of molten pool formed in the groove by a sensor, comparing and calculating this with a preset value and controlling the welding current and welding speed so as to conform to the set value. CONSTITUTION:When welding base metals 1a and 1b by arc of a welding torch 2, the area S, length l and width d of a molten pool 4 formed at the grooge are detected by a visual sensor 21 provided in an obliquely upper part. The detection data 25 are sent to a picture processing system 24, and after subjecting to time equalizing processing and binary-coding picture processing, a binary coded picture 29 is inputted to a picture characteristic measuring device 30 to read the area, length, width etc. The read characteristic value are inputted to an arithmetic unit 31 and compared with preset these values and calculated. The output signals are inputted to a welding power source 37 or a truck driving motor 39 of the welding torch. Thus, welding current and welding speed are controlled properly and an excellent welded joint part is formed.

Description

[Detailed description of the invention] [Industrial application field] TECHNICAL FIELD The present invention relates to improvements in automatic welding methods and apparatus.

[Conventional technology]

As is well known, in order to automate welding, in addition to groove tracing, technology to control the front and back beat shapes and penetration amount, which directly affect joint quality, is essential. For example, as shown in FIG. 2, the groove interval g.

When welding base metals 1m and 1b with a groove angle θ using the welding torch 2, even if the welding conditions are constant and the groove is accurately traced, there will be a change in the groove condition during welding (as shown in the figure). (rotational deformation, shrinkage deformation), and temperature change of base material 1m, lb (
The amount of penetration into the front and back beat shapes varies greatly depending on heat reflection at the start and end of welding, uniformity of preheating, and good welding conditions cannot be obtained. In addition, 3 in FIG. 2 indicates an arc, and 4 indicates a molten pool.

However, until now, there have been almost no sensors or control technology for detecting these parameters that are directly connected to welding quality, and only a few have been developed in Japanese Patent Applications No. 56-i02375 and No. 52-8.
No. 5041, Japanese Patent Application No. 52-85043, etc. have only been filed. In such a known example, as shown in FIG. 3, the temperature from the molten pool is detected by a monochromatic pyrometer, compared with a set temperature, and the difference is output to control the welding heat input.

In FIG. 3, 5 is a visual sensor having a condensing lens 6, 7 is the emitted light from this sensor 5, and 8 to 11 are Si photocells, amplifiers, and Si photocells connected in order to the sensor 5 via an optical fiber 12. 4 shows a temperature converter and a comparison calculation circuit, 12 a temperature setting circuit, and 15 a welding power source.

[Problem that the invention seeks to solve]

However, according to the prior art, there is a constant point as shown below.

(1) Since the temperature is measured and controlled from one point in the molten pool 4 or the continuous bead, the probability of detection errors is high and it is impractical. Specifically, if a small amount of slag, oxide film, etc. is present at the temperature measurement point, the emissivity will change significantly, leading to incorrect measurements and control.

(2) Since the temperature is measured, the influence of the arc 3 is significant. Specifically, when the arc flame spreads due to magnetic blowing or surface conditions, the arc flame overlaps the temperature measuring section. At this time, since the temperature of the arc flame is several thousand degrees Celsius, the measurement results will be fatally incorrect for the molten pool temperature (1500 to 2000 degrees Celsius).

(3) The influence of errors due to measurement positions is large, and it is impractical. Specifically, the measurement results tend to vary depending on the mounting position and angle of the sensor 5 with respect to the welding torch 2, and correction thereof is difficult.

The present invention has been made in view of the above circumstances, and an object of the present invention is to provide an automatic welding method and an apparatus therefor that can reduce costs by automating the welding process and stabilize the quality of welded joints.

[Means for solving problems]

The first invention of the present application detects at least the width of the molten pool formed directly under the welding torch, then compares and calculates this detected data with a preset value, and adjusts the welding current to the set value. By controlling the welding speed and performing welding, we aim to reduce costs and stabilize the quality of welded joints.

A second invention of the present application includes a welding torch, a visual sensor system that detects at least the width of the molten pool by viewing the molten pool formed directly below the welding torch from diagonally above, and detecting data from the visual sensor system. An image processing system for storing and processing, an image characteristic measuring device for reading the shape characteristics of the molten pool, a computing system connected to this image characteristic measuring device, and a controlled system that amplifies and executes the output signal from this computing system. By providing the above, it is intended to obtain the same effect as the first invention of the present application.

[Effect]

First, the molten pool is viewed diagonally from above using a visual sensor, and the detected data is sent to the image processing system.

Here, the photographed image obtained by the visual sensor is subjected to time averaging processing and binarized image processing in an image processing system. Then, when the binarized image is output to an image characteristic measuring device, the shape characteristics of the molten pool are read by this measuring device. Further, the read characteristic values are compared and calculated by a calculation system, and welding is performed by controlling the welding current and welding speed.

〔Example〕

An embodiment of the present invention will be described below with reference to FIG.

It should be noted that the same members as those in FIG. 2 are given the same reference numerals and their explanations will be omitted.

Reference numeral 21 in the figure is a visual sensor system composed of a band 9 pass filter (or sharp cut filter) 22 in the near-infrared region and a visual sensor 23 having sensitivity in the same region.

The visual sensor 23 has an area S1 directly under the welding torch 2.
The molten pool 4 having a length t and a width d is provided so as to be viewed diagonally from above. From the visual sensor system 21, ultraviolet to infrared light with strong radiation intensity that adversely affects imaging is removed, and wavelength light (infrared to infrared range) useful for collecting molten pool information is removed.
only can be obtained.

An image processing system 24 that stores and processes detection data from the sensor system 21 is connected to the visual sensor system 2 . The image processing system 24 includes a video RAM R6 that temporally stores several frames of captured images 25 from the visual sensor 23.
, an image processor 27 connected to this video RAM 26 and performing time averaging processing and binarization image processing, and a threshold level that is set in advance at a level suitable for detecting the outline of the molten pool 4. It is composed of a setting device 28. Note that the above-mentioned time averaging process serves to prevent measurement errors caused by floating slag on the surface of the molten pool.

An image characteristic measuring device 30 to which a binarized image 29 is output from the image processor 27 is connected to the image processing system 24 . This measuring device 30 reads the molten pool characteristics such as the area, length, and width of the molten pool 4 from the binarized image 29.

A calculation system 31 is connected to the image characteristic measuring device 30 . This calculation system 3 is composed of a reference data setter 32, a comparator/calculator 33, and a dead area setter 34. The characteristic values read by the image characteristic measuring device 34 are converted into the standard value (area S6s length tOs width do) 35 from the standard data setting device 32 and the dead value (ΔS) from the dead zone setting device 34 by the calculating device 33. , ΔR). The comparison/calculation method at this time is as shown in FIG.

Here, the measurement data (S, t/d) and the reference value (5OX
output amount Δ to the power supply with respect to the difference amount with (to/do)
The welding temperature can be determined from the characteristic diagram of the molten pool area (Sm) and welding current (I) shown in FIG.

The arithmetic system 31 is connected to a controlled system 36 that amplifies and executes the output signal from the arithmetic system 31. This controlled system 36 is composed of a welding power source 37 connected to the base material 1a, a truck drive motor 39 for driving the traveling truck 78, and a transfer unit 40. An output signal is outputted to the welding power source 37 or the truck drive motor 39 according to the results of comparison and calculation by the calculator 33. Here, the output amount ΔE to the truck drive motor 39 is determined from the characteristic diagram of the molten pool shape ratio (tm/dm) and the truck and drive motor output voltage (E) shown in FIG. In addition, the trolley,
There is a proportional relationship between the drive motor output voltage and the welding speed vw (cm). Further, the above-mentioned data is set in advance in the reference data setter 32.

Next, the operation of the automatic welding device having such a structure will be explained. First, the molten pool 4 is viewed diagonally from above using the visual sensor 23 and detected data is sent to the image processing system 24 . Here, the captured image 25 obtained by the visual sensor 23 is subjected to time averaging processing and binarized image processing by the image processing system 24. Then, when the binarized image 29 is output to the image characteristic measuring device 30, the area S of the molten pool 4 is measured by this measuring device 30.
, length t1 width d, and other molten pool shape characteristics are read. Furthermore, the read characteristic values are compared and calculated by the arithmetic meter 31, and an output signal is outputted to the welding power source 37, the truck, or the drive motor 39 according to the result. In this way, the detected data is compared with a preset value, and welding is performed by controlling the welding current and welding speed so that the set value is achieved.

According to the method of the present invention, the base material 1a.

In order to detect the area, length, and width of the molten pool 4 due to changes in the groove condition of the groove 1b, and to compare this detected data with five preset values and perform calculations, etc., the V When performing Uranami welding on grooved base metals, welding conditions can be adaptively controlled to always obtain a good clean surface bead and Uranami condition, and the welding process can be automated to reduce costs. can. This will be explained in detail below.

(1) In the case of insufficient heat input: In the case of insufficient heat input due to a decrease in the groove gap g, etc., the area S of the molten pool 4 decreases, so compare this with the reference value & according to the procedure shown in Fig. 4, Good welding can be ensured by increasing the welding current in proportion to the difference. At this time, if the molten pool shape ratio t/d becomes poor due to a change in heat input, compare it with the standard shape 16/d (1) and change the welding speed in proportion to the difference to obtain a good surface bead. And a Uranami shape is obtained.

(2) When the heat input is overheated; On the other hand, if the base metal temperature rises due to an increase in the groove gap g, heat reflection at the base metal edge, etc., and the heat input becomes relatively excessive, the melt will not melt. Although adverse effects such as dropping occur, by performing the reverse control of the above (1), it is possible to prevent the melt from dropping and obtain a good top and bottom bead shape.

Further, according to the device of the present invention, the welding chie 2 and the molten pool 4
A visual sensor system 21 that views the image from diagonally above, and an image processing system 24 that stores and processes inspection data from this visual sensor system.
and an image characteristic measuring device 30 that reads the shape characteristics of the molten pool 4.
, a calculation system 31, and a controlled system 36. As mentioned above, it is possible to always obtain good cleanliness, surface bead, and back wave conditions, and also to automate the welding process and reduce costs. It is possible to reduce the

In addition, if the welding conditions are not adaptively controlled, defects such as poor fusion may occur due to insufficient heat input, and conversely, melt drop may occur due to overheating input, and good welding results cannot be obtained.

In the above embodiment, the area, length, and width of the molten pool are detected as the molten pool shape, but the present invention is not limited to this, and at least the width may be detected. That is, since the inventor's experiments have confirmed that the shape of the molten pool and the shape of the Uranami bead have a close relationship with the width, in order to simplify implementation, only the width can be used as the measurement target. However, if it takes
The visual sensor can be simplified using a one-dimensional solid-state image sensor.

Furthermore, in the above embodiments, the case of welding involving an arc has been described, but the present invention is not limited to this.

For example, even in fusion welding that does not involve an arc, such as laser or electron beam welding, similar effects can be expected if the molten pool is measured and controlled using the present device and method.

〔Effect of the invention〕

As detailed above, according to the present invention, automatic welding can reduce costs by automating the welding process, stabilize the quality of welded joints, and is applicable to general fusion welding in which the molten pool is not covered with unmolten black. A method and apparatus thereof can be provided.

[Brief explanation of drawings]

Fig. 1 is an explanatory diagram of an automatic welding device according to an embodiment of the present invention, Fig. 2 is a perspective view for explaining welding between base materials when the groove condition changes, and Fig. 3 is a molten pool. Fig. 4 is a block diagram to explain the comparison/calculation method of the comparison/calculation unit of the device of the present invention, and Fig. 5 shows the welding pool area and FIG. 6 is a characteristic diagram showing the relationship between welding current and FIG. 6 is a characteristic diagram showing the relationship between molten pool shape ratio and truck drive motor output voltage. Its, lb... Base metal, 2... Welding torch, 3...
... Arc, 4... Molten pool, 21... Visual sensor system, 22... Band/matrix filter (or sharp cut filter), 23... Visual sensor, 24... Image processing system, 25... Photographed image, 26... Video RAM,
27... Image processor, 28... Level setting device, 29
...binarized image, 30...image characteristic measuring device, 31
...Arithmetic system, 32...Reference data setter, 33...
・Comparison/calculator, 34... Dead area setting device, 35...
Reference value, 36... Controlled system, 37... Welding power source,
38... Traveling trolley, 39... Trolley drive motor, 40
···Amplifier. Figure 2 Figure 3 Figure 5 (A) Figure 6

Claims (2)

[Claims] (1) After detecting at least the width of the molten pool formed directly under the welding torch, this detected data is compared and calculated with a preset value, and the welding current is adjusted to the set value.
An automatic welding method characterized by controlling welding speed to perform welding. (2) A welding torch, a visual sensor system that detects at least the width of the molten pool by looking at the molten pool formed directly below the welding torch from diagonally above, and storing and processing the detection data from this visual sensor system. Equipped with an image processing system, an image characteristic measuring device that reads the shape characteristics of the molten pool, a calculation system connected to the image characteristic measuring device, and a control target system that amplifies and executes the output signal from the calculation system. An automatic welding device characterized by: