CN115365655A - Method for identifying annular narrow-gap laser filler wire welding by adopting high-speed photography monitoring - Google Patents

Abstract

The invention discloses a method for identifying annular narrow-gap laser filler wire welding by adopting high-speed photography monitoring, and belongs to the technical field of laser welding. Collecting a molten pool motion state, a welding wire melting state and a molten drop transition picture in real time during welding; during collection, the collection equipment, the laser beam and the welding wire end are kept in a relatively static state, and a workpiece to be welded rotates; preprocessing the acquired image; identifying the edge profile morphology of the molten drop through an image processing algorithm; intercepting the identified picture and processing to obtain a simplified cross section diagram of the welding bead wire feeding end; measuring the distance from the welding wire to the two sides of the welding bead, the distance from the welding wire to the position to be welded of the workpiece, the included angle between the molten pool and the side wall and the relative position of the laser light source and the welding wire in the simplified diagram; welding parameters are adjusted based on the measured data. After high-speed photography shooting, collection and processing, qualitatively judging the molten drop transition form and transition in the set range of each welding process parameter, judging the influence of the molten pool motion state on the welding seam, and further dynamically adjusting the welding parameters to ensure the quality and the welding efficiency of a welding joint.

Description

Method for identifying annular narrow-gap laser filler wire welding by adopting high-speed photography monitoring

Technical Field

The invention relates to a narrow gap laser filler wire welding method, in particular to a method for identifying annular narrow gap laser filler wire welding by adopting high-speed photography monitoring, and belongs to the technical field of laser welding.

Background

The low-carbon steel occupies a large proportion in industrial production, and is widely applied to main structural members of mining machinery, pressure vessels, power plants, bridges and the like. At present, a great number of medium and heavy plates (the thickness of the medium plate is more than or equal to 12mm, and the thickness of the heavy plate is more than 20 mm) are difficult to weld in the engineering machinery industry. The traditional electric arc welding has the problems of low efficiency, large energy consumption, high cost, general quality and the like, and adopts narrow-gap laser filler wire welding to replace CO ₂ The gas shielded welding method changes the traditional 45-degree bevel into a narrow-gap bevel (the single-side angle of the bevel is 3 degrees), automatically feeds welding wires into the bevel, melts the welding wires by laser beams, forms a welding seam together with a base metal after cooling, is provided with a gas guide pipe in front of or behind a laser gun body, and blows inert gas on the side to restrain metal plasma and simultaneously protect the welding seam by gas. The welding wire filling amount of the narrow-gap laser wire filling mode can be saved by about 80%, the processing efficiency can be improved by 8-10 times, and the cost and the energy consumption are greatly saved; the heat input is reduced, the problems of welding stress and deformation are solved, and the product quality is improved. At the same time. Under the action of laser beam, the base metal is melted to form keyhole. Therefore, the narrow gap laser wire filling welding technology integrates the advantages of narrow gap welding and laser welding, and becomes one of the most potential research directions in thick plate welding.

At present, the research on the forming rule of a welding seam aiming at different welding process parameters (wire distance from the surface of a workpiece, laser power, welding speed, wire feeding speed, oscillating frequency of an oscillating mirror and oscillating amplitude of the oscillating mirror) of narrow-gap laser wire filling welding and the real-time monitoring, adjusting and optimizing of the welding process parameters aiming at the workpieces with different thicknesses are less, and the cost is higher for the processing of complex structural parts and for the research on the welding process due to excessive tests.

In 29.29.5.2020, the chinese patent application 2020100455290 discloses a system and a method for monitoring molten drop transition of a double-laser-beam bilateral synchronous welding filler wire based on high-speed image pickup, wherein a high-speed image pickup system is used for monitoring the molten drop transition mode and frequency in the process of double-laser-beam bilateral synchronous welding filler wire in real time, so as to achieve the purpose of monitoring the molten drop transition of the double-laser-beam bilateral synchronous welding filler wire at different wire feeding speeds in the whole process. However, the method only monitors the molten drop transition at different wire feeding speeds, and cannot monitor and adjust various welding parameters, so that the quality of a welding joint and the welding efficiency are influenced.

Disclosure of Invention

The invention aims to overcome the defects of the prior art and provides a method for identifying annular narrow-gap laser filler wire welding by adopting high-speed photographic monitoring, which can comprehensively acquire welding state data in the narrow-gap laser filler wire welding process and effectively ensure the quality and the welding efficiency of a welding joint.

In order to solve the technical problem, the method for identifying the annular narrow-gap laser filler wire welding by adopting high-speed photography monitoring provided by the invention comprises the following steps:

s1, collecting a molten pool motion state, a welding wire melting state and a molten drop transition picture in real time during welding; during collection, the collection equipment, the laser beam and the welding wire end are kept in a relatively static state, and a workpiece to be welded rotates;

s2, preprocessing the acquired image;

s3, identifying the edge profile morphology of the molten drop through an image processing algorithm;

s4, intercepting and processing the picture identified in the S3 to obtain a simplified cross section diagram of the welding bead wire feeding end;

s5, measuring the distance from the welding wire to the two sides of the welding bead, the distance from the welding wire to the position to be welded of the workpiece, the included angle between a molten pool and the side wall and the relative position of a laser light source and the welding wire in the simplified diagram in the S4;

and S6, adjusting welding parameters according to the data measured in the S5.

In the invention, a laser light source is used as backlight when an image is collected.

In the invention, the process of S2 is as follows:

s21, carrying out weighted average on the components of the R, G, B channels of the image by using different weights by using a weighted average method to obtain a more reasonable gray image: l = R299/1000G 587/1000B 114/1000;

s22, carrying out image geometric transformation on the gray level image;

and S23, enhancing the image processed in the S22 by adopting a frequency domain method or a space domain method.

In the present invention, the process of S3 is: after the preprocessed pictures are input into the CNN network, the features of the pictures are extracted through a plurality of convolution pooling operations, and then the picture features are sent into a full-connection layer network to finish the classification and identification of the images.

In the present invention, the process of S4 is:

s41, transmitting the picture to an identification system for carrying out gray level processing on the picture;

s42, carrying out binary conversion on the image subjected to gray processing;

s43, searching and drawing the outline on the processed picture.

The invention has the beneficial effects that: (1) After high-speed shooting, collecting and processing, qualitatively judging the transition form and the transition of molten drops in the set range of all welding process parameters, judging the influence of the motion state of a molten pool on a welding seam, and further dynamically adjusting welding parameters to ensure the quality and the welding efficiency of a welding joint; (2) Through the image recognition system, a simplified diagram can be formed on the welding section of the wire feeding end, the welding state data in a welding bead can be measured, and a basis is provided for narrow-gap laser welding of different engineering mechanical structural parts; (3) By comparing the simplified diagram measurement data with the experiment database, the welding process parameters are adjusted and optimized in time, and multiple experiments are avoided, so that the efficiency is improved, and the cost is saved; (4) The laser light source is used as backlight, so that arc light during welding can be attenuated, and clear images can be obtained in the acquisition equipment.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

FIG. 1 is a schematic view of a welding apparatus;

FIG. 2 is a diagram of a convolutional neural network model architecture;

FIG. 3 is a schematic diagram of an image intelligent recognition system;

FIG. 4 is a schematic diagram of an alignment process;

FIG. 5 is a schematic view of the effect of the spacing of the welding wire from the workpiece surface;

FIG. 6 is a diagram illustrating the effect of the oscillating frequency of the galvanometer;

FIG. 7 is a schematic diagram illustrating the effect of the oscillating amplitude of the galvanometer;

FIG. 8 is a diagram of weld macro topography and weld bead topography at different swing amplitudes;

FIG. 9 is a schematic illustration of laser welding power effects;

FIG. 10 is a schematic illustration of the effect of laser welding speed;

FIG. 11 is a schematic view of the wire feed speed effect of the wire feeder.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.

Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.

In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings or the orientations or positional relationships that the products of the present invention are conventionally placed in use, and are only used for convenience in describing the present invention and simplifying the description, but do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus, should not be construed as limiting the present invention. Furthermore, the terms "first," "second," "third," and the like are used solely to distinguish one from another and are not to be construed as indicating or implying relative importance.

In the description of the present invention, it should also be noted that, unless otherwise explicitly specified or limited, the terms "disposed," "mounted," "connected," and "connected" are to be construed broadly and may, for example, be fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.

Some embodiments of the invention are described in detail below with reference to the accompanying drawings. The embodiments described below and the features of the embodiments can be combined with each other without conflict.

As shown in FIG. 1, the welding device used in the method for identifying the annular narrow gap laser filler wire welding by high-speed photography monitoring of the invention comprises a laser welding system 1, a digital wire feeding control system 2, a high-speed photography acquisition system 3 and an intelligent image identification system 4.

The laser welding system 1 adopts an IPG YLS 4000-S2T laser welding device, the maximum output power of the laser is 4kW, the diameter of an optical fiber core is 100um, the focal length is 300mm, IPG D30 Wobble galvanometer equipment is provided, the diameter of a facula when the galvanometer is not started and defocusing is zero is 0.15mm, after the galvanometer is started, the maximum deflection amplitude is 3mm when the defocusing is zero, the maximum bearing laser power of the equipment is 6kW, the maximum swinging frequency is 1000Hz, a FANUC M-20iA type robot is equipped as a laser head operation bearing mechanical arm, the maximum bearing weight of the equipment is 1811mm and is 20kG at most, and the repeated positioning precision can reach 0.04mm.

The digital wire feeding control system 2 adopts a WF25i REEL R wire feeder, and the maximum wire feeding speed is 6m/min.

The high-speed photographic acquisition System 3 adopts a CP80-3-M-540 type high-speed camera manufactured by Optronics company, uses a Diode Laser System 40W Laser light source 5 manufactured by BWT company as backlight, and has the resolution of 1696 multiplied by 1708 and the effective area of a picture of 13.57mm multiplied by 13.68mm.

In this embodiment, the 40W laser light source 5 is used as a backlight to irradiate a laser beam to a welding position, so that the arc during welding can be attenuated, and a clearer image can be obtained in the high-speed photographing device.

The system architecture of the image intelligent recognition system device 4 is shown in fig. 3.

The method for identifying the annular narrow-gap laser filler wire welding by adopting high-speed photography monitoring in the embodiment comprises the following specific processes:

the method comprises the following steps: starting the whole welding system, and automatically carrying out high-frequency shooting by a high-speed camera;

step two: turning on a backlight source;

step three: and preprocessing the acquired image to improve the signal-to-noise ratio and the color of the image.

(1) Graying treatment: the three channels of the image R, G, B are processed sequentially. In order to achieve the purpose of increasing the processing speed of the whole application system, the acquired color image needs to be grayed to reduce the data amount required to be processed. In this embodiment, a weighted average method is used to perform weighted average on the three components with different weights according to importance. In view of the fact that human eyes have the highest sensitivity to green and the lowest sensitivity to blue, in this embodiment, a reasonable gray image can be obtained by performing weighted average on RGB three components according to the following formula, and the conversion algorithm is:

L＝R*299/1000+G*587/1000+B*114/1000

where L is the gray level of the image, R is the red component, G is the green component, and B is the blue component.

(2) Geometric transformation: the images subjected to graying processing are processed through geometric transformations such as translation, transposition, mirroring, rotation, scaling and the like, so that the system error of a high-speed photographic acquisition system and the random error of the instrument position (imaging angle, perspective relation and even the reason of a lens) are corrected. In addition, a gray interpolation algorithm is used, wherein the gray interpolation is a method for redistributing pixels of a grayed image so as to change the number of the pixels, and an interpolation program automatically selects the pixels with better information as a space for adding and compensating blank pixels so as to achieve the effect that the image looks smoother and cleaner when the image is enlarged. The commonly used methods are nearest neighbor interpolation, bilinear interpolation and bicubic interpolation.

3. Image enhancement: the image after geometric transformation is enhanced by using a frequency domain method so as to improve the visual effect of the image, emphasize the overall or local characteristics of the image, change the original unclear image into clear or emphasize some interesting characteristics, enlarge the difference between different object characteristics in the image, inhibit the uninteresting characteristics, improve the image quality and enrich the information content, and enhance the image interpretation and identification effects so as to meet the requirements of subsequent special analysis.

In the actual use process, the image enhancement processing can also be carried out by using a spatial domain method.

Step four: the edge profile morphology of the molten drop is identified through an image processing algorithm, as shown in fig. 2:

(1) And inputting the preprocessed picture into a CNN network, and extracting the features of the picture through a plurality of convolution and pooling operations.

(2) Sending the extracted picture characteristics into a full-connection network for classification and identification;

(3) And the full connection layer combines a plurality of groups of pooled data characteristics into a group of signal data to be output, and the picture category identification is carried out.

Step five: intercepting the picture obtained in the fourth step, and transmitting the picture to an image intelligent identification system to obtain a simplified cross section diagram of the welding bead wire feeding end;

(1) Transmitting the picture to a recognition system for picture gray processing, and using cv2.CvtColor of Opencv (OpenCV is an open-source computer vision library);

(2) Performing binary conversion on the image subjected to gray processing, and selecting cv2.Threshold in Opencv;

(3) Searching for the contour by adopting a cv2.FindContours mode in Opencv;

(4) And drawing the outline, wherein the outline is drawn on the image by using cv2.DrawContours in Opencv.

Step six: and measuring the distance from the welding wire to the two sides of the welding bead, the distance from the welding wire to the position to be welded of the workpiece, the included angle between the molten pool and the side wall and the relative position of the laser light source and the welding wire in the simplified diagram.

Importing the drawn Image into Image J (the Image J is common Image processing software based on java), selecting a straight line tool, drawing a straight line with unit length on a ruler, calibrating, drawing a straight line between two vertexes to be measured, and measuring the length according to Ctrl + M; similarly, an angle measuring tool is selected, and three points are selected on the angle to be measured to obtain an angle value.

Step seven: and comparing the data obtained by the measurement in the step six with a database formed by a large amount of data summarized in the previous experiment, thereby adjusting the welding process parameters.

As shown in fig. 4, a captured picture (a) is shot at high speed, then a weld bead parameter map (b) is obtained, the recognized picture is processed by an image intelligent recognition system device to obtain a weld bead wire-feeding end section simplified map (c), and the parameter value (d) is measured by the section simplified map: firstly, ensuring the welding wire position to be in a symmetrical position in the middle of a welding seam; measuring the delta H value, and comparing with an early-stage experiment database to determine whether the delta H value meets the welding good range; by measuring whether the included angles alpha and beta between the molten pool and the side wall are acute angles, the U-shaped groove beneficial to forming the next welding is met. By measuring the numerical values and comparing with an experimental database, if the fluctuation difference of the measured values is large during welding, the welding process parameters are adjusted in time.

The upper part of fig. 5 is an image of the surface of the workpiece and the lower part is an image of the inside of the gap during welding. Where Δ H is the spacing between the wire and the workpiece surface. It can be seen from fig. 5 that when Δ H is 0.5mm, since the tip of the welding wire is closer to the surface of the workpiece, the laser energy is focused on the welding wire and the workpiece and melts them at the same time, and the area of molten pool formation is large. In the welding process, the distance between the front section of the welding wire and the molten pool is small, the molten pool covers the front end of the welding wire, the welding wire is inserted into the molten pool, the welding wire is heated in the molten pool by laser to be melted for welding, no obvious molten drop transition exists in the welding process, and the molten drop transition form is spreading transition. When delta H is 1 and 1.5mm, the front end of the welding wire has a certain distance from the working surface, the welding wire is irradiated by laser to melt the front end of the welding wire to form a molten drop, part of energy heats a workpiece to melt the workpiece, a formed molten pool has a small gap with the front end of the welding wire, at the moment, the front end of the welding wire is melted to form a small molten drop during welding, the molten drop is in contact with the molten pool when the molten drop is not completely separated from the welding wire, the molten pool which is in dynamic motion has an adsorption effect on the formed incomplete molten drop due to the action of surface tension, the molten drop is stably transited from the front end of the welding wire to the inside of the molten pool, and the molten drop can be observed under high-speed photography, and the splashing generated by welding is small. When the delta H is 2mm, a large bulge is generated in the middle of the welding line, which is not beneficial to the next filling welding, and the defects of non-fusion between layers, air holes and the like can be formed.

In FIG. 6, the upper half is an image of the inside of the gap during welding, and the lower half is a cross-sectional view of the welded portion after welding. It can be seen from fig. 6 that, when swing welding is performed at a low frequency, the fluctuation range of the welding pool is low, and when wire filling welding is performed, the stay time of laser at the front end of the welding wire is short, the welding wire cannot be sufficiently melted, so that molten drops enter the molten pool in a spreading transition mode, the fluidity of the molten pool is increased along with the increase of the swing frequency, liquid weld metal has a tendency of moving towards the side wall under the influence of the swing laser, and the fluctuation of the molten pool is more severe along with the further increase of the swing frequency. The molten drop transition form is from spreading to liquid bridge transition, and can be seen from a high-speed photographic picture, the fluidity of a molten pool is further promoted by the swing of the light source, the laser facula moves along a counterclockwise circular path, the swing frequency is different, and the stay time of the heat source in unit area is reduced along with the increase of the swing frequency. When the 20HZ oscillation frequency is adopted for welding, the molten pool is almost free from fluctuation, and the trend that the molten pool moves towards the direction of the side wall of the groove is more obvious along with the increase of the oscillation frequency.

Compared with the traditional laser wire filling welding, the swinging laser welding method has the advantages that after the swinging laser welding is adopted, the convex part above the welding bead is obviously eliminated after the filling welding, and the welding bead is in a U-shaped groove after the welding. Research shows that when the laser swing frequency is low, the laser heat source heats the same region for a long time in unit time, the energy is concentrated, and the heat is concentrated at the middle part of a welding seam due to the fact that the fluctuation of a molten pool is small and the conduction capability of the heat to the position far away from the welding molten pool is weakened, and cracks appear in the middle part of the welding seam after wire filling welding. Along with the increase of the swing frequency, the molten pool is heated more uniformly, the retention time of a heat source at each position is reduced, the fluctuation of the molten pool is obvious, the tendency of heat transfer to the side wall is obvious, and welding cracks disappear. When the oscillation frequency is increased to 40Hz and 60Hz, a U-shaped welding bead which is favorable for the next layer of wire filling welding can be obtained, and when the oscillation frequency is 40Hz, the cold crack defect appears at the bottom of the molten pool, thereby causing the welding defect. As can be seen from the top view of the weld bead in the lower half of FIG. 6, there are different degrees of protrusion and recession when the laser beam oscillation frequency is 20Hz and 80 Hz. It can be seen from the sectional view of the welding seam that when the oscillation frequency is 80Hz, the middle part of the welding seam has continuous bulges when the oscillation frequency is too high, and the non-fusion between layers and the air hole defect are easy to generate during the wire filling welding of the next layer. It can be seen from the figure that as the laser oscillation frequency increases, the weld width increases linearly, the weld depth decreases first, and then the weld width is kept at the same level, it can be seen that the oscillation frequency has a large influence on the weld width, the weld forming coefficient increases, and the weld forming coefficient is similar when the laser oscillation frequency is 40Hz and 60 Hz. In conclusion, when the oscillating frequency is 60Hz, the weld bead has no obvious defects, the formed U-shaped section is favorable for the wire filling welding of the next pass, and the forming quality of the weld joint is optimal when the oscillating frequency is 60 Hz.

As can be seen from figures 7 and 8, when the swing diameter of the light spot is small, the laser light source is mainly concentrated at the front end of the welding wire, so that the welding wire is fully melted to form molten drops, the molten drops are stably transited to a molten pool, the fluctuation amplitude of the molten pool is small, and the area of a heat source covered by laser is small. When the diameter of a light spot is 1mm and 1.5mm, the melting condition of the welding wire by the laser beam is similar, most energy is transferred into a molten pool when the welding wire is heated to be molten by the laser beam, so that larger penetration depth and melting width are obtained, and the transition form of a molten drop is mainly liquid bridge transition. With the further increase of the diameter of the light spot, when the diameter of the light spot is 2mm, the front end of the welding wire is melted, part of the front end of the welding wire is not melted, the front end of the welding wire enters a molten pool along with the movement of the welding wire, the droplet transition form is spreading transition, the area of the molten pool heated by the light source is further increased, the droplet transition form is biased to spreading transition, the melting range of the molten pool is enlarged, and the fluctuation range of the molten pool is obviously increased. It can be seen from the high-speed photographic image that, under the condition of unchanged swinging frequency, when welding with a small spot diameter, the molten drop is completely melted before entering a molten pool and enters the molten pool in a bridge transition mode, when welding is carried out with a large swinging amplitude, along with the increase of the swinging diameter of the spot, the laser not only acts on the welding wire but also acts on the molten pool, and when the molten drop is not completely formed, the molten drop is merged into the molten pool together with the non-fusion welding wire.

When the diameter of a light spot is small, laser energy is concentrated, the laser energy is mainly concentrated at the front end of a welding wire, a welding wire molten drop is directly merged into a molten pool, laser penetrates into the molten pool better, large penetration depth is obtained in the welding process, the melting behavior of the welding wire blocks the escape of internal air holes, and small air holes are formed between welding layers. Along with the increase of the swing amplitude, the heating time per unit area is reduced in unit time, the laser firstly acts on the welding wire, part of the laser is driven into a molten pool, the depth of the molten pool is correspondingly shallow, and the effect is more obvious along with the further increase of the diameter of the light spot. The change of the spot diameter is suitable for the change of the groove width during the narrow-gap laser wire filling welding, and the welding groove width is increased along with the increase of the welding height and the spot diameter is increased along with the increase of the welding height in the welding process. As can be seen from the top view of the weld bead in fig. 8, when the swing amplitude is 0.5mm, the melt width is narrow, and the defects such as undercut spatter are likely to occur during the cap surface welding. When the swing amplitude is 1mm, 1.5mm and 2mm, the weld bead forming quality is good, no obvious defect exists, and the welding device is suitable for welding.

FIG. 9 is a photograph of the effect of welding power on the molten pool movement state and the welding wire melting state in the groove wire-filling state using a high-speed camera. In fig. 9, the upper half is a schematic view of a weld pool in a groove in a moving state, and the lower half is a cross-sectional view of a welded part after completion of welding. It can be seen from the upper half of fig. 9 that the welding power is small, the wire is not completely melted before entering the molten pool, and the wire is melted mainly by the heat of the laser and the molten pool. When the welding heat input is 3.6kJ/cm, a small-size droplet is formed at the front end of the welding wire as can be seen from high-speed photography, and the droplet is absorbed into the molten pool by surface tension when the molten pool is close to the molten pool, and the molten pool has small fluctuation in the process. With the further increase of laser energy, the front end of the welding wire can be obviously observed to be melted by laser to generate molten drops above the molten pool, when the laser power is 3.92kW, the welding wire is completely melted above the molten pool, the molten drops enter the molten pool under the action of gravity, the molten pool can generate large fluctuation and generate large splashing when entering the molten pool, and therefore, the welding heat input (3.92 kJ/cm) is large and exceeds the requirement of the welding wire melting heat under the wire feeding speed and the welding speed. Under the welding power of 3.8kW, the requirement of welding wire melting is completely met, and no large splashing exists in the welding process. It can be seen that when the power is small, the transition form of the molten drops is spreading transition, and the transition form of the molten drops is changed from liquid bridge transition to particle transition along with the increase of the power.

When the power value is smaller, the welding bead area is smaller, small bulges are easily formed on the welding seam after welding, the welding is not beneficial to next wire filling welding, the welding defect is easily caused, and when the bulges are too high, the welding wire is easily extended to the bottom under the condition of not melting, so that the welding instability is caused. The increase of the welding bead section area is obvious along with the increase of the power, because under the condition of higher power, the melting pool can obtain larger penetration and melting width. As can be seen from the figure, the penetration and the fusion width of the welding seam increase linearly with the increase of the welding power, and the influence of the power on the fusion width is large, which shows that when the power is loaded to a certain degree, the change of the welding bead is mainly reflected in the fusion width. When the power is larger, the welding bead formed after welding is gradually gentle, and under the condition that other parameters are not changed, the power is increased, and the appearance of the welding bead is also improved. It is found that, when the welding heat input is 3.92kJ/cm, weld bead internal defects are easily caused. As can be seen from the weld forming coefficients of the graphs, the weld forming coefficients tend to decrease first and then increase under the condition that the welding power is increased. In conclusion, a weld bead with good forming quality can be obtained when the welding power is 3.8 kW.

Fig. 10 is a schematic view of the upper half of the welding apparatus taken while the weld pool is moving in the groove, and a sectional view of the welded portion after completion of welding in the lower half. As can be seen from the figure, when the welding speed is small, since the wire feeding volume becomes smaller with the increase of the welding speed per unit length, the welding heat input is reduced from 4.89kJ/cm to 3.42kJ/cm, and the welding heat input span is large. When the welding speed is low, most of the welding wire begins to melt near the molten pool, the molten drop transition form is spreading transition, at the moment, the welding wire enters the molten pool under the state of incomplete melting, the fluctuation of the welding molten pool is large, the welding wire gradually separates from the position of the molten pool to melt along with the further improvement of the welding speed, the transition form is liquid bridge transition when the molten drop enters the molten pool, and when the welding speed is 0.48 m/min and 0.54m/min, the molten pool moves stably. With the further improvement of the welding speed, before the welding wire is completely melted, the liquid welding wire metal required to be filled in the bottom molten pool is incomplete, the welding speed is higher than the wire feeding speed required by the welding seam filling, the moving state of the molten pool is unstable at the moment, the molten pool tends to be welded on one side of the welding seam, and the molten drop transition form is particle transition at the moment.

It can be seen from the figure that the welding speed is within 0.42 m/min-0.6 m/min, the welding quality is good, after welding, no defects such as interlayer unmelted welding and sidewall unmelted welding are generated on the cross section, the filler wire welding layer height tends to decrease with the increase of the welding speed, the penetration depth is reduced from 3.97mm to 2.9mm, the welding fusion width variation interval is smaller, and is reduced from 3.87mm to 3.51mm, the fusion width completely meets the welding requirement at the moment, but with the increase of the welding bead, the fusion width is not enough to fuse the sidewall at the moment, the change of the groove width needs to be considered in the actual welding process, and when the welding speed is lower, the stacking height of the welding layer is larger, a larger bulge is formed on the surface of the welding bead, which is not beneficial to the next filler wire welding, and with the increase of the welding speed, the bulge part gradually decreases. As can be seen from the figure, the weld penetration and the weld width have a linear decreasing trend along with the increase of the welding speed, and the influence of the welding speed on the penetration and the weld width is approximate. Along with the increase of welding speed, the welding seam forming coefficient becomes great, and in the multilayer filler wire welding process, welding speed increases, can lead to the reduction of build-up welding height, and every layer of heap height is about 3 ~ 4mm, along with the increase of the welding number of piles, the bevel face width also increases thereupon, when adopting the same welding parameter to weld once more, the actual height of piling up of every layer of welding bead can reduce to some extent, should suitably adjust a speed that send this moment to satisfy the welding requirement.

Fig. 11 is a schematic view of the upper half of the groove taken when the weld pool is in motion, and a sectional view of the welded portion after the lower half of the groove is welded. The diagram is shot under the moving state of the welding pool in the notch, and the effect of changing the welding speed is opposite. Under the unit length, the feeding volume of the welding wire increases along with the increase of the wire feeding speed, the welding heat input is maintained at a certain value, the melting of the front end of the welding wire is influenced to a certain degree at the moment, the heat of the welding wire in the unit volume increases, the heat of the welding wire decreases along with the increase of the wire feeding speed, and the melting of the welding wire is influenced to a certain degree. It can be seen from the figure that when the wire feeding speed is low, most of the welding wire starts to melt when being far away from the molten pool, and the welding wire moves towards the inside of the molten pool along with the welding wire, the molten drop transition form liquid bridge transition is performed, at the moment, because the fluctuation of the welding molten pool is low, along with the further improvement of the wire feeding speed, the welding wire cannot fully form the molten drop transition welding wire before entering the molten pool, the molten drop transition form at the moment is gradually changed into spreading transition from the liquid bridge transition form, the molten pool is influenced by the unmelted welding wire to be discontinuous fluctuation, the instability of the welding process is serious, and therefore, the energy generated by the laser beam is not enough to support the high wire feeding speed.

As can be seen from fig. 11, when the wire feeding speed is small, the defect that the side wall is not fused is easily generated, the weld metal pool cannot sufficiently melt the side wall due to the small filling amount, and the defect is caused, and after welding, the high bulge is generated in the middle of the weld bead, which is not favorable for the next wire filling welding. The appearance of the welding bead is gradually U-shaped after welding, and the bulge generated by the low wire filling speed is also gradually reduced. The weld bead area is enlarged as the wire filling amount increases in a unit length, but when the welding wire feed speed reaches 5.0m/min, cracks are generated in the center of the weld bead, and at this time, welding defects are caused due to a large stacking volume of the welding wire. Within 4.6-4.8 m/min, the welding quality is good, no defects such as interlayer non-melting and side wall non-melting are generated on the welded section, the height of a wire-filling welding layer tends to be gradually increased along with the increase of wire feeding speed, the melting depth is increased from 3.10mm to 3.5mm, and the change interval of the welding melting width is also increased from 3.44mm to 3.75mm. The penetration and the fusion width completely meet the welding requirement at the moment, the width of the groove is increased along with the increase of the number of welding layers, and the welding speed is properly increased to meet the welding bead filling requirement. It can be seen from the figure that, as the wire feeding speed increases, the welding penetration and the fusion width are in a linear increasing trend, and at this time, the influence of the wire feeding speed on the penetration and the fusion width is approximate, and meanwhile, the weld forming coefficient is in a decreasing trend, which indicates that the increase of the wire feeding speed can cause the decrease of the weld forming coefficient, and generally speaking, the influence on the weld forming quality is large, and in summary, under the condition that other welding parameters are not changed, the welding filling effect of the welding groove between 4.6m/min and 4.8m/min is good, and in the single-pass multilayer welding process, the requirement of the welding groove on the welding wire filling amount needs to be considered.

In FIGS. 5 to 11, T is time, T ₀ Is an arbitrary time, ms is a microsecond,

the foregoing is only a preferred embodiment of this invention and it should be noted that modifications can be made by those skilled in the art without departing from the principle of the invention and these modifications should also be considered as the protection scope of the invention.

Claims (5)

1. A method for identifying annular narrow gap laser wire filling welding by adopting high-speed photography monitoring is characterized by comprising the following steps:

s1, collecting a molten pool motion state, a welding wire melting state and a molten drop transition picture in real time during welding; during collection, the collection equipment, the laser beam and the welding wire end are kept in a relatively static state, and a workpiece to be welded rotates;

s2, preprocessing the acquired image;

s3, identifying the edge profile morphology of the molten drop through an image processing algorithm;

s4, intercepting and processing the picture identified in the S3 to obtain a simplified cross section diagram of the welding bead wire feeding end;

s5, measuring the distance between the welding wire and the two sides of the welding bead, the distance between the welding wire and the position to be welded of the workpiece, the included angle between a molten pool and the side wall and the relative position of a laser light source and the welding wire in the simplified diagram S4;

and S6, adjusting welding parameters according to the data measured in the S5.

2. The method for identifying the annular narrow gap laser wire-filling welding by adopting the high-speed photographic monitoring as claimed in claim 1, wherein the method comprises the following steps: when the image is collected, a laser light source is used as backlight.

3. The method for identifying the annular narrow gap laser filler wire welding by adopting the high-speed photographic monitoring as claimed in claim 1 or 2, wherein the process of S2 is as follows:

s21, carrying out weighted average on the components of the R, G, B channels of the image by using different weights by using a weighted average method to obtain a more reasonable gray image: l = R299/1000G 587/1000B 114/1000;

s22, carrying out image geometric transformation on the gray level image;

and S23, enhancing the image processed in the S22 by adopting a frequency domain method or a space domain method.

4. The method for identifying the annular narrow gap laser filler wire welding by adopting the high-speed photographic monitoring as claimed in claim 3, wherein the process of S3 is as follows: after the preprocessed pictures are input into the CNN network, the features of the pictures are extracted through a plurality of convolution pooling operations, and then the picture features are sent into a full-link layer network to finish the classification and identification of the images.

5. The method for identifying the annular narrow gap laser filler wire welding by the high-speed photographic monitoring as claimed in claim 3, wherein the process of S4 is as follows:

s41, transmitting the picture to an identification system for carrying out gray level processing on the picture;

s42, carrying out binary conversion on the image subjected to gray processing;

s43, searching and drawing the outline on the processed picture.

Images