Method for laser welding in-process inspection

Abstract

To provide a method for laser welding in-process inspection that can accurately detect an internal flaw of a molten pool.SOLUTION: A method for laser welding in-process inspection includes a step of measuring a molten pool width in a direction orthogonal to a welding direction from a molten pool image and, when the molten pool width predetermined backward of the welding direction of the molten pool is equal to or less than a lower limit threshold or equal to or more than an upper limit threshold, determining that there is an internal flaw.SELECTED DRAWING: Figure 11

Description

[Technical field]

[0001] The present invention relates to a method for in-process inspection of laser welding. [Background technology]

[0002] Patent document 1 discloses a method for determining the quality of laser welding based on whether the amount of variation is within an appropriate range or not, by calculating the difference between the maximum and minimum values ​​of the molten pool width at a specified distance from the center position of the laser in an image of the molten pool during laser welding. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 6856845 Summary of the Invention [Problem to be solved by the invention]

[0004] However, while the above-mentioned conventional technology can determine whether the quality of laser welding is uniform, it has the problem that it cannot accurately detect internal defects in the molten pool. One object of the present invention is to provide a laser welding in-process inspection method capable of detecting internal defects in a molten pool with high accuracy. [Means for solving the problem]

[0005] In one embodiment of the present invention, a laser welding in-process inspection method measures the width of the molten pool in a direction perpendicular to the welding direction from an image of the molten pool, and determines that there is an internal defect if the width of the molten pool behind a certain point in the welding direction is equal to or less than a threshold value. Effect of the Invention

[0006] Therefore, in the present invention, internal defects in the molten pool can be detected with high accuracy. [Brief description of the drawings]

[0007] [Figure 1] FIG. 1 is a schematic diagram of a laser welding and inspection device according to a first embodiment. [Diagram 2] 2 is an example of an X-ray CT image of an internal defect to be inspected in the first embodiment. [Diagram 3] 1 is an example of an image of a molten pool taken by the camera 12 during normal welding with laser welding. [Figure 4] 1 is an example of an image of a molten pool when an internal defect occurs. [Diagram 5] This is an example of an image of the molten pool immediately after an internal defect occurs. [Figure 6] 1 is an example of an image of a molten pool when fumes are generated. [Figure 7] This is an example of the molten pool length and internal defect occurrence position calculated from the molten pool image. [Figure 8] FIG. 2 is a diagram showing the width of the molten pool in a direction perpendicular to the direction of progress of the molten pool 20 in the molten pool image. [Figure 9] FIG. 13 is a diagram showing the width of the molten pool when fumes are generated. [Figure 10] FIG. 13 is a diagram showing the width of the molten pool immediately after an internal defect occurs. [Figure 11] FIG. 1 is a diagram showing the relationship between the molten pool width (9) calculated from the molten pool image and the defect occurrence position. [Figure 12] FIG. 1 is a diagram showing the relationship between the molten pool width (10) calculated from the molten pool image and the defect occurrence position. [Figure 13] 4 is a flowchart showing the flow of a defect determination process executed by the image processing and inspection device 15. [Figure 14] This is an image of the molten pool when a large fume was generated. [Figure 15] An example of (a) the time waveform of brightness and (b) the frequency distribution of a pixel in a molten pool image before applying a 100 Hz low-pass filter. [Figure 16] This is an example of (a) the time waveform of brightness and (b) the frequency distribution after applying a 100 Hz low-pass filter to a pixel in a molten pool image. [Figure 17]These are examples of molten pool images taken by camera 12 during normal laser welding, (a) an image of the molten pool before applying a low-pass filter, overlaid with the molten pool width obtained after image processing, and (b) an example of a molten pool image in which only the latter half of the molten pool has been binarized after applying a low-pass filter. [Figure 18] FIG. 1 shows (a) an image of the molten pool before low-pass filtering when an internal defect occurs, taken by the camera 12 during laser welding, overlaid with the molten pool width obtained after image processing, and (b) an image of the molten pool after low-pass filtering and binarizing only the latter half of the molten pool, as well as the relationship between the molten pool width (7) calculated from the molten pool image, the upper threshold, and the position of the defect occurrence. [Figure 19] FIG. 11 is a schematic diagram of a laser welding and inspection device according to a fourth embodiment. [Figure 20] FIG. 13 is a diagram showing a determination method for a welding light sensor alone. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] [Embodiment 1] FIG. 1 is a schematic diagram of a laser welding and inspection device according to the first embodiment. The laser welding and inspection device of the first embodiment includes a laser oscillator 1, a laser optical fiber 2, a processing head 3, a collimation lens 4, a partial wavelength transmission mirror 6, an objective lens 7, a rotary spindle 10, a processing stage 11, a camera 12, a mirror 13, an objective lens 14 for the camera, an image processing and inspection device 15, and a laser welding control device 16. The collimation lens 4 widens the beam width in the processing head 3. The partial wavelength transmission mirror 6 sends light emitted from an object to be welded (hereinafter referred to as a work) 5 to the camera 12. The objective lens 7 focuses the laser 8 with a wide beam width in the processing head 3 on the work 5. The mirror 13 guides the light emitted from the work 5 to the camera 12. The image processing and inspection device 15 processes the image captured by the camera 12 and determines whether or not there is an internal defect such as a blowhole.

[0009] The laser oscillator 1 of the first embodiment is a disk laser with a wavelength of about 1,030 nm, and the workpiece 5 is a fuel pump part made of 304 stainless steel. A laser beam 8 generated by a laser oscillator 1 is sent to a processing head 3 through a laser optical fiber 2. The beam width of the laser beam 8 is expanded by a collimation lens 4, passes through a partial wavelength-transmitting mirror 6, is focused by an objective lens 7, and is irradiated onto a workpiece 5. The workpiece 5 is fixed to a rotary spindle 10 and rotates at a predetermined speed in the direction of an arrow 9. The light emitted from the molten pool of the workpiece 5 is reflected by the objective lens 7, the partial wavelength-transmitting mirror 6, and a mirror 13, and then passes through a camera objective lens 14 to enter a camera 12.

[0010] FIG. 2 is an example of an X-ray CT image of an internal defect to be inspected in the first embodiment. 2, weld bead 18 is provided so as to straddle joint surface 17 indicated by a dashed line, and blowhole 19 has occurred inside weld bead 18. Blowhole 19 can be detected by visual inspection if its end is exposed on the surface of weld bead 18, but cannot be detected by visual inspection if its end is blocked.

[0011] FIG. 3 is an example of an image of a molten pool taken by the camera 12 during normal laser welding. The bright white area at the bottom right of the figure is the area that has become hot due to irradiation with the laser 8. There is a thin white area to the left of that, which is the molten metal in the molten pool 20 that has cooled a little but not yet solidified as the workpiece 5 moves from right to left, passing the laser irradiation point and no longer receiving energy. The molten pool 20 becomes thinner as it moves to the left, and will appear black when it cools and solidifies. The white area at the top right of the figure is the light 21 reflected by the workpiece 5.

[0012] Figure 4 is an example of an image of a molten pool when an internal defect occurs. FIG. 4 shows spatters 22 flying out from the molten pool 20. The size of the spatters 22 is close to the width of the molten pool 20 in the direction perpendicular to the direction of travel. Internal defects such as blowholes occur when the space formed by the spatters 22 flying out from the molten pool 20 cannot be filled with the surrounding molten metal. Therefore, if the occurrence of large spatters 22 can be detected, the occurrence of internal defects can be detected. However, the spatters 22 fly out in various directions at high speed, and depending on the shooting interval of the camera 12, the spatters 22 may not be captured. In addition, if the spatters 22 fly in the direction of the camera 12, it is difficult to capture them because they overlap with the molten pool 20. Therefore, it is necessary to capture the occurrence of large spatters 22 in another way.

[0013] Figure 5 is an example of an image of the molten pool immediately after an internal defect occurs. In the molten pool 20, the spatter 22 flies out and the amount of molten metal decreases, so the length of the molten pool 20 in the direction of travel is shortened. After that, the molten pool 20 gradually extends and becomes longer in the direction of travel again as in normal welding shown in Figure 3. This change in the molten pool 20 is slower than the flying spatter 22, so it can be captured even if the interval between shots by the camera 12 is slow. However, as shown in Figure 6, in laser welding, a white "haze" called a fume 24 may be generated from the laser irradiation point. Since this fume 24 also brightens the front of the molten pool 20, the molten pool length calculated from the length of the bright part in the direction of travel of the molten pool 20 varies greatly as shown in Figure 7. The broken line part at rotation angles 312 to 336° in Figure 7 shows the position where an internal defect occurs, and although the molten pool length decreases around 360° a little later, it also decreases in other parts, making it difficult to detect internal defects based on this.

[0014] FIG. 8 is a diagram showing the width of the molten pool in a direction perpendicular to the direction of travel of the molten pool 20 in a molten pool image. In FIG. 8, the rear side (left side in the figure) of the molten pool 20 is located far from the laser irradiation point where the fumes 24 are generated. Therefore, as shown in FIG. 9, even if the fumes 24 are generated, the molten pool widths (9), (10) can be measured with relatively high accuracy. Also, immediately after the occurrence of an internal defect, the molten pool widths (9), (10) at a rotation angle of about 360° become 0 as shown in FIGS. 10 to 12. Therefore, in the first embodiment, the molten pool width is detected from a rotation angle of just under 60° to just under 420°, and when the molten pool width becomes equal to or less than the threshold value of 0.3, it is determined that there is an internal defect.

[0015] FIG. 13 is a flowchart showing the flow of the defect determination process executed by the image processing and inspection device 15. In step S1, the process waits for receipt of a welding start signal from laser welding control device 16, and when the welding start signal is received, the process proceeds to step S2. In step S2, camera images of the molten pool (molten pool images) are repeatedly acquired and stored at regular intervals, and when a welding end signal is received, the process proceeds to step S3.

[0016] In step S3, the contrast enhancement process is performed on the acquired molten pool images. When the contrast enhancement process is completed on all acquired molten pool images, the process proceeds to step S4. In step S4, the width of the molten pool behind the molten pool in the acquired molten pool image is calculated as the distance between the minimum and maximum points of the vertical brightness change rate at specific positions in the image (for example, positions (9) and (10) in Figures 8 to 10). When the calculation of the molten pool width has been completed for all acquired molten pool images, proceed to step S5.

[0017] In step S5, a smoothing process is performed by calculating a moving average of X pieces of the molten pool width obtained in step S4, and the process proceeds to step S6. In step S6, if the width of the molten pool in the rotation angle range to be judged (for example, in Figures 11 and 12, the rotation angle range from just under 60° to just under 420°) continues to be less than the threshold value Y (for example, 0.3 in Figures 11 and 12) for Z consecutive times (Z is a natural number greater than or equal to 2), it is judged that an internal defect has occurred.

[0018] Next, the effects of the first embodiment will be described. In recent years, direct injection gasoline engines that achieve high thermal efficiency by injecting pressurized fuel directly into the combustion chamber have become popular for gasoline engine vehicles, aiming to achieve both fuel economy and improved power output. These direct injection gasoline engines use a high-pressure fuel pump that uses the engine's power to boost the fuel sent from the fuel tank at low pressure and supply it to the injector.

[0019] Laser welding, which allows high-speed, low-distortion welding, is often used in high-pressure fuel pumps and injectors, but internal defects that lead to fuel leaks can occur in high-pressure fuel pumps and injectors. Visual inspection and helium leak testing have traditionally been used to inspect these internal defects, but these require an inspection process separate from the welding process, and there is a problem that internal defects cannot be detected by visual inspection. Therefore, there is a need for technology that uses a camera to inspect internal defects based on information obtained from molten pool images during laser welding without a separate inspection process.

[0020] Conventionally, a method is known in which the width of the molten pool is determined over time at a predetermined position in an area containing liquid metal at a predetermined distance from the laser center position, and the magnitude of change in the width of the molten pool over time is evaluated to determine the quality of the laser welding. Specifically, the difference between the maximum and minimum values ​​of the molten pool width within a predetermined period is calculated as the amount of fluctuation, and if the amount of fluctuation is within an appropriate range for good quality laser welding, the quality of the laser welding is determined to be good, and if it is outside the appropriate range, the quality of the laser welding is determined to be poor. In addition, although fumes, which are dust particles, are generated at the position where the laser light is irradiated, the infrared rays emitted by the fumes are small, so the influence of noise caused by the fumes is suppressed by visualizing the temperature distribution based on the infrared distribution using an infrared camera.

[0021] Here, internal defects in laser welding occur when the molten metal spatters and flies out, reducing the amount of molten metal. When the molten metal is reduced, the molten pool becomes shorter, and it is possible to detect this as a defect when the width behind the molten pool falls below a threshold. However, in the above method, based on the logic that the quality of laser welding is determined to be good when welding is stable, the difference between the maximum and minimum values ​​of the width of the molten pool at a specified distance from the center position of the laser is calculated as the amount of variation, and if it is within an appropriate range, the quality of laser welding is determined to be good.

[0022] Therefore, even if one tries to detect as a defect the occurrence of an internal defect that has reduced the minimum value of the molten pool width, if there is variation in the maximum value, the variation in the amount of fluctuation also increases, and the appropriate range also becomes larger, making it difficult to detect as a defect the reduction in the minimum value. Furthermore, there is a problem in that infrared cameras, which are used to suppress the effects of noise caused by fumes, are in less demand and more expensive than cameras that focus on the visible light range.

[0023] In contrast, in the first embodiment, attention is paid only to the molten pool width in the molten pool image, not to the amount of variation in the molten pool width at a predetermined distance from the laser center position, and when the molten pool width behind the molten pool becomes less than a threshold value, it is determined that there is an internal defect. As described above, the molten pool width behind the molten pool becomes 0 immediately after the occurrence of an internal defect. Therefore, by setting a threshold value close to 0 (e.g., 0.3) and determining that there is an internal defect when the molten pool width behind the molten pool (e.g., (9), (10)) falls below the threshold value, it is possible to detect internal defects in the weld bead that cannot be detected by appearance after welding, and to detect products that have problems such as fuel leakage due to internal defects. In addition, by determining that there is an internal defect only when the molten pool falls below the threshold value Z times in a row, the influence of short-term fluctuations in the molten pool width caused by fumes can be eliminated, improving the accuracy of detecting internal defects.

[0024] [Embodiment 2] The basic configuration of the second embodiment is similar to that of the first embodiment, so only the differences from the first embodiment will be described. As explained in the first embodiment, when fumes are generated, the width of the molten pool varies, which affects the inspection accuracy. Figure 14 shows images of the molten pool when a large fume is generated, (b) is an image when a large fume is generated, (a) is the image immediately before (b), and (c) is the image immediately after (b). Comparing the images, it can be seen that the shape of the molten pool 20 remains almost the same in all three images and changes slowly, whereas the shape of the fume 24 changes significantly in a short period of time.

[0025] Therefore, in embodiment 2, the effect of fumes is reduced by focusing on the change in brightness of each pixel in the molten pool image along the time axis and applying a low-pass filter (e.g., cutoff frequency 100 Hz) that passes only low-frequency changes and does not pass high-frequency changes. Figure 15 shows an example of (a) the brightness waveform over time and (b) the frequency distribution for a pixel in a molten pool image before applying a low-pass filter with a cutoff frequency of 100 Hz. Figure 16 shows an example of (a) the brightness waveform over time and (b) the frequency distribution for a pixel in a molten pool image after applying a low-pass filter with a cutoff frequency of 100 Hz.

[0026] Looking at the frequency distribution, it can be seen that before the low-pass filter was applied, the distribution was almost even up to over 300 Hz, whereas after the low-pass filter was applied, the distribution remained almost unchanged up to 100 Hz, and frequencies above 100 Hz were suppressed. Comparing the time waveforms, it can be seen that the brightness values ​​of 175 or more, which were abundant before the low-pass filter was applied, decreased after the low-pass filter was applied. Therefore, by applying a low-pass filter to the molten pool image as in embodiment 2, the influence of fumes is reduced, the measurement accuracy of the molten pool width is improved, and the detection accuracy of internal defects can be improved. In addition, because demand is low, it is possible to use inexpensive cameras in the visible light range instead of expensive infrared cameras, making it possible to reduce the cost of inspection equipment.

[0027] [Embodiment 3] The basic configuration of the third embodiment is similar to that of the first embodiment, so only the differences from the first embodiment will be described. The third embodiment differs from the first embodiment in that it is determined that there is an internal defect when the molten pool width (7) shown in a rectangular frame in FIG. 17 is equal to or greater than the upper threshold value shown in FIG. In the first embodiment, when the molten pool widths (9), (10) of the molten pool are equal to or smaller than a threshold value, it is determined that there is an internal defect. However, when the molten pool widths (9), (10) vary greatly, the width may be zero in places where there is no internal defect. This may result in an increase in false reports in which internal defects are determined to exist when there are no internal defects. In contrast, in the third embodiment, the presence or absence of an internal defect is judged by adding a condition that the molten pool width (7) is equal to or greater than the upper threshold value. This reduces the above-mentioned false alarms, thereby improving the detection accuracy of internal defects.

[0028] [Embodiment 4] FIG. 19 is a schematic diagram of a laser welding and inspection device according to the fourth embodiment. The laser welding and inspection device of the fourth embodiment is the same as the laser welding and inspection device of the first embodiment, except that mirror 13 is replaced with a partial wavelength transmitting mirror, and mirror 25, objective lens 26 for welding light sensor, welding light sensor 27, and inspection device 28 for welding light sensor are added. Partial wavelength transmitting mirror 13 sends light emitted from work 5 to welding light sensor 27 through mirror 25 and objective lens 26 for welding light sensor. Welding light sensor 27 separates the light into plasma light, infrared light, and reflected laser light and converts them into electrical signals. Inspection device 28 for welding light sensor processes the electrical signals and determines whether or not there is an internal defect.

[0029] Figure 20 is a diagram showing the judgment method for a single welding light sensor. The judgment range is set to a time excluding the part immediately after the start of welding that is prone to variation and the part at the end of the welding where the laser output is gradually reduced after one welding revolution is completed. The median value at each time of the signal value of the welding light sensor of several workpieces that were welded and had no internal defects is calculated, and values ​​that deviate above and below this by a specified percentage are calculated in advance as reference waveforms, and if the signal value of the welding light sensor deviates from the reference waveform by more than a specified percentage, it is judged to be an internal defect. For example, there is a portion slightly to the left of the center in Fig. 20 where the infrared light waveform jumps out below the reference waveform, and in such a case, it is determined to be a defect. Since the rate of false alarms may be high in the determination by the camera in the first to third embodiments, by determining that there is a defect when it is determined to be a defect by both the determination by the camera and the determination by the welding light sensor, the number of false alarms can be reduced.

[0030] The above describes an embodiment for carrying out the present invention, but the specific configuration of the present invention is not limited to the configuration of the embodiment, and design changes and the like that do not deviate from the gist of the invention are also included in the present invention. [Explanation of symbols]

[0031] 1...laser oscillator, 2...optical fiber for laser, 3...processing head, 4...collimation lens, 5...workpiece (material to be welded), 6...partially wavelength-transmitting mirror, 7...objective lens, 8...laser, 10...rotary spindle, 11...processing stage, 12...camera, 13...partially wavelength-transmitting mirror, 14...objective lens for camera, 15...image processing / inspection device, 16...laser welding control device, 25...mirror, 26...objective lens for welding light sensor, 27...welding light sensor, 28...inspection device using welding light sensor

Claims

1. A laser welding in-process inspection method comprising a step of photographing a molten pool formed by irradiating a workpiece with a laser beam, a width of the molten pool in a direction perpendicular to a welding direction is measured from the captured image of the molten pool, and if the width of the molten pool behind a predetermined portion in the welding direction of the molten pool is equal to or smaller than a first lower threshold or equal to or larger than a first upper threshold, it is determined that an internal defect exists; Laser weld in-process inspection method.

2. 2. The laser weld in-process inspection method according to claim 1, detecting a width of the molten pool behind the molten pool based on an image obtained by applying a low-pass filter in the time axis direction to the brightness of each pixel of an image of the molten pool taken by a camera during the laser welding; Laser weld in-process inspection method.

3. 3. The laser welding in-process inspection method according to claim 1, further comprising: determining that there is an internal defect when the width of the molten pool at a position rearward of a predetermined position in the welding direction of the photographed molten pool is equal to or smaller than the first lower threshold or equal to or larger than the first upper threshold, and the signal of the welding light sensor is equal to or smaller than the second lower threshold or equal to or larger than the second upper threshold; Laser weld in-process inspection method.

Images