Method and processing machine for monitoring porosity defects in laser welding process

Abstract

The present invention relates to a method for monitoring porosity defects in a laser welding process for welding two workpieces made of a metallic material by means of a laser beam. The method is particularly used for monitoring a plurality of identical laser welding processes for welding two identical workpieces, each with the same laser power of the laser beam and the same welding duration. When welding the two workpieces, the laser beam is directed at adjacent end faces of the workpieces in order to melt a molten bead on the two end faces, which then solidifies into a weld bead. According to the invention, when welding the two workpieces, the solidification duration from switching off the laser beam until the solidification of the molten bead is determined, the determined solidification duration is compared with a desired solidification duration predefined for a weld without porosity defects, and the solidified weld bead is classified as defective if the determined solidification duration is less than the predefined desired solidification duration.

Description

Technical Field

[0001] The present invention relates to a method for monitoring a laser welding process for welding two workpieces, preferably rod-shaped conductors, made of a metallic material, in particular copper or aluminum, by means of a laser beam. The method is particularly useful for monitoring a plurality of identical laser welding processes, each for welding two identical workpieces with the same laser beam power and the same welding duration. During the welding of the two workpieces, the laser beam is directed at adjacent end faces of the workpieces to melt a molten bead on the two end faces, which subsequently solidifies into a weld bead. Preferably, the workpiece end faces at which the machining laser beam is directed are arranged adjacent to each other at the same height. The present invention also relates to a machining machine suitable for carrying out the method and an associated computer program product. Background Art

[0002] A typical defect in laser welding is the formation of pores, which impair the weld's performance. It's usually not possible to detect the formation of pores in the weld seam or weld bead from the outside. Defective connections can only be verified retrospectively through destructive testing or computed tomography (CT) or X-ray imaging. Therefore, visual inspections by workers or periodic analysis and evaluation of samples using CT or X-ray imaging are typically performed.

[0003] Copper-containing, bent rod-shaped conductors, particularly so-called hairpins, are installed in electric power machines, such as electric motors or generators. The rod-shaped conductors are arranged according to the provided electrical wiring and welded together to form an electromagnet. Electric motors typically have dozens, often hundreds, of bent rod-shaped conductors, which must be welded together in pairs. It is crucial that the welds provide a sufficient cross-sectional area ("attachment surface") through which current can flow from one rod-shaped conductor to the other. If the attachment surface is too small, significant resistance heating, efficiency losses, or even unusable electric power machines can occur during operation.

[0004] Rod-shaped conductors are welded using a laser beam, which is typically directed at the end faces of two adjacent, usually adjacent, rod-shaped conductors. These end faces melt due to the introduced heat and, after solidification, are connected to each other via a resolidified weld bead. The laser beam is usually directed at the rod-shaped conductors at the same power and at the same time, thus achieving a sufficiently large bonding surface.

[0005] However, due to dirt or roughness on the surface of the rod-shaped conductor, the reflectivity of the rod-shaped conductor to the laser beam can fluctuate, and thus the actual energy input can also fluctuate. Similarly, the actual energy input can vary due to incorrect positioning of the rod-shaped conductor, such as gaps or offsets, or due to inaccurate positioning of the laser beam. If the energy input is too low, too little material is melted, resulting in a weld bead that is too small and provides an insufficient attachment surface. If intense spatter forms during laser welding, this can also result in an excessively small weld bead with an insufficient attachment surface.

[0006] In particular, when welding rod-shaped conductors to a single component (e.g., a motor stator) using hundreds of welds, welding defects due to porosity formation are statistically problematic. Large porosities, in particular, often occur when large amounts of melt are ejected from the process area in the form of spatter, subsequently disrupting the flow of current in the rod-shaped conductor. Due to the large number of welds on a single component, visual inspection and analytical evaluation of samples are insufficient to ensure reliable production with low scrap rates. Therefore, for example, if a stator with 500 welds has a functional failure rate of, say, one in 100,000, only one defective weld is acceptable in 50 million. Therefore, 100% inspection and monitoring of the welding process is essential.

[0007] For example, DE 10 2004 016 669 B3 discloses a method for inspecting weld seams introduced into one or more workpieces by means of laser beam welding. Characteristic signals are received from the region of the weld seam by means of a sensor and compared with setpoint values, with only signals received within a characteristic time interval after laser beam welding, beginning at the earliest after solidification of the weld seam, being considered. Summary of the Invention

[0008] The object of the present invention is to provide a method for monitoring a laser welding process, in particular a plurality of identical laser welding processes, each for welding two workpieces, which method can be performed simply, quickly and non-destructively.

[0009] This object is achieved according to the invention in the monitoring method described at the outset by determining the solidification duration from the switching off of the laser beam until the solidification of the weld bead when welding two workpieces, comparing the determined solidification duration with a predetermined expected solidification duration for a weld without porosity defects, and classifying the solidified weld bead as defective ("porosity defect present") if the determined solidification duration is less than the predetermined expected solidification duration.

[0010] The welding duration for a given end cross-section of the two rod-shaped conductors to be connected depends on the available or selected welding power. If the rod cross-section and welding power are defined, a defined melt volume results in a porosity-free weld. Since the laser action time, i.e., the welding time, does not change with repeated welds in practice, all welds have the same duration. This achieves an accuracy of a few milliseconds, which typically corresponds to a time deviation of less than 1%. This approach ensures that the energy content of all welds is the same. Since the mass and end cross-section of the rod-shaped conductors to be connected also experience only minor variations, as the molten weld zone cools, reproducible and precise characteristics are achieved with a consistently identical solidification duration, starting after the laser is switched off.

[0011] If melt is expelled from the weld zone during welding, the expelled melt affects the cooling process, resulting in energy losses in the weld zone along with mass loss. In particular, large pores often occur when large amounts of melt are ejected from the process zone in the form of spatter, which subsequently interfere with the flow of current in the rod-shaped conductor. As a result, during solidification, less energy from the weld zone flows into the heat sink of the rod-shaped conductor, so that the solidification temperature is reached more quickly. According to the present invention, the solidification duration starting after the laser is switched off is evaluated, and if the behavior deviates from the expected solidification curve, the qualitative state of the weld (presence / absence or only insignificant presence of pore defects) is inferred.

[0012] Particularly preferably, starting from when the laser beam is switched off, a detector, in particular a camera, continuously records spatially resolved digital images of the molten bead, and intensity-graded pixel images, in particular grayscale pixel images, are generated from the spatially resolved detector images. The information about whether the melt or the molten bead is still in a liquid state is contained in the scale of the intensity or grayscale values of the individual images. As the temperature of the molten bead decreases, the grayscale values change from light to dark. These images are preferably evaluated in each image section, in particular in an annular image section surrounding the center point of the molten bead.

[0013] In a preferred method variant, an intensity level value averaged over all pixels of each pixel image is determined for each pixel image, and the duration of solidification is determined based on the time profile of this averaged intensity level value. Preferably, the detector images are each evaluated only in a region of the recorded image ("region of interest" (ROI)). The melted region or bead is observed, for example, by the detector, and the development of the intensity values in the so-called "region of interest" (ROI) is evaluated starting from the time "laser off". An algorithm evaluates how long the cooling process lasts in time increments corresponding to the detector frame rate.

[0014] Preferably, the detector images are recorded as a process video with a recording frequency of at least 100 Hz, in particular at least 1 kHz.

[0015] In the case of weld beads classified as defective ("porosity defect present"), these weld beads are automatically rewelded or other actions, in particular warning notifications, are triggered, preferably depending on how much the determined solidification duration falls below a predetermined threshold value.

[0016] The present invention also relates to a processing machine for laser welding two workpieces, preferably rod-shaped conductors, made of a metallic material, in particular copper or aluminum, comprising: a laser beam generator for generating a laser beam; processing optics for directing the laser beam onto adjacent end faces of the two workpieces to melt a bead on the two end faces, which subsequently solidifies into a weld bead; a spatially resolved detector for spatially resolved detection of the weld bead; an image processing unit for evaluating spatially resolved detector images captured by the detector to determine the solidification duration from switching off the laser beam until the weld bead solidifies; and a porosity monitoring device for monitoring or classifying the solidified weld bead for porosity defects based on the determined solidification duration. The detector is advantageously directed coaxially with the laser beam toward the end faces of the workpieces.

[0017] Preferably, the image processing unit has: an intensity-level pixel image generating device for generating an intensity-level pixel image from the captured detector image; and an analysis and evaluation device for analyzing and evaluating the intensity-level pixel image in order to determine the solidification duration from switching off the laser beam until solidification of the weld bead.

[0018] Finally, the present invention also relates to a computer program product having a code medium adapted to carry out all steps of the method according to the present invention when the program is run on a machine control of a processing machine.

[0019] Further advantages and advantageous configurations of the subject matter of the present invention can be derived from the description and the drawings. Likewise, the features mentioned above and those yet to be listed can be used individually or in any desired combination. The embodiments shown and described are not to be understood as a definitive enumeration, but rather as exemplary features for describing the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings show:

[0021] Figure 1 A schematic diagram of a processing machine according to the invention for laser welding two rod-shaped conductors by means of a laser beam;

[0022] Figures 2a-2cThe molten bead generated when laser welding two rod-shaped conductors on their end faces is turned off after the laser beam is turned off ( Figure 2a ), during solidification ( Figure 2b ) and after solidification ( Figure 2c ) images;

[0023] Figure 3 An image of a liquid bead having a ring-shaped image portion surrounding a center point of the bead; and

[0024] 4a, 4b show the time profile of the radiation intensity of the thermal radiation emitted by the molten bead (FIG. 4a) and the time profile of the grayscale values of the image (FIG. 4b), respectively averaged over predetermined image pixels of the recorded image. DETAILED DESCRIPTION

[0025] exist Figure 1 The processing machine 1 schematically shown in FIG is used for laser welding two workpieces made of metal material, which are here exemplarily in the form of two bent copper rod-shaped conductors 2 ("hairpins"), by means of a laser beam 3. The two rod-shaped conductors 2 have identical end faces 4 to be welded with the same cross-section and are arranged side by side with their end faces 4 preferably at the same height.

[0026] The processing machine 1 comprises a laser beam generator 5 for generating a laser beam 3; a processing head 6 having a processing lens assembly 7 for directing the laser beam 3 onto the end faces 4 of two rod-shaped conductors 2 in order to melt a molten bead or a molten zone 8 on the end faces 4; a position-resolving detector, for example in the form of a camera 9, directed at the molten bead 8; an image processing unit 10 for evaluating a digital image recorded by the camera 9 in a position-resolved manner; and a monitoring device 11, which monitors the molten bead 8 solidified into a weld bead 8' with respect to pore formation based on the evaluated camera image.

[0027] The laser beam 3 generated by the laser beam generator 5 strikes a beam splitter 12 (e.g., in the form of a dichroic mirror), which is reflective for the wavelength of the laser beam 3. From the beam splitter 12, the laser beam 3 is reflected by a focusing device (e.g., a focusing lens) (not shown here) onto the processing optics 7 and, from there, is directed onto the two end faces 4. The processing optics 7 can be, for example, a laser scanner having two mirrors that can each be rotated about axes at right angles to each other in order to deflect the laser beam 3 two-dimensionally.

[0028] An image beam 13 emanating from the molten bead 8 is detected by a camera 9. This image beam travels via the processing lens system 7, a beam splitter 12 transparent to the image beam 13, and a further beam splitter 14 reflective to the image beam 13 (e.g., in the form of a dichroic mirror) to the camera 9, where it forms an image of the molten bead 8. As shown, the camera 9 is aligned coaxially with the laser beam 3 by means of the further beam splitter 14. An optical filter 15 and a collimating lens 16 for focusing the image beam 13 are optionally arranged between the further beam splitter 14 and the camera 9. The optical filter 15 blocks the wavelength of the laser beam 3, thereby allowing only the process radiation emitted from the molten bead 8 to pass through, while preventing the laser beam 3 reflected on the workpiece 2 from passing through. The camera 9 can be designed to record a single image or as a video camera for recording a video sequence, with the recording frequency preferably being at least 100 Hz.

[0029] A laser welding process, in particular a plurality of identical laser welding processes carried out on respectively two identical workpieces 2 with the same laser power of the laser beam 3 and the same welding duration, is monitored in the following manner.

[0030] After the workpiece has been processed, that is to say starting from the time the laser beam 3 is switched off, the images 17a-17c of the molten bead 8 are recorded continuously by means of the camera 9. Figures 2a-2c ), wherein the molten bead 8 appears bright in the recorded images 17a-17c. FIG17a shows the liquid molten bead 8 immediately after the laser beam 3 is switched off and before solidification, FIG17b shows the liquid molten bead 17b during solidification, and FIG17c shows the solidified molten bead 8'.

[0031] In the grayscale pixel image generating device 10a of the image processing device 10, grayscale pixel images with pixel values between 0 (dark) and 255 (bright) in the xy pixel grid are generated from the captured images 17a-17c. As the temperature of the molten bead 8 gradually decreases, the grayscale value changes from bright to dark. In these pixel images, the image processing device 10 defines a same image part ("region of interest" (ROI)) ( Figure 3 ), for example in the form of a ring image section around the center point M of the molten bead 8. The pixel image in the region of interest 18 is evaluated in an evaluation device 10b of the image processing device 10 in order to determine the solidification duration Δt from the time the laser beam 3 is switched off until the solidification of the molten bead 8.

[0032] FIG4 a shows the time curve of the radiation intensity I of the thermal radiation emitted by the molten bead 8 after the laser beam 3 is switched off at time t=0. The radiation intensity I decreases after the laser beam 3 is switched off and decreases from a few milliseconds before the solidification of the molten bead 8 until solidification (time t E) remains at a plateau value and then drops to zero.

[0033] After the laser beam 3 is switched off, the grayscale value profile is detected in a time-resolved manner within the spatially averaged region of interest 18. For this purpose, the evaluation device 10b determines the grayscale value averaged over all pixels in the region of interest 18 for each pixel image. And analyze and evaluate the average gray value shown in Figure 4b The time variation curve of The characteristic time variation curve can clearly determine the time point of solidification t E The coagulation duration Δt is thus determined.

[0034] In repeated identical laser welding processes, in which the laser beam 3 always has the same laser power and the same welding duration, in the case of porosity-free welding, all welding events will have the same solidification duration Δt, respectively on two identical rod-shaped conductors 2 with the same end face 4 and the same rod cross section.

[0035] The solidification time Δt thus determined is compared in the monitoring device 11 with the desired solidification time Δt specified for a weld bead 8 without porosity defects. S When the determined coagulation duration Δt is lower than a predetermined threshold value Δt S (Δt<Δt S ), the solidified weld bead 8' is classified as defective ("porosity defect present"). In the case of excessive deviations, automatic re-welding can be initiated or any other action can be triggered.

[0036] To illuminate the molten bead 8, the processing machine 1 can include an illuminating laser 20, whose illuminating beam 21 is coupled coaxially with the laser beam 3 into the processing head 6 through two beam splitters 12, 14 that are transmissive in this direction for the wavelength of the illuminating beam 21 and directed toward the molten bead 8. The illuminating beam 21 reflected by the workpiece 2 returns in the opposite direction to the further beam splitter 14, which is reflective in this direction and deflects the illuminating beam 21 toward the camera 9. In this case, the molten bead 8 appears dark in the recorded image, and the illuminated solid material appears bright.

Claims

1. A method for monitoring porosity defects in a laser welding process for welding two workpieces (2) made of a metallic material by means of a laser beam (3), the method for monitoring a plurality of identical laser welding processes for welding two identical workpieces (2) with the same laser power of the laser beam (3) and the same welding duration, wherein: The two identical workpieces are two identical rod-shaped conductors having identical end faces and identical rod cross-sections, wherein, when welding the two workpieces (2), the laser beam (3) is directed onto the adjacent end faces (4) of the workpieces (2) in order to melt a molten bead (8) on the two end faces (4), which then solidifies into a weld bead (8'), characterized in that When welding two workpieces (2), a solidification time (Δt) is determined from the time the laser beam (3) is switched off until the solidification of the molten bead (8), and the determined solidification time (Δt) is compared with a desired solidification time (Δt) predetermined for a weld without porosity defects. S ) is compared, and when the determined coagulation duration (Δt) is less than the predetermined coagulation duration (Δt S ), the solidified weld bead (8') is classified as having porosity defects.

2. The method according to claim 1, characterized in that Starting from the switching off of the laser beam (3), spatially resolved detector images (17a-17c) of the melt bead (8) are continuously recorded by means of a detector (9), and intensity-grade pixel images are generated from the spatially resolved detector images (17a-17c).

3. The method according to claim 2, characterized in that For each pixel image, an intensity level value averaged over all pixels of the pixel image or over a determined image part (18) of the pixel image is obtained. And based on the averaged intensity level value The coagulation duration (Δt) is determined based on the time variation curve of 4. The method according to claim 2 or 3, characterized in that The detector images (17a-17c) are acquired at an acquisition frequency of at least 100 Hz.

5. The method according to claim 2 or 3, characterized in that The detector images (17a-17c) are each evaluated in an image section (18).

6. The method according to any one of claims 1 to 3, characterized in that In the case of a weld bead (8') classified as defective, the weld bead (8') is automatically re-welded or other actions are triggered.

7. The method according to claim 6, characterized in that According to the determined coagulation time (Δt) and the predetermined expected coagulation time (Δt S ) how much deviation to trigger the re-welding or other action.

8. The method according to claim 1, characterized in that The metal material is copper or aluminum.

9. The method according to claim 2, characterized in that The detector (9) is a camera.

10. The method according to claim 2, characterized in that The intensity level pixel image is a grayscale pixel image.

11. The method according to claim 4, characterized in that The detector images (17a-17c) are acquired at an acquisition frequency of at least 1 kHz.

12. The method according to claim 5, characterized in that The image portion (18) is a ring-shaped image portion surrounding the center point (M) of the molten bead (8).

13. The method according to claim 6, characterized in that The other action is a warning notification.

14. A processing machine (1) for laser welding two workpieces (2) made of metal material, the processing machine being configured to perform the method according to any one of claims 1 to 13, the processing machine comprising: a laser beam generator for generating a laser beam (3); a processing lens assembly (7) for directing the laser beam (3) onto the adjacent end faces (4) of two workpieces (2) in order to melt a molten bead (8) on the two end faces (4) which subsequently solidifies into a welding bead (8'); a position-resolved detector (9) for detecting the molten bead (8) in a position-resolved manner; an image processing unit (10) for evaluating spatially resolved detector images (17a-17c) recorded by the detector (9) in order to determine a solidification duration (Δt) from switching off the laser beam (3) until solidification of the bead (8); as well as A porosity defect monitoring device (11) monitors or classifies a solidified weld bead (8') with respect to porosity defects based on a determined solidification duration (Δt).

15. The processing machine according to claim 14, characterized in that The image processing unit (10) has: - an intensity-level pixel image generating device (10a) for generating an intensity-level pixel image from the captured detector images (17a-17c); and An evaluation device (10b) for evaluating the intensity level pixel image in order to determine a solidification duration (Δt) from the time the laser beam (3) is switched off until the solidification of the molten bead (8).

16. The processing machine according to claim 15, characterized in that The detector (9) is arranged coaxially with the laser beam (3).

17. A computer program carrier having a program stored thereon, the program having a code medium adapted to execute all steps of the method according to any one of claims 1 to 13 when the program is run on a machine control of a processing machine (1).

Images