Method and monitoring device for monitoring laser beam welding process

Abstract

The invention relates to a method for monitoring a laser beam welding process, in which optical signals of at least a region of a liquid melt, in particular of at least a region of a steam capillary (14), are recorded and signals (24, 25) of different optical polarization directions are generated from the optical signals (29), characterized in that the signals (24, 25) of the different optical polarization directions are transmitted to the laser beam welding process. The brightness of the optical signal (24, 25) is evaluated taking into account reference information in such a way that a disturbance to the laser beam welding process is ascertained at a predetermined deviation from the reference information.

Description

Technical Field

[0001] The invention relates to a method for monitoring a laser beam welding process. Background Art

[0002] From the article "Adaptation of topographic measurement systems to beam welding processes with high spatiotemporal resolution" (Journal of Advanced Joining Technology, Issue 1, 2020, Reisgen et al.), a measuring system for electron beam welding systems is known, in which an image of the process zone is recorded with the aid of a goniometer and decomposed along different polarization directions. A high-speed camera records image data of the weld pool along four different polarization directions of the light and an interpretation of the electron beam welding process is made possible by evaluating the image information. Summary of the invention

[0003] In contrast, the method according to the invention for monitoring a laser beam welding process has the advantage that the brightness of the beams of different polarization directions recorded from the observed working area of ​​the laser beam welding process, in particular from the steam capillary of the welding process, is evaluated. Compared to processing the recorded images, evaluating only the brightness has the advantage that only a much smaller amount of data has to be provided and processed for the evaluation. However, by comparing with already existing reference information, it is also possible to determine anomalies in the laser beam welding process and to identify disturbances or errors in the welding process.

[0004] Since the main concern in process monitoring is not to measure and characterize the steam capillaries precisely in terms of geometry, but rather to detect deviations or anomalies. It has been shown that the same result can be achieved without evaluating the camera images, but rather with a simpler evaluation. Thus, a fast and surprisingly sufficiently accurate evaluation can be achieved with a small amount of data.

[0005] Compared to a pure evaluation of brightness when observing the steam capillary, the advantage of evaluating the beams separated by polarization direction is that changes in the surface affect the optical properties and in particular the intensity of the reflected beams for each polarization direction more strongly than the addition of the entire brightness for all polarization directions. In this regard, it is much easier to determine the deviation from the reference value by evaluating the brightness for each polarization direction when the laser beam welding process is disturbed. But especially compared to image evaluation by a high-speed camera, the brightness information in the recorded measurement signal can be compared with the stored reference information with much less computing power. Therefore, in the laser welding process, especially also at high feed speeds, such as when welding bipolar plates of fuel cells or when welding copper parts of electric motors and power electronics, abnormalities in the welding process can be reliably identified and eliminated in time by introducing corresponding measures when necessary. Therefore, it is particularly possible to identify changes in the flow behavior of the molten pool, fluctuations in the steam capillary, molten pool splashes or splashes.

[0006] By evaluating different polarization directions, errors can also be reliably eliminated by scattering, which often occurs during such laser welding. Therefore, for example, it is also possible to identify steam capillary deformation when the melt is partially or completely ejected from the steam capillary when evaluating different polarization directions, while the identification of interference may not be completed due to the hot steam capillary wall when evaluating the complete brightness of all polarization directions. It is possible to evaluate the surface while taking into account the influence of the cutting front angle or other surface topology along with the brightness evaluation for each polarization direction. This detection of welding errors is based on the premise that the capillary shape must also be changed in advance in each weld defect that should and can be detected. Because energy is introduced into the component through the steam capillary. If too little or too much energy is introduced, the steam capillary changes with such an effect that the welding process and the melt flow are disturbed and defects such as splashes, splashes or process pores appear as a result.

[0007] Typical welding errors can be detected with the device according to the invention, because in these welding errors the steam capillary geometry is significantly changed. This is, for example, splashing, in which large molten areas are ejected from the molten pool during the process. This can leave holes or defects in the seam. In addition, splashing is also possible, which separates from the steam capillary and in particular the rear wall. This can contaminate components or the environment or lead to missing melts in the seam. In addition, insufficient connections between gaps can also be detected when gaps between the sheets are not closed in time and thus lead to insufficient connections or no connections. In addition, residual gap heights on the weld seam can also be detected. or insufficient arch They may occur at high welding speeds. In particular, fluctuations in the geometry of the steam capillary and in the flow conditions in the molten metal tail may occur. Welding depth fluctuations may also be detected, in which a decrease or increase in welding depth is associated with fluctuations in the geometry of the steam capillary. If there are interruptions in the welding process or interruptions in the process, such errors can lead to significant changes in the geometry of the steam capillary. In the case of changes in the geometry of the steam capillary, this is particularly visible in different polarization directions and leads to correspondingly different changes in intensity.

[0008] Further advantages are derived from the dependent claims. It is therefore advantageous to evaluate exactly two signals of different polarization directions, since this allows reliable statements on the one hand and limits the computational effort for the evaluation on the other hand. It is particularly advantageous here to select further different polarization directions orthogonally to one another in order to obtain a particularly reliable evaluation.

[0009] It is also advantageous to evaluate signals in the wavelength range between 400 nm and 2000 nm. In this wavelength range, particularly appropriate influences on the melting process by interference occur. The wavelength range between 400 nm and 1100 nm is particularly advantageous for evaluation, since sensitive, fast and powerful silicon diodes can be used in this range.

[0010] It is also advantageous if the signals are respectively detected by a photodiode. Photodiodes have the advantage that they are firstly cost-effective, can be easily adapted to the wavelength range to be observed by selecting the diode material, are fast and sensitive and can reliably provide an integrated signal from the laser interaction region.

[0011] It is also advantageous that the signals are combined with one another in a single or multiple calculation for the evaluation and the combination is evaluated with respect to the reference information. This makes it possible to reduce or at least minimize interference and measurement errors, if necessary. In particular, errors that are believed to be true due to natural scattering can be more easily excluded.

[0012] It is also advantageous to first carry out a trajectory on the target for a laser beam welding process and to empirically determine reference information from the recorded photographs for cases where the laser beam welding process was carried out as intended and to store it in order to use this recorded information to calibrate the monitoring device. Advantageously, use is made of the fact that the recorded signals are differentiated according to the direction of movement of the laser on the target and that the trajectory of the welding process thus influences the recorded intensity to be expected. A corresponding movement-dependent calibration can thus improve the reliability of the assessment of the laser beam welding process.

[0013] It is also advantageous that, when it is found that the laser beam welding process is disturbed based on the deviation determined, the laser power for the laser beam welding process is directly adjusted. By means of this adjustment, the range of the disturbing part can be limited, so that the disturbing part or the welding error remains unimportant for the welding process itself. Therefore, when the material is thrown out of the process area too strongly, the laser power can be reduced in particular. This avoids the process change only when a large temperature increase occurs in the molten pool or the splashes have already flown away. However, the first change in the capillary shape can already be detected with the current method, that is, changes in the process or disturbing processes and energy couplings occur before errors such as splashes. This more timely detection of anomalies opens up an effective feasible solution for regulating the process without welding errors.

[0014] It is also advantageous if the reference information is provided by a machine-learned neural network, so that deviations of the intensity curve can be reliably detected in a well-learned neural network.

[0015] Corresponding advantages result for a monitoring device for carrying out the method according to the invention.

[0016] In another specific embodiment, it is also possible to use a camera for detecting brightness, which camera is optionally also designed in addition to recording image information, for example for compilation.

[0017] It is also advantageous if, in addition to the laser provided for the laser beam welding process, an additional laser is provided which serves as a measuring laser beam, since this laser can provide a measuring laser beam having a wavelength different from that of the working laser beam and thus a particularly advantageous wavelength range for monitoring can be provided by suitable wavelength selection of the measuring laser and influences due to the processing laser can be excluded.

[0018] It is advantageous here that the additional laser is circularly polarized and the reflected back portion of the laser in at least two polarization directions is evaluated. In order to improve the evaluation, further polarization directions (linear combinations of these polarization directions) can also be evaluated here. Alternatively, two additional lasers with different polarization directions and / or wavelengths can also be used. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Exemplary embodiments of the invention are illustrated in the drawings and are explained in detail in the following description.

[0020] Figure 1 A monitoring device according to the invention is shown with a laser for carrying out a laser beam welding process;

[0021] Figure 2 Shows Figure 1An exemplary optical structure of a monitoring device;

[0022] Figure 3a and 3b as well as Figure 4a and 4b Two examples of exemplary recorded reference information for two different trajectory profiles of a laser during a laser welding process are shown;

[0023] Figure 5 ] is an example of an intensity profile of a brightness profile for one polarization direction recorded with an exemplary illustrated disturbance of a laser beam welding process. DETAILED DESCRIPTION

[0024] exist Figure 1 1 shows a laser 10 having a laser unit 11 which radiates laser light 12 onto a material 13. The laser light strikes the material 13 in the region of a so-called steam capillary 14, wherein a melting zone 15 with a higher melting point is formed around the steam capillary 14. The material 13 can thus be welded to the underlying material 16 in the region of the steam capillary 14. The laser unit 11 is controlled by a control device 17 in such a way that the laser beam 12 moves in the direction of the arrow 18 over the surface of the material 19.

[0025] Furthermore, a monitoring device 20 is provided, the optical system 21 of which is directed to the surface 19 of the material 13, wherein the monitoring area 22 is oriented so that it includes the surface area of ​​the melting area 15 and in particular the steam capillary 14. In another embodiment, the monitoring area 22 can be oriented so that it is directed only to the surface of the steam capillary 14. From the optical system unit 21, the received light beam is guided to a beam splitter 23, wherein the received light beam 29 is split into a first light beam 24 with a first polarization direction and a second light beam 25 with a second polarization direction different from the first polarization direction. In one embodiment, the two polarization directions are rotated by 90° and are therefore orthogonal to each other. In other embodiments, it is also possible to select a further division of the polarization directions, for example 45° and 90° and 0°, and thus generate more than two light beams with corresponding polarization directions. The light beams 24, 25 are respectively guided to a first detection unit 26 or a second detection unit 27, which are respectively configured as intensity measuring devices for the light intensity of the light beams 24, 25. The measured intensities are transmitted via the data connection 28 to an evaluation device 30 , in which a computing unit 31 compares the received data with reference information stored in a memory 32 .

[0026] The receiving units 26, 27 are preferably designed as photodiodes or cameras. Photodiodes are more advantageous here and also allow good determination of the brightness. When using a camera, intensity values ​​for different, preferably non-overlapping image areas can be determined in a simple manner. The photodiode can be selected to be adapted to the corresponding wavelength (depending on the radiation used in the process) or to the radiation of the measuring laser, in particular by selecting the diode material. Thus, for example, silicon photodiodes, indium photodiodes, gallium photodiodes, quinoline (also sulfuric acid) photodiodes or germanium photodiodes can be used.

[0027] When selecting different regions, the focus can be on the steam capillary. However, depending on the process, it may also be advantageous to also observe the edge region of the steam capillary or to observe only parts of the steam capillary.

[0028] In order to select the corresponding reference information, the control unit 17 of the laser 10 can transmit information about the intensity of the laser beam 12 and / or the direction of movement 18 of the laser to the evaluation unit 30 via the second data connection 33. If the calculation unit 31 determines that the intensity value received by the monitoring unit 20 deviates from the reference information by more than a predetermined degree, a warning is issued in a first embodiment via the output unit 34. Thus, an erroneous laser welded connection can be recorded and the error can be corrected manually if necessary or the corresponding component can be sorted out as completely faulty or sorted out for manual rechecking. In this case, the welding process can either be terminated first or, alternatively, also be interrupted directly in the event of an error.

[0029] In another embodiment, the evaluation unit 30 transmits corresponding information about deviations, for example information about insufficient intensity due to evaporation of molten material from the vapor capillary 14 due to excessive laser power, via the second data connection 33. The information about the excessive laser power is received by the control unit 17 of the laser 10 and causes the control unit to reduce the laser power by, for example, 10%. If necessary, in a subsequent action, an automatic repetition of the welding process can be prompted at the previously scanned location.

[0030] In a first embodiment, the values ​​for different polarization directions are compared separately from one another with reference information for the corresponding polarization directions. When one of the two intensity values ​​deviates from the expected reference value by more than a predetermined degree, a faulty laser welded connection can be determined in the first embodiment. However, in another embodiment, it is also possible to predict an error when both values ​​deviate from a predetermined reference value.

[0031] In another embodiment, it is also conceivable that the corresponding intensity values ​​do not appear constant in the laser welding program, but that a scattering also occurs when the welding process is carried out as specified. When the limit values ​​are fixedly predetermined, this can lead to either the critical event disappearing in the scattering or a false alarm being generated in the case of too narrow a setting. For improvement, a machine-learned neural network is preferably provided in the memory 32, in which not only the absolute values ​​are used as reference information, but also the corresponding pattern of the intensity measurement is compared with a predetermined pattern. If a deviation from the predetermined pattern occurs here, an error in the welding process is confirmed.

[0032] The reference information can be determined computationally on the one hand, or advantageously experimentally on the other hand, in particular with regard to the respective material to be welded.

[0033] In another embodiment, different intensity values ​​can be evaluated in combination by calculation alternatively or in addition. In one embodiment, the quotient of two intensities can be determined, for example, and compared with a reference value for the quotient. It is possible here that interference can be reduced if necessary and huge deviations can be better identified. Other mathematical functions can also be used for merging signals or for standardization. The corresponding function is determined experimentally for the application purpose of the corresponding material and laser welding process if necessary. Therefore, for example, it may be advantageous to utilize the sum of the intensities. In addition, it is also possible to determine and further use the intensity of the first polarization direction, for example, by using the sum of the intensities of the two polarization directions as a combined signal. It is also possible in another embodiment to divide the intensity signal by the laser power so as to obtain a standardized signal. False alarms can be avoided here, that is, when more than two or more values ​​determined by calculation in this way if necessary, the deviation is determined compared to the corresponding reference value.

[0034] The reference value is preferably determined in this way, that is, different laser processes are carried out, which are classified as optimal or acceptable. The intensity measured in these laser processes is used as reference information for judging further processes. During the evaluation, deviations or anomalies from the expected signal are then determined. This intensity deviation indicates that the steam capillary or the molten pool does not correspond to the reference at the corresponding location of the laser welding process. Therefore, molten pool splashes and other unstable situations in the welding process can be detected as early as possible and eliminated in time if necessary in the following way, that is, the laser and in particular the laser intensity are controlled accordingly. In addition, it is also possible to adjust the beam profile of the laser accordingly or optimize the focus on the welding point.

[0035] In a laser process that follows a geometric pattern such as a curve, rectangle or other geometric shape, such as a wave shape, as a trajectory curve on the material, the orientation of the steam capillary and the corresponding components of the light polarized in different directions starting from the steam capillary changes due to the trajectory movement of the laser beam above the component. In this regard, the intensity of the light for different polarization directions depends on the direction in which the laser is guided above the surface of the material to be processed. In a first embodiment, this correlation can be ignored for evaluation. But it is also possible in an extended design scheme to take this correlation into account. Therefore, in order to better detect errors for the corresponding welding geometry to be performed, calibration is performed. In a first embodiment, the correlation of the light components polarized in different places with the motion vector starting from the origin of the coordinate system is determined. For each motion vector or for the amount of the selected motion vector, the expected signal for different polarization directions can be determined here. In another embodiment, the pattern determined by the laser moves on the material and some portions of the intensity are determined here.

[0036] An exemplary embodiment for this is shown in FIG3A . With reference to a virtual coordinate system 40 , a circular track 42 is moved by the laser at a corresponding angle 41 on the surface. Figure 3b In this case, both the intensity for a first polarization direction 44 and the intensity for a second polarization direction 45 are shown on a y-axis 43 via an angle 41 of the laser along a circular path 42 , which is plotted on an x-axis 46 .

[0037] This determination of the intensity can be performed additionally or alternatively for different feed speeds or different laser powers, ie, calibrated as it were. This is advantageous if the track geometry is processed in this context with partially different parameters.

[0038] 4A shows a second embodiment for a rectangular annular track 50, in which the laser writes a rectangular pattern on the surface of the material, again by means of a virtual cross 40. On the Y axis 43, a first intensity 51 and a second intensity 52 are shown relative to a position 53 on the rectangular track 50 shown on the X axis 56, wherein the position 54 is selected as an example.

[0039] exist Figure 2Detailed description of the optical path and analysis of the laser is shown in FIG. The light of the laser 10 is parallelized by the lens 60, crosses the semi-transparent mirror 61 and hits the focusing lens 62, which focuses the laser light onto the steam capillary 14. The light emitted from the steam capillary 14 reaches the lens 62 and the transparent mirror 61 again, which now deflects the corresponding light into the optical path 63 for evaluation. Here, a filter 64 can be introduced into the optical path 63 in order to limit the light used for evaluation to the desired wavelength range. In addition, the filter can also contain an attenuation structure, which reduces the correspondingly high intensity of the received light in order to avoid damage to the subsequent evaluation unit or so as not to over-control the sensor. The light then hits the beam splitter 65, through which an optical path 66 with a first polarization direction and an optical path 67 with a second polarization direction are generated. The light is bundled to the first evaluation unit 26 and the second evaluation unit 27 through corresponding lenses 68, 69.

[0040] If necessary, further filters 70, 72 can be introduced into the optical path 66, 67. In an alternative embodiment, the second evaluation unit 27 is omitted at the initially specified position. The light beam 67 is then first directed to the second mirror 70, through which the light reaches the convex lens 69' and the second evaluation unit 27'. The second mirror 70 is designed to be semi-transparent, wherein, on the side of the laser unit 71 facing away from the beam splitter 65, the laser light passes through the semi-transparent mirror 70 and the beam splitter 65 and is transmitted to the steam capillary 14 via the first mirror 61 and serves as a measuring laser. The wavelength of the measuring laser 71 is preferably different from the wavelength of the laser 10, wherein the receiving units 26, 27' are designed to measure the intensity in the wavelength range of the measuring laser 71. As a result, the measurement can be carried out independently of the wavelength of the laser 10 or the ambient light, because the intensity measurement is specifically coordinated with the wavelength of the measuring laser 71.

[0041] Figure 5 8 shows an example of an intensity measurement of the quotient of two intensity values ​​for different polarization directions. The intensity value 81 is shown on the y-axis through the path of the weld seam plotted on the x-axis 80. The intensity value of the quotient is usually within a range 82, whereby too short abnormal values ​​83, drifts 84 or too strong fluctuations 85 may occur. Drifts indicate, for example, too low a laser power, strong fluctuations indicate too high a laser power and brief deviations 83 indicate welding errors, which may require manual follow-up checks.

Claims

1. A method for monitoring a laser beam welding process, wherein: An optical signal of at least the region of a liquid melt, in particular at least the region of a steam capillary (14), is recorded and signals (24, 25) of different optical polarization directions are generated from the optical signal (29), characterized in that the brightness of the optical signal (24, 25) is evaluated taking into account reference information so that a disturbance in the laser beam welding process is detected at a predetermined deviation from the reference information.

2. The method according to claim 1, characterized in that Exactly two signals with different polarization directions are generated.

3. The method according to claim 2, characterized in that The different optical polarization directions are selected to be orthogonal to one another.

4. The method according to any one of the preceding claims, characterized in that Two optical signals are detected in the infrared spectrum, in particular in a wavelength range between 400 nm and 2000 nm, in particular between 400 nm and 1100 nm.

5. The method according to any one of the preceding claims, characterized in that The signals are detected by photodiodes, respectively.

6. The method according to any one of the preceding claims, characterized in that For evaluation, the signals are combined with one another in a single or multiple calculations, and the combination is evaluated with respect to the reference information.

7. The method according to any one of the preceding claims, characterized in that The laser beam welding process is carried out according to a predetermined path course on the component (19) and reference information associated with the path course is used for evaluation.

8. The method according to any one of the preceding claims, characterized in that If a disruption in the laser beam welding process is detected, the laser power for the further laser beam welding process is adjusted as a function of the determined deviation.

9. The method according to any one of the preceding claims, characterized in that The reference information is provided by a neural network learned by machine learning.

10. A monitoring device for a laser beam welding process, in particular for carrying out a method according to any of the preceding claims, characterized in that An optical detection device (20) for detecting an optical signal of at least the region of a liquid melt, in particular at least the region of a steam capillary (14) of a laser beam welding process, with a beam splitting device (23) for generating signals (24, 25) of different optical polarization directions from the signal, with detection devices (26, 27) for detecting the two signals and with an evaluation device (30) for evaluating the detected signal taking into account reference information in such a way that, given a predetermined deviation from the reference information, it is detected that the laser beam welding process is disturbed.

11. The monitoring device according to claim 10, characterized in that: The detection device (26, 27) comprises a camera or a photodiode.

12. The monitoring device according to any one of claims 10 to 11, characterized in that A laser (71) which, in addition to a laser (10) for performing a laser beam welding process, directs a measuring laser beam at least towards the region of a liquid melt, in particular at least towards the region of a steam capillary (14).

Images