METHOD AND DEVICE FOR PROCESS MONITORING OF A WELD CREATED BY COLLISION WELDING

Description

[0001] The present invention relates to a method and a device for process monitoring in a weld seam formed by collision welding.

[0002] Collision welding processes, such as magnetic pulse welding, often have the disadvantage that the quality of a formed weld seam can only be checked with considerable effort after the welding process has been completed. In particular, this typically requires additional inspection devices such as ultrasonic devices, which makes the entire manufacturing process longer and therefore less costly.

[0003] Although various approaches for improved capture of the physical processes in such welding processes are known from the prior art (for example, the paper "Numerical Simulation of Impact Welding Processes with LS-DYNA," Peter Groche, Christian Pabst, 10th European LS-DYNA Conference 2015, Würzburg, Germany, presents a method for numerical simulation of such welding processes), it has not yet been possible to perform quality control during the actual welding process. WO 2004 / 073913 A2 discloses the preamble of claims 1 and 6.

[0004] The present invention is therefore based on the object of developing a method and a device which avoid the disadvantages mentioned, with which it is possible to monitor a weld seam forming during collision welding during a welding process.

[0005] This object is achieved according to the invention by a method according to claim 1 and a device according to claim 6. Advantageous embodiments and further developments are contained in the dependent claims.

[0006] A method for process monitoring in a weld seam formed by collision welding has a step in which a first joining partner and a second joining partner are accelerated towards each other by an input of energy and are welded together to form the weld seam. A time-resolved optical recording unit with at least one detector detects a light flash between the first joining partner and the second joining partner during welding. An evaluation unit determines an actual value of a start of the light flash, a duration of the light flash, an intensity of the light flash and / or a temporal intensity distribution of the light flash and compares this with a respective target value of the start of the light flash, the duration of the light flash, the intensity of the light flash and / or the temporal intensity distribution of the light flash.The weld seam should only be qualified as having sufficient quality if a maximum deviation of the actual value from the target value is maintained.

[0007] The described method enables cost-effective and non-destructive inspection of the weld seam during the welding process, as the commonly observed flash of light, also known as a light discharge, allows conclusions to be drawn about the quality of the welding process. Surprisingly, the onset and progression of the flash of light can indicate whether the weld seam was formed as desired. The flash of light is initially caused by electromagnetic radiation from a short-term (approximately 1 µs) hot and dense plasma, followed by longer-lasting electromagnetic radiation from a gas expanding from a joining zone mixed with glowing components of the two joining partners.By comparing the target value of the start of the light flash with the actual value of the start of the light flash, it is possible to determine the quality of the formed weld seam without additional measuring instruments or downstream quality control. Alternatively or additionally, the duration of the light flash can be determined using the time-resolved optical recording unit, and this actual value can be compared with the target value. Finally, this is also possible, alternatively or additionally, using the intensity of the light flash as the actual value and target value. Likewise, the temporal intensity distribution can be taken as the actual value and compared with the corresponding target value. Typically, only one of the four measures mentioned is used for classification, but it is also possible to use more than one or several.all four of the above-mentioned dimensions, i.e. the start of the flash, the duration of the flash, the intensity of the flash and the temporal intensity distribution of the flash, are to be used for the classification with regard to the sufficient quality of the weld.

[0008] It can be provided that a sorting unit only selects components formed by collision welding from the first joining partner and the second joining partner with a weld seam classified as sufficiently high quality for further processing. This ensures that components with a defective weld seam are immediately sorted out, thus allowing the joining and testing process to be carried out with high efficiency.

[0009] Collision welding can be carried out by magnetic pulse welding, explosive welding, water jet spot welding, welding by joining partner acceleration using vaporizing foils and / or welding by joining partner acceleration using laser-induced shock waves.

[0010] The deviation of the actual value from the target value, which must be maintained for a weld of sufficient quality, can be specified, but it can also be provided that this deviation is determined using a statistical measure based on several carried out and qualitatively assessed welding processes. These can be carried out both inside and outside a process window in order to record the start of the light flash, the duration of the light flash, the intensity of the light flash and / or the temporal intensity distribution of the light flash for both qualitatively sufficient and qualitatively unsatisfactory welds. The statistical measure is preferably the standard deviation, particularly preferably twice the standard deviation, i.e. if the deviation is greater than the standard deviation or twice the standard deviation, the weld is no longer classified as qualitatively unsatisfactory.

[0011] The light flash is typically detected by a single detector of the time-resolved optical recording unit, but it can also be detected by at least two detectors of the optical recording unit from at least two different spatial positions. The weld seam is only classified as of sufficient quality if the maximum deviation of the actual value from the target value is maintained in all detectors. The two detectors can detect the start of the light flash, the duration of the light flash, the intensity of the light flash and / or the temporal intensity distribution of the light flash. However, it can also be provided that one of the two detectors detects one of the four dimensions mentioned and the other detector a different one of the four dimensions mentioned. The use of two detectors increases the redundancy and thus the reliability of the process to be carried out.The target values ​​of different detectors do not have to be identical, but can be different and adapted to a joint design.

[0012] A device for process monitoring in a weld seam formed by collision welding has a welding unit for accelerating and welding a first joining partner and a second joining partner by introducing energy to form the weld seam. The device also has a time-resolved optical recording unit with at least one detector for detecting a light flash generated during welding between the first joining partner and the second joining partner and an evaluation unit for determining an actual value of a start of the light flash, a duration of the light flash, an intensity of the light flash and / or a temporal intensity distribution of the light flash. The evaluation unit is also designed to compare the respective determined actual value with a respective target value of the start of the light flash, the duration of the light flash, the intensity of the light flash and / oror a temporal intensity distribution of the light flash, wherein the evaluation unit is set up to classify the weld seam as being of sufficient quality only if the maximum deviation of the actual value from the target value is maintained.

[0013] By maintaining the maximum deviation of the actual value from the target value, i.e. by checking whether the actually determined actual value is close enough to the target value theoretically required for a high-quality weld, a reliable and efficient control of the weld formation is made possible.

[0014] Typically, a sorting unit is provided which is designed to forward only components with a weld seam classified as being of sufficient quality to further processing, so that in and with a single device both quality control or process monitoring and suitable further processing of the components formed from the first joining partner and the second joining partner by the welding process are possible.

[0015] To enable sufficiently fast data acquisition, the optical recording unit can be designed with an image recording frequency of at least 1 MHz, i.e., it can be used as a continuously measuring recording unit. Acceleration of the two joining partners typically takes 10 µs, and the light flash lasts approximately 100 µs. It is advantageous to have at least one data point per microsecond.

[0016] Typically, the optical recording unit or at least one detector of the optical recording unit is designed in the form of a photocell, a photomultiplier, a photodiode, a phototransistor, and / or a photoresistor to reliably and quickly obtain images of the welding process and the resulting light flash. Optical feed lines such as fiber optic cables or reflective surfaces can be connected upstream of the detector in the optical recording unit to create a spatial separation between a detection point of the light flash and a conversion into a measurement signal.

[0017] The optical recording unit can have at least two detectors configured to detect the light flash from at least two different spatial positions. The evaluation unit is configured to classify the weld seam as qualitatively sufficient only if a maximum deviation of the actual value from the target value is maintained at the majority, preferably all, detectors. The target values ​​of each detector can differ depending on the joint design. This minimizes the number of faulty detections and further increases the reliability of the method.

[0018] The device described is suitable for carrying out the described method, ie the method can be carried out with the device presented.

[0019] Embodiments of the invention are illustrated in the drawings and are described below with reference to Figure 1-5 explained.

[0020] They show: Fig. 1 shows a schematic, side view of a device for process monitoring in a weld seam formed by collision welding; Fig. 2 shows a schematic representation of a magnetic pulse welding process using rotationally symmetrical joining partners; Fig. 3 shows a schematic representation of a laser-induced shock wave welding process using sheet metal; Fig. 4 shows a schematic view of an explosive welding process using simultaneous acceleration of both joining partners; and Fig. 5 shows a schematic view of a welding process using joining partner acceleration by vaporizing foils.

[0021] In Figure 1A schematic side view of a collision welding process is shown. This collision welding process is intended to enable locally resolved and non-destructive recording of the time of impact and the weld seam formation in real time during welding with high mechanical energy, i.e., a welding process as defined by DIN EN ISO 4063, 2011, which particularly characterizes magnetic pulse welding (MPW), explosive welding (EXW), collision welding processes through component acceleration using vaporizing foils (Vaporizing Foil Actuator Welding VFAW), laser-induced shock waves (Laser Impact Welding LIW), and water jets (Water Jet Spot Welding WSW).

[0022] In all of these processes, a first joining partner 1 and a second joining partner 2 are spatially adjacent to one another and welded together by applying energy, forming the weld seam. The energy input can vary depending on the type of process used and, for example, in magnetic pulse welding, can be a pulsed magnetic field. For this purpose, a welding unit 5 is provided in each of the processes, which is responsible for applying energy to the first joining partner 1 and / or the second joining partner 2.

[0023] By means of a time-resolved optical recording unit 6 with two detectors 19 and 20 in the Figure 1In the exemplary embodiment shown, a flash of light occurring during welding between the first joining partner 1 and the second joining partner 2 is detected, by means of which conclusions can surprisingly be drawn about the quality of the formed weld seam, without a component formed from the first joining partner 1 and the second joining partner 2 having to be subjected to further examination.

[0024] A recording device 15 receives a signal from both the welding unit 5 and the time-resolved optical recording unit 6. The evaluation unit 16 is connected to the recording device 15 and thus determines an actual value for the start of the light flash, a duration of the light flash, a temporal intensity distribution of the light flash, and / or an intensity of the light flash, and compares these measurements with a respective target value. The weld seam is only classified as qualitatively sufficient if the actual value deviates from the target value within a maximum.

[0025] In the Figure 1In the embodiment shown, a component formed in this way is transported further on a conveyor device 18, for example a conveyor belt, and a sorting unit 17 is provided, which is also controlled by the evaluation unit 16 and is designed to remove from the conveyor unit 18 those components which do not meet the quality requirements.

[0026] The deviation of the actual value from the target value to be maintained can be entered into the evaluation unit before carrying out the process or can be determined by the evaluation unit 16 using a statistical measure, such as the standard deviation, on the basis of several welding processes carried out and subsequently evaluated.

[0027] To ensure continuous image recording, the time-resolved optical recording unit 6 is designed with an image recording frequency of at least 1 MHz. The recording unit 6 can be a photocell, a photomultiplier, a photodiode, a phototransistor, and / or a photoresistor.

[0028] In addition, the time-resolved optical recording unit 6 can have the two detectors 19 and 20 shown, which detect the light flash from at least two different spatial positions. In addition to the four dimensions mentioned, the weld seam is only classified as qualitatively sufficient if a maximum deviation of the actual value from the target value is maintained at both detectors 19, 20. For this purpose, the two detectors 19 and 20 send light intensity signals to the evaluation unit 16, usually via an electrical connection. The two detectors 19 and 20 do not have to be identical; for example, detector 19 can be designed as a photocell, while detector 20 is a photomultiplier. A combination of detector 19 or detector 20, amplifier circuit 14, and evaluation unit 16 is possible, for example, through optical inputs on oscilloscopes in the evaluation unit 16.In further embodiments, of course, only a single detector can be provided in the optical recording unit 6.

[0029] The Figure 1The device shown and the method to be carried out with it make it possible to detect deviations in the shape and position tolerances of the first joining partner 1 and the second joining partner 2 relative to one another and to the welding unit 5, as well as insufficient or faulty fixation of the first joining partner 1 and the second joining partner 2. Furthermore, deviations in the mechanical properties, such as the modulus of elasticity, a yield point, a hardness, a ductility or a transverse contraction of the first joining partner 1 and the second joining partner 2 can also be determined. Deviations in collision speeds, which can occur due to insufficient acceleration, for example in the case of coil defects, generator defects or insufficient explosive, or contaminants on the surfaces to be joined can also be identified as sources of error in the weld seam formation. Finally, the device and method described make it possible toThe described method also detects deviations in the surface properties of the joining partners 1, 2, such as roughness or waviness or temperature deviations in a joining gap between the first joining partner 1 and the second joining partner 2 during the welding process.

[0030] This exploits the fact that the start time of the light flash correlates with the collision time, which reveals asymmetries or shifts when comparing the start times, particularly when using spatially distributed detectors 19 and 20. The intensity and duration of the detected light flash correlate with the impact speed of the joining partners 1 and 2 involved, whereby an appropriate impact speed is essential for correct weld seam formation. Foreign matter on the surface, such as oil, can drastically shorten the duration of the light flash and thus prevent welding. Surface defects, such as roughness, can also negatively influence the intensity.

[0031] In Figure 2A schematic view of a device for magnetic pulse welding is shown. Recurring features are provided with identical reference numerals in this drawing, as well as in the following drawings. The first joining partner 1, which has a flat or curved surface and is to be deformed or moved locally or globally, is held in a first fixture 4. Likewise, the second joining partner 2, which also has a flat or curved surface and is to be deformed or moved locally or globally, is held by a second fixture 3.

[0032] In the illustrated embodiment, an acceleration tool of the welding unit 5 with a defined process start time is in the form of a current-carrying tool coil. The time-resolved measuring optical recording unit 6 in the illustrated embodiment is a photodetector, i.e., an optoelectrical converter that can be sensitive to different wavelengths of the electromagnetic spectrum depending on the application. In the illustrated embodiment, the optical recording unit 6 is provided by two detectors 19 and 20, each of which leads to a converging lens 9 via an optical fiber 8 with an optical connection to the detector 19, 20. The light flash occurring during welding can be detected by the two detectors 19, 20 and the converging lenses 9, of which more than two exist in further embodiments and can be distributed over a circumference or length.For this purpose, the converging lenses 9 are held in a fixation 10 without influencing the joining zone. In the case shown in . Figure 2 In the embodiment shown, a protective glass 11 is also provided in front of the converging lenses 9. The optical recording unit 6 is Figure 2 In the embodiment shown, it is thus formed by the converging lenses 9, the optical waveguides 8 and the detectors 19 and 20.

[0033] A signal detected by the optical recording unit 6 is forwarded by an electrical amplifier circuit 14 connected downstream of each of the detectors to a recording device 15, in the illustrated embodiment an oscilloscope, which requires a temporal zero point (e.g. a start of an electrical current flow in the coil or an ignition time) and, in the illustrated embodiment, records a light intensity in relation to this zero point t 0. In the evaluation unit 16, the start, duration and intensity of the light flash are subsequently compared with predefined target values. Figure 2In the illustrated embodiment, the diagram shows an electrical coil current plotted against time, with the beginning of the light flash being measured at position 1 at time t 3 and the end at time t 4, and the beginning of the light flash being detected at position 2 by the second detector at time t 1 and the end at time t 2. Since time t 1 is before time t 3, incorrect positioning is inferred in the illustrated case.

[0034] Figure 3shows a schematic view of a laser impact welding process. The first joining partner 1 and the second joining partner 2 are again held in their fixations 3 and 4, but now the energy input for deformation and welding is provided by a laser beam 21. In the illustrated embodiment, an optical recording unit 6 is again provided, the only detector 19 of which is now arranged in front of a flat or curved deflecting mirror 7 with a fastening 23 or a flat or curved reflective component surface for reflecting the light discharge or the light flash in the direction of the optical recording unit 6 in order to capture the intensity of the light flash as completely as possible. For this purpose, the recording unit 6 is provided with the fastening 10.In the illustrated embodiment, a protective glass 12 is provided without influencing the joining zone, which is equipped with a protective glass cleaning device 13 that is cleaned, for example, continuously by compressed air or a rinse, or at intervals by a brush, a burst of compressed air, or a burst of detergent. The evaluation unit 16, for example a computer, again compares the target values ​​and the actual values. In the illustrated embodiment, a laser pulse is registered at time t 0 and the start of the light flash at time t 1. However, since the target value of the light flash is at t 3 and the deviation is therefore greater than permitted, it can be concluded that the mechanical properties of the respective joining partner deviate from the intended mechanical properties.

[0035] Figure 4shows a device for explosive welding in which explosives are triggered via a detonator 22 so that the first joining partner 1 and the second joining partner 2 are moved towards one another. The electronic evaluation takes place as in the previously described examples, with the ignition now being detected in the evaluation unit 16 at time t 0. In this case, the duration of the light flash as the target value should cover the time interval between t 3 and t 4 and the light flash should have an intensity I 3 , although in the illustrated embodiment only a length from t 1 to t 2 and an intensity I 1 are measured as the actual value. Since the intensity I 1 is less than I 3 and the measured temporal intensity distribution does not match the expected temporal intensity distribution, the impact speed is too low, which means that incorrect dosing of the explosive medium can be concluded.

[0036] Finally, in Figure 5 A schematic representation of a welding process by component acceleration using vaporizing foils is shown. In the lower part of the Figure 5the arrangement is shown in a side view, while in the upper part a top view of the same arrangement can be seen. In the arrangement shown, a total of three detectors 19, 20 and 24 of the optical recording unit 6 are used, so that the light flash is detected at three positions. In addition, the optical recording unit comprises the optical waveguides 8 and the converging lenses 9. In such processes with vaporizing foils, a wire is heated by resistance heating and completely vaporized by a pulsed electric current, so that the first joining partner 1 is accelerated in the direction of the second joining partner 2 by the pressure of the vaporizing wire. Ignition takes place at a time t 0, and it is then determined from the diagram in the evaluation unit 16 that the detector located at position 3 detects a light flash that is too short, although it starts at the correct time. This indicates impurities orRoughness or an excessively low temperature in the joint gap. In further embodiments, each of the detectors used can also be configured to detect only one of the four dimensions mentioned. For example, detector 19 can detect the onset of the light flash, detector 20 can detect the duration of the light flash, and detector 24 can detect the temporal intensity distribution of the light flash.

Claims

1. Method for monitoring the process for a welding seam formed by means of collision welding, in which a first joining partner (1) and a second joining partner (2) are accelerated towards one another by an introduction of energy and are welded to one another forming a welding seam, characterized in that by means of an optical recording unit (6) with at least one detector (19, 20, 24) that takes time-resolved measurements, during the welding process, a flash of light caused by the electromagnetic radiation of a short-term, hot and dense plasma, followed by longer-lasting electromagnetic radiation of a gas expanding out of a joining zone, mixed with incandescent constituents of the first joining partner (1) and of the second joining partner (2), is detected between the first joining partner (1) and the second joining partner (2) and an actual value of a start of the light flash, a duration of the light flash, an intensity of the light flash and / or an intensity distribution over time of the light flash is / are determined by an evaluation unit (16) and compared with a respective target value of the start of the light flash, the duration of the light flash, the intensity of the light flash and / or an intensity distribution over time of the light flash, wherein the welding seam is classified as being of sufficient quality only if a maximum deviation of the actual value from the target value is maintained.

2. Method according to claim 1, characterized in that only components with a welding seam classified as being of sufficient quality are supplied for further processing by a sorting unit (17).

3. Method according to claim 1 or claim 2, characterized in that magnetic pulse welding, explosion welding, waterjet spot welding, welding by joining partner acceleration by means of vaporizing foils and / or welding by joining partner acceleration by means of laser-induced shock waves is carried out as the collision welding.

4. Method according to any one of the preceding claims, characterized in that the deviation of the actual value from the target value is predetermined or is established via a statistical measure, preferably via the standard deviation, on the basis of a plurality of welding processes which are carried out and evaluated in terms of quality.

5. Method according to any one of the preceding claims, characterized in that the light flash is detected by at least two detectors of the optical recording unit (6) from at least two different spatial positions, wherein the welding seam is only classified as being of sufficient quality if a maximum deviation of the actual value from the target value is maintained at all detectors.

6. Device for monitoring the process for a welding seam formed by means of collision welding, having a welding unit (5) for accelerating and welding a first joining partner (1) and a second joining partner (2) by an introduction of energy to form the welding seam, characterized in that the device has the following features: an optical recording unit (6) with at least one detector (19, 20, 24) that takes time-resolved measurements for detecting a light flash generated during welding by electromagnetic radiation of a short-term, hot and dense plasma, followed by longer-lasting electromagnetic radiation of a gas expanding out of a joining zone, mixed with incandescent constituents of the first joining partner (1) and the second joining partner (2), between the first joining partner (1) and the second joining partner (2), and an evaluation unit (16) for determining an actual value of a start of the light flash, a duration of the light flash, an intensity of the light flash and / or an intensity distribution over time of the light flash, and for comparing the respective determined actual value with a respective target value of the start of the light flash, the duration of the light flash, the intensity of the light flash, and / or an intensity distribution over time of the light flash, wherein the evaluation unit (16) is configured to classify the welding seam as being of sufficient quality only if a maximum deviation of the actual value from the target value is maintained.

7. Device according to claim 6, characterized in that a sorting unit (17) is provided, which is configured to supply only components with a welding seam classified as being of sufficient quality for further processing.

8. Device according to claim 6 or claim 7, characterized in that the optical recording unit (6) is configured to have an image recording frequency of at least 1 MHz.

9. Device according to any one of claims 6 to 8, characterized in that the optical recording unit (6) is configured as a photocell, a photomultiplier, a photodiode, a phototransistor, and / or a photoresistor.

10. Device according to any one of claims 6 to 9, characterized in that the optical recording unit (6) has at least two detectors, which are configured to capture the light flash from at least two different spatial positions, wherein the evaluation unit (16) is configured to classify the welding seam as being of sufficient quality only if a maximum deviation of the actual value from the target value is maintained at the majority, preferably all of the detectors.