Method for processing a hairpin, laser processing machine, method for determining a property in the processing of a hairpin, and inspection machine

Abstract

The invention relates to a method for processing a hairpin (12) by means of a pulsed laser beam (18), wherein the method comprises: processing the hairpin (12) by means of the pulsed laser beam (18); detecting a sound emission from the hairpin (12) during the processing of the hairpin (12); and determining information about the processing of the hairpin (12) by analysing the detected sound emission.

Description

[0001] Title: Method for processing a hairpin, laser processing machine, method for determining a property of processing a hairpin and inspection machine

[0002] Description

[0003] The invention relates to a method for processing a hairpin, a laser processing machine, a method for determining a property of a processing of a hairpin and an inspection machine.

[0004] In the production of stators for motors or generators using hairpin technology, the hairpins used are typically stripped of their coating before welding. The coating removal process can be particularly critical, as incomplete removal leads to poorer mechanical and electrical properties of the weld. To ensure complete coating removal, quality control is often performed during the process to verify that no residue remains.

[0005] Hairpin welding can be a complex process, often involving errors such as mismatched hairpins, uneven energy input, and / or the formation of unwanted intermetallic phases.

[0006] Typically, the paint removal process and hairpin welding are monitored using a camera. This monitoring can include analysis of the images captured by the camera. Camera monitoring allows for verification of complete paint removal and clarification of any processing errors during hairpin welding.

[0007] The invention aims to provide a method for processing a hairpin, a laser processing machine, a method for determining a property of a hairpin processing process, and an inspection machine, each exhibiting improved properties, in particular enabling cost-effective and reliable processing and / or inspection of hairpins. The invention achieves this objective by providing a method with the features of claim 1, a laser processing machine with the features of claim 11, a method with the features of claim 12, and an inspection machine with the features of claim 13. Advantageous embodiments and further developments of the invention are described in the dependent claims.

[0008] A method according to the invention serves to process a hairpin using a pulsed laser beam. The method comprises: processing the hairpin using the pulsed laser beam; detecting sound emission from the hairpin during the processing; and obtaining information about the processing of the hairpin by analyzing the detected sound emission.

[0009] Advantageously, the hairpin can emit sound during machining. This sound emission can be characteristic of the current state of the machining process, the quality of the machining, and / or the occurrence of a machining defect. The sound emission can be detected, for example, using an inexpensive microphone. Furthermore, the analysis of the detected sound emission can be performed with minimal technical effort and resources, thus saving further costs. Therefore, this method can enable cost-effective and reliable machining of the hairpin.

[0010] Another aspect of the procedure is that it is non-invasive, in particular non-destructive, and that the information is obtained based on the detected acoustic emission in near real time during the processing of the hairpin.

[0011] Another aspect of the process may be that the laser beam is used for processing the hairpin and simultaneously for generating the sound emission to determine the information, thus giving the laser beam a dual function.

[0012] The hairpin can be suitable, and in particular intended, for forming a stator for a motor or generator.

[0013] The hairpin can be designed as an elongated conductor. It can have two elongated legs, particularly parallel to each other, which are connected by a connecting section. The hairpin can be U-shaped. In other words, the hairpin can be a U-shaped bent wire. The hairpin can have a core coated with a lacquer. The core can be made of a metal suitable for conducting electricity. In particular, the core can be made of a copper alloy with a copper content of at least 80%, preferably 95%, or of an aluminum alloy with an aluminum content of at least 80%, preferably 95%. The lacquer can serve to electrically insulate the core.

[0014] A laser pulse of the pulsed laser beam can have a pulse duration in the picosecond or nanosecond range, particularly in the range of 1 picosecond to 950 nanoseconds. The pulse repetition rate of the pulsed laser beam can range from 1 kHz (kilohertz) to 4 MHz (megahertz). The pulse repetition rate can also be referred to as the pulse frequency or pulse repetition frequency.

[0015] Sound emission can be detected by detecting structure-borne sound. This detection can be achieved using a number of sound transducers, for example, one, two, or three, particularly microphones. A sound transducer can be configured to convert the detected sound emission into an electrical signal.

[0016] The information can be retrieved using an electronic processing unit. This unit can be a computer, a microcontroller, and / or a programmable integrated circuit.

[0017] The information can contain details about the processing status of the hairpin. In particular, the information can contain details about the processing status, processing progress, and / or processing errors.

[0018] The information can include details about the condition of a laser processing machine performing the process. Preferably, the information can indicate whether the laser processing machine has a defect or not. The defect could be, for example, a defective nozzle and / or a defective protective glass.

[0019] The processing of the hairpin, the detection of the acoustic emission, and the acquisition of the information can all be performed simultaneously. Advantageously, this allows the information to be acquired in near real-time during the processing of the hairpin.

[0020] The analysis of acoustic emissions can include a Fourier analysis, specifically a Fourier transform, of the detected acoustic emission. The information can be determined based on at least one additional data set. This additional data set could, for example, contain image data from a camera and / or at least one beam parameter of the pulsed laser beam. By determining the information based on this additional data set and analyzing the detected acoustic emission, the information can be determined more precisely. For example, potential information describing processing states that are incompatible with the additional data set can be excluded.

[0021] Another aspect could be that the information enables optimization of the processing.

[0022] In a further development of the method, the information is obtained using a trained machine learning model. Advantageously, characteristic patterns in the detected sound emission can be reliably and accurately identified using the trained machine learning model.

[0023] The trained machine learning model can also be referred to as artificial intelligence, in particular as having been trained as artificial intelligence.

[0024] Obtaining the information using the trained machine learning model can involve applying the trained machine learning model to the detected sound emission.

[0025] The trained machine learning model can be trained to recognize, and in particular identify, characteristic patterns in the detected sound emission, and assigns the information to the pattern of the detected sound emission.

[0026] The pattern can be a pattern in the frequency spectrum of the detected sound emission. The frequency spectrum can also be referred to as the sound spectrum. The sound spectrum can be a representation of the sound pressure amplitude as a function of frequency.

[0027] In other words, the trained machine learning model can identify characteristic patterns in the sound spectrum that correlate with process behavior during hairpin processing. The identified patterns can be grouped. Each group can be assigned a label that describes the process behavior. Each group can describe a process state. The information can describe the process state.

[0028] The trained machine learning model can be trained using multiple acoustic emissions, each corresponding to a piece of information about the processing of the hairpin. This allows the trained machine learning model to determine the information based on the detected acoustic emission.

[0029] The trained machine learning model can be an artificial neural network, in particular a self-organizing map (SOM).

[0030] Another aspect of the trained machine learning model can be that additional sensor readings can be incorporated into the information gathering process.

[0031] In a further development of the process, the procedure, after determining the information, includes: adjusting at least one beam parameter of the pulsed laser beam based on the determined information. Advantageously, this allows for optimization of the processing process.

[0032] The beam parameter of the pulsed laser beam can be adjusted during hairpin processing. This helps to counteract the occurrence of processing errors and allows for monitoring of the processing process. Alternatively, the beam parameter of the pulsed laser beam can be adjusted after the hairpin processing is complete, thus optimizing subsequent processing steps.

[0033] The beam parameter can be an average power of the laser beam, a pulse duration, a focus diameter, a laser power, a nozzle focus distance, a nozzle workpiece distance, a feed rate, a gas pressure, a nozzle diameter and / or a gas type.

[0034] In a further development of the process, processing the hairpin involves removing the lacquer from the hairpin, particularly in a specific area, using a pulsed laser beam. The information obtained is about the lacquer removal, specifically about the lacquer removal process. Advantageously, the process can be particularly suitable for lacquer removal. Lacquer removal can also be referred to as stripping. Lacquer removal can also refer to the removal of the lacquer from the hairpin.

[0035] The information obtained can indicate whether the lacquer of the hairpin is completely, partially, or not removed from the core of the hairpin by means of the laser beam, particularly within a specified area.

[0036] The area can be an end section of the hairpin. The area can extend from a free end of the hairpin. The area can have a length ranging from 0.5 cm (centimeters) to 4 cm.

[0037] In a further development of the method, the analysis of the detected sound emission includes the analysis of its frequency spectrum. Advantageously, the information can be determined with particular precision using the frequency spectrum. The frequency spectrum can be determined by a Fourier transform of the detected sound emission.

[0038] In a further development of the method, the information is determined as a function of a sound level, in particular a sound pressure level, of the sound emission at a frequency whose value is equal to a pulse repetition frequency of the pulsed laser beam. Advantageously, this can reduce or completely eliminate the influence of disturbances or artifacts in the detected sound emission, which occur at a frequency other than the pulse repetition frequency of the pulsed laser beam, on the determination of the information.

[0039] The sound level, in particular the volume, at a frequency equal to the pulse repetition frequency of the pulsed laser beam, can decrease with increasing distance from the lacquer. Advantageously, this allows the determination, based on the sound level, of whether the lacquer of the hairpin has been completely, partially, or not at all removed from the core of the hairpin by the laser beam.

[0040] The pulse repetition rate can be a pulse rate, a repetition rate and / or a pulse repetition rate.

[0041] In a further development of the procedure, determining the information includes determining the information based on an assignment of a pattern of the frequency spectrum to a paint removal state, in particular by means of the trained machine learning model.

[0042] The paint removal state can describe a state of paint removal. For example, the paint removal state can describe the state that the paint has not been removed, the state that the paint has been partially removed, or the state that the paint has been completely removed.

[0043] The pattern of the frequency spectrum can be understood as a pattern in the progression of an envelope within the frequency spectrum. The pattern of the frequency spectrum can be formed, for example, by frequency ranges or intensity fluctuations.

[0044] The information can be representative of the paint removal condition.

[0045] In a further development of the process, processing the hairpin involves electrically joining one hairpin to another by welding with a pulsed laser beam. The information obtained is about the welding process, specifically the welding procedure itself. Advantageously, this method can be particularly well-suited for welding hairpins. The method can contribute to improving the quality of the welded joint and making the welding process more efficient.

[0046] The information obtained can be about the weld pool dynamics during welding. Additionally or alternatively, the information can indicate whether or not there is a mismatch in the hairpins, uneven energy input, an unwanted intermetallic bond, and / or distortion of the hairpins.

[0047] In a further development of the method, the analysis of the detected sound emission includes the analysis of its frequency spectrum. Advantageously, the information can be determined with particular precision using the frequency spectrum. The frequency spectrum can be determined by a Fourier transform of the detected sound emission.

[0048] In a further development of the procedure, determining the information includes determining the information based on an assignment of a pattern of the frequency spectrum to a welding state of the welding process, in particular by means of the trained machine learning model.

[0049] The welding condition can describe a state of the welding process. For example, the welding condition can describe whether the hairpins are not welded together, whether they are partially welded together, or whether they are fully welded together. Additionally or alternatively, the welding condition can describe the formation of a vapor capillary or a weld pool. Additionally or alternatively, the welding condition can describe whether the welding is flawless or results in processing defects.

[0050] The pattern of the frequency spectrum can be understood as a pattern in the progression of an envelope within the frequency spectrum. The pattern of the frequency spectrum can be formed, for example, by frequency ranges or intensity fluctuations.

[0051] The information can be representative of the welding condition.

[0052] A laser processing machine according to the invention is designed for processing a hairpin using a pulsed laser beam. The laser processing machine comprises a laser beam source for generating the pulsed laser beam; a detector device for detecting acoustic emission from the hairpin during processing; and an analysis device. The analysis device is designed to obtain information about the processing of the hairpin by analyzing the detected acoustic emission.

[0053] The laser processing machine is specifically designed and configured to perform a previously described process. The previously given description of the process can also apply to laser processing machines with identical or functionally equivalent features.

[0054] The laser processing machine can at least partially strip the paint from the hairpin and / or weld it.

[0055] The detector device can have a frequency resolution that determines the number of frequency components that can be detected in the sound spectrum. A method according to the invention serves to determine a property of processing a hairpin with a pulsed detector laser beam. The method comprises: irradiating the hairpin with the pulsed detector laser beam; detecting sound emission from the hairpin during irradiation; and determining the property of the hairpin processing by analyzing the detected sound emission.

[0056] Advantageously, the hairpin can emit sound when irradiated with the pulsed laser beam. This sound emission can be characteristic of the hairpin's processing characteristics. In particular, the sound emission can indicate the quality of the processing. Specifically, the presence or absence of a processing defect can be determined based on the sound emission.

[0057] The procedure for determining the properties of the hairpin machining can be carried out after the hairpin machining has taken place.

[0058] The method for determining the hairpin machining property may differ from the method for machining the hairpin using a pulsed laser beam in that the detector laser beam is not suitable for machining the hairpin. The previously given description of the hairpin machining method may apply to identical or functionally equivalent features of the method for determining the hairpin machining property.

[0059] The procedure for determining the properties of the hairpin machining can be carried out immediately after the hairpin machining.

[0060] A laser pulse of the pulsed detector laser beam can have a pulse duration in the picosecond or nanosecond range, particularly in the range of 1 picosecond to 950 nanoseconds. The pulse repetition frequency of the pulsed detector laser beam can be from 1 kHz to 4 MHz.

[0061] The determined property can describe the quality of the machining, the presence of a machining defect, or the absence of a machining defect.

[0062] This property can be determined by comparing the detected acoustic emission from the hairpin with the detected acoustic emission of a correctly machined hairpin, especially a reference hairpin. The property can be determined based on the sound level at the repetition rate of the detector laser beam.

[0063] An inspection machine according to the invention is configured to perform a previously described method for determining the machining properties of the hairpin. The inspection machine cannot be configured to machine the hairpin itself.

[0064] Further advantages and advantageous embodiments of the invention can be seen from the figures, their description, and the claims. All features disclosed in the figures, their description, and the claims can be essential to the invention, both individually and in any combination. The figures show:

[0065] Fig. 1 shows a schematic representation of a paint stripping machine during the paint removal process on a hairpin.

[0066] Fig. 2 shows a schematic representation of a welding machine during the welding of a hairpin, and

[0067] Fig. 3 shows a schematic representation of an inspection machine.

[0068] Fig. 1 shows a laser processing machine in the form of a paint stripping machine 10 during the processing of a hairpin 12.

[0069] The hairpin 12 is designed as a U-shaped conductor. The hairpin 12 has a core 14 and a lacquer 16. The core 14 is coated with the lacquer 16. The core 14 is designed to conduct electricity. The core 14 is made of a copper alloy with a copper content of at least 90%. The lacquer 16 serves to electrically insulate the core 14.

[0070] A motor stator is to be manufactured using a large number of identical hairpins 12. To manufacture the stator, it is necessary to electrically connect the large number of identical hairpins 12. The electrical connection is made by means of welds, each of which is located in a welding area. To create the welds, each hairpin 12 is stripped of its coating in an area 15 before welding. Area 15 is an end region of the hairpin 12. Stripping the coating removes the coating 16 from area 15.

[0071] Paint removal is particularly critical, as incomplete paint removal from the weld area leads to poorer mechanical and electrical properties of the weld. The processing of the hairpin 12 shown in Fig. 1 is paint removal using a laser beam 18 of the paint removal machine 10.

[0072] The absorption coefficient of the varnish 16 for a wavelength of the laser beam 18 is higher than the absorption coefficient of the material from which the core 14 is formed.

[0073] Laser beam 18 is a pulsed laser beam. A laser pulse of pulsed laser beam 18 has a pulse duration in the range of 1 picosecond to 950 nanoseconds. The pulse repetition frequency of pulsed laser beam 18 is 20 kHz.

[0074] The paint stripping machine 10 has a laser beam source 20 that generates the pulsed laser beam 18. The laser beam source 20 can be designed as a disk laser, fiber laser, or CO2 laser.

[0075] The paint stripping machine 10 has a processing head 22 from which the laser beam 18 emerges. The laser beam 18 is directed onto the hairpin 12 by means of the processing head 22.

[0076] The paint stripping machine 10 has a clamping device 24 in which the hairpin 12 is clamped for paint stripping.

[0077] The paint stripping machine 10 has a detector device 26 in the form of a transducer, in particular a MEMS microphone. The detector device 26 is attached to the clamping device 24. The detector device 26 has a mechanical contact with the clamping device 24.

[0078] The detector device 26 is designed to detect sound emission, in particular structure-borne sound, emitted by the hairpin 12 during the paint stripping process. The sound transducer 26 converts the detected sound emission into an electrical signal, which is received by an analysis device 28 of the paint stripping machine 10.

[0079] The detector unit 26 is configured to determine a frequency spectrum of the detected acoustic emission by means of a Fourier transform, in particular FFT decomposition. In other words, converting the detected acoustic emission into the electrical signal includes determining the frequency spectrum. The frequency spectrum is output in the form of the electrical signal. The analysis unit 28 is configured to determine information about the removal of the paint from the hairpin 12 based on an analysis of the detected acoustic emission, in particular the electrical signal. The determined information indicates whether the paint 16 has been completely, partially, or not completely removed from the core 14 by means of the laser beam 18.

[0080] The analysis device 28 determines the information as a function of a sound pressure level of the sound emission at a frequency whose value is equal to a pulse repetition frequency of the pulsed laser beam 18. In particular, the pulse repetition frequency of the pulsed laser beam 18 is specified for the analysis device 28.

[0081] The analysis device 28 filters out all frequencies except the pulse repetition frequency of the pulsed laser beam 18 from the frequency spectrum of the sound emission and determines a sound pressure level based on the filtered frequency spectrum.

[0082] Since the absorption coefficient of the lacquer 16 is higher than that of the core 14, the absorbed power of the laser beam 18 is higher in the presence of lacquer 16 than in its absence. Due to the higher absorbed power of the laser beam 18 in the presence of lacquer 16, the sound pressure level at the frequency equal to the pulse repetition frequency of the laser beam 18 is higher than in the absence of lacquer 16. Thus, as the lacquer 16 is removed, the sound pressure level at the frequency of the laser beam 18 decreases continuously and drops abruptly when the lacquer 16 is completely removed and the laser beam is striking, in particular, only the core 14. In other words, the sound level at the frequency equal to the pulse repetition frequency of the laser beam 18 decreases as the lacquer 16 is removed.

[0083] Depending on the progress of the paint removal, the analysis unit 28 determines the information from "Paint 16 not removed" to "Paint 16 partially removed" to "Paint 16 removed". If the paint is not completely removed, the analysis unit 28 determines the information "Paint 16 not removed" or "Paint 16 partially removed". This allows the information to indicate whether paint 16 has been completely removed or not.

[0084] A section of the hairpin 12 illuminated by the laser beam 18 can be irradiated with the laser beam 18 until the analysis device 28 detects the information "Paint 16 removed". Advantageously, this protects the core 14 from unnecessary irradiation with the laser beam 18. Furthermore, it prevents paint residues or other contaminants that could negatively affect subsequent welding.

[0085] The analysis device 28 can include a trained machine learning model. The analysis device 28 can be trained to determine the information about the paint removal of the hairpin 12 using the trained machine learning model. The trained machine learning model can be provided with the sound pressure level of the acoustic emission at the pulse repetition frequency of the laser beam 18 and additional data about the hairpin 12, for example, in the form of image data of the hairpin 12 and / or beam parameters of the pulsed laser beam 18. Based on the provided data, the trained machine learning model can determine the information about the paint removal of the hairpin 12.

[0086] In an alternative embodiment not shown, the detector device can contact the hairpin for the purpose of detecting the sound emission.

[0087] In another alternative embodiment not shown, the analysis device 28 can be configured to determine the frequency spectrum.

[0088] Fig. 2 shows a laser processing machine in the form of a welding machine 30. The same reference numerals are used for identical and functionally equivalent elements, and in this respect reference can be made to the above explanations of the embodiment of Fig. 1, so that essentially only the existing differences are discussed.

[0089] Fig. 2 shows the welding machine 30 during machining of the hairpin 12. Machining the hairpin 12 is the creation of a welded joint 34 between the hairpin 12 and another hairpin 32.

[0090] Hairpin 12 and the other hairpin 32 are identical in construction. The other hairpin 32 has a core 36 and a lacquer 38. The previously given description of the core 14 and the lacquer 16 of hairpin 12 applies accordingly to the core 36 and the lacquer 38 of the other hairpin 32.

[0091] The weld 34 is produced by melting the two cores 14, 36 using the laser beam 18. The hairpin 12 and the other hairpin 32 are arranged in a housing 25 of the stator. The detector device 26 contacts the housing 25 for the purpose of detecting the sound emission.

[0092] The laser beam strikes the two cores 14, 36 and melts them to create the weld 34. During the melting of the two cores 14, 36, the analysis unit 28 determines information about the welding process based on the analysis of the acoustic emission detected by the detector unit 26.

[0093] The information obtained about the welding process can include information about the weld pool dynamics during welding, information about a mismatch of the hairpins, information about uneven energy input, information about an unwanted intermetallic bond and / or information about distortion of the hairpins.

[0094] Analysis facility 28 has a trained machine learning model. The trained machine learning model is trained to analyze the frequency spectrum of the detected sound emission during the production of the weld joint 34. The trained machine learning model is a self-organizing map.

[0095] The trained machine learning model is trained to recognize, and in particular identify, characteristic patterns in the frequency spectrum. In other words, the trained machine learning model is trained to recognize patterns in the frequency spectrum that correlate with process behavior.

[0096] The trained machine learning model assigns information to the frequency spectrum pattern. This assignment can be achieved by dividing the identified pattern into groups. Each group is assigned a label that describes the process behavior. The information then describes this process behavior.

[0097] For example, a machining defect can occur during welding. This defect could be, for instance, strong weld pool dynamics, a mismatch in the hairpins, uneven energy input, unwanted intermetallic bonding, and / or hairpin distortion, each of which produces a characteristic pattern in the frequency spectrum. If a machining defect is present, the trained machine learning model detects its presence based on the frequency spectrum pattern and determines the information related to the detected defect. In other words, the information represents the detected machining defect. The trained machine learning model assigns the information based on the recognized frequency spectrum pattern.

[0098] The analysis unit 28 is designed to adjust at least one beam parameter of the pulsed laser beam 18 based on the determined information and / or to control the processing head 22.

[0099] The beam parameter can be an average power of the laser beam, a pulse duration, a focus diameter, a laser power, a nozzle focus distance, a nozzle workpiece distance, a feed rate, a gas pressure, a nozzle diameter and / or a gas type.

[0100] The adjustment of the beam parameter and / or the control of the processing head 22 take place during the welding process. This allows the analysis unit 28 to regulate the welding based on the information it has gathered.

[0101] For example, the information may include that there is excessive melt pool dynamics during the production of the weld joint 34. Based on this information, the analysis device 28 can reduce the laser power of the laser beam 18 by controlling the laser beam source 20.

[0102] Fig. 3 shows an inspection machine 40. The same reference numerals are used for identical and functionally equivalent elements, and in this respect reference can be made to the above explanations of the embodiment of Fig. 1, so that essentially only the existing differences are discussed.

[0103] The hairpin 12 has undergone a machining process. The inspection machine 40 is used to check whether the hairpin 12 has been machined correctly. The inspection machine 40 is not designed to machine the hairpin 12. The machining process was paint stripping. The inspection machine 40 is designed to check whether the hairpin 12 has been stripped of paint correctly, in particular as desired. The inspection machine 40 has a laser beam source 42 for generating a detector laser beam 44. The detector laser beam 44 is a pulsed laser beam. A laser pulse of the pulsed detector laser beam 44 has a pulse duration in the range of 1 picosecond to 950 nanoseconds. A pulse repetition frequency of the pulsed detector laser beam 44 has a value in the range of 1 kHz to 4 MHz. The detector laser beam 44 is not suitable for machining the hairpin 12.

[0104] The inspection machine 40 irradiates the hairpin 12 to be checked with the pulsed detector laser beam 44.

[0105] The inspection machine 40 detected an acoustic emission from the hairpin 12 by means of the detector device 26 during the irradiation of the hairpin 12 with the detector laser beam 44.

[0106] The analysis unit 28 of the inspection machine 40 determines a property of the machining process by analyzing the detected acoustic emission. This property of the machining process provides information about whether the hairpin 12 has been correctly stripped of paint or not. In this way, the inspection machine 40 checks whether a machining defect is present or not.

[0107] For example, the inspection machine 40 compares the level of the detected sound level at the repetition rate of the detector laser beam 44 with a level of sound of a correctly machined hairpin specified by the analysis device 28.

[0108] If the detected sound level deviates from the specified sound level by more than a specified limit, the analysis unit 28 determines that the hairpin 12 has been machined incorrectly. If the detected sound level deviates from the specified sound level by less than or equal to the specified limit, the analysis unit 28 determines that the hairpin 12 has been machined correctly.

Claims

Patent claims 1. Method for processing a hairpin (12) using a pulsed laser beam (18), the method comprising: Machining the hairpin (12) using the pulsed laser beam (18), Detecting an acoustic emission from the hairpin (12) during machining of the hairpin (12), and Determining information about the processing of the hairpin (12) by analyzing the detected acoustic emission.

2. Method according to claim 1, the information is obtained using a trained machine learning model.

3. Method according to any one of the preceding claims 1 to 2, wherein the procedure after obtaining the information comprises: adjusting at least one beam parameter of the pulsed laser beam (18) based on the information obtained.

4. Method according to any one of the preceding claims 1 to 3, wherein the processing of the hairpin (12) is a lacquer removal from the hairpin (12) using the pulsed laser beam (18), the information obtained is information about paint removal.

5. Method according to claim 4, where the analysis of the detected sound emission includes an analysis of a frequency spectrum of the detected sound emission.

6. Method according to any one of the preceding claims 4 to 5, wherein the information is determined as a function of a sound level of the sound emission at a frequency whose value is equal to a value of a pulse repetition frequency of the pulsed laser beam (18).

7. Method according to any one of the preceding claims 4 to 6, where determining the information includes: determining the information based on an assignment of a pattern of the frequency spectrum to a paint removal state.

8. Method according to any one of the preceding claims 1 to 3, wherein the processing of the hairpin (12) is an electrical connection of the hairpin (12) to another hairpin (32) by welding using the pulsed laser beam (18), the information obtained is information about welding.

9. Method according to claim 8, where the analysis of the detected sound emission includes an analysis of a frequency spectrum of the detected sound emission.

10. Method according to claim 2 and claim 9, where determining the information includes: determining the information based on an assignment of a pattern of the frequency spectrum to a welding state of the welding process.

11. Laser processing machine (10, 30) for processing a hairpin (12) using a pulsed laser beam (18), comprising: a laser beam source (20) for generating the pulsed laser beam (18), a detector device (26) for detecting an acoustic emission from the hairpin (12) during the machining of the hairpin (12), an analysis device (28) which is trained to determine information about the processing of the hairpin (12) by analyzing the detected acoustic emission.

12. Method for determining a property of a processing of a hairpin (12) using a pulsed detector laser beam (44), the method comprising: Irradiating the hairpin (12) with a pulsed detector laser beam (44), detecting an acoustic emission from the hairpin (12) during the irradiation of the hairpin (12), and Determining the properties of the processing of the hairpin (12) by analyzing the detected acoustic emission.

13. Inspection machine (40) configured to perform a method according to claim 12.

Images