METHOD FOR STANDARDIZING SENSOR SIGNALS FOR MONITORING A LASER PROCESS, METHOD FOR MONITORING A LASER PROCESS AND LASER PROCESSING SYSTEM

Description

[0001] The invention relates to a method for normalizing sensor signals for monitoring a laser processing process, a method for monitoring a laser processing process and a laser processing system which is set up to carry out the monitoring method. Background and state of the art

[0002] In a laser processing system, also called a laser processing unit or simply a system, the laser beam emitted from a laser beam source or the end of a laser fiber is focused onto the workpiece using beam guidance and focusing optics to heat it locally to its melting temperature. The processing can include laser welding or laser cutting. The laser processing system can include a laser processing head, such as a laser welding head or a laser cutting head. The laser processing process can include both laser welding and laser cutting.

[0003] To ensure the quality of the processing, it is essential to continuously monitor the laser processing process. Monitoring is typically achieved by capturing and evaluating process emissions generated during the laser processing. These process emissions include optical emissions, such as laser radiation backscattered or reflected from the workpiece, and infrared emissions, such as thermal radiation, as well as visible emissions, such as radiation from a plasma generated during processing. Optical process emissions can be referred to as process radiation. Process emissions may also include acoustic emissions.

[0004] Process emissions are typically detected by at least one sensor, such as a photodiode. The sensor measures the intensity of the process emissions at a predefined wavelength or within a predefined wavelength range and generates a corresponding sensor signal. For monitoring purposes, the sensor signal is compared with predefined envelopes and / or threshold values, and an error is generated if the sensor signal falls outside the envelope or exceeds or falls below a threshold value.

[0005] When an identical laser processing process is carried out by different units of the same or identical type of laser processing system, differences or variations occur in the resulting and / or measured process emissions. This is caused, among other things, by tolerances in the components of identical laser processing systems. For example, a laser beam source, deflection units or scanners, or optical elements of the laser processing head such as lenses, mirrors, etc., can exhibit manufacturing tolerances. Tolerances in the sensor units can also lead to differences in the measured process emissions. Corresponding sensor signals can therefore differ from one laser processing system to another. For example, the average signal levels of the sensor signals may differ.

[0006] These differences in the sensor signals make comparing identical laser processing systems difficult or impossible. Furthermore, the same parameters, such as envelopes or thresholds, cannot be used to monitor a given laser processing process on identical laser processing systems.

[0007] DE 10 2019 111501 A1 relates to a method for optimizing laser processing performed with a laser processing device, wherein, during laser processing, laser radiation is generated by a laser radiation source and emitted along an optical path towards a working plane, wherein the laser radiation is modulated by means of a spatial light modulator arranged downstream of the laser radiation source in the optical path, and wherein the modulated laser radiation is focused onto the working plane, comprising the following steps, executed in a closed-loop control loop: i. Determining at least one actual characteristic of the laser radiation and / or the modulated laser radiation using at least one camera unit; ii. Comparing the at least one actual characteristic of the laser radiation and / or the modulated laser radiation with at least one predetermined target characteristic and determining a respective characteristic deviation; iii.Evaluate the characteristic deviation and, if necessary, adjust at least one control parameter. Disclosure of the invention

[0008] It is therefore an objective of the present disclosure to specify a method by which identical or similar laser processing systems or identical or similar laser processing processes can be compared.

[0009] Furthermore, it is an object of the present disclosure to specify a method by which the monitoring of an identical laser processing process with identical monitoring parameters on identical or similar laser processing systems is made possible. In particular, it is an object of the present disclosure to specify a method by which sensor signals of a process monitoring system can be normalized.

[0010] Furthermore, it is an object of the present disclosure to specify a laser processing system configured to carry out such a process. It is also an object of the present disclosure to specify a method for monitoring a given laser processing process using standardized sensor signals.

[0011] One or more of these tasks are solved by the subject matter of independent claims. Advantageous designs and further developments are the subject matter of dependent claims.

[0012] The invention is based on the concept of normalizing sensor signals for monitoring a laser processing process, which are based on detected process emissions, particularly optical process emissions, of the laser processing process, by means of scaling. For calibration, a predefined laser processing process is carried out by a laser processing system, and a sensor signal is generated based on the process emissions produced. Scaling factors for this predefined laser processing process and for this laser processing system are determined based on the generated sensor signal. The scaling factors are then used to scale sensor signals when monitoring subsequent, identical laser processing processes on the same laser processing system (autoscaling).

[0013] Normalization makes the sensor signals generated by multiple identical laser processing systems during the laser processing process comparable across these systems. Therefore, differences in the generated sensor signals between identical laser processing systems can be compensated for through normalization. Consequently, the same monitoring parameters, procedures, or programs can be used on all identical or structurally similar laser processing systems to monitor the same laser processing process.

[0014] Identical or similar laser processing systems, as defined in this disclosure, mean multiple instances of the same or similar type of a given laser processing system or structurally identical instances of a laser processing system. Identical or the same laser processing processes mean laser processing processes that are or were carried out with identical process parameters as a given laser processing process.

[0015] According to a first aspect of the present disclosure, a method for normalizing sensor signals for monitoring a laser processing process is specified, comprising the steps of: performing a predetermined laser processing process using a laser processing system for processing at least one workpiece, wherein the laser processing process includes processing at least one processing area; generating a first sensor signal based on process emissions detected for the processing area by a first sensor; determining a first mean value of the first sensor signal; and determining a first scaling factor for the first sensor signal for the processing area based on the first mean value.

[0016] The specified laser processing process can include or define the processing of at least one workpiece by shining a laser beam along a specified processing path. The specified processing path can be defined with respect to the at least one workpiece, in particular with respect to the surface of one or more workpieces. The at least one processing area of ​​the laser processing process can specify an area of ​​the at least one workpiece onto which the laser beam is shining. The at least one processing area can correspond to a path segment of the specified processing path on the workpiece. The path segment can be defined by a start point and an end point along the processing path, or by a start point and a length along the processing path.

[0017] The laser processing process can include laser welding, also called a laser welding process. In this case, processing a work area can correspond to forming at least one weld seam.

[0018] The laser processing process can alternatively or additionally include laser cutting, also called a laser cutting process. In this case, processing the at least one processing area can correspond to forming one or more cut edges. The at least one processing area can also correspond to a piercing hole created by the laser cutting process. In this case, processing the processing area can include a piercing operation, i.e., forming the piercing hole.

[0019] The process emissions generated during laser processing can include optical process emissions, in particular at least one of the following: thermal radiation in the infrared wavelength range, plasma radiation in the visible wavelength range, and laser radiation reflected from the workpiece. The reflected laser radiation can be in the infrared or visible wavelength range.

[0020] The first sensor signal can be based on the intensity of process emissions detected within a predefined wavelength range or at a specific wavelength. The first sensor can therefore exhibit spectral sensitivity within the predefined wavelength range or at the predefined wavelength. For example, the first sensor signal can be based on the detected intensity of thermal radiation, the detected intensity of plasma radiation, and / or the detected intensity of reflected laser radiation.

[0021] The first sensor signal can be a one-dimensional, time-varying signal. It can comprise multiple sensor values, each assigned to a specific time point. A sensor value can correspond to the intensity of process emissions detected at the assigned time point during the processing of a work area within the specified wavelength range or at the specified wavelength. The first sensor signal, or a segment thereof, can correspond to the processing of the work area in time. In other words, the first sensor signal can be based on process emissions detected during the processing of the work area.

[0022] The method can further include: generating a second sensor signal based on the process emissions detected by a second sensor for the processing area, determining a second mean value of the second sensor signal, and determining a second scaling factor for the second sensor signal for the processing area of ​​the laser processing process based on the second mean value. The first and second sensors may differ in their respective spectral sensitivity ranges.

[0023] The method can generally comprise generating N sensor signals based on the process emissions detected by sensors for the processing area and determining N corresponding mean values ​​for the at least one processing area, where N is a natural number greater than 1. The N sensor signals can differ in their respective underlying wavelength ranges or the respective underlying wavelengths of the detected process emissions. Thus, N scaling factors corresponding to N sensor signals can be determined for the processing area. According to embodiments, the N sensor signals can be generated by N corresponding sensors. According to further embodiments, the N sensor signals can be generated by fewer than N sensors. For example, a single sensor can have spectral sensitivity in several wavelength ranges or at several wavelengths.

[0024] The method can include the simultaneous generation of multiple sensor signals based on the process emissions emitted during the processing of a machining area. Preferably, the intensity of the process emissions can be detected in a first wavelength range, for example, in a visible wavelength range corresponding to plasma radiation, in a second wavelength range corresponding to reflected laser radiation, and in a third wavelength range, for example, in an infrared wavelength range corresponding to thermal radiation, and corresponding first to third sensor signals can be generated. In this case, N = 3.

[0025] The sensors can each include an optical sensor, in particular a photodiode. The photodiode can output a time-varying voltage signal. This time-varying voltage signal can be converted into the corresponding sensor signal by the respective sensor or a control unit.

[0026] The laser processing process can, according to embodiments, include the processing of one or more processing areas. For example, the laser processing process can include the processing of M processing areas, where M is a natural number greater than 1. The process emissions for each of the M processing areas can be detected, and corresponding sensor signals can be generated based on these detected process emissions. Thus, sensor signals can be generated for each of the M processing areas. Therefore, N sensor signals can be generated for each of the M processing areas. For example, the laser processing process can include the processing of three processing areas to form three weld seams, and the first and second sensor signals can be generated for each of these three processing areas.

[0027] The method can further include assigning the sensor signals to the corresponding processing areas. This assignment can be performed manually, for example by a user of the laser processing system, or automatically, for example by a control unit of the laser processing system. The assignment of the sensor signals to the corresponding processing areas can be based on a predetermined or known number of processing areas, a predetermined or known duration for processing each processing area, and / or a predetermined start and / or end time for processing each processing area.

[0028] Preferably, the at least one scaling factor is determined for a given laser processing process and a given instance of a given type of laser processing system. Furthermore, a single scaling factor is determined for a given processing area of ​​this laser processing process and for a type of sensor signal or sensor with a spectral sensitivity in a given wavelength range or at a given wavelength.

[0029] Determining the scaling factor can involve calculating the ratio between the respective mean and a corresponding reference value. For example, determining the first scaling factor can involve calculating the ratio between a first reference value and the first mean, and determining the second scaling factor can involve calculating the ratio between a second reference value and the second mean, and so on.

[0030] Determining the quotient between a reference value and a mean can involve dividing the mean by the reference value. Therefore, the scaling factor can be dimensionless.

[0031] Preferably, the reference values ​​themselves are mean values ​​determined according to the above-described method for normalizing sensor signals for a reference laser processing system. The reference laser processing system can be identical or of the same construction as the laser processing system under consideration.

[0032] The mean can be an arithmetic mean or a geometric mean. Within the scope of this disclosure, the mean can also be defined as a median.

[0033] Determining the respective mean value can involve: removing outliers, i.e., removing the smallest signal values ​​and / or removing the largest signal values ​​of the respective sensor signal, and determining the mean value based on the remaining signal values ​​of the sensor signal. This can also be referred to as cleaning the respective sensor signal.

[0034] The method may further include: repeating the step of performing the specified laser processing operation, generating a corresponding first sensor signal for each repetition, determining a corresponding first mean for each repetition, and determining a higher-level first mean based on the first means of the repetitions, with the first scaling factor being determined based on the first higher-level mean. The method may further include: generating a corresponding second sensor signal for each repetition, determining a corresponding second mean for each repetition, and determining a higher-level second mean based on the second means of the repetitions, with the second scaling factor being determined based on the second higher-level mean. The method may include corresponding steps for the third sensor signal, and so on.

[0035] Repeating the step of performing the specified laser processing procedure can involve performing the identical laser processing procedure on additional workpieces and / or performing the identical laser processing procedure on at least one of the same workpieces. By repeating the steps and repeatedly performing the identical laser processing procedure, the scaling factors can be determined more accurately and reliably.

[0036] The scaling factors, in particular the first and / or second scaling factor, can be stored for monitoring subsequent laser processing processes that are identical to the specified laser processing process. This storage can be performed by a control unit.

[0037] Preferably, the method can be carried out during the commissioning of a laser processing system and / or after the replacement of a component of the laser processing system. The replaced component can be one of the following or include: a laser processing head, a process monitoring system, an optical element, a lens, collimation optics, focusing optics, a mirror, a deflection unit for deflecting the laser beam, an optical fiber, a protective glass, or a laser beam source.

[0038] According to another aspect of the present disclosure, a method for monitoring a laser processing process is specified, the method comprising: performing a laser processing process using a laser processing system for processing at least one workpiece, wherein the laser processing process includes processing at least one processing area; generating a first sensor signal based on process emissions detected for the processing area by a first sensor; scaling the first sensor signal using a first scaling factor determined for this laser processing system and this laser processing process; and monitoring the laser processing process using the first sensor signal scaled by the first scaling factor, wherein the first scaling factor was determined by a normalization method according to the present disclosure.Monitoring can be performed using at least one predefined monitoring parameter for the first sensor signal.

[0039] The sensor signal can also be referred to as the monitoring sensor signal. The sensor itself can also be referred to as the monitoring sensor. Monitoring the laser processing process using the scaled first sensor signal can involve comparing the scaled first sensor signal with monitoring parameters.

[0040] Furthermore, the method for monitoring a laser processing process can include: generating a second sensor signal based on process emissions detected by a second sensor for the processing area; scaling the second sensor signal using a second scaling factor determined for this laser processing system and this laser processing process; and monitoring the laser processing process using the scaled second sensor signal, wherein the second scaling factor was determined by a normalization method according to the present disclosure. The method for monitoring the laser processing process can further include corresponding steps for a third sensor signal, and so on. Monitoring can be performed using at least one predetermined monitoring parameter for the second sensor signal.

[0041] Scaling the sensor signal preferably involves dividing the sensor signal by the appropriate scaling factor.

[0042] Alternatively or additionally, the method for monitoring a laser processing process may include the following steps: scaling the at least one predetermined monitoring parameter for the first sensor signal using the first scaling factor determined for this laser processing system and this laser processing process, and monitoring the laser processing process using the first sensor signal and using the at least one monitoring parameter for the first sensor signal scaled based on the first scaling factor.

[0043] Furthermore, the method for monitoring a laser processing process can include the following steps: scaling the at least one predetermined monitoring parameter for the second sensor signal using the second scaling factor determined for this laser processing system and this laser processing process, and monitoring the laser processing process using the second sensor signal and the at least one monitoring parameter for the second sensor signal scaled based on the second scaling factor.

[0044] Scaling the monitoring parameter preferably involves multiplying the monitoring parameter by the corresponding scaling factor.

[0045] By scaling the sensor signals and / or at least one monitoring parameter using scaling factors specific to this laser processing system and this laser processing process, monitoring of the laser processing process can be carried out based on monitoring parameters determined for another laser processing system that is structurally identical or the same as this laser processing system, or that were determined for an identical or structurally similar reference laser processing system. In particular, monitoring of the laser processing process can be carried out based on monitoring parameters determined for the reference laser processing system.

[0046] The monitoring parameters can include predefined upper and / or lower envelopes, upper and / or lower thresholds, reference curves, and / or average values ​​for the sensor signals. Monitoring can specifically include comparing the generated sensor signals with the monitoring parameters. An error message can be generated if the sensor signal lies outside the envelopes or exceeds or falls below a threshold.

[0047] According to a further aspect of the present disclosure, a laser processing system is specified which is configured to perform the method for normalizing sensor signals and / or the method for monitoring a laser processing process according to embodiments of the present disclosure. The laser processing system comprises: a laser processing head configured for directing a laser beam onto at least one workpiece for performing a laser processing process, and a process monitoring system configured for monitoring the laser processing process, comprising a sensor unit with at least one sensor for detecting process emissions, and a control unit.

[0048] The control unit can be configured to control the laser processing system, to perform the method for normalizing sensor signals and / or the method for monitoring the laser processing process according to embodiments of the present disclosure. The control unit can be configured to control the laser processing head, the sensor unit, and / or the at least one sensor.

[0049] The sensor unit comprises at least one sensor for capturing or detecting process emissions from the laser processing. The sensor may include at least one photodiode and / or a photodiode array and / or a camera, for example a CMOS or CCD-based camera, and / or an acoustic sensor and / or a sound sensor and / or a pressure sensor.

[0050] The sensor can be configured to detect an intensity within a predetermined wavelength range or at a predetermined wavelength. The sensor can exhibit spectral sensitivity within the predetermined wavelength range or at a predetermined wavelength. The sensor can be sensitive only at a specific wavelength or within a specific wavelength range. Alternatively, the sensor can be sensitive within several predetermined wavelength ranges or at several predetermined wavelengths. Furthermore, the sensor can be configured to output a sensor signal based on the detected process emissions, according to embodiments of this disclosure.

[0051] Preferably, the sensor unit comprises a first sensor configured to generate a first sensor signal based on the detected intensity of plasma radiation emitted by the workpiece, a second sensor configured to generate a second sensor signal based on the detected intensity of laser radiation reflected by the workpiece, and / or a third sensor configured to generate a third sensor signal based on the detected intensity of thermal radiation emitted by the workpiece.

[0052] The laser processing head can be a laser welding head or a laser cutting head. The sensor unit and / or at least one sensor can be mounted or arranged on a housing of the laser processing head.

[0053] The process emissions generated by the laser processing can enter the laser processing head, particularly its housing. The laser processing head can have a beam splitter to decouple the process emissions from the housing and the laser beam path. A beam path of the process emissions between the at least one workpiece and the at least one sensor can therefore run at least partially within the laser processing head and / or coaxially with the laser beam path. Alternatively, the sensor unit or the sensor can be arranged such that a beam path of the process emissions between the at least one workpiece and the sensor runs completely outside the laser processing head.

[0054] The control unit can be configured to receive a sensor signal from the at least one sensor or from the sensor unit. Alternatively or additionally, the at least one sensor can output an analog voltage signal, and the control unit can be configured to receive an analog voltage signal from the at least one sensor and to convert the analog voltage signal into a sensor signal according to embodiments of the present disclosure.

[0055] The laser beam emitted by the laser beam source can have a wavelength in the infrared spectral range, particularly in a range between 1030 nm and 1070 nm, or in the visible green spectral range, particularly in a range between 500 nm and 570 nm, preferably at 515 nm, or in the visible blue spectral range, particularly in a range from 400 nm to 500 nm, or in a range between 440 nm and 460 nm, preferably at 450 nm. Brief description of the drawings

[0056] The invention is described in detail below with reference to figures. Fig. 1A shows a schematic representation of a laser processing system for carrying out a laser processing process and for monitoring the laser processing process according to embodiments; Fig. 1B shows a schematic representation of a laser processing system for carrying out a laser processing process and for monitoring the laser processing process according to further embodiments; Fig. 2 shows a schematic representation of two workpieces to be machined in a top view with a machining path and machining areas; Fig. 3 shows a flowchart of a procedure for normalizing sensor signals for monitoring a laser processing process; Fig. 4A shows a schematic representation of two workpieces to be machined in a top view with three machining paths; Fig. 4Bshows a schematic representation of three groups of two workpieces each to be machined, in top view, each with a machining path; Fig. 5A shows the progression of monitoring sensor signals without scaling; Fig. 5B shows the waveforms of monitoring sensor signals that have been scaled with scaling factors. Detailed description of the drawings

[0057] Unless otherwise noted, the same reference symbols are used for identical and equivalent elements in the following.

[0058] Fig. 1A shows a schematic representation of a laser processing system for carrying out a laser processing process and for monitoring the laser processing process according to embodiments.

[0059] The laser processing system 10 comprises a laser processing head 12 and a process monitoring system 13 with a sensor unit 14 and a control unit 16. The following describes embodiments in which the laser processing head 12 is configured as a welding head and in which the laser processing system 10 is set up to perform a laser welding process; however, the present disclosure is not limited to these embodiments. For example, the laser processing head 12 can also be configured as a cutting head, and the laser processing system 10 can be set up to perform a laser cutting process. The laser processing system 10 can also be referred to simply as the system.

[0060] The laser welding head 12 is configured to focus a laser beam (not shown), also referred to as a processing laser beam or simply processing beam, emitting from a laser beam source or the end of a laser guide fiber, using beam guidance and focusing optics (not shown), and to direct the beam onto at least one processing area B1, B2, B3 of at least one workpiece Wa, Wb to perform the laser welding process. The workpieces Wa, Wb are in Fig. 2 shown in a schematic top view.

[0061] The laser welding head 12 can be configured as a so-called fixed-optics laser welding head or as a so-called scanner laser welding head. The scanner laser welding head can have a deflection unit for deflecting the laser beam relative to the workpieces Wa, Wb and for directing the laser beam along a processing path P relative to the workpieces Wa, Wb. The deflection unit can include scanner optics, a scanner system, a scanner mirror, and / or an electroplating scanner. In the fixed-optics laser welding head, the laser beam can be moved relative to the workpiece Wa, Wb by moving the laser welding head 12 itself, or the workpiece Wa, Wb can be moved relative to the laser welding head 12.

[0062] When the processing beam is directed onto at least one of the workpieces Wa, Wb to perform the laser welding process, process emissions 22 are generated and emitted by the workpieces Wa, Wb. These process emissions include radiation in the visible wavelength range, in particular plasma radiation emitted by the workpieces Wa, Wb; radiation in the infrared wavelength range, in particular infrared or thermal radiation emitted by the workpieces Wa, Wb; and laser radiation from the incident laser beam reflected by the workpieces Wa, Wb. According to embodiments, the process emissions can also include acoustic process emissions, for example, sound waves and / or pressure waves emitted by the workpieces Wa, Wb.

[0063] The at least one sensor of the sensor unit 14 is configured to detect or measure the intensity of the process emissions and, based on this, to generate and output a sensor signal. With reference to the method for monitoring the laser processing process according to embodiments of the present disclosure, the generated sensor signal can also be referred to as a monitoring sensor signal. According to embodiments, the generated sensor signal is a one-dimensional, time-varying signal. The at least one sensor can comprise a photodiode or a photodiode or pixel array.According to embodiments, the sensor unit 14 comprises a first sensor for measuring the intensity of the plasma radiation, a second sensor for measuring the intensity of the reflected laser radiation, and a third sensor for measuring the intensity of the thermal radiation, wherein the three sensors each comprise photodiodes and generate and output corresponding first to third sensor signals. Generally speaking, the sensor unit can have N sensors, where N is a natural number greater than 1.

[0064] The control unit 16 is connected to the sensor unit 14 and receives the sensor signal from the at least one sensor. The control unit 16 can be configured to record the sensor signal. The control unit 16 is configured to monitor the laser welding process based on the sensor signal from the at least one sensor. The control unit 16 is further configured to control the laser processing system 10, in particular the sensor unit 14, the at least one sensor, and / or the laser welding head 12, in order to carry out a method for normalizing sensor signals according to embodiments of the present disclosure.

[0065] At the in Fig. 1AIn the illustrated embodiment, the process emissions 22 enter the laser welding head 12 and are coupled out of the beam path of the processing laser beam or out of the laser welding head 12 by a beam splitter 24. The process emissions are then directed into the sensor unit 14 and encounter at least one sensor (not shown). The beam path of the process emissions 22 between the workpieces Wa, Wb and the sensor runs at least partially within the laser processing head 12 and / or coaxially with the beam path of the processing laser beam. Fig. 1A The process emissions 22 entering the laser welding head 12 are deflected at a beam splitter 24 towards the sensor or sensor unit 14. Alternatively, the processing laser beam can of course be deflected at the beam splitter 24, and the beam path of the process emissions 22 can be straight between the workpieces Wa, Wb and the sensor.

[0066] Fig. 1B Figure 1 shows a schematic representation of a laser processing system for carrying out a laser processing process and for monitoring the laser processing process according to further embodiments. The figure in Fig. 1B The laser processing system 10 shown is the one in Fig. 1A The laser processing system shown is similar to 10, therefore only the differences will be described below. As in Fig. 1B The sensor unit 14 or the sensor is shown arranged such that the beam path of the process emissions 22 between the at least one workpiece Wa, Wb and the sensor lies completely outside the laser processing head 12. In this case, the Fig. 1A The beam splitter 24 shown, as well as an interface between the sensor unit 14 and the laser welding head 12, is not required.

[0067] The laser welding process involves shining a laser beam onto one or both of the workpieces Wa, Wb to be welded. This laser beam introduces energy into at least one of the workpieces Wa, Wb, causing the material to melt or vaporize. This process generates thermal and plasma radiation. A portion of the laser beam is reflected by the workpiece Wa, Wb. As the material cools, a weld seam forms. The weld seam can form on one or both of the workpieces Wa, Wb. The resulting weld seam joins the workpieces Wa, Wb together.

[0068] An area of ​​the workpieces Wa, Wb onto which the laser beam is directed to form a weld can be referred to as a processing area. The processing area can be, for example, linear or planar. A processing area can correspond to one or more welds. Processing areas and / or welds can be arranged contiguously or separately on the workpieces Wa, Wb. The laser beam is directed onto one or more processing areas to perform the laser welding process along a predefined processing path P, as shown in Fig. 2 As shown, this can be done. Accordingly, path segments of the processing path P can correspond to the processing areas. The speed at which the laser beam is directed along the processing path P can be referred to as the feed rate.

[0069] As in Fig. 2As shown, a laser processing process includes, for example, the processing of three processing areas B1, B2, B3 to form three weld seams. When processing one of the processing areas B1, B2, B3 by shining the laser beam upon it, the previously described sensor signal of the at least one sensor for that processing area B1, B2, B3 can be generated. If the sensor unit 14 has, for example, three sensors, then three sensor signals can be recorded for each of the three processing areas B1, B2, B3. In general, a laser welding process can include the processing of M processing areas, where M is a natural number greater than 1.

[0070] A predefined laser welding process defines one or more process parameters to form the weld seams in a predetermined manner. These process parameters can include the areas to be processed, a predefined processing path, a focus position, laser power, feed rate, laser beam diameter, and / or the distance and / or orientation of a laser welding head to a workpiece. Process parameters such as focus position, feed rate, laser power, and / or laser beam diameter can be variable or constant during the processing of a given area.

[0071] The monitoring of the specified laser welding process is carried out using predefined monitoring parameters. For example, sensor signals are compared with predefined envelopes and / or threshold values, and an error can be generated if the sensor signals lie outside the envelopes or exceed or fall below a threshold value.

[0072] The inventors discovered that when performing a specific laser welding process using identical laser processing systems, differences in the generated sensor signals still occur. In other words, the process emissions generated by this laser processing process vary between identical laser processing systems. Identical laser processing systems could, for example, refer to multiple units of the same commercially available type or model of laser processing system. Due to the variation in process emissions, the resulting sensor signals also vary. In particular, the average signal levels or average signal values ​​of the sensor signals can vary.

[0073] The reason for this lies, among other things, in tolerances of the laser processing system, particularly manufacturing and material tolerances. For example, the laser beam source and optical elements of the laser processing head, such as focusing optics, collimator optics, lenses, mirrors, deflection units, etc., can exhibit tolerances. Tolerances of the sensor unit or the sensors themselves can also lead to differences in the detected process emissions and the generated sensor signals. Therefore, the sensor signals can differ from one laser processing system to another.

[0074] Using the method described herein for normalizing sensor signals for monitoring a given laser processing process on identical laser processing systems, variations in the sensor signals can be reduced or eliminated. This makes the sensor signals of identical laser processing systems comparable. Consequently, process monitoring with identical monitoring parameters is possible for multiple identical laser processing systems.

[0075] This eliminates the need to individually determine monitoring parameters for a given laser processing process for each of several identical laser processing systems. For example, monitoring parameters for all identical laser processing systems sold to customers can be adopted from a reference laser processing system. This reference laser processing system can then remain with the manufacturer of the laser processing systems. This significantly simplifies and shortens the manufacturing, commissioning, and maintenance of the laser processing systems.

[0076] Fig. 3 Figure 1 shows a flowchart of a method for normalizing sensor signals for monitoring a laser processing process according to embodiments. The method can also be referred to as autoscaling or autoscaling.

[0077] It is assumed below that a processing area corresponds to a weld seam being formed. However, the disclosure is not limited to this. A processing area can also correspond to several weld seams being formed. The processing areas and / or the weld seams can be arranged contiguously or separately on a workpiece. The method for normalizing sensor signals and / or the method described below for monitoring the laser processing process can be performed individually for each weld seam or across all weld seams.

[0078] The procedure is demonstrated using the example with reference to Fig. 1A and 1BThe laser processing system described in section 10 is described. The method begins (S1) with the execution of a predefined laser processing process, wherein the laser processing process includes the processing of at least one processing area. For example, the predefined laser processing process can be a laser welding process and the processing of the three processing areas B1, B2, B3 of Fig. 2 to form three weld seams.

[0079] During the processing of at least one processing area, the generated process emissions are detected by at least one sensor of the sensor unit, and a corresponding sensor signal is generated (S2). If the sensor unit 14 comprises, for example, three sensors, three sensor signals are generated for the processing of each of the three processing areas B1, B2, and B3. Thus, a corresponding first sensor signal, a second sensor signal, and a third sensor signal are generated for each of the three processing areas B1, B2, and B3.

[0080] The operator can manually define how many individual processing areas or weld seams a single workpiece comprises to the laser processing system. Alternatively, or in addition, the laser processing system can automatically detect the individual processing areas or weld seams. The definition of the processing areas, such as a start time, a stop time, and / or the number of processing areas, can be manually entered by the operator or automatically detected by the laser processing system.

[0081] Subsequently (S3), mean values ​​are determined for the three sensor signals. According to embodiments, a corresponding mean value can be determined for each of the processing areas and for each of the generated sensor signals. Thus, a total of N x M mean values ​​can be determined. Referring to the previous example, a mean value M 1,1 of the first sensor signal is determined for the first processing area B 1, a mean value M 1,2 of the first sensor signal is determined for the second processing area B 2, and a mean value M 1,3 of the first sensor signal is determined for the third processing area B 3. Furthermore, three mean values ​​M 2,1, M 2,2, M 2,3 of the second sensor signal are determined for the first to third processing areas B 1, B 2, B 3, and three mean values ​​M 3,1, M 3,2, M 3,3 of the third sensor signal are determined for the first to third processing areas B 1, B 2, B 3. In total, nine mean values ​​are determined in this example.

[0082] According to embodiments, the calculated mean values ​​are arithmetic mean values ​​of the sensor values ​​contained in the sensor signals for the respective processing areas; however, the present disclosure is not limited to this. In other words, an arithmetic mean calculation is performed for each sensor and processing area (e.g., weld seam).

[0083] Before the mean values ​​are determined, the procedure may include a step to clean the sensor signals. This cleaning is performed separately for each sensor signal in each processing area. Outliers in the sensor values ​​are identified and removed. Outliers include, for example, particularly large or small sensor values. This step may also involve sorting the sensor values.

[0084] This involves removing the smallest and / or largest sensor values ​​based on predefined and / or adjustable percentages. For example, x percent of the smallest sensor values ​​of a sensor signal for a processing area and / or y percent of the largest sensor values ​​of a sensor signal for a processing area can be removed, where x and y are rational numbers between 0 and 100. Determining the mean value of the sensor signal for the processing area is then performed as described previously, based on the remaining signal values ​​of the sensor signal, i.e., based on the cleaned sensor signal, for that processing area.

[0085] For example, the sensor values ​​of the first sensor signal are cleaned of outliers generated during the processing of the first processing area B1 based on the detected process emissions from the first sensor. Subsequently, the remaining sensor values ​​of the first sensor signal are used to determine the mean value M1.1 of the first sensor signal for the first processing area B1.

[0086] Removing or discarding outliers has the advantage that the step of determining the means, and thus the determination of the scaling factors and the normalization of the sensor values, can be carried out stably and free of disturbances. Therefore, outliers have little or no influence on the determined scaling factors.

[0087] Next (S4), scaling factors for the sensor signals and for the processing areas are determined based on the determined mean values. Determining the scaling factors may involve normalizing the determined mean values. According to embodiments, determining the scaling factors may involve dividing the mean values ​​by appropriate reference values.

[0088] According to embodiments, the reference values ​​can be mean values ​​determined by steps S1-S3 for a previously described reference laser processing system that is identical or structurally equivalent to the laser processing system under consideration. Accordingly, the mean values ​​of the laser processing system under consideration can be related to the mean values ​​of the identical reference laser processing system in step S4.

[0089] Referring to the previously described example, a scaling factor F 1.1 for the first sensor signal and for the first processing area B 1 is determined by dividing the mean value M 1.1 by a predefined reference value R 1.1. In total, nine scaling factors F 1.1 to F 3.3 are determined in this example, based on nine reference values ​​R 1.1 to R 3.3.

[0090] The sensor signal normalization process determines or calculates individual scaling factors for each processing area and each sensor signal or sensor. This allows the sensor signals to be brought to a uniform signal level for identical laser processing systems and to a defined target value.

[0091] The scaling factors determined using the above-described method for normalizing sensor values ​​can be stored for monitoring subsequent, identical laser processing operations. This storage can be performed, for example, by the control unit 16.

[0092] The above procedure can be carried out during the commissioning of the laser processing system and after the replacement of components of the laser processing system 10, for example the laser processing head 12, the process monitoring system 12, or the components described above.

[0093] The above-described method for normalizing the sensor signals can, for example, be implemented by the control unit 16 of the monitoring system 13 with reference to Fig. 1A and 1B described laser processing system, and can in particular be carried out by software of the control unit 16.

[0094] According to further embodiments, an operator of the method or the laser processing system can select whether the laser processing process for normalizing the sensor values ​​should be performed multiple times. Accordingly, steps S1, S2, and S3 can be performed multiple times or repeatedly. In total, steps S1-S3 can be performed Z times, where Z is a natural number greater than 1. Z can, for example, be between 2 and 100, particularly between 5 and 50, including the extremes, and can, for example, be 10.

[0095] Step S1 of the laser processing process, which involves processing at least one area, is therefore performed multiple times. The laser processing process can, as Fig. 4A shown, the process can be carried out multiple times on the same workpiece Wa, Wb. Alternatively or additionally, the laser processing can be carried out as shown in Fig. 4BAs shown, this process can be carried out multiple times on different workpieces or workpiece groups Wa 1 / Wb 1 , Wa 2 / Wb 2 , Wa 3 / Wb 3. As shown in Fig. 4A and 4B As shown, the identical laser processing process was performed three times along three identical processing paths P1, P2, P3. Each processing path P1, P2, P3 comprises three processing areas. The first processing path P1 includes processing areas B1,1, B1,2, B1,3; the second processing path P2 includes processing areas B2,1, B2,2, B2,3; and the third processing path P3 includes processing areas B3,1, B3,2, B3,3. Therefore, the first processing areas B1,1, B2,1, B3,1 were processed with identical process parameters. The same applies to the second processing areas B 1,2 , B 2,2 , B 3,2 and to the third processing areas B 1,3 , B 2,3 , B 3,3 .

[0096] Similarly, steps S2 and S3 are executed Z times. Accordingly, for example, a first sensor signal is generated for the processing area B1,1, a first sensor signal for the processing area B2,1, and a first sensor signal for the processing area B3,1 using a first sensor, and an average value M1,1,1 is determined for the first sensor signal for the processing area B1,1, an average value M1,1,2 for the first sensor signal for the processing area B2,1, and an average value M1,1,3 for the first sensor signal for the processing area B3,1.

[0097] The same applies to the first sensor signal for processing areas B 2,1 , B 2,2 , B 2,3 and processing areas B 3,1 , B 3,2 , B 3,3. Furthermore, the same applies to the second sensor signal and to the third sensor signal for processing areas B 1,1 to B 3,3. In total, N x M x Z, in this example 27, mean values ​​are determined.

[0098] Based on the mean values ​​M1,1,1, M1,1,2, and M1,1,3, a higher-level mean value M'1,1 is then determined for the first sensor signal for the identical first processing areas B1,1, B2,1, and B3,1. According to embodiments, the higher-level mean value M'1,1 can be determined as the arithmetic mean of the mean values ​​M1,1,1, M1,1,2, and M1,1,3. Similarly, a higher-level mean value M'1,2 can be obtained for the first sensor signal for the identical second processing areas B1,2, B2,2, and B3,2, and a higher-level mean value M'1,3 can be obtained for the first sensor signal for the identical processing areas B1,3, B2,3, and B3,3. Thus, nine higher-level mean values ​​M'1,1 to M'3,3 are obtained.

[0099] Finally, the scaling factors are determined as in step S4 of the procedure according to Fig. 3determined. In the present example, the nine scaling factors F 1.1 to F 3.3 are determined based on the higher-order means M' 1.1 to M' 3.3.

[0100] It is possible that the laser welding process may not be error-free. For example, the resulting welds may contain defects. Therefore, it is advantageous to perform the laser processing multiple times, as described above, and to determine the scaling factors based on the overall mean values. This ensures that faulty laser welding processes and the sensor signals recorded during them do not have an undue influence on the determined scaling factors. The risk can be further minimized by having the operator select, after step S1, which of the processed areas or workpieces should be used to determine the scaling factors. When making this selection, the operator should ensure that the processed areas or workpieces have been processed without defects.

[0101] Using the scaling factors determined by the sensor signal normalization procedure, monitoring parameters for one laser processing system can also be used to monitor another, identical laser processing system. This is demonstrated with reference to Fig. 5A and 5B illustrated.

[0102] Fig. 5A and 5B The diagrams show the time courses SE1, SE2, SE3 of three monitoring sensor signals. The three monitoring sensor signals were obtained by a method for monitoring a laser processing operation, which is performed by a sample B of the device with reference to Fig. 1A and 1B described laser processing system 10 was carried out.

[0103] The laser processing process is a laser welding process encompassing the processing of three processing areas B1, B2, and B3, each corresponding to a weld seam. During the laser processing process, three monitoring sensor signals were generated based on process emissions, which were detected by three corresponding sensors with spectral sensitivities in different wavelength ranges.

[0104] In Fig. 5A and 5B In the SE1, SE2, and SE3 traces, the three monitoring sensor signals for each of the three processing areas B1, B2, and B3 are identified. Each of the three monitoring sensor signals was therefore generated and recorded for each of the three processing areas B1, B2, and B3. According to the embodiment, process monitoring can only take place within the identified monitoring sensor signals.

[0105] In Fig. 5AThe diagram illustrates a case where the procedure for normalizing the sensor signals for laser processing system B has not yet been performed. The diagrams also show the envelope curves used for monitoring of a single unit A of the laser processing system. These envelope curves describe a permissible tolerance band, which was determined for laser processing system A and for the identical laser welding process after autoscaling of laser processing system A.

[0106] If, as described, the identical laser welding process is carried out by laser processing system B, and an operator of the laser welding system, also referred to as a system operator, ensures that the laser welding system runs stably and without errors, the SE1, SE2, and SE3 curves of the monitoring sensor signals nevertheless show apparent differences in the recorded process emissions. This is due to the described tolerances of laser processing system B compared to A, such as tolerances of the installed optics, the laser used, or the scanner. These differences mean that the laser welding process, carried out identically by laser processing system B, cannot be monitored using the monitoring parameters, such as the envelope curves, of laser processing system A.

[0107] In Fig. 5BThe case is presented in which the method for normalizing the sensor signal values ​​according to embodiments of the present disclosure was carried out for the laser processing system B. For this purpose, the three monitoring sensor signals were scaled using scaling factors specific to the laser processing system B and to the given laser welding process.

[0108] The scaling factors determined in the normalization procedure for laser processing system B (individually for each processing area and per sensor) ensure that the sensor signals of laser processing system B remain within the envelope curves of laser processing system A. Therefore, the monitoring parameters of laser processing system A can be used to monitor laser processing system B thanks to these scaled sensor signals.

Claims

1. Method for normalizing sensor signals for monitoring a laser machining process, comprising the steps: - carrying out (S1) a predefined laser machining process by means of a laser machining system (10) for machining at least one workpiece (Wa, Wb), wherein the laser machining process comprises the machining of at least one machining region (B1, B2, B3); - generating (S2) a first sensor signal based on process emissions (22), in particular optical process emissions, detected by a first sensor for the machining region (B1, B2, B3), - determining (S3) a first mean value of the first sensor signal, - determining (S4) a first scaling factor for the first sensor signal for the machining region (B1, B2, B3) based on the first mean value.

2. Method according to claim 1, wherein the determination of the first scaling factor (S4) comprises the determination of a quotient between a first reference value and the first mean value.

3. Method according to either of the preceding claims, further comprising: - repeating the step of carrying out the predetermined laser machining process (S1), - generating a corresponding first sensor signal (S2) for each repetition, and determining a corresponding first mean value for each repetition, - determining a superordinate first mean value based on the corresponding first mean values for each repetition, wherein the determination of the first scaling factor is carried out based on the first superordinate mean value.

4. Method according to one of the preceding claims, comprising: - generating a second sensor signal based on process emissions (22) detected by a second sensor for the machining region, wherein the first sensor and the second sensor have different spectral sensitivity ranges; - determining a second mean value of the second sensor signal, - determining a second scaling factor for the second sensor signal for the machining region (B1, B2, B3) based on the second mean value.

5. Method according to one of the preceding claims, wherein the first sensor signal is based on an intensity of the process emissions (22) detected in a predefined wavelength range or at a predefined wavelength.

6. Method according to one of the preceding claims, wherein the first sensor signal is based on a detected intensity of temperature radiation in the infrared wavelength range, a detected intensity of plasma radiation in the visible wavelength range or a detected intensity of reflected laser radiation.

7. Method according to one of the preceding claims, wherein the determination of the mean value comprises: - removing the smallest signal values and / or removing the largest signal values of the sensor signal, and - determining the mean value based on the remaining signal values of the sensor signal.

8. Method according to one of the preceding claims, wherein the machining region (B1, B2, B3) corresponds to a path section of a predefined machining path (P) of the laser machining process on the at least one workpiece (Wa, Wb).

9. Method according to one of the preceding claims, wherein the predefined laser machining process comprises a laser welding process, and the at least one machining region (B1, B2, B3) corresponds to a weld seam, and / or wherein the predefined laser machining process comprises a laser cutting process and the at least one machining region (B1, B2, B3) corresponds to a cutting edge.

10. Method according to one of the preceding claims, wherein the first scaling factor and / or second scaling factor is stored for monitoring subsequent, identical laser machining processes.

11. Method according to one of the preceding claims, wherein the method is carried out during the start-up of the laser machining system (10) and / or after replacement of a component of the laser machining system (10).

12. Method for monitoring a laser machining process, the method comprising: - carrying out a laser machining process by means of a laser machining system for machining at least one workpiece (Wa, Wb), wherein the laser machining process comprises the machining of at least one machining region (B1, B2, B3); - generating a first sensor signal based on process emissions (22), in particular optical process emissions, detected by a first sensor for the machining region (B1, B2, B3); - scaling the first sensor signal by means of a first scaling factor determined for this laser machining system (10) and this laser machining process, and monitoring the laser machining process on the basis of the first sensor signal scaled based on the first scaling factor and on the basis of at least one predefined monitoring parameter for the first sensor signal; or - scaling the at least one predefined monitoring parameter for the first sensor signal by means of the first scaling factor determined for this laser machining system (10) and this laser machining process, and monitoring the laser machining process on the basis of the first sensor signal and on the basis of the at least one monitoring parameter scaled based on the first scaling factor.

13. Method according to claim 12, wherein the at least one monitoring parameter for the first sensor signal is selected from the group comprising: an upper envelope curve, a lower envelope curve, an upper threshold value, a lower threshold value, a reference profile, a mean value.

14. Laser machining system (10), comprising: - a laser machining head (12), configured for irradiating a laser beam onto a workpiece (Wa, Wb) for carrying out a laser machining process, and - process monitoring system (13), configured for monitoring the laser machining process, comprising a sensor unit (14) with at least one sensor for detecting process emissions, and a control unit (16), configured for carrying out a method according to claim 12 or 13.

15. Laser machining system according to claim 14, wherein the at least one sensor comprises a photodiode with a spectral sensitivity in a predefined wavelength range.